Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v5.15/checkpoint-11264/ema.safetensors +3 -0
- checkpoints-v5.15/checkpoint-11264/eval_state.json +3 -0
- checkpoints-v5.15/checkpoint-11264/model.safetensors +3 -0
- checkpoints-v5.15/checkpoint-11264/optimizer.pt +3 -0
- checkpoints-v5.15/checkpoint-11264/rng_state.pth +3 -0
- checkpoints-v5.15/checkpoint-11264/scaler.pt +3 -0
- checkpoints-v5.15/checkpoint-11264/scheduler.pt +3 -0
- checkpoints-v5.15/checkpoint-11264/trainer_state.json +452 -0
- checkpoints-v5.15/checkpoint-11264/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -70,3 +70,4 @@ checkpoints-v5.14/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 70 |
checkpoints-v4.6++/checkpoint-16384/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 71 |
checkpoints-v4.6++/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 72 |
checkpoints-v5.14-b/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 70 |
checkpoints-v4.6++/checkpoint-16384/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 71 |
checkpoints-v4.6++/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 72 |
checkpoints-v5.14-b/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 73 |
+
checkpoints-v5.15/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v5.15/checkpoint-11264/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8aeae678d96b2bc552efb96492aae35461bf9b44c404caa2365069f6b0d1c4d6
|
| 3 |
+
size 54599376
|
checkpoints-v5.15/checkpoint-11264/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24b2b0179d94205419f3b3deaea6251d75c9eed58ff9bf83f7a0efd3f58c5610
|
| 3 |
+
size 57008049
|
checkpoints-v5.15/checkpoint-11264/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e091cb48566b9e7967ae291f0dd1032103cff51ab136369c12516ada6f083a23
|
| 3 |
+
size 54599408
|
checkpoints-v5.15/checkpoint-11264/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ca487b195a1d0ad02251c04d8b768b492e5ef6a5270ab45557231dcd171da93
|
| 3 |
+
size 76550347
|
checkpoints-v5.15/checkpoint-11264/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db58e1fa1043771ede7e059026a171217e359b9877ce2b1ad7350e695e5e52ca
|
| 3 |
+
size 14645
|
checkpoints-v5.15/checkpoint-11264/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6723ae794957af182ae5058283117171219221fff5d286943166a33ff0964826
|
| 3 |
+
size 1383
|
checkpoints-v5.15/checkpoint-11264/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d4fb09a735654b7310c6b8e8c4da0b9a798d3d75aadadc24eb1116ac42560a1
|
| 3 |
+
size 1465
|
checkpoints-v5.15/checkpoint-11264/trainer_state.json
ADDED
|
@@ -0,0 +1,452 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.4854440062921542,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 11264,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.04413127329928675,
|
| 14 |
+
"grad_norm": 8.887663841247559,
|
| 15 |
+
"learning_rate": 9.990234375e-05,
|
| 16 |
+
"loss": 8.19105339050293,
|
| 17 |
+
"step": 1024
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.04413127329928675,
|
| 21 |
+
"eval_bleu": 0.2371850032485889,
|
| 22 |
+
"eval_ce_clean_loss": 3.3862914774717807,
|
| 23 |
+
"eval_ce_pred_loss": 5.394671515361078,
|
| 24 |
+
"eval_flow_latent_mse_loss": 0.07054425241437548,
|
| 25 |
+
"eval_flow_velocity_mse_loss": 0.657677960929586,
|
| 26 |
+
"eval_loss": 4.653980879387113,
|
| 27 |
+
"flow/cos_sim": 0.6173963620464431,
|
| 28 |
+
"flow/improvement_ratio": 0.9840314367940939,
|
| 29 |
+
"flow/mag_ratio_mean": 0.5731897604490902,
|
| 30 |
+
"flow/mag_ratio_std": 0.14144727260446244,
|
| 31 |
+
"step": 1024
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04413127329928675,
|
| 35 |
+
"eval_bleu": 0.2371850032485889,
|
| 36 |
+
"eval_ce_clean_loss": 3.3862914774717807,
|
| 37 |
+
"eval_ce_pred_loss": 5.394671515361078,
|
| 38 |
+
"eval_flow_latent_mse_loss": 0.07054425241437548,
|
| 39 |
+
"eval_flow_velocity_mse_loss": 0.657677960929586,
|
| 40 |
+
"eval_loss": 4.653980879387113,
|
| 41 |
+
"eval_runtime": 211.7869,
|
| 42 |
+
"eval_samples_per_second": 141.652,
|
| 43 |
+
"eval_steps_per_second": 2.214,
|
| 44 |
+
"flow/cos_sim": 0.6173963620464431,
|
| 45 |
+
"flow/improvement_ratio": 0.9840314367940939,
|
| 46 |
+
"flow/mag_ratio_mean": 0.5731897604490902,
|
| 47 |
+
"flow/mag_ratio_std": 0.14144727260446244,
|
| 48 |
+
"step": 1024
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.0882625465985735,
|
| 52 |
+
"grad_norm": 18.40336799621582,
|
| 53 |
+
"learning_rate": 9.947705025097448e-05,
|
| 54 |
+
"loss": 3.3889715671539307,
|
| 55 |
+
"step": 2048
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.0882625465985735,
|
| 59 |
+
"eval_bleu": 0.24089355021843578,
|
| 60 |
+
"eval_ce_clean_loss": 1.5524053565983071,
|
| 61 |
+
"eval_ce_pred_loss": 5.091436030259773,
|
| 62 |
+
"eval_flow_latent_mse_loss": 0.048218625540068664,
|
| 63 |
+
"eval_flow_velocity_mse_loss": 0.3526596663984409,
|
| 64 |
+
"eval_loss": 2.462427265862666,
|
| 65 |
+
"flow/cos_sim": 0.8220945212886785,
|
| 66 |
+
"flow/improvement_ratio": 0.9908133576165384,
|
| 67 |
+
"flow/mag_ratio_mean": 0.7910515288554275,
|
| 68 |
+
"flow/mag_ratio_std": 0.1580576471078879,
|
| 69 |
+
"step": 2048
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.0882625465985735,
|
| 73 |
+
"eval_bleu": 0.24089355021843578,
|
| 74 |
+
"eval_ce_clean_loss": 1.5524053565983071,
|
| 75 |
+
"eval_ce_pred_loss": 5.091436030259773,
|
| 76 |
+
"eval_flow_latent_mse_loss": 0.048218625540068664,
|
| 77 |
+
"eval_flow_velocity_mse_loss": 0.3526596663984409,
|
| 78 |
+
"eval_loss": 2.462427265862666,
|
| 79 |
+
"eval_runtime": 207.3942,
|
| 80 |
+
"eval_samples_per_second": 144.652,
|
| 81 |
+
"eval_steps_per_second": 2.261,
|
| 82 |
+
"flow/cos_sim": 0.8220945212886785,
|
| 83 |
+
"flow/improvement_ratio": 0.9908133576165384,
|
| 84 |
+
"flow/mag_ratio_mean": 0.7910515288554275,
|
| 85 |
+
"flow/mag_ratio_std": 0.1580576471078879,
|
| 86 |
+
"step": 2048
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.13239381989786023,
|
| 90 |
+
"grad_norm": 13.780325889587402,
|
| 91 |
+
"learning_rate": 9.7915094488941e-05,
|
| 92 |
+
"loss": 2.019826889038086,
|
| 93 |
+
"step": 3072
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13239381989786023,
|
| 97 |
+
"eval_bleu": 0.20682060276074551,
|
| 98 |
+
"eval_ce_clean_loss": 0.7854123648041601,
|
| 99 |
+
"eval_ce_pred_loss": 5.334758354910909,
|
| 100 |
+
"eval_flow_latent_mse_loss": 0.040881452071212375,
|
| 101 |
+
"eval_flow_velocity_mse_loss": 0.2279702990548189,
|
| 102 |
+
"eval_loss": 1.5877399614879064,
|
| 103 |
+
"flow/cos_sim": 0.8912065430744879,
|
| 104 |
+
"flow/improvement_ratio": 0.9926800377079165,
|
| 105 |
+
"flow/mag_ratio_mean": 0.8669626872930953,
|
| 106 |
+
"flow/mag_ratio_std": 0.15283272580614984,
|
| 107 |
+
"step": 3072
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.13239381989786023,
|
| 111 |
+
"eval_bleu": 0.20682060276074551,
|
| 112 |
+
"eval_ce_clean_loss": 0.7854123648041601,
|
| 113 |
+
"eval_ce_pred_loss": 5.334758354910909,
|
| 114 |
+
"eval_flow_latent_mse_loss": 0.040881452071212375,
|
| 115 |
+
"eval_flow_velocity_mse_loss": 0.2279702990548189,
|
| 116 |
+
"eval_loss": 1.5877399614879064,
|
| 117 |
+
"eval_runtime": 205.7686,
|
| 118 |
+
"eval_samples_per_second": 145.795,
|
| 119 |
+
"eval_steps_per_second": 2.279,
|
| 120 |
+
"flow/cos_sim": 0.8912065430744879,
|
| 121 |
+
"flow/improvement_ratio": 0.9926800377079165,
|
| 122 |
+
"flow/mag_ratio_mean": 0.8669626872930953,
|
| 123 |
+
"flow/mag_ratio_std": 0.15283272580614984,
|
| 124 |
+
"step": 3072
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.176525093197147,
|
| 128 |
+
"grad_norm": 11.012340545654297,
|
| 129 |
+
"learning_rate": 9.534991440649608e-05,
|
| 130 |
+
"loss": 1.3743938207626343,
|
| 131 |
+
"step": 4096
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.176525093197147,
|
| 135 |
+
"eval_bleu": 0.16363188503894788,
|
| 136 |
+
"eval_ce_clean_loss": 0.3785287282868489,
|
| 137 |
+
"eval_ce_pred_loss": 5.528324539981671,
|
| 138 |
+
"eval_flow_latent_mse_loss": 0.03347295840411806,
|
| 139 |
+
"eval_flow_velocity_mse_loss": 0.16226537120558304,
|
| 140 |
+
"eval_loss": 1.1270995217599848,
|
| 141 |
+
"flow/cos_sim": 0.9286127779275369,
|
| 142 |
+
"flow/improvement_ratio": 0.9932153086418282,
|
| 143 |
+
"flow/mag_ratio_mean": 0.9108342735497936,
|
| 144 |
+
"flow/mag_ratio_std": 0.15295166635055785,
|
| 145 |
+
"step": 4096
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.176525093197147,
|
| 149 |
+
"eval_bleu": 0.16363188503894788,
|
| 150 |
+
"eval_ce_clean_loss": 0.3785287282868489,
|
| 151 |
+
"eval_ce_pred_loss": 5.528324539981671,
|
| 152 |
+
"eval_flow_latent_mse_loss": 0.03347295840411806,
|
| 153 |
+
"eval_flow_velocity_mse_loss": 0.16226537120558304,
|
| 154 |
+
"eval_loss": 1.1270995217599848,
|
| 155 |
+
"eval_runtime": 204.563,
|
| 156 |
+
"eval_samples_per_second": 146.654,
|
| 157 |
+
"eval_steps_per_second": 2.293,
|
| 158 |
+
"flow/cos_sim": 0.9286127779275369,
|
| 159 |
+
"flow/improvement_ratio": 0.9932153086418282,
|
| 160 |
+
"flow/mag_ratio_mean": 0.9108342735497936,
|
| 161 |
+
"flow/mag_ratio_std": 0.15295166635055785,
|
| 162 |
+
"step": 4096
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.22065636649643372,
|
| 166 |
+
"grad_norm": 11.745284080505371,
|
| 167 |
+
"learning_rate": 9.183037205346935e-05,
|
| 168 |
+
"loss": 1.0636177062988281,
|
| 169 |
+
"step": 5120
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.22065636649643372,
|
| 173 |
+
"eval_bleu": 0.13738491075539824,
|
| 174 |
+
"eval_ce_clean_loss": 0.1879858059733153,
|
| 175 |
+
"eval_ce_pred_loss": 5.7349623474739255,
|
| 176 |
+
"eval_flow_latent_mse_loss": 0.02759745604654492,
|
| 177 |
+
"eval_flow_velocity_mse_loss": 0.14143581146688095,
|
| 178 |
+
"eval_loss": 0.9305153198079514,
|
| 179 |
+
"flow/cos_sim": 0.941998773038006,
|
| 180 |
+
"flow/improvement_ratio": 0.9937093091417731,
|
| 181 |
+
"flow/mag_ratio_mean": 0.9254698810546891,
|
| 182 |
+
"flow/mag_ratio_std": 0.14320005217531342,
|
| 183 |
+
"step": 5120
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.22065636649643372,
|
| 187 |
+
"eval_bleu": 0.13738491075539824,
|
| 188 |
+
"eval_ce_clean_loss": 0.1879858059733153,
|
| 189 |
+
"eval_ce_pred_loss": 5.7349623474739255,
|
| 190 |
+
"eval_flow_latent_mse_loss": 0.02759745604654492,
|
| 191 |
+
"eval_flow_velocity_mse_loss": 0.14143581146688095,
|
| 192 |
+
"eval_loss": 0.9305153198079514,
|
| 193 |
+
"eval_runtime": 202.9046,
|
| 194 |
+
"eval_samples_per_second": 147.853,
|
| 195 |
+
"eval_steps_per_second": 2.311,
|
| 196 |
+
"flow/cos_sim": 0.941998773038006,
|
| 197 |
+
"flow/improvement_ratio": 0.9937093091417731,
|
| 198 |
+
"flow/mag_ratio_mean": 0.9254698810546891,
|
| 199 |
+
"flow/mag_ratio_std": 0.14320005217531342,
|
| 200 |
+
"step": 5120
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.26478763979572045,
|
| 204 |
+
"grad_norm": 9.8949556350708,
|
| 205 |
+
"learning_rate": 8.74324003722993e-05,
|
| 206 |
+
"loss": 0.9332119226455688,
|
| 207 |
+
"step": 6144
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.26478763979572045,
|
| 211 |
+
"eval_bleu": 0.10485989517125303,
|
| 212 |
+
"eval_ce_clean_loss": 0.10314439222820278,
|
| 213 |
+
"eval_ce_pred_loss": 6.026585411161248,
|
| 214 |
+
"eval_flow_latent_mse_loss": 0.02462343667059946,
|
| 215 |
+
"eval_flow_velocity_mse_loss": 0.12707700184794632,
|
| 216 |
+
"eval_loss": 0.8575033856861627,
|
| 217 |
+
"flow/cos_sim": 0.9490489955904133,
|
| 218 |
+
"flow/improvement_ratio": 0.9939835194839852,
|
| 219 |
+
"flow/mag_ratio_mean": 0.9330783666832361,
|
| 220 |
+
"flow/mag_ratio_std": 0.1343777290126408,
|
| 221 |
+
"step": 6144
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.26478763979572045,
|
| 225 |
+
"eval_bleu": 0.10485989517125303,
|
| 226 |
+
"eval_ce_clean_loss": 0.10314439222820278,
|
| 227 |
+
"eval_ce_pred_loss": 6.026585411161248,
|
| 228 |
+
"eval_flow_latent_mse_loss": 0.02462343667059946,
|
| 229 |
+
"eval_flow_velocity_mse_loss": 0.12707700184794632,
|
| 230 |
+
"eval_loss": 0.8575033856861627,
|
| 231 |
+
"eval_runtime": 201.985,
|
| 232 |
+
"eval_samples_per_second": 148.526,
|
| 233 |
+
"eval_steps_per_second": 2.322,
|
| 234 |
+
"flow/cos_sim": 0.9490489955904133,
|
| 235 |
+
"flow/improvement_ratio": 0.9939835194839852,
|
| 236 |
+
"flow/mag_ratio_mean": 0.9330783666832361,
|
| 237 |
+
"flow/mag_ratio_std": 0.1343777290126408,
|
| 238 |
+
"step": 6144
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.30891891309500724,
|
| 242 |
+
"grad_norm": 8.900321960449219,
|
| 243 |
+
"learning_rate": 8.22483558761947e-05,
|
| 244 |
+
"loss": 0.8709867000579834,
|
| 245 |
+
"step": 7168
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.30891891309500724,
|
| 249 |
+
"eval_bleu": 0.08651694305130357,
|
| 250 |
+
"eval_ce_clean_loss": 0.0600688684104221,
|
| 251 |
+
"eval_ce_pred_loss": 6.218749198832238,
|
| 252 |
+
"eval_flow_latent_mse_loss": 0.02104399689356052,
|
| 253 |
+
"eval_flow_velocity_mse_loss": 0.11542461483653929,
|
| 254 |
+
"eval_loss": 0.8184124124583914,
|
| 255 |
+
"flow/cos_sim": 0.9532103971885973,
|
| 256 |
+
"flow/improvement_ratio": 0.9942541531662443,
|
| 257 |
+
"flow/mag_ratio_mean": 0.9374735507883751,
|
| 258 |
+
"flow/mag_ratio_std": 0.1255933871306082,
|
| 259 |
+
"step": 7168
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.30891891309500724,
|
| 263 |
+
"eval_bleu": 0.08651694305130357,
|
| 264 |
+
"eval_ce_clean_loss": 0.0600688684104221,
|
| 265 |
+
"eval_ce_pred_loss": 6.218749198832238,
|
| 266 |
+
"eval_flow_latent_mse_loss": 0.02104399689356052,
|
| 267 |
+
"eval_flow_velocity_mse_loss": 0.11542461483653929,
|
| 268 |
+
"eval_loss": 0.8184124124583914,
|
| 269 |
+
"eval_runtime": 201.994,
|
| 270 |
+
"eval_samples_per_second": 148.519,
|
| 271 |
+
"eval_steps_per_second": 2.322,
|
| 272 |
+
"flow/cos_sim": 0.9532103971885973,
|
| 273 |
+
"flow/improvement_ratio": 0.9942541531662443,
|
| 274 |
+
"flow/mag_ratio_mean": 0.9374735507883751,
|
| 275 |
+
"flow/mag_ratio_std": 0.1255933871306082,
|
| 276 |
+
"step": 7168
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.353050186394294,
|
| 280 |
+
"grad_norm": 6.8332929611206055,
|
| 281 |
+
"learning_rate": 7.639311770076283e-05,
|
| 282 |
+
"loss": 0.8521633744239807,
|
| 283 |
+
"step": 8192
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.353050186394294,
|
| 287 |
+
"eval_bleu": 0.07066854474381863,
|
| 288 |
+
"eval_ce_clean_loss": 0.03517881183942625,
|
| 289 |
+
"eval_ce_pred_loss": 6.361157763995595,
|
| 290 |
+
"eval_flow_latent_mse_loss": 0.017622319229844727,
|
| 291 |
+
"eval_flow_velocity_mse_loss": 0.10679940540971024,
|
| 292 |
+
"eval_loss": 0.7957163247218264,
|
| 293 |
+
"flow/cos_sim": 0.9565253009928315,
|
| 294 |
+
"flow/improvement_ratio": 0.9950958989830668,
|
| 295 |
+
"flow/mag_ratio_mean": 0.9426390433362298,
|
| 296 |
+
"flow/mag_ratio_std": 0.11702151938097309,
|
| 297 |
+
"step": 8192
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.353050186394294,
|
| 301 |
+
"eval_bleu": 0.07066854474381863,
|
| 302 |
+
"eval_ce_clean_loss": 0.03517881183942625,
|
| 303 |
+
"eval_ce_pred_loss": 6.361157763995595,
|
| 304 |
+
"eval_flow_latent_mse_loss": 0.017622319229844727,
|
| 305 |
+
"eval_flow_velocity_mse_loss": 0.10679940540971024,
|
| 306 |
+
"eval_loss": 0.7957163247218264,
|
| 307 |
+
"eval_runtime": 201.4435,
|
| 308 |
+
"eval_samples_per_second": 148.925,
|
| 309 |
+
"eval_steps_per_second": 2.328,
|
| 310 |
+
"flow/cos_sim": 0.9565253009928315,
|
| 311 |
+
"flow/improvement_ratio": 0.9950958989830668,
|
| 312 |
+
"flow/mag_ratio_mean": 0.9426390433362298,
|
| 313 |
+
"flow/mag_ratio_std": 0.11702151938097309,
|
| 314 |
+
"step": 8192
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.3971814596935807,
|
| 318 |
+
"grad_norm": 6.2701029777526855,
|
| 319 |
+
"learning_rate": 6.997821756319211e-05,
|
| 320 |
+
"loss": 0.8328163623809814,
|
| 321 |
+
"step": 9216
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.3971814596935807,
|
| 325 |
+
"eval_bleu": 0.059063499314682966,
|
| 326 |
+
"eval_ce_clean_loss": 0.02335210119102047,
|
| 327 |
+
"eval_ce_pred_loss": 6.471723645989067,
|
| 328 |
+
"eval_flow_latent_mse_loss": 0.016289353997174547,
|
| 329 |
+
"eval_flow_velocity_mse_loss": 0.10239806742683402,
|
| 330 |
+
"eval_loss": 0.789211899614029,
|
| 331 |
+
"flow/cos_sim": 0.9578453517163487,
|
| 332 |
+
"flow/improvement_ratio": 0.9947574119578039,
|
| 333 |
+
"flow/mag_ratio_mean": 0.9445628934323407,
|
| 334 |
+
"flow/mag_ratio_std": 0.11213315162323177,
|
| 335 |
+
"step": 9216
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.3971814596935807,
|
| 339 |
+
"eval_bleu": 0.059063499314682966,
|
| 340 |
+
"eval_ce_clean_loss": 0.02335210119102047,
|
| 341 |
+
"eval_ce_pred_loss": 6.471723645989067,
|
| 342 |
+
"eval_flow_latent_mse_loss": 0.016289353997174547,
|
| 343 |
+
"eval_flow_velocity_mse_loss": 0.10239806742683402,
|
| 344 |
+
"eval_loss": 0.789211899614029,
|
| 345 |
+
"eval_runtime": 200.3456,
|
| 346 |
+
"eval_samples_per_second": 149.741,
|
| 347 |
+
"eval_steps_per_second": 2.341,
|
| 348 |
+
"flow/cos_sim": 0.9578453517163487,
|
| 349 |
+
"flow/improvement_ratio": 0.9947574119578039,
|
| 350 |
+
"flow/mag_ratio_mean": 0.9445628934323407,
|
| 351 |
+
"flow/mag_ratio_std": 0.11213315162323177,
|
| 352 |
+
"step": 9216
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.44131273299286744,
|
| 356 |
+
"grad_norm": 6.149109363555908,
|
| 357 |
+
"learning_rate": 6.314377890922702e-05,
|
| 358 |
+
"loss": 0.8299998044967651,
|
| 359 |
+
"step": 10240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.44131273299286744,
|
| 363 |
+
"eval_bleu": 0.04646173709857315,
|
| 364 |
+
"eval_ce_clean_loss": 0.016398947276889897,
|
| 365 |
+
"eval_ce_pred_loss": 6.611309008811837,
|
| 366 |
+
"eval_flow_latent_mse_loss": 0.01444387674006,
|
| 367 |
+
"eval_flow_velocity_mse_loss": 0.09797465415206799,
|
| 368 |
+
"eval_loss": 0.7899483920160387,
|
| 369 |
+
"flow/cos_sim": 0.9596936283335249,
|
| 370 |
+
"flow/improvement_ratio": 0.9951455176257884,
|
| 371 |
+
"flow/mag_ratio_mean": 0.9418040758002796,
|
| 372 |
+
"flow/mag_ratio_std": 0.10655801122122482,
|
| 373 |
+
"step": 10240
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.44131273299286744,
|
| 377 |
+
"eval_bleu": 0.04646173709857315,
|
| 378 |
+
"eval_ce_clean_loss": 0.016398947276889897,
|
| 379 |
+
"eval_ce_pred_loss": 6.611309008811837,
|
| 380 |
+
"eval_flow_latent_mse_loss": 0.01444387674006,
|
| 381 |
+
"eval_flow_velocity_mse_loss": 0.09797465415206799,
|
| 382 |
+
"eval_loss": 0.7899483920160387,
|
| 383 |
+
"eval_runtime": 199.7885,
|
| 384 |
+
"eval_samples_per_second": 150.159,
|
| 385 |
+
"eval_steps_per_second": 2.347,
|
| 386 |
+
"flow/cos_sim": 0.9596936283335249,
|
| 387 |
+
"flow/improvement_ratio": 0.9951455176257884,
|
| 388 |
+
"flow/mag_ratio_mean": 0.9418040758002796,
|
| 389 |
+
"flow/mag_ratio_std": 0.10655801122122482,
|
| 390 |
+
"step": 10240
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.4854440062921542,
|
| 394 |
+
"grad_norm": 3.603848695755005,
|
| 395 |
+
"learning_rate": 5.604035379537632e-05,
|
| 396 |
+
"loss": 0.8279342651367188,
|
| 397 |
+
"step": 11264
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.4854440062921542,
|
| 401 |
+
"eval_bleu": 0.042570233416081196,
|
| 402 |
+
"eval_ce_clean_loss": 0.011810618673147423,
|
| 403 |
+
"eval_ce_pred_loss": 6.648376730205154,
|
| 404 |
+
"eval_flow_latent_mse_loss": 0.012516131068565953,
|
| 405 |
+
"eval_flow_velocity_mse_loss": 0.09491073330645876,
|
| 406 |
+
"eval_loss": 0.7840751668791781,
|
| 407 |
+
"flow/cos_sim": 0.9602375737131278,
|
| 408 |
+
"flow/improvement_ratio": 0.9958161113104587,
|
| 409 |
+
"flow/mag_ratio_mean": 0.9454934674539546,
|
| 410 |
+
"flow/mag_ratio_std": 0.10616235840104536,
|
| 411 |
+
"step": 11264
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.4854440062921542,
|
| 415 |
+
"eval_bleu": 0.042570233416081196,
|
| 416 |
+
"eval_ce_clean_loss": 0.011810618673147423,
|
| 417 |
+
"eval_ce_pred_loss": 6.648376730205154,
|
| 418 |
+
"eval_flow_latent_mse_loss": 0.012516131068565953,
|
| 419 |
+
"eval_flow_velocity_mse_loss": 0.09491073330645876,
|
| 420 |
+
"eval_loss": 0.7840751668791781,
|
| 421 |
+
"eval_runtime": 200.0632,
|
| 422 |
+
"eval_samples_per_second": 149.953,
|
| 423 |
+
"eval_steps_per_second": 2.344,
|
| 424 |
+
"flow/cos_sim": 0.9602375737131278,
|
| 425 |
+
"flow/improvement_ratio": 0.9958161113104587,
|
| 426 |
+
"flow/mag_ratio_mean": 0.9454934674539546,
|
| 427 |
+
"flow/mag_ratio_std": 0.10616235840104536,
|
| 428 |
+
"step": 11264
|
| 429 |
+
}
|
| 430 |
+
],
|
| 431 |
+
"logging_steps": 1024,
|
| 432 |
+
"max_steps": 23204,
|
| 433 |
+
"num_input_tokens_seen": 0,
|
| 434 |
+
"num_train_epochs": 1,
|
| 435 |
+
"save_steps": 1024,
|
| 436 |
+
"stateful_callbacks": {
|
| 437 |
+
"TrainerControl": {
|
| 438 |
+
"args": {
|
| 439 |
+
"should_epoch_stop": false,
|
| 440 |
+
"should_evaluate": false,
|
| 441 |
+
"should_log": false,
|
| 442 |
+
"should_save": true,
|
| 443 |
+
"should_training_stop": false
|
| 444 |
+
},
|
| 445 |
+
"attributes": {}
|
| 446 |
+
}
|
| 447 |
+
},
|
| 448 |
+
"total_flos": 0.0,
|
| 449 |
+
"train_batch_size": 64,
|
| 450 |
+
"trial_name": null,
|
| 451 |
+
"trial_params": null
|
| 452 |
+
}
|
checkpoints-v5.15/checkpoint-11264/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8469bbc81a2ba0be2f5b44007faafd15c75615abe30f4f4e56171816d31caa5b
|
| 3 |
+
size 5137
|