librarian-bots/model_cards_with_metadata · Datasets at Hugging Face

modelId stringlengths 5 139	author stringlengths 2 42	last_modified timestamp[us, tz=UTC]date 2020-02-15 11:33:14 2025-08-26 06:27:36	downloads int64 0 223M	likes int64 0 11.7k	library_name stringclasses 521 values	tags listlengths 1 4.05k	pipeline_tag stringclasses 55 values	createdAt timestamp[us, tz=UTC]date 2022-03-02 23:29:04 2025-08-26 06:26:46	card stringlengths 11 1.01M
KDHyun08/TAACO_STS	KDHyun08	2022-08-01T05:00:14Z	2,406	2	sentence-transformers	[ "sentence-transformers", "pytorch", "bert", "feature-extraction", "sentence-similarity", "transformers", "TAACO", "ko", "autotrain_compatible", "text-embeddings-inference", "endpoints_compatible", "region:us" ]	sentence-similarity	2022-07-25T08:19:31Z	--- pipeline_tag: sentence-similarity tags: - sentence-transformers - sentence-similarity - transformers - TAACO language: ko --- # TAACO_Similarity 본 모델은 [Sentence-transformers](https://www.SBERT.net)를 기반으로 하며 KLUE의 STS(Sentence Textual Similarity) 데이터셋을 통해 훈련을 진행한 모델입니다. 필자가 제작하고 있는 한국어 문장간 결속성 측정 도구인 K-TAACO(가제)의 지표 중 하나인 문장 간 의미적 결속성을 측정하기 위해 본 모델을 제작하였습니다. 또한 모두의 말뭉치의 문장간 유사도 데이터 등 다양한 데이터를 구해 추가 훈련을 진행할 예정입니다. ## Train Data KLUE-sts-v1.1._train.json NLI-sts-train.tsv ## Usage (Sentence-Transformers) 본 모델을 사용하기 위해서는 [Sentence-transformers](https://www.SBERT.net)를 설치하여야 합니다. ``` pip install -U sentence-transformers ``` 모델을 사용하기 위해서는 아래 코드를 참조하시길 바랍니다. ```python from sentence_transformers import SentenceTransformer, models sentences = ["This is an example sentence", "Each sentence is converted"] embedding_model = models.Transformer( model_name_or_path="KDHyun08/TAACO_STS", max_seq_length=256, do_lower_case=True ) pooling_model = models.Pooling( embedding_model.get_word_embedding_dimension(), pooling_mode_mean_tokens=True, pooling_mode_cls_token=False, pooling_mode_max_tokens=False, ) model = SentenceTransformer(modules=[embedding_model, pooling_model]) embeddings = model.encode(sentences) print(embeddings) ``` ## Usage (실제 문장 간 유사도 비교) [Sentence-transformers](https://www.SBERT.net) 를 설치한 후 아래 내용과 같이 문장 간 유사도를 비교할 수 있습니다. query 변수는 비교 기준이 되는 문장(Source Sentence)이고 비교를 진행할 문장은 docs에 list 형식으로 구성하시면 됩니다. ```python from sentence_transformers import SentenceTransformer, models embedding_model = models.Transformer( model_name_or_path="KDHyun08/TAACO_STS", max_seq_length=256, do_lower_case=True ) pooling_model = models.Pooling( embedding_model.get_word_embedding_dimension(), pooling_mode_mean_tokens=True, pooling_mode_cls_token=False, pooling_mode_max_tokens=False, ) model = SentenceTransformer(modules=[embedding_model, pooling_model]) docs = ['어제는 아내의 생일이었다', '생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다. 주된 메뉴는 스테이크와 낙지볶음, 미역국, 잡채, 소야 등이었다', '스테이크는 자주 하는 음식이어서 자신이 준비하려고 했다', '앞뒤도 1분씩 3번 뒤집고 래스팅을 잘 하면 육즙이 가득한 스테이크가 준비되다', '아내도 그런 스테이크를 좋아한다. 그런데 상상도 못한 일이 벌이지고 말았다', '보통 시즈닝이 되지 않은 원육을 사서 스테이크를 했는데, 이번에는 시즈닝이 된 부챗살을 구입해서 했다', '그런데 케이스 안에 방부제가 들어있는 것을 인지하지 못하고 방부제와 동시에 프라이팬에 올려놓을 것이다', '그것도 인지 못한 체... 앞면을 센 불에 1분을 굽고 뒤집는 순간 방부제가 함께 구어진 것을 알았다', '아내의 생일이라 맛있게 구워보고 싶었는데 어처구니없는 상황이 발생한 것이다', '방부제가 센 불에 녹아서 그런지 물처럼 흘러내렸다', ' 고민을 했다. 방부제가 묻은 부문만 제거하고 다시 구울까 했는데 방부제에 절대 먹지 말라는 문구가 있어서 아깝지만 버리는 방향을 했다', '너무나 안타까웠다', '아침 일찍 아내가 좋아하는 스테이크를 준비하고 그것을 맛있게 먹는 아내의 모습을 보고 싶었는데 전혀 생각지도 못한 상황이 발생해서... 하지만 정신을 추스르고 바로 다른 메뉴로 변경했다', '소야, 소시지 야채볶음..', '아내가 좋아하는지 모르겠지만 냉장고 안에 있는 후랑크소세지를 보니 바로 소야를 해야겠다는 생각이 들었다. 음식은 성공적으로 완성이 되었다', '40번째를 맞이하는 아내의 생일은 성공적으로 준비가 되었다', '맛있게 먹어 준 아내에게도 감사했다', '매년 아내의 생일에 맞이하면 아침마다 생일을 차려야겠다. 오늘도 즐거운 하루가 되었으면 좋겠다', '생일이니까~'] #각 문장의 vector값 encoding document_embeddings = model.encode(docs) query = '생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다' query_embedding = model.encode(query) top_k = min(10, len(docs)) # 코사인 유사도 계산 후, cos_scores = util.pytorch_cos_sim(query_embedding, document_embeddings)[0] # 코사인 유사도 순으로 문장 추출 top_results = torch.topk(cos_scores, k=top_k) print(f"입력 문장: {query}") print(f"\n<입력 문장과 유사한 {top_k} 개의 문장>\n") for i, (score, idx) in enumerate(zip(top_results[0], top_results[1])): print(f"{i+1}: {docs[idx]} {'(유사도: {:.4f})'.format(score)}\n") ``` ## Evaluation Results 위 Usage를 실행하게 되면 아래와 같은 결과가 도출됩니다. 1에 가까울수록 유사한 문장입니다. ``` 입력 문장: 생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다 <입력 문장과 유사한 10 개의 문장> 1: 생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다. 주된 메뉴는 스테이크와 낙지볶음, 미역국, 잡채, 소야 등이었다 (유사도: 0.6687) 2: 매년 아내의 생일에 맞이하면 아침마다 생일을 차려야겠다. 오늘도 즐거운 하루가 되었으면 좋겠다 (유사도: 0.6468) 3: 40번째를 맞이하는 아내의 생일은 성공적으로 준비가 되었다 (유사도: 0.4647) 4: 아내의 생일이라 맛있게 구워보고 싶었는데 어처구니없는 상황이 발생한 것이다 (유사도: 0.4469) 5: 생일이니까~ (유사도: 0.4218) 6: 어제는 아내의 생일이었다 (유사도: 0.4192) 7: 아침 일찍 아내가 좋아하는 스테이크를 준비하고 그것을 맛있게 먹는 아내의 모습을 보고 싶었는데 전혀 생각지도 못한 상황이 발생해서... 하지만 정신을 추스르고 바로 다른 메뉴로 변경했다 (유사도: 0.4156) 8: 맛있게 먹어 준 아내에게도 감사했다 (유사도: 0.3093) 9: 아내가 좋아하는지 모르겠지만 냉장고 안에 있는 후랑크소세지를 보니 바로 소야를 해야겠다는 생각이 들었다. 음식은 성공적으로 완성이 되었다 (유사도: 0.2259) 10: 아내도 그런 스테이크를 좋아한다. 그런데 상상도 못한 일이 벌이지고 말았다 (유사도: 0.1967) ``` DataLoader: `torch.utils.data.dataloader.DataLoader` of length 142 with parameters: ``` {'batch_size': 32, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'} ``` Loss: `sentence_transformers.losses.CosineSimilarityLoss.CosineSimilarityLoss` Parameters of the fit()-Method: ``` { "epochs": 4, "evaluation_steps": 1000, "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator", "max_grad_norm": 1, "optimizer_class": "<class 'transformers.optimization.AdamW'>", "optimizer_params": { "lr": 2e-05 }, "scheduler": "WarmupLinear", "steps_per_epoch": null, "warmup_steps": 10000, "weight_decay": 0.01 } ``` ## Full Model Architecture ``` SentenceTransformer( (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False}) ) ``` ## Citing & Authors <!--- Describe where people can find more information -->
wenkai-li/distilroberta-base-finetuned-marktextepoch_n200	wenkai-li	2022-08-01T04:07:13Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "roberta", "fill-mask", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-31T18:33:57Z	--- license: apache-2.0 tags: - generated_from_trainer model-index: - name: distilroberta-base-finetuned-marktextepoch_n200 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilroberta-base-finetuned-marktextepoch_n200 This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the None dataset. It achieves the following results on the evaluation set: - Loss: 2.0531 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 200 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:------:\|:---------------:\| \| 2.2313 \| 1.0 \| 1500 \| 2.1592 \| \| 2.1731 \| 2.0 \| 3000 \| 2.1277 \| \| 2.153 \| 3.0 \| 4500 \| 2.1144 \| \| 2.1469 \| 4.0 \| 6000 \| 2.1141 \| \| 2.1281 \| 5.0 \| 7500 \| 2.1374 \| \| 2.1043 \| 6.0 \| 9000 \| 2.1069 \| \| 2.0834 \| 7.0 \| 10500 \| 2.0993 \| \| 2.0602 \| 8.0 \| 12000 \| 2.0817 \| \| 2.024 \| 9.0 \| 13500 \| 2.0918 \| \| 2.0261 \| 10.0 \| 15000 \| 2.0793 \| \| 1.9889 \| 11.0 \| 16500 \| 2.0567 \| \| 1.9915 \| 12.0 \| 18000 \| 2.0700 \| \| 1.9532 \| 13.0 \| 19500 \| 2.0436 \| \| 1.9362 \| 14.0 \| 21000 \| 2.0596 \| \| 1.9024 \| 15.0 \| 22500 \| 2.0189 \| \| 1.9262 \| 16.0 \| 24000 \| 2.0435 \| \| 1.8883 \| 17.0 \| 25500 \| 2.0430 \| \| 1.8867 \| 18.0 \| 27000 \| 2.0416 \| \| 1.8807 \| 19.0 \| 28500 \| 2.0051 \| \| 1.8517 \| 20.0 \| 30000 \| 2.0338 \| \| 1.8357 \| 21.0 \| 31500 \| 2.0166 \| \| 1.8241 \| 22.0 \| 33000 \| 2.0355 \| \| 1.7985 \| 23.0 \| 34500 \| 2.0073 \| \| 1.8061 \| 24.0 \| 36000 \| 2.0473 \| \| 1.7996 \| 25.0 \| 37500 \| 2.0446 \| \| 1.7786 \| 26.0 \| 39000 \| 2.0086 \| \| 1.771 \| 27.0 \| 40500 \| 2.0294 \| \| 1.7549 \| 28.0 \| 42000 \| 2.0127 \| \| 1.7726 \| 29.0 \| 43500 \| 2.0191 \| \| 1.7275 \| 30.0 \| 45000 \| 2.0182 \| \| 1.708 \| 31.0 \| 46500 \| 2.0130 \| \| 1.7345 \| 32.0 \| 48000 \| 2.0155 \| \| 1.7044 \| 33.0 \| 49500 \| 1.9898 \| \| 1.7126 \| 34.0 \| 51000 \| 2.0166 \| \| 1.698 \| 35.0 \| 52500 \| 1.9879 \| \| 1.6637 \| 36.0 \| 54000 \| 2.0311 \| \| 1.6854 \| 37.0 \| 55500 \| 2.0355 \| \| 1.6585 \| 38.0 \| 57000 \| 2.0094 \| \| 1.6418 \| 39.0 \| 58500 \| 2.0042 \| \| 1.667 \| 40.0 \| 60000 \| 2.0116 \| \| 1.6507 \| 41.0 \| 61500 \| 2.0095 \| \| 1.622 \| 42.0 \| 63000 \| 2.0158 \| \| 1.6381 \| 43.0 \| 64500 \| 2.0339 \| \| 1.6099 \| 44.0 \| 66000 \| 2.0082 \| \| 1.6076 \| 45.0 \| 67500 \| 2.0207 \| \| 1.5805 \| 46.0 \| 69000 \| 2.0172 \| \| 1.5862 \| 47.0 \| 70500 \| 2.0132 \| \| 1.5806 \| 48.0 \| 72000 \| 2.0198 \| \| 1.574 \| 49.0 \| 73500 \| 2.0181 \| \| 1.5718 \| 50.0 \| 75000 \| 2.0086 \| \| 1.5591 \| 51.0 \| 76500 \| 1.9832 \| \| 1.5468 \| 52.0 \| 78000 \| 2.0167 \| \| 1.5637 \| 53.0 \| 79500 \| 2.0118 \| \| 1.5117 \| 54.0 \| 81000 \| 2.0290 \| \| 1.5363 \| 55.0 \| 82500 \| 2.0011 \| \| 1.4976 \| 56.0 \| 84000 \| 2.0160 \| \| 1.5129 \| 57.0 \| 85500 \| 2.0224 \| \| 1.4964 \| 58.0 \| 87000 \| 2.0219 \| \| 1.4906 \| 59.0 \| 88500 \| 2.0212 \| \| 1.4941 \| 60.0 \| 90000 \| 2.0255 \| \| 1.4876 \| 61.0 \| 91500 \| 2.0116 \| \| 1.4837 \| 62.0 \| 93000 \| 2.0176 \| \| 1.4661 \| 63.0 \| 94500 \| 2.0388 \| \| 1.4634 \| 64.0 \| 96000 \| 2.0165 \| \| 1.4449 \| 65.0 \| 97500 \| 2.0185 \| \| 1.468 \| 66.0 \| 99000 \| 2.0246 \| \| 1.4567 \| 67.0 \| 100500 \| 2.0244 \| \| 1.4367 \| 68.0 \| 102000 \| 2.0093 \| \| 1.4471 \| 69.0 \| 103500 \| 2.0101 \| \| 1.4255 \| 70.0 \| 105000 \| 2.0248 \| \| 1.4203 \| 71.0 \| 106500 \| 2.0224 \| \| 1.42 \| 72.0 \| 108000 \| 2.0279 \| \| 1.4239 \| 73.0 \| 109500 \| 2.0295 \| \| 1.4126 \| 74.0 \| 111000 \| 2.0196 \| \| 1.4038 \| 75.0 \| 112500 \| 2.0225 \| \| 1.3874 \| 76.0 \| 114000 \| 2.0456 \| \| 1.3758 \| 77.0 \| 115500 \| 2.0423 \| \| 1.3924 \| 78.0 \| 117000 \| 2.0184 \| \| 1.3744 \| 79.0 \| 118500 \| 2.0555 \| \| 1.3622 \| 80.0 \| 120000 \| 2.0387 \| \| 1.3653 \| 81.0 \| 121500 \| 2.0344 \| \| 1.3724 \| 82.0 \| 123000 \| 2.0184 \| \| 1.3684 \| 83.0 \| 124500 \| 2.0285 \| \| 1.3576 \| 84.0 \| 126000 \| 2.0544 \| \| 1.348 \| 85.0 \| 127500 \| 2.0412 \| \| 1.3387 \| 86.0 \| 129000 \| 2.0459 \| \| 1.3416 \| 87.0 \| 130500 \| 2.0329 \| \| 1.3421 \| 88.0 \| 132000 \| 2.0274 \| \| 1.3266 \| 89.0 \| 133500 \| 2.0233 \| \| 1.3183 \| 90.0 \| 135000 \| 2.0319 \| \| 1.322 \| 91.0 \| 136500 \| 2.0080 \| \| 1.32 \| 92.0 \| 138000 \| 2.0472 \| \| 1.304 \| 93.0 \| 139500 \| 2.0538 \| \| 1.3061 \| 94.0 \| 141000 \| 2.0340 \| \| 1.3199 \| 95.0 \| 142500 \| 2.0456 \| \| 1.2985 \| 96.0 \| 144000 \| 2.0167 \| \| 1.3021 \| 97.0 \| 145500 \| 2.0204 \| \| 1.2787 \| 98.0 \| 147000 \| 2.0645 \| \| 1.2879 \| 99.0 \| 148500 \| 2.0345 \| \| 1.2695 \| 100.0 \| 150000 \| 2.0340 \| \| 1.2884 \| 101.0 \| 151500 \| 2.0602 \| \| 1.2747 \| 102.0 \| 153000 \| 2.0667 \| \| 1.2607 \| 103.0 \| 154500 \| 2.0551 \| \| 1.2551 \| 104.0 \| 156000 \| 2.0544 \| \| 1.2557 \| 105.0 \| 157500 \| 2.0553 \| \| 1.2495 \| 106.0 \| 159000 \| 2.0370 \| \| 1.26 \| 107.0 \| 160500 \| 2.0568 \| \| 1.2499 \| 108.0 \| 162000 \| 2.0427 \| \| 1.2438 \| 109.0 \| 163500 \| 2.0184 \| \| 1.2496 \| 110.0 \| 165000 \| 2.0227 \| \| 1.2332 \| 111.0 \| 166500 \| 2.0621 \| \| 1.2231 \| 112.0 \| 168000 \| 2.0661 \| \| 1.211 \| 113.0 \| 169500 \| 2.0673 \| \| 1.217 \| 114.0 \| 171000 \| 2.0544 \| \| 1.2206 \| 115.0 \| 172500 \| 2.0542 \| \| 1.2083 \| 116.0 \| 174000 \| 2.0592 \| \| 1.2205 \| 117.0 \| 175500 \| 2.0451 \| \| 1.2065 \| 118.0 \| 177000 \| 2.0402 \| \| 1.1988 \| 119.0 \| 178500 \| 2.0615 \| \| 1.218 \| 120.0 \| 180000 \| 2.0374 \| \| 1.1917 \| 121.0 \| 181500 \| 2.0349 \| \| 1.1854 \| 122.0 \| 183000 \| 2.0790 \| \| 1.1819 \| 123.0 \| 184500 \| 2.0766 \| \| 1.2029 \| 124.0 \| 186000 \| 2.0364 \| \| 1.1851 \| 125.0 \| 187500 \| 2.0568 \| \| 1.1734 \| 126.0 \| 189000 \| 2.0445 \| \| 1.1701 \| 127.0 \| 190500 \| 2.0770 \| \| 1.1824 \| 128.0 \| 192000 \| 2.0566 \| \| 1.1604 \| 129.0 \| 193500 \| 2.0542 \| \| 1.1733 \| 130.0 \| 195000 \| 2.0525 \| \| 1.1743 \| 131.0 \| 196500 \| 2.0577 \| \| 1.1692 \| 132.0 \| 198000 \| 2.0723 \| \| 1.1519 \| 133.0 \| 199500 \| 2.0567 \| \| 1.1401 \| 134.0 \| 201000 \| 2.0795 \| \| 1.1692 \| 135.0 \| 202500 \| 2.0625 \| \| 1.157 \| 136.0 \| 204000 \| 2.0793 \| \| 1.1495 \| 137.0 \| 205500 \| 2.0782 \| \| 1.1479 \| 138.0 \| 207000 \| 2.0392 \| \| 1.1247 \| 139.0 \| 208500 \| 2.0796 \| \| 1.143 \| 140.0 \| 210000 \| 2.0369 \| \| 1.1324 \| 141.0 \| 211500 \| 2.0699 \| \| 1.1341 \| 142.0 \| 213000 \| 2.0694 \| \| 1.1317 \| 143.0 \| 214500 \| 2.0569 \| \| 1.1254 \| 144.0 \| 216000 \| 2.0545 \| \| 1.1156 \| 145.0 \| 217500 \| 2.0708 \| \| 1.1353 \| 146.0 \| 219000 \| 2.0767 \| \| 1.1312 \| 147.0 \| 220500 \| 2.0523 \| \| 1.1224 \| 148.0 \| 222000 \| 2.0565 \| \| 1.106 \| 149.0 \| 223500 \| 2.0696 \| \| 1.1069 \| 150.0 \| 225000 \| 2.0478 \| \| 1.1011 \| 151.0 \| 226500 \| 2.0475 \| \| 1.0985 \| 152.0 \| 228000 \| 2.0888 \| \| 1.1107 \| 153.0 \| 229500 \| 2.0756 \| \| 1.1058 \| 154.0 \| 231000 \| 2.0812 \| \| 1.1027 \| 155.0 \| 232500 \| 2.0597 \| \| 1.0996 \| 156.0 \| 234000 \| 2.0684 \| \| 1.0987 \| 157.0 \| 235500 \| 2.0629 \| \| 1.0881 \| 158.0 \| 237000 \| 2.0701 \| \| 1.1143 \| 159.0 \| 238500 \| 2.0740 \| \| 1.0823 \| 160.0 \| 240000 \| 2.0869 \| \| 1.0925 \| 161.0 \| 241500 \| 2.0567 \| \| 1.1034 \| 162.0 \| 243000 \| 2.0833 \| \| 1.0759 \| 163.0 \| 244500 \| 2.0585 \| \| 1.0998 \| 164.0 \| 246000 \| 2.0293 \| \| 1.0891 \| 165.0 \| 247500 \| 2.0608 \| \| 1.1036 \| 166.0 \| 249000 \| 2.0831 \| \| 1.076 \| 167.0 \| 250500 \| 2.0979 \| \| 1.0895 \| 168.0 \| 252000 \| 2.0882 \| \| 1.0825 \| 169.0 \| 253500 \| 2.0742 \| \| 1.0793 \| 170.0 \| 255000 \| 2.0841 \| \| 1.079 \| 171.0 \| 256500 \| 2.0829 \| \| 1.0653 \| 172.0 \| 258000 \| 2.0888 \| \| 1.0834 \| 173.0 \| 259500 \| 2.0784 \| \| 1.0721 \| 174.0 \| 261000 \| 2.0859 \| \| 1.0712 \| 175.0 \| 262500 \| 2.0810 \| \| 1.0494 \| 176.0 \| 264000 \| 2.0605 \| \| 1.0654 \| 177.0 \| 265500 \| 2.0623 \| \| 1.077 \| 178.0 \| 267000 \| 2.0756 \| \| 1.056 \| 179.0 \| 268500 \| 2.0782 \| \| 1.0523 \| 180.0 \| 270000 \| 2.0966 \| \| 1.0656 \| 181.0 \| 271500 \| 2.0750 \| \| 1.0636 \| 182.0 \| 273000 \| 2.0769 \| \| 1.0851 \| 183.0 \| 274500 \| 2.0872 \| \| 1.0562 \| 184.0 \| 276000 \| 2.0893 \| \| 1.0534 \| 185.0 \| 277500 \| 2.0661 \| \| 1.0514 \| 186.0 \| 279000 \| 2.0712 \| \| 1.062 \| 187.0 \| 280500 \| 2.0769 \| \| 1.0683 \| 188.0 \| 282000 \| 2.0765 \| \| 1.0606 \| 189.0 \| 283500 \| 2.0735 \| \| 1.0555 \| 190.0 \| 285000 \| 2.0710 \| \| 1.0568 \| 191.0 \| 286500 \| 2.0860 \| \| 1.0502 \| 192.0 \| 288000 \| 2.0587 \| \| 1.0437 \| 193.0 \| 289500 \| 2.0998 \| \| 1.0534 \| 194.0 \| 291000 \| 2.0418 \| \| 1.062 \| 195.0 \| 292500 \| 2.0724 \| \| 1.0457 \| 196.0 \| 294000 \| 2.0612 \| \| 1.0501 \| 197.0 \| 295500 \| 2.1012 \| \| 1.0728 \| 198.0 \| 297000 \| 2.0721 \| \| 1.0413 \| 199.0 \| 298500 \| 2.0535 \| \| 1.0461 \| 200.0 \| 300000 \| 2.0531 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
Izarel/distilbert-base-uncased_fine_tuned_body_text	Izarel	2022-08-01T03:52:20Z	4	0	transformers	[ "transformers", "pytorch", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T19:03:36Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - accuracy - recall - precision - f1 model-index: - name: distilbert-base-uncased_fine_tuned_body_text results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilbert-base-uncased_fine_tuned_body_text This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 1.2153 - Accuracy: {'accuracy': 0.8827265261428963} - Recall: {'recall': 0.8641975308641975} - Precision: {'precision': 0.8900034993584509} - F1: {'f1': 0.8769106999195494} ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 1000 - num_epochs: 15 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Accuracy \| Recall \| Precision \| F1 \| \|:-------------:\|:-----:\|:-----:\|:---------------:\|:--------------------------------:\|:------------------------------:\|:---------------------------------:\|:--------------------------:\| \| 0.3056 \| 1.0 \| 2284 \| 0.3040 \| {'accuracy': 0.8874897344648235} \| {'recall': 0.8466417487824216} \| {'precision': 0.914261252446184} \| {'f1': 0.8791531902381653} \| \| 0.2279 \| 2.0 \| 4568 \| 0.2891 \| {'accuracy': 0.8908294552422666} \| {'recall': 0.8606863744478424} \| {'precision': 0.9086452230060983} \| {'f1': 0.8840158213122382} \| \| 0.1467 \| 3.0 \| 6852 \| 0.3580 \| {'accuracy': 0.8882562277580072} \| {'recall': 0.8452825914599615} \| {'precision': 0.9170557876628164} \| {'f1': 0.8797076678257796} \| \| 0.0921 \| 4.0 \| 9136 \| 0.4560 \| {'accuracy': 0.8754448398576512} \| {'recall': 0.8948918337297542} \| {'precision': 0.8543468858131488} \| {'f1': 0.8741494717043756} \| \| 0.0587 \| 5.0 \| 11420 \| 0.5701 \| {'accuracy': 0.8768135778811935} \| {'recall': 0.8139087099331748} \| {'precision': 0.9221095855254716} \| {'f1': 0.8646372277704246} \| \| 0.0448 \| 6.0 \| 13704 \| 0.6738 \| {'accuracy': 0.8767040788393101} \| {'recall': 0.8794880507418734} \| {'precision': 0.8673070479168994} \| {'f1': 0.873355078168935} \| \| 0.0289 \| 7.0 \| 15988 \| 0.7965 \| {'accuracy': 0.8798248015329866} \| {'recall': 0.8491335372069317} \| {'precision': 0.8967703349282297} \| {'f1': 0.8723020536389552} \| \| 0.0214 \| 8.0 \| 18272 \| 0.8244 \| {'accuracy': 0.8811387900355871} \| {'recall': 0.8576282704723072} \| {'precision': 0.8922931887815225} \| {'f1': 0.8746173837712965} \| \| 0.0147 \| 9.0 \| 20556 \| 0.8740 \| {'accuracy': 0.8806460443471119} \| {'recall': 0.8669158455091177} \| {'precision': 0.8839357893521191} \| {'f1': 0.8753430924062213} \| \| 0.0099 \| 10.0 \| 22840 \| 0.9716 \| {'accuracy': 0.8788940596769779} \| {'recall': 0.8694076339336279} \| {'precision': 0.8787635947338294} \| {'f1': 0.8740605784559327} \| \| 0.0092 \| 11.0 \| 25124 \| 1.0296 \| {'accuracy': 0.8822885299753627} \| {'recall': 0.8669158455091177} \| {'precision': 0.8870089233978444} \| {'f1': 0.876847290640394} \| \| 0.0039 \| 12.0 \| 27408 \| 1.0974 \| {'accuracy': 0.8787845606350945} \| {'recall': 0.8628383735417374} \| {'precision': 0.8836561883772184} \| {'f1': 0.8731232091690544} \| \| 0.0053 \| 13.0 \| 29692 \| 1.0833 \| {'accuracy': 0.8799890500958116} \| {'recall': 0.8503794314191868} \| {'precision': 0.8960496479293472} \| {'f1': 0.8726173872617387} \| \| 0.0032 \| 14.0 \| 31976 \| 1.1731 \| {'accuracy': 0.8813030385984123} \| {'recall': 0.8705402650356778} \| {'precision': 0.8823326828148318} \| {'f1': 0.8763968072976055} \| \| 0.0017 \| 15.0 \| 34260 \| 1.2153 \| {'accuracy': 0.8827265261428963} \| {'recall': 0.8641975308641975} \| {'precision': 0.8900034993584509} \| {'f1': 0.8769106999195494} \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
keithanpai/resnet-50-finetuned-eurosat	keithanpai	2022-07-31T23:54:26Z	16	0	transformers	[ "transformers", "pytorch", "tensorboard", "resnet", "image-classification", "generated_from_trainer", "dataset:imagefolder", "license:apache-2.0", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	image-classification	2022-07-31T23:46:46Z	--- license: apache-2.0 tags: - generated_from_trainer datasets: - imagefolder metrics: - accuracy model-index: - name: resnet-50-finetuned-eurosat results: - task: name: Image Classification type: image-classification dataset: name: imagefolder type: imagefolder config: default split: train args: default metrics: - name: Accuracy type: accuracy value: 0.6676646706586826 --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # resnet-50-finetuned-eurosat This model is a fine-tuned version of [microsoft/resnet-50](https://huggingface.co/microsoft/resnet-50) on the imagefolder dataset. It achieves the following results on the evaluation set: - Loss: 1.1981 - Accuracy: 0.6677 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 128 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 3 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Accuracy \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------:\| \| 1.5279 \| 0.99 \| 70 \| 1.5218 \| 0.6677 \| \| 1.1982 \| 1.99 \| 140 \| 1.2405 \| 0.6677 \| \| 1.0836 \| 2.99 \| 210 \| 1.1981 \| 0.6677 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/distilbert-base-uncased_fold_3_ternary	elopezlopez	2022-07-31T23:52:36Z	12	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T23:35:15Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_3_ternary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilbert-base-uncased_fold_3_ternary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.7987 - F1: 0.7460 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 289 \| 0.5903 \| 0.6893 \| \| 0.5417 \| 2.0 \| 578 \| 0.5822 \| 0.7130 \| \| 0.5417 \| 3.0 \| 867 \| 0.6471 \| 0.7385 \| \| 0.2298 \| 4.0 \| 1156 \| 0.8933 \| 0.7322 \| \| 0.2298 \| 5.0 \| 1445 \| 1.1002 \| 0.7147 \| \| 0.1012 \| 6.0 \| 1734 \| 1.2041 \| 0.7249 \| \| 0.0508 \| 7.0 \| 2023 \| 1.3575 \| 0.7195 \| \| 0.0508 \| 8.0 \| 2312 \| 1.3896 \| 0.7385 \| \| 0.018 \| 9.0 \| 2601 \| 1.5363 \| 0.7238 \| \| 0.018 \| 10.0 \| 2890 \| 1.5336 \| 0.7364 \| \| 0.0142 \| 11.0 \| 3179 \| 1.6335 \| 0.7308 \| \| 0.0142 \| 12.0 \| 3468 \| 1.6915 \| 0.7295 \| \| 0.0047 \| 13.0 \| 3757 \| 1.7087 \| 0.7427 \| \| 0.0058 \| 14.0 \| 4046 \| 1.7875 \| 0.7378 \| \| 0.0058 \| 15.0 \| 4335 \| 1.7649 \| 0.7438 \| \| 0.0051 \| 16.0 \| 4624 \| 1.7987 \| 0.7460 \| \| 0.0051 \| 17.0 \| 4913 \| 1.8435 \| 0.7404 \| \| 0.0025 \| 18.0 \| 5202 \| 1.9623 \| 0.7257 \| \| 0.0025 \| 19.0 \| 5491 \| 1.9005 \| 0.7304 \| \| 0.0029 \| 20.0 \| 5780 \| 1.9437 \| 0.7374 \| \| 0.0011 \| 21.0 \| 6069 \| 1.9840 \| 0.7268 \| \| 0.0011 \| 22.0 \| 6358 \| 1.9411 \| 0.7346 \| \| 0.0025 \| 23.0 \| 6647 \| 1.9233 \| 0.7438 \| \| 0.0025 \| 24.0 \| 6936 \| 1.9415 \| 0.7395 \| \| 0.0015 \| 25.0 \| 7225 \| 1.9481 \| 0.7411 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/xlnet-base-cased_fold_2_binary	elopezlopez	2022-07-31T23:13:47Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "xlnet", "text-classification", "generated_from_trainer", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T22:50:03Z	--- license: mit tags: - generated_from_trainer metrics: - f1 model-index: - name: xlnet-base-cased_fold_2_binary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # xlnet-base-cased_fold_2_binary This model is a fine-tuned version of [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.4858 - F1: 0.7648 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 290 \| 0.4361 \| 0.7404 \| \| 0.4403 \| 2.0 \| 580 \| 0.5363 \| 0.7515 \| \| 0.4403 \| 3.0 \| 870 \| 0.4858 \| 0.7648 \| \| 0.2505 \| 4.0 \| 1160 \| 0.7127 \| 0.7612 \| \| 0.2505 \| 5.0 \| 1450 \| 0.8930 \| 0.7554 \| \| 0.1425 \| 6.0 \| 1740 \| 0.9897 \| 0.7580 \| \| 0.0869 \| 7.0 \| 2030 \| 1.2683 \| 0.7615 \| \| 0.0869 \| 8.0 \| 2320 \| 1.4988 \| 0.7343 \| \| 0.0411 \| 9.0 \| 2610 \| 1.5082 \| 0.7492 \| \| 0.0411 \| 10.0 \| 2900 \| 1.4974 \| 0.7450 \| \| 0.0306 \| 11.0 \| 3190 \| 1.5723 \| 0.7435 \| \| 0.0306 \| 12.0 \| 3480 \| 1.8446 \| 0.7432 \| \| 0.0291 \| 13.0 \| 3770 \| 1.7113 \| 0.7639 \| \| 0.0207 \| 14.0 \| 4060 \| 1.8073 \| 0.7394 \| \| 0.0207 \| 15.0 \| 4350 \| 1.7524 \| 0.7585 \| \| 0.0171 \| 16.0 \| 4640 \| 1.8751 \| 0.7374 \| \| 0.0171 \| 17.0 \| 4930 \| 1.7849 \| 0.7561 \| \| 0.0084 \| 18.0 \| 5220 \| 1.8618 \| 0.7441 \| \| 0.0064 \| 19.0 \| 5510 \| 1.9613 \| 0.7437 \| \| 0.0064 \| 20.0 \| 5800 \| 1.8898 \| 0.7430 \| \| 0.006 \| 21.0 \| 6090 \| 1.9889 \| 0.7409 \| \| 0.006 \| 22.0 \| 6380 \| 1.9949 \| 0.7488 \| \| 0.0049 \| 23.0 \| 6670 \| 1.9453 \| 0.7488 \| \| 0.0049 \| 24.0 \| 6960 \| 1.9754 \| 0.7472 \| \| 0.002 \| 25.0 \| 7250 \| 1.9946 \| 0.7504 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/xlnet-base-cased_fold_1_binary	elopezlopez	2022-07-31T22:49:49Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "xlnet", "text-classification", "generated_from_trainer", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T22:26:16Z	--- license: mit tags: - generated_from_trainer metrics: - f1 model-index: - name: xlnet-base-cased_fold_1_binary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # xlnet-base-cased_fold_1_binary This model is a fine-tuned version of [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.7607 - F1: 0.7778 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 288 \| 0.4111 \| 0.7555 \| \| 0.4387 \| 2.0 \| 576 \| 0.4075 \| 0.7540 \| \| 0.4387 \| 3.0 \| 864 \| 0.5344 \| 0.7567 \| \| 0.2471 \| 4.0 \| 1152 \| 0.7405 \| 0.7597 \| \| 0.2471 \| 5.0 \| 1440 \| 1.0564 \| 0.7508 \| \| 0.1419 \| 6.0 \| 1728 \| 1.0703 \| 0.7751 \| \| 0.0845 \| 7.0 \| 2016 \| 1.0866 \| 0.7609 \| \| 0.0845 \| 8.0 \| 2304 \| 1.2135 \| 0.7751 \| \| 0.05 \| 9.0 \| 2592 \| 1.3649 \| 0.7516 \| \| 0.05 \| 10.0 \| 2880 \| 1.4943 \| 0.7590 \| \| 0.0267 \| 11.0 \| 3168 \| 1.5174 \| 0.7412 \| \| 0.0267 \| 12.0 \| 3456 \| 1.4884 \| 0.7559 \| \| 0.0278 \| 13.0 \| 3744 \| 1.5109 \| 0.7405 \| \| 0.0201 \| 14.0 \| 4032 \| 1.7251 \| 0.7409 \| \| 0.0201 \| 15.0 \| 4320 \| 1.5833 \| 0.7354 \| \| 0.0185 \| 16.0 \| 4608 \| 1.7744 \| 0.7598 \| \| 0.0185 \| 17.0 \| 4896 \| 1.8283 \| 0.7619 \| \| 0.0066 \| 18.0 \| 5184 \| 1.7607 \| 0.7778 \| \| 0.0066 \| 19.0 \| 5472 \| 1.7503 \| 0.7719 \| \| 0.0078 \| 20.0 \| 5760 \| 1.7807 \| 0.7508 \| \| 0.006 \| 21.0 \| 6048 \| 1.6887 \| 0.7629 \| \| 0.006 \| 22.0 \| 6336 \| 1.7041 \| 0.7678 \| \| 0.0074 \| 23.0 \| 6624 \| 1.7337 \| 0.7633 \| \| 0.0074 \| 24.0 \| 6912 \| 1.7548 \| 0.7645 \| \| 0.0035 \| 25.0 \| 7200 \| 1.7685 \| 0.7621 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/distilbert-base-uncased_fold_6_binary	elopezlopez	2022-07-31T22:25:18Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T22:14:59Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_6_binary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilbert-base-uncased_fold_6_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.6838 - F1: 0.7881 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 290 \| 0.4181 \| 0.7732 \| \| 0.4097 \| 2.0 \| 580 \| 0.3967 \| 0.7697 \| \| 0.4097 \| 3.0 \| 870 \| 0.5811 \| 0.7797 \| \| 0.2034 \| 4.0 \| 1160 \| 0.8684 \| 0.7320 \| \| 0.2034 \| 5.0 \| 1450 \| 0.9116 \| 0.7718 \| \| 0.0794 \| 6.0 \| 1740 \| 1.0588 \| 0.7690 \| \| 0.0278 \| 7.0 \| 2030 \| 1.2092 \| 0.7738 \| \| 0.0278 \| 8.0 \| 2320 \| 1.2180 \| 0.7685 \| \| 0.0233 \| 9.0 \| 2610 \| 1.3005 \| 0.7676 \| \| 0.0233 \| 10.0 \| 2900 \| 1.4009 \| 0.7634 \| \| 0.0093 \| 11.0 \| 3190 \| 1.4528 \| 0.7805 \| \| 0.0093 \| 12.0 \| 3480 \| 1.4803 \| 0.7859 \| \| 0.0088 \| 13.0 \| 3770 \| 1.4775 \| 0.7750 \| \| 0.0077 \| 14.0 \| 4060 \| 1.6171 \| 0.7699 \| \| 0.0077 \| 15.0 \| 4350 \| 1.6429 \| 0.7636 \| \| 0.0047 \| 16.0 \| 4640 \| 1.5619 \| 0.7819 \| \| 0.0047 \| 17.0 \| 4930 \| 1.5833 \| 0.7724 \| \| 0.0034 \| 18.0 \| 5220 \| 1.6400 \| 0.7853 \| \| 0.0008 \| 19.0 \| 5510 \| 1.6508 \| 0.7792 \| \| 0.0008 \| 20.0 \| 5800 \| 1.6838 \| 0.7881 \| \| 0.0009 \| 21.0 \| 6090 \| 1.6339 \| 0.7829 \| \| 0.0009 \| 22.0 \| 6380 \| 1.6824 \| 0.7806 \| \| 0.0016 \| 23.0 \| 6670 \| 1.6867 \| 0.7876 \| \| 0.0016 \| 24.0 \| 6960 \| 1.7107 \| 0.7877 \| \| 0.0013 \| 25.0 \| 7250 \| 1.6933 \| 0.7812 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/distilbert-base-uncased_fold_5_binary	elopezlopez	2022-07-31T22:14:52Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T22:04:19Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_5_binary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilbert-base-uncased_fold_5_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.5093 - F1: 0.7801 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 288 \| 0.4760 \| 0.7315 \| \| 0.3992 \| 2.0 \| 576 \| 0.4428 \| 0.7785 \| \| 0.3992 \| 3.0 \| 864 \| 0.5093 \| 0.7801 \| \| 0.2021 \| 4.0 \| 1152 \| 0.6588 \| 0.7634 \| \| 0.2021 \| 5.0 \| 1440 \| 0.9174 \| 0.7713 \| \| 0.0945 \| 6.0 \| 1728 \| 0.9832 \| 0.7726 \| \| 0.0321 \| 7.0 \| 2016 \| 1.2103 \| 0.7672 \| \| 0.0321 \| 8.0 \| 2304 \| 1.3759 \| 0.7616 \| \| 0.0134 \| 9.0 \| 2592 \| 1.4405 \| 0.7570 \| \| 0.0134 \| 10.0 \| 2880 \| 1.4591 \| 0.7710 \| \| 0.0117 \| 11.0 \| 3168 \| 1.4947 \| 0.7713 \| \| 0.0117 \| 12.0 \| 3456 \| 1.6224 \| 0.7419 \| \| 0.0081 \| 13.0 \| 3744 \| 1.6462 \| 0.7520 \| \| 0.0083 \| 14.0 \| 4032 \| 1.6880 \| 0.7637 \| \| 0.0083 \| 15.0 \| 4320 \| 1.7080 \| 0.7380 \| \| 0.0048 \| 16.0 \| 4608 \| 1.7352 \| 0.7551 \| \| 0.0048 \| 17.0 \| 4896 \| 1.6761 \| 0.7713 \| \| 0.0024 \| 18.0 \| 5184 \| 1.7553 \| 0.76 \| \| 0.0024 \| 19.0 \| 5472 \| 1.7312 \| 0.7673 \| \| 0.005 \| 20.0 \| 5760 \| 1.7334 \| 0.7713 \| \| 0.0032 \| 21.0 \| 6048 \| 1.7963 \| 0.7578 \| \| 0.0032 \| 22.0 \| 6336 \| 1.7529 \| 0.7679 \| \| 0.0025 \| 23.0 \| 6624 \| 1.7741 \| 0.7662 \| \| 0.0025 \| 24.0 \| 6912 \| 1.7515 \| 0.7679 \| \| 0.0004 \| 25.0 \| 7200 \| 1.7370 \| 0.7765 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/distilbert-base-uncased_fold_3_binary	elopezlopez	2022-07-31T21:53:55Z	4	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T21:43:33Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_3_binary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilbert-base-uncased_fold_3_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.8310 - F1: 0.7584 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 289 \| 0.4560 \| 0.7522 \| \| 0.4008 \| 2.0 \| 578 \| 0.4790 \| 0.7567 \| \| 0.4008 \| 3.0 \| 867 \| 0.6368 \| 0.7557 \| \| 0.1967 \| 4.0 \| 1156 \| 0.6854 \| 0.7534 \| \| 0.1967 \| 5.0 \| 1445 \| 0.9823 \| 0.7309 \| \| 0.0768 \| 6.0 \| 1734 \| 1.2531 \| 0.7511 \| \| 0.0202 \| 7.0 \| 2023 \| 1.2906 \| 0.7391 \| \| 0.0202 \| 8.0 \| 2312 \| 1.4025 \| 0.7460 \| \| 0.0087 \| 9.0 \| 2601 \| 1.5713 \| 0.7507 \| \| 0.0087 \| 10.0 \| 2890 \| 1.4212 \| 0.7528 \| \| 0.0162 \| 11.0 \| 3179 \| 1.5775 \| 0.7511 \| \| 0.0162 \| 12.0 \| 3468 \| 1.6361 \| 0.7377 \| \| 0.0048 \| 13.0 \| 3757 \| 1.6972 \| 0.7542 \| \| 0.0098 \| 14.0 \| 4046 \| 1.6569 \| 0.7565 \| \| 0.0098 \| 15.0 \| 4335 \| 1.7547 \| 0.7325 \| \| 0.0042 \| 16.0 \| 4624 \| 1.8108 \| 0.7544 \| \| 0.0042 \| 17.0 \| 4913 \| 1.7593 \| 0.7554 \| \| 0.0041 \| 18.0 \| 5202 \| 1.7582 \| 0.7551 \| \| 0.0041 \| 19.0 \| 5491 \| 1.8200 \| 0.7512 \| \| 0.0029 \| 20.0 \| 5780 \| 1.8310 \| 0.7584 \| \| 0.0018 \| 21.0 \| 6069 \| 1.8146 \| 0.7568 \| \| 0.0018 \| 22.0 \| 6358 \| 1.7870 \| 0.7558 \| \| 0.0029 \| 23.0 \| 6647 \| 1.8508 \| 0.7530 \| \| 0.0029 \| 24.0 \| 6936 \| 1.8327 \| 0.7543 \| \| 0.0001 \| 25.0 \| 7225 \| 1.8546 \| 0.7561 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
elopezlopez/distilbert-base-uncased_fold_1_binary	elopezlopez	2022-07-31T21:33:03Z	6	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T20:57:24Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_1_binary results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilbert-base-uncased_fold_1_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.5992 - F1: 0.7687 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| No log \| 1.0 \| 288 \| 0.3960 \| 0.7467 \| \| 0.3988 \| 2.0 \| 576 \| 0.3947 \| 0.7487 \| \| 0.3988 \| 3.0 \| 864 \| 0.4511 \| 0.7662 \| \| 0.1853 \| 4.0 \| 1152 \| 0.7226 \| 0.7285 \| \| 0.1853 \| 5.0 \| 1440 \| 0.9398 \| 0.7334 \| \| 0.0827 \| 6.0 \| 1728 \| 1.0547 \| 0.7427 \| \| 0.0287 \| 7.0 \| 2016 \| 1.1602 \| 0.7563 \| \| 0.0287 \| 8.0 \| 2304 \| 1.3332 \| 0.7171 \| \| 0.0219 \| 9.0 \| 2592 \| 1.3429 \| 0.7420 \| \| 0.0219 \| 10.0 \| 2880 \| 1.2603 \| 0.7648 \| \| 0.0139 \| 11.0 \| 3168 \| 1.4126 \| 0.7569 \| \| 0.0139 \| 12.0 \| 3456 \| 1.3195 \| 0.7483 \| \| 0.0115 \| 13.0 \| 3744 \| 1.4356 \| 0.7491 \| \| 0.0035 \| 14.0 \| 4032 \| 1.5693 \| 0.7636 \| \| 0.0035 \| 15.0 \| 4320 \| 1.4071 \| 0.7662 \| \| 0.0071 \| 16.0 \| 4608 \| 1.4561 \| 0.7579 \| \| 0.0071 \| 17.0 \| 4896 \| 1.5405 \| 0.7634 \| \| 0.0041 \| 18.0 \| 5184 \| 1.5862 \| 0.7589 \| \| 0.0041 \| 19.0 \| 5472 \| 1.6782 \| 0.76 \| \| 0.0024 \| 20.0 \| 5760 \| 1.5699 \| 0.7677 \| \| 0.0006 \| 21.0 \| 6048 \| 1.5991 \| 0.7467 \| \| 0.0006 \| 22.0 \| 6336 \| 1.6205 \| 0.7682 \| \| 0.0003 \| 23.0 \| 6624 \| 1.6334 \| 0.7643 \| \| 0.0003 \| 24.0 \| 6912 \| 1.5992 \| 0.7687 \| \| 0.0011 \| 25.0 \| 7200 \| 1.6053 \| 0.7624 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
DS-20202/DoubleHardDebias	DS-20202	2022-07-31T20:32:45Z	0	0	null	[ "license:mit", "region:us" ]	null	2022-07-31T12:08:09Z	--- title: Double Hard Debiasing emoji: 👁 colorFrom: blue colorTo: pink sdk: gradio sdk_version: 3.1.1 app_file: app.py pinned: false license: mit --- Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
neuralmagic/oBERT-6-downstream-pruned-block4-90-QAT-squadv1	neuralmagic	2022-07-31T19:52:34Z	5	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T19:21:02Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-pruned-block4-90-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - Sparsity 90% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 6 ``` The dev-set performance of this model: ``` EM = 76.56 F1 = 84.59 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-6-downstream-pruned-unstructured-90-squadv1	neuralmagic	2022-07-31T19:52:34Z	4	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T14:00:05Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-pruned-unstructured-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - Sparsity 90% - unstructured`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 6 ``` The dev-set performance of this model: ``` EM = 79.16 F1 = 86.78 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-teacher-squadv1	neuralmagic	2022-07-31T19:52:34Z	396	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:47:26Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # SQuADv1 teacher This model is used as a teacher for all runs on the SQuADv1 downstream task in the paper [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). SQuADv1 dev-set: ``` EM = 81.41 F1 = 88.54 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-3-downstream-pruned-block4-90-squadv1	neuralmagic	2022-07-31T19:52:33Z	5	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T14:01:41Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-block4-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 90% - 4-block`. ``` Pruning method: oBERT downstream block-4 Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 71.36 F1 = 80.69 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-3-downstream-dense-squadv1	neuralmagic	2022-07-31T19:52:33Z	5	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T14:00:43Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-dense-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - 0% Sparsity`, and it represents an upper bound for performance of the corresponding pruned models: - 80% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-80-squadv1` - 80% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-80-squadv1` - 90% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-90-squadv1` - 90% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-90-squadv1` SQuADv1 dev-set: ``` EM = 76.62 F1 = 84.65 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-6-downstream-dense-QAT-squadv1	neuralmagic	2022-07-31T19:52:33Z	2	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T19:20:36Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-dense-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - 0% Sparsity - QAT`, and it represents an upper bound for performance of the corresponding pruned and quantized models: - 80% unstructured QAT: `neuralmagic/oBERT-6-downstream-pruned-unstructured-80-QAT-squadv1` - 80% block-4 QAT: `neuralmagic/oBERT-6-downstream-pruned-block4-80-QAT-squadv1` - 90% unstructured QAT: `neuralmagic/oBERT-6-downstream-pruned-unstructured-90-QAT-squadv1` - 90% block-4 QAT: `neuralmagic/oBERT-6-downstream-pruned-block4-90-QAT-squadv1` SQuADv1 dev-set: ``` EM = 80.85 F1 = 87.94 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-3-downstream-pruned-block4-80-QAT-squadv1	neuralmagic	2022-07-31T19:52:33Z	5	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T19:21:28Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-block4-80-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 80% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 80% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 72.70 F1 = 82.04 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-6-downstream-dense-squadv1	neuralmagic	2022-07-31T19:52:33Z	8	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:59:35Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-dense-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - 0% Sparsity`, and it represents an upper bound for performance of the corresponding pruned models: - 80% unstructured: `neuralmagic/oBERT-6-downstream-pruned-unstructured-80-squadv1` - 80% block-4: `neuralmagic/oBERT-6-downstream-pruned-block4-80-squadv1` - 90% unstructured: `neuralmagic/oBERT-6-downstream-pruned-unstructured-90-squadv1` - 90% block-4: `neuralmagic/oBERT-6-downstream-pruned-block4-90-squadv1` SQuADv1 dev-set: ``` EM = 81.17 F1 = 88.32 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-3-downstream-pruned-block4-90-QAT-squadv1	neuralmagic	2022-07-31T19:52:33Z	14	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T19:21:41Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-block4-90-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 90% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 70.00 F1 = 79.66 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-3-downstream-pruned-unstructured-90-squadv1	neuralmagic	2022-07-31T19:52:33Z	13	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T14:01:15Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-unstructured-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 90% - unstructured`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 73.61 F1 = 82.50 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-3-upstream-pretrained-dense	neuralmagic	2022-07-31T19:52:33Z	14	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:56:43Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-3-upstream-pretrained-dense This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to 3 layers from `neuralmagic/oBERT-12-upstream-pretrained-dense`, pretrained with knowledge distillation. This model is used as a starting point for downstream finetuning and pruning runs presented in the `Table 3 - 3 Layers`. The model can also be used for finetuning on any downstream task, as a starting point instead of the three times larger `bert-base-uncased` model. Finetuned and pruned versions of this model on the SQuADv1 downstream task, as described in the paper: - 0%: `neuralmagic/oBERT-3-downstream-dense-squadv1` - 80% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-80-squadv1` - 80% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-80-squadv1` - 90% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-90-squadv1` - 90% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-90-squadv1` ``` Training objective: masked language modeling (MLM) + knowledge distillation Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 0% Number of layers: 3 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pruned-unstructured-97-v2	neuralmagic	2022-07-31T19:52:32Z	8	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-06-17T07:25:30Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-12-upstream-pruned-unstructured-97-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the upstream pruned model used as a starting point for sparse-transfer learning to downstream tasks presented in the `Table 2 - oBERT - {SQuADv1, MNLI, QQP} - 97%` (in the upcoming updated version of the paper). Finetuned versions of this model for each downstream task are: - SQuADv1: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-squadv1-v2` - MNLI: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-mnli-v2` - QQP: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-qqp-v2` ``` Pruning method: oBERT upstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 97% Number of layers: 12 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-squadv1-v2	neuralmagic	2022-07-31T19:52:32Z	5	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-06-17T07:30:56Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-upstream-pruned-unstructured-97-finetuned-squadv1-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - SQuADv1 97%` (in the upcoming updated version of the paper). ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 97% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `()`): ``` \| oBERT 97% \| F1 \| EM \| \| ------------- \| ----- \| ----- \| \| seed=42 \| 84.92 \| 76.94 \| \| seed=3407 \| 84.87 \| 76.71 \| \| seed=123 \| 84.95 \| 77.06 \| \| seed=12345 ()\| 84.95 \| 76.90 \| \| ------------- \| ----- \| ----- \| \| mean \| 84.92 \| 76.90 \| \| stdev \| 0.037 \| 0.145 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp-v2	neuralmagic	2022-07-31T19:52:32Z	7	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:qqp", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-06-17T07:31:44Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: qqp --- # oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - QQP 90%` (in the upcoming updated version of the paper). ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: QQP Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `()`): ``` \| oBERT 90% \| acc \| F1 \| \| ------------- \| ----- \| ----- \| \| seed=42 \| 90.94 \| 87.79 \| \| seed=3407 \| 91.00 \| 87.81 \| \| seed=123 \| 90.94 \| 87.73 \| \| seed=12345 ()\| 91.07 \| 87.92 \| \| ------------- \| ----- \| ----- \| \| mean \| 90.99 \| 87.81 \| \| stdev \| 0.061 \| 0.079 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2	neuralmagic	2022-07-31T19:52:32Z	7	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-06-17T07:30:41Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - SQuADv1 90%` (in the upcoming updated version of the paper). ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `()`): ``` \| oBERT 90% \| F1 \| EM \| \| ------------ \| ----- \| ----- \| \| seed=42 \| 88.55 \| 81.48 \| \| seed=3407 \| 88.34 \| 81.25 \| \| seed=123 ()\| 88.64 \| 81.57 \| \| seed=12345 \| 88.44 \| 81.43 \| \| ------------ \| ----- \| ----- \| \| mean \| 88.49 \| 81.43 \| \| stdev \| 0.130 \| 0.134 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pruned-unstructured-90-v2	neuralmagic	2022-07-31T19:52:32Z	4	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-06-17T07:22:37Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-12-upstream-pruned-unstructured-90-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the upstream pruned model used as a starting point for sparse-transfer learning to downstream tasks presented in the `Table 2 - oBERT - {SQuADv1, MNLI, QQP} - 90%` (in the upcoming updated version of the paper). Finetuned versions of this model for each downstream task are: - SQuADv1: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2` - MNLI: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-mnli-v2` - QQP: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp-v2` ``` Pruning method: oBERT upstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 90% Number of layers: 12 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp	neuralmagic	2022-07-31T19:52:32Z	10	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:qqp", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:58:30Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: qqp --- # oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - QQP 90%`. ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: QQP Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 90% \| acc \| F1 \| \| ------------ \| ----- \| ----- \| \| seed=42 ()\| 90.93 \| 87.77 \| \| seed=3407 \| 90.70 \| 87.49 \| \| seed=54321 \| 90.86 \| 87.68 \| \| ------------ \| ----- \| ----- \| \| mean \| 90.83 \| 87.65 \| \| stdev \| 0.117 \| 0.143 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-unstructured-90-mnli	neuralmagic	2022-07-31T19:52:31Z	19	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:mnli", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:54:55Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: mnli --- # oBERT-12-downstream-pruned-unstructured-90-mnli This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - MNLI 90%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: MNLI Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 90% \| m-acc \| mm-acc\| \| ------------ \| ----- \| ----- \| \| seed=42 \| 83.74 \| 84.31 \| \| seed=3407 ()\| 83.85 \| 84.40 \| \| seed=54321 \| 83.77 \| 84.33 \| \| ------------ \| ----- \| ----- \| \| mean \| 83.79 \| 84.35 \| \| stdev \| 0.056 \| 0.047 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-unstructured-97-mnli	neuralmagic	2022-07-31T19:52:31Z	8	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:mnli", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:55:09Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: mnli --- # oBERT-12-downstream-pruned-unstructured-97-mnli This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - MNLI 97%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: MNLI Sparsity: 97% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 97% \| m-acc \| mm-acc\| \| ------------ \| ----- \| ----- \| \| seed=42 ()\| 82.10 \| 81.94 \| \| seed=3407 \| 81.81 \| 82.27 \| \| seed=54321 \| 81.40 \| 81.83 \| \| ------------ \| ----- \| ----- \| \| mean \| 81.77 \| 82.01 \| \| stdev \| 0.351 \| 0.228 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-unstructured-90-squadv1	neuralmagic	2022-07-31T19:52:31Z	7	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:53:32Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-downstream-pruned-unstructured-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - SQuADv1 90%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 90% \| F1 \| EM \| \| ------------ \| ----- \| ----- \| \| seed=42 \| 88.22 \| 81.10 \| \| seed=3407 ()\| 88.46 \| 81.26 \| \| seed=54321 \| 88.26 \| 81.00 \| \| ------------ \| ----- \| ----- \| \| mean \| 88.31 \| 81.12 \| \| stdev \| 0.128 \| 0.131 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-unstructured-90-qqp	neuralmagic	2022-07-31T19:52:31Z	12	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:qqp", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:55:50Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: qqp --- # oBERT-12-downstream-pruned-unstructured-90-qqp This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - QQP 90%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: QQP Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 90% \| acc \| F1 \| \| ------------ \| ----- \| ----- \| \| seed=42 \| 91.30 \| 88.24 \| \| seed=3407 ()\| 91.39 \| 88.36 \| \| seed=54321 \| 91.36 \| 88.29 \| \| ------------ \| ----- \| ----- \| \| mean \| 91.35 \| 88.30 \| \| stdev \| 0.045 \| 0.060 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-upstream-pretrained-dense	neuralmagic	2022-07-31T19:52:31Z	4	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:56:17Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-12-upstream-pretrained-dense This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the pretrained dense model used as a teacher for upstream pruning runs, as described in the paper. The model can be finetuned on any downstream task, just like the standard `bert-base-uncased` model which is used as initialization for training of this model. Sparse versions of this model: - 90% sparse: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90` - 97% sparse: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97` ``` Training objective: masked language modeling (MLM) Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 0% Number of layers: 12 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-unstructured-80-squadv1	neuralmagic	2022-07-31T19:52:31Z	11	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:53:16Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-downstream-pruned-unstructured-80-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - SQuADv1 80%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 80% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 80% \| F1 \| EM \| \| ------------ \| ----- \| ----- \| \| seed=42 \| 88.95 \| 82.08 \| \| seed=3407 ()\| 89.16 \| 82.05 \| \| seed=54321 \| 89.01 \| 82.12 \| \| ------------ \| ----- \| ----- \| \| mean \| 89.04 \| 82.08 \| \| stdev \| 0.108 \| 0.035 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-block4-90-QAT-squadv1	neuralmagic	2022-07-31T19:52:30Z	4	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T19:20:22Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-downstream-pruned-block4-90-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 12 Layers - Sparsity 90% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 12 ``` The dev-set performance of this model: ``` EM = 78.84 F1 = 86.68 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
neuralmagic/oBERT-12-downstream-pruned-unstructured-80-mnli	neuralmagic	2022-07-31T19:52:30Z	7	0	transformers	[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:mnli", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]	null	2022-05-25T13:54:40Z	--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: mnli --- # oBERT-12-downstream-pruned-unstructured-80-mnli This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - MNLI 80%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: MNLI Sparsity: 80% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `()`): ``` \| oBERT 80% \| m-acc \| mm-acc\| \| ------------ \| ----- \| ----- \| \| seed=42 \| 84.30 \| 84.98 \| \| seed=3407 ()\| 84.46 \| 84.99 \| \| seed=54321 \| 84.18 \| 84.76 \| \| ------------ \| ----- \| ----- \| \| mean \| 84.32 \| 84.91 \| \| stdev \| 0.140 \| 0.133 \| ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```
Ebuu/Aaaaa	Ebuu	2022-07-31T19:00:30Z	0	0	null	[ "license:bigscience-bloom-rail-1.0", "region:us" ]	null	2022-07-31T19:00:30Z	--- license: bigscience-bloom-rail-1.0 ---
SummerChiam/rust_image_classification_12	SummerChiam	2022-07-31T17:33:58Z	48	0	transformers	[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	image-classification	2022-07-31T17:33:47Z	--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: rust_image_classification_12 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9679595232009888 --- # rust_image_classification_12 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for anything by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### nonrust0 ![nonrust0](images/nonrust0.png) #### rust0 ![rust0](images/rust0.png)
QuickSilver007/Reinforce-Pong-PLE-v0	QuickSilver007	2022-07-31T16:23:22Z	0	0	null	[ "Pong-PLE-v0", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]	reinforcement-learning	2022-07-31T16:23:13Z	--- tags: - Pong-PLE-v0 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-Pong-PLE-v0 results: - metrics: - type: mean_reward value: -16.00 +/- 0.00 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: Pong-PLE-v0 type: Pong-PLE-v0 --- # Reinforce Agent playing Pong-PLE-v0 This is a trained model of a Reinforce agent playing Pong-PLE-v0 . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5
anneke/finetuning-distilbert-base-uncased-finetuned-sst-2-english-5000-samples-final	anneke	2022-07-31T16:05:59Z	7	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T15:49:08Z	--- license: apache-2.0 tags: - generated_from_trainer metrics: - accuracy - f1 model-index: - name: finetuning-distilbert-base-uncased-finetuned-sst-2-english-5000-samples-final results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # finetuning-distilbert-base-uncased-finetuned-sst-2-english-5000-samples-final This model is a fine-tuned version of [distilbert-base-uncased-finetuned-sst-2-english](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.1289 - Accuracy: 0.977 - F1: 0.9878 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 5 ### Training results ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
SummerChiam/pond_image_classification_11	SummerChiam	2022-07-31T15:36:10Z	50	0	transformers	[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	image-classification	2022-07-31T15:35:57Z	--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: pond_image_classification_11 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9951980710029602 --- # pond_image_classification_11 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for anything by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### Algae0 ![Algae0](images/Algae0.png) #### Boiling0 ![Boiling0](images/Boiling0.png) #### BoilingNight0 ![BoilingNight0](images/BoilingNight0.png) #### Normal0 ![Normal0](images/Normal0.png) #### NormalCement0 ![NormalCement0](images/NormalCement0.png) #### NormalNight0 ![NormalNight0](images/NormalNight0.png) #### NormalRain0 ![NormalRain0](images/NormalRain0.png)
samwit/ddpm-afhq-cats-128	samwit	2022-07-31T15:31:53Z	2	0	diffusers	[ "diffusers", "tensorboard", "en", "dataset:imagefolder", "license:apache-2.0", "diffusers:DDPMPipeline", "region:us" ]	null	2022-07-31T00:49:28Z	--- language: en license: apache-2.0 library_name: diffusers tags: [] datasets: imagefolder metrics: [] --- <!-- This model card has been generated automatically according to the information the training script had access to. You should probably proofread and complete it, then remove this comment. --> # ddpm-afhq-cats-128 ## Model description This diffusion model is trained with the [🤗 Diffusers](https://github.com/huggingface/diffusers) library on the `imagefolder` dataset. ## Intended uses & limitations #### How to use ```python # TODO: add an example code snippet for running this diffusion pipeline ``` #### Limitations and bias [TODO: provide examples of latent issues and potential remediations] ## Training data [TODO: describe the data used to train the model] ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001 - train_batch_size: 16 - eval_batch_size: 16 - gradient_accumulation_steps: 1 - optimizer: AdamW with betas=(None, None), weight_decay=None and epsilon=None - lr_scheduler: None - lr_warmup_steps: 500 - ema_inv_gamma: None - ema_inv_gamma: None - ema_inv_gamma: None - mixed_precision: fp16 ### Training results 📈 [TensorBoard logs](https://huggingface.co/samwit/ddpm-afhq-cats-128/tensorboard?#scalars)
CuteBlack/gfp_guided_diffusion_200k	CuteBlack	2022-07-31T15:10:42Z	0	6	null	[ "license:mit", "region:us" ]	null	2022-07-15T22:24:34Z	--- license: mit --- 256x256 Diffusion model trained on 1000+ NSFW gay furry pics (with same composition) 'attention_resolutions': '16', 'class_cond': False, 'diffusion_steps': 1000, 'rescale_timesteps': True, 'timestep_respacing': 'ddim100', 'image_size': 256, 'learn_sigma': True, 'noise_schedule': 'linear', 'num_channels': 128, 'num_heads': 1, 'num_res_blocks': 2, 'use_checkpoint': use_checkpoint, 'use_fp16': True, 'use_scale_shift_norm': False,
Kinahem/Reinforce-3	Kinahem	2022-07-31T13:02:51Z	0	0	null	[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]	reinforcement-learning	2022-07-31T13:02:35Z	--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-3 results: - metrics: - type: mean_reward value: 471.20 +/- 86.40 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 --- # Reinforce Agent playing CartPole-v1 This is a trained model of a Reinforce agent playing CartPole-v1 . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5
Vasanth/bert_emo_classifier	Vasanth	2022-07-31T12:34:43Z	4	1	transformers	[ "transformers", "pytorch", "tensorboard", "bert", "text-classification", "generated_from_trainer", "dataset:emotion", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-30T23:30:12Z	--- license: apache-2.0 tags: - generated_from_trainer datasets: - emotion model-index: - name: bert_emo_classifier results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # bert_emo_classifier This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the emotion dataset. It achieves the following results on the evaluation set: - Loss: 0.2748 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 4 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:----:\|:---------------:\| \| 0.9063 \| 0.25 \| 500 \| 0.4845 \| \| 0.3362 \| 0.5 \| 1000 \| 0.3492 \| \| 0.2759 \| 0.75 \| 1500 \| 0.2819 \| \| 0.2521 \| 1.0 \| 2000 \| 0.2464 \| \| 0.1705 \| 1.25 \| 2500 \| 0.2345 \| \| 0.1841 \| 1.5 \| 3000 \| 0.2013 \| \| 0.1428 \| 1.75 \| 3500 \| 0.1926 \| \| 0.1747 \| 2.0 \| 4000 \| 0.1866 \| \| 0.1082 \| 2.25 \| 4500 \| 0.2302 \| \| 0.1142 \| 2.5 \| 5000 \| 0.2118 \| \| 0.1205 \| 2.75 \| 5500 \| 0.2318 \| \| 0.1135 \| 3.0 \| 6000 \| 0.2306 \| \| 0.0803 \| 3.25 \| 6500 \| 0.2625 \| \| 0.0745 \| 3.5 \| 7000 \| 0.2850 \| \| 0.085 \| 3.75 \| 7500 \| 0.2719 \| \| 0.0701 \| 4.0 \| 8000 \| 0.2748 \| ### Framework versions - Transformers 4.15.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.10.3
QuickSilver007/Reinforce-CartPole-v1	QuickSilver007	2022-07-31T12:21:39Z	0	0	null	[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]	reinforcement-learning	2022-07-31T12:21:29Z	--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-CartPole-v1 results: - metrics: - type: mean_reward value: 123.40 +/- 12.38 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 --- # Reinforce Agent playing CartPole-v1 This is a trained model of a Reinforce agent playing CartPole-v1 . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5
Kinahem/Reinforce-1	Kinahem	2022-07-31T12:07:01Z	0	0	null	[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]	reinforcement-learning	2022-07-31T12:06:53Z	--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-1 results: - metrics: - type: mean_reward value: 18.30 +/- 7.93 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 --- # Reinforce Agent playing CartPole-v1 This is a trained model of a Reinforce agent playing CartPole-v1 . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5
fabf21/finetuning-sentiment-model-3000-samples	fabf21	2022-07-31T11:16:46Z	5	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "dataset:imdb", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text-classification	2022-07-31T11:05:03Z	--- license: apache-2.0 tags: - generated_from_trainer datasets: - imdb model-index: - name: finetuning-sentiment-model-3000-samples results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # finetuning-sentiment-model-3000-samples This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the imdb dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
Neha2608/xlm-roberta-base-finetuned-panx-en	Neha2608	2022-07-31T10:42:20Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "xlm-roberta", "generated_from_trainer", "dataset:xtreme", "license:mit", "endpoints_compatible", "region:us" ]	null	2022-07-02T12:17:17Z	--- license: mit tags: - generated_from_trainer datasets: - xtreme metrics: - f1 model-index: - name: xlm-roberta-base-finetuned-panx-en results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # xlm-roberta-base-finetuned-panx-en This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset. It achieves the following results on the evaluation set: - Loss: 0.4329 - F1: 0.6431 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 24 - eval_batch_size: 24 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| 1.1554 \| 1.0 \| 50 \| 0.5989 \| 0.4571 \| \| 0.5361 \| 2.0 \| 100 \| 0.4329 \| 0.6431 \| ### Framework versions - Transformers 4.11.3 - Pytorch 1.12.0+cu113 - Datasets 1.16.1 - Tokenizers 0.10.3
Okyx/finetuned-amazon-en-es	Okyx	2022-07-31T10:33:05Z	10	0	transformers	[ "transformers", "tf", "mt5", "text2text-generation", "generated_from_keras_callback", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]	text2text-generation	2022-07-31T09:41:05Z	--- license: apache-2.0 tags: - generated_from_keras_callback model-index: - name: Okyx/finetuned-amazon-en-es results: [] --- <!-- This model card has been generated automatically according to the information Keras had access to. You should probably proofread and complete it, then remove this comment. --> # Okyx/finetuned-amazon-en-es This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on an unknown dataset. It achieves the following results on the evaluation set: - Train Loss: 4.0154 - Validation Loss: 3.3292 - Epoch: 7 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5.6e-05, 'decay_steps': 9672, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01} - training_precision: float32 ### Training results \| Train Loss \| Validation Loss \| Epoch \| \|:----------:\|:---------------:\|:-----:\| \| 9.2009 \| 4.0465 \| 0 \| \| 5.7436 \| 3.6640 \| 1 \| \| 5.0419 \| 3.5296 \| 2 \| \| 4.6412 \| 3.4582 \| 3 \| \| 4.3722 \| 3.3943 \| 4 \| \| 4.1947 \| 3.3610 \| 5 \| \| 4.0747 \| 3.3295 \| 6 \| \| 4.0154 \| 3.3292 \| 7 \| ### Framework versions - Transformers 4.21.0 - TensorFlow 2.8.2 - Datasets 2.4.0 - Tokenizers 0.12.1
Neha2608/xlm-roberta-base-finetuned-panx-it	Neha2608	2022-07-31T10:26:20Z	7	0	transformers	[ "transformers", "pytorch", "xlm-roberta", "generated_from_trainer", "dataset:xtreme", "license:mit", "endpoints_compatible", "region:us" ]	null	2022-07-02T11:59:49Z	--- license: mit tags: - generated_from_trainer datasets: - xtreme metrics: - f1 model-index: - name: xlm-roberta-base-finetuned-panx-it results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # xlm-roberta-base-finetuned-panx-it This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset. It achieves the following results on the evaluation set: - Loss: 0.2740 - F1: 0.7919 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 24 - eval_batch_size: 24 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| F1 \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:------:\| \| 0.8185 \| 1.0 \| 70 \| 0.3369 \| 0.7449 \| \| 0.2899 \| 2.0 \| 140 \| 0.2740 \| 0.7919 \| ### Framework versions - Transformers 4.11.3 - Pytorch 1.12.0+cu113 - Datasets 1.16.1 - Tokenizers 0.10.3
CuteBlack/gfp_guided_diffusion_v4	CuteBlack	2022-07-31T10:04:18Z	0	9	null	[ "license:mit", "region:us" ]	null	2022-07-31T09:48:47Z	--- license: mit --- Open AI diffusion model that has trained on every single NSFW gay furry illustrations on e621.net that’s over the community score of 100. Excluding extreme fetishes and underage contents. 'attention_resolutions': '32, 16, 8', 'class_cond': False, 'diffusion_steps': 1000, 'rescale_timesteps': True, 'image_size': 256, 'learn_sigma': True, 'noise_schedule': 'linear', 'num_channels': 128, 'num_heads': 4, 'num_res_blocks': 2, 'resblock_updown': True, 'use_checkpoint': use_checkpoint, 'use_fp16': True, 'use_scale_shift_norm': True
ijnekonasa/ppo-LunarLander-v2	ijnekonasa	2022-07-31T03:58:07Z	0	0	stable-baselines3	[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2022-07-31T03:57:44Z	--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - metrics: - type: mean_reward value: 252.64 +/- 18.29 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 --- # PPO Agent playing LunarLander-v2 This is a trained model of a PPO agent playing LunarLander-v2 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```
Frikallo/DeepLeffen-TSM_Leffen	Frikallo	2022-07-31T01:31:44Z	3	0	transformers	[ "transformers", "pytorch", "tensorboard", "gpt2", "text-generation", "generated_from_trainer", "license:mit", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2022-07-31T01:27:45Z	--- license: mit tags: - generated_from_trainer model-index: - name: DeepLeffen-TSM_Leffen results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # DeepLeffen-TSM_Leffen This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001372 - train_batch_size: 1 - eval_batch_size: 8 - seed: 2780791035 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 1.0 ### Training results ### Framework versions - Transformers 4.20.1 - Pytorch 1.9.1+cu111 - Datasets 2.3.2 - Tokenizers 0.12.1
keithanpai/vit-base-patch16-224-finetuned-eurosat	keithanpai	2022-07-31T00:07:31Z	55	0	transformers	[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "generated_from_trainer", "dataset:imagefolder", "license:apache-2.0", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	image-classification	2022-07-30T23:42:57Z	--- license: apache-2.0 tags: - generated_from_trainer datasets: - imagefolder metrics: - accuracy model-index: - name: vit-base-patch16-224-finetuned-eurosat results: - task: name: Image Classification type: image-classification dataset: name: imagefolder type: imagefolder config: default split: train args: default metrics: - name: Accuracy type: accuracy value: 0.8632734530938124 --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # vit-base-patch16-224-finetuned-eurosat This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the imagefolder dataset. It achieves the following results on the evaluation set: - Loss: 0.3953 - Accuracy: 0.8633 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 128 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 3 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Accuracy \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:--------:\| \| 0.6081 \| 0.99 \| 70 \| 0.5482 \| 0.8004 \| \| 0.4515 \| 1.99 \| 140 \| 0.4245 \| 0.8533 \| \| 0.3967 \| 2.99 \| 210 \| 0.3953 \| 0.8633 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1
sophiestein/experiment_2	sophiestein	2022-07-30T17:57:00Z	6	0	transformers	[ "transformers", "pytorch", "tensorboard", "distilbert", "token-classification", "generated_from_trainer", "dataset:conll2003", "license:apache-2.0", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	token-classification	2022-07-30T10:21:23Z	--- license: apache-2.0 tags: - generated_from_trainer datasets: - conll2003 metrics: - precision - recall - f1 - accuracy model-index: - name: experiment_2 results: - task: name: Token Classification type: token-classification dataset: name: conll2003 type: conll2003 config: conll2003 split: train args: conll2003 metrics: - name: Precision type: precision value: 0.8840954508052192 - name: Recall type: recall value: 0.8925943508188939 - name: F1 type: f1 value: 0.8883245733183724 - name: Accuracy type: accuracy value: 0.9746737103791174 --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # experiment_2 This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the conll2003 dataset. It achieves the following results on the evaluation set: - Loss: 0.1211 - Precision: 0.8841 - Recall: 0.8926 - F1: 0.8883 - Accuracy: 0.9747 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| Precision \| Recall \| F1 \| Accuracy \| \|:-------------:\|:-----:\|:----:\|:---------------:\|:---------:\|:------:\|:------:\|:--------:\| \| 0.2418 \| 1.0 \| 878 \| 0.0695 \| 0.9159 \| 0.9255 \| 0.9207 \| 0.9816 \| \| 0.0541 \| 2.0 \| 1756 \| 0.0592 \| 0.9244 \| 0.9343 \| 0.9293 \| 0.9833 \| \| 0.0303 \| 3.0 \| 2634 \| 0.0602 \| 0.9260 \| 0.9388 \| 0.9323 \| 0.9838 \| ### Framework versions - Transformers 4.21.0 - Pytorch 1.11.0+cpu - Datasets 2.4.0 - Tokenizers 0.12.1
anzorq/kbd_lat-ru_char_tokenizer	anzorq	2022-07-30T16:16:55Z	10	0	transformers	[ "transformers", "pytorch", "jax", "t5", "text2text-generation", "translation", "ru", "kbd", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	translation	2022-07-29T10:31:32Z	--- language: - ru - kbd tags: - translation ---
comodoro/testpyramidsrnd2	comodoro	2022-07-30T15:58:53Z	4	0	ml-agents	[ "ml-agents", "tensorboard", "onnx", "unity-ml-agents", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-Pyramids", "region:us" ]	reinforcement-learning	2022-07-30T15:58:47Z	--- tags: - unity-ml-agents - ml-agents - deep-reinforcement-learning - reinforcement-learning - ML-Agents-Pyramids library_name: ml-agents --- # ppo Agent playing Pyramids This is a trained model of a ppo agent playing Pyramids using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://github.com/huggingface/ml-agents#get-started We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: ### Resume the training ``` mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent playing directly in your browser:. 1. Go to https://huggingface.co/spaces/unity/ML-Agents-Pyramids 2. Step 1: Write your model_id: comodoro/testpyramidsrnd2 3. Step 2: Select your .nn /.onnx file 4. Click on Watch the agent play 👀
constanter/PPO-LunarLander-v2	constanter	2022-07-30T13:34:25Z	0	0	stable-baselines3	[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2022-07-30T13:33:54Z	--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - metrics: - type: mean_reward value: 268.37 +/- 20.32 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 --- # PPO Agent playing LunarLander-v2 This is a trained model of a PPO agent playing LunarLander-v2 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```
SummerChiam/rust_image_classification_7	SummerChiam	2022-07-30T12:04:23Z	57	0	transformers	[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	image-classification	2022-07-30T12:04:11Z	--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: rust_image_classification_7 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9645569324493408 --- # rust_image_classification_7 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for anything by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### nonrust ![nonrust](images/nonrust.png) #### rust ![rust](images/rust.png)
devetle/a2c-AntBulletEnv-v0	devetle	2022-07-30T10:14:08Z	2	0	stable-baselines3	[ "stable-baselines3", "AntBulletEnv-v0", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2022-07-30T10:13:05Z	--- library_name: stable-baselines3 tags: - AntBulletEnv-v0 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: A2C results: - metrics: - type: mean_reward value: 1098.81 +/- 321.12 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: AntBulletEnv-v0 type: AntBulletEnv-v0 --- # A2C Agent playing AntBulletEnv-v0 This is a trained model of a A2C agent playing AntBulletEnv-v0 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```
SummerChiam/pond_image_classification_10	SummerChiam	2022-07-30T08:57:50Z	50	0	transformers	[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]	image-classification	2022-07-30T08:57:38Z	--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: pond_image_classification_10 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9948979616165161 --- # pond_image_classification_10 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for anything by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### Algae ![Algae](images/Algae.png) #### Boiling ![Boiling](images/Boiling.png) #### BoilingNight ![BoilingNight](images/BoilingNight.png) #### Normal ![Normal](images/Normal.png) #### NormalCement ![NormalCement](images/NormalCement.png) #### NormalNight ![NormalNight](images/NormalNight.png) #### NormalRain ![NormalRain](images/NormalRain.png)
mbarnig/lb-de-fr-en-pt-coqui-vits-tts	mbarnig	2022-07-30T06:00:58Z	222	7	transformers	[ "transformers", "tensorboard", "TTS", "audio", "synthesis", "yourTTS", "speech", "coqui.ai", "lb", "de", "fr", "en", "pt", "dataset:mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS", "license:cc-by-nc-sa-4.0", "endpoints_compatible", "region:us" ]	null	2022-07-08T20:42:32Z	--- license: cc-by-nc-sa-4.0 language: - lb - de - fr - en - pt tags: - TTS - audio - synthesis - yourTTS - speech - coqui.ai datasets: - mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS --- #### This model has been trained from scratch with my customized dataset [mbarnig/lb-de-fr-en-pt-12800-TTS_CORPUS](https://huggingface.co/datasets/mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS) and the 🐸 [Coqui-TTS multilingual VITS-model recipe](https://github.com/coqui-ai/TTS/tree/dev/recipes/multilingual/vits_tts) (version 0.7.1). The model was trained without phonemes with the following character-set: ``` characters="abcdefghijklmnopqrstuvwxyz ßàáâãäçèéêëíîïóôõöùúûü", punctuations="!'(),-.:;? ", phonemes=None, ``` #### A live inference-demo of the model is available in my HuggingFace space ⌨️ 🇱🇺 🔈 [mbarnig/lb_de_fr_en_pt_COQUI_VITS_TTS](https://huggingface.co/spaces/mbarnig/lb_de_fr_en_pt_COQUI_VITS_TTS). #### Click the tab training metrics above to view the live Tensorboard of the model training. ![tensorboard](tensorboard.png)
vinitharaj/distilbert-base-uncased-finetuned-squad2	vinitharaj	2022-07-30T05:47:35Z	5	0	transformers	[ "transformers", "tf", "tensorboard", "distilbert", "question-answering", "generated_from_keras_callback", "license:apache-2.0", "endpoints_compatible", "region:us" ]	question-answering	2022-07-29T07:47:14Z	--- license: apache-2.0 tags: - generated_from_keras_callback model-index: - name: vinitharaj/distilbert-base-uncased-finetuned-squad2 results: [] --- <!-- This model card has been generated automatically according to the information Keras had access to. You should probably proofread and complete it, then remove this comment. --> # vinitharaj/distilbert-base-uncased-finetuned-squad2 This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset. It achieves the following results on the evaluation set: - Train Loss: 0.4953 - Validation Loss: 0.3885 - Epoch: 1 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - optimizer: {'name': 'Adam', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 1602, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False} - training_precision: float32 ### Training results \| Train Loss \| Validation Loss \| Epoch \| \|:----------:\|:---------------:\|:-----:\| \| 0.7037 \| 0.4222 \| 0 \| \| 0.4953 \| 0.3885 \| 1 \| ### Framework versions - Transformers 4.21.0 - TensorFlow 2.8.2 - Datasets 2.4.0 - Tokenizers 0.12.1
Migga/ViT-BERT-Chess-V4	Migga	2022-07-30T04:26:03Z	1	0	transformers	[ "transformers", "pytorch", "vision-encoder-decoder", "image-text-to-text", "generated_from_trainer", "endpoints_compatible", "region:us" ]	image-text-to-text	2022-07-29T16:57:48Z	--- tags: - generated_from_trainer model-index: - name: ViT-BERT-Chess-V4 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # ViT-BERT-Chess-V4 This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.3213 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 4 - seed: 42 - gradient_accumulation_steps: 8 - total_train_batch_size: 64 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 5 - mixed_precision_training: Native AMP ### Training results \| Training Loss \| Epoch \| Step \| Validation Loss \| \|:-------------:\|:-----:\|:-----:\|:---------------:\| \| 3.705 \| 1.0 \| 3895 \| 3.5686 \| \| 3.5139 \| 2.0 \| 7790 \| 3.4288 \| \| 3.4156 \| 3.0 \| 11685 \| 3.3663 \| \| 3.3661 \| 4.0 \| 15580 \| 3.3331 \| \| 3.3352 \| 5.0 \| 19475 \| 3.3213 \| ### Framework versions - Transformers 4.20.1 - Pytorch 1.12.0+cu116 - Datasets 2.3.2 - Tokenizers 0.12.1
reachrkr/testpyramidsrnd	reachrkr	2022-07-30T02:46:18Z	5	0	ml-agents	[ "ml-agents", "tensorboard", "onnx", "unity-ml-agents", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-Pyramids", "region:us" ]	reinforcement-learning	2022-07-28T06:59:12Z	--- tags: - unity-ml-agents - ml-agents - deep-reinforcement-learning - reinforcement-learning - ML-Agents-Pyramids library_name: ml-agents --- # ppo Agent playing Pyramids This is a trained model of a ppo agent playing Pyramids using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://github.com/huggingface/ml-agents#get-started We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: ### Resume the training ``` mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent playing directly in your browser:. 1. Go to https://huggingface.co/spaces/unity/ML-Agents-Pyramids 2. Step 1: Write your model_id: reachrkr/testpyramidsrnd 3. Step 2: Select your .nn /.onnx file 4. Click on Watch the agent play 👀
huggingtweets/dags	huggingtweets	2022-07-30T01:32:18Z	4	0	transformers	[ "transformers", "pytorch", "gpt2", "text-generation", "huggingtweets", "en", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]	text-generation	2022-07-30T01:30:26Z	--- language: en thumbnail: http://www.huggingtweets.com/dags/1659144733206/predictions.png tags: - huggingtweets widget: - text: "My dream is" --- <div class="inline-flex flex-col" style="line-height: 1.5;"> <div class="flex"> <div style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/722815128501026817/IMWCRzEn_400x400.jpg')"> </div> <div style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')"> </div> <div style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')"> </div> </div> <div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div> <div style="text-align: center; font-size: 16px; font-weight: 800">DAGs</div> <div style="text-align: center; font-size: 14px;">@dags</div> </div> I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets). Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)! ## How does it work? The model uses the following pipeline. ![pipeline](https://github.com/borisdayma/huggingtweets/blob/master/img/pipeline.png?raw=true) To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI). ## Training data The model was trained on tweets from DAGs. \| Data \| DAGs \| \| --- \| --- \| \| Tweets downloaded \| 3003 \| \| Retweets \| 31 \| \| Short tweets \| 158 \| \| Tweets kept \| 2814 \| [Explore the data](https://wandb.ai/wandb/huggingtweets/runs/3qyk6uzo/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline. ## Training procedure The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @dags's tweets. Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/18qzuqjb) for full transparency and reproducibility. At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/18qzuqjb/artifacts) is logged and versioned. ## How to use You can use this model directly with a pipeline for text generation: ```python from transformers import pipeline generator = pipeline('text-generation', model='huggingtweets/dags') generator("My dream is", num_return_sequences=5) ``` ## Limitations and bias The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias). In addition, the data present in the user's tweets further affects the text generated by the model. ## About Built by Boris Dayma [![Follow](https://img.shields.io/twitter/follow/borisdayma?style=social)](https://twitter.com/intent/follow?screen_name=borisdayma) For more details, visit the project repository. [![GitHub stars](https://img.shields.io/github/stars/borisdayma/huggingtweets?style=social)](https://github.com/borisdayma/huggingtweets)
rebolforces/ppo-LunarLander-v2	rebolforces	2022-07-30T00:43:21Z	1	0	stable-baselines3	[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]	reinforcement-learning	2022-07-23T09:28:37Z	--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - metrics: - type: mean_reward value: 285.83 +/- 15.59 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 --- # PPO Agent playing LunarLander-v2 This is a trained model of a PPO agent playing LunarLander-v2 using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```
yanaiela/roberta-base-epoch_80	yanaiela	2022-07-29T23:08:59Z	7	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_80", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T18:03:25Z	--- language: en tags: - roberta-base - roberta-base-epoch_80 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 80 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_80. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_78	yanaiela	2022-07-29T23:08:15Z	12	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_78", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T18:01:03Z	--- language: en tags: - roberta-base - roberta-base-epoch_78 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 78 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_78. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_77	yanaiela	2022-07-29T23:07:53Z	10	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_77", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:59:57Z	--- language: en tags: - roberta-base - roberta-base-epoch_77 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 77 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_77. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_73	yanaiela	2022-07-29T23:06:21Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_73", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:55:51Z	--- language: en tags: - roberta-base - roberta-base-epoch_73 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 73 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_73. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_71	yanaiela	2022-07-29T23:05:36Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_71", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:53:19Z	--- language: en tags: - roberta-base - roberta-base-epoch_71 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 71 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_71. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_67	yanaiela	2022-07-29T23:04:02Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_67", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:48:39Z	--- language: en tags: - roberta-base - roberta-base-epoch_67 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 67 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_67. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_66	yanaiela	2022-07-29T23:03:37Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_66", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:46:45Z	--- language: en tags: - roberta-base - roberta-base-epoch_66 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 66 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_66. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_64	yanaiela	2022-07-29T23:02:45Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_64", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:43:33Z	--- language: en tags: - roberta-base - roberta-base-epoch_64 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 64 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_64. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_61	yanaiela	2022-07-29T23:01:44Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_61", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:39:32Z	--- language: en tags: - roberta-base - roberta-base-epoch_61 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 61 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_61. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_59	yanaiela	2022-07-29T23:01:00Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_59", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:35:53Z	--- language: en tags: - roberta-base - roberta-base-epoch_59 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 59 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_59. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_57	yanaiela	2022-07-29T23:00:18Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_57", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:34:22Z	--- language: en tags: - roberta-base - roberta-base-epoch_57 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 57 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_57. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_56	yanaiela	2022-07-29T22:59:56Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_56", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:33:29Z	--- language: en tags: - roberta-base - roberta-base-epoch_56 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 56 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_56. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_54	yanaiela	2022-07-29T22:59:09Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_54", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:31:39Z	--- language: en tags: - roberta-base - roberta-base-epoch_54 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 54 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_54. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_52	yanaiela	2022-07-29T22:58:22Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_52", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:30:02Z	--- language: en tags: - roberta-base - roberta-base-epoch_52 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 52 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_52. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_50	yanaiela	2022-07-29T22:57:31Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_50", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:28:26Z	--- language: en tags: - roberta-base - roberta-base-epoch_50 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 50 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_50. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_47	yanaiela	2022-07-29T22:56:19Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_47", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:26:12Z	--- language: en tags: - roberta-base - roberta-base-epoch_47 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 47 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_47. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_46	yanaiela	2022-07-29T22:55:54Z	3	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_46", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:25:28Z	--- language: en tags: - roberta-base - roberta-base-epoch_46 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 46 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_46. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_45	yanaiela	2022-07-29T22:55:32Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_45", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:24:44Z	--- language: en tags: - roberta-base - roberta-base-epoch_45 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 45 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_45. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_43	yanaiela	2022-07-29T22:54:43Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_43", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:23:18Z	--- language: en tags: - roberta-base - roberta-base-epoch_43 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 43 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_43. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_39	yanaiela	2022-07-29T22:53:02Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_39", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:20:23Z	--- language: en tags: - roberta-base - roberta-base-epoch_39 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 39 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_39. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_31	yanaiela	2022-07-29T22:50:29Z	6	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_31", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:14:05Z	--- language: en tags: - roberta-base - roberta-base-epoch_31 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 31 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_31. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_27	yanaiela	2022-07-29T22:49:16Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_27", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:10:38Z	--- language: en tags: - roberta-base - roberta-base-epoch_27 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 27 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_27. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_26	yanaiela	2022-07-29T22:48:56Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_26", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:09:55Z	--- language: en tags: - roberta-base - roberta-base-epoch_26 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 26 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_26. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_21	yanaiela	2022-07-29T22:47:23Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_21", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:06:01Z	--- language: en tags: - roberta-base - roberta-base-epoch_21 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 21 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_21. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_20	yanaiela	2022-07-29T22:47:06Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_20", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:05:11Z	--- language: en tags: - roberta-base - roberta-base-epoch_20 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 20 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_20. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_17	yanaiela	2022-07-29T22:46:08Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_17", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:02:47Z	--- language: en tags: - roberta-base - roberta-base-epoch_17 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 17 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_17. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_15	yanaiela	2022-07-29T22:45:30Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_15", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:01:23Z	--- language: en tags: - roberta-base - roberta-base-epoch_15 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 15 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_15. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_14	yanaiela	2022-07-29T22:45:11Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_14", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T17:00:38Z	--- language: en tags: - roberta-base - roberta-base-epoch_14 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 14 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_14. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_9	yanaiela	2022-07-29T22:43:40Z	4	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_9", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T16:56:14Z	--- language: en tags: - roberta-base - roberta-base-epoch_9 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 9 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_9. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_7	yanaiela	2022-07-29T22:43:03Z	7	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_7", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T16:54:40Z	--- language: en tags: - roberta-base - roberta-base-epoch_7 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 7 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_7. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```
yanaiela/roberta-base-epoch_3	yanaiela	2022-07-29T22:41:43Z	5	0	transformers	[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_3", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]	fill-mask	2022-07-28T16:51:25Z	--- language: en tags: - roberta-base - roberta-base-epoch_3 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 3 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_3. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

KDHyun08/TAACO_STS

KDHyun08

2022-08-01T05:00:14Z

2,406

2

sentence-transformers

[ "sentence-transformers", "pytorch", "bert", "feature-extraction", "sentence-similarity", "transformers", "TAACO", "ko", "autotrain_compatible", "text-embeddings-inference", "endpoints_compatible", "region:us" ]

sentence-similarity

2022-07-25T08:19:31Z

--- pipeline_tag: sentence-similarity tags: - sentence-transformers - sentence-similarity - transformers - TAACO language: ko --- # TAACO_Similarity 본 모델은 [Sentence-transformers](https://www.SBERT.net)를 기반으로 하며 KLUE의 STS(Sentence Textual Similarity) 데이터셋을 통해 훈련을 진행한 모델입니다. 필자가 제작하고 있는 한국어 문장간 결속성 측정 도구인 K-TAACO(가제)의 지표 중 하나인 문장 간 의미적 결속성을 측정하기 위해 본 모델을 제작하였습니다. 또한 모두의 말뭉치의 문장간 유사도 데이터 등 다양한 데이터를 구해 추가 훈련을 진행할 예정입니다. ## Train Data KLUE-sts-v1.1._train.json NLI-sts-train.tsv ## Usage (Sentence-Transformers) 본 모델을 사용하기 위해서는 [Sentence-transformers](https://www.SBERT.net)를 설치하여야 합니다. ``` pip install -U sentence-transformers ``` 모델을 사용하기 위해서는 아래 코드를 참조하시길 바랍니다. ```python from sentence_transformers import SentenceTransformer, models sentences = ["This is an example sentence", "Each sentence is converted"] embedding_model = models.Transformer( model_name_or_path="KDHyun08/TAACO_STS", max_seq_length=256, do_lower_case=True ) pooling_model = models.Pooling( embedding_model.get_word_embedding_dimension(), pooling_mode_mean_tokens=True, pooling_mode_cls_token=False, pooling_mode_max_tokens=False, ) model = SentenceTransformer(modules=[embedding_model, pooling_model]) embeddings = model.encode(sentences) print(embeddings) ``` ## Usage (실제 문장 간 유사도 비교) [Sentence-transformers](https://www.SBERT.net) 를 설치한 후 아래 내용과 같이 문장 간 유사도를 비교할 수 있습니다. query 변수는 비교 기준이 되는 문장(Source Sentence)이고 비교를 진행할 문장은 docs에 list 형식으로 구성하시면 됩니다. ```python from sentence_transformers import SentenceTransformer, models embedding_model = models.Transformer( model_name_or_path="KDHyun08/TAACO_STS", max_seq_length=256, do_lower_case=True ) pooling_model = models.Pooling( embedding_model.get_word_embedding_dimension(), pooling_mode_mean_tokens=True, pooling_mode_cls_token=False, pooling_mode_max_tokens=False, ) model = SentenceTransformer(modules=[embedding_model, pooling_model]) docs = ['어제는 아내의 생일이었다', '생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다. 주된 메뉴는 스테이크와 낙지볶음, 미역국, 잡채, 소야 등이었다', '스테이크는 자주 하는 음식이어서 자신이 준비하려고 했다', '앞뒤도 1분씩 3번 뒤집고 래스팅을 잘 하면 육즙이 가득한 스테이크가 준비되다', '아내도 그런 스테이크를 좋아한다. 그런데 상상도 못한 일이 벌이지고 말았다', '보통 시즈닝이 되지 않은 원육을 사서 스테이크를 했는데, 이번에는 시즈닝이 된 부챗살을 구입해서 했다', '그런데 케이스 안에 방부제가 들어있는 것을 인지하지 못하고 방부제와 동시에 프라이팬에 올려놓을 것이다', '그것도 인지 못한 체... 앞면을 센 불에 1분을 굽고 뒤집는 순간 방부제가 함께 구어진 것을 알았다', '아내의 생일이라 맛있게 구워보고 싶었는데 어처구니없는 상황이 발생한 것이다', '방부제가 센 불에 녹아서 그런지 물처럼 흘러내렸다', ' 고민을 했다. 방부제가 묻은 부문만 제거하고 다시 구울까 했는데 방부제에 절대 먹지 말라는 문구가 있어서 아깝지만 버리는 방향을 했다', '너무나 안타까웠다', '아침 일찍 아내가 좋아하는 스테이크를 준비하고 그것을 맛있게 먹는 아내의 모습을 보고 싶었는데 전혀 생각지도 못한 상황이 발생해서... 하지만 정신을 추스르고 바로 다른 메뉴로 변경했다', '소야, 소시지 야채볶음..', '아내가 좋아하는지 모르겠지만 냉장고 안에 있는 후랑크소세지를 보니 바로 소야를 해야겠다는 생각이 들었다. 음식은 성공적으로 완성이 되었다', '40번째를 맞이하는 아내의 생일은 성공적으로 준비가 되었다', '맛있게 먹어 준 아내에게도 감사했다', '매년 아내의 생일에 맞이하면 아침마다 생일을 차려야겠다. 오늘도 즐거운 하루가 되었으면 좋겠다', '생일이니까~'] #각 문장의 vector값 encoding document_embeddings = model.encode(docs) query = '생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다' query_embedding = model.encode(query) top_k = min(10, len(docs)) # 코사인 유사도 계산 후, cos_scores = util.pytorch_cos_sim(query_embedding, document_embeddings)[0] # 코사인 유사도 순으로 문장 추출 top_results = torch.topk(cos_scores, k=top_k) print(f"입력 문장: {query}") print(f"\n<입력 문장과 유사한 {top_k} 개의 문장>\n") for i, (score, idx) in enumerate(zip(top_results[0], top_results[1])): print(f"{i+1}: {docs[idx]} {'(유사도: {:.4f})'.format(score)}\n") ``` ## Evaluation Results 위 Usage를 실행하게 되면 아래와 같은 결과가 도출됩니다. 1에 가까울수록 유사한 문장입니다. ``` 입력 문장: 생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다 <입력 문장과 유사한 10 개의 문장> 1: 생일을 맞이하여 아침을 준비하겠다고 오전 8시 30분부터 음식을 준비하였다. 주된 메뉴는 스테이크와 낙지볶음, 미역국, 잡채, 소야 등이었다 (유사도: 0.6687) 2: 매년 아내의 생일에 맞이하면 아침마다 생일을 차려야겠다. 오늘도 즐거운 하루가 되었으면 좋겠다 (유사도: 0.6468) 3: 40번째를 맞이하는 아내의 생일은 성공적으로 준비가 되었다 (유사도: 0.4647) 4: 아내의 생일이라 맛있게 구워보고 싶었는데 어처구니없는 상황이 발생한 것이다 (유사도: 0.4469) 5: 생일이니까~ (유사도: 0.4218) 6: 어제는 아내의 생일이었다 (유사도: 0.4192) 7: 아침 일찍 아내가 좋아하는 스테이크를 준비하고 그것을 맛있게 먹는 아내의 모습을 보고 싶었는데 전혀 생각지도 못한 상황이 발생해서... 하지만 정신을 추스르고 바로 다른 메뉴로 변경했다 (유사도: 0.4156) 8: 맛있게 먹어 준 아내에게도 감사했다 (유사도: 0.3093) 9: 아내가 좋아하는지 모르겠지만 냉장고 안에 있는 후랑크소세지를 보니 바로 소야를 해야겠다는 생각이 들었다. 음식은 성공적으로 완성이 되었다 (유사도: 0.2259) 10: 아내도 그런 스테이크를 좋아한다. 그런데 상상도 못한 일이 벌이지고 말았다 (유사도: 0.1967) ``` **DataLoader**: `torch.utils.data.dataloader.DataLoader` of length 142 with parameters: ``` {'batch_size': 32, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'} ``` **Loss**: `sentence_transformers.losses.CosineSimilarityLoss.CosineSimilarityLoss` Parameters of the fit()-Method: ``` { "epochs": 4, "evaluation_steps": 1000, "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator", "max_grad_norm": 1, "optimizer_class": "<class 'transformers.optimization.AdamW'>", "optimizer_params": { "lr": 2e-05 }, "scheduler": "WarmupLinear", "steps_per_epoch": null, "warmup_steps": 10000, "weight_decay": 0.01 } ``` ## Full Model Architecture ``` SentenceTransformer( (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False}) ) ``` ## Citing & Authors

wenkai-li/distilroberta-base-finetuned-marktextepoch_n200

wenkai-li

2022-08-01T04:07:13Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "roberta", "fill-mask", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-31T18:33:57Z

--- license: apache-2.0 tags: - generated_from_trainer model-index: - name: distilroberta-base-finetuned-marktextepoch_n200 results: [] ---  # distilroberta-base-finetuned-marktextepoch_n200 This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the None dataset. It achieves the following results on the evaluation set: - Loss: 2.0531 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 200 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:------:|:---------------:| | 2.2313 | 1.0 | 1500 | 2.1592 | | 2.1731 | 2.0 | 3000 | 2.1277 | | 2.153 | 3.0 | 4500 | 2.1144 | | 2.1469 | 4.0 | 6000 | 2.1141 | | 2.1281 | 5.0 | 7500 | 2.1374 | | 2.1043 | 6.0 | 9000 | 2.1069 | | 2.0834 | 7.0 | 10500 | 2.0993 | | 2.0602 | 8.0 | 12000 | 2.0817 | | 2.024 | 9.0 | 13500 | 2.0918 | | 2.0261 | 10.0 | 15000 | 2.0793 | | 1.9889 | 11.0 | 16500 | 2.0567 | | 1.9915 | 12.0 | 18000 | 2.0700 | | 1.9532 | 13.0 | 19500 | 2.0436 | | 1.9362 | 14.0 | 21000 | 2.0596 | | 1.9024 | 15.0 | 22500 | 2.0189 | | 1.9262 | 16.0 | 24000 | 2.0435 | | 1.8883 | 17.0 | 25500 | 2.0430 | | 1.8867 | 18.0 | 27000 | 2.0416 | | 1.8807 | 19.0 | 28500 | 2.0051 | | 1.8517 | 20.0 | 30000 | 2.0338 | | 1.8357 | 21.0 | 31500 | 2.0166 | | 1.8241 | 22.0 | 33000 | 2.0355 | | 1.7985 | 23.0 | 34500 | 2.0073 | | 1.8061 | 24.0 | 36000 | 2.0473 | | 1.7996 | 25.0 | 37500 | 2.0446 | | 1.7786 | 26.0 | 39000 | 2.0086 | | 1.771 | 27.0 | 40500 | 2.0294 | | 1.7549 | 28.0 | 42000 | 2.0127 | | 1.7726 | 29.0 | 43500 | 2.0191 | | 1.7275 | 30.0 | 45000 | 2.0182 | | 1.708 | 31.0 | 46500 | 2.0130 | | 1.7345 | 32.0 | 48000 | 2.0155 | | 1.7044 | 33.0 | 49500 | 1.9898 | | 1.7126 | 34.0 | 51000 | 2.0166 | | 1.698 | 35.0 | 52500 | 1.9879 | | 1.6637 | 36.0 | 54000 | 2.0311 | | 1.6854 | 37.0 | 55500 | 2.0355 | | 1.6585 | 38.0 | 57000 | 2.0094 | | 1.6418 | 39.0 | 58500 | 2.0042 | | 1.667 | 40.0 | 60000 | 2.0116 | | 1.6507 | 41.0 | 61500 | 2.0095 | | 1.622 | 42.0 | 63000 | 2.0158 | | 1.6381 | 43.0 | 64500 | 2.0339 | | 1.6099 | 44.0 | 66000 | 2.0082 | | 1.6076 | 45.0 | 67500 | 2.0207 | | 1.5805 | 46.0 | 69000 | 2.0172 | | 1.5862 | 47.0 | 70500 | 2.0132 | | 1.5806 | 48.0 | 72000 | 2.0198 | | 1.574 | 49.0 | 73500 | 2.0181 | | 1.5718 | 50.0 | 75000 | 2.0086 | | 1.5591 | 51.0 | 76500 | 1.9832 | | 1.5468 | 52.0 | 78000 | 2.0167 | | 1.5637 | 53.0 | 79500 | 2.0118 | | 1.5117 | 54.0 | 81000 | 2.0290 | | 1.5363 | 55.0 | 82500 | 2.0011 | | 1.4976 | 56.0 | 84000 | 2.0160 | | 1.5129 | 57.0 | 85500 | 2.0224 | | 1.4964 | 58.0 | 87000 | 2.0219 | | 1.4906 | 59.0 | 88500 | 2.0212 | | 1.4941 | 60.0 | 90000 | 2.0255 | | 1.4876 | 61.0 | 91500 | 2.0116 | | 1.4837 | 62.0 | 93000 | 2.0176 | | 1.4661 | 63.0 | 94500 | 2.0388 | | 1.4634 | 64.0 | 96000 | 2.0165 | | 1.4449 | 65.0 | 97500 | 2.0185 | | 1.468 | 66.0 | 99000 | 2.0246 | | 1.4567 | 67.0 | 100500 | 2.0244 | | 1.4367 | 68.0 | 102000 | 2.0093 | | 1.4471 | 69.0 | 103500 | 2.0101 | | 1.4255 | 70.0 | 105000 | 2.0248 | | 1.4203 | 71.0 | 106500 | 2.0224 | | 1.42 | 72.0 | 108000 | 2.0279 | | 1.4239 | 73.0 | 109500 | 2.0295 | | 1.4126 | 74.0 | 111000 | 2.0196 | | 1.4038 | 75.0 | 112500 | 2.0225 | | 1.3874 | 76.0 | 114000 | 2.0456 | | 1.3758 | 77.0 | 115500 | 2.0423 | | 1.3924 | 78.0 | 117000 | 2.0184 | | 1.3744 | 79.0 | 118500 | 2.0555 | | 1.3622 | 80.0 | 120000 | 2.0387 | | 1.3653 | 81.0 | 121500 | 2.0344 | | 1.3724 | 82.0 | 123000 | 2.0184 | | 1.3684 | 83.0 | 124500 | 2.0285 | | 1.3576 | 84.0 | 126000 | 2.0544 | | 1.348 | 85.0 | 127500 | 2.0412 | | 1.3387 | 86.0 | 129000 | 2.0459 | | 1.3416 | 87.0 | 130500 | 2.0329 | | 1.3421 | 88.0 | 132000 | 2.0274 | | 1.3266 | 89.0 | 133500 | 2.0233 | | 1.3183 | 90.0 | 135000 | 2.0319 | | 1.322 | 91.0 | 136500 | 2.0080 | | 1.32 | 92.0 | 138000 | 2.0472 | | 1.304 | 93.0 | 139500 | 2.0538 | | 1.3061 | 94.0 | 141000 | 2.0340 | | 1.3199 | 95.0 | 142500 | 2.0456 | | 1.2985 | 96.0 | 144000 | 2.0167 | | 1.3021 | 97.0 | 145500 | 2.0204 | | 1.2787 | 98.0 | 147000 | 2.0645 | | 1.2879 | 99.0 | 148500 | 2.0345 | | 1.2695 | 100.0 | 150000 | 2.0340 | | 1.2884 | 101.0 | 151500 | 2.0602 | | 1.2747 | 102.0 | 153000 | 2.0667 | | 1.2607 | 103.0 | 154500 | 2.0551 | | 1.2551 | 104.0 | 156000 | 2.0544 | | 1.2557 | 105.0 | 157500 | 2.0553 | | 1.2495 | 106.0 | 159000 | 2.0370 | | 1.26 | 107.0 | 160500 | 2.0568 | | 1.2499 | 108.0 | 162000 | 2.0427 | | 1.2438 | 109.0 | 163500 | 2.0184 | | 1.2496 | 110.0 | 165000 | 2.0227 | | 1.2332 | 111.0 | 166500 | 2.0621 | | 1.2231 | 112.0 | 168000 | 2.0661 | | 1.211 | 113.0 | 169500 | 2.0673 | | 1.217 | 114.0 | 171000 | 2.0544 | | 1.2206 | 115.0 | 172500 | 2.0542 | | 1.2083 | 116.0 | 174000 | 2.0592 | | 1.2205 | 117.0 | 175500 | 2.0451 | | 1.2065 | 118.0 | 177000 | 2.0402 | | 1.1988 | 119.0 | 178500 | 2.0615 | | 1.218 | 120.0 | 180000 | 2.0374 | | 1.1917 | 121.0 | 181500 | 2.0349 | | 1.1854 | 122.0 | 183000 | 2.0790 | | 1.1819 | 123.0 | 184500 | 2.0766 | | 1.2029 | 124.0 | 186000 | 2.0364 | | 1.1851 | 125.0 | 187500 | 2.0568 | | 1.1734 | 126.0 | 189000 | 2.0445 | | 1.1701 | 127.0 | 190500 | 2.0770 | | 1.1824 | 128.0 | 192000 | 2.0566 | | 1.1604 | 129.0 | 193500 | 2.0542 | | 1.1733 | 130.0 | 195000 | 2.0525 | | 1.1743 | 131.0 | 196500 | 2.0577 | | 1.1692 | 132.0 | 198000 | 2.0723 | | 1.1519 | 133.0 | 199500 | 2.0567 | | 1.1401 | 134.0 | 201000 | 2.0795 | | 1.1692 | 135.0 | 202500 | 2.0625 | | 1.157 | 136.0 | 204000 | 2.0793 | | 1.1495 | 137.0 | 205500 | 2.0782 | | 1.1479 | 138.0 | 207000 | 2.0392 | | 1.1247 | 139.0 | 208500 | 2.0796 | | 1.143 | 140.0 | 210000 | 2.0369 | | 1.1324 | 141.0 | 211500 | 2.0699 | | 1.1341 | 142.0 | 213000 | 2.0694 | | 1.1317 | 143.0 | 214500 | 2.0569 | | 1.1254 | 144.0 | 216000 | 2.0545 | | 1.1156 | 145.0 | 217500 | 2.0708 | | 1.1353 | 146.0 | 219000 | 2.0767 | | 1.1312 | 147.0 | 220500 | 2.0523 | | 1.1224 | 148.0 | 222000 | 2.0565 | | 1.106 | 149.0 | 223500 | 2.0696 | | 1.1069 | 150.0 | 225000 | 2.0478 | | 1.1011 | 151.0 | 226500 | 2.0475 | | 1.0985 | 152.0 | 228000 | 2.0888 | | 1.1107 | 153.0 | 229500 | 2.0756 | | 1.1058 | 154.0 | 231000 | 2.0812 | | 1.1027 | 155.0 | 232500 | 2.0597 | | 1.0996 | 156.0 | 234000 | 2.0684 | | 1.0987 | 157.0 | 235500 | 2.0629 | | 1.0881 | 158.0 | 237000 | 2.0701 | | 1.1143 | 159.0 | 238500 | 2.0740 | | 1.0823 | 160.0 | 240000 | 2.0869 | | 1.0925 | 161.0 | 241500 | 2.0567 | | 1.1034 | 162.0 | 243000 | 2.0833 | | 1.0759 | 163.0 | 244500 | 2.0585 | | 1.0998 | 164.0 | 246000 | 2.0293 | | 1.0891 | 165.0 | 247500 | 2.0608 | | 1.1036 | 166.0 | 249000 | 2.0831 | | 1.076 | 167.0 | 250500 | 2.0979 | | 1.0895 | 168.0 | 252000 | 2.0882 | | 1.0825 | 169.0 | 253500 | 2.0742 | | 1.0793 | 170.0 | 255000 | 2.0841 | | 1.079 | 171.0 | 256500 | 2.0829 | | 1.0653 | 172.0 | 258000 | 2.0888 | | 1.0834 | 173.0 | 259500 | 2.0784 | | 1.0721 | 174.0 | 261000 | 2.0859 | | 1.0712 | 175.0 | 262500 | 2.0810 | | 1.0494 | 176.0 | 264000 | 2.0605 | | 1.0654 | 177.0 | 265500 | 2.0623 | | 1.077 | 178.0 | 267000 | 2.0756 | | 1.056 | 179.0 | 268500 | 2.0782 | | 1.0523 | 180.0 | 270000 | 2.0966 | | 1.0656 | 181.0 | 271500 | 2.0750 | | 1.0636 | 182.0 | 273000 | 2.0769 | | 1.0851 | 183.0 | 274500 | 2.0872 | | 1.0562 | 184.0 | 276000 | 2.0893 | | 1.0534 | 185.0 | 277500 | 2.0661 | | 1.0514 | 186.0 | 279000 | 2.0712 | | 1.062 | 187.0 | 280500 | 2.0769 | | 1.0683 | 188.0 | 282000 | 2.0765 | | 1.0606 | 189.0 | 283500 | 2.0735 | | 1.0555 | 190.0 | 285000 | 2.0710 | | 1.0568 | 191.0 | 286500 | 2.0860 | | 1.0502 | 192.0 | 288000 | 2.0587 | | 1.0437 | 193.0 | 289500 | 2.0998 | | 1.0534 | 194.0 | 291000 | 2.0418 | | 1.062 | 195.0 | 292500 | 2.0724 | | 1.0457 | 196.0 | 294000 | 2.0612 | | 1.0501 | 197.0 | 295500 | 2.1012 | | 1.0728 | 198.0 | 297000 | 2.0721 | | 1.0413 | 199.0 | 298500 | 2.0535 | | 1.0461 | 200.0 | 300000 | 2.0531 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

Izarel/distilbert-base-uncased_fine_tuned_body_text

Izarel

2022-08-01T03:52:20Z

4

0

transformers

[ "transformers", "pytorch", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T19:03:36Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - accuracy - recall - precision - f1 model-index: - name: distilbert-base-uncased_fine_tuned_body_text results: [] ---  # distilbert-base-uncased_fine_tuned_body_text This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 1.2153 - Accuracy: {'accuracy': 0.8827265261428963} - Recall: {'recall': 0.8641975308641975} - Precision: {'precision': 0.8900034993584509} - F1: {'f1': 0.8769106999195494} ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 1000 - num_epochs: 15 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | Recall | Precision | F1 | |:-------------:|:-----:|:-----:|:---------------:|:--------------------------------:|:------------------------------:|:---------------------------------:|:--------------------------:| | 0.3056 | 1.0 | 2284 | 0.3040 | {'accuracy': 0.8874897344648235} | {'recall': 0.8466417487824216} | {'precision': 0.914261252446184} | {'f1': 0.8791531902381653} | | 0.2279 | 2.0 | 4568 | 0.2891 | {'accuracy': 0.8908294552422666} | {'recall': 0.8606863744478424} | {'precision': 0.9086452230060983} | {'f1': 0.8840158213122382} | | 0.1467 | 3.0 | 6852 | 0.3580 | {'accuracy': 0.8882562277580072} | {'recall': 0.8452825914599615} | {'precision': 0.9170557876628164} | {'f1': 0.8797076678257796} | | 0.0921 | 4.0 | 9136 | 0.4560 | {'accuracy': 0.8754448398576512} | {'recall': 0.8948918337297542} | {'precision': 0.8543468858131488} | {'f1': 0.8741494717043756} | | 0.0587 | 5.0 | 11420 | 0.5701 | {'accuracy': 0.8768135778811935} | {'recall': 0.8139087099331748} | {'precision': 0.9221095855254716} | {'f1': 0.8646372277704246} | | 0.0448 | 6.0 | 13704 | 0.6738 | {'accuracy': 0.8767040788393101} | {'recall': 0.8794880507418734} | {'precision': 0.8673070479168994} | {'f1': 0.873355078168935} | | 0.0289 | 7.0 | 15988 | 0.7965 | {'accuracy': 0.8798248015329866} | {'recall': 0.8491335372069317} | {'precision': 0.8967703349282297} | {'f1': 0.8723020536389552} | | 0.0214 | 8.0 | 18272 | 0.8244 | {'accuracy': 0.8811387900355871} | {'recall': 0.8576282704723072} | {'precision': 0.8922931887815225} | {'f1': 0.8746173837712965} | | 0.0147 | 9.0 | 20556 | 0.8740 | {'accuracy': 0.8806460443471119} | {'recall': 0.8669158455091177} | {'precision': 0.8839357893521191} | {'f1': 0.8753430924062213} | | 0.0099 | 10.0 | 22840 | 0.9716 | {'accuracy': 0.8788940596769779} | {'recall': 0.8694076339336279} | {'precision': 0.8787635947338294} | {'f1': 0.8740605784559327} | | 0.0092 | 11.0 | 25124 | 1.0296 | {'accuracy': 0.8822885299753627} | {'recall': 0.8669158455091177} | {'precision': 0.8870089233978444} | {'f1': 0.876847290640394} | | 0.0039 | 12.0 | 27408 | 1.0974 | {'accuracy': 0.8787845606350945} | {'recall': 0.8628383735417374} | {'precision': 0.8836561883772184} | {'f1': 0.8731232091690544} | | 0.0053 | 13.0 | 29692 | 1.0833 | {'accuracy': 0.8799890500958116} | {'recall': 0.8503794314191868} | {'precision': 0.8960496479293472} | {'f1': 0.8726173872617387} | | 0.0032 | 14.0 | 31976 | 1.1731 | {'accuracy': 0.8813030385984123} | {'recall': 0.8705402650356778} | {'precision': 0.8823326828148318} | {'f1': 0.8763968072976055} | | 0.0017 | 15.0 | 34260 | 1.2153 | {'accuracy': 0.8827265261428963} | {'recall': 0.8641975308641975} | {'precision': 0.8900034993584509} | {'f1': 0.8769106999195494} | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

keithanpai/resnet-50-finetuned-eurosat

keithanpai

2022-07-31T23:54:26Z

16

0

transformers

[ "transformers", "pytorch", "tensorboard", "resnet", "image-classification", "generated_from_trainer", "dataset:imagefolder", "license:apache-2.0", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

image-classification

2022-07-31T23:46:46Z

--- license: apache-2.0 tags: - generated_from_trainer datasets: - imagefolder metrics: - accuracy model-index: - name: resnet-50-finetuned-eurosat results: - task: name: Image Classification type: image-classification dataset: name: imagefolder type: imagefolder config: default split: train args: default metrics: - name: Accuracy type: accuracy value: 0.6676646706586826 ---  # resnet-50-finetuned-eurosat This model is a fine-tuned version of [microsoft/resnet-50](https://huggingface.co/microsoft/resnet-50) on the imagefolder dataset. It achieves the following results on the evaluation set: - Loss: 1.1981 - Accuracy: 0.6677 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 128 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 3 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | 1.5279 | 0.99 | 70 | 1.5218 | 0.6677 | | 1.1982 | 1.99 | 140 | 1.2405 | 0.6677 | | 1.0836 | 2.99 | 210 | 1.1981 | 0.6677 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/distilbert-base-uncased_fold_3_ternary

elopezlopez

2022-07-31T23:52:36Z

12

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T23:35:15Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_3_ternary results: [] ---  # distilbert-base-uncased_fold_3_ternary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.7987 - F1: 0.7460 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 289 | 0.5903 | 0.6893 | | 0.5417 | 2.0 | 578 | 0.5822 | 0.7130 | | 0.5417 | 3.0 | 867 | 0.6471 | 0.7385 | | 0.2298 | 4.0 | 1156 | 0.8933 | 0.7322 | | 0.2298 | 5.0 | 1445 | 1.1002 | 0.7147 | | 0.1012 | 6.0 | 1734 | 1.2041 | 0.7249 | | 0.0508 | 7.0 | 2023 | 1.3575 | 0.7195 | | 0.0508 | 8.0 | 2312 | 1.3896 | 0.7385 | | 0.018 | 9.0 | 2601 | 1.5363 | 0.7238 | | 0.018 | 10.0 | 2890 | 1.5336 | 0.7364 | | 0.0142 | 11.0 | 3179 | 1.6335 | 0.7308 | | 0.0142 | 12.0 | 3468 | 1.6915 | 0.7295 | | 0.0047 | 13.0 | 3757 | 1.7087 | 0.7427 | | 0.0058 | 14.0 | 4046 | 1.7875 | 0.7378 | | 0.0058 | 15.0 | 4335 | 1.7649 | 0.7438 | | 0.0051 | 16.0 | 4624 | 1.7987 | 0.7460 | | 0.0051 | 17.0 | 4913 | 1.8435 | 0.7404 | | 0.0025 | 18.0 | 5202 | 1.9623 | 0.7257 | | 0.0025 | 19.0 | 5491 | 1.9005 | 0.7304 | | 0.0029 | 20.0 | 5780 | 1.9437 | 0.7374 | | 0.0011 | 21.0 | 6069 | 1.9840 | 0.7268 | | 0.0011 | 22.0 | 6358 | 1.9411 | 0.7346 | | 0.0025 | 23.0 | 6647 | 1.9233 | 0.7438 | | 0.0025 | 24.0 | 6936 | 1.9415 | 0.7395 | | 0.0015 | 25.0 | 7225 | 1.9481 | 0.7411 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/xlnet-base-cased_fold_2_binary

elopezlopez

2022-07-31T23:13:47Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "xlnet", "text-classification", "generated_from_trainer", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T22:50:03Z

--- license: mit tags: - generated_from_trainer metrics: - f1 model-index: - name: xlnet-base-cased_fold_2_binary results: [] ---  # xlnet-base-cased_fold_2_binary This model is a fine-tuned version of [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.4858 - F1: 0.7648 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 290 | 0.4361 | 0.7404 | | 0.4403 | 2.0 | 580 | 0.5363 | 0.7515 | | 0.4403 | 3.0 | 870 | 0.4858 | 0.7648 | | 0.2505 | 4.0 | 1160 | 0.7127 | 0.7612 | | 0.2505 | 5.0 | 1450 | 0.8930 | 0.7554 | | 0.1425 | 6.0 | 1740 | 0.9897 | 0.7580 | | 0.0869 | 7.0 | 2030 | 1.2683 | 0.7615 | | 0.0869 | 8.0 | 2320 | 1.4988 | 0.7343 | | 0.0411 | 9.0 | 2610 | 1.5082 | 0.7492 | | 0.0411 | 10.0 | 2900 | 1.4974 | 0.7450 | | 0.0306 | 11.0 | 3190 | 1.5723 | 0.7435 | | 0.0306 | 12.0 | 3480 | 1.8446 | 0.7432 | | 0.0291 | 13.0 | 3770 | 1.7113 | 0.7639 | | 0.0207 | 14.0 | 4060 | 1.8073 | 0.7394 | | 0.0207 | 15.0 | 4350 | 1.7524 | 0.7585 | | 0.0171 | 16.0 | 4640 | 1.8751 | 0.7374 | | 0.0171 | 17.0 | 4930 | 1.7849 | 0.7561 | | 0.0084 | 18.0 | 5220 | 1.8618 | 0.7441 | | 0.0064 | 19.0 | 5510 | 1.9613 | 0.7437 | | 0.0064 | 20.0 | 5800 | 1.8898 | 0.7430 | | 0.006 | 21.0 | 6090 | 1.9889 | 0.7409 | | 0.006 | 22.0 | 6380 | 1.9949 | 0.7488 | | 0.0049 | 23.0 | 6670 | 1.9453 | 0.7488 | | 0.0049 | 24.0 | 6960 | 1.9754 | 0.7472 | | 0.002 | 25.0 | 7250 | 1.9946 | 0.7504 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/xlnet-base-cased_fold_1_binary

elopezlopez

2022-07-31T22:49:49Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "xlnet", "text-classification", "generated_from_trainer", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T22:26:16Z

--- license: mit tags: - generated_from_trainer metrics: - f1 model-index: - name: xlnet-base-cased_fold_1_binary results: [] ---  # xlnet-base-cased_fold_1_binary This model is a fine-tuned version of [xlnet-base-cased](https://huggingface.co/xlnet-base-cased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.7607 - F1: 0.7778 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 288 | 0.4111 | 0.7555 | | 0.4387 | 2.0 | 576 | 0.4075 | 0.7540 | | 0.4387 | 3.0 | 864 | 0.5344 | 0.7567 | | 0.2471 | 4.0 | 1152 | 0.7405 | 0.7597 | | 0.2471 | 5.0 | 1440 | 1.0564 | 0.7508 | | 0.1419 | 6.0 | 1728 | 1.0703 | 0.7751 | | 0.0845 | 7.0 | 2016 | 1.0866 | 0.7609 | | 0.0845 | 8.0 | 2304 | 1.2135 | 0.7751 | | 0.05 | 9.0 | 2592 | 1.3649 | 0.7516 | | 0.05 | 10.0 | 2880 | 1.4943 | 0.7590 | | 0.0267 | 11.0 | 3168 | 1.5174 | 0.7412 | | 0.0267 | 12.0 | 3456 | 1.4884 | 0.7559 | | 0.0278 | 13.0 | 3744 | 1.5109 | 0.7405 | | 0.0201 | 14.0 | 4032 | 1.7251 | 0.7409 | | 0.0201 | 15.0 | 4320 | 1.5833 | 0.7354 | | 0.0185 | 16.0 | 4608 | 1.7744 | 0.7598 | | 0.0185 | 17.0 | 4896 | 1.8283 | 0.7619 | | 0.0066 | 18.0 | 5184 | 1.7607 | 0.7778 | | 0.0066 | 19.0 | 5472 | 1.7503 | 0.7719 | | 0.0078 | 20.0 | 5760 | 1.7807 | 0.7508 | | 0.006 | 21.0 | 6048 | 1.6887 | 0.7629 | | 0.006 | 22.0 | 6336 | 1.7041 | 0.7678 | | 0.0074 | 23.0 | 6624 | 1.7337 | 0.7633 | | 0.0074 | 24.0 | 6912 | 1.7548 | 0.7645 | | 0.0035 | 25.0 | 7200 | 1.7685 | 0.7621 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/distilbert-base-uncased_fold_6_binary

elopezlopez

2022-07-31T22:25:18Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T22:14:59Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_6_binary results: [] ---  # distilbert-base-uncased_fold_6_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.6838 - F1: 0.7881 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 290 | 0.4181 | 0.7732 | | 0.4097 | 2.0 | 580 | 0.3967 | 0.7697 | | 0.4097 | 3.0 | 870 | 0.5811 | 0.7797 | | 0.2034 | 4.0 | 1160 | 0.8684 | 0.7320 | | 0.2034 | 5.0 | 1450 | 0.9116 | 0.7718 | | 0.0794 | 6.0 | 1740 | 1.0588 | 0.7690 | | 0.0278 | 7.0 | 2030 | 1.2092 | 0.7738 | | 0.0278 | 8.0 | 2320 | 1.2180 | 0.7685 | | 0.0233 | 9.0 | 2610 | 1.3005 | 0.7676 | | 0.0233 | 10.0 | 2900 | 1.4009 | 0.7634 | | 0.0093 | 11.0 | 3190 | 1.4528 | 0.7805 | | 0.0093 | 12.0 | 3480 | 1.4803 | 0.7859 | | 0.0088 | 13.0 | 3770 | 1.4775 | 0.7750 | | 0.0077 | 14.0 | 4060 | 1.6171 | 0.7699 | | 0.0077 | 15.0 | 4350 | 1.6429 | 0.7636 | | 0.0047 | 16.0 | 4640 | 1.5619 | 0.7819 | | 0.0047 | 17.0 | 4930 | 1.5833 | 0.7724 | | 0.0034 | 18.0 | 5220 | 1.6400 | 0.7853 | | 0.0008 | 19.0 | 5510 | 1.6508 | 0.7792 | | 0.0008 | 20.0 | 5800 | 1.6838 | 0.7881 | | 0.0009 | 21.0 | 6090 | 1.6339 | 0.7829 | | 0.0009 | 22.0 | 6380 | 1.6824 | 0.7806 | | 0.0016 | 23.0 | 6670 | 1.6867 | 0.7876 | | 0.0016 | 24.0 | 6960 | 1.7107 | 0.7877 | | 0.0013 | 25.0 | 7250 | 1.6933 | 0.7812 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/distilbert-base-uncased_fold_5_binary

elopezlopez

2022-07-31T22:14:52Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T22:04:19Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_5_binary results: [] ---  # distilbert-base-uncased_fold_5_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.5093 - F1: 0.7801 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 288 | 0.4760 | 0.7315 | | 0.3992 | 2.0 | 576 | 0.4428 | 0.7785 | | 0.3992 | 3.0 | 864 | 0.5093 | 0.7801 | | 0.2021 | 4.0 | 1152 | 0.6588 | 0.7634 | | 0.2021 | 5.0 | 1440 | 0.9174 | 0.7713 | | 0.0945 | 6.0 | 1728 | 0.9832 | 0.7726 | | 0.0321 | 7.0 | 2016 | 1.2103 | 0.7672 | | 0.0321 | 8.0 | 2304 | 1.3759 | 0.7616 | | 0.0134 | 9.0 | 2592 | 1.4405 | 0.7570 | | 0.0134 | 10.0 | 2880 | 1.4591 | 0.7710 | | 0.0117 | 11.0 | 3168 | 1.4947 | 0.7713 | | 0.0117 | 12.0 | 3456 | 1.6224 | 0.7419 | | 0.0081 | 13.0 | 3744 | 1.6462 | 0.7520 | | 0.0083 | 14.0 | 4032 | 1.6880 | 0.7637 | | 0.0083 | 15.0 | 4320 | 1.7080 | 0.7380 | | 0.0048 | 16.0 | 4608 | 1.7352 | 0.7551 | | 0.0048 | 17.0 | 4896 | 1.6761 | 0.7713 | | 0.0024 | 18.0 | 5184 | 1.7553 | 0.76 | | 0.0024 | 19.0 | 5472 | 1.7312 | 0.7673 | | 0.005 | 20.0 | 5760 | 1.7334 | 0.7713 | | 0.0032 | 21.0 | 6048 | 1.7963 | 0.7578 | | 0.0032 | 22.0 | 6336 | 1.7529 | 0.7679 | | 0.0025 | 23.0 | 6624 | 1.7741 | 0.7662 | | 0.0025 | 24.0 | 6912 | 1.7515 | 0.7679 | | 0.0004 | 25.0 | 7200 | 1.7370 | 0.7765 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/distilbert-base-uncased_fold_3_binary

elopezlopez

2022-07-31T21:53:55Z

4

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T21:43:33Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_3_binary results: [] ---  # distilbert-base-uncased_fold_3_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.8310 - F1: 0.7584 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 289 | 0.4560 | 0.7522 | | 0.4008 | 2.0 | 578 | 0.4790 | 0.7567 | | 0.4008 | 3.0 | 867 | 0.6368 | 0.7557 | | 0.1967 | 4.0 | 1156 | 0.6854 | 0.7534 | | 0.1967 | 5.0 | 1445 | 0.9823 | 0.7309 | | 0.0768 | 6.0 | 1734 | 1.2531 | 0.7511 | | 0.0202 | 7.0 | 2023 | 1.2906 | 0.7391 | | 0.0202 | 8.0 | 2312 | 1.4025 | 0.7460 | | 0.0087 | 9.0 | 2601 | 1.5713 | 0.7507 | | 0.0087 | 10.0 | 2890 | 1.4212 | 0.7528 | | 0.0162 | 11.0 | 3179 | 1.5775 | 0.7511 | | 0.0162 | 12.0 | 3468 | 1.6361 | 0.7377 | | 0.0048 | 13.0 | 3757 | 1.6972 | 0.7542 | | 0.0098 | 14.0 | 4046 | 1.6569 | 0.7565 | | 0.0098 | 15.0 | 4335 | 1.7547 | 0.7325 | | 0.0042 | 16.0 | 4624 | 1.8108 | 0.7544 | | 0.0042 | 17.0 | 4913 | 1.7593 | 0.7554 | | 0.0041 | 18.0 | 5202 | 1.7582 | 0.7551 | | 0.0041 | 19.0 | 5491 | 1.8200 | 0.7512 | | 0.0029 | 20.0 | 5780 | 1.8310 | 0.7584 | | 0.0018 | 21.0 | 6069 | 1.8146 | 0.7568 | | 0.0018 | 22.0 | 6358 | 1.7870 | 0.7558 | | 0.0029 | 23.0 | 6647 | 1.8508 | 0.7530 | | 0.0029 | 24.0 | 6936 | 1.8327 | 0.7543 | | 0.0001 | 25.0 | 7225 | 1.8546 | 0.7561 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

elopezlopez/distilbert-base-uncased_fold_1_binary

elopezlopez

2022-07-31T21:33:03Z

6

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T20:57:24Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - f1 model-index: - name: distilbert-base-uncased_fold_1_binary results: [] ---  # distilbert-base-uncased_fold_1_binary This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 1.5992 - F1: 0.7687 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 25 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | No log | 1.0 | 288 | 0.3960 | 0.7467 | | 0.3988 | 2.0 | 576 | 0.3947 | 0.7487 | | 0.3988 | 3.0 | 864 | 0.4511 | 0.7662 | | 0.1853 | 4.0 | 1152 | 0.7226 | 0.7285 | | 0.1853 | 5.0 | 1440 | 0.9398 | 0.7334 | | 0.0827 | 6.0 | 1728 | 1.0547 | 0.7427 | | 0.0287 | 7.0 | 2016 | 1.1602 | 0.7563 | | 0.0287 | 8.0 | 2304 | 1.3332 | 0.7171 | | 0.0219 | 9.0 | 2592 | 1.3429 | 0.7420 | | 0.0219 | 10.0 | 2880 | 1.2603 | 0.7648 | | 0.0139 | 11.0 | 3168 | 1.4126 | 0.7569 | | 0.0139 | 12.0 | 3456 | 1.3195 | 0.7483 | | 0.0115 | 13.0 | 3744 | 1.4356 | 0.7491 | | 0.0035 | 14.0 | 4032 | 1.5693 | 0.7636 | | 0.0035 | 15.0 | 4320 | 1.4071 | 0.7662 | | 0.0071 | 16.0 | 4608 | 1.4561 | 0.7579 | | 0.0071 | 17.0 | 4896 | 1.5405 | 0.7634 | | 0.0041 | 18.0 | 5184 | 1.5862 | 0.7589 | | 0.0041 | 19.0 | 5472 | 1.6782 | 0.76 | | 0.0024 | 20.0 | 5760 | 1.5699 | 0.7677 | | 0.0006 | 21.0 | 6048 | 1.5991 | 0.7467 | | 0.0006 | 22.0 | 6336 | 1.6205 | 0.7682 | | 0.0003 | 23.0 | 6624 | 1.6334 | 0.7643 | | 0.0003 | 24.0 | 6912 | 1.5992 | 0.7687 | | 0.0011 | 25.0 | 7200 | 1.6053 | 0.7624 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

DS-20202/DoubleHardDebias

DS-20202

2022-07-31T20:32:45Z

0

null

[ "license:mit", "region:us" ]

null

2022-07-31T12:08:09Z

--- title: Double Hard Debiasing emoji: 👁 colorFrom: blue colorTo: pink sdk: gradio sdk_version: 3.1.1 app_file: app.py pinned: false license: mit --- Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

neuralmagic/oBERT-6-downstream-pruned-block4-90-QAT-squadv1

neuralmagic

2022-07-31T19:52:34Z

5

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T19:21:02Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-pruned-block4-90-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - Sparsity 90% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 6 ``` The dev-set performance of this model: ``` EM = 76.56 F1 = 84.59 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-6-downstream-pruned-unstructured-90-squadv1

neuralmagic

2022-07-31T19:52:34Z

4

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T14:00:05Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-pruned-unstructured-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - Sparsity 90% - unstructured`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 6 ``` The dev-set performance of this model: ``` EM = 79.16 F1 = 86.78 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-teacher-squadv1

neuralmagic

2022-07-31T19:52:34Z

396

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:47:26Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # SQuADv1 teacher This model is used as a teacher for all runs on the SQuADv1 downstream task in the paper [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). SQuADv1 dev-set: ``` EM = 81.41 F1 = 88.54 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-3-downstream-pruned-block4-90-squadv1

neuralmagic

2022-07-31T19:52:33Z

5

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T14:01:41Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-block4-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 90% - 4-block`. ``` Pruning method: oBERT downstream block-4 Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 71.36 F1 = 80.69 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-3-downstream-dense-squadv1

neuralmagic

2022-07-31T19:52:33Z

5

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T14:00:43Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-dense-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - 0% Sparsity`, and it represents an upper bound for performance of the corresponding pruned models: - 80% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-80-squadv1` - 80% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-80-squadv1` - 90% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-90-squadv1` - 90% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-90-squadv1` SQuADv1 dev-set: ``` EM = 76.62 F1 = 84.65 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-6-downstream-dense-QAT-squadv1

neuralmagic

2022-07-31T19:52:33Z

2

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T19:20:36Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-dense-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - 0% Sparsity - QAT`, and it represents an upper bound for performance of the corresponding pruned and quantized models: - 80% unstructured QAT: `neuralmagic/oBERT-6-downstream-pruned-unstructured-80-QAT-squadv1` - 80% block-4 QAT: `neuralmagic/oBERT-6-downstream-pruned-block4-80-QAT-squadv1` - 90% unstructured QAT: `neuralmagic/oBERT-6-downstream-pruned-unstructured-90-QAT-squadv1` - 90% block-4 QAT: `neuralmagic/oBERT-6-downstream-pruned-block4-90-QAT-squadv1` SQuADv1 dev-set: ``` EM = 80.85 F1 = 87.94 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-3-downstream-pruned-block4-80-QAT-squadv1

neuralmagic

2022-07-31T19:52:33Z

5

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T19:21:28Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-block4-80-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 80% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 80% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 72.70 F1 = 82.04 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-6-downstream-dense-squadv1

neuralmagic

2022-07-31T19:52:33Z

8

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:59:35Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-6-downstream-dense-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 6 Layers - 0% Sparsity`, and it represents an upper bound for performance of the corresponding pruned models: - 80% unstructured: `neuralmagic/oBERT-6-downstream-pruned-unstructured-80-squadv1` - 80% block-4: `neuralmagic/oBERT-6-downstream-pruned-block4-80-squadv1` - 90% unstructured: `neuralmagic/oBERT-6-downstream-pruned-unstructured-90-squadv1` - 90% block-4: `neuralmagic/oBERT-6-downstream-pruned-block4-90-squadv1` SQuADv1 dev-set: ``` EM = 81.17 F1 = 88.32 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-3-downstream-pruned-block4-90-QAT-squadv1

neuralmagic

2022-07-31T19:52:33Z

14

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T19:21:41Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-block4-90-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 90% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 70.00 F1 = 79.66 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-3-downstream-pruned-unstructured-90-squadv1

neuralmagic

2022-07-31T19:52:33Z

13

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T14:01:15Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-3-downstream-pruned-unstructured-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 3 Layers - Sparsity 90% - unstructured`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 3 ``` The dev-set performance of this model: ``` EM = 73.61 F1 = 82.50 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-3-upstream-pretrained-dense

neuralmagic

2022-07-31T19:52:33Z

14

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:56:43Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-3-upstream-pretrained-dense This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to 3 layers from `neuralmagic/oBERT-12-upstream-pretrained-dense`, pretrained with knowledge distillation. This model is used as a starting point for downstream finetuning and pruning runs presented in the `Table 3 - 3 Layers`. The model can also be used for finetuning on any downstream task, as a starting point instead of the three times larger `bert-base-uncased` model. Finetuned and pruned versions of this model on the SQuADv1 downstream task, as described in the paper: - 0%: `neuralmagic/oBERT-3-downstream-dense-squadv1` - 80% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-80-squadv1` - 80% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-80-squadv1` - 90% unstructured: `neuralmagic/oBERT-3-downstream-pruned-unstructured-90-squadv1` - 90% block-4: `neuralmagic/oBERT-3-downstream-pruned-block4-90-squadv1` ``` Training objective: masked language modeling (MLM) + knowledge distillation Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 0% Number of layers: 3 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pruned-unstructured-97-v2

neuralmagic

2022-07-31T19:52:32Z

8

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-06-17T07:25:30Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-12-upstream-pruned-unstructured-97-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the upstream pruned model used as a starting point for sparse-transfer learning to downstream tasks presented in the `Table 2 - oBERT - {SQuADv1, MNLI, QQP} - 97%` (in the upcoming updated version of the paper). Finetuned versions of this model for each downstream task are: - SQuADv1: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-squadv1-v2` - MNLI: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-mnli-v2` - QQP: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-qqp-v2` ``` Pruning method: oBERT upstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 97% Number of layers: 12 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pruned-unstructured-97-finetuned-squadv1-v2

neuralmagic

2022-07-31T19:52:32Z

5

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-06-17T07:30:56Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-upstream-pruned-unstructured-97-finetuned-squadv1-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - SQuADv1 97%` (in the upcoming updated version of the paper). ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 97% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `(*)`): ``` | oBERT 97% | F1 | EM | | ------------- | ----- | ----- | | seed=42 | 84.92 | 76.94 | | seed=3407 | 84.87 | 76.71 | | seed=123 | 84.95 | 77.06 | | seed=12345 (*)| 84.95 | 76.90 | | ------------- | ----- | ----- | | mean | 84.92 | 76.90 | | stdev | 0.037 | 0.145 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp-v2

neuralmagic

2022-07-31T19:52:32Z

7

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:qqp", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-06-17T07:31:44Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: qqp --- # oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - QQP 90%` (in the upcoming updated version of the paper). ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: QQP Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `(*)`): ``` | oBERT 90% | acc | F1 | | ------------- | ----- | ----- | | seed=42 | 90.94 | 87.79 | | seed=3407 | 91.00 | 87.81 | | seed=123 | 90.94 | 87.73 | | seed=12345 (*)| 91.07 | 87.92 | | ------------- | ----- | ----- | | mean | 90.99 | 87.81 | | stdev | 0.061 | 0.079 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2

neuralmagic

2022-07-31T19:52:32Z

7

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-06-17T07:30:41Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - SQuADv1 90%` (in the upcoming updated version of the paper). ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `(*)`): ``` | oBERT 90% | F1 | EM | | ------------ | ----- | ----- | | seed=42 | 88.55 | 81.48 | | seed=3407 | 88.34 | 81.25 | | seed=123 (*)| 88.64 | 81.57 | | seed=12345 | 88.44 | 81.43 | | ------------ | ----- | ----- | | mean | 88.49 | 81.43 | | stdev | 0.130 | 0.134 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pruned-unstructured-90-v2

neuralmagic

2022-07-31T19:52:32Z

4

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-06-17T07:22:37Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-12-upstream-pruned-unstructured-90-v2 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the upstream pruned model used as a starting point for sparse-transfer learning to downstream tasks presented in the `Table 2 - oBERT - {SQuADv1, MNLI, QQP} - 90%` (in the upcoming updated version of the paper). Finetuned versions of this model for each downstream task are: - SQuADv1: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2` - MNLI: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-mnli-v2` - QQP: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp-v2` ``` Pruning method: oBERT upstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 90% Number of layers: 12 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp

neuralmagic

2022-07-31T19:52:32Z

10

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:qqp", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:58:30Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: qqp --- # oBERT-12-upstream-pruned-unstructured-90-finetuned-qqp This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 2 - oBERT - QQP 90%`. ``` Pruning method: oBERT upstream unstructured + sparse-transfer to downstream Paper: https://arxiv.org/abs/2203.07259 Dataset: QQP Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 90% | acc | F1 | | ------------ | ----- | ----- | | seed=42 (*)| 90.93 | 87.77 | | seed=3407 | 90.70 | 87.49 | | seed=54321 | 90.86 | 87.68 | | ------------ | ----- | ----- | | mean | 90.83 | 87.65 | | stdev | 0.117 | 0.143 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-unstructured-90-mnli

neuralmagic

2022-07-31T19:52:31Z

19

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:mnli", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:54:55Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: mnli --- # oBERT-12-downstream-pruned-unstructured-90-mnli This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - MNLI 90%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: MNLI Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 90% | m-acc | mm-acc| | ------------ | ----- | ----- | | seed=42 | 83.74 | 84.31 | | seed=3407 (*)| 83.85 | 84.40 | | seed=54321 | 83.77 | 84.33 | | ------------ | ----- | ----- | | mean | 83.79 | 84.35 | | stdev | 0.056 | 0.047 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-unstructured-97-mnli

neuralmagic

2022-07-31T19:52:31Z

8

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:mnli", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:55:09Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: mnli --- # oBERT-12-downstream-pruned-unstructured-97-mnli This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - MNLI 97%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: MNLI Sparsity: 97% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 97% | m-acc | mm-acc| | ------------ | ----- | ----- | | seed=42 (*)| 82.10 | 81.94 | | seed=3407 | 81.81 | 82.27 | | seed=54321 | 81.40 | 81.83 | | ------------ | ----- | ----- | | mean | 81.77 | 82.01 | | stdev | 0.351 | 0.228 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-unstructured-90-squadv1

neuralmagic

2022-07-31T19:52:31Z

7

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:53:32Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-downstream-pruned-unstructured-90-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - SQuADv1 90%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 90% | F1 | EM | | ------------ | ----- | ----- | | seed=42 | 88.22 | 81.10 | | seed=3407 (*)| 88.46 | 81.26 | | seed=54321 | 88.26 | 81.00 | | ------------ | ----- | ----- | | mean | 88.31 | 81.12 | | stdev | 0.128 | 0.131 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-unstructured-90-qqp

neuralmagic

2022-07-31T19:52:31Z

12

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:qqp", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:55:50Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: qqp --- # oBERT-12-downstream-pruned-unstructured-90-qqp This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - QQP 90%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: QQP Sparsity: 90% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 90% | acc | F1 | | ------------ | ----- | ----- | | seed=42 | 91.30 | 88.24 | | seed=3407 (*)| 91.39 | 88.36 | | seed=54321 | 91.36 | 88.29 | | ------------ | ----- | ----- | | mean | 91.35 | 88.30 | | stdev | 0.045 | 0.060 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-upstream-pretrained-dense

neuralmagic

2022-07-31T19:52:31Z

4

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:bookcorpus", "dataset:wikipedia", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:56:17Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: - bookcorpus - wikipedia --- # oBERT-12-upstream-pretrained-dense This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the pretrained dense model used as a teacher for upstream pruning runs, as described in the paper. The model can be finetuned on any downstream task, just like the standard `bert-base-uncased` model which is used as initialization for training of this model. Sparse versions of this model: - 90% sparse: `neuralmagic/oBERT-12-upstream-pruned-unstructured-90` - 97% sparse: `neuralmagic/oBERT-12-upstream-pruned-unstructured-97` ``` Training objective: masked language modeling (MLM) Paper: https://arxiv.org/abs/2203.07259 Dataset: BookCorpus and English Wikipedia Sparsity: 0% Number of layers: 12 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-unstructured-80-squadv1

neuralmagic

2022-07-31T19:52:31Z

11

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:53:16Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-downstream-pruned-unstructured-80-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - SQuADv1 80%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 80% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 80% | F1 | EM | | ------------ | ----- | ----- | | seed=42 | 88.95 | 82.08 | | seed=3407 (*)| 89.16 | 82.05 | | seed=54321 | 89.01 | 82.12 | | ------------ | ----- | ----- | | mean | 89.04 | 82.08 | | stdev | 0.108 | 0.035 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-block4-90-QAT-squadv1

neuralmagic

2022-07-31T19:52:30Z

4

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:squad", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T19:20:22Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: squad --- # oBERT-12-downstream-pruned-block4-90-QAT-squadv1 This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 3 - 12 Layers - Sparsity 90% - 4-block + QAT`. ``` Pruning method: oBERT downstream block-4 + QAT Paper: https://arxiv.org/abs/2203.07259 Dataset: SQuADv1 Sparsity: 90% Number of layers: 12 ``` The dev-set performance of this model: ``` EM = 78.84 F1 = 86.68 ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

neuralmagic/oBERT-12-downstream-pruned-unstructured-80-mnli

neuralmagic

2022-07-31T19:52:30Z

7

0

transformers

[ "transformers", "pytorch", "bert", "oBERT", "sparsity", "pruning", "compression", "en", "dataset:mnli", "arxiv:2203.07259", "endpoints_compatible", "region:us" ]

null

2022-05-25T13:54:40Z

--- tags: - bert - oBERT - sparsity - pruning - compression language: en datasets: mnli --- # oBERT-12-downstream-pruned-unstructured-80-mnli This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259). It corresponds to the model presented in the `Table 1 - 30 Epochs - oBERT - MNLI 80%`. ``` Pruning method: oBERT downstream unstructured Paper: https://arxiv.org/abs/2203.07259 Dataset: MNLI Sparsity: 80% Number of layers: 12 ``` The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`): ``` | oBERT 80% | m-acc | mm-acc| | ------------ | ----- | ----- | | seed=42 | 84.30 | 84.98 | | seed=3407 (*)| 84.46 | 84.99 | | seed=54321 | 84.18 | 84.76 | | ------------ | ----- | ----- | | mean | 84.32 | 84.91 | | stdev | 0.140 | 0.133 | ``` Code: [https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT](https://github.com/neuralmagic/sparseml/tree/main/research/optimal_BERT_surgeon_oBERT) If you find the model useful, please consider citing our work. ## Citation info ```bibtex @article{kurtic2022optimal, title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models}, author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan}, journal={arXiv preprint arXiv:2203.07259}, year={2022} } ```

Ebuu/Aaaaa

Ebuu

2022-07-31T19:00:30Z

0

null

[ "license:bigscience-bloom-rail-1.0", "region:us" ]

null

2022-07-31T19:00:30Z

--- license: bigscience-bloom-rail-1.0 ---

SummerChiam/rust_image_classification_12

SummerChiam

2022-07-31T17:33:58Z

48

0

transformers

[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

image-classification

2022-07-31T17:33:47Z

--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: rust_image_classification_12 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9679595232009888 --- # rust_image_classification_12 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for **anything** by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### nonrust0 ![nonrust0](images/nonrust0.png) #### rust0 ![rust0](images/rust0.png)

QuickSilver007/Reinforce-Pong-PLE-v0

QuickSilver007

2022-07-31T16:23:22Z

0

null

[ "Pong-PLE-v0", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]

reinforcement-learning

2022-07-31T16:23:13Z

--- tags: - Pong-PLE-v0 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-Pong-PLE-v0 results: - metrics: - type: mean_reward value: -16.00 +/- 0.00 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: Pong-PLE-v0 type: Pong-PLE-v0 --- # **Reinforce** Agent playing **Pong-PLE-v0** This is a trained model of a **Reinforce** agent playing **Pong-PLE-v0** . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5

anneke/finetuning-distilbert-base-uncased-finetuned-sst-2-english-5000-samples-final

anneke

2022-07-31T16:05:59Z

7

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T15:49:08Z

--- license: apache-2.0 tags: - generated_from_trainer metrics: - accuracy - f1 model-index: - name: finetuning-distilbert-base-uncased-finetuned-sst-2-english-5000-samples-final results: [] ---  # finetuning-distilbert-base-uncased-finetuned-sst-2-english-5000-samples-final This model is a fine-tuned version of [distilbert-base-uncased-finetuned-sst-2-english](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.1289 - Accuracy: 0.977 - F1: 0.9878 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 5 ### Training results ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

SummerChiam/pond_image_classification_11

SummerChiam

2022-07-31T15:36:10Z

50

0

transformers

[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

image-classification

2022-07-31T15:35:57Z

--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: pond_image_classification_11 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9951980710029602 --- # pond_image_classification_11 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for **anything** by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### Algae0 ![Algae0](images/Algae0.png) #### Boiling0 ![Boiling0](images/Boiling0.png) #### BoilingNight0 ![BoilingNight0](images/BoilingNight0.png) #### Normal0 ![Normal0](images/Normal0.png) #### NormalCement0 ![NormalCement0](images/NormalCement0.png) #### NormalNight0 ![NormalNight0](images/NormalNight0.png) #### NormalRain0 ![NormalRain0](images/NormalRain0.png)

samwit/ddpm-afhq-cats-128

samwit

2022-07-31T15:31:53Z

2

0

diffusers

[ "diffusers", "tensorboard", "en", "dataset:imagefolder", "license:apache-2.0", "diffusers:DDPMPipeline", "region:us" ]

null

2022-07-31T00:49:28Z

--- language: en license: apache-2.0 library_name: diffusers tags: [] datasets: imagefolder metrics: [] ---  # ddpm-afhq-cats-128 ## Model description This diffusion model is trained with the [🤗 Diffusers](https://github.com/huggingface/diffusers) library on the `imagefolder` dataset. ## Intended uses & limitations #### How to use ```python # TODO: add an example code snippet for running this diffusion pipeline ``` #### Limitations and bias [TODO: provide examples of latent issues and potential remediations] ## Training data [TODO: describe the data used to train the model] ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001 - train_batch_size: 16 - eval_batch_size: 16 - gradient_accumulation_steps: 1 - optimizer: AdamW with betas=(None, None), weight_decay=None and epsilon=None - lr_scheduler: None - lr_warmup_steps: 500 - ema_inv_gamma: None - ema_inv_gamma: None - ema_inv_gamma: None - mixed_precision: fp16 ### Training results 📈 [TensorBoard logs](https://huggingface.co/samwit/ddpm-afhq-cats-128/tensorboard?#scalars)

CuteBlack/gfp_guided_diffusion_200k

CuteBlack

2022-07-31T15:10:42Z

0

6

null

[ "license:mit", "region:us" ]

null

2022-07-15T22:24:34Z

--- license: mit --- 256x256 Diffusion model trained on 1000+ NSFW gay furry pics (with same composition) 'attention_resolutions': '16', 'class_cond': False, 'diffusion_steps': 1000, 'rescale_timesteps': True, 'timestep_respacing': 'ddim100', 'image_size': 256, 'learn_sigma': True, 'noise_schedule': 'linear', 'num_channels': 128, 'num_heads': 1, 'num_res_blocks': 2, 'use_checkpoint': use_checkpoint, 'use_fp16': True, 'use_scale_shift_norm': False,

Kinahem/Reinforce-3

Kinahem

2022-07-31T13:02:51Z

0

null

[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]

reinforcement-learning

2022-07-31T13:02:35Z

--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-3 results: - metrics: - type: mean_reward value: 471.20 +/- 86.40 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 --- # **Reinforce** Agent playing **CartPole-v1** This is a trained model of a **Reinforce** agent playing **CartPole-v1** . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5

Vasanth/bert_emo_classifier

Vasanth

2022-07-31T12:34:43Z

4

1

transformers

[ "transformers", "pytorch", "tensorboard", "bert", "text-classification", "generated_from_trainer", "dataset:emotion", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-30T23:30:12Z

--- license: apache-2.0 tags: - generated_from_trainer datasets: - emotion model-index: - name: bert_emo_classifier results: [] ---  # bert_emo_classifier This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the emotion dataset. It achieves the following results on the evaluation set: - Loss: 0.2748 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 4 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:----:|:---------------:| | 0.9063 | 0.25 | 500 | 0.4845 | | 0.3362 | 0.5 | 1000 | 0.3492 | | 0.2759 | 0.75 | 1500 | 0.2819 | | 0.2521 | 1.0 | 2000 | 0.2464 | | 0.1705 | 1.25 | 2500 | 0.2345 | | 0.1841 | 1.5 | 3000 | 0.2013 | | 0.1428 | 1.75 | 3500 | 0.1926 | | 0.1747 | 2.0 | 4000 | 0.1866 | | 0.1082 | 2.25 | 4500 | 0.2302 | | 0.1142 | 2.5 | 5000 | 0.2118 | | 0.1205 | 2.75 | 5500 | 0.2318 | | 0.1135 | 3.0 | 6000 | 0.2306 | | 0.0803 | 3.25 | 6500 | 0.2625 | | 0.0745 | 3.5 | 7000 | 0.2850 | | 0.085 | 3.75 | 7500 | 0.2719 | | 0.0701 | 4.0 | 8000 | 0.2748 | ### Framework versions - Transformers 4.15.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.10.3

QuickSilver007/Reinforce-CartPole-v1

QuickSilver007

2022-07-31T12:21:39Z

0

null

[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]

reinforcement-learning

2022-07-31T12:21:29Z

--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-CartPole-v1 results: - metrics: - type: mean_reward value: 123.40 +/- 12.38 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 --- # **Reinforce** Agent playing **CartPole-v1** This is a trained model of a **Reinforce** agent playing **CartPole-v1** . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5

Kinahem/Reinforce-1

Kinahem

2022-07-31T12:07:01Z

0

null

[ "CartPole-v1", "reinforce", "reinforcement-learning", "custom-implementation", "deep-rl-class", "model-index", "region:us" ]

reinforcement-learning

2022-07-31T12:06:53Z

--- tags: - CartPole-v1 - reinforce - reinforcement-learning - custom-implementation - deep-rl-class model-index: - name: Reinforce-1 results: - metrics: - type: mean_reward value: 18.30 +/- 7.93 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: CartPole-v1 type: CartPole-v1 --- # **Reinforce** Agent playing **CartPole-v1** This is a trained model of a **Reinforce** agent playing **CartPole-v1** . To learn to use this model and train yours check Unit 5 of the Deep Reinforcement Learning Class: https://github.com/huggingface/deep-rl-class/tree/main/unit5

fabf21/finetuning-sentiment-model-3000-samples

fabf21

2022-07-31T11:16:46Z

5

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "text-classification", "generated_from_trainer", "dataset:imdb", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text-classification

2022-07-31T11:05:03Z

--- license: apache-2.0 tags: - generated_from_trainer datasets: - imdb model-index: - name: finetuning-sentiment-model-3000-samples results: [] ---  # finetuning-sentiment-model-3000-samples This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the imdb dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

Neha2608/xlm-roberta-base-finetuned-panx-en

Neha2608

2022-07-31T10:42:20Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "xlm-roberta", "generated_from_trainer", "dataset:xtreme", "license:mit", "endpoints_compatible", "region:us" ]

null

2022-07-02T12:17:17Z

--- license: mit tags: - generated_from_trainer datasets: - xtreme metrics: - f1 model-index: - name: xlm-roberta-base-finetuned-panx-en results: [] ---  # xlm-roberta-base-finetuned-panx-en This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset. It achieves the following results on the evaluation set: - Loss: 0.4329 - F1: 0.6431 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 24 - eval_batch_size: 24 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | 1.1554 | 1.0 | 50 | 0.5989 | 0.4571 | | 0.5361 | 2.0 | 100 | 0.4329 | 0.6431 | ### Framework versions - Transformers 4.11.3 - Pytorch 1.12.0+cu113 - Datasets 1.16.1 - Tokenizers 0.10.3

Okyx/finetuned-amazon-en-es

Okyx

2022-07-31T10:33:05Z

10

0

transformers

[ "transformers", "tf", "mt5", "text2text-generation", "generated_from_keras_callback", "license:apache-2.0", "autotrain_compatible", "endpoints_compatible", "region:us" ]

text2text-generation

2022-07-31T09:41:05Z

--- license: apache-2.0 tags: - generated_from_keras_callback model-index: - name: Okyx/finetuned-amazon-en-es results: [] ---  # Okyx/finetuned-amazon-en-es This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on an unknown dataset. It achieves the following results on the evaluation set: - Train Loss: 4.0154 - Validation Loss: 3.3292 - Epoch: 7 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5.6e-05, 'decay_steps': 9672, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01} - training_precision: float32 ### Training results | Train Loss | Validation Loss | Epoch | |:----------:|:---------------:|:-----:| | 9.2009 | 4.0465 | 0 | | 5.7436 | 3.6640 | 1 | | 5.0419 | 3.5296 | 2 | | 4.6412 | 3.4582 | 3 | | 4.3722 | 3.3943 | 4 | | 4.1947 | 3.3610 | 5 | | 4.0747 | 3.3295 | 6 | | 4.0154 | 3.3292 | 7 | ### Framework versions - Transformers 4.21.0 - TensorFlow 2.8.2 - Datasets 2.4.0 - Tokenizers 0.12.1

Neha2608/xlm-roberta-base-finetuned-panx-it

Neha2608

2022-07-31T10:26:20Z

7

0

transformers

[ "transformers", "pytorch", "xlm-roberta", "generated_from_trainer", "dataset:xtreme", "license:mit", "endpoints_compatible", "region:us" ]

null

2022-07-02T11:59:49Z

--- license: mit tags: - generated_from_trainer datasets: - xtreme metrics: - f1 model-index: - name: xlm-roberta-base-finetuned-panx-it results: [] ---  # xlm-roberta-base-finetuned-panx-it This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the xtreme dataset. It achieves the following results on the evaluation set: - Loss: 0.2740 - F1: 0.7919 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 24 - eval_batch_size: 24 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 2 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:----:|:---------------:|:------:| | 0.8185 | 1.0 | 70 | 0.3369 | 0.7449 | | 0.2899 | 2.0 | 140 | 0.2740 | 0.7919 | ### Framework versions - Transformers 4.11.3 - Pytorch 1.12.0+cu113 - Datasets 1.16.1 - Tokenizers 0.10.3

CuteBlack/gfp_guided_diffusion_v4

CuteBlack

2022-07-31T10:04:18Z

0

9

null

[ "license:mit", "region:us" ]

null

2022-07-31T09:48:47Z

--- license: mit --- Open AI diffusion model that has trained on every single NSFW gay furry illustrations on e621.net that’s over the community score of 100. Excluding extreme fetishes and underage contents. 'attention_resolutions': '32, 16, 8', 'class_cond': False, 'diffusion_steps': 1000, 'rescale_timesteps': True, 'image_size': 256, 'learn_sigma': True, 'noise_schedule': 'linear', 'num_channels': 128, 'num_heads': 4, 'num_res_blocks': 2, 'resblock_updown': True, 'use_checkpoint': use_checkpoint, 'use_fp16': True, 'use_scale_shift_norm': True

ijnekonasa/ppo-LunarLander-v2

ijnekonasa

2022-07-31T03:58:07Z

0

stable-baselines3

[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]

reinforcement-learning

2022-07-31T03:57:44Z

--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - metrics: - type: mean_reward value: 252.64 +/- 18.29 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 --- # **PPO** Agent playing **LunarLander-v2** This is a trained model of a **PPO** agent playing **LunarLander-v2** using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```

Frikallo/DeepLeffen-TSM_Leffen

Frikallo

2022-07-31T01:31:44Z

3

0

transformers

[ "transformers", "pytorch", "tensorboard", "gpt2", "text-generation", "generated_from_trainer", "license:mit", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]

text-generation

2022-07-31T01:27:45Z

--- license: mit tags: - generated_from_trainer model-index: - name: DeepLeffen-TSM_Leffen results: [] ---  # DeepLeffen-TSM_Leffen This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 0.0001372 - train_batch_size: 1 - eval_batch_size: 8 - seed: 2780791035 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 1.0 ### Training results ### Framework versions - Transformers 4.20.1 - Pytorch 1.9.1+cu111 - Datasets 2.3.2 - Tokenizers 0.12.1

keithanpai/vit-base-patch16-224-finetuned-eurosat

keithanpai

2022-07-31T00:07:31Z

55

0

transformers

[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "generated_from_trainer", "dataset:imagefolder", "license:apache-2.0", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

image-classification

2022-07-30T23:42:57Z

--- license: apache-2.0 tags: - generated_from_trainer datasets: - imagefolder metrics: - accuracy model-index: - name: vit-base-patch16-224-finetuned-eurosat results: - task: name: Image Classification type: image-classification dataset: name: imagefolder type: imagefolder config: default split: train args: default metrics: - name: Accuracy type: accuracy value: 0.8632734530938124 ---  # vit-base-patch16-224-finetuned-eurosat This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the imagefolder dataset. It achieves the following results on the evaluation set: - Loss: 0.3953 - Accuracy: 0.8633 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 32 - eval_batch_size: 32 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 128 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_ratio: 0.1 - num_epochs: 3 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | 0.6081 | 0.99 | 70 | 0.5482 | 0.8004 | | 0.4515 | 1.99 | 140 | 0.4245 | 0.8533 | | 0.3967 | 2.99 | 210 | 0.3953 | 0.8633 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.12.0+cu113 - Datasets 2.4.0 - Tokenizers 0.12.1

sophiestein/experiment_2

sophiestein

2022-07-30T17:57:00Z

6

0

transformers

[ "transformers", "pytorch", "tensorboard", "distilbert", "token-classification", "generated_from_trainer", "dataset:conll2003", "license:apache-2.0", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

token-classification

2022-07-30T10:21:23Z

--- license: apache-2.0 tags: - generated_from_trainer datasets: - conll2003 metrics: - precision - recall - f1 - accuracy model-index: - name: experiment_2 results: - task: name: Token Classification type: token-classification dataset: name: conll2003 type: conll2003 config: conll2003 split: train args: conll2003 metrics: - name: Precision type: precision value: 0.8840954508052192 - name: Recall type: recall value: 0.8925943508188939 - name: F1 type: f1 value: 0.8883245733183724 - name: Accuracy type: accuracy value: 0.9746737103791174 ---  # experiment_2 This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the conll2003 dataset. It achieves the following results on the evaluation set: - Loss: 0.1211 - Precision: 0.8841 - Recall: 0.8926 - F1: 0.8883 - Accuracy: 0.9747 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 ### Training results | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| | 0.2418 | 1.0 | 878 | 0.0695 | 0.9159 | 0.9255 | 0.9207 | 0.9816 | | 0.0541 | 2.0 | 1756 | 0.0592 | 0.9244 | 0.9343 | 0.9293 | 0.9833 | | 0.0303 | 3.0 | 2634 | 0.0602 | 0.9260 | 0.9388 | 0.9323 | 0.9838 | ### Framework versions - Transformers 4.21.0 - Pytorch 1.11.0+cpu - Datasets 2.4.0 - Tokenizers 0.12.1

anzorq/kbd_lat-ru_char_tokenizer

anzorq

2022-07-30T16:16:55Z

10

0

transformers

[ "transformers", "pytorch", "jax", "t5", "text2text-generation", "translation", "ru", "kbd", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]

translation

2022-07-29T10:31:32Z

--- language: - ru - kbd tags: - translation ---

comodoro/testpyramidsrnd2

comodoro

2022-07-30T15:58:53Z

4

0

ml-agents

[ "ml-agents", "tensorboard", "onnx", "unity-ml-agents", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-Pyramids", "region:us" ]

reinforcement-learning

2022-07-30T15:58:47Z

--- tags: - unity-ml-agents - ml-agents - deep-reinforcement-learning - reinforcement-learning - ML-Agents-Pyramids library_name: ml-agents --- # **ppo** Agent playing **Pyramids** This is a trained model of a **ppo** agent playing **Pyramids** using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://github.com/huggingface/ml-agents#get-started We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: ### Resume the training ``` mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent **playing directly in your browser:**. 1. Go to https://huggingface.co/spaces/unity/ML-Agents-Pyramids 2. Step 1: Write your model_id: comodoro/testpyramidsrnd2 3. Step 2: Select your *.nn /*.onnx file 4. Click on Watch the agent play 👀

constanter/PPO-LunarLander-v2

constanter

2022-07-30T13:34:25Z

0

stable-baselines3

[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]

reinforcement-learning

2022-07-30T13:33:54Z

--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - metrics: - type: mean_reward value: 268.37 +/- 20.32 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 --- # **PPO** Agent playing **LunarLander-v2** This is a trained model of a **PPO** agent playing **LunarLander-v2** using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```

SummerChiam/rust_image_classification_7

SummerChiam

2022-07-30T12:04:23Z

57

0

transformers

[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

image-classification

2022-07-30T12:04:11Z

--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: rust_image_classification_7 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9645569324493408 --- # rust_image_classification_7 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for **anything** by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### nonrust ![nonrust](images/nonrust.png) #### rust ![rust](images/rust.png)

devetle/a2c-AntBulletEnv-v0

devetle

2022-07-30T10:14:08Z

2

0

stable-baselines3

[ "stable-baselines3", "AntBulletEnv-v0", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]

reinforcement-learning

2022-07-30T10:13:05Z

--- library_name: stable-baselines3 tags: - AntBulletEnv-v0 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: A2C results: - metrics: - type: mean_reward value: 1098.81 +/- 321.12 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: AntBulletEnv-v0 type: AntBulletEnv-v0 --- # **A2C** Agent playing **AntBulletEnv-v0** This is a trained model of a **A2C** agent playing **AntBulletEnv-v0** using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```

SummerChiam/pond_image_classification_10

SummerChiam

2022-07-30T08:57:50Z

50

0

transformers

[ "transformers", "pytorch", "tensorboard", "vit", "image-classification", "huggingpics", "model-index", "autotrain_compatible", "endpoints_compatible", "region:us" ]

image-classification

2022-07-30T08:57:38Z

--- tags: - image-classification - pytorch - huggingpics metrics: - accuracy model-index: - name: pond_image_classification_10 results: - task: name: Image Classification type: image-classification metrics: - name: Accuracy type: accuracy value: 0.9948979616165161 --- # pond_image_classification_10 Autogenerated by HuggingPics🤗🖼️ Create your own image classifier for **anything** by running [the demo on Google Colab](https://colab.research.google.com/github/nateraw/huggingpics/blob/main/HuggingPics.ipynb). Report any issues with the demo at the [github repo](https://github.com/nateraw/huggingpics). ## Example Images #### Algae ![Algae](images/Algae.png) #### Boiling ![Boiling](images/Boiling.png) #### BoilingNight ![BoilingNight](images/BoilingNight.png) #### Normal ![Normal](images/Normal.png) #### NormalCement ![NormalCement](images/NormalCement.png) #### NormalNight ![NormalNight](images/NormalNight.png) #### NormalRain ![NormalRain](images/NormalRain.png)

mbarnig/lb-de-fr-en-pt-coqui-vits-tts

mbarnig

2022-07-30T06:00:58Z

222

7

transformers

[ "transformers", "tensorboard", "TTS", "audio", "synthesis", "yourTTS", "speech", "coqui.ai", "lb", "de", "fr", "en", "pt", "dataset:mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS", "license:cc-by-nc-sa-4.0", "endpoints_compatible", "region:us" ]

null

2022-07-08T20:42:32Z

--- license: cc-by-nc-sa-4.0 language: - lb - de - fr - en - pt tags: - TTS - audio - synthesis - yourTTS - speech - coqui.ai datasets: - mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS --- #### This model has been trained from scratch with my customized dataset [mbarnig/lb-de-fr-en-pt-12800-TTS_CORPUS](https://huggingface.co/datasets/mbarnig/lb-de-fr-en-pt-12800-TTS-CORPUS) and the 🐸 [Coqui-TTS multilingual VITS-model recipe](https://github.com/coqui-ai/TTS/tree/dev/recipes/multilingual/vits_tts) (version 0.7.1). The model was trained without phonemes with the following character-set: ``` characters="abcdefghijklmnopqrstuvwxyz ßàáâãäçèéêëíîïóôõöùúûü", punctuations="!'(),-.:;? ", phonemes=None, ``` #### A live inference-demo of the model is available in my HuggingFace space ⌨️ 🇱🇺 🔈 [mbarnig/lb_de_fr_en_pt_COQUI_VITS_TTS](https://huggingface.co/spaces/mbarnig/lb_de_fr_en_pt_COQUI_VITS_TTS). #### Click the tab *training metrics* above to view the live Tensorboard of the model training. ![tensorboard](tensorboard.png)

vinitharaj/distilbert-base-uncased-finetuned-squad2

vinitharaj

2022-07-30T05:47:35Z

5

0

transformers

[ "transformers", "tf", "tensorboard", "distilbert", "question-answering", "generated_from_keras_callback", "license:apache-2.0", "endpoints_compatible", "region:us" ]

question-answering

2022-07-29T07:47:14Z

--- license: apache-2.0 tags: - generated_from_keras_callback model-index: - name: vinitharaj/distilbert-base-uncased-finetuned-squad2 results: [] ---  # vinitharaj/distilbert-base-uncased-finetuned-squad2 This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on an unknown dataset. It achieves the following results on the evaluation set: - Train Loss: 0.4953 - Validation Loss: 0.3885 - Epoch: 1 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - optimizer: {'name': 'Adam', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 1602, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False} - training_precision: float32 ### Training results | Train Loss | Validation Loss | Epoch | |:----------:|:---------------:|:-----:| | 0.7037 | 0.4222 | 0 | | 0.4953 | 0.3885 | 1 | ### Framework versions - Transformers 4.21.0 - TensorFlow 2.8.2 - Datasets 2.4.0 - Tokenizers 0.12.1

Migga/ViT-BERT-Chess-V4

Migga

2022-07-30T04:26:03Z

1

0

transformers

[ "transformers", "pytorch", "vision-encoder-decoder", "image-text-to-text", "generated_from_trainer", "endpoints_compatible", "region:us" ]

image-text-to-text

2022-07-29T16:57:48Z

--- tags: - generated_from_trainer model-index: - name: ViT-BERT-Chess-V4 results: [] ---  # ViT-BERT-Chess-V4 This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.3213 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 8 - eval_batch_size: 4 - seed: 42 - gradient_accumulation_steps: 8 - total_train_batch_size: 64 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 5 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:-----:|:---------------:| | 3.705 | 1.0 | 3895 | 3.5686 | | 3.5139 | 2.0 | 7790 | 3.4288 | | 3.4156 | 3.0 | 11685 | 3.3663 | | 3.3661 | 4.0 | 15580 | 3.3331 | | 3.3352 | 5.0 | 19475 | 3.3213 | ### Framework versions - Transformers 4.20.1 - Pytorch 1.12.0+cu116 - Datasets 2.3.2 - Tokenizers 0.12.1

reachrkr/testpyramidsrnd

reachrkr

2022-07-30T02:46:18Z

5

0

ml-agents

[ "ml-agents", "tensorboard", "onnx", "unity-ml-agents", "deep-reinforcement-learning", "reinforcement-learning", "ML-Agents-Pyramids", "region:us" ]

reinforcement-learning

2022-07-28T06:59:12Z

--- tags: - unity-ml-agents - ml-agents - deep-reinforcement-learning - reinforcement-learning - ML-Agents-Pyramids library_name: ml-agents --- # **ppo** Agent playing **Pyramids** This is a trained model of a **ppo** agent playing **Pyramids** using the [Unity ML-Agents Library](https://github.com/Unity-Technologies/ml-agents). ## Usage (with ML-Agents) The Documentation: https://github.com/huggingface/ml-agents#get-started We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub: ### Resume the training ``` mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume ``` ### Watch your Agent play You can watch your agent **playing directly in your browser:**. 1. Go to https://huggingface.co/spaces/unity/ML-Agents-Pyramids 2. Step 1: Write your model_id: reachrkr/testpyramidsrnd 3. Step 2: Select your *.nn /*.onnx file 4. Click on Watch the agent play 👀

huggingtweets/dags

huggingtweets

2022-07-30T01:32:18Z

4

0

transformers

[ "transformers", "pytorch", "gpt2", "text-generation", "huggingtweets", "en", "autotrain_compatible", "text-generation-inference", "endpoints_compatible", "region:us" ]

text-generation

2022-07-30T01:30:26Z

--- language: en thumbnail: http://www.huggingtweets.com/dags/1659144733206/predictions.png tags: - huggingtweets widget: - text: "My dream is" --- <div class="inline-flex flex-col" style="line-height: 1.5;"> <div class="flex"> <div style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/722815128501026817/IMWCRzEn_400x400.jpg')"> </div> <div style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')"> </div> <div style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')"> </div> </div> <div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div> <div style="text-align: center; font-size: 16px; font-weight: 800">DAGs</div> <div style="text-align: center; font-size: 14px;">@dags</div> </div> I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets). Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)! ## How does it work? The model uses the following pipeline. ![pipeline](https://github.com/borisdayma/huggingtweets/blob/master/img/pipeline.png?raw=true) To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI). ## Training data The model was trained on tweets from DAGs. | Data | DAGs | | --- | --- | | Tweets downloaded | 3003 | | Retweets | 31 | | Short tweets | 158 | | Tweets kept | 2814 | [Explore the data](https://wandb.ai/wandb/huggingtweets/runs/3qyk6uzo/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline. ## Training procedure The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @dags's tweets. Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/18qzuqjb) for full transparency and reproducibility. At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/18qzuqjb/artifacts) is logged and versioned. ## How to use You can use this model directly with a pipeline for text generation: ```python from transformers import pipeline generator = pipeline('text-generation', model='huggingtweets/dags') generator("My dream is", num_return_sequences=5) ``` ## Limitations and bias The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias). In addition, the data present in the user's tweets further affects the text generated by the model. ## About *Built by Boris Dayma* [![Follow](https://img.shields.io/twitter/follow/borisdayma?style=social)](https://twitter.com/intent/follow?screen_name=borisdayma) For more details, visit the project repository. [![GitHub stars](https://img.shields.io/github/stars/borisdayma/huggingtweets?style=social)](https://github.com/borisdayma/huggingtweets)

rebolforces/ppo-LunarLander-v2

rebolforces

2022-07-30T00:43:21Z

1

0

stable-baselines3

[ "stable-baselines3", "LunarLander-v2", "deep-reinforcement-learning", "reinforcement-learning", "model-index", "region:us" ]

reinforcement-learning

2022-07-23T09:28:37Z

--- library_name: stable-baselines3 tags: - LunarLander-v2 - deep-reinforcement-learning - reinforcement-learning - stable-baselines3 model-index: - name: PPO results: - metrics: - type: mean_reward value: 285.83 +/- 15.59 name: mean_reward task: type: reinforcement-learning name: reinforcement-learning dataset: name: LunarLander-v2 type: LunarLander-v2 --- # **PPO** Agent playing **LunarLander-v2** This is a trained model of a **PPO** agent playing **LunarLander-v2** using the [stable-baselines3 library](https://github.com/DLR-RM/stable-baselines3). ## Usage (with Stable-baselines3) TODO: Add your code ```python from stable_baselines3 import ... from huggingface_sb3 import load_from_hub ... ```

yanaiela/roberta-base-epoch_80

yanaiela

2022-07-29T23:08:59Z

7

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_80", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T18:03:25Z

--- language: en tags: - roberta-base - roberta-base-epoch_80 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 80 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_80. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_78

yanaiela

2022-07-29T23:08:15Z

12

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_78", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T18:01:03Z

--- language: en tags: - roberta-base - roberta-base-epoch_78 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 78 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_78. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_77

yanaiela

2022-07-29T23:07:53Z

10

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_77", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:59:57Z

--- language: en tags: - roberta-base - roberta-base-epoch_77 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 77 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_77. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_73

yanaiela

2022-07-29T23:06:21Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_73", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:55:51Z

--- language: en tags: - roberta-base - roberta-base-epoch_73 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 73 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_73. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_71

yanaiela

2022-07-29T23:05:36Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_71", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:53:19Z

--- language: en tags: - roberta-base - roberta-base-epoch_71 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 71 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_71. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_67

yanaiela

2022-07-29T23:04:02Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_67", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:48:39Z

--- language: en tags: - roberta-base - roberta-base-epoch_67 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 67 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_67. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_66

yanaiela

2022-07-29T23:03:37Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_66", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:46:45Z

--- language: en tags: - roberta-base - roberta-base-epoch_66 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 66 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_66. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_64

yanaiela

2022-07-29T23:02:45Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_64", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:43:33Z

--- language: en tags: - roberta-base - roberta-base-epoch_64 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 64 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_64. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_61

yanaiela

2022-07-29T23:01:44Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_61", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:39:32Z

--- language: en tags: - roberta-base - roberta-base-epoch_61 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 61 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_61. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_59

yanaiela

2022-07-29T23:01:00Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_59", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:35:53Z

--- language: en tags: - roberta-base - roberta-base-epoch_59 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 59 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_59. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_57

yanaiela

2022-07-29T23:00:18Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_57", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:34:22Z

--- language: en tags: - roberta-base - roberta-base-epoch_57 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 57 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_57. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_56

yanaiela

2022-07-29T22:59:56Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_56", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:33:29Z

--- language: en tags: - roberta-base - roberta-base-epoch_56 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 56 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_56. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_54

yanaiela

2022-07-29T22:59:09Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_54", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:31:39Z

--- language: en tags: - roberta-base - roberta-base-epoch_54 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 54 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_54. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_52

yanaiela

2022-07-29T22:58:22Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_52", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:30:02Z

--- language: en tags: - roberta-base - roberta-base-epoch_52 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 52 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_52. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_50

yanaiela

2022-07-29T22:57:31Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_50", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:28:26Z

--- language: en tags: - roberta-base - roberta-base-epoch_50 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 50 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_50. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_47

yanaiela

2022-07-29T22:56:19Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_47", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:26:12Z

--- language: en tags: - roberta-base - roberta-base-epoch_47 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 47 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_47. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_46

yanaiela

2022-07-29T22:55:54Z

3

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_46", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:25:28Z

--- language: en tags: - roberta-base - roberta-base-epoch_46 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 46 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_46. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_45

yanaiela

2022-07-29T22:55:32Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_45", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:24:44Z

--- language: en tags: - roberta-base - roberta-base-epoch_45 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 45 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_45. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_43

yanaiela

2022-07-29T22:54:43Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_43", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:23:18Z

--- language: en tags: - roberta-base - roberta-base-epoch_43 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 43 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_43. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_39

yanaiela

2022-07-29T22:53:02Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_39", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:20:23Z

--- language: en tags: - roberta-base - roberta-base-epoch_39 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 39 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_39. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_31

yanaiela

2022-07-29T22:50:29Z

6

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_31", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:14:05Z

--- language: en tags: - roberta-base - roberta-base-epoch_31 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 31 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_31. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_27

yanaiela

2022-07-29T22:49:16Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_27", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:10:38Z

--- language: en tags: - roberta-base - roberta-base-epoch_27 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 27 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_27. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_26

yanaiela

2022-07-29T22:48:56Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_26", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:09:55Z

--- language: en tags: - roberta-base - roberta-base-epoch_26 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 26 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_26. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_21

yanaiela

2022-07-29T22:47:23Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_21", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:06:01Z

--- language: en tags: - roberta-base - roberta-base-epoch_21 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 21 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_21. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_20

yanaiela

2022-07-29T22:47:06Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_20", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:05:11Z

--- language: en tags: - roberta-base - roberta-base-epoch_20 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 20 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_20. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_17

yanaiela

2022-07-29T22:46:08Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_17", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:02:47Z

--- language: en tags: - roberta-base - roberta-base-epoch_17 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 17 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_17. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_15

yanaiela

2022-07-29T22:45:30Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_15", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:01:23Z

--- language: en tags: - roberta-base - roberta-base-epoch_15 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 15 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_15. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_14

yanaiela

2022-07-29T22:45:11Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_14", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T17:00:38Z

--- language: en tags: - roberta-base - roberta-base-epoch_14 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 14 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_14. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_9

yanaiela

2022-07-29T22:43:40Z

4

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_9", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T16:56:14Z

--- language: en tags: - roberta-base - roberta-base-epoch_9 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 9 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_9. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_7

yanaiela

2022-07-29T22:43:03Z

7

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_7", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T16:54:40Z

--- language: en tags: - roberta-base - roberta-base-epoch_7 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 7 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_7. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```

yanaiela/roberta-base-epoch_3

yanaiela

2022-07-29T22:41:43Z

5

0

transformers

[ "transformers", "pytorch", "roberta", "fill-mask", "roberta-base", "roberta-base-epoch_3", "en", "dataset:wikipedia", "dataset:bookcorpus", "arxiv:1907.11692", "arxiv:2207.14251", "license:mit", "autotrain_compatible", "endpoints_compatible", "region:us" ]

fill-mask

2022-07-28T16:51:25Z

--- language: en tags: - roberta-base - roberta-base-epoch_3 license: mit datasets: - wikipedia - bookcorpus --- # RoBERTa, Intermediate Checkpoint - Epoch 3 This model is part of our reimplementation of the [RoBERTa model](https://arxiv.org/abs/1907.11692), trained on Wikipedia and the Book Corpus only. We train this model for almost 100K steps, corresponding to 83 epochs. We provide the 84 checkpoints (including the randomly initialized weights before the training) to provide the ability to study the training dynamics of such models, and other possible use-cases. These models were trained in part of a work that studies how simple statistics from data, such as co-occurrences affects model predictions, which are described in the paper [Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions](https://arxiv.org/abs/2207.14251). This is RoBERTa-base epoch_3. ## Model Description This model was captured during a reproduction of [RoBERTa-base](https://huggingface.co/roberta-base), for English: it is a Transformers model pretrained on a large corpus of English data, using the Masked Language Modelling (MLM). The intended uses, limitations, training data and training procedure for the fully trained model are similar to [RoBERTa-base](https://huggingface.co/roberta-base). Two major differences with the original model: * We trained our model for 100K steps, instead of 500K * We only use Wikipedia and the Book Corpus, as corpora which are publicly available. ### How to use Using code from [RoBERTa-base](https://huggingface.co/roberta-base), here is an example based on PyTorch: ``` from transformers import pipeline model = pipeline("fill-mask", model='yanaiela/roberta-base-epoch_83', device=-1, top_k=10) model("Hello, I'm the <mask> RoBERTa-base language model") ``` ## Citation info ```bibtex @article{2207.14251, Author = {Yanai Elazar and Nora Kassner and Shauli Ravfogel and Amir Feder and Abhilasha Ravichander and Marius Mosbach and Yonatan Belinkov and Hinrich Schütze and Yoav Goldberg}, Title = {Measuring Causal Effects of Data Statistics on Language Model's `Factual' Predictions}, Year = {2022}, Eprint = {arXiv:2207.14251}, } ```