End of training
Browse files- README.md +1 -1
- all_results.json +4 -4
- model-00001-of-00004.safetensors +1 -1
- model-00002-of-00004.safetensors +1 -1
- model-00003-of-00004.safetensors +1 -1
- model-00004-of-00004.safetensors +1 -1
- runs/May18_05-21-42_w-jerom-inter-play-sim-94c6890b9ccf44ea86f033a3db8a5dbd-78pqvlh/events.out.tfevents.1747545955.w-jerom-inter-play-sim-94c6890b9ccf44ea86f033a3db8a5dbd-78pqvlh.90146.0 +3 -0
- special_tokens_map.json +7 -1
- tokenizer.json +2 -2
- train_results.json +4 -4
- trainer_state.json +587 -587
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -27,7 +27,7 @@ print(output["generated_text"])
|
|
| 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/jerome-ramos-20/huggingface/runs/
|
| 31 |
|
| 32 |
|
| 33 |
This model was trained with SFT.
|
|
|
|
| 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/jerome-ramos-20/huggingface/runs/puy8bq5a)
|
| 31 |
|
| 32 |
|
| 33 |
This model was trained with SFT.
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9982692973347178,
|
| 3 |
"total_flos": 6.774706456143856e+18,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 46218,
|
| 7 |
-
"train_samples_per_second": 6.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9982692973347178,
|
| 3 |
"total_flos": 6.774706456143856e+18,
|
| 4 |
+
"train_loss": 0.6556608070957364,
|
| 5 |
+
"train_runtime": 13724.1641,
|
| 6 |
"train_samples": 46218,
|
| 7 |
+
"train_samples_per_second": 6.735,
|
| 8 |
+
"train_steps_per_second": 0.105
|
| 9 |
}
|
model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4977222960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:119c6e9db805bf7cbe5b7cddbe818e4bf4da2748d34508d82daf25935136683a
|
| 3 |
size 4977222960
|
model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4999802720
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0e00d37684b3ac74a1c2fb14ecd59ffbbc93837c2ed27a70556ac5969f6deda
|
| 3 |
size 4999802720
|
model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4915916176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1a94e05c1e8ec6e46afea28d429473ef7fa2edf701ed9e447e06e3436466994
|
| 3 |
size 4915916176
|
model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1168663096
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1dd969a6ceac8efec02f60fc1144dd9312ae80a3117e4de82e6ca9972576cdb
|
| 3 |
size 1168663096
|
runs/May18_05-21-42_w-jerom-inter-play-sim-94c6890b9ccf44ea86f033a3db8a5dbd-78pqvlh/events.out.tfevents.1747545955.w-jerom-inter-play-sim-94c6890b9ccf44ea86f033a3db8a5dbd-78pqvlh.90146.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:075338236141c3fc1a2e7ec144407edf30b57e7e9ca49d0dd5f675c681ea9c13
|
| 3 |
+
size 67519
|
special_tokens_map.json
CHANGED
|
@@ -99,5 +99,11 @@
|
|
| 99 |
"rstrip": false,
|
| 100 |
"single_word": false
|
| 101 |
},
|
| 102 |
-
"pad_token":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
}
|
|
|
|
| 99 |
"rstrip": false,
|
| 100 |
"single_word": false
|
| 101 |
},
|
| 102 |
+
"pad_token": {
|
| 103 |
+
"content": "<|finetune_right_pad_id|>",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": false,
|
| 106 |
+
"rstrip": false,
|
| 107 |
+
"single_word": false
|
| 108 |
+
}
|
| 109 |
}
|
tokenizer.json
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 sha256:7dec9f5ce81a37e00b41babd3428f22ccb961fb8ba1ff9ca5e6979e50456e106
|
| 3 |
+
size 17212647
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9982692973347178,
|
| 3 |
"total_flos": 6.774706456143856e+18,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 46218,
|
| 7 |
-
"train_samples_per_second": 6.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9982692973347178,
|
| 3 |
"total_flos": 6.774706456143856e+18,
|
| 4 |
+
"train_loss": 0.6556608070957364,
|
| 5 |
+
"train_runtime": 13724.1641,
|
| 6 |
"train_samples": 46218,
|
| 7 |
+
"train_samples_per_second": 6.735,
|
| 8 |
+
"train_steps_per_second": 0.105
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -11,2051 +11,2051 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0013845621322256837,
|
| 14 |
-
"grad_norm": 73.
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
"loss": 3.4696,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.006922810661128418,
|
| 21 |
-
"grad_norm": 30.
|
| 22 |
"learning_rate": 5.517241379310345e-06,
|
| 23 |
-
"loss": 3.
|
| 24 |
"step": 5
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.013845621322256836,
|
| 28 |
-
"grad_norm":
|
| 29 |
"learning_rate": 1.2413793103448277e-05,
|
| 30 |
-
"loss": 2.
|
| 31 |
"step": 10
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.020768431983385256,
|
| 35 |
-
"grad_norm": 7.
|
| 36 |
"learning_rate": 1.9310344827586207e-05,
|
| 37 |
-
"loss": 2.
|
| 38 |
"step": 15
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.027691242644513673,
|
| 42 |
-
"grad_norm": 7.
|
| 43 |
"learning_rate": 2.620689655172414e-05,
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 20
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.034614053305642094,
|
| 49 |
-
"grad_norm": 7.
|
| 50 |
"learning_rate": 3.310344827586207e-05,
|
| 51 |
-
"loss": 1.
|
| 52 |
"step": 25
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.04153686396677051,
|
| 56 |
-
"grad_norm": 7.
|
| 57 |
"learning_rate": 4e-05,
|
| 58 |
-
"loss": 1.
|
| 59 |
"step": 30
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.04845967462789893,
|
| 63 |
-
"grad_norm": 3.
|
| 64 |
"learning_rate": 4.689655172413793e-05,
|
| 65 |
-
"loss": 1.
|
| 66 |
"step": 35
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.055382485289027346,
|
| 70 |
-
"grad_norm":
|
| 71 |
"learning_rate": 5.379310344827586e-05,
|
| 72 |
-
"loss": 1.
|
| 73 |
"step": 40
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.06230529595015576,
|
| 77 |
-
"grad_norm":
|
| 78 |
"learning_rate": 6.068965517241379e-05,
|
| 79 |
-
"loss": 1.
|
| 80 |
"step": 45
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.06922810661128419,
|
| 84 |
-
"grad_norm": 3.
|
| 85 |
"learning_rate": 6.758620689655173e-05,
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 50
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.0761509172724126,
|
| 91 |
-
"grad_norm":
|
| 92 |
"learning_rate": 7.448275862068966e-05,
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 55
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.08307372793354102,
|
| 98 |
-
"grad_norm":
|
| 99 |
"learning_rate": 8.137931034482759e-05,
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 60
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.08999653859466944,
|
| 105 |
-
"grad_norm":
|
| 106 |
"learning_rate": 8.827586206896552e-05,
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 65
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.09691934925579786,
|
| 112 |
-
"grad_norm":
|
| 113 |
"learning_rate": 9.517241379310345e-05,
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 70
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.10384215991692627,
|
| 119 |
-
"grad_norm":
|
| 120 |
"learning_rate": 0.0001020689655172414,
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 75
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.11076497057805469,
|
| 126 |
-
"grad_norm":
|
| 127 |
"learning_rate": 0.00010896551724137931,
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 80
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.11768778123918311,
|
| 133 |
-
"grad_norm":
|
| 134 |
"learning_rate": 0.00011586206896551725,
|
| 135 |
-
"loss":
|
| 136 |
"step": 85
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.12461059190031153,
|
| 140 |
-
"grad_norm":
|
| 141 |
"learning_rate": 0.00012275862068965518,
|
| 142 |
-
"loss":
|
| 143 |
"step": 90
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.13153340256143994,
|
| 147 |
-
"grad_norm":
|
| 148 |
"learning_rate": 0.0001296551724137931,
|
| 149 |
-
"loss":
|
| 150 |
"step": 95
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.13845621322256838,
|
| 154 |
-
"grad_norm":
|
| 155 |
"learning_rate": 0.00013655172413793104,
|
| 156 |
-
"loss":
|
| 157 |
"step": 100
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.14537902388369678,
|
| 161 |
-
"grad_norm":
|
| 162 |
"learning_rate": 0.00014344827586206896,
|
| 163 |
-
"loss":
|
| 164 |
"step": 105
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.1523018345448252,
|
| 168 |
-
"grad_norm":
|
| 169 |
"learning_rate": 0.0001503448275862069,
|
| 170 |
-
"loss":
|
| 171 |
"step": 110
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.1592246452059536,
|
| 175 |
-
"grad_norm":
|
| 176 |
"learning_rate": 0.00015724137931034485,
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 115
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.16614745586708204,
|
| 182 |
-
"grad_norm": 2.
|
| 183 |
"learning_rate": 0.00016413793103448276,
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 120
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.17307026652821045,
|
| 189 |
-
"grad_norm":
|
| 190 |
"learning_rate": 0.0001710344827586207,
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 125
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.17999307718933888,
|
| 196 |
-
"grad_norm":
|
| 197 |
"learning_rate": 0.00017793103448275862,
|
| 198 |
-
"loss":
|
| 199 |
"step": 130
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.18691588785046728,
|
| 203 |
-
"grad_norm":
|
| 204 |
"learning_rate": 0.00018482758620689654,
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 135
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.19383869851159571,
|
| 210 |
-
"grad_norm":
|
| 211 |
"learning_rate": 0.0001917241379310345,
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 140
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.20076150917272412,
|
| 217 |
-
"grad_norm": 2.
|
| 218 |
"learning_rate": 0.00019862068965517243,
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 145
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.20768431983385255,
|
| 224 |
-
"grad_norm": 1.
|
| 225 |
"learning_rate": 0.0001999953208384105,
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 150
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.21460713049498095,
|
| 231 |
-
"grad_norm":
|
| 232 |
"learning_rate": 0.0001999763124949378,
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 155
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.22152994115610938,
|
| 238 |
-
"grad_norm": 1.
|
| 239 |
"learning_rate": 0.0001999426852993044,
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 160
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.2284527518172378,
|
| 245 |
-
"grad_norm":
|
| 246 |
"learning_rate": 0.00019989444416858502,
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 165
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.23537556247836622,
|
| 252 |
-
"grad_norm":
|
| 253 |
"learning_rate": 0.0001998315961567502,
|
| 254 |
-
"loss":
|
| 255 |
"step": 170
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.24229837313949462,
|
| 259 |
-
"grad_norm":
|
| 260 |
"learning_rate": 0.00019975415045363467,
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 175
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.24922118380062305,
|
| 266 |
-
"grad_norm":
|
| 267 |
"learning_rate": 0.00019966211838359378,
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 180
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.25614399446175146,
|
| 273 |
-
"grad_norm":
|
| 274 |
"learning_rate": 0.00019955551340384743,
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 185
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.2630668051228799,
|
| 280 |
-
"grad_norm":
|
| 281 |
"learning_rate": 0.00019943435110251248,
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 190
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.2699896157840083,
|
| 287 |
-
"grad_norm":
|
| 288 |
"learning_rate": 0.0001992986491963232,
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 195
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.27691242644513675,
|
| 294 |
-
"grad_norm":
|
| 295 |
"learning_rate": 0.00019914842752804103,
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 200
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.2838352371062651,
|
| 301 |
-
"grad_norm": 1.
|
| 302 |
"learning_rate": 0.00019898370806355263,
|
| 303 |
-
"loss": 0.
|
| 304 |
"step": 205
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.29075804776739356,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
"learning_rate": 0.00019880451488865836,
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 210
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.297680858428522,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
"learning_rate": 0.00019861087420555018,
|
| 317 |
-
"loss": 0.
|
| 318 |
"step": 215
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.3046036690896504,
|
| 322 |
-
"grad_norm":
|
| 323 |
"learning_rate": 0.0001984028143289803,
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 220
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.3115264797507788,
|
| 329 |
-
"grad_norm":
|
| 330 |
"learning_rate": 0.00019818036568212106,
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 225
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.3184492904119072,
|
| 336 |
-
"grad_norm":
|
| 337 |
"learning_rate": 0.00019794356079211604,
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 230
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.32537210107303566,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
"learning_rate": 0.00019769243428532422,
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 235
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.3322949117341641,
|
| 350 |
-
"grad_norm":
|
| 351 |
"learning_rate": 0.00019742702288225652,
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 240
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.33921772239529246,
|
| 357 |
-
"grad_norm":
|
| 358 |
"learning_rate": 0.00019714736539220648,
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 245
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.3461405330564209,
|
| 364 |
-
"grad_norm":
|
| 365 |
"learning_rate": 0.00019685350270757555,
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 250
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.3530633437175493,
|
| 371 |
-
"grad_norm":
|
| 372 |
"learning_rate": 0.0001965454777978936,
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 255
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.35998615437867776,
|
| 378 |
-
"grad_norm": 1.
|
| 379 |
"learning_rate": 0.00019622333570353567,
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 260
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.36690896503980613,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
"learning_rate": 0.00019588712352913625,
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 265
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.37383177570093457,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
"learning_rate": 0.00019553689043670127,
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 270
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.380754586362063,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
"learning_rate": 0.00019517268763841962,
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 275
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.38767739702319143,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
"learning_rate": 0.00019479456838917473,
|
| 408 |
-
"loss": 0.
|
| 409 |
"step": 280
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.39460020768431986,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
"learning_rate": 0.0001944025879787574,
|
| 415 |
-
"loss": 0.
|
| 416 |
"step": 285
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.40152301834544823,
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
"learning_rate": 0.0001939968037237812,
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 290
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.40844582900657667,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
"learning_rate": 0.00019357727495930146,
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 295
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.4153686396677051,
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
"learning_rate": 0.00019314406303013904,
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 300
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.42229145032883353,
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
"learning_rate": 0.00019269723128191048,
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 305
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.4292142609899619,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
"learning_rate": 0.00019223684505176517,
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 310
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.43613707165109034,
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
"learning_rate": 0.00019176297165883165,
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 315
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.44305988231221877,
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
"learning_rate": 0.00019127568039437406,
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 320
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.4499826929733472,
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
"learning_rate": 0.00019077504251166006,
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 325
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.4569055036344756,
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
"learning_rate": 0.00019026113121554204,
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 330
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.463828314295604,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
"learning_rate": 0.00018973402165175268,
|
| 485 |
-
"loss": 0.
|
| 486 |
"step": 335
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.47075112495673244,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
"learning_rate": 0.0001891937908959172,
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 340
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.47767393561786087,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
"learning_rate": 0.00018864051794228282,
|
| 499 |
-
"loss": 0.
|
| 500 |
"step": 345
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.48459674627898924,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
"learning_rate": 0.00018807428369216822,
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 350
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.4915195569401177,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
"learning_rate": 0.00018749517094213375,
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 355
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.4984423676012461,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
"learning_rate": 0.00018690326437187472,
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 360
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.5053651782623745,
|
| 525 |
-
"grad_norm":
|
| 526 |
"learning_rate": 0.00018629865053183911,
|
| 527 |
-
"loss": 0.
|
| 528 |
"step": 365
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.5122879889235029,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
"learning_rate": 0.0001856814178305722,
|
| 534 |
-
"loss": 0.
|
| 535 |
"step": 370
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.5192107995846313,
|
| 539 |
-
"grad_norm": 0.
|
| 540 |
"learning_rate": 0.00018505165652178893,
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 375
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.5261336102457598,
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
"learning_rate": 0.00018440945869117675,
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 380
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.5330564209068882,
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
"learning_rate": 0.00018375491824293067,
|
| 555 |
-
"loss": 0.
|
| 556 |
"step": 385
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.5399792315680166,
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
"learning_rate": 0.00018308813088602227,
|
| 562 |
-
"loss": 0.
|
| 563 |
"step": 390
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.5469020422291451,
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
"learning_rate": 0.00018240919412020466,
|
| 569 |
-
"loss": 0.
|
| 570 |
"step": 395
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.5538248528902735,
|
| 574 |
-
"grad_norm":
|
| 575 |
"learning_rate": 0.00018171820722175604,
|
| 576 |
-
"loss": 0.
|
| 577 |
"step": 400
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.5607476635514018,
|
| 581 |
-
"grad_norm": 0.
|
| 582 |
"learning_rate": 0.00018101527122896293,
|
| 583 |
-
"loss": 0.
|
| 584 |
"step": 405
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.5676704742125303,
|
| 588 |
-
"grad_norm": 0.
|
| 589 |
"learning_rate": 0.0001803004889273463,
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 410
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.5745932848736587,
|
| 595 |
-
"grad_norm": 0.
|
| 596 |
"learning_rate": 0.0001795739648346317,
|
| 597 |
-
"loss": 0.
|
| 598 |
"step": 415
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.5815160955347871,
|
| 602 |
-
"grad_norm":
|
| 603 |
"learning_rate": 0.00017883580518546647,
|
| 604 |
-
"loss": 0.
|
| 605 |
"step": 420
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.5884389061959155,
|
| 609 |
-
"grad_norm":
|
| 610 |
"learning_rate": 0.00017808611791588584,
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 425
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.595361716857044,
|
| 616 |
-
"grad_norm": 0.
|
| 617 |
"learning_rate": 0.00017732501264752996,
|
| 618 |
-
"loss": 0.
|
| 619 |
"step": 430
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.6022845275181724,
|
| 623 |
-
"grad_norm": 0.
|
| 624 |
"learning_rate": 0.00017655260067161503,
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 435
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.6092073381793008,
|
| 630 |
-
"grad_norm": 0.
|
| 631 |
"learning_rate": 0.00017576899493265954,
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 440
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.6161301488404292,
|
| 637 |
-
"grad_norm": 0.
|
| 638 |
"learning_rate": 0.00017497431001196943,
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 445
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.6230529595015576,
|
| 644 |
-
"grad_norm": 0.
|
| 645 |
"learning_rate": 0.00017416866211088366,
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 450
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.629975770162686,
|
| 651 |
-
"grad_norm": 0.
|
| 652 |
"learning_rate": 0.00017335216903378267,
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 455
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.6368985808238145,
|
| 658 |
-
"grad_norm": 0.
|
| 659 |
"learning_rate": 0.00017252495017086296,
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 460
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.6438213914849429,
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
"learning_rate": 0.0001716871264806791,
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 465
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.6507442021460713,
|
| 672 |
-
"grad_norm":
|
| 673 |
"learning_rate": 0.0001708388204724572,
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 470
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.6576670128071997,
|
| 679 |
-
"grad_norm": 0.
|
| 680 |
"learning_rate": 0.00016998015618818087,
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 475
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.6645898234683282,
|
| 686 |
-
"grad_norm": 0.
|
| 687 |
"learning_rate": 0.00016911125918445363,
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 480
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.6715126341294566,
|
| 693 |
-
"grad_norm": 0.
|
| 694 |
"learning_rate": 0.00016823225651413953,
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 485
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.6784354447905849,
|
| 700 |
-
"grad_norm": 0.
|
| 701 |
"learning_rate": 0.00016734327670778508,
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 490
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.6853582554517134,
|
| 707 |
-
"grad_norm": 0.
|
| 708 |
"learning_rate": 0.00016644444975482523,
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 495
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.6922810661128418,
|
| 714 |
-
"grad_norm": 0.
|
| 715 |
"learning_rate": 0.0001655359070845757,
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 500
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.6992038767739702,
|
| 721 |
-
"grad_norm": 0.
|
| 722 |
"learning_rate": 0.0001646177815470151,
|
| 723 |
-
"loss": 0.
|
| 724 |
"step": 505
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.7061266874350987,
|
| 728 |
-
"grad_norm": 0.
|
| 729 |
"learning_rate": 0.00016369020739335924,
|
| 730 |
-
"loss": 0.
|
| 731 |
"step": 510
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.7130494980962271,
|
| 735 |
-
"grad_norm": 0.
|
| 736 |
"learning_rate": 0.00016275332025643028,
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 515
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.7199723087573555,
|
| 742 |
-
"grad_norm": 0.
|
| 743 |
"learning_rate": 0.00016180725713082434,
|
| 744 |
-
"loss": 0.
|
| 745 |
"step": 520
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.726895119418484,
|
| 749 |
-
"grad_norm": 0.
|
| 750 |
"learning_rate": 0.0001608521563528795,
|
| 751 |
-
"loss": 0.
|
| 752 |
"step": 525
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.7338179300796123,
|
| 756 |
-
"grad_norm": 0.
|
| 757 |
"learning_rate": 0.00015988815758044792,
|
| 758 |
-
"loss": 0.
|
| 759 |
"step": 530
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.7407407407407407,
|
| 763 |
-
"grad_norm": 0.
|
| 764 |
"learning_rate": 0.00015891540177247467,
|
| 765 |
-
"loss": 0.
|
| 766 |
"step": 535
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.7476635514018691,
|
| 770 |
-
"grad_norm": 0.
|
| 771 |
"learning_rate": 0.00015793403116838613,
|
| 772 |
-
"loss": 0.
|
| 773 |
"step": 540
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.7545863620629976,
|
| 777 |
-
"grad_norm": 0.
|
| 778 |
"learning_rate": 0.00015694418926729146,
|
| 779 |
-
"loss": 0.
|
| 780 |
"step": 545
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.761509172724126,
|
| 784 |
-
"grad_norm": 0.
|
| 785 |
"learning_rate": 0.00015594602080699956,
|
| 786 |
-
"loss": 0.
|
| 787 |
"step": 550
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.7684319833852544,
|
| 791 |
-
"grad_norm": 0.
|
| 792 |
"learning_rate": 0.00015493967174285512,
|
| 793 |
-
"loss": 0.
|
| 794 |
"step": 555
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.7753547940463829,
|
| 798 |
-
"grad_norm": 0.
|
| 799 |
"learning_rate": 0.00015392528922639662,
|
| 800 |
-
"loss": 0.
|
| 801 |
"step": 560
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.7822776047075113,
|
| 805 |
-
"grad_norm": 0.
|
| 806 |
"learning_rate": 0.00015290302158383923,
|
| 807 |
-
"loss": 0.
|
| 808 |
"step": 565
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.7892004153686397,
|
| 812 |
-
"grad_norm": 0.
|
| 813 |
"learning_rate": 0.0001518730182943862,
|
| 814 |
-
"loss": 0.
|
| 815 |
"step": 570
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.796123226029768,
|
| 819 |
-
"grad_norm": 0.
|
| 820 |
"learning_rate": 0.0001508354299683715,
|
| 821 |
-
"loss": 0.
|
| 822 |
"step": 575
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 0.8030460366908965,
|
| 826 |
-
"grad_norm": 0.
|
| 827 |
"learning_rate": 0.0001497904083252371,
|
| 828 |
-
"loss": 0.
|
| 829 |
"step": 580
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 0.8099688473520249,
|
| 833 |
-
"grad_norm": 0.
|
| 834 |
"learning_rate": 0.00014873810617134806,
|
| 835 |
-
"loss": 0.
|
| 836 |
"step": 585
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.8168916580131533,
|
| 840 |
-
"grad_norm": 0.
|
| 841 |
"learning_rate": 0.00014767867737764876,
|
| 842 |
-
"loss": 0.
|
| 843 |
"step": 590
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 0.8238144686742818,
|
| 847 |
-
"grad_norm": 0.
|
| 848 |
"learning_rate": 0.0001466122768571633,
|
| 849 |
-
"loss": 0.
|
| 850 |
"step": 595
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 0.8307372793354102,
|
| 854 |
-
"grad_norm": 0.
|
| 855 |
"learning_rate": 0.0001455390605423437,
|
| 856 |
-
"loss": 0.
|
| 857 |
"step": 600
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.8376600899965386,
|
| 861 |
-
"grad_norm": 0.
|
| 862 |
"learning_rate": 0.00014445918536226904,
|
| 863 |
-
"loss": 0.
|
| 864 |
"step": 605
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.8445829006576671,
|
| 868 |
-
"grad_norm": 0.
|
| 869 |
"learning_rate": 0.00014337280921969858,
|
| 870 |
-
"loss": 0.
|
| 871 |
"step": 610
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 0.8515057113187954,
|
| 875 |
-
"grad_norm": 0.
|
| 876 |
"learning_rate": 0.0001422800909679829,
|
| 877 |
-
"loss": 0.
|
| 878 |
"step": 615
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.8584285219799238,
|
| 882 |
-
"grad_norm": 0.
|
| 883 |
"learning_rate": 0.00014118119038783588,
|
| 884 |
-
"loss": 0.
|
| 885 |
"step": 620
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 0.8653513326410522,
|
| 889 |
-
"grad_norm": 0.
|
| 890 |
"learning_rate": 0.00014007626816397087,
|
| 891 |
-
"loss": 0.
|
| 892 |
"step": 625
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 0.8722741433021807,
|
| 896 |
-
"grad_norm": 0.
|
| 897 |
"learning_rate": 0.00013896548586160503,
|
| 898 |
-
"loss": 0.
|
| 899 |
"step": 630
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 0.8791969539633091,
|
| 903 |
-
"grad_norm": 0.
|
| 904 |
"learning_rate": 0.00013784900590283473,
|
| 905 |
-
"loss": 0.
|
| 906 |
"step": 635
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 0.8861197646244375,
|
| 910 |
-
"grad_norm": 0.
|
| 911 |
"learning_rate": 0.00013672699154288569,
|
| 912 |
-
"loss": 0.
|
| 913 |
"step": 640
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 0.893042575285566,
|
| 917 |
-
"grad_norm": 0.
|
| 918 |
"learning_rate": 0.00013559960684624112,
|
| 919 |
-
"loss": 0.
|
| 920 |
"step": 645
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 0.8999653859466944,
|
| 924 |
-
"grad_norm": 0.
|
| 925 |
"learning_rate": 0.0001344670166626519,
|
| 926 |
-
"loss": 0.
|
| 927 |
"step": 650
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.9068881966078228,
|
| 931 |
-
"grad_norm": 0.
|
| 932 |
"learning_rate": 0.0001333293866030317,
|
| 933 |
-
"loss": 0.
|
| 934 |
"step": 655
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 0.9138110072689511,
|
| 938 |
-
"grad_norm": 0.
|
| 939 |
"learning_rate": 0.0001321868830152407,
|
| 940 |
-
"loss": 0.
|
| 941 |
"step": 660
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.9207338179300796,
|
| 945 |
-
"grad_norm": 0.
|
| 946 |
"learning_rate": 0.00013103967295976179,
|
| 947 |
-
"loss": 0.
|
| 948 |
"step": 665
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 0.927656628591208,
|
| 952 |
-
"grad_norm": 0.
|
| 953 |
"learning_rate": 0.0001298879241852724,
|
| 954 |
-
"loss": 0.
|
| 955 |
"step": 670
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 0.9345794392523364,
|
| 959 |
-
"grad_norm": 0.
|
| 960 |
"learning_rate": 0.0001287318051041159,
|
| 961 |
-
"loss": 0.
|
| 962 |
"step": 675
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.9415022499134649,
|
| 966 |
-
"grad_norm": 0.
|
| 967 |
"learning_rate": 0.00012757148476767553,
|
| 968 |
-
"loss": 0.
|
| 969 |
"step": 680
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.9484250605745933,
|
| 973 |
-
"grad_norm": 0.
|
| 974 |
"learning_rate": 0.0001264071328416555,
|
| 975 |
-
"loss": 0.
|
| 976 |
"step": 685
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 0.9553478712357217,
|
| 980 |
-
"grad_norm": 0.
|
| 981 |
"learning_rate": 0.00012523891958127168,
|
| 982 |
-
"loss": 0.
|
| 983 |
"step": 690
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 0.9622706818968502,
|
| 987 |
-
"grad_norm": 0.
|
| 988 |
"learning_rate": 0.0001240670158063565,
|
| 989 |
-
"loss": 0.
|
| 990 |
"step": 695
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.9691934925579785,
|
| 994 |
-
"grad_norm": 0.
|
| 995 |
"learning_rate": 0.00012289159287638105,
|
| 996 |
-
"loss": 0.
|
| 997 |
"step": 700
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 0.9761163032191069,
|
| 1001 |
-
"grad_norm": 0.
|
| 1002 |
"learning_rate": 0.00012171282266539845,
|
| 1003 |
-
"loss": 0.
|
| 1004 |
"step": 705
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 0.9830391138802353,
|
| 1008 |
-
"grad_norm": 0.
|
| 1009 |
"learning_rate": 0.00012053087753691172,
|
| 1010 |
-
"loss": 0.
|
| 1011 |
"step": 710
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 0.9899619245413638,
|
| 1015 |
-
"grad_norm": 0.
|
| 1016 |
"learning_rate": 0.00011934593031867038,
|
| 1017 |
-
"loss": 0.
|
| 1018 |
"step": 715
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 0.9968847352024922,
|
| 1022 |
-
"grad_norm": 0.
|
| 1023 |
"learning_rate": 0.00011815815427739911,
|
| 1024 |
-
"loss": 0.
|
| 1025 |
"step": 720
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 1.0,
|
| 1029 |
-
"eval_loss": 0.
|
| 1030 |
-
"eval_runtime": 75.
|
| 1031 |
-
"eval_samples_per_second": 27.
|
| 1032 |
-
"eval_steps_per_second": 1.
|
| 1033 |
"step": 723
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 1.0027691242644514,
|
| 1037 |
-
"grad_norm": 0.
|
| 1038 |
"learning_rate": 0.00011696772309346182,
|
| 1039 |
-
"loss": 0.
|
| 1040 |
"step": 725
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
"epoch": 1.0096919349255797,
|
| 1044 |
-
"grad_norm": 0.
|
| 1045 |
"learning_rate": 0.0001157748108354659,
|
| 1046 |
-
"loss": 0.
|
| 1047 |
"step": 730
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
"epoch": 1.0166147455867083,
|
| 1051 |
-
"grad_norm": 0.
|
| 1052 |
"learning_rate": 0.00011457959193480925,
|
| 1053 |
-
"loss": 0.
|
| 1054 |
"step": 735
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
"epoch": 1.0235375562478366,
|
| 1058 |
-
"grad_norm": 0.
|
| 1059 |
"learning_rate": 0.00011338224116017423,
|
| 1060 |
-
"loss": 0.
|
| 1061 |
"step": 740
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
"epoch": 1.0304603669089651,
|
| 1065 |
-
"grad_norm": 0.
|
| 1066 |
"learning_rate": 0.00011218293359197267,
|
| 1067 |
-
"loss": 0.
|
| 1068 |
"step": 745
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
"epoch": 1.0373831775700935,
|
| 1072 |
-
"grad_norm": 0.
|
| 1073 |
"learning_rate": 0.00011098184459674484,
|
| 1074 |
-
"loss": 0.
|
| 1075 |
"step": 750
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
"epoch": 1.0443059882312218,
|
| 1079 |
-
"grad_norm": 0.
|
| 1080 |
"learning_rate": 0.00010977914980151691,
|
| 1081 |
-
"loss": 0.
|
| 1082 |
"step": 755
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
"epoch": 1.0512287988923503,
|
| 1086 |
-
"grad_norm": 0.
|
| 1087 |
"learning_rate": 0.00010857502506812029,
|
| 1088 |
-
"loss": 0.
|
| 1089 |
"step": 760
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 1.0581516095534786,
|
| 1093 |
-
"grad_norm": 0.
|
| 1094 |
"learning_rate": 0.00010736964646747629,
|
| 1095 |
-
"loss": 0.
|
| 1096 |
"step": 765
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
"epoch": 1.0650744202146072,
|
| 1100 |
-
"grad_norm": 0.
|
| 1101 |
"learning_rate": 0.00010616319025385089,
|
| 1102 |
-
"loss": 0.
|
| 1103 |
"step": 770
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
"epoch": 1.0719972308757355,
|
| 1107 |
-
"grad_norm": 0.
|
| 1108 |
"learning_rate": 0.00010495583283908177,
|
| 1109 |
-
"loss": 0.
|
| 1110 |
"step": 775
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
"epoch": 1.078920041536864,
|
| 1114 |
-
"grad_norm": 0.
|
| 1115 |
"learning_rate": 0.00010374775076678332,
|
| 1116 |
-
"loss": 0.
|
| 1117 |
"step": 780
|
| 1118 |
},
|
| 1119 |
{
|
| 1120 |
"epoch": 1.0858428521979924,
|
| 1121 |
-
"grad_norm": 0.
|
| 1122 |
"learning_rate": 0.00010253912068653146,
|
| 1123 |
-
"loss": 0.
|
| 1124 |
"step": 785
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
"epoch": 1.092765662859121,
|
| 1128 |
-
"grad_norm": 0.
|
| 1129 |
"learning_rate": 0.00010133011932803378,
|
| 1130 |
-
"loss": 0.
|
| 1131 |
"step": 790
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
"epoch": 1.0996884735202492,
|
| 1135 |
-
"grad_norm": 0.
|
| 1136 |
"learning_rate": 0.00010012092347528705,
|
| 1137 |
-
"loss": 0.
|
| 1138 |
"step": 795
|
| 1139 |
},
|
| 1140 |
{
|
| 1141 |
"epoch": 1.1066112841813776,
|
| 1142 |
-
"grad_norm": 0.
|
| 1143 |
"learning_rate": 9.891170994072793e-05,
|
| 1144 |
-
"loss": 0.
|
| 1145 |
"step": 800
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"epoch": 1.113534094842506,
|
| 1149 |
-
"grad_norm": 0.
|
| 1150 |
"learning_rate": 9.770265553937831e-05,
|
| 1151 |
-
"loss": 0.
|
| 1152 |
"step": 805
|
| 1153 |
},
|
| 1154 |
{
|
| 1155 |
"epoch": 1.1204569055036344,
|
| 1156 |
-
"grad_norm": 0.
|
| 1157 |
"learning_rate": 9.649393706299127e-05,
|
| 1158 |
-
"loss": 0.
|
| 1159 |
"step": 810
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
"epoch": 1.127379716164763,
|
| 1163 |
-
"grad_norm": 0.
|
| 1164 |
"learning_rate": 9.52857312541998e-05,
|
| 1165 |
-
"loss": 0.
|
| 1166 |
"step": 815
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
"epoch": 1.1343025268258913,
|
| 1170 |
-
"grad_norm": 0.
|
| 1171 |
"learning_rate": 9.407821478067304e-05,
|
| 1172 |
-
"loss": 0.
|
| 1173 |
"step": 820
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
"epoch": 1.1412253374870198,
|
| 1177 |
-
"grad_norm": 0.
|
| 1178 |
"learning_rate": 9.287156420928342e-05,
|
| 1179 |
-
"loss": 0.
|
| 1180 |
"step": 825
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
"epoch": 1.1481481481481481,
|
| 1184 |
-
"grad_norm": 0.
|
| 1185 |
"learning_rate": 9.166595598028832e-05,
|
| 1186 |
-
"loss": 0.
|
| 1187 |
"step": 830
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
"epoch": 1.1550709588092767,
|
| 1191 |
-
"grad_norm": 0.
|
| 1192 |
"learning_rate": 9.046156638153056e-05,
|
| 1193 |
-
"loss": 0.
|
| 1194 |
"step": 835
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
"epoch": 1.161993769470405,
|
| 1198 |
-
"grad_norm": 0.
|
| 1199 |
"learning_rate": 8.925857152266103e-05,
|
| 1200 |
-
"loss": 0.
|
| 1201 |
"step": 840
|
| 1202 |
},
|
| 1203 |
{
|
| 1204 |
"epoch": 1.1689165801315333,
|
| 1205 |
-
"grad_norm": 0.
|
| 1206 |
"learning_rate": 8.805714730938728e-05,
|
| 1207 |
-
"loss": 0.
|
| 1208 |
"step": 845
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
"epoch": 1.1758393907926619,
|
| 1212 |
-
"grad_norm": 0.
|
| 1213 |
"learning_rate": 8.685746941775219e-05,
|
| 1214 |
-
"loss": 0.
|
| 1215 |
"step": 850
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
"epoch": 1.1827622014537902,
|
| 1219 |
-
"grad_norm": 0.
|
| 1220 |
"learning_rate": 8.565971326844584e-05,
|
| 1221 |
-
"loss": 0.
|
| 1222 |
"step": 855
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
"epoch": 1.1896850121149187,
|
| 1226 |
-
"grad_norm": 0.
|
| 1227 |
"learning_rate": 8.44640540011553e-05,
|
| 1228 |
-
"loss": 0.
|
| 1229 |
"step": 860
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
"epoch": 1.196607822776047,
|
| 1233 |
-
"grad_norm": 0.
|
| 1234 |
"learning_rate": 8.327066644895469e-05,
|
| 1235 |
-
"loss": 0.
|
| 1236 |
"step": 865
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
"epoch": 1.2035306334371756,
|
| 1240 |
-
"grad_norm": 0.
|
| 1241 |
"learning_rate": 8.207972511274095e-05,
|
| 1242 |
-
"loss": 0.
|
| 1243 |
"step": 870
|
| 1244 |
},
|
| 1245 |
{
|
| 1246 |
"epoch": 1.210453444098304,
|
| 1247 |
-
"grad_norm": 0.
|
| 1248 |
"learning_rate": 8.089140413571747e-05,
|
| 1249 |
-
"loss": 0.
|
| 1250 |
"step": 875
|
| 1251 |
},
|
| 1252 |
{
|
| 1253 |
"epoch": 1.2173762547594325,
|
| 1254 |
-
"grad_norm": 0.
|
| 1255 |
"learning_rate": 7.970587727793051e-05,
|
| 1256 |
-
"loss": 0.
|
| 1257 |
"step": 880
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"epoch": 1.2242990654205608,
|
| 1261 |
-
"grad_norm": 0.
|
| 1262 |
"learning_rate": 7.852331789086135e-05,
|
| 1263 |
-
"loss": 0.
|
| 1264 |
"step": 885
|
| 1265 |
},
|
| 1266 |
{
|
| 1267 |
"epoch": 1.231221876081689,
|
| 1268 |
-
"grad_norm": 0.
|
| 1269 |
"learning_rate": 7.73438988920784e-05,
|
| 1270 |
-
"loss": 0.
|
| 1271 |
"step": 890
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
"epoch": 1.2381446867428176,
|
| 1275 |
-
"grad_norm": 0.
|
| 1276 |
"learning_rate": 7.616779273995252e-05,
|
| 1277 |
-
"loss": 0.
|
| 1278 |
"step": 895
|
| 1279 |
},
|
| 1280 |
{
|
| 1281 |
"epoch": 1.245067497403946,
|
| 1282 |
-
"grad_norm": 0.
|
| 1283 |
"learning_rate": 7.499517140843968e-05,
|
| 1284 |
-
"loss": 0.
|
| 1285 |
"step": 900
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 1.2519903080650745,
|
| 1289 |
-
"grad_norm": 0.
|
| 1290 |
"learning_rate": 7.382620636193438e-05,
|
| 1291 |
-
"loss": 0.
|
| 1292 |
"step": 905
|
| 1293 |
},
|
| 1294 |
{
|
| 1295 |
"epoch": 1.2589131187262028,
|
| 1296 |
-
"grad_norm": 0.
|
| 1297 |
"learning_rate": 7.266106853019759e-05,
|
| 1298 |
-
"loss": 0.
|
| 1299 |
"step": 910
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
"epoch": 1.2658359293873311,
|
| 1303 |
-
"grad_norm": 0.
|
| 1304 |
"learning_rate": 7.149992828336274e-05,
|
| 1305 |
-
"loss": 0.
|
| 1306 |
"step": 915
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
"epoch": 1.2727587400484597,
|
| 1310 |
-
"grad_norm": 0.
|
| 1311 |
"learning_rate": 7.034295540702397e-05,
|
| 1312 |
-
"loss": 0.
|
| 1313 |
"step": 920
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
"epoch": 1.2796815507095882,
|
| 1317 |
-
"grad_norm": 0.
|
| 1318 |
"learning_rate": 6.919031907740922e-05,
|
| 1319 |
-
"loss": 0.
|
| 1320 |
"step": 925
|
| 1321 |
},
|
| 1322 |
{
|
| 1323 |
"epoch": 1.2866043613707165,
|
| 1324 |
-
"grad_norm": 0.
|
| 1325 |
"learning_rate": 6.804218783664288e-05,
|
| 1326 |
-
"loss": 0.
|
| 1327 |
"step": 930
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
"epoch": 1.2935271720318449,
|
| 1331 |
-
"grad_norm": 0.
|
| 1332 |
"learning_rate": 6.689872956810102e-05,
|
| 1333 |
-
"loss": 0.
|
| 1334 |
"step": 935
|
| 1335 |
},
|
| 1336 |
{
|
| 1337 |
"epoch": 1.3004499826929734,
|
| 1338 |
-
"grad_norm": 0.
|
| 1339 |
"learning_rate": 6.5760111471863e-05,
|
| 1340 |
-
"loss": 0.
|
| 1341 |
"step": 940
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
"epoch": 1.3073727933541017,
|
| 1345 |
-
"grad_norm": 0.
|
| 1346 |
"learning_rate": 6.46265000402628e-05,
|
| 1347 |
-
"loss": 0.
|
| 1348 |
"step": 945
|
| 1349 |
},
|
| 1350 |
{
|
| 1351 |
"epoch": 1.3142956040152303,
|
| 1352 |
-
"grad_norm": 0.
|
| 1353 |
"learning_rate": 6.349806103354417e-05,
|
| 1354 |
-
"loss": 0.
|
| 1355 |
"step": 950
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 1.3212184146763586,
|
| 1359 |
-
"grad_norm": 0.
|
| 1360 |
"learning_rate": 6.23749594556226e-05,
|
| 1361 |
-
"loss": 0.
|
| 1362 |
"step": 955
|
| 1363 |
},
|
| 1364 |
{
|
| 1365 |
"epoch": 1.328141225337487,
|
| 1366 |
-
"grad_norm": 0.
|
| 1367 |
"learning_rate": 6.125735952995775e-05,
|
| 1368 |
-
"loss": 0.
|
| 1369 |
"step": 960
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 1.3350640359986155,
|
| 1373 |
-
"grad_norm": 0.
|
| 1374 |
"learning_rate": 6.0145424675540394e-05,
|
| 1375 |
-
"loss": 0.
|
| 1376 |
"step": 965
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
"epoch": 1.341986846659744,
|
| 1380 |
-
"grad_norm": 0.
|
| 1381 |
"learning_rate": 5.9039317482996735e-05,
|
| 1382 |
-
"loss": 0.
|
| 1383 |
"step": 970
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 1.3489096573208723,
|
| 1387 |
-
"grad_norm": 0.
|
| 1388 |
"learning_rate": 5.793919969081374e-05,
|
| 1389 |
-
"loss": 0.
|
| 1390 |
"step": 975
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 1.3558324679820006,
|
| 1394 |
-
"grad_norm": 0.
|
| 1395 |
"learning_rate": 5.68452321616894e-05,
|
| 1396 |
-
"loss": 0.
|
| 1397 |
"step": 980
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 1.3627552786431292,
|
| 1401 |
-
"grad_norm": 0.
|
| 1402 |
"learning_rate": 5.5757574859010886e-05,
|
| 1403 |
-
"loss": 0.
|
| 1404 |
"step": 985
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 1.3696780893042575,
|
| 1408 |
-
"grad_norm": 0.
|
| 1409 |
"learning_rate": 5.467638682346403e-05,
|
| 1410 |
-
"loss": 0.
|
| 1411 |
"step": 990
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 1.376600899965386,
|
| 1415 |
-
"grad_norm": 0.
|
| 1416 |
"learning_rate": 5.3601826149777966e-05,
|
| 1417 |
-
"loss": 0.
|
| 1418 |
"step": 995
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
"epoch": 1.3835237106265144,
|
| 1422 |
-
"grad_norm": 0.
|
| 1423 |
"learning_rate": 5.25340499636082e-05,
|
| 1424 |
"loss": 0.485,
|
| 1425 |
"step": 1000
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 1.3904465212876427,
|
| 1429 |
-
"grad_norm": 0.
|
| 1430 |
"learning_rate": 5.147321439856091e-05,
|
| 1431 |
-
"loss": 0.
|
| 1432 |
"step": 1005
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 1.3973693319487712,
|
| 1436 |
-
"grad_norm": 0.
|
| 1437 |
"learning_rate": 5.041947457336274e-05,
|
| 1438 |
-
"loss": 0.
|
| 1439 |
"step": 1010
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 1.4042921426098995,
|
| 1443 |
-
"grad_norm": 0.
|
| 1444 |
"learning_rate": 4.937298456917889e-05,
|
| 1445 |
-
"loss": 0.
|
| 1446 |
"step": 1015
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 1.411214953271028,
|
| 1450 |
-
"grad_norm": 0.
|
| 1451 |
"learning_rate": 4.8333897407082896e-05,
|
| 1452 |
-
"loss": 0.
|
| 1453 |
"step": 1020
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 1.4181377639321564,
|
| 1457 |
-
"grad_norm": 0.
|
| 1458 |
"learning_rate": 4.7302365025681206e-05,
|
| 1459 |
-
"loss": 0.
|
| 1460 |
"step": 1025
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 1.425060574593285,
|
| 1464 |
-
"grad_norm": 0.
|
| 1465 |
"learning_rate": 4.627853825889642e-05,
|
| 1466 |
-
"loss": 0.
|
| 1467 |
"step": 1030
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 1.4319833852544133,
|
| 1471 |
-
"grad_norm": 0.
|
| 1472 |
"learning_rate": 4.526256681391192e-05,
|
| 1473 |
-
"loss": 0.
|
| 1474 |
"step": 1035
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 1.4389061959155418,
|
| 1478 |
-
"grad_norm": 0.
|
| 1479 |
"learning_rate": 4.4254599249281016e-05,
|
| 1480 |
-
"loss": 0.
|
| 1481 |
"step": 1040
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 1.4458290065766701,
|
| 1485 |
-
"grad_norm": 0.
|
| 1486 |
"learning_rate": 4.3254782953204375e-05,
|
| 1487 |
-
"loss": 0.
|
| 1488 |
"step": 1045
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 1.4527518172377984,
|
| 1492 |
-
"grad_norm": 0.
|
| 1493 |
"learning_rate": 4.2263264121978505e-05,
|
| 1494 |
-
"loss": 0.
|
| 1495 |
"step": 1050
|
| 1496 |
},
|
| 1497 |
{
|
| 1498 |
"epoch": 1.459674627898927,
|
| 1499 |
-
"grad_norm": 0.
|
| 1500 |
"learning_rate": 4.12801877386183e-05,
|
| 1501 |
-
"loss": 0.
|
| 1502 |
"step": 1055
|
| 1503 |
},
|
| 1504 |
{
|
| 1505 |
"epoch": 1.4665974385600553,
|
| 1506 |
-
"grad_norm": 0.
|
| 1507 |
"learning_rate": 4.030569755165736e-05,
|
| 1508 |
-
"loss": 0.
|
| 1509 |
"step": 1060
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
"epoch": 1.4735202492211839,
|
| 1513 |
-
"grad_norm": 0.
|
| 1514 |
"learning_rate": 3.9339936054128466e-05,
|
| 1515 |
-
"loss": 0.
|
| 1516 |
"step": 1065
|
| 1517 |
},
|
| 1518 |
{
|
| 1519 |
"epoch": 1.4804430598823122,
|
| 1520 |
-
"grad_norm": 0.
|
| 1521 |
"learning_rate": 3.8383044462727826e-05,
|
| 1522 |
-
"loss": 0.
|
| 1523 |
"step": 1070
|
| 1524 |
},
|
| 1525 |
{
|
| 1526 |
"epoch": 1.4873658705434407,
|
| 1527 |
-
"grad_norm": 0.
|
| 1528 |
"learning_rate": 3.7435162697166107e-05,
|
| 1529 |
-
"loss": 0.
|
| 1530 |
"step": 1075
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 1.494288681204569,
|
| 1534 |
-
"grad_norm": 0.
|
| 1535 |
"learning_rate": 3.649642935970859e-05,
|
| 1536 |
-
"loss": 0.
|
| 1537 |
"step": 1080
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
"epoch": 1.5012114918656976,
|
| 1541 |
-
"grad_norm": 0.
|
| 1542 |
"learning_rate": 3.556698171490871e-05,
|
| 1543 |
-
"loss": 0.
|
| 1544 |
"step": 1085
|
| 1545 |
},
|
| 1546 |
{
|
| 1547 |
"epoch": 1.508134302526826,
|
| 1548 |
-
"grad_norm": 0.
|
| 1549 |
"learning_rate": 3.464695566953644e-05,
|
| 1550 |
-
"loss": 0.
|
| 1551 |
"step": 1090
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
"epoch": 1.5150571131879542,
|
| 1555 |
-
"grad_norm": 0.
|
| 1556 |
"learning_rate": 3.373648575270576e-05,
|
| 1557 |
-
"loss": 0.
|
| 1558 |
"step": 1095
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
"epoch": 1.5219799238490828,
|
| 1562 |
-
"grad_norm": 0.
|
| 1563 |
"learning_rate": 3.283570509620344e-05,
|
| 1564 |
-
"loss": 0.
|
| 1565 |
"step": 1100
|
| 1566 |
},
|
| 1567 |
{
|
| 1568 |
"epoch": 1.5289027345102113,
|
| 1569 |
-
"grad_norm": 0.
|
| 1570 |
"learning_rate": 3.1944745415021915e-05,
|
| 1571 |
-
"loss": 0.
|
| 1572 |
"step": 1105
|
| 1573 |
},
|
| 1574 |
{
|
| 1575 |
"epoch": 1.5358255451713396,
|
| 1576 |
-
"grad_norm": 0.
|
| 1577 |
"learning_rate": 3.106373698809958e-05,
|
| 1578 |
-
"loss": 0.
|
| 1579 |
"step": 1110
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
"epoch": 1.542748355832468,
|
| 1583 |
-
"grad_norm": 0.
|
| 1584 |
"learning_rate": 3.0192808639271065e-05,
|
| 1585 |
-
"loss": 0.
|
| 1586 |
"step": 1115
|
| 1587 |
},
|
| 1588 |
{
|
| 1589 |
"epoch": 1.5496711664935963,
|
| 1590 |
-
"grad_norm": 0.
|
| 1591 |
"learning_rate": 2.933208771843018e-05,
|
| 1592 |
-
"loss": 0.
|
| 1593 |
"step": 1120
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 1.5565939771547248,
|
| 1597 |
-
"grad_norm": 0.
|
| 1598 |
"learning_rate": 2.8481700082908268e-05,
|
| 1599 |
-
"loss": 0.
|
| 1600 |
"step": 1125
|
| 1601 |
},
|
| 1602 |
{
|
| 1603 |
"epoch": 1.5635167878158533,
|
| 1604 |
-
"grad_norm": 0.
|
| 1605 |
"learning_rate": 2.7641770079071127e-05,
|
| 1606 |
-
"loss": 0.
|
| 1607 |
"step": 1130
|
| 1608 |
},
|
| 1609 |
{
|
| 1610 |
"epoch": 1.5704395984769817,
|
| 1611 |
-
"grad_norm": 0.
|
| 1612 |
"learning_rate": 2.681242052413656e-05,
|
| 1613 |
-
"loss": 0.
|
| 1614 |
"step": 1135
|
| 1615 |
},
|
| 1616 |
{
|
| 1617 |
"epoch": 1.57736240913811,
|
| 1618 |
-
"grad_norm": 0.
|
| 1619 |
"learning_rate": 2.59937726882158e-05,
|
| 1620 |
-
"loss": 0.
|
| 1621 |
"step": 1140
|
| 1622 |
},
|
| 1623 |
{
|
| 1624 |
"epoch": 1.5842852197992385,
|
| 1625 |
-
"grad_norm": 0.
|
| 1626 |
"learning_rate": 2.5185946276580918e-05,
|
| 1627 |
-
"loss": 0.
|
| 1628 |
"step": 1145
|
| 1629 |
},
|
| 1630 |
{
|
| 1631 |
"epoch": 1.591208030460367,
|
| 1632 |
-
"grad_norm": 0.
|
| 1633 |
"learning_rate": 2.4389059412161087e-05,
|
| 1634 |
"loss": 0.4736,
|
| 1635 |
"step": 1150
|
| 1636 |
},
|
| 1637 |
{
|
| 1638 |
"epoch": 1.5981308411214954,
|
| 1639 |
-
"grad_norm": 0.
|
| 1640 |
"learning_rate": 2.3603228618270356e-05,
|
| 1641 |
-
"loss": 0.
|
| 1642 |
"step": 1155
|
| 1643 |
},
|
| 1644 |
{
|
| 1645 |
"epoch": 1.6050536517826237,
|
| 1646 |
-
"grad_norm": 0.
|
| 1647 |
"learning_rate": 2.2828568801569283e-05,
|
| 1648 |
-
"loss": 0.
|
| 1649 |
"step": 1160
|
| 1650 |
},
|
| 1651 |
{
|
| 1652 |
"epoch": 1.611976462443752,
|
| 1653 |
-
"grad_norm": 0.
|
| 1654 |
"learning_rate": 2.2065193235262782e-05,
|
| 1655 |
-
"loss": 0.
|
| 1656 |
"step": 1165
|
| 1657 |
},
|
| 1658 |
{
|
| 1659 |
"epoch": 1.6188992731048806,
|
| 1660 |
-
"grad_norm": 0.
|
| 1661 |
"learning_rate": 2.1313213542537058e-05,
|
| 1662 |
-
"loss": 0.
|
| 1663 |
"step": 1170
|
| 1664 |
},
|
| 1665 |
{
|
| 1666 |
"epoch": 1.6258220837660091,
|
| 1667 |
-
"grad_norm": 0.
|
| 1668 |
"learning_rate": 2.0572739680237717e-05,
|
| 1669 |
-
"loss": 0.
|
| 1670 |
"step": 1175
|
| 1671 |
},
|
| 1672 |
{
|
| 1673 |
"epoch": 1.6327448944271374,
|
| 1674 |
-
"grad_norm": 0.
|
| 1675 |
"learning_rate": 1.984387992279153e-05,
|
| 1676 |
-
"loss": 0.
|
| 1677 |
"step": 1180
|
| 1678 |
},
|
| 1679 |
{
|
| 1680 |
"epoch": 1.6396677050882658,
|
| 1681 |
-
"grad_norm": 0.
|
| 1682 |
"learning_rate": 1.912674084637409e-05,
|
| 1683 |
-
"loss": 0.
|
| 1684 |
"step": 1185
|
| 1685 |
},
|
| 1686 |
{
|
| 1687 |
"epoch": 1.6465905157493943,
|
| 1688 |
-
"grad_norm": 0.
|
| 1689 |
"learning_rate": 1.8421427313326046e-05,
|
| 1690 |
-
"loss": 0.
|
| 1691 |
"step": 1190
|
| 1692 |
},
|
| 1693 |
{
|
| 1694 |
"epoch": 1.6535133264105226,
|
| 1695 |
-
"grad_norm": 0.
|
| 1696 |
"learning_rate": 1.7728042456819704e-05,
|
| 1697 |
-
"loss": 0.
|
| 1698 |
"step": 1195
|
| 1699 |
},
|
| 1700 |
{
|
| 1701 |
"epoch": 1.6604361370716512,
|
| 1702 |
-
"grad_norm": 0.
|
| 1703 |
"learning_rate": 1.7046687665778717e-05,
|
| 1704 |
-
"loss": 0.
|
| 1705 |
"step": 1200
|
| 1706 |
},
|
| 1707 |
{
|
| 1708 |
"epoch": 1.6673589477327795,
|
| 1709 |
-
"grad_norm": 0.
|
| 1710 |
"learning_rate": 1.6377462570052438e-05,
|
| 1711 |
-
"loss": 0.
|
| 1712 |
"step": 1205
|
| 1713 |
},
|
| 1714 |
{
|
| 1715 |
"epoch": 1.6742817583939078,
|
| 1716 |
-
"grad_norm": 0.
|
| 1717 |
"learning_rate": 1.5720465025847987e-05,
|
| 1718 |
-
"loss": 0.
|
| 1719 |
"step": 1210
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
"epoch": 1.6812045690550363,
|
| 1723 |
-
"grad_norm": 0.
|
| 1724 |
"learning_rate": 1.5075791101421166e-05,
|
| 1725 |
-
"loss": 0.
|
| 1726 |
"step": 1215
|
| 1727 |
},
|
| 1728 |
{
|
| 1729 |
"epoch": 1.6881273797161649,
|
| 1730 |
-
"grad_norm": 0.
|
| 1731 |
"learning_rate": 1.4443535063029279e-05,
|
| 1732 |
-
"loss": 0.
|
| 1733 |
"step": 1220
|
| 1734 |
},
|
| 1735 |
{
|
| 1736 |
"epoch": 1.6950501903772932,
|
| 1737 |
-
"grad_norm": 0.
|
| 1738 |
"learning_rate": 1.3823789361147044e-05,
|
| 1739 |
-
"loss": 0.
|
| 1740 |
"step": 1225
|
| 1741 |
},
|
| 1742 |
{
|
| 1743 |
"epoch": 1.7019730010384215,
|
| 1744 |
-
"grad_norm": 0.
|
| 1745 |
"learning_rate": 1.3216644616948282e-05,
|
| 1746 |
-
"loss": 0.
|
| 1747 |
"step": 1230
|
| 1748 |
},
|
| 1749 |
{
|
| 1750 |
"epoch": 1.70889581169955,
|
| 1751 |
-
"grad_norm": 0.
|
| 1752 |
"learning_rate": 1.2622189609054979e-05,
|
| 1753 |
-
"loss": 0.
|
| 1754 |
"step": 1235
|
| 1755 |
},
|
| 1756 |
{
|
| 1757 |
"epoch": 1.7158186223606784,
|
| 1758 |
-
"grad_norm": 0.
|
| 1759 |
"learning_rate": 1.2040511260555831e-05,
|
| 1760 |
-
"loss": 0.
|
| 1761 |
"step": 1240
|
| 1762 |
},
|
| 1763 |
{
|
| 1764 |
"epoch": 1.722741433021807,
|
| 1765 |
-
"grad_norm": 0.
|
| 1766 |
"learning_rate": 1.1471694626296049e-05,
|
| 1767 |
-
"loss": 0.
|
| 1768 |
"step": 1245
|
| 1769 |
},
|
| 1770 |
{
|
| 1771 |
"epoch": 1.7296642436829353,
|
| 1772 |
-
"grad_norm": 0.
|
| 1773 |
"learning_rate": 1.0915822880440308e-05,
|
| 1774 |
-
"loss": 0.
|
| 1775 |
"step": 1250
|
| 1776 |
},
|
| 1777 |
{
|
| 1778 |
"epoch": 1.7365870543440636,
|
| 1779 |
-
"grad_norm": 0.
|
| 1780 |
"learning_rate": 1.037297730431085e-05,
|
| 1781 |
-
"loss": 0.
|
| 1782 |
"step": 1255
|
| 1783 |
},
|
| 1784 |
{
|
| 1785 |
"epoch": 1.7435098650051921,
|
| 1786 |
-
"grad_norm": 0.
|
| 1787 |
"learning_rate": 9.843237274502282e-06,
|
| 1788 |
-
"loss": 0.
|
| 1789 |
"step": 1260
|
| 1790 |
},
|
| 1791 |
{
|
| 1792 |
"epoch": 1.7504326756663207,
|
| 1793 |
-
"grad_norm": 0.
|
| 1794 |
"learning_rate": 9.326680251274777e-06,
|
| 1795 |
-
"loss": 0.
|
| 1796 |
"step": 1265
|
| 1797 |
},
|
| 1798 |
{
|
| 1799 |
"epoch": 1.757355486327449,
|
| 1800 |
-
"grad_norm": 0.
|
| 1801 |
"learning_rate": 8.823381767227667e-06,
|
| 1802 |
-
"loss": 0.
|
| 1803 |
"step": 1270
|
| 1804 |
},
|
| 1805 |
{
|
| 1806 |
"epoch": 1.7642782969885773,
|
| 1807 |
-
"grad_norm": 0.
|
| 1808 |
"learning_rate": 8.333415416254831e-06,
|
| 1809 |
-
"loss": 0.
|
| 1810 |
"step": 1275
|
| 1811 |
},
|
| 1812 |
{
|
| 1813 |
"epoch": 1.7712011076497056,
|
| 1814 |
-
"grad_norm": 0.
|
| 1815 |
"learning_rate": 7.856852842783547e-06,
|
| 1816 |
-
"loss": 0.
|
| 1817 |
"step": 1280
|
| 1818 |
},
|
| 1819 |
{
|
| 1820 |
"epoch": 1.7781239183108342,
|
| 1821 |
-
"grad_norm": 0.
|
| 1822 |
"learning_rate": 7.393763731298387e-06,
|
| 1823 |
-
"loss": 0.
|
| 1824 |
"step": 1285
|
| 1825 |
},
|
| 1826 |
{
|
| 1827 |
"epoch": 1.7850467289719627,
|
| 1828 |
-
"grad_norm": 0.
|
| 1829 |
"learning_rate": 6.944215796151765e-06,
|
| 1830 |
-
"loss": 0.
|
| 1831 |
"step": 1290
|
| 1832 |
},
|
| 1833 |
{
|
| 1834 |
"epoch": 1.791969539633091,
|
| 1835 |
-
"grad_norm": 0.
|
| 1836 |
"learning_rate": 6.5082747716625255e-06,
|
| 1837 |
-
"loss": 0.
|
| 1838 |
"step": 1295
|
| 1839 |
},
|
| 1840 |
{
|
| 1841 |
"epoch": 1.7988923502942193,
|
| 1842 |
-
"grad_norm": 0.
|
| 1843 |
"learning_rate": 6.086004402504098e-06,
|
| 1844 |
-
"loss": 0.
|
| 1845 |
"step": 1300
|
| 1846 |
},
|
| 1847 |
{
|
| 1848 |
"epoch": 1.8058151609553479,
|
| 1849 |
-
"grad_norm": 0.
|
| 1850 |
"learning_rate": 5.6774664343834495e-06,
|
| 1851 |
-
"loss": 0.
|
| 1852 |
"step": 1305
|
| 1853 |
},
|
| 1854 |
{
|
| 1855 |
"epoch": 1.8127379716164764,
|
| 1856 |
-
"grad_norm": 0.
|
| 1857 |
"learning_rate": 5.2827206050125585e-06,
|
| 1858 |
-
"loss": 0.
|
| 1859 |
"step": 1310
|
| 1860 |
},
|
| 1861 |
{
|
| 1862 |
"epoch": 1.8196607822776047,
|
| 1863 |
-
"grad_norm": 0.
|
| 1864 |
"learning_rate": 4.9018246353732755e-06,
|
| 1865 |
-
"loss": 0.
|
| 1866 |
"step": 1315
|
| 1867 |
},
|
| 1868 |
{
|
| 1869 |
"epoch": 1.826583592938733,
|
| 1870 |
-
"grad_norm": 0.
|
| 1871 |
"learning_rate": 4.534834221277262e-06,
|
| 1872 |
-
"loss": 0.
|
| 1873 |
"step": 1320
|
| 1874 |
},
|
| 1875 |
{
|
| 1876 |
"epoch": 1.8335064035998614,
|
| 1877 |
-
"grad_norm": 0.
|
| 1878 |
"learning_rate": 4.181803025221898e-06,
|
| 1879 |
-
"loss": 0.
|
| 1880 |
"step": 1325
|
| 1881 |
},
|
| 1882 |
{
|
| 1883 |
"epoch": 1.84042921426099,
|
| 1884 |
-
"grad_norm": 0.
|
| 1885 |
"learning_rate": 3.842782668543599e-06,
|
| 1886 |
-
"loss": 0.
|
| 1887 |
"step": 1330
|
| 1888 |
},
|
| 1889 |
{
|
| 1890 |
"epoch": 1.8473520249221185,
|
| 1891 |
-
"grad_norm": 0.
|
| 1892 |
"learning_rate": 3.5178227238696347e-06,
|
| 1893 |
-
"loss": 0.
|
| 1894 |
"step": 1335
|
| 1895 |
},
|
| 1896 |
{
|
| 1897 |
"epoch": 1.8542748355832468,
|
| 1898 |
-
"grad_norm": 0.
|
| 1899 |
"learning_rate": 3.2069707078694057e-06,
|
| 1900 |
-
"loss": 0.
|
| 1901 |
"step": 1340
|
| 1902 |
},
|
| 1903 |
{
|
| 1904 |
"epoch": 1.8611976462443751,
|
| 1905 |
-
"grad_norm": 0.
|
| 1906 |
"learning_rate": 2.9102720743064527e-06,
|
| 1907 |
-
"loss": 0.
|
| 1908 |
"step": 1345
|
| 1909 |
},
|
| 1910 |
{
|
| 1911 |
"epoch": 1.8681204569055037,
|
| 1912 |
-
"grad_norm": 0.
|
| 1913 |
"learning_rate": 2.6277702073919997e-06,
|
| 1914 |
-
"loss": 0.
|
| 1915 |
"step": 1350
|
| 1916 |
},
|
| 1917 |
{
|
| 1918 |
"epoch": 1.8750432675666322,
|
| 1919 |
-
"grad_norm": 0.
|
| 1920 |
"learning_rate": 2.3595064154412374e-06,
|
| 1921 |
-
"loss": 0.
|
| 1922 |
"step": 1355
|
| 1923 |
},
|
| 1924 |
{
|
| 1925 |
"epoch": 1.8819660782277605,
|
| 1926 |
-
"grad_norm": 0.
|
| 1927 |
"learning_rate": 2.1055199248330546e-06,
|
| 1928 |
-
"loss": 0.
|
| 1929 |
"step": 1360
|
| 1930 |
},
|
| 1931 |
{
|
| 1932 |
"epoch": 1.8888888888888888,
|
| 1933 |
-
"grad_norm": 0.
|
| 1934 |
"learning_rate": 1.8658478742742625e-06,
|
| 1935 |
-
"loss": 0.
|
| 1936 |
"step": 1365
|
| 1937 |
},
|
| 1938 |
{
|
| 1939 |
"epoch": 1.8958116995500172,
|
| 1940 |
-
"grad_norm": 0.
|
| 1941 |
"learning_rate": 1.6405253093690343e-06,
|
| 1942 |
-
"loss": 0.
|
| 1943 |
"step": 1370
|
| 1944 |
},
|
| 1945 |
{
|
| 1946 |
"epoch": 1.9027345102111457,
|
| 1947 |
-
"grad_norm": 0.
|
| 1948 |
"learning_rate": 1.429585177494419e-06,
|
| 1949 |
-
"loss": 0.
|
| 1950 |
"step": 1375
|
| 1951 |
},
|
| 1952 |
{
|
| 1953 |
"epoch": 1.9096573208722742,
|
| 1954 |
-
"grad_norm": 0.
|
| 1955 |
"learning_rate": 1.2330583229827053e-06,
|
| 1956 |
-
"loss": 0.
|
| 1957 |
"step": 1380
|
| 1958 |
},
|
| 1959 |
{
|
| 1960 |
"epoch": 1.9165801315334026,
|
| 1961 |
-
"grad_norm": 0.
|
| 1962 |
"learning_rate": 1.0509734826112394e-06,
|
| 1963 |
-
"loss": 0.
|
| 1964 |
"step": 1385
|
| 1965 |
},
|
| 1966 |
{
|
| 1967 |
"epoch": 1.9235029421945309,
|
| 1968 |
-
"grad_norm": 0.
|
| 1969 |
"learning_rate": 8.833572814004321e-07,
|
| 1970 |
-
"loss": 0.
|
| 1971 |
"step": 1390
|
| 1972 |
},
|
| 1973 |
{
|
| 1974 |
"epoch": 1.9304257528556594,
|
| 1975 |
-
"grad_norm": 0.
|
| 1976 |
"learning_rate": 7.302342287205943e-07,
|
| 1977 |
-
"loss": 0.
|
| 1978 |
"step": 1395
|
| 1979 |
},
|
| 1980 |
{
|
| 1981 |
"epoch": 1.937348563516788,
|
| 1982 |
-
"grad_norm": 0.
|
| 1983 |
"learning_rate": 5.916267147080934e-07,
|
| 1984 |
-
"loss": 0.
|
| 1985 |
"step": 1400
|
| 1986 |
},
|
| 1987 |
{
|
| 1988 |
"epoch": 1.9442713741779163,
|
| 1989 |
-
"grad_norm": 0.
|
| 1990 |
"learning_rate": 4.6755500699138253e-07,
|
| 1991 |
-
"loss": 0.
|
| 1992 |
"step": 1405
|
| 1993 |
},
|
| 1994 |
{
|
| 1995 |
"epoch": 1.9511941848390446,
|
| 1996 |
-
"grad_norm": 0.
|
| 1997 |
"learning_rate": 3.5803724772740523e-07,
|
| 1998 |
-
"loss": 0.
|
| 1999 |
"step": 1410
|
| 2000 |
},
|
| 2001 |
{
|
| 2002 |
"epoch": 1.958116995500173,
|
| 2003 |
-
"grad_norm": 0.
|
| 2004 |
"learning_rate": 2.630894509488058e-07,
|
| 2005 |
-
"loss": 0.
|
| 2006 |
"step": 1415
|
| 2007 |
},
|
| 2008 |
{
|
| 2009 |
"epoch": 1.9650398061613015,
|
| 2010 |
-
"grad_norm": 0.
|
| 2011 |
"learning_rate": 1.8272550022230272e-07,
|
| 2012 |
-
"loss": 0.
|
| 2013 |
"step": 1420
|
| 2014 |
},
|
| 2015 |
{
|
| 2016 |
"epoch": 1.97196261682243,
|
| 2017 |
-
"grad_norm": 0.
|
| 2018 |
"learning_rate": 1.1695714661856816e-07,
|
| 2019 |
-
"loss": 0.
|
| 2020 |
"step": 1425
|
| 2021 |
},
|
| 2022 |
{
|
| 2023 |
"epoch": 1.9788854274835583,
|
| 2024 |
-
"grad_norm": 0.
|
| 2025 |
"learning_rate": 6.579400699397998e-08,
|
| 2026 |
-
"loss": 0.
|
| 2027 |
"step": 1430
|
| 2028 |
},
|
| 2029 |
{
|
| 2030 |
"epoch": 1.9858082381446867,
|
| 2031 |
-
"grad_norm": 0.
|
| 2032 |
"learning_rate": 2.9243562584402305e-08,
|
| 2033 |
-
"loss": 0.
|
| 2034 |
"step": 1435
|
| 2035 |
},
|
| 2036 |
{
|
| 2037 |
"epoch": 1.9927310488058152,
|
| 2038 |
-
"grad_norm": 0.
|
| 2039 |
"learning_rate": 7.311157911249478e-09,
|
| 2040 |
-
"loss": 0.
|
| 2041 |
"step": 1440
|
| 2042 |
},
|
| 2043 |
{
|
| 2044 |
"epoch": 1.9982692973347178,
|
| 2045 |
-
"eval_loss": 0.
|
| 2046 |
-
"eval_runtime": 75.
|
| 2047 |
-
"eval_samples_per_second": 27.
|
| 2048 |
-
"eval_steps_per_second": 1.
|
| 2049 |
"step": 1444
|
| 2050 |
},
|
| 2051 |
{
|
| 2052 |
"epoch": 1.9982692973347178,
|
| 2053 |
"step": 1444,
|
| 2054 |
"total_flos": 6.774706456143856e+18,
|
| 2055 |
-
"train_loss": 0.
|
| 2056 |
-
"train_runtime":
|
| 2057 |
-
"train_samples_per_second": 6.
|
| 2058 |
-
"train_steps_per_second": 0.
|
| 2059 |
}
|
| 2060 |
],
|
| 2061 |
"logging_steps": 5,
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.0013845621322256837,
|
| 14 |
+
"grad_norm": 73.5276107788086,
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
"loss": 3.4696,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.006922810661128418,
|
| 21 |
+
"grad_norm": 30.179004669189453,
|
| 22 |
"learning_rate": 5.517241379310345e-06,
|
| 23 |
+
"loss": 3.2174,
|
| 24 |
"step": 5
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.013845621322256836,
|
| 28 |
+
"grad_norm": 18.46404266357422,
|
| 29 |
"learning_rate": 1.2413793103448277e-05,
|
| 30 |
+
"loss": 2.7354,
|
| 31 |
"step": 10
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.020768431983385256,
|
| 35 |
+
"grad_norm": 7.723433017730713,
|
| 36 |
"learning_rate": 1.9310344827586207e-05,
|
| 37 |
+
"loss": 2.4492,
|
| 38 |
"step": 15
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.027691242644513673,
|
| 42 |
+
"grad_norm": 7.463802337646484,
|
| 43 |
"learning_rate": 2.620689655172414e-05,
|
| 44 |
+
"loss": 2.2143,
|
| 45 |
"step": 20
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.034614053305642094,
|
| 49 |
+
"grad_norm": 7.357604503631592,
|
| 50 |
"learning_rate": 3.310344827586207e-05,
|
| 51 |
+
"loss": 1.921,
|
| 52 |
"step": 25
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.04153686396677051,
|
| 56 |
+
"grad_norm": 7.084353446960449,
|
| 57 |
"learning_rate": 4e-05,
|
| 58 |
+
"loss": 1.5084,
|
| 59 |
"step": 30
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.04845967462789893,
|
| 63 |
+
"grad_norm": 3.4993274211883545,
|
| 64 |
"learning_rate": 4.689655172413793e-05,
|
| 65 |
+
"loss": 1.0997,
|
| 66 |
"step": 35
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.055382485289027346,
|
| 70 |
+
"grad_norm": 2.801764965057373,
|
| 71 |
"learning_rate": 5.379310344827586e-05,
|
| 72 |
+
"loss": 1.0037,
|
| 73 |
"step": 40
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.06230529595015576,
|
| 77 |
+
"grad_norm": 7.829431533813477,
|
| 78 |
"learning_rate": 6.068965517241379e-05,
|
| 79 |
+
"loss": 1.0,
|
| 80 |
"step": 45
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.06922810661128419,
|
| 84 |
+
"grad_norm": 3.6705822944641113,
|
| 85 |
"learning_rate": 6.758620689655173e-05,
|
| 86 |
+
"loss": 0.9834,
|
| 87 |
"step": 50
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.0761509172724126,
|
| 91 |
+
"grad_norm": 2.8807718753814697,
|
| 92 |
"learning_rate": 7.448275862068966e-05,
|
| 93 |
+
"loss": 0.8439,
|
| 94 |
"step": 55
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.08307372793354102,
|
| 98 |
+
"grad_norm": 2.0084006786346436,
|
| 99 |
"learning_rate": 8.137931034482759e-05,
|
| 100 |
+
"loss": 0.754,
|
| 101 |
"step": 60
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.08999653859466944,
|
| 105 |
+
"grad_norm": 9.466357231140137,
|
| 106 |
"learning_rate": 8.827586206896552e-05,
|
| 107 |
+
"loss": 0.9603,
|
| 108 |
"step": 65
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.09691934925579786,
|
| 112 |
+
"grad_norm": 2.5492539405822754,
|
| 113 |
"learning_rate": 9.517241379310345e-05,
|
| 114 |
+
"loss": 0.8366,
|
| 115 |
"step": 70
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.10384215991692627,
|
| 119 |
+
"grad_norm": 22.056821823120117,
|
| 120 |
"learning_rate": 0.0001020689655172414,
|
| 121 |
+
"loss": 0.7592,
|
| 122 |
"step": 75
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.11076497057805469,
|
| 126 |
+
"grad_norm": 29.231815338134766,
|
| 127 |
"learning_rate": 0.00010896551724137931,
|
| 128 |
+
"loss": 0.8805,
|
| 129 |
"step": 80
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.11768778123918311,
|
| 133 |
+
"grad_norm": 34.974769592285156,
|
| 134 |
"learning_rate": 0.00011586206896551725,
|
| 135 |
+
"loss": 1.0534,
|
| 136 |
"step": 85
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.12461059190031153,
|
| 140 |
+
"grad_norm": 5.458140850067139,
|
| 141 |
"learning_rate": 0.00012275862068965518,
|
| 142 |
+
"loss": 1.0965,
|
| 143 |
"step": 90
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.13153340256143994,
|
| 147 |
+
"grad_norm": 4.925862789154053,
|
| 148 |
"learning_rate": 0.0001296551724137931,
|
| 149 |
+
"loss": 0.9976,
|
| 150 |
"step": 95
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.13845621322256838,
|
| 154 |
+
"grad_norm": 43.03465270996094,
|
| 155 |
"learning_rate": 0.00013655172413793104,
|
| 156 |
+
"loss": 0.821,
|
| 157 |
"step": 100
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.14537902388369678,
|
| 161 |
+
"grad_norm": 17.449140548706055,
|
| 162 |
"learning_rate": 0.00014344827586206896,
|
| 163 |
+
"loss": 0.9602,
|
| 164 |
"step": 105
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.1523018345448252,
|
| 168 |
+
"grad_norm": 1.8036819696426392,
|
| 169 |
"learning_rate": 0.0001503448275862069,
|
| 170 |
+
"loss": 0.859,
|
| 171 |
"step": 110
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.1592246452059536,
|
| 175 |
+
"grad_norm": 1.6082074642181396,
|
| 176 |
"learning_rate": 0.00015724137931034485,
|
| 177 |
+
"loss": 0.7846,
|
| 178 |
"step": 115
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.16614745586708204,
|
| 182 |
+
"grad_norm": 2.0686991214752197,
|
| 183 |
"learning_rate": 0.00016413793103448276,
|
| 184 |
+
"loss": 0.8044,
|
| 185 |
"step": 120
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.17307026652821045,
|
| 189 |
+
"grad_norm": 25.762229919433594,
|
| 190 |
"learning_rate": 0.0001710344827586207,
|
| 191 |
+
"loss": 0.8199,
|
| 192 |
"step": 125
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.17999307718933888,
|
| 196 |
+
"grad_norm": 3.0691945552825928,
|
| 197 |
"learning_rate": 0.00017793103448275862,
|
| 198 |
+
"loss": 0.8383,
|
| 199 |
"step": 130
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.18691588785046728,
|
| 203 |
+
"grad_norm": 1.7510746717453003,
|
| 204 |
"learning_rate": 0.00018482758620689654,
|
| 205 |
+
"loss": 0.8982,
|
| 206 |
"step": 135
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.19383869851159571,
|
| 210 |
+
"grad_norm": 1.7901217937469482,
|
| 211 |
"learning_rate": 0.0001917241379310345,
|
| 212 |
+
"loss": 0.8741,
|
| 213 |
"step": 140
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.20076150917272412,
|
| 217 |
+
"grad_norm": 2.0487852096557617,
|
| 218 |
"learning_rate": 0.00019862068965517243,
|
| 219 |
+
"loss": 0.8404,
|
| 220 |
"step": 145
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.20768431983385255,
|
| 224 |
+
"grad_norm": 1.5697190761566162,
|
| 225 |
"learning_rate": 0.0001999953208384105,
|
| 226 |
+
"loss": 0.8212,
|
| 227 |
"step": 150
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.21460713049498095,
|
| 231 |
+
"grad_norm": 11.699938774108887,
|
| 232 |
"learning_rate": 0.0001999763124949378,
|
| 233 |
+
"loss": 0.8841,
|
| 234 |
"step": 155
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.22152994115610938,
|
| 238 |
+
"grad_norm": 1.6109673976898193,
|
| 239 |
"learning_rate": 0.0001999426852993044,
|
| 240 |
+
"loss": 0.9349,
|
| 241 |
"step": 160
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.2284527518172378,
|
| 245 |
+
"grad_norm": 2.988689661026001,
|
| 246 |
"learning_rate": 0.00019989444416858502,
|
| 247 |
+
"loss": 0.9006,
|
| 248 |
"step": 165
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.23537556247836622,
|
| 252 |
+
"grad_norm": 6.933545112609863,
|
| 253 |
"learning_rate": 0.0001998315961567502,
|
| 254 |
+
"loss": 1.0421,
|
| 255 |
"step": 170
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.24229837313949462,
|
| 259 |
+
"grad_norm": 5.933894634246826,
|
| 260 |
"learning_rate": 0.00019975415045363467,
|
| 261 |
+
"loss": 0.9528,
|
| 262 |
"step": 175
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.24922118380062305,
|
| 266 |
+
"grad_norm": 82.2666244506836,
|
| 267 |
"learning_rate": 0.00019966211838359378,
|
| 268 |
+
"loss": 0.947,
|
| 269 |
"step": 180
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.25614399446175146,
|
| 273 |
+
"grad_norm": 2.0857582092285156,
|
| 274 |
"learning_rate": 0.00019955551340384743,
|
| 275 |
+
"loss": 0.9174,
|
| 276 |
"step": 185
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.2630668051228799,
|
| 280 |
+
"grad_norm": 4.074726104736328,
|
| 281 |
"learning_rate": 0.00019943435110251248,
|
| 282 |
+
"loss": 0.8472,
|
| 283 |
"step": 190
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.2699896157840083,
|
| 287 |
+
"grad_norm": 2.122100591659546,
|
| 288 |
"learning_rate": 0.0001992986491963232,
|
| 289 |
+
"loss": 0.8403,
|
| 290 |
"step": 195
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.27691242644513675,
|
| 294 |
+
"grad_norm": 1.1407779455184937,
|
| 295 |
"learning_rate": 0.00019914842752804103,
|
| 296 |
+
"loss": 0.8065,
|
| 297 |
"step": 200
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.2838352371062651,
|
| 301 |
+
"grad_norm": 1.0739364624023438,
|
| 302 |
"learning_rate": 0.00019898370806355263,
|
| 303 |
+
"loss": 0.7784,
|
| 304 |
"step": 205
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.29075804776739356,
|
| 308 |
+
"grad_norm": 0.9849103093147278,
|
| 309 |
"learning_rate": 0.00019880451488865836,
|
| 310 |
+
"loss": 0.8174,
|
| 311 |
"step": 210
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.297680858428522,
|
| 315 |
+
"grad_norm": 0.840072751045227,
|
| 316 |
"learning_rate": 0.00019861087420555018,
|
| 317 |
+
"loss": 0.7934,
|
| 318 |
"step": 215
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.3046036690896504,
|
| 322 |
+
"grad_norm": 0.996491551399231,
|
| 323 |
"learning_rate": 0.0001984028143289803,
|
| 324 |
+
"loss": 0.7948,
|
| 325 |
"step": 220
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.3115264797507788,
|
| 329 |
+
"grad_norm": 1.0046900510787964,
|
| 330 |
"learning_rate": 0.00019818036568212106,
|
| 331 |
+
"loss": 0.7789,
|
| 332 |
"step": 225
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.3184492904119072,
|
| 336 |
+
"grad_norm": 1.3844839334487915,
|
| 337 |
"learning_rate": 0.00019794356079211604,
|
| 338 |
+
"loss": 0.8206,
|
| 339 |
"step": 230
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.32537210107303566,
|
| 343 |
+
"grad_norm": 0.9604908227920532,
|
| 344 |
"learning_rate": 0.00019769243428532422,
|
| 345 |
+
"loss": 0.7827,
|
| 346 |
"step": 235
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.3322949117341641,
|
| 350 |
+
"grad_norm": 14.684309959411621,
|
| 351 |
"learning_rate": 0.00019742702288225652,
|
| 352 |
+
"loss": 0.8219,
|
| 353 |
"step": 240
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.33921772239529246,
|
| 357 |
+
"grad_norm": 1.059198260307312,
|
| 358 |
"learning_rate": 0.00019714736539220648,
|
| 359 |
+
"loss": 0.825,
|
| 360 |
"step": 245
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.3461405330564209,
|
| 364 |
+
"grad_norm": 0.8264622688293457,
|
| 365 |
"learning_rate": 0.00019685350270757555,
|
| 366 |
+
"loss": 0.7993,
|
| 367 |
"step": 250
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.3530633437175493,
|
| 371 |
+
"grad_norm": 0.8702566623687744,
|
| 372 |
"learning_rate": 0.0001965454777978936,
|
| 373 |
+
"loss": 0.8209,
|
| 374 |
"step": 255
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.35998615437867776,
|
| 378 |
+
"grad_norm": 1.260245680809021,
|
| 379 |
"learning_rate": 0.00019622333570353567,
|
| 380 |
+
"loss": 0.8117,
|
| 381 |
"step": 260
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.36690896503980613,
|
| 385 |
+
"grad_norm": 0.8236448764801025,
|
| 386 |
"learning_rate": 0.00019588712352913625,
|
| 387 |
+
"loss": 0.7933,
|
| 388 |
"step": 265
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.37383177570093457,
|
| 392 |
+
"grad_norm": 0.7921434640884399,
|
| 393 |
"learning_rate": 0.00019553689043670127,
|
| 394 |
+
"loss": 0.7715,
|
| 395 |
"step": 270
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.380754586362063,
|
| 399 |
+
"grad_norm": 0.8058004975318909,
|
| 400 |
"learning_rate": 0.00019517268763841962,
|
| 401 |
+
"loss": 0.7636,
|
| 402 |
"step": 275
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.38767739702319143,
|
| 406 |
+
"grad_norm": 0.7768955826759338,
|
| 407 |
"learning_rate": 0.00019479456838917473,
|
| 408 |
+
"loss": 0.774,
|
| 409 |
"step": 280
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.39460020768431986,
|
| 413 |
+
"grad_norm": 0.7238219380378723,
|
| 414 |
"learning_rate": 0.0001944025879787574,
|
| 415 |
+
"loss": 0.7903,
|
| 416 |
"step": 285
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.40152301834544823,
|
| 420 |
+
"grad_norm": 0.9116212725639343,
|
| 421 |
"learning_rate": 0.0001939968037237812,
|
| 422 |
+
"loss": 0.7678,
|
| 423 |
"step": 290
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.40844582900657667,
|
| 427 |
+
"grad_norm": 0.9523202180862427,
|
| 428 |
"learning_rate": 0.00019357727495930146,
|
| 429 |
+
"loss": 0.7683,
|
| 430 |
"step": 295
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.4153686396677051,
|
| 434 |
+
"grad_norm": 0.902203381061554,
|
| 435 |
"learning_rate": 0.00019314406303013904,
|
| 436 |
+
"loss": 0.7641,
|
| 437 |
"step": 300
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.42229145032883353,
|
| 441 |
+
"grad_norm": 0.6467769742012024,
|
| 442 |
"learning_rate": 0.00019269723128191048,
|
| 443 |
+
"loss": 0.7636,
|
| 444 |
"step": 305
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.4292142609899619,
|
| 448 |
+
"grad_norm": 0.7330191135406494,
|
| 449 |
"learning_rate": 0.00019223684505176517,
|
| 450 |
+
"loss": 0.7551,
|
| 451 |
"step": 310
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.43613707165109034,
|
| 455 |
+
"grad_norm": 0.8263298273086548,
|
| 456 |
"learning_rate": 0.00019176297165883165,
|
| 457 |
+
"loss": 0.7655,
|
| 458 |
"step": 315
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.44305988231221877,
|
| 462 |
+
"grad_norm": 0.6522659063339233,
|
| 463 |
"learning_rate": 0.00019127568039437406,
|
| 464 |
+
"loss": 0.7607,
|
| 465 |
"step": 320
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.4499826929733472,
|
| 469 |
+
"grad_norm": 0.6384459733963013,
|
| 470 |
"learning_rate": 0.00019077504251166006,
|
| 471 |
+
"loss": 0.7559,
|
| 472 |
"step": 325
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.4569055036344756,
|
| 476 |
+
"grad_norm": 0.7025325298309326,
|
| 477 |
"learning_rate": 0.00019026113121554204,
|
| 478 |
+
"loss": 0.7348,
|
| 479 |
"step": 330
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.463828314295604,
|
| 483 |
+
"grad_norm": 0.6380446553230286,
|
| 484 |
"learning_rate": 0.00018973402165175268,
|
| 485 |
+
"loss": 0.7585,
|
| 486 |
"step": 335
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.47075112495673244,
|
| 490 |
+
"grad_norm": 0.6707427501678467,
|
| 491 |
"learning_rate": 0.0001891937908959172,
|
| 492 |
+
"loss": 0.7546,
|
| 493 |
"step": 340
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.47767393561786087,
|
| 497 |
+
"grad_norm": 0.7454755306243896,
|
| 498 |
"learning_rate": 0.00018864051794228282,
|
| 499 |
+
"loss": 0.7639,
|
| 500 |
"step": 345
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.48459674627898924,
|
| 504 |
+
"grad_norm": 0.8577032685279846,
|
| 505 |
"learning_rate": 0.00018807428369216822,
|
| 506 |
+
"loss": 0.7323,
|
| 507 |
"step": 350
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.4915195569401177,
|
| 511 |
+
"grad_norm": 0.655467689037323,
|
| 512 |
"learning_rate": 0.00018749517094213375,
|
| 513 |
+
"loss": 0.7396,
|
| 514 |
"step": 355
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.4984423676012461,
|
| 518 |
+
"grad_norm": 0.6951208114624023,
|
| 519 |
"learning_rate": 0.00018690326437187472,
|
| 520 |
+
"loss": 0.7385,
|
| 521 |
"step": 360
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.5053651782623745,
|
| 525 |
+
"grad_norm": 1.1481677293777466,
|
| 526 |
"learning_rate": 0.00018629865053183911,
|
| 527 |
+
"loss": 0.7453,
|
| 528 |
"step": 365
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.5122879889235029,
|
| 532 |
+
"grad_norm": 0.7248277068138123,
|
| 533 |
"learning_rate": 0.0001856814178305722,
|
| 534 |
+
"loss": 0.7435,
|
| 535 |
"step": 370
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.5192107995846313,
|
| 539 |
+
"grad_norm": 0.787811279296875,
|
| 540 |
"learning_rate": 0.00018505165652178893,
|
| 541 |
+
"loss": 0.7515,
|
| 542 |
"step": 375
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.5261336102457598,
|
| 546 |
+
"grad_norm": 0.6863015294075012,
|
| 547 |
"learning_rate": 0.00018440945869117675,
|
| 548 |
+
"loss": 0.7192,
|
| 549 |
"step": 380
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.5330564209068882,
|
| 553 |
+
"grad_norm": 0.6457992196083069,
|
| 554 |
"learning_rate": 0.00018375491824293067,
|
| 555 |
+
"loss": 0.7192,
|
| 556 |
"step": 385
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.5399792315680166,
|
| 560 |
+
"grad_norm": 0.7503763437271118,
|
| 561 |
"learning_rate": 0.00018308813088602227,
|
| 562 |
+
"loss": 0.7396,
|
| 563 |
"step": 390
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.5469020422291451,
|
| 567 |
+
"grad_norm": 0.5836992263793945,
|
| 568 |
"learning_rate": 0.00018240919412020466,
|
| 569 |
+
"loss": 0.7118,
|
| 570 |
"step": 395
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.5538248528902735,
|
| 574 |
+
"grad_norm": 5.260194778442383,
|
| 575 |
"learning_rate": 0.00018171820722175604,
|
| 576 |
+
"loss": 0.7462,
|
| 577 |
"step": 400
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.5607476635514018,
|
| 581 |
+
"grad_norm": 0.6316019296646118,
|
| 582 |
"learning_rate": 0.00018101527122896293,
|
| 583 |
+
"loss": 0.7215,
|
| 584 |
"step": 405
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.5676704742125303,
|
| 588 |
+
"grad_norm": 0.6845203638076782,
|
| 589 |
"learning_rate": 0.0001803004889273463,
|
| 590 |
+
"loss": 0.7099,
|
| 591 |
"step": 410
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.5745932848736587,
|
| 595 |
+
"grad_norm": 0.6993951797485352,
|
| 596 |
"learning_rate": 0.0001795739648346317,
|
| 597 |
+
"loss": 0.7139,
|
| 598 |
"step": 415
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.5815160955347871,
|
| 602 |
+
"grad_norm": 0.5756929516792297,
|
| 603 |
"learning_rate": 0.00017883580518546647,
|
| 604 |
+
"loss": 0.6847,
|
| 605 |
"step": 420
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.5884389061959155,
|
| 609 |
+
"grad_norm": 0.6658014059066772,
|
| 610 |
"learning_rate": 0.00017808611791588584,
|
| 611 |
+
"loss": 0.7278,
|
| 612 |
"step": 425
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.595361716857044,
|
| 616 |
+
"grad_norm": 0.6488052010536194,
|
| 617 |
"learning_rate": 0.00017732501264752996,
|
| 618 |
+
"loss": 0.7142,
|
| 619 |
"step": 430
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.6022845275181724,
|
| 623 |
+
"grad_norm": 0.5289434790611267,
|
| 624 |
"learning_rate": 0.00017655260067161503,
|
| 625 |
+
"loss": 0.7167,
|
| 626 |
"step": 435
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.6092073381793008,
|
| 630 |
+
"grad_norm": 0.6898224949836731,
|
| 631 |
"learning_rate": 0.00017576899493265954,
|
| 632 |
+
"loss": 0.7165,
|
| 633 |
"step": 440
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.6161301488404292,
|
| 637 |
+
"grad_norm": 0.6195114850997925,
|
| 638 |
"learning_rate": 0.00017497431001196943,
|
| 639 |
+
"loss": 0.699,
|
| 640 |
"step": 445
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.6230529595015576,
|
| 644 |
+
"grad_norm": 0.6543487906455994,
|
| 645 |
"learning_rate": 0.00017416866211088366,
|
| 646 |
+
"loss": 0.7115,
|
| 647 |
"step": 450
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.629975770162686,
|
| 651 |
+
"grad_norm": 0.5665395259857178,
|
| 652 |
"learning_rate": 0.00017335216903378267,
|
| 653 |
+
"loss": 0.7074,
|
| 654 |
"step": 455
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.6368985808238145,
|
| 658 |
+
"grad_norm": 0.6460801959037781,
|
| 659 |
"learning_rate": 0.00017252495017086296,
|
| 660 |
+
"loss": 0.7199,
|
| 661 |
"step": 460
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.6438213914849429,
|
| 665 |
+
"grad_norm": 0.7826030254364014,
|
| 666 |
"learning_rate": 0.0001716871264806791,
|
| 667 |
+
"loss": 0.7027,
|
| 668 |
"step": 465
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.6507442021460713,
|
| 672 |
+
"grad_norm": 5.1860504150390625,
|
| 673 |
"learning_rate": 0.0001708388204724572,
|
| 674 |
+
"loss": 0.7063,
|
| 675 |
"step": 470
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.6576670128071997,
|
| 679 |
+
"grad_norm": 0.5539765954017639,
|
| 680 |
"learning_rate": 0.00016998015618818087,
|
| 681 |
+
"loss": 0.7021,
|
| 682 |
"step": 475
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.6645898234683282,
|
| 686 |
+
"grad_norm": 0.6178136467933655,
|
| 687 |
"learning_rate": 0.00016911125918445363,
|
| 688 |
+
"loss": 0.7047,
|
| 689 |
"step": 480
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.6715126341294566,
|
| 693 |
+
"grad_norm": 0.5850706696510315,
|
| 694 |
"learning_rate": 0.00016823225651413953,
|
| 695 |
+
"loss": 0.7163,
|
| 696 |
"step": 485
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.6784354447905849,
|
| 700 |
+
"grad_norm": 0.5677686929702759,
|
| 701 |
"learning_rate": 0.00016734327670778508,
|
| 702 |
+
"loss": 0.6866,
|
| 703 |
"step": 490
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.6853582554517134,
|
| 707 |
+
"grad_norm": 0.5619258880615234,
|
| 708 |
"learning_rate": 0.00016644444975482523,
|
| 709 |
+
"loss": 0.7048,
|
| 710 |
"step": 495
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.6922810661128418,
|
| 714 |
+
"grad_norm": 0.6009318232536316,
|
| 715 |
"learning_rate": 0.0001655359070845757,
|
| 716 |
+
"loss": 0.7023,
|
| 717 |
"step": 500
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.6992038767739702,
|
| 721 |
+
"grad_norm": 0.5337724089622498,
|
| 722 |
"learning_rate": 0.0001646177815470151,
|
| 723 |
+
"loss": 0.6858,
|
| 724 |
"step": 505
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.7061266874350987,
|
| 728 |
+
"grad_norm": 0.550845205783844,
|
| 729 |
"learning_rate": 0.00016369020739335924,
|
| 730 |
+
"loss": 0.6807,
|
| 731 |
"step": 510
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.7130494980962271,
|
| 735 |
+
"grad_norm": 0.5907328724861145,
|
| 736 |
"learning_rate": 0.00016275332025643028,
|
| 737 |
+
"loss": 0.7041,
|
| 738 |
"step": 515
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.7199723087573555,
|
| 742 |
+
"grad_norm": 0.5490245223045349,
|
| 743 |
"learning_rate": 0.00016180725713082434,
|
| 744 |
+
"loss": 0.6875,
|
| 745 |
"step": 520
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.726895119418484,
|
| 749 |
+
"grad_norm": 0.5875672101974487,
|
| 750 |
"learning_rate": 0.0001608521563528795,
|
| 751 |
+
"loss": 0.6942,
|
| 752 |
"step": 525
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.7338179300796123,
|
| 756 |
+
"grad_norm": 0.4903637170791626,
|
| 757 |
"learning_rate": 0.00015988815758044792,
|
| 758 |
+
"loss": 0.667,
|
| 759 |
"step": 530
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.7407407407407407,
|
| 763 |
+
"grad_norm": 0.5825007557868958,
|
| 764 |
"learning_rate": 0.00015891540177247467,
|
| 765 |
+
"loss": 0.697,
|
| 766 |
"step": 535
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.7476635514018691,
|
| 770 |
+
"grad_norm": 0.6213587522506714,
|
| 771 |
"learning_rate": 0.00015793403116838613,
|
| 772 |
+
"loss": 0.6751,
|
| 773 |
"step": 540
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.7545863620629976,
|
| 777 |
+
"grad_norm": 0.6950411796569824,
|
| 778 |
"learning_rate": 0.00015694418926729146,
|
| 779 |
+
"loss": 0.6913,
|
| 780 |
"step": 545
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.761509172724126,
|
| 784 |
+
"grad_norm": 0.5191546678543091,
|
| 785 |
"learning_rate": 0.00015594602080699956,
|
| 786 |
+
"loss": 0.6758,
|
| 787 |
"step": 550
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.7684319833852544,
|
| 791 |
+
"grad_norm": 0.5811347961425781,
|
| 792 |
"learning_rate": 0.00015493967174285512,
|
| 793 |
+
"loss": 0.6892,
|
| 794 |
"step": 555
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.7753547940463829,
|
| 798 |
+
"grad_norm": 0.495950847864151,
|
| 799 |
"learning_rate": 0.00015392528922639662,
|
| 800 |
+
"loss": 0.6768,
|
| 801 |
"step": 560
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.7822776047075113,
|
| 805 |
+
"grad_norm": 0.5069176554679871,
|
| 806 |
"learning_rate": 0.00015290302158383923,
|
| 807 |
+
"loss": 0.6716,
|
| 808 |
"step": 565
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.7892004153686397,
|
| 812 |
+
"grad_norm": 0.5323678255081177,
|
| 813 |
"learning_rate": 0.0001518730182943862,
|
| 814 |
+
"loss": 0.6698,
|
| 815 |
"step": 570
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.796123226029768,
|
| 819 |
+
"grad_norm": 0.7031868100166321,
|
| 820 |
"learning_rate": 0.0001508354299683715,
|
| 821 |
+
"loss": 0.6673,
|
| 822 |
"step": 575
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 0.8030460366908965,
|
| 826 |
+
"grad_norm": 0.55892014503479,
|
| 827 |
"learning_rate": 0.0001497904083252371,
|
| 828 |
+
"loss": 0.6675,
|
| 829 |
"step": 580
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 0.8099688473520249,
|
| 833 |
+
"grad_norm": 0.4777008593082428,
|
| 834 |
"learning_rate": 0.00014873810617134806,
|
| 835 |
+
"loss": 0.6807,
|
| 836 |
"step": 585
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.8168916580131533,
|
| 840 |
+
"grad_norm": 0.543347954750061,
|
| 841 |
"learning_rate": 0.00014767867737764876,
|
| 842 |
+
"loss": 0.6814,
|
| 843 |
"step": 590
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 0.8238144686742818,
|
| 847 |
+
"grad_norm": 0.4696509838104248,
|
| 848 |
"learning_rate": 0.0001466122768571633,
|
| 849 |
+
"loss": 0.6506,
|
| 850 |
"step": 595
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 0.8307372793354102,
|
| 854 |
+
"grad_norm": 0.5306531190872192,
|
| 855 |
"learning_rate": 0.0001455390605423437,
|
| 856 |
+
"loss": 0.6731,
|
| 857 |
"step": 600
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.8376600899965386,
|
| 861 |
+
"grad_norm": 0.6365358233451843,
|
| 862 |
"learning_rate": 0.00014445918536226904,
|
| 863 |
+
"loss": 0.6441,
|
| 864 |
"step": 605
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.8445829006576671,
|
| 868 |
+
"grad_norm": 0.5179509520530701,
|
| 869 |
"learning_rate": 0.00014337280921969858,
|
| 870 |
+
"loss": 0.6852,
|
| 871 |
"step": 610
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 0.8515057113187954,
|
| 875 |
+
"grad_norm": 0.5093657970428467,
|
| 876 |
"learning_rate": 0.0001422800909679829,
|
| 877 |
+
"loss": 0.6648,
|
| 878 |
"step": 615
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.8584285219799238,
|
| 882 |
+
"grad_norm": 0.4981916546821594,
|
| 883 |
"learning_rate": 0.00014118119038783588,
|
| 884 |
+
"loss": 0.6563,
|
| 885 |
"step": 620
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 0.8653513326410522,
|
| 889 |
+
"grad_norm": 0.5044928789138794,
|
| 890 |
"learning_rate": 0.00014007626816397087,
|
| 891 |
+
"loss": 0.6661,
|
| 892 |
"step": 625
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 0.8722741433021807,
|
| 896 |
+
"grad_norm": 0.5227713584899902,
|
| 897 |
"learning_rate": 0.00013896548586160503,
|
| 898 |
+
"loss": 0.6885,
|
| 899 |
"step": 630
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 0.8791969539633091,
|
| 903 |
+
"grad_norm": 0.5463073253631592,
|
| 904 |
"learning_rate": 0.00013784900590283473,
|
| 905 |
+
"loss": 0.6609,
|
| 906 |
"step": 635
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 0.8861197646244375,
|
| 910 |
+
"grad_norm": 0.49851828813552856,
|
| 911 |
"learning_rate": 0.00013672699154288569,
|
| 912 |
+
"loss": 0.6501,
|
| 913 |
"step": 640
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 0.893042575285566,
|
| 917 |
+
"grad_norm": 0.5655946731567383,
|
| 918 |
"learning_rate": 0.00013559960684624112,
|
| 919 |
+
"loss": 0.6635,
|
| 920 |
"step": 645
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 0.8999653859466944,
|
| 924 |
+
"grad_norm": 0.5413991808891296,
|
| 925 |
"learning_rate": 0.0001344670166626519,
|
| 926 |
+
"loss": 0.6705,
|
| 927 |
"step": 650
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.9068881966078228,
|
| 931 |
+
"grad_norm": 0.5494153499603271,
|
| 932 |
"learning_rate": 0.0001333293866030317,
|
| 933 |
+
"loss": 0.6723,
|
| 934 |
"step": 655
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 0.9138110072689511,
|
| 938 |
+
"grad_norm": 0.5189006328582764,
|
| 939 |
"learning_rate": 0.0001321868830152407,
|
| 940 |
+
"loss": 0.6562,
|
| 941 |
"step": 660
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.9207338179300796,
|
| 945 |
+
"grad_norm": 0.5886821746826172,
|
| 946 |
"learning_rate": 0.00013103967295976179,
|
| 947 |
+
"loss": 0.6506,
|
| 948 |
"step": 665
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 0.927656628591208,
|
| 952 |
+
"grad_norm": 0.44091060757637024,
|
| 953 |
"learning_rate": 0.0001298879241852724,
|
| 954 |
+
"loss": 0.6543,
|
| 955 |
"step": 670
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 0.9345794392523364,
|
| 959 |
+
"grad_norm": 0.5451465249061584,
|
| 960 |
"learning_rate": 0.0001287318051041159,
|
| 961 |
+
"loss": 0.6436,
|
| 962 |
"step": 675
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.9415022499134649,
|
| 966 |
+
"grad_norm": 0.48705631494522095,
|
| 967 |
"learning_rate": 0.00012757148476767553,
|
| 968 |
+
"loss": 0.6486,
|
| 969 |
"step": 680
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.9484250605745933,
|
| 973 |
+
"grad_norm": 0.5340054035186768,
|
| 974 |
"learning_rate": 0.0001264071328416555,
|
| 975 |
+
"loss": 0.6415,
|
| 976 |
"step": 685
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 0.9553478712357217,
|
| 980 |
+
"grad_norm": 0.5495091080665588,
|
| 981 |
"learning_rate": 0.00012523891958127168,
|
| 982 |
+
"loss": 0.6737,
|
| 983 |
"step": 690
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 0.9622706818968502,
|
| 987 |
+
"grad_norm": 0.4970749020576477,
|
| 988 |
"learning_rate": 0.0001240670158063565,
|
| 989 |
+
"loss": 0.6546,
|
| 990 |
"step": 695
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.9691934925579785,
|
| 994 |
+
"grad_norm": 0.47575998306274414,
|
| 995 |
"learning_rate": 0.00012289159287638105,
|
| 996 |
+
"loss": 0.6449,
|
| 997 |
"step": 700
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 0.9761163032191069,
|
| 1001 |
+
"grad_norm": 0.4925188422203064,
|
| 1002 |
"learning_rate": 0.00012171282266539845,
|
| 1003 |
+
"loss": 0.6609,
|
| 1004 |
"step": 705
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 0.9830391138802353,
|
| 1008 |
+
"grad_norm": 0.47044652700424194,
|
| 1009 |
"learning_rate": 0.00012053087753691172,
|
| 1010 |
+
"loss": 0.653,
|
| 1011 |
"step": 710
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 0.9899619245413638,
|
| 1015 |
+
"grad_norm": 0.47281110286712646,
|
| 1016 |
"learning_rate": 0.00011934593031867038,
|
| 1017 |
+
"loss": 0.6552,
|
| 1018 |
"step": 715
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 0.9968847352024922,
|
| 1022 |
+
"grad_norm": 0.4654919505119324,
|
| 1023 |
"learning_rate": 0.00011815815427739911,
|
| 1024 |
+
"loss": 0.639,
|
| 1025 |
"step": 720
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 1.0,
|
| 1029 |
+
"eval_loss": 0.63841313123703,
|
| 1030 |
+
"eval_runtime": 75.465,
|
| 1031 |
+
"eval_samples_per_second": 27.417,
|
| 1032 |
+
"eval_steps_per_second": 1.723,
|
| 1033 |
"step": 723
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
"epoch": 1.0027691242644514,
|
| 1037 |
+
"grad_norm": 0.476882666349411,
|
| 1038 |
"learning_rate": 0.00011696772309346182,
|
| 1039 |
+
"loss": 0.591,
|
| 1040 |
"step": 725
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
"epoch": 1.0096919349255797,
|
| 1044 |
+
"grad_norm": 0.5042451620101929,
|
| 1045 |
"learning_rate": 0.0001157748108354659,
|
| 1046 |
+
"loss": 0.5285,
|
| 1047 |
"step": 730
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
"epoch": 1.0166147455867083,
|
| 1051 |
+
"grad_norm": 0.4443126916885376,
|
| 1052 |
"learning_rate": 0.00011457959193480925,
|
| 1053 |
+
"loss": 0.5251,
|
| 1054 |
"step": 735
|
| 1055 |
},
|
| 1056 |
{
|
| 1057 |
"epoch": 1.0235375562478366,
|
| 1058 |
+
"grad_norm": 0.47023993730545044,
|
| 1059 |
"learning_rate": 0.00011338224116017423,
|
| 1060 |
+
"loss": 0.5307,
|
| 1061 |
"step": 740
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
"epoch": 1.0304603669089651,
|
| 1065 |
+
"grad_norm": 0.4469102621078491,
|
| 1066 |
"learning_rate": 0.00011218293359197267,
|
| 1067 |
+
"loss": 0.5406,
|
| 1068 |
"step": 745
|
| 1069 |
},
|
| 1070 |
{
|
| 1071 |
"epoch": 1.0373831775700935,
|
| 1072 |
+
"grad_norm": 0.5018590092658997,
|
| 1073 |
"learning_rate": 0.00011098184459674484,
|
| 1074 |
+
"loss": 0.5217,
|
| 1075 |
"step": 750
|
| 1076 |
},
|
| 1077 |
{
|
| 1078 |
"epoch": 1.0443059882312218,
|
| 1079 |
+
"grad_norm": 0.44223928451538086,
|
| 1080 |
"learning_rate": 0.00010977914980151691,
|
| 1081 |
+
"loss": 0.5215,
|
| 1082 |
"step": 755
|
| 1083 |
},
|
| 1084 |
{
|
| 1085 |
"epoch": 1.0512287988923503,
|
| 1086 |
+
"grad_norm": 0.4535723030567169,
|
| 1087 |
"learning_rate": 0.00010857502506812029,
|
| 1088 |
+
"loss": 0.4987,
|
| 1089 |
"step": 760
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 1.0581516095534786,
|
| 1093 |
+
"grad_norm": 0.4797036647796631,
|
| 1094 |
"learning_rate": 0.00010736964646747629,
|
| 1095 |
+
"loss": 0.5018,
|
| 1096 |
"step": 765
|
| 1097 |
},
|
| 1098 |
{
|
| 1099 |
"epoch": 1.0650744202146072,
|
| 1100 |
+
"grad_norm": 0.4599474370479584,
|
| 1101 |
"learning_rate": 0.00010616319025385089,
|
| 1102 |
+
"loss": 0.5222,
|
| 1103 |
"step": 770
|
| 1104 |
},
|
| 1105 |
{
|
| 1106 |
"epoch": 1.0719972308757355,
|
| 1107 |
+
"grad_norm": 0.43609824776649475,
|
| 1108 |
"learning_rate": 0.00010495583283908177,
|
| 1109 |
+
"loss": 0.5227,
|
| 1110 |
"step": 775
|
| 1111 |
},
|
| 1112 |
{
|
| 1113 |
"epoch": 1.078920041536864,
|
| 1114 |
+
"grad_norm": 0.4024893045425415,
|
| 1115 |
"learning_rate": 0.00010374775076678332,
|
| 1116 |
+
"loss": 0.5148,
|
| 1117 |
"step": 780
|
| 1118 |
},
|
| 1119 |
{
|
| 1120 |
"epoch": 1.0858428521979924,
|
| 1121 |
+
"grad_norm": 0.41304945945739746,
|
| 1122 |
"learning_rate": 0.00010253912068653146,
|
| 1123 |
+
"loss": 0.5222,
|
| 1124 |
"step": 785
|
| 1125 |
},
|
| 1126 |
{
|
| 1127 |
"epoch": 1.092765662859121,
|
| 1128 |
+
"grad_norm": 0.4150070250034332,
|
| 1129 |
"learning_rate": 0.00010133011932803378,
|
| 1130 |
+
"loss": 0.5287,
|
| 1131 |
"step": 790
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
"epoch": 1.0996884735202492,
|
| 1135 |
+
"grad_norm": 0.4497414827346802,
|
| 1136 |
"learning_rate": 0.00010012092347528705,
|
| 1137 |
+
"loss": 0.5287,
|
| 1138 |
"step": 795
|
| 1139 |
},
|
| 1140 |
{
|
| 1141 |
"epoch": 1.1066112841813776,
|
| 1142 |
+
"grad_norm": 0.4617963433265686,
|
| 1143 |
"learning_rate": 9.891170994072793e-05,
|
| 1144 |
+
"loss": 0.5142,
|
| 1145 |
"step": 800
|
| 1146 |
},
|
| 1147 |
{
|
| 1148 |
"epoch": 1.113534094842506,
|
| 1149 |
+
"grad_norm": 0.4907746911048889,
|
| 1150 |
"learning_rate": 9.770265553937831e-05,
|
| 1151 |
+
"loss": 0.5154,
|
| 1152 |
"step": 805
|
| 1153 |
},
|
| 1154 |
{
|
| 1155 |
"epoch": 1.1204569055036344,
|
| 1156 |
+
"grad_norm": 0.4437863230705261,
|
| 1157 |
"learning_rate": 9.649393706299127e-05,
|
| 1158 |
+
"loss": 0.5114,
|
| 1159 |
"step": 810
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
"epoch": 1.127379716164763,
|
| 1163 |
+
"grad_norm": 0.4392574429512024,
|
| 1164 |
"learning_rate": 9.52857312541998e-05,
|
| 1165 |
+
"loss": 0.5054,
|
| 1166 |
"step": 815
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
"epoch": 1.1343025268258913,
|
| 1170 |
+
"grad_norm": 0.41205236315727234,
|
| 1171 |
"learning_rate": 9.407821478067304e-05,
|
| 1172 |
+
"loss": 0.5043,
|
| 1173 |
"step": 820
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
"epoch": 1.1412253374870198,
|
| 1177 |
+
"grad_norm": 0.4570033848285675,
|
| 1178 |
"learning_rate": 9.287156420928342e-05,
|
| 1179 |
+
"loss": 0.5022,
|
| 1180 |
"step": 825
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
"epoch": 1.1481481481481481,
|
| 1184 |
+
"grad_norm": 0.43261808156967163,
|
| 1185 |
"learning_rate": 9.166595598028832e-05,
|
| 1186 |
+
"loss": 0.5218,
|
| 1187 |
"step": 830
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
"epoch": 1.1550709588092767,
|
| 1191 |
+
"grad_norm": 0.4350537657737732,
|
| 1192 |
"learning_rate": 9.046156638153056e-05,
|
| 1193 |
+
"loss": 0.5034,
|
| 1194 |
"step": 835
|
| 1195 |
},
|
| 1196 |
{
|
| 1197 |
"epoch": 1.161993769470405,
|
| 1198 |
+
"grad_norm": 0.41802477836608887,
|
| 1199 |
"learning_rate": 8.925857152266103e-05,
|
| 1200 |
+
"loss": 0.4959,
|
| 1201 |
"step": 840
|
| 1202 |
},
|
| 1203 |
{
|
| 1204 |
"epoch": 1.1689165801315333,
|
| 1205 |
+
"grad_norm": 0.4566584527492523,
|
| 1206 |
"learning_rate": 8.805714730938728e-05,
|
| 1207 |
+
"loss": 0.5095,
|
| 1208 |
"step": 845
|
| 1209 |
},
|
| 1210 |
{
|
| 1211 |
"epoch": 1.1758393907926619,
|
| 1212 |
+
"grad_norm": 0.3882467448711395,
|
| 1213 |
"learning_rate": 8.685746941775219e-05,
|
| 1214 |
+
"loss": 0.5031,
|
| 1215 |
"step": 850
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
"epoch": 1.1827622014537902,
|
| 1219 |
+
"grad_norm": 0.42582014203071594,
|
| 1220 |
"learning_rate": 8.565971326844584e-05,
|
| 1221 |
+
"loss": 0.5144,
|
| 1222 |
"step": 855
|
| 1223 |
},
|
| 1224 |
{
|
| 1225 |
"epoch": 1.1896850121149187,
|
| 1226 |
+
"grad_norm": 0.4243321418762207,
|
| 1227 |
"learning_rate": 8.44640540011553e-05,
|
| 1228 |
+
"loss": 0.5188,
|
| 1229 |
"step": 860
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
"epoch": 1.196607822776047,
|
| 1233 |
+
"grad_norm": 0.4075254797935486,
|
| 1234 |
"learning_rate": 8.327066644895469e-05,
|
| 1235 |
+
"loss": 0.5082,
|
| 1236 |
"step": 865
|
| 1237 |
},
|
| 1238 |
{
|
| 1239 |
"epoch": 1.2035306334371756,
|
| 1240 |
+
"grad_norm": 0.43225353956222534,
|
| 1241 |
"learning_rate": 8.207972511274095e-05,
|
| 1242 |
+
"loss": 0.5182,
|
| 1243 |
"step": 870
|
| 1244 |
},
|
| 1245 |
{
|
| 1246 |
"epoch": 1.210453444098304,
|
| 1247 |
+
"grad_norm": 0.4101962447166443,
|
| 1248 |
"learning_rate": 8.089140413571747e-05,
|
| 1249 |
+
"loss": 0.5116,
|
| 1250 |
"step": 875
|
| 1251 |
},
|
| 1252 |
{
|
| 1253 |
"epoch": 1.2173762547594325,
|
| 1254 |
+
"grad_norm": 0.4450133144855499,
|
| 1255 |
"learning_rate": 7.970587727793051e-05,
|
| 1256 |
+
"loss": 0.4971,
|
| 1257 |
"step": 880
|
| 1258 |
},
|
| 1259 |
{
|
| 1260 |
"epoch": 1.2242990654205608,
|
| 1261 |
+
"grad_norm": 0.43541422486305237,
|
| 1262 |
"learning_rate": 7.852331789086135e-05,
|
| 1263 |
+
"loss": 0.4993,
|
| 1264 |
"step": 885
|
| 1265 |
},
|
| 1266 |
{
|
| 1267 |
"epoch": 1.231221876081689,
|
| 1268 |
+
"grad_norm": 0.39591750502586365,
|
| 1269 |
"learning_rate": 7.73438988920784e-05,
|
| 1270 |
+
"loss": 0.5075,
|
| 1271 |
"step": 890
|
| 1272 |
},
|
| 1273 |
{
|
| 1274 |
"epoch": 1.2381446867428176,
|
| 1275 |
+
"grad_norm": 0.46820273995399475,
|
| 1276 |
"learning_rate": 7.616779273995252e-05,
|
| 1277 |
+
"loss": 0.5218,
|
| 1278 |
"step": 895
|
| 1279 |
},
|
| 1280 |
{
|
| 1281 |
"epoch": 1.245067497403946,
|
| 1282 |
+
"grad_norm": 0.425655722618103,
|
| 1283 |
"learning_rate": 7.499517140843968e-05,
|
| 1284 |
+
"loss": 0.5159,
|
| 1285 |
"step": 900
|
| 1286 |
},
|
| 1287 |
{
|
| 1288 |
"epoch": 1.2519903080650745,
|
| 1289 |
+
"grad_norm": 0.40865471959114075,
|
| 1290 |
"learning_rate": 7.382620636193438e-05,
|
| 1291 |
+
"loss": 0.494,
|
| 1292 |
"step": 905
|
| 1293 |
},
|
| 1294 |
{
|
| 1295 |
"epoch": 1.2589131187262028,
|
| 1296 |
+
"grad_norm": 0.4230162501335144,
|
| 1297 |
"learning_rate": 7.266106853019759e-05,
|
| 1298 |
+
"loss": 0.5007,
|
| 1299 |
"step": 910
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
"epoch": 1.2658359293873311,
|
| 1303 |
+
"grad_norm": 0.4216570258140564,
|
| 1304 |
"learning_rate": 7.149992828336274e-05,
|
| 1305 |
+
"loss": 0.5048,
|
| 1306 |
"step": 915
|
| 1307 |
},
|
| 1308 |
{
|
| 1309 |
"epoch": 1.2727587400484597,
|
| 1310 |
+
"grad_norm": 0.41065993905067444,
|
| 1311 |
"learning_rate": 7.034295540702397e-05,
|
| 1312 |
+
"loss": 0.4963,
|
| 1313 |
"step": 920
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
"epoch": 1.2796815507095882,
|
| 1317 |
+
"grad_norm": 0.4277260899543762,
|
| 1318 |
"learning_rate": 6.919031907740922e-05,
|
| 1319 |
+
"loss": 0.5049,
|
| 1320 |
"step": 925
|
| 1321 |
},
|
| 1322 |
{
|
| 1323 |
"epoch": 1.2866043613707165,
|
| 1324 |
+
"grad_norm": 0.4380500018596649,
|
| 1325 |
"learning_rate": 6.804218783664288e-05,
|
| 1326 |
+
"loss": 0.5168,
|
| 1327 |
"step": 930
|
| 1328 |
},
|
| 1329 |
{
|
| 1330 |
"epoch": 1.2935271720318449,
|
| 1331 |
+
"grad_norm": 0.40834662318229675,
|
| 1332 |
"learning_rate": 6.689872956810102e-05,
|
| 1333 |
+
"loss": 0.507,
|
| 1334 |
"step": 935
|
| 1335 |
},
|
| 1336 |
{
|
| 1337 |
"epoch": 1.3004499826929734,
|
| 1338 |
+
"grad_norm": 0.39029037952423096,
|
| 1339 |
"learning_rate": 6.5760111471863e-05,
|
| 1340 |
+
"loss": 0.5016,
|
| 1341 |
"step": 940
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
"epoch": 1.3073727933541017,
|
| 1345 |
+
"grad_norm": 0.40143686532974243,
|
| 1346 |
"learning_rate": 6.46265000402628e-05,
|
| 1347 |
+
"loss": 0.5116,
|
| 1348 |
"step": 945
|
| 1349 |
},
|
| 1350 |
{
|
| 1351 |
"epoch": 1.3142956040152303,
|
| 1352 |
+
"grad_norm": 0.4155568480491638,
|
| 1353 |
"learning_rate": 6.349806103354417e-05,
|
| 1354 |
+
"loss": 0.499,
|
| 1355 |
"step": 950
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 1.3212184146763586,
|
| 1359 |
+
"grad_norm": 0.44305872917175293,
|
| 1360 |
"learning_rate": 6.23749594556226e-05,
|
| 1361 |
+
"loss": 0.4959,
|
| 1362 |
"step": 955
|
| 1363 |
},
|
| 1364 |
{
|
| 1365 |
"epoch": 1.328141225337487,
|
| 1366 |
+
"grad_norm": 0.410986065864563,
|
| 1367 |
"learning_rate": 6.125735952995775e-05,
|
| 1368 |
+
"loss": 0.4927,
|
| 1369 |
"step": 960
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 1.3350640359986155,
|
| 1373 |
+
"grad_norm": 0.42151662707328796,
|
| 1374 |
"learning_rate": 6.0145424675540394e-05,
|
| 1375 |
+
"loss": 0.4949,
|
| 1376 |
"step": 965
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
"epoch": 1.341986846659744,
|
| 1380 |
+
"grad_norm": 0.3956366181373596,
|
| 1381 |
"learning_rate": 5.9039317482996735e-05,
|
| 1382 |
+
"loss": 0.5057,
|
| 1383 |
"step": 970
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 1.3489096573208723,
|
| 1387 |
+
"grad_norm": 0.4053477644920349,
|
| 1388 |
"learning_rate": 5.793919969081374e-05,
|
| 1389 |
+
"loss": 0.4875,
|
| 1390 |
"step": 975
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 1.3558324679820006,
|
| 1394 |
+
"grad_norm": 0.3788825273513794,
|
| 1395 |
"learning_rate": 5.68452321616894e-05,
|
| 1396 |
+
"loss": 0.4808,
|
| 1397 |
"step": 980
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 1.3627552786431292,
|
| 1401 |
+
"grad_norm": 0.3918589651584625,
|
| 1402 |
"learning_rate": 5.5757574859010886e-05,
|
| 1403 |
+
"loss": 0.4953,
|
| 1404 |
"step": 985
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 1.3696780893042575,
|
| 1408 |
+
"grad_norm": 0.39359208941459656,
|
| 1409 |
"learning_rate": 5.467638682346403e-05,
|
| 1410 |
+
"loss": 0.4777,
|
| 1411 |
"step": 990
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 1.376600899965386,
|
| 1415 |
+
"grad_norm": 0.3955044448375702,
|
| 1416 |
"learning_rate": 5.3601826149777966e-05,
|
| 1417 |
+
"loss": 0.4992,
|
| 1418 |
"step": 995
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
"epoch": 1.3835237106265144,
|
| 1422 |
+
"grad_norm": 0.4178461730480194,
|
| 1423 |
"learning_rate": 5.25340499636082e-05,
|
| 1424 |
"loss": 0.485,
|
| 1425 |
"step": 1000
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 1.3904465212876427,
|
| 1429 |
+
"grad_norm": 0.40446537733078003,
|
| 1430 |
"learning_rate": 5.147321439856091e-05,
|
| 1431 |
+
"loss": 0.4945,
|
| 1432 |
"step": 1005
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 1.3973693319487712,
|
| 1436 |
+
"grad_norm": 0.40754058957099915,
|
| 1437 |
"learning_rate": 5.041947457336274e-05,
|
| 1438 |
+
"loss": 0.5124,
|
| 1439 |
"step": 1010
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 1.4042921426098995,
|
| 1443 |
+
"grad_norm": 0.3956260681152344,
|
| 1444 |
"learning_rate": 4.937298456917889e-05,
|
| 1445 |
+
"loss": 0.4949,
|
| 1446 |
"step": 1015
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 1.411214953271028,
|
| 1450 |
+
"grad_norm": 0.45337051153182983,
|
| 1451 |
"learning_rate": 4.8333897407082896e-05,
|
| 1452 |
+
"loss": 0.4888,
|
| 1453 |
"step": 1020
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 1.4181377639321564,
|
| 1457 |
+
"grad_norm": 0.38214102387428284,
|
| 1458 |
"learning_rate": 4.7302365025681206e-05,
|
| 1459 |
+
"loss": 0.4924,
|
| 1460 |
"step": 1025
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 1.425060574593285,
|
| 1464 |
+
"grad_norm": 0.4063718020915985,
|
| 1465 |
"learning_rate": 4.627853825889642e-05,
|
| 1466 |
+
"loss": 0.4932,
|
| 1467 |
"step": 1030
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 1.4319833852544133,
|
| 1471 |
+
"grad_norm": 0.42218196392059326,
|
| 1472 |
"learning_rate": 4.526256681391192e-05,
|
| 1473 |
+
"loss": 0.4843,
|
| 1474 |
"step": 1035
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 1.4389061959155418,
|
| 1478 |
+
"grad_norm": 0.43437570333480835,
|
| 1479 |
"learning_rate": 4.4254599249281016e-05,
|
| 1480 |
+
"loss": 0.4744,
|
| 1481 |
"step": 1040
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 1.4458290065766701,
|
| 1485 |
+
"grad_norm": 0.3987582325935364,
|
| 1486 |
"learning_rate": 4.3254782953204375e-05,
|
| 1487 |
+
"loss": 0.4845,
|
| 1488 |
"step": 1045
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 1.4527518172377984,
|
| 1492 |
+
"grad_norm": 0.42256399989128113,
|
| 1493 |
"learning_rate": 4.2263264121978505e-05,
|
| 1494 |
+
"loss": 0.5006,
|
| 1495 |
"step": 1050
|
| 1496 |
},
|
| 1497 |
{
|
| 1498 |
"epoch": 1.459674627898927,
|
| 1499 |
+
"grad_norm": 0.37744051218032837,
|
| 1500 |
"learning_rate": 4.12801877386183e-05,
|
| 1501 |
+
"loss": 0.477,
|
| 1502 |
"step": 1055
|
| 1503 |
},
|
| 1504 |
{
|
| 1505 |
"epoch": 1.4665974385600553,
|
| 1506 |
+
"grad_norm": 0.4157800078392029,
|
| 1507 |
"learning_rate": 4.030569755165736e-05,
|
| 1508 |
+
"loss": 0.4891,
|
| 1509 |
"step": 1060
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
"epoch": 1.4735202492211839,
|
| 1513 |
+
"grad_norm": 0.39397871494293213,
|
| 1514 |
"learning_rate": 3.9339936054128466e-05,
|
| 1515 |
+
"loss": 0.4797,
|
| 1516 |
"step": 1065
|
| 1517 |
},
|
| 1518 |
{
|
| 1519 |
"epoch": 1.4804430598823122,
|
| 1520 |
+
"grad_norm": 0.39519259333610535,
|
| 1521 |
"learning_rate": 3.8383044462727826e-05,
|
| 1522 |
+
"loss": 0.4682,
|
| 1523 |
"step": 1070
|
| 1524 |
},
|
| 1525 |
{
|
| 1526 |
"epoch": 1.4873658705434407,
|
| 1527 |
+
"grad_norm": 0.4494350850582123,
|
| 1528 |
"learning_rate": 3.7435162697166107e-05,
|
| 1529 |
+
"loss": 0.4904,
|
| 1530 |
"step": 1075
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 1.494288681204569,
|
| 1534 |
+
"grad_norm": 0.3966675400733948,
|
| 1535 |
"learning_rate": 3.649642935970859e-05,
|
| 1536 |
+
"loss": 0.4864,
|
| 1537 |
"step": 1080
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
"epoch": 1.5012114918656976,
|
| 1541 |
+
"grad_norm": 0.3766285181045532,
|
| 1542 |
"learning_rate": 3.556698171490871e-05,
|
| 1543 |
+
"loss": 0.478,
|
| 1544 |
"step": 1085
|
| 1545 |
},
|
| 1546 |
{
|
| 1547 |
"epoch": 1.508134302526826,
|
| 1548 |
+
"grad_norm": 0.3993706703186035,
|
| 1549 |
"learning_rate": 3.464695566953644e-05,
|
| 1550 |
+
"loss": 0.4912,
|
| 1551 |
"step": 1090
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
"epoch": 1.5150571131879542,
|
| 1555 |
+
"grad_norm": 0.41815367341041565,
|
| 1556 |
"learning_rate": 3.373648575270576e-05,
|
| 1557 |
+
"loss": 0.486,
|
| 1558 |
"step": 1095
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
"epoch": 1.5219799238490828,
|
| 1562 |
+
"grad_norm": 0.4036204516887665,
|
| 1563 |
"learning_rate": 3.283570509620344e-05,
|
| 1564 |
+
"loss": 0.4851,
|
| 1565 |
"step": 1100
|
| 1566 |
},
|
| 1567 |
{
|
| 1568 |
"epoch": 1.5289027345102113,
|
| 1569 |
+
"grad_norm": 0.39286836981773376,
|
| 1570 |
"learning_rate": 3.1944745415021915e-05,
|
| 1571 |
+
"loss": 0.4807,
|
| 1572 |
"step": 1105
|
| 1573 |
},
|
| 1574 |
{
|
| 1575 |
"epoch": 1.5358255451713396,
|
| 1576 |
+
"grad_norm": 0.3978467881679535,
|
| 1577 |
"learning_rate": 3.106373698809958e-05,
|
| 1578 |
+
"loss": 0.4763,
|
| 1579 |
"step": 1110
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
"epoch": 1.542748355832468,
|
| 1583 |
+
"grad_norm": 0.389644056558609,
|
| 1584 |
"learning_rate": 3.0192808639271065e-05,
|
| 1585 |
+
"loss": 0.4812,
|
| 1586 |
"step": 1115
|
| 1587 |
},
|
| 1588 |
{
|
| 1589 |
"epoch": 1.5496711664935963,
|
| 1590 |
+
"grad_norm": 0.3576400876045227,
|
| 1591 |
"learning_rate": 2.933208771843018e-05,
|
| 1592 |
+
"loss": 0.4883,
|
| 1593 |
"step": 1120
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 1.5565939771547248,
|
| 1597 |
+
"grad_norm": 0.3846583068370819,
|
| 1598 |
"learning_rate": 2.8481700082908268e-05,
|
| 1599 |
+
"loss": 0.4731,
|
| 1600 |
"step": 1125
|
| 1601 |
},
|
| 1602 |
{
|
| 1603 |
"epoch": 1.5635167878158533,
|
| 1604 |
+
"grad_norm": 0.3483330309391022,
|
| 1605 |
"learning_rate": 2.7641770079071127e-05,
|
| 1606 |
+
"loss": 0.4695,
|
| 1607 |
"step": 1130
|
| 1608 |
},
|
| 1609 |
{
|
| 1610 |
"epoch": 1.5704395984769817,
|
| 1611 |
+
"grad_norm": 0.3988693654537201,
|
| 1612 |
"learning_rate": 2.681242052413656e-05,
|
| 1613 |
+
"loss": 0.4717,
|
| 1614 |
"step": 1135
|
| 1615 |
},
|
| 1616 |
{
|
| 1617 |
"epoch": 1.57736240913811,
|
| 1618 |
+
"grad_norm": 0.3845638334751129,
|
| 1619 |
"learning_rate": 2.59937726882158e-05,
|
| 1620 |
+
"loss": 0.4769,
|
| 1621 |
"step": 1140
|
| 1622 |
},
|
| 1623 |
{
|
| 1624 |
"epoch": 1.5842852197992385,
|
| 1625 |
+
"grad_norm": 0.39838650822639465,
|
| 1626 |
"learning_rate": 2.5185946276580918e-05,
|
| 1627 |
+
"loss": 0.4745,
|
| 1628 |
"step": 1145
|
| 1629 |
},
|
| 1630 |
{
|
| 1631 |
"epoch": 1.591208030460367,
|
| 1632 |
+
"grad_norm": 0.404542475938797,
|
| 1633 |
"learning_rate": 2.4389059412161087e-05,
|
| 1634 |
"loss": 0.4736,
|
| 1635 |
"step": 1150
|
| 1636 |
},
|
| 1637 |
{
|
| 1638 |
"epoch": 1.5981308411214954,
|
| 1639 |
+
"grad_norm": 0.385309100151062,
|
| 1640 |
"learning_rate": 2.3603228618270356e-05,
|
| 1641 |
+
"loss": 0.4769,
|
| 1642 |
"step": 1155
|
| 1643 |
},
|
| 1644 |
{
|
| 1645 |
"epoch": 1.6050536517826237,
|
| 1646 |
+
"grad_norm": 0.4078046679496765,
|
| 1647 |
"learning_rate": 2.2828568801569283e-05,
|
| 1648 |
+
"loss": 0.4577,
|
| 1649 |
"step": 1160
|
| 1650 |
},
|
| 1651 |
{
|
| 1652 |
"epoch": 1.611976462443752,
|
| 1653 |
+
"grad_norm": 0.38839876651763916,
|
| 1654 |
"learning_rate": 2.2065193235262782e-05,
|
| 1655 |
+
"loss": 0.4712,
|
| 1656 |
"step": 1165
|
| 1657 |
},
|
| 1658 |
{
|
| 1659 |
"epoch": 1.6188992731048806,
|
| 1660 |
+
"grad_norm": 0.41136986017227173,
|
| 1661 |
"learning_rate": 2.1313213542537058e-05,
|
| 1662 |
+
"loss": 0.4677,
|
| 1663 |
"step": 1170
|
| 1664 |
},
|
| 1665 |
{
|
| 1666 |
"epoch": 1.6258220837660091,
|
| 1667 |
+
"grad_norm": 0.41455721855163574,
|
| 1668 |
"learning_rate": 2.0572739680237717e-05,
|
| 1669 |
+
"loss": 0.4788,
|
| 1670 |
"step": 1175
|
| 1671 |
},
|
| 1672 |
{
|
| 1673 |
"epoch": 1.6327448944271374,
|
| 1674 |
+
"grad_norm": 0.39108678698539734,
|
| 1675 |
"learning_rate": 1.984387992279153e-05,
|
| 1676 |
+
"loss": 0.4776,
|
| 1677 |
"step": 1180
|
| 1678 |
},
|
| 1679 |
{
|
| 1680 |
"epoch": 1.6396677050882658,
|
| 1681 |
+
"grad_norm": 0.40750446915626526,
|
| 1682 |
"learning_rate": 1.912674084637409e-05,
|
| 1683 |
+
"loss": 0.4618,
|
| 1684 |
"step": 1185
|
| 1685 |
},
|
| 1686 |
{
|
| 1687 |
"epoch": 1.6465905157493943,
|
| 1688 |
+
"grad_norm": 0.389186829328537,
|
| 1689 |
"learning_rate": 1.8421427313326046e-05,
|
| 1690 |
+
"loss": 0.4724,
|
| 1691 |
"step": 1190
|
| 1692 |
},
|
| 1693 |
{
|
| 1694 |
"epoch": 1.6535133264105226,
|
| 1695 |
+
"grad_norm": 0.40712910890579224,
|
| 1696 |
"learning_rate": 1.7728042456819704e-05,
|
| 1697 |
+
"loss": 0.4702,
|
| 1698 |
"step": 1195
|
| 1699 |
},
|
| 1700 |
{
|
| 1701 |
"epoch": 1.6604361370716512,
|
| 1702 |
+
"grad_norm": 0.37883758544921875,
|
| 1703 |
"learning_rate": 1.7046687665778717e-05,
|
| 1704 |
+
"loss": 0.4648,
|
| 1705 |
"step": 1200
|
| 1706 |
},
|
| 1707 |
{
|
| 1708 |
"epoch": 1.6673589477327795,
|
| 1709 |
+
"grad_norm": 0.3762923777103424,
|
| 1710 |
"learning_rate": 1.6377462570052438e-05,
|
| 1711 |
+
"loss": 0.4645,
|
| 1712 |
"step": 1205
|
| 1713 |
},
|
| 1714 |
{
|
| 1715 |
"epoch": 1.6742817583939078,
|
| 1716 |
+
"grad_norm": 0.3841410279273987,
|
| 1717 |
"learning_rate": 1.5720465025847987e-05,
|
| 1718 |
+
"loss": 0.4706,
|
| 1719 |
"step": 1210
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
"epoch": 1.6812045690550363,
|
| 1723 |
+
"grad_norm": 0.3858813941478729,
|
| 1724 |
"learning_rate": 1.5075791101421166e-05,
|
| 1725 |
+
"loss": 0.468,
|
| 1726 |
"step": 1215
|
| 1727 |
},
|
| 1728 |
{
|
| 1729 |
"epoch": 1.6881273797161649,
|
| 1730 |
+
"grad_norm": 0.3854951858520508,
|
| 1731 |
"learning_rate": 1.4443535063029279e-05,
|
| 1732 |
+
"loss": 0.4646,
|
| 1733 |
"step": 1220
|
| 1734 |
},
|
| 1735 |
{
|
| 1736 |
"epoch": 1.6950501903772932,
|
| 1737 |
+
"grad_norm": 0.39483922719955444,
|
| 1738 |
"learning_rate": 1.3823789361147044e-05,
|
| 1739 |
+
"loss": 0.4657,
|
| 1740 |
"step": 1225
|
| 1741 |
},
|
| 1742 |
{
|
| 1743 |
"epoch": 1.7019730010384215,
|
| 1744 |
+
"grad_norm": 0.3647611737251282,
|
| 1745 |
"learning_rate": 1.3216644616948282e-05,
|
| 1746 |
+
"loss": 0.4665,
|
| 1747 |
"step": 1230
|
| 1748 |
},
|
| 1749 |
{
|
| 1750 |
"epoch": 1.70889581169955,
|
| 1751 |
+
"grad_norm": 0.3819540739059448,
|
| 1752 |
"learning_rate": 1.2622189609054979e-05,
|
| 1753 |
+
"loss": 0.4631,
|
| 1754 |
"step": 1235
|
| 1755 |
},
|
| 1756 |
{
|
| 1757 |
"epoch": 1.7158186223606784,
|
| 1758 |
+
"grad_norm": 0.394875168800354,
|
| 1759 |
"learning_rate": 1.2040511260555831e-05,
|
| 1760 |
+
"loss": 0.4546,
|
| 1761 |
"step": 1240
|
| 1762 |
},
|
| 1763 |
{
|
| 1764 |
"epoch": 1.722741433021807,
|
| 1765 |
+
"grad_norm": 0.4007471203804016,
|
| 1766 |
"learning_rate": 1.1471694626296049e-05,
|
| 1767 |
+
"loss": 0.4604,
|
| 1768 |
"step": 1245
|
| 1769 |
},
|
| 1770 |
{
|
| 1771 |
"epoch": 1.7296642436829353,
|
| 1772 |
+
"grad_norm": 0.39787620306015015,
|
| 1773 |
"learning_rate": 1.0915822880440308e-05,
|
| 1774 |
+
"loss": 0.4568,
|
| 1775 |
"step": 1250
|
| 1776 |
},
|
| 1777 |
{
|
| 1778 |
"epoch": 1.7365870543440636,
|
| 1779 |
+
"grad_norm": 0.37953585386276245,
|
| 1780 |
"learning_rate": 1.037297730431085e-05,
|
| 1781 |
+
"loss": 0.443,
|
| 1782 |
"step": 1255
|
| 1783 |
},
|
| 1784 |
{
|
| 1785 |
"epoch": 1.7435098650051921,
|
| 1786 |
+
"grad_norm": 0.38411709666252136,
|
| 1787 |
"learning_rate": 9.843237274502282e-06,
|
| 1788 |
+
"loss": 0.4734,
|
| 1789 |
"step": 1260
|
| 1790 |
},
|
| 1791 |
{
|
| 1792 |
"epoch": 1.7504326756663207,
|
| 1793 |
+
"grad_norm": 0.40413138270378113,
|
| 1794 |
"learning_rate": 9.326680251274777e-06,
|
| 1795 |
+
"loss": 0.4563,
|
| 1796 |
"step": 1265
|
| 1797 |
},
|
| 1798 |
{
|
| 1799 |
"epoch": 1.757355486327449,
|
| 1800 |
+
"grad_norm": 0.39860308170318604,
|
| 1801 |
"learning_rate": 8.823381767227667e-06,
|
| 1802 |
+
"loss": 0.4668,
|
| 1803 |
"step": 1270
|
| 1804 |
},
|
| 1805 |
{
|
| 1806 |
"epoch": 1.7642782969885773,
|
| 1807 |
+
"grad_norm": 0.37807101011276245,
|
| 1808 |
"learning_rate": 8.333415416254831e-06,
|
| 1809 |
+
"loss": 0.4599,
|
| 1810 |
"step": 1275
|
| 1811 |
},
|
| 1812 |
{
|
| 1813 |
"epoch": 1.7712011076497056,
|
| 1814 |
+
"grad_norm": 0.3738378882408142,
|
| 1815 |
"learning_rate": 7.856852842783547e-06,
|
| 1816 |
+
"loss": 0.4649,
|
| 1817 |
"step": 1280
|
| 1818 |
},
|
| 1819 |
{
|
| 1820 |
"epoch": 1.7781239183108342,
|
| 1821 |
+
"grad_norm": 0.38058996200561523,
|
| 1822 |
"learning_rate": 7.393763731298387e-06,
|
| 1823 |
+
"loss": 0.4811,
|
| 1824 |
"step": 1285
|
| 1825 |
},
|
| 1826 |
{
|
| 1827 |
"epoch": 1.7850467289719627,
|
| 1828 |
+
"grad_norm": 0.3849298059940338,
|
| 1829 |
"learning_rate": 6.944215796151765e-06,
|
| 1830 |
+
"loss": 0.4643,
|
| 1831 |
"step": 1290
|
| 1832 |
},
|
| 1833 |
{
|
| 1834 |
"epoch": 1.791969539633091,
|
| 1835 |
+
"grad_norm": 0.3965122699737549,
|
| 1836 |
"learning_rate": 6.5082747716625255e-06,
|
| 1837 |
+
"loss": 0.4534,
|
| 1838 |
"step": 1295
|
| 1839 |
},
|
| 1840 |
{
|
| 1841 |
"epoch": 1.7988923502942193,
|
| 1842 |
+
"grad_norm": 0.3806658387184143,
|
| 1843 |
"learning_rate": 6.086004402504098e-06,
|
| 1844 |
+
"loss": 0.4668,
|
| 1845 |
"step": 1300
|
| 1846 |
},
|
| 1847 |
{
|
| 1848 |
"epoch": 1.8058151609553479,
|
| 1849 |
+
"grad_norm": 0.3792901039123535,
|
| 1850 |
"learning_rate": 5.6774664343834495e-06,
|
| 1851 |
+
"loss": 0.4648,
|
| 1852 |
"step": 1305
|
| 1853 |
},
|
| 1854 |
{
|
| 1855 |
"epoch": 1.8127379716164764,
|
| 1856 |
+
"grad_norm": 0.37997448444366455,
|
| 1857 |
"learning_rate": 5.2827206050125585e-06,
|
| 1858 |
+
"loss": 0.4771,
|
| 1859 |
"step": 1310
|
| 1860 |
},
|
| 1861 |
{
|
| 1862 |
"epoch": 1.8196607822776047,
|
| 1863 |
+
"grad_norm": 0.3826618492603302,
|
| 1864 |
"learning_rate": 4.9018246353732755e-06,
|
| 1865 |
+
"loss": 0.475,
|
| 1866 |
"step": 1315
|
| 1867 |
},
|
| 1868 |
{
|
| 1869 |
"epoch": 1.826583592938733,
|
| 1870 |
+
"grad_norm": 0.3670499622821808,
|
| 1871 |
"learning_rate": 4.534834221277262e-06,
|
| 1872 |
+
"loss": 0.4569,
|
| 1873 |
"step": 1320
|
| 1874 |
},
|
| 1875 |
{
|
| 1876 |
"epoch": 1.8335064035998614,
|
| 1877 |
+
"grad_norm": 0.3672844469547272,
|
| 1878 |
"learning_rate": 4.181803025221898e-06,
|
| 1879 |
+
"loss": 0.4691,
|
| 1880 |
"step": 1325
|
| 1881 |
},
|
| 1882 |
{
|
| 1883 |
"epoch": 1.84042921426099,
|
| 1884 |
+
"grad_norm": 0.37513473629951477,
|
| 1885 |
"learning_rate": 3.842782668543599e-06,
|
| 1886 |
+
"loss": 0.4564,
|
| 1887 |
"step": 1330
|
| 1888 |
},
|
| 1889 |
{
|
| 1890 |
"epoch": 1.8473520249221185,
|
| 1891 |
+
"grad_norm": 0.3402206301689148,
|
| 1892 |
"learning_rate": 3.5178227238696347e-06,
|
| 1893 |
+
"loss": 0.4712,
|
| 1894 |
"step": 1335
|
| 1895 |
},
|
| 1896 |
{
|
| 1897 |
"epoch": 1.8542748355832468,
|
| 1898 |
+
"grad_norm": 0.35512781143188477,
|
| 1899 |
"learning_rate": 3.2069707078694057e-06,
|
| 1900 |
+
"loss": 0.4505,
|
| 1901 |
"step": 1340
|
| 1902 |
},
|
| 1903 |
{
|
| 1904 |
"epoch": 1.8611976462443751,
|
| 1905 |
+
"grad_norm": 0.3555276095867157,
|
| 1906 |
"learning_rate": 2.9102720743064527e-06,
|
| 1907 |
+
"loss": 0.4526,
|
| 1908 |
"step": 1345
|
| 1909 |
},
|
| 1910 |
{
|
| 1911 |
"epoch": 1.8681204569055037,
|
| 1912 |
+
"grad_norm": 0.3737949728965759,
|
| 1913 |
"learning_rate": 2.6277702073919997e-06,
|
| 1914 |
+
"loss": 0.456,
|
| 1915 |
"step": 1350
|
| 1916 |
},
|
| 1917 |
{
|
| 1918 |
"epoch": 1.8750432675666322,
|
| 1919 |
+
"grad_norm": 0.3693467378616333,
|
| 1920 |
"learning_rate": 2.3595064154412374e-06,
|
| 1921 |
+
"loss": 0.4629,
|
| 1922 |
"step": 1355
|
| 1923 |
},
|
| 1924 |
{
|
| 1925 |
"epoch": 1.8819660782277605,
|
| 1926 |
+
"grad_norm": 0.3651539385318756,
|
| 1927 |
"learning_rate": 2.1055199248330546e-06,
|
| 1928 |
+
"loss": 0.449,
|
| 1929 |
"step": 1360
|
| 1930 |
},
|
| 1931 |
{
|
| 1932 |
"epoch": 1.8888888888888888,
|
| 1933 |
+
"grad_norm": 0.38853177428245544,
|
| 1934 |
"learning_rate": 1.8658478742742625e-06,
|
| 1935 |
+
"loss": 0.4595,
|
| 1936 |
"step": 1365
|
| 1937 |
},
|
| 1938 |
{
|
| 1939 |
"epoch": 1.8958116995500172,
|
| 1940 |
+
"grad_norm": 0.3724125325679779,
|
| 1941 |
"learning_rate": 1.6405253093690343e-06,
|
| 1942 |
+
"loss": 0.4637,
|
| 1943 |
"step": 1370
|
| 1944 |
},
|
| 1945 |
{
|
| 1946 |
"epoch": 1.9027345102111457,
|
| 1947 |
+
"grad_norm": 0.39097484946250916,
|
| 1948 |
"learning_rate": 1.429585177494419e-06,
|
| 1949 |
+
"loss": 0.46,
|
| 1950 |
"step": 1375
|
| 1951 |
},
|
| 1952 |
{
|
| 1953 |
"epoch": 1.9096573208722742,
|
| 1954 |
+
"grad_norm": 0.40443167090415955,
|
| 1955 |
"learning_rate": 1.2330583229827053e-06,
|
| 1956 |
+
"loss": 0.4738,
|
| 1957 |
"step": 1380
|
| 1958 |
},
|
| 1959 |
{
|
| 1960 |
"epoch": 1.9165801315334026,
|
| 1961 |
+
"grad_norm": 0.37939175963401794,
|
| 1962 |
"learning_rate": 1.0509734826112394e-06,
|
| 1963 |
+
"loss": 0.4544,
|
| 1964 |
"step": 1385
|
| 1965 |
},
|
| 1966 |
{
|
| 1967 |
"epoch": 1.9235029421945309,
|
| 1968 |
+
"grad_norm": 0.39131584763526917,
|
| 1969 |
"learning_rate": 8.833572814004321e-07,
|
| 1970 |
+
"loss": 0.4436,
|
| 1971 |
"step": 1390
|
| 1972 |
},
|
| 1973 |
{
|
| 1974 |
"epoch": 1.9304257528556594,
|
| 1975 |
+
"grad_norm": 0.3646048903465271,
|
| 1976 |
"learning_rate": 7.302342287205943e-07,
|
| 1977 |
+
"loss": 0.4543,
|
| 1978 |
"step": 1395
|
| 1979 |
},
|
| 1980 |
{
|
| 1981 |
"epoch": 1.937348563516788,
|
| 1982 |
+
"grad_norm": 0.36286523938179016,
|
| 1983 |
"learning_rate": 5.916267147080934e-07,
|
| 1984 |
+
"loss": 0.4688,
|
| 1985 |
"step": 1400
|
| 1986 |
},
|
| 1987 |
{
|
| 1988 |
"epoch": 1.9442713741779163,
|
| 1989 |
+
"grad_norm": 0.39855965971946716,
|
| 1990 |
"learning_rate": 4.6755500699138253e-07,
|
| 1991 |
+
"loss": 0.4583,
|
| 1992 |
"step": 1405
|
| 1993 |
},
|
| 1994 |
{
|
| 1995 |
"epoch": 1.9511941848390446,
|
| 1996 |
+
"grad_norm": 0.37583374977111816,
|
| 1997 |
"learning_rate": 3.5803724772740523e-07,
|
| 1998 |
+
"loss": 0.4489,
|
| 1999 |
"step": 1410
|
| 2000 |
},
|
| 2001 |
{
|
| 2002 |
"epoch": 1.958116995500173,
|
| 2003 |
+
"grad_norm": 0.3726235330104828,
|
| 2004 |
"learning_rate": 2.630894509488058e-07,
|
| 2005 |
+
"loss": 0.4689,
|
| 2006 |
"step": 1415
|
| 2007 |
},
|
| 2008 |
{
|
| 2009 |
"epoch": 1.9650398061613015,
|
| 2010 |
+
"grad_norm": 0.3750331997871399,
|
| 2011 |
"learning_rate": 1.8272550022230272e-07,
|
| 2012 |
+
"loss": 0.4647,
|
| 2013 |
"step": 1420
|
| 2014 |
},
|
| 2015 |
{
|
| 2016 |
"epoch": 1.97196261682243,
|
| 2017 |
+
"grad_norm": 0.3834379017353058,
|
| 2018 |
"learning_rate": 1.1695714661856816e-07,
|
| 2019 |
+
"loss": 0.4665,
|
| 2020 |
"step": 1425
|
| 2021 |
},
|
| 2022 |
{
|
| 2023 |
"epoch": 1.9788854274835583,
|
| 2024 |
+
"grad_norm": 0.37510913610458374,
|
| 2025 |
"learning_rate": 6.579400699397998e-08,
|
| 2026 |
+
"loss": 0.459,
|
| 2027 |
"step": 1430
|
| 2028 |
},
|
| 2029 |
{
|
| 2030 |
"epoch": 1.9858082381446867,
|
| 2031 |
+
"grad_norm": 0.36773183941841125,
|
| 2032 |
"learning_rate": 2.9243562584402305e-08,
|
| 2033 |
+
"loss": 0.4565,
|
| 2034 |
"step": 1435
|
| 2035 |
},
|
| 2036 |
{
|
| 2037 |
"epoch": 1.9927310488058152,
|
| 2038 |
+
"grad_norm": 0.3669570982456207,
|
| 2039 |
"learning_rate": 7.311157911249478e-09,
|
| 2040 |
+
"loss": 0.4612,
|
| 2041 |
"step": 1440
|
| 2042 |
},
|
| 2043 |
{
|
| 2044 |
"epoch": 1.9982692973347178,
|
| 2045 |
+
"eval_loss": 0.5933529138565063,
|
| 2046 |
+
"eval_runtime": 75.4661,
|
| 2047 |
+
"eval_samples_per_second": 27.416,
|
| 2048 |
+
"eval_steps_per_second": 1.723,
|
| 2049 |
"step": 1444
|
| 2050 |
},
|
| 2051 |
{
|
| 2052 |
"epoch": 1.9982692973347178,
|
| 2053 |
"step": 1444,
|
| 2054 |
"total_flos": 6.774706456143856e+18,
|
| 2055 |
+
"train_loss": 0.6556608070957364,
|
| 2056 |
+
"train_runtime": 13724.1641,
|
| 2057 |
+
"train_samples_per_second": 6.735,
|
| 2058 |
+
"train_steps_per_second": 0.105
|
| 2059 |
}
|
| 2060 |
],
|
| 2061 |
"logging_steps": 5,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7096
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72fb31e267db59548aa7312200eb1ac0e3106c16bc436709ceea41c5e8b30ac1
|
| 3 |
size 7096
|