Model save
Browse files- README.md +2 -6
- all_results.json +5 -10
- config.json +1 -1
- model.safetensors +1 -1
- train_results.json +5 -5
- trainer_state.json +610 -610
- training_args.bin +2 -2
README.md
CHANGED
|
@@ -1,13 +1,9 @@
|
|
| 1 |
---
|
| 2 |
base_model: HuggingFaceTB/SmolVLM2-2.2B-Instruct
|
| 3 |
-
datasets: smolagents/aguvis-stage-2
|
| 4 |
library_name: transformers
|
| 5 |
model_name: SmolVLM2-2.2B-Instruct-Agentic-GUI
|
| 6 |
tags:
|
| 7 |
- generated_from_trainer
|
| 8 |
-
- open-r1
|
| 9 |
-
- vision-language
|
| 10 |
-
- vlm
|
| 11 |
- trl
|
| 12 |
- sft
|
| 13 |
licence: license
|
|
@@ -15,7 +11,7 @@ licence: license
|
|
| 15 |
|
| 16 |
# Model Card for SmolVLM2-2.2B-Instruct-Agentic-GUI
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [HuggingFaceTB/SmolVLM2-2.2B-Instruct](https://huggingface.co/HuggingFaceTB/SmolVLM2-2.2B-Instruct)
|
| 19 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 20 |
|
| 21 |
## Quick start
|
|
@@ -31,7 +27,7 @@ print(output["generated_text"])
|
|
| 31 |
|
| 32 |
## Training procedure
|
| 33 |
|
| 34 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/huggingface/smolagents/runs/
|
| 35 |
|
| 36 |
|
| 37 |
This model was trained with SFT.
|
|
|
|
| 1 |
---
|
| 2 |
base_model: HuggingFaceTB/SmolVLM2-2.2B-Instruct
|
|
|
|
| 3 |
library_name: transformers
|
| 4 |
model_name: SmolVLM2-2.2B-Instruct-Agentic-GUI
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
|
| 7 |
- trl
|
| 8 |
- sft
|
| 9 |
licence: license
|
|
|
|
| 11 |
|
| 12 |
# Model Card for SmolVLM2-2.2B-Instruct-Agentic-GUI
|
| 13 |
|
| 14 |
+
This model is a fine-tuned version of [HuggingFaceTB/SmolVLM2-2.2B-Instruct](https://huggingface.co/HuggingFaceTB/SmolVLM2-2.2B-Instruct).
|
| 15 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
|
| 17 |
## Quick start
|
|
|
|
| 27 |
|
| 28 |
## Training procedure
|
| 29 |
|
| 30 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/huggingface/smolagents/runs/p3hj27rk)
|
| 31 |
|
| 32 |
|
| 33 |
This model was trained with SFT.
|
all_results.json
CHANGED
|
@@ -1,13 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"eval_samples_per_second": 24.509,
|
| 6 |
-
"eval_steps_per_second": 0.768,
|
| 7 |
-
"total_flos": 6033817814958080.0,
|
| 8 |
-
"train_loss": 0.6046972061422738,
|
| 9 |
-
"train_runtime": 27367.7567,
|
| 10 |
"train_samples": 360252,
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"total_flos": 6035679775555584.0,
|
| 3 |
+
"train_loss": 0.6108071732927453,
|
| 4 |
+
"train_runtime": 30814.6397,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"train_samples": 360252,
|
| 6 |
+
"train_samples_per_second": 11.691,
|
| 7 |
+
"train_steps_per_second": 0.023
|
| 8 |
}
|
config.json
CHANGED
|
@@ -119,7 +119,7 @@
|
|
| 119 |
"tie_word_embeddings": false,
|
| 120 |
"torch_dtype": "bfloat16",
|
| 121 |
"transformers_version": "4.52.3",
|
| 122 |
-
"use_cache":
|
| 123 |
"use_reentrant_checkpointing": false,
|
| 124 |
"vision_config": {
|
| 125 |
"attention_dropout": 0.0,
|
|
|
|
| 119 |
"tie_word_embeddings": false,
|
| 120 |
"torch_dtype": "bfloat16",
|
| 121 |
"transformers_version": "4.52.3",
|
| 122 |
+
"use_cache": false,
|
| 123 |
"use_reentrant_checkpointing": false,
|
| 124 |
"vision_config": {
|
| 125 |
"attention_dropout": 0.0,
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4493654912
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a465552e2210eb810bf823961356cd0eaa88951d6da6e736e10050175f5a494
|
| 3 |
size 4493654912
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"total_flos":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 360252,
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"total_flos": 6035679775555584.0,
|
| 3 |
+
"train_loss": 0.6108071732927453,
|
| 4 |
+
"train_runtime": 30814.6397,
|
| 5 |
"train_samples": 360252,
|
| 6 |
+
"train_samples_per_second": 11.691,
|
| 7 |
+
"train_steps_per_second": 0.023
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -11,1344 +11,1344 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.007106057914372002,
|
| 14 |
-
"grad_norm":
|
| 15 |
"learning_rate": 3.6363636363636366e-06,
|
| 16 |
-
"loss": 4.
|
| 17 |
-
"mean_token_accuracy": 0.
|
| 18 |
-
"num_tokens":
|
| 19 |
"step": 5
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 0.014212115828744005,
|
| 23 |
-
"grad_norm": 14.
|
| 24 |
"learning_rate": 8.181818181818183e-06,
|
| 25 |
-
"loss": 3.
|
| 26 |
-
"mean_token_accuracy": 0.
|
| 27 |
-
"num_tokens":
|
| 28 |
"step": 10
|
| 29 |
},
|
| 30 |
{
|
| 31 |
"epoch": 0.021318173743116006,
|
| 32 |
-
"grad_norm":
|
| 33 |
"learning_rate": 1.2727272727272728e-05,
|
| 34 |
-
"loss": 1.
|
| 35 |
-
"mean_token_accuracy": 0.
|
| 36 |
-
"num_tokens":
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02842423165748801,
|
| 41 |
-
"grad_norm": 2.
|
| 42 |
"learning_rate": 1.7272727272727274e-05,
|
| 43 |
-
"loss": 1.
|
| 44 |
-
"mean_token_accuracy": 0.
|
| 45 |
-
"num_tokens":
|
| 46 |
"step": 20
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.03553028957186001,
|
| 50 |
-
"grad_norm": 3.
|
| 51 |
"learning_rate": 1.999961805535155e-05,
|
| 52 |
-
"loss": 0.
|
| 53 |
-
"mean_token_accuracy": 0.
|
| 54 |
-
"num_tokens":
|
| 55 |
"step": 25
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 0.04263634748623201,
|
| 59 |
-
"grad_norm": 3.
|
| 60 |
"learning_rate": 1.9995321550350065e-05,
|
| 61 |
-
"loss": 0.
|
| 62 |
-
"mean_token_accuracy": 0.
|
| 63 |
-
"num_tokens":
|
| 64 |
"step": 30
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"epoch": 0.04974240540060401,
|
| 68 |
-
"grad_norm": 2.
|
| 69 |
"learning_rate": 1.998625339625423e-05,
|
| 70 |
-
"loss": 0.
|
| 71 |
-
"mean_token_accuracy": 0.
|
| 72 |
-
"num_tokens":
|
| 73 |
"step": 35
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.05684846331497602,
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
"learning_rate": 1.9972418403347817e-05,
|
| 79 |
-
"loss": 0.
|
| 80 |
-
"mean_token_accuracy": 0.
|
| 81 |
-
"num_tokens":
|
| 82 |
"step": 40
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 0.06395452122934801,
|
| 86 |
-
"grad_norm": 0.
|
| 87 |
"learning_rate": 1.9953823910527057e-05,
|
| 88 |
-
"loss": 0.
|
| 89 |
-
"mean_token_accuracy": 0.
|
| 90 |
-
"num_tokens":
|
| 91 |
"step": 45
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 0.07106057914372002,
|
| 95 |
-
"grad_norm": 0.
|
| 96 |
"learning_rate": 1.993047978140764e-05,
|
| 97 |
-
"loss": 0.
|
| 98 |
-
"mean_token_accuracy": 0.
|
| 99 |
-
"num_tokens":
|
| 100 |
"step": 50
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.07816663705809203,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
"learning_rate": 1.9902398399092494e-05,
|
| 106 |
-
"loss": 0.
|
| 107 |
-
"mean_token_accuracy": 0.
|
| 108 |
-
"num_tokens":
|
| 109 |
"step": 55
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 0.08527269497246402,
|
| 113 |
-
"grad_norm": 0.
|
| 114 |
"learning_rate": 1.9869594659603032e-05,
|
| 115 |
-
"loss": 0.
|
| 116 |
-
"mean_token_accuracy": 0.
|
| 117 |
-
"num_tokens":
|
| 118 |
"step": 60
|
| 119 |
},
|
| 120 |
{
|
| 121 |
"epoch": 0.09237875288683603,
|
| 122 |
-
"grad_norm": 0.
|
| 123 |
"learning_rate": 1.9832085963977445e-05,
|
| 124 |
-
"loss": 0.
|
| 125 |
-
"mean_token_accuracy": 0.
|
| 126 |
-
"num_tokens":
|
| 127 |
"step": 65
|
| 128 |
},
|
| 129 |
{
|
| 130 |
"epoch": 0.09948481080120802,
|
| 131 |
-
"grad_norm": 0.
|
| 132 |
"learning_rate": 1.978989220904016e-05,
|
| 133 |
-
"loss": 0.
|
| 134 |
-
"mean_token_accuracy": 0.
|
| 135 |
-
"num_tokens":
|
| 136 |
"step": 70
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.10659086871558003,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
"learning_rate": 1.9743035776847377e-05,
|
| 142 |
-
"loss": 0.
|
| 143 |
-
"mean_token_accuracy": 0.
|
| 144 |
-
"num_tokens":
|
| 145 |
"step": 75
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 0.11369692662995204,
|
| 149 |
-
"grad_norm": 0.
|
| 150 |
"learning_rate": 1.9691541522814327e-05,
|
| 151 |
-
"loss": 0.
|
| 152 |
-
"mean_token_accuracy": 0.
|
| 153 |
-
"num_tokens":
|
| 154 |
"step": 80
|
| 155 |
},
|
| 156 |
{
|
| 157 |
"epoch": 0.12080298454432403,
|
| 158 |
-
"grad_norm": 0.
|
| 159 |
"learning_rate": 1.963543676253048e-05,
|
| 160 |
-
"loss": 0.
|
| 161 |
-
"mean_token_accuracy": 0.
|
| 162 |
-
"num_tokens":
|
| 163 |
"step": 85
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.12790904245869603,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
"learning_rate": 1.9574751257269748e-05,
|
| 169 |
-
"loss": 0.
|
| 170 |
-
"mean_token_accuracy": 0.
|
| 171 |
-
"num_tokens":
|
| 172 |
"step": 90
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.13501510037306805,
|
| 176 |
-
"grad_norm": 0.
|
| 177 |
"learning_rate": 1.950951719820335e-05,
|
| 178 |
-
"loss": 0.
|
| 179 |
-
"mean_token_accuracy": 0.
|
| 180 |
-
"num_tokens":
|
| 181 |
"step": 95
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 0.14212115828744004,
|
| 185 |
-
"grad_norm": 0.
|
| 186 |
"learning_rate": 1.9439769189323727e-05,
|
| 187 |
-
"loss": 0.
|
| 188 |
-
"mean_token_accuracy": 0.
|
| 189 |
-
"num_tokens":
|
| 190 |
"step": 100
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 0.14212115828744004,
|
| 194 |
-
"eval_loss": 0.
|
| 195 |
-
"eval_mean_token_accuracy": 0.
|
| 196 |
-
"eval_num_tokens":
|
| 197 |
-
"eval_runtime":
|
| 198 |
-
"eval_samples_per_second":
|
| 199 |
-
"eval_steps_per_second": 0.
|
| 200 |
"step": 100
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.14922721620181204,
|
| 204 |
-
"grad_norm": 0.
|
| 205 |
"learning_rate": 1.9365544229088517e-05,
|
| 206 |
-
"loss": 0.
|
| 207 |
-
"mean_token_accuracy": 0.
|
| 208 |
-
"num_tokens":
|
| 209 |
"step": 105
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 0.15633327411618406,
|
| 213 |
-
"grad_norm": 0.
|
| 214 |
"learning_rate": 1.9286881690794425e-05,
|
| 215 |
-
"loss": 0.
|
| 216 |
-
"mean_token_accuracy": 0.
|
| 217 |
-
"num_tokens":
|
| 218 |
"step": 110
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.16343933203055605,
|
| 222 |
-
"grad_norm": 0.
|
| 223 |
"learning_rate": 1.9203823301691272e-05,
|
| 224 |
-
"loss": 0.
|
| 225 |
-
"mean_token_accuracy": 0.
|
| 226 |
-
"num_tokens":
|
| 227 |
"step": 115
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.17054538994492804,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
"learning_rate": 1.9116413120847425e-05,
|
| 233 |
-
"loss": 0.
|
| 234 |
-
"mean_token_accuracy": 0.
|
| 235 |
-
"num_tokens":
|
| 236 |
"step": 120
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.17765144785930007,
|
| 240 |
-
"grad_norm": 0.
|
| 241 |
"learning_rate": 1.902469751577826e-05,
|
| 242 |
-
"loss": 0.
|
| 243 |
-
"mean_token_accuracy": 0.
|
| 244 |
-
"num_tokens":
|
| 245 |
"step": 125
|
| 246 |
},
|
| 247 |
{
|
| 248 |
"epoch": 0.18475750577367206,
|
| 249 |
-
"grad_norm": 0.
|
| 250 |
"learning_rate": 1.892872513785008e-05,
|
| 251 |
-
"loss": 0.
|
| 252 |
-
"mean_token_accuracy": 0.
|
| 253 |
-
"num_tokens":
|
| 254 |
"step": 130
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.19186356368804405,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 1.88285468964726e-05,
|
| 260 |
-
"loss": 0.
|
| 261 |
-
"mean_token_accuracy": 0.
|
| 262 |
-
"num_tokens":
|
| 263 |
"step": 135
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.19896962160241605,
|
| 267 |
-
"grad_norm": 0.
|
| 268 |
"learning_rate": 1.872421593209355e-05,
|
| 269 |
-
"loss": 0.
|
| 270 |
-
"mean_token_accuracy": 0.
|
| 271 |
-
"num_tokens":
|
| 272 |
"step": 140
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 0.20607567951678807,
|
| 276 |
-
"grad_norm": 0.
|
| 277 |
"learning_rate": 1.861578758800989e-05,
|
| 278 |
-
"loss": 0.
|
| 279 |
-
"mean_token_accuracy": 0.
|
| 280 |
-
"num_tokens":
|
| 281 |
"step": 145
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 0.21318173743116006,
|
| 285 |
-
"grad_norm": 0.
|
| 286 |
"learning_rate": 1.8503319381010414e-05,
|
| 287 |
-
"loss": 0.
|
| 288 |
-
"mean_token_accuracy": 0.
|
| 289 |
-
"num_tokens":
|
| 290 |
"step": 150
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.22028779534553206,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
"learning_rate": 1.8386870970865488e-05,
|
| 296 |
-
"loss": 0.
|
| 297 |
-
"mean_token_accuracy": 0.
|
| 298 |
-
"num_tokens":
|
| 299 |
"step": 155
|
| 300 |
},
|
| 301 |
{
|
| 302 |
"epoch": 0.22739385325990408,
|
| 303 |
-
"grad_norm": 0.
|
| 304 |
"learning_rate": 1.8266504128679988e-05,
|
| 305 |
-
"loss": 0.
|
| 306 |
-
"mean_token_accuracy": 0.
|
| 307 |
-
"num_tokens":
|
| 308 |
"step": 160
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.23449991117427607,
|
| 312 |
-
"grad_norm": 0.
|
| 313 |
"learning_rate": 1.814228270412624e-05,
|
| 314 |
-
"loss": 0.
|
| 315 |
-
"mean_token_accuracy": 0.
|
| 316 |
-
"num_tokens":
|
| 317 |
"step": 165
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.24160596908864806,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
"learning_rate": 1.8014272591574405e-05,
|
| 323 |
-
"loss": 0.
|
| 324 |
-
"mean_token_accuracy": 0.
|
| 325 |
-
"num_tokens":
|
| 326 |
"step": 170
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.2487120270030201,
|
| 330 |
-
"grad_norm": 0.
|
| 331 |
"learning_rate": 1.7882541695138224e-05,
|
| 332 |
-
"loss": 0.
|
| 333 |
-
"mean_token_accuracy": 0.
|
| 334 |
-
"num_tokens":
|
| 335 |
"step": 175
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 0.25581808491739205,
|
| 339 |
-
"grad_norm": 0.
|
| 340 |
"learning_rate": 1.7747159892654646e-05,
|
| 341 |
-
"loss": 0.
|
| 342 |
-
"mean_token_accuracy": 0.
|
| 343 |
-
"num_tokens":
|
| 344 |
"step": 180
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 0.2629241428317641,
|
| 348 |
-
"grad_norm": 0.
|
| 349 |
"learning_rate": 1.7608198998616533e-05,
|
| 350 |
-
"loss": 0.
|
| 351 |
-
"mean_token_accuracy": 0.
|
| 352 |
-
"num_tokens":
|
| 353 |
"step": 185
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.2700302007461361,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
"learning_rate": 1.7465732726077993e-05,
|
| 359 |
-
"loss": 0.
|
| 360 |
-
"mean_token_accuracy": 0.
|
| 361 |
-
"num_tokens":
|
| 362 |
"step": 190
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.27713625866050806,
|
| 366 |
-
"grad_norm": 0.
|
| 367 |
"learning_rate": 1.731983664755264e-05,
|
| 368 |
-
"loss": 0.
|
| 369 |
-
"mean_token_accuracy": 0.
|
| 370 |
-
"num_tokens":
|
| 371 |
"step": 195
|
| 372 |
},
|
| 373 |
{
|
| 374 |
"epoch": 0.2842423165748801,
|
| 375 |
-
"grad_norm": 0.
|
| 376 |
"learning_rate": 1.717058815492548e-05,
|
| 377 |
-
"loss": 0.
|
| 378 |
-
"mean_token_accuracy": 0.
|
| 379 |
-
"num_tokens":
|
| 380 |
"step": 200
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.2842423165748801,
|
| 384 |
-
"eval_loss": 0.
|
| 385 |
-
"eval_mean_token_accuracy": 0.
|
| 386 |
-
"eval_num_tokens":
|
| 387 |
-
"eval_runtime":
|
| 388 |
-
"eval_samples_per_second":
|
| 389 |
-
"eval_steps_per_second": 0.
|
| 390 |
"step": 200
|
| 391 |
},
|
| 392 |
{
|
| 393 |
"epoch": 0.2913483744892521,
|
| 394 |
-
"grad_norm": 0.
|
| 395 |
"learning_rate": 1.701806641839967e-05,
|
| 396 |
-
"loss": 0.
|
| 397 |
-
"mean_token_accuracy": 0.
|
| 398 |
-
"num_tokens":
|
| 399 |
"step": 205
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 0.29845443240362407,
|
| 403 |
-
"grad_norm": 0.
|
| 404 |
"learning_rate": 1.6862352344500004e-05,
|
| 405 |
-
"loss": 0.
|
| 406 |
-
"mean_token_accuracy": 0.
|
| 407 |
-
"num_tokens":
|
| 408 |
"step": 210
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.3055604903179961,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
"learning_rate": 1.6703528533155283e-05,
|
| 414 |
-
"loss": 0.
|
| 415 |
-
"mean_token_accuracy": 0.
|
| 416 |
-
"num_tokens":
|
| 417 |
"step": 215
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.3126665482323681,
|
| 421 |
-
"grad_norm": 0.
|
| 422 |
"learning_rate": 1.6541679233882477e-05,
|
| 423 |
-
"loss": 0.
|
| 424 |
-
"mean_token_accuracy": 0.
|
| 425 |
-
"num_tokens":
|
| 426 |
"step": 220
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 0.3197726061467401,
|
| 430 |
-
"grad_norm": 0.
|
| 431 |
"learning_rate": 1.63768903010958e-05,
|
| 432 |
-
"loss": 0.
|
| 433 |
-
"mean_token_accuracy": 0.
|
| 434 |
-
"num_tokens":
|
| 435 |
"step": 225
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 0.3268786640611121,
|
| 439 |
-
"grad_norm": 0.
|
| 440 |
"learning_rate": 1.6209249148564437e-05,
|
| 441 |
-
"loss": 0.
|
| 442 |
-
"mean_token_accuracy": 0.
|
| 443 |
-
"num_tokens":
|
| 444 |
"step": 230
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.3339847219754841,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
"learning_rate": 1.603884470304318e-05,
|
| 450 |
-
"loss": 0.
|
| 451 |
-
"mean_token_accuracy": 0.
|
| 452 |
-
"num_tokens":
|
| 453 |
"step": 235
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 0.3410907798898561,
|
| 457 |
-
"grad_norm": 0.
|
| 458 |
"learning_rate": 1.5865767357100383e-05,
|
| 459 |
-
"loss": 0.
|
| 460 |
-
"mean_token_accuracy": 0.
|
| 461 |
-
"num_tokens":
|
| 462 |
"step": 240
|
| 463 |
},
|
| 464 |
{
|
| 465 |
"epoch": 0.3481968378042281,
|
| 466 |
-
"grad_norm": 0.
|
| 467 |
"learning_rate": 1.5690108921168428e-05,
|
| 468 |
-
"loss": 0.
|
| 469 |
-
"mean_token_accuracy": 0.
|
| 470 |
-
"num_tokens":
|
| 471 |
"step": 245
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.35530289571860013,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
"learning_rate": 1.5511962574842073e-05,
|
| 477 |
-
"loss": 0.
|
| 478 |
-
"mean_token_accuracy": 0.
|
| 479 |
-
"num_tokens":
|
| 480 |
"step": 250
|
| 481 |
},
|
| 482 |
{
|
| 483 |
"epoch": 0.3624089536329721,
|
| 484 |
-
"grad_norm": 0.
|
| 485 |
"learning_rate": 1.5331422817450485e-05,
|
| 486 |
-
"loss": 0.
|
| 487 |
-
"mean_token_accuracy": 0.
|
| 488 |
-
"num_tokens":
|
| 489 |
"step": 255
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.3695150115473441,
|
| 493 |
-
"grad_norm": 0.
|
| 494 |
"learning_rate": 1.5148585417929212e-05,
|
| 495 |
-
"loss": 0.
|
| 496 |
-
"mean_token_accuracy": 0.
|
| 497 |
-
"num_tokens":
|
| 498 |
"step": 260
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 0.37662106946171614,
|
| 502 |
-
"grad_norm": 0.
|
| 503 |
"learning_rate": 1.4963547364018711e-05,
|
| 504 |
-
"loss": 0.
|
| 505 |
-
"mean_token_accuracy": 0.
|
| 506 |
-
"num_tokens":
|
| 507 |
"step": 265
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.3837271273760881,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
"learning_rate": 1.477640681081632e-05,
|
| 513 |
-
"loss": 0.
|
| 514 |
-
"mean_token_accuracy": 0.
|
| 515 |
-
"num_tokens":
|
| 516 |
"step": 270
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 0.39083318529046013,
|
| 520 |
-
"grad_norm": 0.
|
| 521 |
"learning_rate": 1.4587263028709013e-05,
|
| 522 |
-
"loss": 0.
|
| 523 |
-
"mean_token_accuracy": 0.
|
| 524 |
-
"num_tokens":
|
| 525 |
"step": 275
|
| 526 |
},
|
| 527 |
{
|
| 528 |
"epoch": 0.3979392432048321,
|
| 529 |
-
"grad_norm": 0.
|
| 530 |
"learning_rate": 1.4396216350714512e-05,
|
| 531 |
-
"loss": 0.
|
| 532 |
-
"mean_token_accuracy": 0.
|
| 533 |
-
"num_tokens":
|
| 534 |
"step": 280
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.4050453011192041,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
"learning_rate": 1.4203368119258759e-05,
|
| 540 |
-
"loss": 0.
|
| 541 |
-
"mean_token_accuracy": 0.
|
| 542 |
-
"num_tokens":
|
| 543 |
"step": 285
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 0.41215135903357614,
|
| 547 |
-
"grad_norm": 0.
|
| 548 |
"learning_rate": 1.4008820632417906e-05,
|
| 549 |
-
"loss": 0.
|
| 550 |
-
"mean_token_accuracy": 0.
|
| 551 |
-
"num_tokens":
|
| 552 |
"step": 290
|
| 553 |
},
|
| 554 |
{
|
| 555 |
"epoch": 0.4192574169479481,
|
| 556 |
-
"grad_norm": 0.
|
| 557 |
"learning_rate": 1.381267708965339e-05,
|
| 558 |
-
"loss": 0.
|
| 559 |
-
"mean_token_accuracy": 0.
|
| 560 |
-
"num_tokens":
|
| 561 |
"step": 295
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 0.4263634748623201,
|
| 565 |
-
"grad_norm": 0.
|
| 566 |
"learning_rate": 1.3615041537068831e-05,
|
| 567 |
-
"loss": 0.
|
| 568 |
-
"mean_token_accuracy": 0.
|
| 569 |
-
"num_tokens":
|
| 570 |
"step": 300
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.4263634748623201,
|
| 574 |
-
"eval_loss": 0.
|
| 575 |
-
"eval_mean_token_accuracy": 0.
|
| 576 |
-
"eval_num_tokens":
|
| 577 |
-
"eval_runtime":
|
| 578 |
-
"eval_samples_per_second":
|
| 579 |
-
"eval_steps_per_second": 0.
|
| 580 |
"step": 300
|
| 581 |
},
|
| 582 |
{
|
| 583 |
"epoch": 0.43346953277669215,
|
| 584 |
-
"grad_norm": 0.
|
| 585 |
"learning_rate": 1.3416018812217866e-05,
|
| 586 |
-
"loss": 0.
|
| 587 |
-
"mean_token_accuracy": 0.
|
| 588 |
-
"num_tokens":
|
| 589 |
"step": 305
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 0.4405755906910641,
|
| 593 |
-
"grad_norm": 0.
|
| 594 |
"learning_rate": 1.3215714488492121e-05,
|
| 595 |
-
"loss": 0.
|
| 596 |
-
"mean_token_accuracy": 0.
|
| 597 |
-
"num_tokens":
|
| 598 |
"step": 310
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.44768164860543613,
|
| 602 |
-
"grad_norm": 0.
|
| 603 |
"learning_rate": 1.3014234819118846e-05,
|
| 604 |
-
"loss": 0.
|
| 605 |
-
"mean_token_accuracy": 0.
|
| 606 |
-
"num_tokens":
|
| 607 |
"step": 315
|
| 608 |
},
|
| 609 |
{
|
| 610 |
"epoch": 0.45478770651980815,
|
| 611 |
-
"grad_norm": 0.
|
| 612 |
"learning_rate": 1.2811686680797942e-05,
|
| 613 |
-
"loss": 0.
|
| 614 |
-
"mean_token_accuracy": 0.
|
| 615 |
-
"num_tokens":
|
| 616 |
"step": 320
|
| 617 |
},
|
| 618 |
{
|
| 619 |
"epoch": 0.4618937644341801,
|
| 620 |
-
"grad_norm": 0.
|
| 621 |
"learning_rate": 1.2608177517008268e-05,
|
| 622 |
-
"loss": 0.
|
| 623 |
-
"mean_token_accuracy": 0.
|
| 624 |
-
"num_tokens":
|
| 625 |
"step": 325
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.46899982234855214,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
"learning_rate": 1.240381528101327e-05,
|
| 631 |
-
"loss": 0.
|
| 632 |
-
"mean_token_accuracy": 0.
|
| 633 |
-
"num_tokens":
|
| 634 |
"step": 330
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 0.47610588026292416,
|
| 638 |
-
"grad_norm": 0.
|
| 639 |
"learning_rate": 1.2198708378596198e-05,
|
| 640 |
-
"loss": 0.
|
| 641 |
-
"mean_token_accuracy": 0.
|
| 642 |
-
"num_tokens":
|
| 643 |
"step": 335
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 0.48321193817729613,
|
| 647 |
-
"grad_norm": 0.
|
| 648 |
"learning_rate": 1.19929656105553e-05,
|
| 649 |
-
"loss": 0.
|
| 650 |
-
"mean_token_accuracy": 0.
|
| 651 |
-
"num_tokens":
|
| 652 |
"step": 340
|
| 653 |
},
|
| 654 |
{
|
| 655 |
"epoch": 0.49031799609166815,
|
| 656 |
-
"grad_norm": 0.
|
| 657 |
"learning_rate": 1.1786696114989455e-05,
|
| 658 |
-
"loss": 0.
|
| 659 |
-
"mean_token_accuracy": 0.
|
| 660 |
-
"num_tokens":
|
| 661 |
"step": 345
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.4974240540060402,
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
"learning_rate": 1.1580009309404887e-05,
|
| 667 |
-
"loss": 0.
|
| 668 |
-
"mean_token_accuracy": 0.
|
| 669 |
-
"num_tokens":
|
| 670 |
"step": 350
|
| 671 |
},
|
| 672 |
{
|
| 673 |
"epoch": 0.5045301119204122,
|
| 674 |
-
"grad_norm": 0.
|
| 675 |
"learning_rate": 1.1373014832673661e-05,
|
| 676 |
-
"loss": 0.
|
| 677 |
-
"mean_token_accuracy": 0.
|
| 678 |
-
"num_tokens":
|
| 679 |
"step": 355
|
| 680 |
},
|
| 681 |
{
|
| 682 |
"epoch": 0.5116361698347841,
|
| 683 |
-
"grad_norm": 0.
|
| 684 |
"learning_rate": 1.1165822486874773e-05,
|
| 685 |
-
"loss": 0.
|
| 686 |
-
"mean_token_accuracy": 0.
|
| 687 |
-
"num_tokens":
|
| 688 |
"step": 360
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.5187422277491561,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
"learning_rate": 1.0958542179048637e-05,
|
| 694 |
-
"loss": 0.
|
| 695 |
-
"mean_token_accuracy": 0.
|
| 696 |
-
"num_tokens":
|
| 697 |
"step": 365
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.5258482856635281,
|
| 701 |
-
"grad_norm": 0.
|
| 702 |
"learning_rate": 1.0751283862895914e-05,
|
| 703 |
-
"loss": 0.
|
| 704 |
-
"mean_token_accuracy": 0.
|
| 705 |
-
"num_tokens":
|
| 706 |
"step": 370
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 0.5329543435779002,
|
| 710 |
-
"grad_norm": 0.
|
| 711 |
"learning_rate": 1.0544157480451586e-05,
|
| 712 |
-
"loss": 0.
|
| 713 |
-
"mean_token_accuracy": 0.
|
| 714 |
-
"num_tokens":
|
| 715 |
"step": 375
|
| 716 |
},
|
| 717 |
{
|
| 718 |
"epoch": 0.5400604014922722,
|
| 719 |
-
"grad_norm": 0.
|
| 720 |
"learning_rate": 1.033727290376522e-05,
|
| 721 |
-
"loss": 0.
|
| 722 |
-
"mean_token_accuracy": 0.
|
| 723 |
-
"num_tokens":
|
| 724 |
"step": 380
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.5471664594066442,
|
| 728 |
-
"grad_norm": 0.
|
| 729 |
"learning_rate": 1.013073987661834e-05,
|
| 730 |
-
"loss": 0.
|
| 731 |
-
"mean_token_accuracy": 0.
|
| 732 |
-
"num_tokens":
|
| 733 |
"step": 385
|
| 734 |
},
|
| 735 |
{
|
| 736 |
"epoch": 0.5542725173210161,
|
| 737 |
-
"grad_norm": 0.
|
| 738 |
"learning_rate": 9.924667956309862e-06,
|
| 739 |
-
"loss": 0.
|
| 740 |
-
"mean_token_accuracy": 0.
|
| 741 |
-
"num_tokens":
|
| 742 |
"step": 390
|
| 743 |
},
|
| 744 |
{
|
| 745 |
"epoch": 0.5613785752353881,
|
| 746 |
-
"grad_norm": 0.
|
| 747 |
"learning_rate": 9.719166455540437e-06,
|
| 748 |
-
"loss": 0.
|
| 749 |
-
"mean_token_accuracy": 0.
|
| 750 |
-
"num_tokens":
|
| 751 |
"step": 395
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.5684846331497602,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
"learning_rate": 9.51434438442655e-06,
|
| 757 |
-
"loss": 0.
|
| 758 |
-
"mean_token_accuracy": 0.
|
| 759 |
-
"num_tokens":
|
| 760 |
"step": 400
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 0.5684846331497602,
|
| 764 |
-
"eval_loss": 0.
|
| 765 |
-
"eval_mean_token_accuracy": 0.
|
| 766 |
-
"eval_num_tokens":
|
| 767 |
-
"eval_runtime":
|
| 768 |
-
"eval_samples_per_second":
|
| 769 |
-
"eval_steps_per_second": 0.
|
| 770 |
"step": 400
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.5755906910641322,
|
| 774 |
-
"grad_norm": 0.
|
| 775 |
"learning_rate": 9.310310392675132e-06,
|
| 776 |
-
"loss": 0.
|
| 777 |
-
"mean_token_accuracy": 0.
|
| 778 |
-
"num_tokens":
|
| 779 |
"step": 405
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.5826967489785042,
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
"learning_rate": 9.107172711949324e-06,
|
| 785 |
-
"loss": 0.
|
| 786 |
-
"mean_token_accuracy": 0.
|
| 787 |
-
"num_tokens":
|
| 788 |
"step": 410
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 0.5898028068928762,
|
| 792 |
-
"grad_norm": 0.
|
| 793 |
"learning_rate": 8.905039098456049e-06,
|
| 794 |
-
"loss": 0.
|
| 795 |
-
"mean_token_accuracy": 0.
|
| 796 |
-
"num_tokens":
|
| 797 |
"step": 415
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 0.5969088648072481,
|
| 801 |
-
"grad_norm": 0.
|
| 802 |
"learning_rate": 8.704016775785742e-06,
|
| 803 |
-
"loss": 0.
|
| 804 |
-
"mean_token_accuracy": 0.
|
| 805 |
-
"num_tokens":
|
| 806 |
"step": 420
|
| 807 |
},
|
| 808 |
{
|
| 809 |
"epoch": 0.6040149227216202,
|
| 810 |
-
"grad_norm": 0.
|
| 811 |
"learning_rate": 8.50421237803464e-06,
|
| 812 |
-
"loss": 0.
|
| 813 |
-
"mean_token_accuracy": 0.
|
| 814 |
-
"num_tokens":
|
| 815 |
"step": 425
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.6111209806359922,
|
| 819 |
-
"grad_norm": 0.
|
| 820 |
"learning_rate": 8.30573189323978e-06,
|
| 821 |
-
"loss": 0.
|
| 822 |
-
"mean_token_accuracy": 0.
|
| 823 |
-
"num_tokens":
|
| 824 |
"step": 430
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 0.6182270385503642,
|
| 828 |
-
"grad_norm": 0.
|
| 829 |
"learning_rate": 8.108680607156669e-06,
|
| 830 |
-
"loss": 0.
|
| 831 |
-
"mean_token_accuracy": 0.
|
| 832 |
-
"num_tokens":
|
| 833 |
"step": 435
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 0.6253330964647362,
|
| 837 |
-
"grad_norm": 0.
|
| 838 |
"learning_rate": 7.913163047409533e-06,
|
| 839 |
-
"loss": 0.
|
| 840 |
-
"mean_token_accuracy": 0.
|
| 841 |
-
"num_tokens":
|
| 842 |
"step": 440
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.6324391543791081,
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
"learning_rate": 7.719282928043688e-06,
|
| 848 |
-
"loss": 0.
|
| 849 |
-
"mean_token_accuracy": 0.
|
| 850 |
-
"num_tokens":
|
| 851 |
"step": 445
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 0.6395452122934802,
|
| 855 |
-
"grad_norm": 0.
|
| 856 |
"learning_rate": 7.527143094509492e-06,
|
| 857 |
-
"loss": 0.
|
| 858 |
-
"mean_token_accuracy": 0.
|
| 859 |
-
"num_tokens":
|
| 860 |
"step": 450
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 0.6466512702078522,
|
| 864 |
-
"grad_norm": 0.
|
| 865 |
"learning_rate": 7.336845469107061e-06,
|
| 866 |
-
"loss": 0.
|
| 867 |
-
"mean_token_accuracy": 0.
|
| 868 |
-
"num_tokens":
|
| 869 |
"step": 455
|
| 870 |
},
|
| 871 |
{
|
| 872 |
"epoch": 0.6537573281222242,
|
| 873 |
-
"grad_norm": 0.
|
| 874 |
"learning_rate": 7.148490996920661e-06,
|
| 875 |
-
"loss": 0.
|
| 876 |
-
"mean_token_accuracy": 0.
|
| 877 |
-
"num_tokens":
|
| 878 |
"step": 460
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.6608633860365962,
|
| 882 |
-
"grad_norm": 0.
|
| 883 |
"learning_rate": 6.9621795922714805e-06,
|
| 884 |
-
"loss": 0.
|
| 885 |
-
"mean_token_accuracy": 0.
|
| 886 |
-
"num_tokens":
|
| 887 |
"step": 465
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 0.6679694439509682,
|
| 891 |
-
"grad_norm": 0.
|
| 892 |
"learning_rate": 6.778010085717202e-06,
|
| 893 |
-
"loss": 0.
|
| 894 |
-
"mean_token_accuracy": 0.
|
| 895 |
-
"num_tokens":
|
| 896 |
"step": 470
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.6750755018653402,
|
| 900 |
-
"grad_norm": 0.
|
| 901 |
"learning_rate": 6.596080171626409e-06,
|
| 902 |
-
"loss": 0.
|
| 903 |
-
"mean_token_accuracy": 0.
|
| 904 |
-
"num_tokens":
|
| 905 |
"step": 475
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.6821815597797122,
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
"learning_rate": 6.416486356355769e-06,
|
| 911 |
-
"loss": 0.
|
| 912 |
-
"mean_token_accuracy": 0.
|
| 913 |
-
"num_tokens":
|
| 914 |
"step": 480
|
| 915 |
},
|
| 916 |
{
|
| 917 |
"epoch": 0.6892876176940842,
|
| 918 |
-
"grad_norm": 0.
|
| 919 |
"learning_rate": 6.239323907057342e-06,
|
| 920 |
-
"loss": 0.
|
| 921 |
-
"mean_token_accuracy": 0.
|
| 922 |
-
"num_tokens":
|
| 923 |
"step": 485
|
| 924 |
},
|
| 925 |
{
|
| 926 |
"epoch": 0.6963936756084562,
|
| 927 |
-
"grad_norm": 0.
|
| 928 |
"learning_rate": 6.064686801143271e-06,
|
| 929 |
-
"loss": 0.
|
| 930 |
-
"mean_token_accuracy": 0.
|
| 931 |
-
"num_tokens":
|
| 932 |
"step": 490
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 0.7034997335228282,
|
| 936 |
-
"grad_norm": 0.
|
| 937 |
"learning_rate": 5.892667676434633e-06,
|
| 938 |
-
"loss": 0.
|
| 939 |
-
"mean_token_accuracy": 0.
|
| 940 |
-
"num_tokens":
|
| 941 |
"step": 495
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.7106057914372003,
|
| 945 |
-
"grad_norm": 0.
|
| 946 |
"learning_rate": 5.723357782020867e-06,
|
| 947 |
-
"loss": 0.
|
| 948 |
-
"mean_token_accuracy": 0.
|
| 949 |
-
"num_tokens":
|
| 950 |
"step": 500
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 0.7106057914372003,
|
| 954 |
-
"eval_loss": 0.
|
| 955 |
-
"eval_mean_token_accuracy": 0.
|
| 956 |
-
"eval_num_tokens":
|
| 957 |
-
"eval_runtime":
|
| 958 |
-
"eval_samples_per_second":
|
| 959 |
-
"eval_steps_per_second": 0.
|
| 960 |
"step": 500
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 0.7177118493515722,
|
| 964 |
-
"grad_norm": 0.
|
| 965 |
"learning_rate": 5.556846929855857e-06,
|
| 966 |
-
"loss": 0.
|
| 967 |
-
"mean_token_accuracy": 0.
|
| 968 |
-
"num_tokens":
|
| 969 |
"step": 505
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.7248179072659442,
|
| 973 |
-
"grad_norm": 0.
|
| 974 |
"learning_rate": 5.393223447116409e-06,
|
| 975 |
-
"loss": 0.
|
| 976 |
-
"mean_token_accuracy": 0.
|
| 977 |
-
"num_tokens":
|
| 978 |
"step": 510
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 0.7319239651803162,
|
| 982 |
-
"grad_norm": 0.
|
| 983 |
"learning_rate": 5.232574129348278e-06,
|
| 984 |
-
"loss": 0.
|
| 985 |
-
"mean_token_accuracy": 0.
|
| 986 |
-
"num_tokens":
|
| 987 |
"step": 515
|
| 988 |
},
|
| 989 |
{
|
| 990 |
"epoch": 0.7390300230946882,
|
| 991 |
-
"grad_norm": 0.
|
| 992 |
"learning_rate": 5.0749841944247e-06,
|
| 993 |
-
"loss": 0.
|
| 994 |
-
"mean_token_accuracy": 0.
|
| 995 |
-
"num_tokens":
|
| 996 |
"step": 520
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.7461360810090603,
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
"learning_rate": 4.92053723734182e-06,
|
| 1002 |
-
"loss": 0.
|
| 1003 |
-
"mean_token_accuracy": 0.
|
| 1004 |
-
"num_tokens":
|
| 1005 |
"step": 525
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
"epoch": 0.7532421389234323,
|
| 1009 |
-
"grad_norm": 0.
|
| 1010 |
"learning_rate": 4.769315185874951e-06,
|
| 1011 |
-
"loss": 0.
|
| 1012 |
-
"mean_token_accuracy": 0.
|
| 1013 |
-
"num_tokens":
|
| 1014 |
"step": 530
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
"epoch": 0.7603481968378042,
|
| 1018 |
-
"grad_norm": 0.
|
| 1019 |
"learning_rate": 4.621398257119266e-06,
|
| 1020 |
-
"loss": 0.
|
| 1021 |
-
"mean_token_accuracy": 0.
|
| 1022 |
-
"num_tokens":
|
| 1023 |
"step": 535
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 0.7674542547521762,
|
| 1027 |
-
"grad_norm": 0.
|
| 1028 |
"learning_rate": 4.476864914937923e-06,
|
| 1029 |
-
"loss": 0.
|
| 1030 |
-
"mean_token_accuracy": 0.
|
| 1031 |
-
"num_tokens":
|
| 1032 |
"step": 540
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 0.7745603126665482,
|
| 1036 |
-
"grad_norm": 0.
|
| 1037 |
"learning_rate": 4.335791828340183e-06,
|
| 1038 |
-
"loss": 0.
|
| 1039 |
-
"mean_token_accuracy": 0.
|
| 1040 |
-
"num_tokens":
|
| 1041 |
"step": 545
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 0.7816663705809203,
|
| 1045 |
-
"grad_norm": 0.
|
| 1046 |
"learning_rate": 4.1982538308116775e-06,
|
| 1047 |
-
"loss": 0.
|
| 1048 |
-
"mean_token_accuracy": 0.
|
| 1049 |
-
"num_tokens":
|
| 1050 |
"step": 550
|
| 1051 |
},
|
| 1052 |
{
|
| 1053 |
"epoch": 0.7887724284952923,
|
| 1054 |
-
"grad_norm": 0.
|
| 1055 |
"learning_rate": 4.064323880618279e-06,
|
| 1056 |
-
"loss": 0.
|
| 1057 |
-
"mean_token_accuracy": 0.
|
| 1058 |
-
"num_tokens":
|
| 1059 |
"step": 555
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.7958784864096642,
|
| 1063 |
-
"grad_norm": 0.
|
| 1064 |
"learning_rate": 3.934073022104759e-06,
|
| 1065 |
-
"loss": 0.
|
| 1066 |
-
"mean_token_accuracy": 0.
|
| 1067 |
-
"num_tokens":
|
| 1068 |
"step": 560
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
"epoch": 0.8029845443240362,
|
| 1072 |
-
"grad_norm": 0.
|
| 1073 |
"learning_rate": 3.807570348008672e-06,
|
| 1074 |
-
"loss": 0.
|
| 1075 |
-
"mean_token_accuracy": 0.
|
| 1076 |
-
"num_tokens":
|
| 1077 |
"step": 565
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
"epoch": 0.8100906022384082,
|
| 1081 |
-
"grad_norm": 0.
|
| 1082 |
"learning_rate": 3.684882962809484e-06,
|
| 1083 |
-
"loss": 0.
|
| 1084 |
-
"mean_token_accuracy": 0.
|
| 1085 |
-
"num_tokens":
|
| 1086 |
"step": 570
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 0.8171966601527803,
|
| 1090 |
-
"grad_norm": 0.
|
| 1091 |
"learning_rate": 3.5660759471324037e-06,
|
| 1092 |
-
"loss": 0.
|
| 1093 |
-
"mean_token_accuracy": 0.
|
| 1094 |
-
"num_tokens":
|
| 1095 |
"step": 575
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 0.8243027180671523,
|
| 1099 |
-
"grad_norm": 0.
|
| 1100 |
"learning_rate": 3.451212323225786e-06,
|
| 1101 |
-
"loss": 0.
|
| 1102 |
-
"mean_token_accuracy": 0.
|
| 1103 |
-
"num_tokens":
|
| 1104 |
"step": 580
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 0.8314087759815243,
|
| 1108 |
-
"grad_norm": 0.
|
| 1109 |
"learning_rate": 3.340353021530409e-06,
|
| 1110 |
-
"loss": 0.
|
| 1111 |
-
"mean_token_accuracy": 0.
|
| 1112 |
-
"num_tokens":
|
| 1113 |
"step": 585
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"epoch": 0.8385148338958962,
|
| 1117 |
-
"grad_norm": 0.
|
| 1118 |
"learning_rate": 3.2335568483583708e-06,
|
| 1119 |
-
"loss": 0.
|
| 1120 |
-
"mean_token_accuracy": 0.
|
| 1121 |
-
"num_tokens":
|
| 1122 |
"step": 590
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.8456208918102682,
|
| 1126 |
-
"grad_norm": 0.
|
| 1127 |
"learning_rate": 3.1308804546987615e-06,
|
| 1128 |
-
"loss": 0.
|
| 1129 |
-
"mean_token_accuracy": 0.
|
| 1130 |
-
"num_tokens":
|
| 1131 |
"step": 595
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
"epoch": 0.8527269497246402,
|
| 1135 |
-
"grad_norm": 0.
|
| 1136 |
"learning_rate": 3.0323783061666307e-06,
|
| 1137 |
-
"loss": 0.
|
| 1138 |
-
"mean_token_accuracy": 0.
|
| 1139 |
-
"num_tokens":
|
| 1140 |
"step": 600
|
| 1141 |
},
|
| 1142 |
{
|
| 1143 |
"epoch": 0.8527269497246402,
|
| 1144 |
-
"eval_loss": 0.
|
| 1145 |
-
"eval_mean_token_accuracy": 0.
|
| 1146 |
-
"eval_num_tokens":
|
| 1147 |
-
"eval_runtime":
|
| 1148 |
-
"eval_samples_per_second":
|
| 1149 |
-
"eval_steps_per_second": 0.
|
| 1150 |
"step": 600
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.8598330076390123,
|
| 1154 |
-
"grad_norm": 0.
|
| 1155 |
"learning_rate": 2.9381026541112145e-06,
|
| 1156 |
-
"loss": 0.
|
| 1157 |
-
"mean_token_accuracy": 0.
|
| 1158 |
-
"num_tokens":
|
| 1159 |
"step": 605
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
"epoch": 0.8669390655533843,
|
| 1163 |
-
"grad_norm": 0.
|
| 1164 |
"learning_rate": 2.848103507898745e-06,
|
| 1165 |
-
"loss": 0.
|
| 1166 |
-
"mean_token_accuracy": 0.
|
| 1167 |
-
"num_tokens":
|
| 1168 |
"step": 610
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 0.8740451234677563,
|
| 1172 |
-
"grad_norm": 0.
|
| 1173 |
"learning_rate": 2.7624286083845187e-06,
|
| 1174 |
-
"loss": 0.
|
| 1175 |
-
"mean_token_accuracy": 0.
|
| 1176 |
-
"num_tokens":
|
| 1177 |
"step": 615
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"epoch": 0.8811511813821282,
|
| 1181 |
-
"grad_norm": 0.
|
| 1182 |
"learning_rate": 2.6811234025883457e-06,
|
| 1183 |
-
"loss": 0.
|
| 1184 |
-
"mean_token_accuracy": 0.
|
| 1185 |
-
"num_tokens":
|
| 1186 |
"step": 620
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 0.8882572392965002,
|
| 1190 |
-
"grad_norm": 0.
|
| 1191 |
"learning_rate": 2.604231019586761e-06,
|
| 1192 |
-
"loss": 0.
|
| 1193 |
-
"mean_token_accuracy": 0.
|
| 1194 |
-
"num_tokens":
|
| 1195 |
"step": 625
|
| 1196 |
},
|
| 1197 |
{
|
| 1198 |
"epoch": 0.8953632972108723,
|
| 1199 |
-
"grad_norm": 0.
|
| 1200 |
"learning_rate": 2.5317922476348194e-06,
|
| 1201 |
-
"loss": 0.
|
| 1202 |
-
"mean_token_accuracy": 0.
|
| 1203 |
-
"num_tokens":
|
| 1204 |
"step": 630
|
| 1205 |
},
|
| 1206 |
{
|
| 1207 |
"epoch": 0.9024693551252443,
|
| 1208 |
-
"grad_norm": 0.
|
| 1209 |
"learning_rate": 2.4638455125296043e-06,
|
| 1210 |
-
"loss": 0.
|
| 1211 |
-
"mean_token_accuracy": 0.
|
| 1212 |
-
"num_tokens":
|
| 1213 |
"step": 635
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.9095754130396163,
|
| 1217 |
-
"grad_norm": 0.
|
| 1218 |
"learning_rate": 2.400426857226914e-06,
|
| 1219 |
-
"loss": 0.
|
| 1220 |
-
"mean_token_accuracy": 0.
|
| 1221 |
-
"num_tokens":
|
| 1222 |
"step": 640
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
"epoch": 0.9166814709539882,
|
| 1226 |
-
"grad_norm": 0.
|
| 1227 |
"learning_rate": 2.3415699227219517e-06,
|
| 1228 |
-
"loss": 0.
|
| 1229 |
-
"mean_token_accuracy": 0.
|
| 1230 |
-
"num_tokens":
|
| 1231 |
"step": 645
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 0.9237875288683602,
|
| 1235 |
-
"grad_norm": 0.
|
| 1236 |
"learning_rate": 2.2873059302041627e-06,
|
| 1237 |
-
"loss": 0.
|
| 1238 |
-
"mean_token_accuracy": 0.
|
| 1239 |
-
"num_tokens":
|
| 1240 |
"step": 650
|
| 1241 |
},
|
| 1242 |
{
|
| 1243 |
"epoch": 0.9308935867827323,
|
| 1244 |
-
"grad_norm": 0.
|
| 1245 |
"learning_rate": 2.2376636644956656e-06,
|
| 1246 |
-
"loss": 0.
|
| 1247 |
-
"mean_token_accuracy": 0.
|
| 1248 |
-
"num_tokens":
|
| 1249 |
"step": 655
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.9379996446971043,
|
| 1253 |
-
"grad_norm": 0.
|
| 1254 |
"learning_rate": 2.192669458782096e-06,
|
| 1255 |
-
"loss": 0.
|
| 1256 |
-
"mean_token_accuracy": 0.
|
| 1257 |
-
"num_tokens":
|
| 1258 |
"step": 660
|
| 1259 |
},
|
| 1260 |
{
|
| 1261 |
"epoch": 0.9451057026114763,
|
| 1262 |
-
"grad_norm": 0.
|
| 1263 |
"learning_rate": 2.1523471806439205e-06,
|
| 1264 |
-
"loss": 0.
|
| 1265 |
-
"mean_token_accuracy": 0.
|
| 1266 |
-
"num_tokens":
|
| 1267 |
"step": 665
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
"epoch": 0.9522117605258483,
|
| 1271 |
-
"grad_norm": 0.
|
| 1272 |
"learning_rate": 2.1167182193956738e-06,
|
| 1273 |
-
"loss": 0.
|
| 1274 |
-
"mean_token_accuracy": 0.
|
| 1275 |
-
"num_tokens":
|
| 1276 |
"step": 670
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
"epoch": 0.9593178184402202,
|
| 1280 |
-
"grad_norm": 0.
|
| 1281 |
"learning_rate": 2.0858014747397952e-06,
|
| 1282 |
-
"loss": 0.
|
| 1283 |
-
"mean_token_accuracy": 0.
|
| 1284 |
-
"num_tokens":
|
| 1285 |
"step": 675
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 0.9664238763545923,
|
| 1289 |
-
"grad_norm": 0.
|
| 1290 |
"learning_rate": 2.0596133467411213e-06,
|
| 1291 |
-
"loss": 0.
|
| 1292 |
-
"mean_token_accuracy": 0.
|
| 1293 |
-
"num_tokens":
|
| 1294 |
"step": 680
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 0.9735299342689643,
|
| 1298 |
-
"grad_norm": 0.
|
| 1299 |
"learning_rate": 2.0381677271273177e-06,
|
| 1300 |
-
"loss": 0.
|
| 1301 |
-
"mean_token_accuracy": 0.
|
| 1302 |
-
"num_tokens":
|
| 1303 |
"step": 685
|
| 1304 |
},
|
| 1305 |
{
|
| 1306 |
"epoch": 0.9806359921833363,
|
| 1307 |
-
"grad_norm": 0.
|
| 1308 |
"learning_rate": 2.0214759919198904e-06,
|
| 1309 |
-
"loss": 0.
|
| 1310 |
-
"mean_token_accuracy": 0.
|
| 1311 |
-
"num_tokens":
|
| 1312 |
"step": 690
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.9877420500977083,
|
| 1316 |
-
"grad_norm": 0.
|
| 1317 |
"learning_rate": 2.0095469953996724e-06,
|
| 1318 |
-
"loss": 0.
|
| 1319 |
-
"mean_token_accuracy": 0.
|
| 1320 |
-
"num_tokens":
|
| 1321 |
"step": 695
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"epoch": 0.9948481080120803,
|
| 1325 |
-
"grad_norm": 0.
|
| 1326 |
"learning_rate": 2.002387065409989e-06,
|
| 1327 |
-
"loss": 0.
|
| 1328 |
-
"mean_token_accuracy": 0.
|
| 1329 |
-
"num_tokens":
|
| 1330 |
"step": 700
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 0.9948481080120803,
|
| 1334 |
-
"eval_loss": 0.
|
| 1335 |
-
"eval_mean_token_accuracy": 0.
|
| 1336 |
-
"eval_num_tokens":
|
| 1337 |
-
"eval_runtime":
|
| 1338 |
-
"eval_samples_per_second":
|
| 1339 |
-
"eval_steps_per_second": 0.
|
| 1340 |
"step": 700
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 1.0,
|
| 1344 |
-
"mean_token_accuracy": 0.
|
| 1345 |
-
"num_tokens":
|
| 1346 |
"step": 704,
|
| 1347 |
-
"total_flos":
|
| 1348 |
-
"train_loss": 0.
|
| 1349 |
-
"train_runtime":
|
| 1350 |
-
"train_samples_per_second":
|
| 1351 |
-
"train_steps_per_second": 0.
|
| 1352 |
}
|
| 1353 |
],
|
| 1354 |
"logging_steps": 5,
|
|
@@ -1368,7 +1368,7 @@
|
|
| 1368 |
"attributes": {}
|
| 1369 |
}
|
| 1370 |
},
|
| 1371 |
-
"total_flos":
|
| 1372 |
"train_batch_size": 4,
|
| 1373 |
"trial_name": null,
|
| 1374 |
"trial_params": null
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.007106057914372002,
|
| 14 |
+
"grad_norm": 48.05719398745313,
|
| 15 |
"learning_rate": 3.6363636363636366e-06,
|
| 16 |
+
"loss": 4.3873,
|
| 17 |
+
"mean_token_accuracy": 0.44012689143419265,
|
| 18 |
+
"num_tokens": 5475610.0,
|
| 19 |
"step": 5
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 0.014212115828744005,
|
| 23 |
+
"grad_norm": 14.175302664795455,
|
| 24 |
"learning_rate": 8.181818181818183e-06,
|
| 25 |
+
"loss": 3.1519,
|
| 26 |
+
"mean_token_accuracy": 0.4974055245518684,
|
| 27 |
+
"num_tokens": 10990469.0,
|
| 28 |
"step": 10
|
| 29 |
},
|
| 30 |
{
|
| 31 |
"epoch": 0.021318173743116006,
|
| 32 |
+
"grad_norm": 4.9875540402165965,
|
| 33 |
"learning_rate": 1.2727272727272728e-05,
|
| 34 |
+
"loss": 1.6073,
|
| 35 |
+
"mean_token_accuracy": 0.6541870839893817,
|
| 36 |
+
"num_tokens": 16509954.0,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02842423165748801,
|
| 41 |
+
"grad_norm": 2.7892792489854905,
|
| 42 |
"learning_rate": 1.7272727272727274e-05,
|
| 43 |
+
"loss": 1.054,
|
| 44 |
+
"mean_token_accuracy": 0.7446420036256314,
|
| 45 |
+
"num_tokens": 22025438.0,
|
| 46 |
"step": 20
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.03553028957186001,
|
| 50 |
+
"grad_norm": 3.458274666427455,
|
| 51 |
"learning_rate": 1.999961805535155e-05,
|
| 52 |
+
"loss": 0.8975,
|
| 53 |
+
"mean_token_accuracy": 0.7689281158149243,
|
| 54 |
+
"num_tokens": 27537030.0,
|
| 55 |
"step": 25
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 0.04263634748623201,
|
| 59 |
+
"grad_norm": 3.116160123738211,
|
| 60 |
"learning_rate": 1.9995321550350065e-05,
|
| 61 |
+
"loss": 0.8077,
|
| 62 |
+
"mean_token_accuracy": 0.7813956893980503,
|
| 63 |
+
"num_tokens": 33069854.0,
|
| 64 |
"step": 30
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"epoch": 0.04974240540060401,
|
| 68 |
+
"grad_norm": 2.495268086238599,
|
| 69 |
"learning_rate": 1.998625339625423e-05,
|
| 70 |
+
"loss": 0.772,
|
| 71 |
+
"mean_token_accuracy": 0.7832800410687923,
|
| 72 |
+
"num_tokens": 38592423.0,
|
| 73 |
"step": 35
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.05684846331497602,
|
| 77 |
+
"grad_norm": 0.7719664512735186,
|
| 78 |
"learning_rate": 1.9972418403347817e-05,
|
| 79 |
+
"loss": 0.7213,
|
| 80 |
+
"mean_token_accuracy": 0.790926855057478,
|
| 81 |
+
"num_tokens": 44102692.0,
|
| 82 |
"step": 40
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 0.06395452122934801,
|
| 86 |
+
"grad_norm": 0.3786668012815538,
|
| 87 |
"learning_rate": 1.9953823910527057e-05,
|
| 88 |
+
"loss": 0.6845,
|
| 89 |
+
"mean_token_accuracy": 0.8018673352897168,
|
| 90 |
+
"num_tokens": 49606158.0,
|
| 91 |
"step": 45
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 0.07106057914372002,
|
| 95 |
+
"grad_norm": 0.35013514756439346,
|
| 96 |
"learning_rate": 1.993047978140764e-05,
|
| 97 |
+
"loss": 0.6652,
|
| 98 |
+
"mean_token_accuracy": 0.805324524641037,
|
| 99 |
+
"num_tokens": 55124799.0,
|
| 100 |
"step": 50
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.07816663705809203,
|
| 104 |
+
"grad_norm": 0.34309264528308187,
|
| 105 |
"learning_rate": 1.9902398399092494e-05,
|
| 106 |
+
"loss": 0.633,
|
| 107 |
+
"mean_token_accuracy": 0.8133919551968575,
|
| 108 |
+
"num_tokens": 60636117.0,
|
| 109 |
"step": 55
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 0.08527269497246402,
|
| 113 |
+
"grad_norm": 0.31280126819499715,
|
| 114 |
"learning_rate": 1.9869594659603032e-05,
|
| 115 |
+
"loss": 0.6376,
|
| 116 |
+
"mean_token_accuracy": 0.8120212636888027,
|
| 117 |
+
"num_tokens": 66154262.0,
|
| 118 |
"step": 60
|
| 119 |
},
|
| 120 |
{
|
| 121 |
"epoch": 0.09237875288683603,
|
| 122 |
+
"grad_norm": 0.3060483250258842,
|
| 123 |
"learning_rate": 1.9832085963977445e-05,
|
| 124 |
+
"loss": 0.6315,
|
| 125 |
+
"mean_token_accuracy": 0.8132476054131985,
|
| 126 |
+
"num_tokens": 71679661.0,
|
| 127 |
"step": 65
|
| 128 |
},
|
| 129 |
{
|
| 130 |
"epoch": 0.09948481080120802,
|
| 131 |
+
"grad_norm": 0.3345141724039098,
|
| 132 |
"learning_rate": 1.978989220904016e-05,
|
| 133 |
+
"loss": 0.6212,
|
| 134 |
+
"mean_token_accuracy": 0.8146577157080174,
|
| 135 |
+
"num_tokens": 77204132.0,
|
| 136 |
"step": 70
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.10659086871558003,
|
| 140 |
+
"grad_norm": 0.3005501022852846,
|
| 141 |
"learning_rate": 1.9743035776847377e-05,
|
| 142 |
+
"loss": 0.6155,
|
| 143 |
+
"mean_token_accuracy": 0.8163603387773037,
|
| 144 |
+
"num_tokens": 82747984.0,
|
| 145 |
"step": 75
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 0.11369692662995204,
|
| 149 |
+
"grad_norm": 0.2900120545658077,
|
| 150 |
"learning_rate": 1.9691541522814327e-05,
|
| 151 |
+
"loss": 0.5976,
|
| 152 |
+
"mean_token_accuracy": 0.8202282235026359,
|
| 153 |
+
"num_tokens": 88267430.0,
|
| 154 |
"step": 80
|
| 155 |
},
|
| 156 |
{
|
| 157 |
"epoch": 0.12080298454432403,
|
| 158 |
+
"grad_norm": 0.2885088462838806,
|
| 159 |
"learning_rate": 1.963543676253048e-05,
|
| 160 |
+
"loss": 0.6042,
|
| 161 |
+
"mean_token_accuracy": 0.8186872370541096,
|
| 162 |
+
"num_tokens": 93790180.0,
|
| 163 |
"step": 85
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.12790904245869603,
|
| 167 |
+
"grad_norm": 0.2976868963338895,
|
| 168 |
"learning_rate": 1.9574751257269748e-05,
|
| 169 |
+
"loss": 0.597,
|
| 170 |
+
"mean_token_accuracy": 0.8198762819170952,
|
| 171 |
+
"num_tokens": 99314107.0,
|
| 172 |
"step": 90
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.13501510037306805,
|
| 176 |
+
"grad_norm": 0.29708913020397487,
|
| 177 |
"learning_rate": 1.950951719820335e-05,
|
| 178 |
+
"loss": 0.5899,
|
| 179 |
+
"mean_token_accuracy": 0.8221348404884339,
|
| 180 |
+
"num_tokens": 104809146.0,
|
| 181 |
"step": 95
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 0.14212115828744004,
|
| 185 |
+
"grad_norm": 0.37260508301931367,
|
| 186 |
"learning_rate": 1.9439769189323727e-05,
|
| 187 |
+
"loss": 0.5987,
|
| 188 |
+
"mean_token_accuracy": 0.8204578839242458,
|
| 189 |
+
"num_tokens": 110323602.0,
|
| 190 |
"step": 100
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 0.14212115828744004,
|
| 194 |
+
"eval_loss": 0.5712670683860779,
|
| 195 |
+
"eval_mean_token_accuracy": 0.8227657189494685,
|
| 196 |
+
"eval_num_tokens": 110323602.0,
|
| 197 |
+
"eval_runtime": 172.5249,
|
| 198 |
+
"eval_samples_per_second": 21.093,
|
| 199 |
+
"eval_steps_per_second": 0.661,
|
| 200 |
"step": 100
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.14922721620181204,
|
| 204 |
+
"grad_norm": 0.33324601875994775,
|
| 205 |
"learning_rate": 1.9365544229088517e-05,
|
| 206 |
+
"loss": 0.5946,
|
| 207 |
+
"mean_token_accuracy": 0.8209397733211518,
|
| 208 |
+
"num_tokens": 115858100.0,
|
| 209 |
"step": 105
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 0.15633327411618406,
|
| 213 |
+
"grad_norm": 0.39239762510667864,
|
| 214 |
"learning_rate": 1.9286881690794425e-05,
|
| 215 |
+
"loss": 0.5834,
|
| 216 |
+
"mean_token_accuracy": 0.8240870274603367,
|
| 217 |
+
"num_tokens": 121393955.0,
|
| 218 |
"step": 110
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.16343933203055605,
|
| 222 |
+
"grad_norm": 0.32015369228112756,
|
| 223 |
"learning_rate": 1.9203823301691272e-05,
|
| 224 |
+
"loss": 0.5943,
|
| 225 |
+
"mean_token_accuracy": 0.8197322949767113,
|
| 226 |
+
"num_tokens": 126941522.0,
|
| 227 |
"step": 115
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.17054538994492804,
|
| 231 |
+
"grad_norm": 0.307698826802217,
|
| 232 |
"learning_rate": 1.9116413120847425e-05,
|
| 233 |
+
"loss": 0.5854,
|
| 234 |
+
"mean_token_accuracy": 0.8233826741576195,
|
| 235 |
+
"num_tokens": 132467661.0,
|
| 236 |
"step": 120
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.17765144785930007,
|
| 240 |
+
"grad_norm": 0.2976076833401805,
|
| 241 |
"learning_rate": 1.902469751577826e-05,
|
| 242 |
+
"loss": 0.5786,
|
| 243 |
+
"mean_token_accuracy": 0.8247542798519134,
|
| 244 |
+
"num_tokens": 137980426.0,
|
| 245 |
"step": 125
|
| 246 |
},
|
| 247 |
{
|
| 248 |
"epoch": 0.18475750577367206,
|
| 249 |
+
"grad_norm": 0.27117409007655724,
|
| 250 |
"learning_rate": 1.892872513785008e-05,
|
| 251 |
+
"loss": 0.568,
|
| 252 |
+
"mean_token_accuracy": 0.8273683197796344,
|
| 253 |
+
"num_tokens": 143490685.0,
|
| 254 |
"step": 130
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.19186356368804405,
|
| 258 |
+
"grad_norm": 0.29664627267584126,
|
| 259 |
"learning_rate": 1.88285468964726e-05,
|
| 260 |
+
"loss": 0.5733,
|
| 261 |
+
"mean_token_accuracy": 0.8260110780596733,
|
| 262 |
+
"num_tokens": 149018034.0,
|
| 263 |
"step": 135
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.19896962160241605,
|
| 267 |
+
"grad_norm": 0.36485817809652166,
|
| 268 |
"learning_rate": 1.872421593209355e-05,
|
| 269 |
+
"loss": 0.568,
|
| 270 |
+
"mean_token_accuracy": 0.8266257472336292,
|
| 271 |
+
"num_tokens": 154549711.0,
|
| 272 |
"step": 140
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 0.20607567951678807,
|
| 276 |
+
"grad_norm": 0.3079345676026577,
|
| 277 |
"learning_rate": 1.861578758800989e-05,
|
| 278 |
+
"loss": 0.5747,
|
| 279 |
+
"mean_token_accuracy": 0.8259860560297966,
|
| 280 |
+
"num_tokens": 160056919.0,
|
| 281 |
"step": 145
|
| 282 |
},
|
| 283 |
{
|
| 284 |
"epoch": 0.21318173743116006,
|
| 285 |
+
"grad_norm": 0.2852424873621744,
|
| 286 |
"learning_rate": 1.8503319381010414e-05,
|
| 287 |
+
"loss": 0.5688,
|
| 288 |
+
"mean_token_accuracy": 0.8270646832883358,
|
| 289 |
+
"num_tokens": 165584428.0,
|
| 290 |
"step": 150
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.22028779534553206,
|
| 294 |
+
"grad_norm": 0.2960389000145404,
|
| 295 |
"learning_rate": 1.8386870970865488e-05,
|
| 296 |
+
"loss": 0.5619,
|
| 297 |
+
"mean_token_accuracy": 0.8287320241332055,
|
| 298 |
+
"num_tokens": 171107691.0,
|
| 299 |
"step": 155
|
| 300 |
},
|
| 301 |
{
|
| 302 |
"epoch": 0.22739385325990408,
|
| 303 |
+
"grad_norm": 0.3101042392971617,
|
| 304 |
"learning_rate": 1.8266504128679988e-05,
|
| 305 |
+
"loss": 0.5625,
|
| 306 |
+
"mean_token_accuracy": 0.8290504030883312,
|
| 307 |
+
"num_tokens": 176626158.0,
|
| 308 |
"step": 160
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.23449991117427607,
|
| 312 |
+
"grad_norm": 0.2858220649217893,
|
| 313 |
"learning_rate": 1.814228270412624e-05,
|
| 314 |
+
"loss": 0.5771,
|
| 315 |
+
"mean_token_accuracy": 0.824916198849678,
|
| 316 |
+
"num_tokens": 182151744.0,
|
| 317 |
"step": 165
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.24160596908864806,
|
| 321 |
+
"grad_norm": 0.2808854592518143,
|
| 322 |
"learning_rate": 1.8014272591574405e-05,
|
| 323 |
+
"loss": 0.5703,
|
| 324 |
+
"mean_token_accuracy": 0.8272286102175712,
|
| 325 |
+
"num_tokens": 187670198.0,
|
| 326 |
"step": 170
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.2487120270030201,
|
| 330 |
+
"grad_norm": 0.26700783789442195,
|
| 331 |
"learning_rate": 1.7882541695138224e-05,
|
| 332 |
+
"loss": 0.5579,
|
| 333 |
+
"mean_token_accuracy": 0.8301700547337532,
|
| 334 |
+
"num_tokens": 193189696.0,
|
| 335 |
"step": 175
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 0.25581808491739205,
|
| 339 |
+
"grad_norm": 0.2915536966638598,
|
| 340 |
"learning_rate": 1.7747159892654646e-05,
|
| 341 |
+
"loss": 0.5566,
|
| 342 |
+
"mean_token_accuracy": 0.8293862745165825,
|
| 343 |
+
"num_tokens": 198706595.0,
|
| 344 |
"step": 180
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 0.2629241428317641,
|
| 348 |
+
"grad_norm": 0.28557136856744886,
|
| 349 |
"learning_rate": 1.7608198998616533e-05,
|
| 350 |
+
"loss": 0.5636,
|
| 351 |
+
"mean_token_accuracy": 0.8277892582118511,
|
| 352 |
+
"num_tokens": 204263565.0,
|
| 353 |
"step": 185
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.2700302007461361,
|
| 357 |
+
"grad_norm": 0.32075246526065804,
|
| 358 |
"learning_rate": 1.7465732726077993e-05,
|
| 359 |
+
"loss": 0.5595,
|
| 360 |
+
"mean_token_accuracy": 0.8280965454876423,
|
| 361 |
+
"num_tokens": 209753822.0,
|
| 362 |
"step": 190
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.27713625866050806,
|
| 366 |
+
"grad_norm": 0.2660822595268641,
|
| 367 |
"learning_rate": 1.731983664755264e-05,
|
| 368 |
+
"loss": 0.5618,
|
| 369 |
+
"mean_token_accuracy": 0.8285780437290668,
|
| 370 |
+
"num_tokens": 215275472.0,
|
| 371 |
"step": 195
|
| 372 |
},
|
| 373 |
{
|
| 374 |
"epoch": 0.2842423165748801,
|
| 375 |
+
"grad_norm": 0.29789618111303534,
|
| 376 |
"learning_rate": 1.717058815492548e-05,
|
| 377 |
+
"loss": 0.5622,
|
| 378 |
+
"mean_token_accuracy": 0.8278822645545005,
|
| 379 |
+
"num_tokens": 220789450.0,
|
| 380 |
"step": 200
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.2842423165748801,
|
| 384 |
+
"eval_loss": 0.5402741432189941,
|
| 385 |
+
"eval_mean_token_accuracy": 0.8298060622131616,
|
| 386 |
+
"eval_num_tokens": 220789450.0,
|
| 387 |
+
"eval_runtime": 174.4644,
|
| 388 |
+
"eval_samples_per_second": 20.858,
|
| 389 |
+
"eval_steps_per_second": 0.653,
|
| 390 |
"step": 200
|
| 391 |
},
|
| 392 |
{
|
| 393 |
"epoch": 0.2913483744892521,
|
| 394 |
+
"grad_norm": 0.2767591231881183,
|
| 395 |
"learning_rate": 1.701806641839967e-05,
|
| 396 |
+
"loss": 0.5614,
|
| 397 |
+
"mean_token_accuracy": 0.8294874578714371,
|
| 398 |
+
"num_tokens": 226318335.0,
|
| 399 |
"step": 205
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 0.29845443240362407,
|
| 403 |
+
"grad_norm": 0.2603276094371961,
|
| 404 |
"learning_rate": 1.6862352344500004e-05,
|
| 405 |
+
"loss": 0.5602,
|
| 406 |
+
"mean_token_accuracy": 0.8292225509881973,
|
| 407 |
+
"num_tokens": 231830330.0,
|
| 408 |
"step": 210
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.3055604903179961,
|
| 412 |
+
"grad_norm": 0.27620804402617855,
|
| 413 |
"learning_rate": 1.6703528533155283e-05,
|
| 414 |
+
"loss": 0.5557,
|
| 415 |
+
"mean_token_accuracy": 0.829853817075491,
|
| 416 |
+
"num_tokens": 237344871.0,
|
| 417 |
"step": 215
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.3126665482323681,
|
| 421 |
+
"grad_norm": 0.3101017179335764,
|
| 422 |
"learning_rate": 1.6541679233882477e-05,
|
| 423 |
+
"loss": 0.5518,
|
| 424 |
+
"mean_token_accuracy": 0.8312768168747425,
|
| 425 |
+
"num_tokens": 242868508.0,
|
| 426 |
"step": 220
|
| 427 |
},
|
| 428 |
{
|
| 429 |
"epoch": 0.3197726061467401,
|
| 430 |
+
"grad_norm": 0.29395402232070805,
|
| 431 |
"learning_rate": 1.63768903010958e-05,
|
| 432 |
+
"loss": 0.5553,
|
| 433 |
+
"mean_token_accuracy": 0.829937931895256,
|
| 434 |
+
"num_tokens": 248407721.0,
|
| 435 |
"step": 225
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 0.3268786640611121,
|
| 439 |
+
"grad_norm": 0.2934429422246518,
|
| 440 |
"learning_rate": 1.6209249148564437e-05,
|
| 441 |
+
"loss": 0.5503,
|
| 442 |
+
"mean_token_accuracy": 0.8317292295396328,
|
| 443 |
+
"num_tokens": 253911671.0,
|
| 444 |
"step": 230
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.3339847219754841,
|
| 448 |
+
"grad_norm": 0.27828524021977424,
|
| 449 |
"learning_rate": 1.603884470304318e-05,
|
| 450 |
+
"loss": 0.5627,
|
| 451 |
+
"mean_token_accuracy": 0.8284983985126019,
|
| 452 |
+
"num_tokens": 259442838.0,
|
| 453 |
"step": 235
|
| 454 |
},
|
| 455 |
{
|
| 456 |
"epoch": 0.3410907798898561,
|
| 457 |
+
"grad_norm": 0.2833160524998765,
|
| 458 |
"learning_rate": 1.5865767357100383e-05,
|
| 459 |
+
"loss": 0.5453,
|
| 460 |
+
"mean_token_accuracy": 0.8329578995704651,
|
| 461 |
+
"num_tokens": 264975290.0,
|
| 462 |
"step": 240
|
| 463 |
},
|
| 464 |
{
|
| 465 |
"epoch": 0.3481968378042281,
|
| 466 |
+
"grad_norm": 0.3111860491256293,
|
| 467 |
"learning_rate": 1.5690108921168428e-05,
|
| 468 |
+
"loss": 0.5511,
|
| 469 |
+
"mean_token_accuracy": 0.8315815955400467,
|
| 470 |
+
"num_tokens": 270498217.0,
|
| 471 |
"step": 245
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.35530289571860013,
|
| 475 |
+
"grad_norm": 0.28195517442993695,
|
| 476 |
"learning_rate": 1.5511962574842073e-05,
|
| 477 |
+
"loss": 0.5509,
|
| 478 |
+
"mean_token_accuracy": 0.831289553642273,
|
| 479 |
+
"num_tokens": 276014342.0,
|
| 480 |
"step": 250
|
| 481 |
},
|
| 482 |
{
|
| 483 |
"epoch": 0.3624089536329721,
|
| 484 |
+
"grad_norm": 0.28567658174923455,
|
| 485 |
"learning_rate": 1.5331422817450485e-05,
|
| 486 |
+
"loss": 0.5541,
|
| 487 |
+
"mean_token_accuracy": 0.8301701284945011,
|
| 488 |
+
"num_tokens": 281549490.0,
|
| 489 |
"step": 255
|
| 490 |
},
|
| 491 |
{
|
| 492 |
"epoch": 0.3695150115473441,
|
| 493 |
+
"grad_norm": 0.28520328994929606,
|
| 494 |
"learning_rate": 1.5148585417929212e-05,
|
| 495 |
+
"loss": 0.5499,
|
| 496 |
+
"mean_token_accuracy": 0.8319707162678242,
|
| 497 |
+
"num_tokens": 287067865.0,
|
| 498 |
"step": 260
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 0.37662106946171614,
|
| 502 |
+
"grad_norm": 0.2998874255910539,
|
| 503 |
"learning_rate": 1.4963547364018711e-05,
|
| 504 |
+
"loss": 0.5464,
|
| 505 |
+
"mean_token_accuracy": 0.8325809933245182,
|
| 506 |
+
"num_tokens": 292575698.0,
|
| 507 |
"step": 265
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.3837271273760881,
|
| 511 |
+
"grad_norm": 0.27817741706032906,
|
| 512 |
"learning_rate": 1.477640681081632e-05,
|
| 513 |
+
"loss": 0.5492,
|
| 514 |
+
"mean_token_accuracy": 0.831634771823883,
|
| 515 |
+
"num_tokens": 298104461.0,
|
| 516 |
"step": 270
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 0.39083318529046013,
|
| 520 |
+
"grad_norm": 0.28190739794699604,
|
| 521 |
"learning_rate": 1.4587263028709013e-05,
|
| 522 |
+
"loss": 0.5461,
|
| 523 |
+
"mean_token_accuracy": 0.832448897510767,
|
| 524 |
+
"num_tokens": 303615295.0,
|
| 525 |
"step": 275
|
| 526 |
},
|
| 527 |
{
|
| 528 |
"epoch": 0.3979392432048321,
|
| 529 |
+
"grad_norm": 0.2638975088745085,
|
| 530 |
"learning_rate": 1.4396216350714512e-05,
|
| 531 |
+
"loss": 0.5474,
|
| 532 |
+
"mean_token_accuracy": 0.83254414498806,
|
| 533 |
+
"num_tokens": 309131939.0,
|
| 534 |
"step": 280
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.4050453011192041,
|
| 538 |
+
"grad_norm": 0.274569570192276,
|
| 539 |
"learning_rate": 1.4203368119258759e-05,
|
| 540 |
+
"loss": 0.5434,
|
| 541 |
+
"mean_token_accuracy": 0.8325955301523209,
|
| 542 |
+
"num_tokens": 314661318.0,
|
| 543 |
"step": 285
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 0.41215135903357614,
|
| 547 |
+
"grad_norm": 0.28456274011218696,
|
| 548 |
"learning_rate": 1.4008820632417906e-05,
|
| 549 |
+
"loss": 0.5401,
|
| 550 |
+
"mean_token_accuracy": 0.8341820329427719,
|
| 551 |
+
"num_tokens": 320185466.0,
|
| 552 |
"step": 290
|
| 553 |
},
|
| 554 |
{
|
| 555 |
"epoch": 0.4192574169479481,
|
| 556 |
+
"grad_norm": 0.3091981931200518,
|
| 557 |
"learning_rate": 1.381267708965339e-05,
|
| 558 |
+
"loss": 0.5441,
|
| 559 |
+
"mean_token_accuracy": 0.8333158262073994,
|
| 560 |
+
"num_tokens": 325709519.0,
|
| 561 |
"step": 295
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 0.4263634748623201,
|
| 565 |
+
"grad_norm": 0.27485967054454014,
|
| 566 |
"learning_rate": 1.3615041537068831e-05,
|
| 567 |
+
"loss": 0.5514,
|
| 568 |
+
"mean_token_accuracy": 0.8306196145713329,
|
| 569 |
+
"num_tokens": 331248121.0,
|
| 570 |
"step": 300
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.4263634748623201,
|
| 574 |
+
"eval_loss": 0.5253133177757263,
|
| 575 |
+
"eval_mean_token_accuracy": 0.833632708118673,
|
| 576 |
+
"eval_num_tokens": 331248121.0,
|
| 577 |
+
"eval_runtime": 186.3523,
|
| 578 |
+
"eval_samples_per_second": 19.528,
|
| 579 |
+
"eval_steps_per_second": 0.612,
|
| 580 |
"step": 300
|
| 581 |
},
|
| 582 |
{
|
| 583 |
"epoch": 0.43346953277669215,
|
| 584 |
+
"grad_norm": 0.2619898012502233,
|
| 585 |
"learning_rate": 1.3416018812217866e-05,
|
| 586 |
+
"loss": 0.5499,
|
| 587 |
+
"mean_token_accuracy": 0.8310498617589474,
|
| 588 |
+
"num_tokens": 336772706.0,
|
| 589 |
"step": 305
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 0.4405755906910641,
|
| 593 |
+
"grad_norm": 0.26704589829703823,
|
| 594 |
"learning_rate": 1.3215714488492121e-05,
|
| 595 |
+
"loss": 0.5346,
|
| 596 |
+
"mean_token_accuracy": 0.8354850515723229,
|
| 597 |
+
"num_tokens": 342302933.0,
|
| 598 |
"step": 310
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.44768164860543613,
|
| 602 |
+
"grad_norm": 0.2742618507652577,
|
| 603 |
"learning_rate": 1.3014234819118846e-05,
|
| 604 |
+
"loss": 0.5332,
|
| 605 |
+
"mean_token_accuracy": 0.8360144823789597,
|
| 606 |
+
"num_tokens": 347832122.0,
|
| 607 |
"step": 315
|
| 608 |
},
|
| 609 |
{
|
| 610 |
"epoch": 0.45478770651980815,
|
| 611 |
+
"grad_norm": 0.2898062481172225,
|
| 612 |
"learning_rate": 1.2811686680797942e-05,
|
| 613 |
+
"loss": 0.5462,
|
| 614 |
+
"mean_token_accuracy": 0.8317469418048858,
|
| 615 |
+
"num_tokens": 353357495.0,
|
| 616 |
"step": 320
|
| 617 |
},
|
| 618 |
{
|
| 619 |
"epoch": 0.4618937644341801,
|
| 620 |
+
"grad_norm": 0.4126439738262578,
|
| 621 |
"learning_rate": 1.2608177517008268e-05,
|
| 622 |
+
"loss": 0.5378,
|
| 623 |
+
"mean_token_accuracy": 0.834320318698883,
|
| 624 |
+
"num_tokens": 358876427.0,
|
| 625 |
"step": 325
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.46899982234855214,
|
| 629 |
+
"grad_norm": 0.2869066692681731,
|
| 630 |
"learning_rate": 1.240381528101327e-05,
|
| 631 |
+
"loss": 0.5313,
|
| 632 |
+
"mean_token_accuracy": 0.8357653059065342,
|
| 633 |
+
"num_tokens": 364395480.0,
|
| 634 |
"step": 330
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 0.47610588026292416,
|
| 638 |
+
"grad_norm": 0.29613504825068343,
|
| 639 |
"learning_rate": 1.2198708378596198e-05,
|
| 640 |
+
"loss": 0.527,
|
| 641 |
+
"mean_token_accuracy": 0.8370734937489033,
|
| 642 |
+
"num_tokens": 369904006.0,
|
| 643 |
"step": 335
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 0.48321193817729613,
|
| 647 |
+
"grad_norm": 0.3006154330770784,
|
| 648 |
"learning_rate": 1.19929656105553e-05,
|
| 649 |
+
"loss": 0.5314,
|
| 650 |
+
"mean_token_accuracy": 0.835692385584116,
|
| 651 |
+
"num_tokens": 375415817.0,
|
| 652 |
"step": 340
|
| 653 |
},
|
| 654 |
{
|
| 655 |
"epoch": 0.49031799609166815,
|
| 656 |
+
"grad_norm": 0.2904851996175041,
|
| 657 |
"learning_rate": 1.1786696114989455e-05,
|
| 658 |
+
"loss": 0.5324,
|
| 659 |
+
"mean_token_accuracy": 0.835996475070715,
|
| 660 |
+
"num_tokens": 380931196.0,
|
| 661 |
"step": 345
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.4974240540060402,
|
| 665 |
+
"grad_norm": 0.3182709137321541,
|
| 666 |
"learning_rate": 1.1580009309404887e-05,
|
| 667 |
+
"loss": 0.5338,
|
| 668 |
+
"mean_token_accuracy": 0.8356778532266617,
|
| 669 |
+
"num_tokens": 386462024.0,
|
| 670 |
"step": 350
|
| 671 |
},
|
| 672 |
{
|
| 673 |
"epoch": 0.5045301119204122,
|
| 674 |
+
"grad_norm": 0.2859791129483927,
|
| 675 |
"learning_rate": 1.1373014832673661e-05,
|
| 676 |
+
"loss": 0.5362,
|
| 677 |
+
"mean_token_accuracy": 0.8355107352137565,
|
| 678 |
+
"num_tokens": 391971484.0,
|
| 679 |
"step": 355
|
| 680 |
},
|
| 681 |
{
|
| 682 |
"epoch": 0.5116361698347841,
|
| 683 |
+
"grad_norm": 0.2819679337968885,
|
| 684 |
"learning_rate": 1.1165822486874773e-05,
|
| 685 |
+
"loss": 0.5289,
|
| 686 |
+
"mean_token_accuracy": 0.8362454734742641,
|
| 687 |
+
"num_tokens": 397503191.0,
|
| 688 |
"step": 360
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.5187422277491561,
|
| 692 |
+
"grad_norm": 0.2990821176795346,
|
| 693 |
"learning_rate": 1.0958542179048637e-05,
|
| 694 |
+
"loss": 0.5309,
|
| 695 |
+
"mean_token_accuracy": 0.8369539521634579,
|
| 696 |
+
"num_tokens": 403001146.0,
|
| 697 |
"step": 365
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.5258482856635281,
|
| 701 |
+
"grad_norm": 0.522961425134485,
|
| 702 |
"learning_rate": 1.0751283862895914e-05,
|
| 703 |
+
"loss": 0.5423,
|
| 704 |
+
"mean_token_accuracy": 0.8337171785533428,
|
| 705 |
+
"num_tokens": 408526858.0,
|
| 706 |
"step": 370
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 0.5329543435779002,
|
| 710 |
+
"grad_norm": 0.3022320840496628,
|
| 711 |
"learning_rate": 1.0544157480451586e-05,
|
| 712 |
+
"loss": 0.5401,
|
| 713 |
+
"mean_token_accuracy": 0.8333135344088077,
|
| 714 |
+
"num_tokens": 414051639.0,
|
| 715 |
"step": 375
|
| 716 |
},
|
| 717 |
{
|
| 718 |
"epoch": 0.5400604014922722,
|
| 719 |
+
"grad_norm": 0.28660946482078536,
|
| 720 |
"learning_rate": 1.033727290376522e-05,
|
| 721 |
+
"loss": 0.5419,
|
| 722 |
+
"mean_token_accuracy": 0.8340378671884536,
|
| 723 |
+
"num_tokens": 419571564.0,
|
| 724 |
"step": 380
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.5471664594066442,
|
| 728 |
+
"grad_norm": 0.2913510346763107,
|
| 729 |
"learning_rate": 1.013073987661834e-05,
|
| 730 |
+
"loss": 0.5393,
|
| 731 |
+
"mean_token_accuracy": 0.8341242715716362,
|
| 732 |
+
"num_tokens": 425096738.0,
|
| 733 |
"step": 385
|
| 734 |
},
|
| 735 |
{
|
| 736 |
"epoch": 0.5542725173210161,
|
| 737 |
+
"grad_norm": 0.2769745617343647,
|
| 738 |
"learning_rate": 9.924667956309862e-06,
|
| 739 |
+
"loss": 0.5314,
|
| 740 |
+
"mean_token_accuracy": 0.8363712728023529,
|
| 741 |
+
"num_tokens": 430620174.0,
|
| 742 |
"step": 390
|
| 743 |
},
|
| 744 |
{
|
| 745 |
"epoch": 0.5613785752353881,
|
| 746 |
+
"grad_norm": 0.2583947844645885,
|
| 747 |
"learning_rate": 9.719166455540437e-06,
|
| 748 |
+
"loss": 0.5373,
|
| 749 |
+
"mean_token_accuracy": 0.8345833919942379,
|
| 750 |
+
"num_tokens": 436140181.0,
|
| 751 |
"step": 395
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.5684846331497602,
|
| 755 |
+
"grad_norm": 0.33571136388004397,
|
| 756 |
"learning_rate": 9.51434438442655e-06,
|
| 757 |
+
"loss": 0.5348,
|
| 758 |
+
"mean_token_accuracy": 0.8359192557632923,
|
| 759 |
+
"num_tokens": 441662675.0,
|
| 760 |
"step": 400
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 0.5684846331497602,
|
| 764 |
+
"eval_loss": 0.5156883597373962,
|
| 765 |
+
"eval_mean_token_accuracy": 0.8363660657614992,
|
| 766 |
+
"eval_num_tokens": 441662675.0,
|
| 767 |
+
"eval_runtime": 200.4555,
|
| 768 |
+
"eval_samples_per_second": 18.154,
|
| 769 |
+
"eval_steps_per_second": 0.569,
|
| 770 |
"step": 400
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.5755906910641322,
|
| 774 |
+
"grad_norm": 0.27890733469247325,
|
| 775 |
"learning_rate": 9.310310392675132e-06,
|
| 776 |
+
"loss": 0.5232,
|
| 777 |
+
"mean_token_accuracy": 0.8381346166133881,
|
| 778 |
+
"num_tokens": 447154493.0,
|
| 779 |
"step": 405
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.5826967489785042,
|
| 783 |
+
"grad_norm": 0.3083123958703242,
|
| 784 |
"learning_rate": 9.107172711949324e-06,
|
| 785 |
+
"loss": 0.5379,
|
| 786 |
+
"mean_token_accuracy": 0.8339590892195702,
|
| 787 |
+
"num_tokens": 452684121.0,
|
| 788 |
"step": 410
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 0.5898028068928762,
|
| 792 |
+
"grad_norm": 0.2595437811318788,
|
| 793 |
"learning_rate": 8.905039098456049e-06,
|
| 794 |
+
"loss": 0.53,
|
| 795 |
+
"mean_token_accuracy": 0.8357990562915802,
|
| 796 |
+
"num_tokens": 458209573.0,
|
| 797 |
"step": 415
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 0.5969088648072481,
|
| 801 |
+
"grad_norm": 0.36529180881846035,
|
| 802 |
"learning_rate": 8.704016775785742e-06,
|
| 803 |
+
"loss": 0.5348,
|
| 804 |
+
"mean_token_accuracy": 0.8356608338654041,
|
| 805 |
+
"num_tokens": 463743007.0,
|
| 806 |
"step": 420
|
| 807 |
},
|
| 808 |
{
|
| 809 |
"epoch": 0.6040149227216202,
|
| 810 |
+
"grad_norm": 0.30289810672979556,
|
| 811 |
"learning_rate": 8.50421237803464e-06,
|
| 812 |
+
"loss": 0.5291,
|
| 813 |
+
"mean_token_accuracy": 0.8363065294921398,
|
| 814 |
+
"num_tokens": 469259672.0,
|
| 815 |
"step": 425
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.6111209806359922,
|
| 819 |
+
"grad_norm": 0.26457343656286747,
|
| 820 |
"learning_rate": 8.30573189323978e-06,
|
| 821 |
+
"loss": 0.5231,
|
| 822 |
+
"mean_token_accuracy": 0.838748998194933,
|
| 823 |
+
"num_tokens": 474761309.0,
|
| 824 |
"step": 430
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 0.6182270385503642,
|
| 828 |
+
"grad_norm": 0.2689617228112552,
|
| 829 |
"learning_rate": 8.108680607156669e-06,
|
| 830 |
+
"loss": 0.5373,
|
| 831 |
+
"mean_token_accuracy": 0.8346136786043644,
|
| 832 |
+
"num_tokens": 480283239.0,
|
| 833 |
"step": 435
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 0.6253330964647362,
|
| 837 |
+
"grad_norm": 0.3088485273327652,
|
| 838 |
"learning_rate": 7.913163047409533e-06,
|
| 839 |
+
"loss": 0.5296,
|
| 840 |
+
"mean_token_accuracy": 0.8361807405948639,
|
| 841 |
+
"num_tokens": 485803053.0,
|
| 842 |
"step": 440
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.6324391543791081,
|
| 846 |
+
"grad_norm": 0.2721839437620559,
|
| 847 |
"learning_rate": 7.719282928043688e-06,
|
| 848 |
+
"loss": 0.5312,
|
| 849 |
+
"mean_token_accuracy": 0.835797655582428,
|
| 850 |
+
"num_tokens": 491312112.0,
|
| 851 |
"step": 445
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 0.6395452122934802,
|
| 855 |
+
"grad_norm": 0.2520862488552354,
|
| 856 |
"learning_rate": 7.527143094509492e-06,
|
| 857 |
+
"loss": 0.5294,
|
| 858 |
+
"mean_token_accuracy": 0.8369647547602653,
|
| 859 |
+
"num_tokens": 496828932.0,
|
| 860 |
"step": 450
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 0.6466512702078522,
|
| 864 |
+
"grad_norm": 0.25758192949076364,
|
| 865 |
"learning_rate": 7.336845469107061e-06,
|
| 866 |
+
"loss": 0.5293,
|
| 867 |
+
"mean_token_accuracy": 0.8363292142748833,
|
| 868 |
+
"num_tokens": 502329483.0,
|
| 869 |
"step": 455
|
| 870 |
},
|
| 871 |
{
|
| 872 |
"epoch": 0.6537573281222242,
|
| 873 |
+
"grad_norm": 0.2820371395553448,
|
| 874 |
"learning_rate": 7.148490996920661e-06,
|
| 875 |
+
"loss": 0.5315,
|
| 876 |
+
"mean_token_accuracy": 0.8359401389956475,
|
| 877 |
+
"num_tokens": 507853922.0,
|
| 878 |
"step": 460
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.6608633860365962,
|
| 882 |
+
"grad_norm": 0.34193322970792595,
|
| 883 |
"learning_rate": 6.9621795922714805e-06,
|
| 884 |
+
"loss": 0.5283,
|
| 885 |
+
"mean_token_accuracy": 0.837289647012949,
|
| 886 |
+
"num_tokens": 513366033.0,
|
| 887 |
"step": 465
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 0.6679694439509682,
|
| 891 |
+
"grad_norm": 0.2909689799049006,
|
| 892 |
"learning_rate": 6.778010085717202e-06,
|
| 893 |
+
"loss": 0.5278,
|
| 894 |
+
"mean_token_accuracy": 0.8373917371034623,
|
| 895 |
+
"num_tokens": 518888798.0,
|
| 896 |
"step": 470
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.6750755018653402,
|
| 900 |
+
"grad_norm": 0.24946404604971933,
|
| 901 |
"learning_rate": 6.596080171626409e-06,
|
| 902 |
+
"loss": 0.5306,
|
| 903 |
+
"mean_token_accuracy": 0.8356848947703839,
|
| 904 |
+
"num_tokens": 524392285.0,
|
| 905 |
"step": 475
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.6821815597797122,
|
| 909 |
+
"grad_norm": 0.2622556922485006,
|
| 910 |
"learning_rate": 6.416486356355769e-06,
|
| 911 |
+
"loss": 0.5365,
|
| 912 |
+
"mean_token_accuracy": 0.8342930674552917,
|
| 913 |
+
"num_tokens": 529904641.0,
|
| 914 |
"step": 480
|
| 915 |
},
|
| 916 |
{
|
| 917 |
"epoch": 0.6892876176940842,
|
| 918 |
+
"grad_norm": 0.24646929820568078,
|
| 919 |
"learning_rate": 6.239323907057342e-06,
|
| 920 |
+
"loss": 0.5334,
|
| 921 |
+
"mean_token_accuracy": 0.8358374811708927,
|
| 922 |
+
"num_tokens": 535417362.0,
|
| 923 |
"step": 485
|
| 924 |
},
|
| 925 |
{
|
| 926 |
"epoch": 0.6963936756084562,
|
| 927 |
+
"grad_norm": 0.2614785261373714,
|
| 928 |
"learning_rate": 6.064686801143271e-06,
|
| 929 |
+
"loss": 0.5166,
|
| 930 |
+
"mean_token_accuracy": 0.8401089183986187,
|
| 931 |
+
"num_tokens": 540908836.0,
|
| 932 |
"step": 490
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 0.7034997335228282,
|
| 936 |
+
"grad_norm": 0.25886556075681727,
|
| 937 |
"learning_rate": 5.892667676434633e-06,
|
| 938 |
+
"loss": 0.5241,
|
| 939 |
+
"mean_token_accuracy": 0.8378406457602978,
|
| 940 |
+
"num_tokens": 546444796.0,
|
| 941 |
"step": 495
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.7106057914372003,
|
| 945 |
+
"grad_norm": 0.2742344703816926,
|
| 946 |
"learning_rate": 5.723357782020867e-06,
|
| 947 |
+
"loss": 0.5224,
|
| 948 |
+
"mean_token_accuracy": 0.8380300879478455,
|
| 949 |
+
"num_tokens": 551952947.0,
|
| 950 |
"step": 500
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 0.7106057914372003,
|
| 954 |
+
"eval_loss": 0.509369969367981,
|
| 955 |
+
"eval_mean_token_accuracy": 0.8376399942657404,
|
| 956 |
+
"eval_num_tokens": 551952947.0,
|
| 957 |
+
"eval_runtime": 195.3032,
|
| 958 |
+
"eval_samples_per_second": 18.633,
|
| 959 |
+
"eval_steps_per_second": 0.584,
|
| 960 |
"step": 500
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 0.7177118493515722,
|
| 964 |
+
"grad_norm": 0.2834434176358039,
|
| 965 |
"learning_rate": 5.556846929855857e-06,
|
| 966 |
+
"loss": 0.5199,
|
| 967 |
+
"mean_token_accuracy": 0.8389093585312366,
|
| 968 |
+
"num_tokens": 557467246.0,
|
| 969 |
"step": 505
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.7248179072659442,
|
| 973 |
+
"grad_norm": 0.26408949367317563,
|
| 974 |
"learning_rate": 5.393223447116409e-06,
|
| 975 |
+
"loss": 0.5344,
|
| 976 |
+
"mean_token_accuracy": 0.8358340613543987,
|
| 977 |
+
"num_tokens": 562988077.0,
|
| 978 |
"step": 510
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 0.7319239651803162,
|
| 982 |
+
"grad_norm": 0.2750963122440575,
|
| 983 |
"learning_rate": 5.232574129348278e-06,
|
| 984 |
+
"loss": 0.5234,
|
| 985 |
+
"mean_token_accuracy": 0.8388362295925618,
|
| 986 |
+
"num_tokens": 568506583.0,
|
| 987 |
"step": 515
|
| 988 |
},
|
| 989 |
{
|
| 990 |
"epoch": 0.7390300230946882,
|
| 991 |
+
"grad_norm": 0.2688348127137177,
|
| 992 |
"learning_rate": 5.0749841944247e-06,
|
| 993 |
+
"loss": 0.5335,
|
| 994 |
+
"mean_token_accuracy": 0.8347815096378326,
|
| 995 |
+
"num_tokens": 574039672.0,
|
| 996 |
"step": 520
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.7461360810090603,
|
| 1000 |
+
"grad_norm": 0.2665245986172324,
|
| 1001 |
"learning_rate": 4.92053723734182e-06,
|
| 1002 |
+
"loss": 0.5312,
|
| 1003 |
+
"mean_token_accuracy": 0.8360748492181301,
|
| 1004 |
+
"num_tokens": 579544417.0,
|
| 1005 |
"step": 525
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
"epoch": 0.7532421389234323,
|
| 1009 |
+
"grad_norm": 0.27943620542156355,
|
| 1010 |
"learning_rate": 4.769315185874951e-06,
|
| 1011 |
+
"loss": 0.5268,
|
| 1012 |
+
"mean_token_accuracy": 0.8377416066825389,
|
| 1013 |
+
"num_tokens": 585066703.0,
|
| 1014 |
"step": 530
|
| 1015 |
},
|
| 1016 |
{
|
| 1017 |
"epoch": 0.7603481968378042,
|
| 1018 |
+
"grad_norm": 0.24611383998383726,
|
| 1019 |
"learning_rate": 4.621398257119266e-06,
|
| 1020 |
+
"loss": 0.5258,
|
| 1021 |
+
"mean_token_accuracy": 0.8379491157829762,
|
| 1022 |
+
"num_tokens": 590596017.0,
|
| 1023 |
"step": 535
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 0.7674542547521762,
|
| 1027 |
+
"grad_norm": 0.2803451376767719,
|
| 1028 |
"learning_rate": 4.476864914937923e-06,
|
| 1029 |
+
"loss": 0.5193,
|
| 1030 |
+
"mean_token_accuracy": 0.8394668839871884,
|
| 1031 |
+
"num_tokens": 596113106.0,
|
| 1032 |
"step": 540
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 0.7745603126665482,
|
| 1036 |
+
"grad_norm": 0.2629240113229047,
|
| 1037 |
"learning_rate": 4.335791828340183e-06,
|
| 1038 |
+
"loss": 0.5295,
|
| 1039 |
+
"mean_token_accuracy": 0.8369358383119107,
|
| 1040 |
+
"num_tokens": 601659151.0,
|
| 1041 |
"step": 545
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 0.7816663705809203,
|
| 1045 |
+
"grad_norm": 0.25281118878900727,
|
| 1046 |
"learning_rate": 4.1982538308116775e-06,
|
| 1047 |
+
"loss": 0.5247,
|
| 1048 |
+
"mean_token_accuracy": 0.8361569255590439,
|
| 1049 |
+
"num_tokens": 607175134.0,
|
| 1050 |
"step": 550
|
| 1051 |
},
|
| 1052 |
{
|
| 1053 |
"epoch": 0.7887724284952923,
|
| 1054 |
+
"grad_norm": 0.2546111373713822,
|
| 1055 |
"learning_rate": 4.064323880618279e-06,
|
| 1056 |
+
"loss": 0.5277,
|
| 1057 |
+
"mean_token_accuracy": 0.8377487383782863,
|
| 1058 |
+
"num_tokens": 612699036.0,
|
| 1059 |
"step": 555
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.7958784864096642,
|
| 1063 |
+
"grad_norm": 0.2669669649381399,
|
| 1064 |
"learning_rate": 3.934073022104759e-06,
|
| 1065 |
+
"loss": 0.5232,
|
| 1066 |
+
"mean_token_accuracy": 0.8382376879453659,
|
| 1067 |
+
"num_tokens": 618232472.0,
|
| 1068 |
"step": 560
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
"epoch": 0.8029845443240362,
|
| 1072 |
+
"grad_norm": 0.28999220785660285,
|
| 1073 |
"learning_rate": 3.807570348008672e-06,
|
| 1074 |
+
"loss": 0.5243,
|
| 1075 |
+
"mean_token_accuracy": 0.8378099992871284,
|
| 1076 |
+
"num_tokens": 623766598.0,
|
| 1077 |
"step": 565
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
"epoch": 0.8100906022384082,
|
| 1081 |
+
"grad_norm": 0.3148725227834589,
|
| 1082 |
"learning_rate": 3.684882962809484e-06,
|
| 1083 |
+
"loss": 0.5303,
|
| 1084 |
+
"mean_token_accuracy": 0.8364369817078113,
|
| 1085 |
+
"num_tokens": 629297697.0,
|
| 1086 |
"step": 570
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 0.8171966601527803,
|
| 1090 |
+
"grad_norm": 0.2532113025757247,
|
| 1091 |
"learning_rate": 3.5660759471324037e-06,
|
| 1092 |
+
"loss": 0.5287,
|
| 1093 |
+
"mean_token_accuracy": 0.8368946433067321,
|
| 1094 |
+
"num_tokens": 634808630.0,
|
| 1095 |
"step": 575
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 0.8243027180671523,
|
| 1099 |
+
"grad_norm": 0.270452120656795,
|
| 1100 |
"learning_rate": 3.451212323225786e-06,
|
| 1101 |
+
"loss": 0.5203,
|
| 1102 |
+
"mean_token_accuracy": 0.8387155674397946,
|
| 1103 |
+
"num_tokens": 640315396.0,
|
| 1104 |
"step": 580
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 0.8314087759815243,
|
| 1108 |
+
"grad_norm": 0.24111734087156292,
|
| 1109 |
"learning_rate": 3.340353021530409e-06,
|
| 1110 |
+
"loss": 0.5211,
|
| 1111 |
+
"mean_token_accuracy": 0.8376603581011295,
|
| 1112 |
+
"num_tokens": 645841425.0,
|
| 1113 |
"step": 585
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
"epoch": 0.8385148338958962,
|
| 1117 |
+
"grad_norm": 0.25659082714673837,
|
| 1118 |
"learning_rate": 3.2335568483583708e-06,
|
| 1119 |
+
"loss": 0.5162,
|
| 1120 |
+
"mean_token_accuracy": 0.8415209501981735,
|
| 1121 |
+
"num_tokens": 651349089.0,
|
| 1122 |
"step": 590
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.8456208918102682,
|
| 1126 |
+
"grad_norm": 0.2536382351794681,
|
| 1127 |
"learning_rate": 3.1308804546987615e-06,
|
| 1128 |
+
"loss": 0.5305,
|
| 1129 |
+
"mean_token_accuracy": 0.8366158396005631,
|
| 1130 |
+
"num_tokens": 656881938.0,
|
| 1131 |
"step": 595
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
"epoch": 0.8527269497246402,
|
| 1135 |
+
"grad_norm": 0.3569014386612136,
|
| 1136 |
"learning_rate": 3.0323783061666307e-06,
|
| 1137 |
+
"loss": 0.5219,
|
| 1138 |
+
"mean_token_accuracy": 0.838913232088089,
|
| 1139 |
+
"num_tokens": 662399294.0,
|
| 1140 |
"step": 600
|
| 1141 |
},
|
| 1142 |
{
|
| 1143 |
"epoch": 0.8527269497246402,
|
| 1144 |
+
"eval_loss": 0.5057865977287292,
|
| 1145 |
+
"eval_mean_token_accuracy": 0.8385337261777175,
|
| 1146 |
+
"eval_num_tokens": 662399294.0,
|
| 1147 |
+
"eval_runtime": 177.6102,
|
| 1148 |
+
"eval_samples_per_second": 20.489,
|
| 1149 |
+
"eval_steps_per_second": 0.642,
|
| 1150 |
"step": 600
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.8598330076390123,
|
| 1154 |
+
"grad_norm": 0.26005579437743775,
|
| 1155 |
"learning_rate": 2.9381026541112145e-06,
|
| 1156 |
+
"loss": 0.5238,
|
| 1157 |
+
"mean_token_accuracy": 0.8379264928400516,
|
| 1158 |
+
"num_tokens": 667931430.0,
|
| 1159 |
"step": 605
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
"epoch": 0.8669390655533843,
|
| 1163 |
+
"grad_norm": 0.26240817562208274,
|
| 1164 |
"learning_rate": 2.848103507898745e-06,
|
| 1165 |
+
"loss": 0.5271,
|
| 1166 |
+
"mean_token_accuracy": 0.8367306478321552,
|
| 1167 |
+
"num_tokens": 673461286.0,
|
| 1168 |
"step": 610
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 0.8740451234677563,
|
| 1172 |
+
"grad_norm": 0.2769009955982679,
|
| 1173 |
"learning_rate": 2.7624286083845187e-06,
|
| 1174 |
+
"loss": 0.522,
|
| 1175 |
+
"mean_token_accuracy": 0.8375455342233181,
|
| 1176 |
+
"num_tokens": 678984176.0,
|
| 1177 |
"step": 615
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
"epoch": 0.8811511813821282,
|
| 1181 |
+
"grad_norm": 0.2694784557320761,
|
| 1182 |
"learning_rate": 2.6811234025883457e-06,
|
| 1183 |
+
"loss": 0.5173,
|
| 1184 |
+
"mean_token_accuracy": 0.8397197715938092,
|
| 1185 |
+
"num_tokens": 684519341.0,
|
| 1186 |
"step": 620
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 0.8882572392965002,
|
| 1190 |
+
"grad_norm": 0.25312087877008377,
|
| 1191 |
"learning_rate": 2.604231019586761e-06,
|
| 1192 |
+
"loss": 0.521,
|
| 1193 |
+
"mean_token_accuracy": 0.8392728187143803,
|
| 1194 |
+
"num_tokens": 690037943.0,
|
| 1195 |
"step": 625
|
| 1196 |
},
|
| 1197 |
{
|
| 1198 |
"epoch": 0.8953632972108723,
|
| 1199 |
+
"grad_norm": 0.2662458048336358,
|
| 1200 |
"learning_rate": 2.5317922476348194e-06,
|
| 1201 |
+
"loss": 0.5226,
|
| 1202 |
+
"mean_token_accuracy": 0.838124718517065,
|
| 1203 |
+
"num_tokens": 695563755.0,
|
| 1204 |
"step": 630
|
| 1205 |
},
|
| 1206 |
{
|
| 1207 |
"epoch": 0.9024693551252443,
|
| 1208 |
+
"grad_norm": 0.25072250566051796,
|
| 1209 |
"learning_rate": 2.4638455125296043e-06,
|
| 1210 |
+
"loss": 0.5246,
|
| 1211 |
+
"mean_token_accuracy": 0.8378940619528293,
|
| 1212 |
+
"num_tokens": 701088593.0,
|
| 1213 |
"step": 635
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.9095754130396163,
|
| 1217 |
+
"grad_norm": 0.2500242495687067,
|
| 1218 |
"learning_rate": 2.400426857226914e-06,
|
| 1219 |
+
"loss": 0.518,
|
| 1220 |
+
"mean_token_accuracy": 0.8393492937088013,
|
| 1221 |
+
"num_tokens": 706621161.0,
|
| 1222 |
"step": 640
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
"epoch": 0.9166814709539882,
|
| 1226 |
+
"grad_norm": 0.25909853709039266,
|
| 1227 |
"learning_rate": 2.3415699227219517e-06,
|
| 1228 |
+
"loss": 0.5297,
|
| 1229 |
+
"mean_token_accuracy": 0.836426030099392,
|
| 1230 |
+
"num_tokens": 712135507.0,
|
| 1231 |
"step": 645
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 0.9237875288683602,
|
| 1235 |
+
"grad_norm": 0.24930834384146366,
|
| 1236 |
"learning_rate": 2.2873059302041627e-06,
|
| 1237 |
+
"loss": 0.52,
|
| 1238 |
+
"mean_token_accuracy": 0.8393146157264709,
|
| 1239 |
+
"num_tokens": 717654276.0,
|
| 1240 |
"step": 650
|
| 1241 |
},
|
| 1242 |
{
|
| 1243 |
"epoch": 0.9308935867827323,
|
| 1244 |
+
"grad_norm": 0.3669296287397151,
|
| 1245 |
"learning_rate": 2.2376636644956656e-06,
|
| 1246 |
+
"loss": 0.5177,
|
| 1247 |
+
"mean_token_accuracy": 0.8393071293830872,
|
| 1248 |
+
"num_tokens": 723171964.0,
|
| 1249 |
"step": 655
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.9379996446971043,
|
| 1253 |
+
"grad_norm": 0.2791419593414452,
|
| 1254 |
"learning_rate": 2.192669458782096e-06,
|
| 1255 |
+
"loss": 0.5255,
|
| 1256 |
+
"mean_token_accuracy": 0.8377047665417194,
|
| 1257 |
+
"num_tokens": 728678520.0,
|
| 1258 |
"step": 660
|
| 1259 |
},
|
| 1260 |
{
|
| 1261 |
"epoch": 0.9451057026114763,
|
| 1262 |
+
"grad_norm": 0.25640575855730413,
|
| 1263 |
"learning_rate": 2.1523471806439205e-06,
|
| 1264 |
+
"loss": 0.5346,
|
| 1265 |
+
"mean_token_accuracy": 0.8354143664240837,
|
| 1266 |
+
"num_tokens": 734210332.0,
|
| 1267 |
"step": 665
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
"epoch": 0.9522117605258483,
|
| 1271 |
+
"grad_norm": 0.25608979314758057,
|
| 1272 |
"learning_rate": 2.1167182193956738e-06,
|
| 1273 |
+
"loss": 0.516,
|
| 1274 |
+
"mean_token_accuracy": 0.8407890982925892,
|
| 1275 |
+
"num_tokens": 739702586.0,
|
| 1276 |
"step": 670
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
"epoch": 0.9593178184402202,
|
| 1280 |
+
"grad_norm": 0.2552144728023743,
|
| 1281 |
"learning_rate": 2.0858014747397952e-06,
|
| 1282 |
+
"loss": 0.5242,
|
| 1283 |
+
"mean_token_accuracy": 0.8386999048292637,
|
| 1284 |
+
"num_tokens": 745219812.0,
|
| 1285 |
"step": 675
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 0.9664238763545923,
|
| 1289 |
+
"grad_norm": 0.2458787780145255,
|
| 1290 |
"learning_rate": 2.0596133467411213e-06,
|
| 1291 |
+
"loss": 0.5175,
|
| 1292 |
+
"mean_token_accuracy": 0.8397213242948055,
|
| 1293 |
+
"num_tokens": 750717984.0,
|
| 1294 |
"step": 680
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 0.9735299342689643,
|
| 1298 |
+
"grad_norm": 0.26076663120485116,
|
| 1299 |
"learning_rate": 2.0381677271273177e-06,
|
| 1300 |
+
"loss": 0.521,
|
| 1301 |
+
"mean_token_accuracy": 0.8379410050809384,
|
| 1302 |
+
"num_tokens": 756252357.0,
|
| 1303 |
"step": 685
|
| 1304 |
},
|
| 1305 |
{
|
| 1306 |
"epoch": 0.9806359921833363,
|
| 1307 |
+
"grad_norm": 0.2752948731550225,
|
| 1308 |
"learning_rate": 2.0214759919198904e-06,
|
| 1309 |
+
"loss": 0.515,
|
| 1310 |
+
"mean_token_accuracy": 0.8394489504396916,
|
| 1311 |
+
"num_tokens": 761750059.0,
|
| 1312 |
"step": 690
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.9877420500977083,
|
| 1316 |
+
"grad_norm": 0.671568399451163,
|
| 1317 |
"learning_rate": 2.0095469953996724e-06,
|
| 1318 |
+
"loss": 0.5242,
|
| 1319 |
+
"mean_token_accuracy": 0.8378232717514038,
|
| 1320 |
+
"num_tokens": 767275567.0,
|
| 1321 |
"step": 695
|
| 1322 |
},
|
| 1323 |
{
|
| 1324 |
"epoch": 0.9948481080120803,
|
| 1325 |
+
"grad_norm": 0.2422210886510521,
|
| 1326 |
"learning_rate": 2.002387065409989e-06,
|
| 1327 |
+
"loss": 0.5271,
|
| 1328 |
+
"mean_token_accuracy": 0.8371209591627121,
|
| 1329 |
+
"num_tokens": 772815373.0,
|
| 1330 |
"step": 700
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 0.9948481080120803,
|
| 1334 |
+
"eval_loss": 0.5038756132125854,
|
| 1335 |
+
"eval_mean_token_accuracy": 0.8392674577863592,
|
| 1336 |
+
"eval_num_tokens": 772815373.0,
|
| 1337 |
+
"eval_runtime": 189.9146,
|
| 1338 |
+
"eval_samples_per_second": 19.161,
|
| 1339 |
+
"eval_steps_per_second": 0.6,
|
| 1340 |
"step": 700
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 1.0,
|
| 1344 |
+
"mean_token_accuracy": 0.8374893572823755,
|
| 1345 |
+
"num_tokens": 776817701.0,
|
| 1346 |
"step": 704,
|
| 1347 |
+
"total_flos": 6035679775555584.0,
|
| 1348 |
+
"train_loss": 0.6108071732927453,
|
| 1349 |
+
"train_runtime": 30814.6397,
|
| 1350 |
+
"train_samples_per_second": 11.691,
|
| 1351 |
+
"train_steps_per_second": 0.023
|
| 1352 |
}
|
| 1353 |
],
|
| 1354 |
"logging_steps": 5,
|
|
|
|
| 1368 |
"attributes": {}
|
| 1369 |
}
|
| 1370 |
},
|
| 1371 |
+
"total_flos": 6035679775555584.0,
|
| 1372 |
"train_batch_size": 4,
|
| 1373 |
"trial_name": null,
|
| 1374 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06e473fb8ea01f99ee7530caf95335065acd44fc3c9db811bee45f3718a42c37
|
| 3 |
+
size 7480
|