dots.llm1.base / README.md
mrfakename's picture
Duplicate from rednote-hilab/dots.llm1.base
9d5c891 verified
---
license: mit
license_link: https://huggingface.co/rednote-hilab/dots.llm1.base/blob/main/LICENSE
library_name: transformers
language:
- en
- zh
---
# dots1
<p align="center">
<img src="figures/new_logo2.png" width="300"/>
<p>
<p align="center">
&nbsp&nbsp🤗 <a href="https://huggingface.co/rednote-hilab">Hugging Face</a>&nbsp&nbsp | &nbsp&nbsp 📑 <a href="https://github.com/rednote-hilab/dots.llm1/blob/main/dots1_tech_report.pdf">Paper</a> &nbsp&nbsp
<br>
🖥️ <a href="https://huggingface.co/spaces/rednote-hilab/dots-demo">Demo</a>&nbsp&nbsp | &nbsp&nbsp💬 <a href="figures/wechat.png">WeChat (微信)</a>&nbsp&nbsp | &nbsp&nbsp📕 <a href="https://www.xiaohongshu.com/user/profile/683ffe42000000001d021a4c">rednote</a>&nbsp&nbsp
</p>
Visit our Hugging Face (click links above), search checkpoints with names starting with `dots.llm1` or visit the [dots1 collection](https://huggingface.co/collections/rednote-hilab/dotsllm1-68246aaaaba3363374a8aa7c), and you will find all you need! Enjoy!
## News
- 2025.06.06: We released the `dots.llm1` series. Check our [report](https://github.com/rednote-hilab/dots.llm1/blob/main/dots1_tech_report.pdf) for more details!
## 1. Introduction
The `dots.llm1` model is a large-scale MoE model that activates 14B parameters out of a total of 142B parameters, delivering performance on par with state-of-the-art models.
Leveraging our meticulously crafted and efficient data processing pipeline, `dots.llm1` achieves performance comparable to Qwen2.5-72B after pretrained on 11.2T high-quality tokens without synthetic data. To foster further research, we open-source intermediate training checkpoints at every one trillion tokens, providing valuable insights into the learning dynamics of large language models.
<p align="center">
<img width="90%" src="./figures/performance.png">
</p>
## 2. Model Summary
**This repo contains the base and instruction-tuned `dots.llm1` model**. which has the following features:
- Type: A MoE model with 14B activated and 142B total parameters trained on 11.2T tokens.
- Training Stages: Pretraining and SFT.
- Architecture: Multi-head Attention with QK-Norm in attention Layer, fine-grained MoE utilizing top-6 out of 128 routed experts, plus 2 shared experts.
- Number of Layers: 62
- Number of Attention Heads: 32
- Supported Languages: English, Chinese
- Context Length: 32,768 tokens
- License: MIT
The highlights from `dots.llm1` include:
- **Enhanced Data Processing**: We propose a scalable and fine-grained *three-stage* data processing framework designed to generate large-scale, high-quality and diverse data for pretraining.
- **No Synthetic Data during Pretraining**: *11.2 trillion* high-quality non-synthetic tokens was used in base model pretraining.
- **Performance and Cost Efficiency**: `dots.llm1` is an open-source model that activates only *14B* parameters at inference, delivering both comprehensive capabilities and high computational efficiency.
- **Infrastructure**: We introduce an innovative MoE all-to-all communication and computation overlapping recipe based on interleaved 1F1B pipeline scheduling and an efficient grouped GEMM implementation to boost computational efficiency.
- **Open Accessibility to Model Dynamics**: Intermediate model checkpoints for *every 1T tokens* trained are released, facilitating future research into the learning dynamics of large language models.
## 3. Example Usage
### Model Downloads
<div align="center">
| **Model** | **#Total Params** | **#Activated Params** | **Context Length** | **Download Link** |
| :------------: | :------------: | :------------: | :------------: | :------------: |
| dots.llm1.base | 142B | 14B | 32K | [🤗 Hugging Face](https://huggingface.co/rednote-hilab/dots.llm1.base) |
| dots.llm1.inst | 142B | 14B | 32K | [🤗 Hugging Face](https://huggingface.co/rednote-hilab/dots.llm1.inst) |
</div>
### Docker (recommended)
The docker images are available on [Docker Hub](https://hub.docker.com/repository/docker/rednotehilab/dots1/tags), based on the official images.
You can start a server via vllm.
```shell
docker run --gpus all \
-v ~/.cache/huggingface:/root/.cache/huggingface \
-p 8000:8000 \
--ipc=host \
rednotehilab/dots1:vllm-openai-v0.9.0.1 \
--model rednote-hilab/dots.llm1.inst \
--tensor-parallel-size 8 \
--trust-remote-code \
--served-model-name dots1
```
Then you can verify whether the model is running successfully in the following way.
```shell
curl http://localhost:8000/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "dots1",
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Who won the world series in 2020?"}
],
"max_tokens": 32,
"temperature": 0
}'
```
### Inference with huggingface
We are working to merge it into Transformers ([PR #38143](https://github.com/huggingface/transformers/pull/38143)).
#### Text Completion
```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
model_name = "rednote-hilab/dots.llm1.base"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype=torch.bfloat16)
text = "An attention function can be described as mapping a query and a set of key-value pairs to an output, where the query, keys, values, and output are all vectors. The output is"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs.to(model.device), max_new_tokens=100)
result = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(result)
```
#### Chat Completion
```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
model_name = "rednote-hilab/dots.llm1.inst"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype=torch.bfloat16)
messages = [
{"role": "user", "content": "Write a piece of quicksort code in C++"}
]
input_tensor = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt")
outputs = model.generate(input_tensor.to(model.device), max_new_tokens=200)
result = tokenizer.decode(outputs[0][input_tensor.shape[1]:], skip_special_tokens=True)
print(result)
```
### Inference with vllm
[vLLM](https://github.com/vllm-project/vllm) is a high-throughput and memory-efficient inference and serving engine for LLMs. Official support for this feature is covered in [PR #18254](https://github.com/vllm-project/vllm/pull/18254).
```shell
vllm serve dots.llm1.inst --port 8000 --tensor-parallel-size 8
```
An OpenAI-compatible API will be available at `http://localhost:8000/v1`.
### Inference with sglang
[SGLang](https://github.com/sgl-project/sglang) is a fast serving framework for large language models and vision language models. SGLang could be used to launch a server with OpenAI-compatible API service. Official support for this feature is covered in [PR #6471](https://github.com/sgl-project/sglang/pull/6471).
Getting started is as simple as running:
```shell
python -m sglang.launch_server --model-path dots.llm1.inst --tp 8 --host 0.0.0.0 --port 8000
```
An OpenAI-compatible API will be available at `http://localhost:8000/v1`.
## 4. Evaluation Results
Detailed evaluation results are reported in this [📑 report](https://github.com/rednote-hilab/dots.llm1/blob/main/dots1_tech_report.pdf).
## Citation
If you find `dots.llm1` is useful or want to use in your projects, please kindly cite our paper:
```
@article{dots1,
title={dots.llm1 Technical Report},
author={rednote-hilab},
journal={arXiv preprint arXiv:TBD},
year={2025}
}
```