End of training
Browse files- README.md +26 -46
- model.safetensors +1 -1
- runs/Dec24_15-59-58_DESKTOP-A45193E/events.out.tfevents.1703426400.DESKTOP-A45193E +3 -0
- trainer_state.json +934 -1128
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.
|
| 19 |
- Mean Distance: 0
|
| 20 |
- Max Distance: 1
|
| 21 |
|
|
@@ -37,58 +37,38 @@ More information needed
|
|
| 37 |
|
| 38 |
The following hyperparameters were used during training:
|
| 39 |
- learning_rate: 0.0001
|
| 40 |
-
- train_batch_size:
|
| 41 |
-
- eval_batch_size:
|
| 42 |
- seed: 42
|
| 43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
- lr_scheduler_warmup_ratio: 0.1
|
| 46 |
-
- num_epochs:
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
-
| Training Loss | Epoch | Step
|
| 51 |
-
|:-------------:|:-----:|:-----
|
| 52 |
-
| 0.
|
| 53 |
-
| 0.
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
| 62 |
-
| 0.
|
| 63 |
-
| 0.
|
| 64 |
-
| 0.
|
| 65 |
-
| 0.
|
| 66 |
-
| 0.
|
| 67 |
-
| 0.
|
| 68 |
-
| 0.
|
| 69 |
-
| 0.
|
| 70 |
-
| 0.
|
| 71 |
-
| 0.
|
| 72 |
-
| 0.0006 | 21.0 | 55944 | 0.0305 | 0 | 1 |
|
| 73 |
-
| 0.001 | 22.0 | 58608 | 0.0292 | 0 | 1 |
|
| 74 |
-
| 0.0008 | 23.0 | 61272 | 0.0373 | 0 | 1 |
|
| 75 |
-
| 0.0008 | 24.0 | 63936 | 0.0309 | 0 | 1 |
|
| 76 |
-
| 0.0008 | 25.0 | 66600 | 0.0385 | 0 | 1 |
|
| 77 |
-
| 0.0014 | 26.0 | 69264 | 0.0134 | 0 | 1 |
|
| 78 |
-
| 0.0004 | 27.0 | 71928 | 0.0239 | 0 | 1 |
|
| 79 |
-
| 0.0011 | 28.0 | 74592 | 0.0164 | 0 | 1 |
|
| 80 |
-
| 0.0002 | 29.0 | 77256 | 0.0186 | 0 | 1 |
|
| 81 |
-
| 0.0001 | 30.0 | 79920 | 0.0298 | 0 | 1 |
|
| 82 |
-
| 0.0008 | 31.0 | 82584 | 0.0277 | 0 | 1 |
|
| 83 |
-
| 0.0003 | 32.0 | 85248 | 0.0377 | 0 | 1 |
|
| 84 |
-
| 0.0003 | 33.0 | 87912 | 0.0354 | 0 | 1 |
|
| 85 |
-
| 0.0007 | 34.0 | 90576 | 0.0585 | 0 | 1 |
|
| 86 |
-
| 0.0005 | 35.0 | 93240 | 0.0568 | 0 | 1 |
|
| 87 |
-
| 0.0001 | 36.0 | 95904 | 0.0567 | 0 | 1 |
|
| 88 |
-
| 0.0009 | 37.0 | 98568 | 0.0605 | 0 | 1 |
|
| 89 |
-
| 0.0002 | 38.0 | 101232 | 0.0613 | 0 | 1 |
|
| 90 |
-
| 0.0002 | 39.0 | 103896 | 0.0563 | 0 | 1 |
|
| 91 |
-
| 0.0002 | 40.0 | 106560 | 0.0632 | 0 | 1 |
|
| 92 |
|
| 93 |
|
| 94 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.0102
|
| 19 |
- Mean Distance: 0
|
| 20 |
- Max Distance: 1
|
| 21 |
|
|
|
|
| 37 |
|
| 38 |
The following hyperparameters were used during training:
|
| 39 |
- learning_rate: 0.0001
|
| 40 |
+
- train_batch_size: 32
|
| 41 |
+
- eval_batch_size: 32
|
| 42 |
- seed: 42
|
| 43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
- lr_scheduler_warmup_ratio: 0.1
|
| 46 |
+
- num_epochs: 20
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
+
| Training Loss | Epoch | Step | Validation Loss | Mean Distance | Max Distance |
|
| 51 |
+
|:-------------:|:-----:|:-----:|:---------------:|:-------------:|:------------:|
|
| 52 |
+
| 0.2779 | 1.0 | 1050 | 0.3371 | 0 | 1 |
|
| 53 |
+
| 0.0346 | 2.0 | 2100 | 0.0173 | 0 | 1 |
|
| 54 |
+
| 0.0072 | 3.0 | 3150 | 0.0077 | 0 | 1 |
|
| 55 |
+
| 0.0057 | 4.0 | 4200 | 0.0093 | 0 | 1 |
|
| 56 |
+
| 0.005 | 5.0 | 5250 | 0.0053 | 0 | 1 |
|
| 57 |
+
| 0.002 | 6.0 | 6300 | 0.0056 | 0 | 1 |
|
| 58 |
+
| 0.002 | 7.0 | 7350 | 0.0091 | 0 | 1 |
|
| 59 |
+
| 0.0018 | 8.0 | 8400 | 0.0010 | 0 | 0 |
|
| 60 |
+
| 0.0044 | 9.0 | 9450 | 0.0043 | 0 | 1 |
|
| 61 |
+
| 0.0024 | 10.0 | 10500 | 0.0048 | 0 | 1 |
|
| 62 |
+
| 0.0032 | 11.0 | 11550 | 0.0023 | 0 | 1 |
|
| 63 |
+
| 0.0028 | 12.0 | 12600 | 0.0003 | 0 | 0 |
|
| 64 |
+
| 0.0005 | 13.0 | 13650 | 0.0008 | 0 | 0 |
|
| 65 |
+
| 0.0026 | 14.0 | 14700 | 0.0012 | 0 | 1 |
|
| 66 |
+
| 0.0011 | 15.0 | 15750 | 0.0171 | 0 | 1 |
|
| 67 |
+
| 0.0 | 16.0 | 16800 | 0.0120 | 0 | 1 |
|
| 68 |
+
| 0.0001 | 17.0 | 17850 | 0.0031 | 0 | 1 |
|
| 69 |
+
| 0.0007 | 18.0 | 18900 | 0.0165 | 0 | 1 |
|
| 70 |
+
| 0.001 | 19.0 | 19950 | 0.0194 | 0 | 1 |
|
| 71 |
+
| 0.0001 | 20.0 | 21000 | 0.0102 | 0 | 1 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
|
| 73 |
|
| 74 |
### Framework versions
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1200729512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3180de15bd1c09f657d097a414082e5c3e98ebfdc101611626b6fa3b1747f1b1
|
| 3 |
size 1200729512
|
runs/Dec24_15-59-58_DESKTOP-A45193E/events.out.tfevents.1703426400.DESKTOP-A45193E
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7842d3f87b74287c68edf2154dcc5ab5d5b83fbe3845a0ab47a267e345b37245
|
| 3 |
+
size 44083
|
trainer_state.json
CHANGED
|
@@ -1,1628 +1,1434 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 30.
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate":
|
| 20 |
-
"loss":
|
| 21 |
-
"step":
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate":
|
| 26 |
-
"loss":
|
| 27 |
-
"step":
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 1.
|
| 32 |
-
"loss":
|
| 33 |
-
"step":
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss":
|
| 39 |
-
"step":
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch":
|
| 43 |
-
"
|
| 44 |
-
"
|
| 45 |
-
"
|
| 46 |
-
"eval_runtime": 1.1154,
|
| 47 |
-
"eval_samples_per_second": 320.968,
|
| 48 |
-
"eval_steps_per_second": 21.517,
|
| 49 |
-
"step": 2664
|
| 50 |
},
|
| 51 |
{
|
| 52 |
-
"epoch":
|
| 53 |
-
"learning_rate":
|
| 54 |
-
"loss":
|
| 55 |
-
"step":
|
| 56 |
},
|
| 57 |
{
|
| 58 |
-
"epoch":
|
| 59 |
-
"learning_rate": 3.
|
| 60 |
-
"loss":
|
| 61 |
-
"step":
|
| 62 |
},
|
| 63 |
{
|
| 64 |
-
"epoch":
|
| 65 |
-
"learning_rate":
|
| 66 |
-
"loss": 0.
|
| 67 |
-
"step":
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch":
|
| 71 |
-
"learning_rate": 4.
|
| 72 |
-
"loss": 0.
|
| 73 |
-
"step":
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 1.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
-
"step":
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch":
|
| 83 |
-
"eval_loss": 0.
|
| 84 |
"eval_max_distance": 1,
|
| 85 |
"eval_mean_distance": 0,
|
| 86 |
-
"eval_runtime": 0.
|
| 87 |
-
"eval_samples_per_second":
|
| 88 |
-
"eval_steps_per_second":
|
| 89 |
-
"step":
|
| 90 |
},
|
| 91 |
{
|
| 92 |
-
"epoch":
|
| 93 |
-
"learning_rate": 5.
|
| 94 |
-
"loss": 0.
|
| 95 |
-
"step":
|
| 96 |
},
|
| 97 |
{
|
| 98 |
-
"epoch":
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss": 0.
|
| 101 |
-
"step":
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch":
|
| 105 |
-
"learning_rate": 6.
|
| 106 |
-
"loss": 0.
|
| 107 |
-
"step":
|
| 108 |
},
|
| 109 |
{
|
| 110 |
-
"epoch":
|
| 111 |
-
"learning_rate":
|
| 112 |
-
"loss": 0.
|
| 113 |
-
"step":
|
| 114 |
},
|
| 115 |
{
|
| 116 |
-
"epoch":
|
| 117 |
-
"learning_rate": 7.
|
| 118 |
-
"loss": 0.
|
| 119 |
-
"step":
|
| 120 |
},
|
| 121 |
{
|
| 122 |
-
"epoch":
|
| 123 |
-
"
|
| 124 |
-
"
|
| 125 |
-
"
|
| 126 |
-
"eval_runtime": 0.9146,
|
| 127 |
-
"eval_samples_per_second": 391.418,
|
| 128 |
-
"eval_steps_per_second": 26.24,
|
| 129 |
-
"step": 7992
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 0.
|
| 135 |
-
"step":
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss": 0.
|
| 141 |
-
"step":
|
| 142 |
-
},
|
| 143 |
-
{
|
| 144 |
-
"epoch": 3.4,
|
| 145 |
-
"learning_rate": 8.50319069069069e-05,
|
| 146 |
-
"loss": 0.012,
|
| 147 |
-
"step": 9061
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch":
|
| 151 |
-
"learning_rate": 9.
|
| 152 |
-
"loss": 0.
|
| 153 |
-
"step":
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch":
|
| 157 |
-
"learning_rate":
|
| 158 |
-
"loss": 0.
|
| 159 |
-
"step":
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"eval_loss": 0.
|
| 164 |
"eval_max_distance": 1,
|
| 165 |
"eval_mean_distance": 0,
|
| 166 |
-
"eval_runtime": 0.
|
| 167 |
-
"eval_samples_per_second":
|
| 168 |
-
"eval_steps_per_second":
|
| 169 |
-
"step":
|
| 170 |
},
|
| 171 |
{
|
| 172 |
-
"epoch":
|
| 173 |
-
"learning_rate": 9.
|
| 174 |
-
"loss": 0.
|
| 175 |
-
"step":
|
| 176 |
},
|
| 177 |
{
|
| 178 |
-
"epoch":
|
| 179 |
-
"learning_rate": 9.
|
| 180 |
-
"loss": 0.
|
| 181 |
-
"step":
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 4.4,
|
| 185 |
-
"learning_rate": 9.88843009676343e-05,
|
| 186 |
-
"loss": 0.0072,
|
| 187 |
-
"step": 11726
|
| 188 |
},
|
| 189 |
{
|
| 190 |
-
"epoch":
|
| 191 |
-
"learning_rate": 9.
|
| 192 |
-
"loss": 0.
|
| 193 |
-
"step":
|
| 194 |
},
|
| 195 |
{
|
| 196 |
-
"epoch":
|
| 197 |
-
"learning_rate": 9.
|
| 198 |
-
"loss": 0.
|
| 199 |
-
"step":
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch":
|
| 203 |
-
"
|
| 204 |
-
"
|
| 205 |
-
"
|
| 206 |
-
"eval_runtime": 0.9141,
|
| 207 |
-
"eval_samples_per_second": 391.658,
|
| 208 |
-
"eval_steps_per_second": 26.256,
|
| 209 |
-
"step": 13320
|
| 210 |
},
|
| 211 |
{
|
| 212 |
-
"epoch":
|
| 213 |
-
"learning_rate": 9.
|
| 214 |
-
"loss": 0.
|
| 215 |
-
"step":
|
| 216 |
},
|
| 217 |
{
|
| 218 |
-
"epoch":
|
| 219 |
-
"learning_rate": 9.
|
| 220 |
-
"loss": 0.
|
| 221 |
-
"step":
|
| 222 |
},
|
| 223 |
{
|
| 224 |
-
"epoch":
|
| 225 |
-
"learning_rate": 9.
|
| 226 |
-
"loss": 0.
|
| 227 |
-
"step":
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch":
|
| 231 |
-
"learning_rate": 9.
|
| 232 |
-
"loss": 0.
|
| 233 |
-
"step":
|
| 234 |
},
|
| 235 |
{
|
| 236 |
-
"epoch":
|
| 237 |
-
"learning_rate": 9.
|
| 238 |
-
"loss": 0.
|
| 239 |
-
"step":
|
| 240 |
},
|
| 241 |
{
|
| 242 |
-
"epoch":
|
| 243 |
-
"eval_loss": 0.
|
| 244 |
"eval_max_distance": 1,
|
| 245 |
"eval_mean_distance": 0,
|
| 246 |
-
"eval_runtime": 0.
|
| 247 |
-
"eval_samples_per_second":
|
| 248 |
-
"eval_steps_per_second":
|
| 249 |
-
"step":
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 6.0,
|
| 253 |
-
"learning_rate": 9.44381881881882e-05,
|
| 254 |
-
"loss": 0.0048,
|
| 255 |
-
"step": 15990
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
-
"learning_rate": 9.
|
| 260 |
-
"loss": 0.
|
| 261 |
-
"step":
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
-
"learning_rate": 9.
|
| 266 |
-
"loss": 0.
|
| 267 |
-
"step":
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch":
|
| 271 |
-
"learning_rate": 9.
|
| 272 |
-
"loss": 0.
|
| 273 |
-
"step":
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch":
|
| 277 |
-
"learning_rate": 9.
|
| 278 |
-
"loss": 0.
|
| 279 |
-
"step":
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch":
|
| 283 |
-
"
|
| 284 |
-
"
|
| 285 |
-
"
|
| 286 |
-
"eval_runtime": 0.9187,
|
| 287 |
-
"eval_samples_per_second": 389.697,
|
| 288 |
-
"eval_steps_per_second": 26.125,
|
| 289 |
-
"step": 18648
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch":
|
| 293 |
-
"learning_rate": 9.
|
| 294 |
-
"loss": 0.
|
| 295 |
-
"step":
|
| 296 |
},
|
| 297 |
{
|
| 298 |
-
"epoch":
|
| 299 |
-
"learning_rate": 9.
|
| 300 |
-
"loss": 0.
|
| 301 |
-
"step":
|
| 302 |
},
|
| 303 |
{
|
| 304 |
-
"epoch":
|
| 305 |
-
"learning_rate":
|
| 306 |
-
"loss": 0.
|
| 307 |
-
"step":
|
| 308 |
},
|
| 309 |
{
|
| 310 |
-
"epoch":
|
| 311 |
-
"learning_rate": 8.
|
| 312 |
-
"loss": 0.
|
| 313 |
-
"step":
|
| 314 |
},
|
| 315 |
{
|
| 316 |
-
"epoch":
|
| 317 |
-
"learning_rate": 8.
|
| 318 |
-
"loss": 0.
|
| 319 |
-
"step":
|
| 320 |
},
|
| 321 |
{
|
| 322 |
-
"epoch":
|
| 323 |
-
"eval_loss": 0.
|
| 324 |
"eval_max_distance": 1,
|
| 325 |
"eval_mean_distance": 0,
|
| 326 |
-
"eval_runtime": 0.
|
| 327 |
-
"eval_samples_per_second":
|
| 328 |
-
"eval_steps_per_second":
|
| 329 |
-
"step":
|
| 330 |
},
|
| 331 |
{
|
| 332 |
-
"epoch":
|
| 333 |
-
"learning_rate": 8.
|
| 334 |
-
"loss": 0.
|
| 335 |
-
"step":
|
| 336 |
-
},
|
| 337 |
-
{
|
| 338 |
-
"epoch": 8.2,
|
| 339 |
-
"learning_rate": 8.832478311644978e-05,
|
| 340 |
-
"loss": 0.0021,
|
| 341 |
-
"step": 21853
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch":
|
| 345 |
-
"learning_rate": 8.
|
| 346 |
-
"loss": 0.
|
| 347 |
-
"step":
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch":
|
| 351 |
-
"learning_rate": 8.
|
| 352 |
-
"loss": 0.
|
| 353 |
-
"step":
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch":
|
| 357 |
-
"learning_rate": 8.
|
| 358 |
-
"loss": 0.
|
| 359 |
-
"step":
|
| 360 |
},
|
| 361 |
{
|
| 362 |
-
"epoch":
|
| 363 |
-
"
|
| 364 |
-
"
|
| 365 |
-
"
|
| 366 |
-
"eval_runtime": 0.9126,
|
| 367 |
-
"eval_samples_per_second": 392.285,
|
| 368 |
-
"eval_steps_per_second": 26.298,
|
| 369 |
-
"step": 23976
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch":
|
| 373 |
-
"learning_rate": 8.
|
| 374 |
-
"loss": 0.
|
| 375 |
-
"step":
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch":
|
| 379 |
-
"learning_rate": 8.
|
| 380 |
-
"loss": 0.
|
| 381 |
-
"step":
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch":
|
| 385 |
-
"learning_rate": 8.
|
| 386 |
-
"loss": 0.
|
| 387 |
-
"step":
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch":
|
| 391 |
-
"learning_rate": 8.
|
| 392 |
-
"loss": 0.
|
| 393 |
-
"step":
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch":
|
| 397 |
-
"learning_rate": 8.
|
| 398 |
-
"loss": 0.
|
| 399 |
-
"step":
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch":
|
| 403 |
-
"eval_loss": 0.
|
| 404 |
"eval_max_distance": 1,
|
| 405 |
"eval_mean_distance": 0,
|
| 406 |
-
"eval_runtime": 0.
|
| 407 |
-
"eval_samples_per_second":
|
| 408 |
-
"eval_steps_per_second":
|
| 409 |
-
"step":
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 10.0,
|
| 413 |
-
"learning_rate": 8.332290623957291e-05,
|
| 414 |
-
"loss": 0.0014,
|
| 415 |
-
"step": 26650
|
| 416 |
},
|
| 417 |
{
|
| 418 |
-
"epoch":
|
| 419 |
-
"learning_rate": 8.
|
| 420 |
-
"loss": 0.
|
| 421 |
-
"step":
|
| 422 |
},
|
| 423 |
{
|
| 424 |
-
"epoch":
|
| 425 |
-
"learning_rate": 8.
|
| 426 |
-
"loss": 0.
|
| 427 |
-
"step":
|
| 428 |
-
},
|
| 429 |
-
{
|
| 430 |
-
"epoch": 10.6,
|
| 431 |
-
"learning_rate": 8.165561394728062e-05,
|
| 432 |
-
"loss": 0.0017,
|
| 433 |
-
"step": 28249
|
| 434 |
-
},
|
| 435 |
-
{
|
| 436 |
-
"epoch": 10.8,
|
| 437 |
-
"learning_rate": 8.109984984984985e-05,
|
| 438 |
-
"loss": 0.0026,
|
| 439 |
-
"step": 28782
|
| 440 |
},
|
| 441 |
{
|
| 442 |
-
"epoch":
|
| 443 |
-
"
|
| 444 |
-
"
|
| 445 |
-
"
|
| 446 |
-
"eval_runtime": 0.9104,
|
| 447 |
-
"eval_samples_per_second": 393.245,
|
| 448 |
-
"eval_steps_per_second": 26.363,
|
| 449 |
-
"step": 29304
|
| 450 |
},
|
| 451 |
{
|
| 452 |
-
"epoch":
|
| 453 |
-
"learning_rate": 8.
|
| 454 |
"loss": 0.0016,
|
| 455 |
-
"step":
|
| 456 |
-
},
|
| 457 |
-
{
|
| 458 |
-
"epoch": 11.2,
|
| 459 |
-
"learning_rate": 7.998832165498832e-05,
|
| 460 |
-
"loss": 0.0006,
|
| 461 |
-
"step": 29848
|
| 462 |
-
},
|
| 463 |
-
{
|
| 464 |
-
"epoch": 11.4,
|
| 465 |
-
"learning_rate": 7.943255755755756e-05,
|
| 466 |
-
"loss": 0.0011,
|
| 467 |
-
"step": 30381
|
| 468 |
-
},
|
| 469 |
-
{
|
| 470 |
-
"epoch": 11.6,
|
| 471 |
-
"learning_rate": 7.88767934601268e-05,
|
| 472 |
-
"loss": 0.0009,
|
| 473 |
-
"step": 30914
|
| 474 |
},
|
| 475 |
{
|
| 476 |
-
"epoch":
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss": 0.
|
| 479 |
-
"step":
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch":
|
| 483 |
-
"
|
| 484 |
-
"eval_max_distance": 1,
|
| 485 |
-
"eval_mean_distance": 0,
|
| 486 |
-
"eval_runtime": 0.9125,
|
| 487 |
-
"eval_samples_per_second": 392.323,
|
| 488 |
-
"eval_steps_per_second": 26.301,
|
| 489 |
-
"step": 31968
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 12.0,
|
| 493 |
-
"learning_rate": 7.776526526526526e-05,
|
| 494 |
"loss": 0.0029,
|
| 495 |
-
"step":
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
-
"learning_rate": 7.
|
| 500 |
-
"loss": 0.
|
| 501 |
-
"step":
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch":
|
| 505 |
-
"learning_rate": 7.
|
| 506 |
-
"loss": 0.
|
| 507 |
-
"step":
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
-
"learning_rate": 7.
|
| 512 |
-
"loss": 0.
|
| 513 |
-
"step":
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch":
|
| 517 |
-
"learning_rate": 7.
|
| 518 |
-
"loss": 0.
|
| 519 |
-
"step":
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"eval_loss": 0.
|
| 524 |
"eval_max_distance": 1,
|
| 525 |
"eval_mean_distance": 0,
|
| 526 |
-
"eval_runtime": 0.
|
| 527 |
-
"eval_samples_per_second":
|
| 528 |
-
"eval_steps_per_second":
|
| 529 |
-
"step":
|
| 530 |
},
|
| 531 |
{
|
| 532 |
-
"epoch":
|
| 533 |
-
"learning_rate": 7.
|
| 534 |
-
"loss": 0.
|
| 535 |
-
"step":
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 13.2,
|
| 539 |
-
"learning_rate": 7.443068068068069e-05,
|
| 540 |
-
"loss": 0.0014,
|
| 541 |
-
"step": 35178
|
| 542 |
},
|
| 543 |
{
|
| 544 |
-
"epoch":
|
| 545 |
-
"learning_rate": 7.
|
| 546 |
-
"loss": 0.
|
| 547 |
-
"step":
|
| 548 |
},
|
| 549 |
{
|
| 550 |
-
"epoch":
|
| 551 |
-
"learning_rate": 7.
|
| 552 |
-
"loss": 0.
|
| 553 |
-
"step":
|
| 554 |
},
|
| 555 |
{
|
| 556 |
-
"epoch":
|
| 557 |
-
"learning_rate": 7.
|
| 558 |
-
"loss": 0.
|
| 559 |
-
"step":
|
| 560 |
},
|
| 561 |
{
|
| 562 |
-
"epoch":
|
| 563 |
-
"
|
| 564 |
-
"
|
| 565 |
-
"
|
| 566 |
-
"eval_runtime": 0.9178,
|
| 567 |
-
"eval_samples_per_second": 390.083,
|
| 568 |
-
"eval_steps_per_second": 26.151,
|
| 569 |
-
"step": 37296
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch":
|
| 573 |
-
"learning_rate": 7.
|
| 574 |
-
"loss": 0.
|
| 575 |
-
"step":
|
| 576 |
},
|
| 577 |
{
|
| 578 |
-
"epoch":
|
| 579 |
-
"learning_rate": 7.
|
| 580 |
-
"loss": 0.
|
| 581 |
-
"step":
|
| 582 |
},
|
| 583 |
{
|
| 584 |
-
"epoch":
|
| 585 |
-
"learning_rate": 7.
|
| 586 |
-
"loss": 0.
|
| 587 |
-
"step":
|
| 588 |
},
|
| 589 |
{
|
| 590 |
-
"epoch":
|
| 591 |
-
"learning_rate": 7.
|
| 592 |
-
"loss": 0.
|
| 593 |
-
"step":
|
| 594 |
},
|
| 595 |
{
|
| 596 |
-
"epoch":
|
| 597 |
-
"learning_rate":
|
| 598 |
-
"loss": 0.
|
| 599 |
-
"step":
|
| 600 |
},
|
| 601 |
{
|
| 602 |
-
"epoch":
|
| 603 |
-
"eval_loss": 0.
|
| 604 |
"eval_max_distance": 1,
|
| 605 |
"eval_mean_distance": 0,
|
| 606 |
-
"eval_runtime": 0.
|
| 607 |
-
"eval_samples_per_second":
|
| 608 |
-
"eval_steps_per_second":
|
| 609 |
-
"step":
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 15.01,
|
| 613 |
-
"learning_rate": 6.94288038038038e-05,
|
| 614 |
-
"loss": 0.001,
|
| 615 |
-
"step": 39975
|
| 616 |
},
|
| 617 |
{
|
| 618 |
-
"epoch":
|
| 619 |
-
"learning_rate":
|
| 620 |
-
"loss": 0.
|
| 621 |
-
"step":
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch":
|
| 625 |
-
"learning_rate":
|
| 626 |
-
"loss": 0.
|
| 627 |
-
"step":
|
| 628 |
},
|
| 629 |
{
|
| 630 |
-
"epoch":
|
| 631 |
-
"learning_rate":
|
| 632 |
-
"loss": 0.
|
| 633 |
-
"step":
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss": 0.
|
| 639 |
-
"step":
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch":
|
| 643 |
-
"
|
| 644 |
-
"
|
| 645 |
-
"
|
| 646 |
-
"eval_runtime": 0.9131,
|
| 647 |
-
"eval_samples_per_second": 392.074,
|
| 648 |
-
"eval_steps_per_second": 26.284,
|
| 649 |
-
"step": 42624
|
| 650 |
},
|
| 651 |
{
|
| 652 |
-
"epoch":
|
| 653 |
-
"learning_rate": 6.
|
| 654 |
-
"loss": 0.
|
| 655 |
-
"step":
|
| 656 |
},
|
| 657 |
{
|
| 658 |
-
"epoch":
|
| 659 |
-
"learning_rate": 6.
|
| 660 |
-
"loss": 0.
|
| 661 |
-
"step":
|
| 662 |
},
|
| 663 |
{
|
| 664 |
-
"epoch":
|
| 665 |
-
"learning_rate": 6.
|
| 666 |
-
"loss": 0.
|
| 667 |
-
"step":
|
| 668 |
},
|
| 669 |
{
|
| 670 |
-
"epoch":
|
| 671 |
-
"learning_rate": 6.
|
| 672 |
-
"loss": 0.
|
| 673 |
-
"step":
|
| 674 |
},
|
| 675 |
{
|
| 676 |
-
"epoch":
|
| 677 |
-
"learning_rate": 6.
|
| 678 |
-
"loss": 0.
|
| 679 |
-
"step":
|
| 680 |
},
|
| 681 |
{
|
| 682 |
-
"epoch":
|
| 683 |
-
"eval_loss": 0.
|
| 684 |
-
"eval_max_distance":
|
| 685 |
"eval_mean_distance": 0,
|
| 686 |
-
"eval_runtime": 0.
|
| 687 |
-
"eval_samples_per_second":
|
| 688 |
-
"eval_steps_per_second":
|
| 689 |
-
"step":
|
| 690 |
},
|
| 691 |
{
|
| 692 |
-
"epoch":
|
| 693 |
-
"learning_rate": 6.
|
| 694 |
-
"loss": 0.
|
| 695 |
-
"step":
|
| 696 |
},
|
| 697 |
{
|
| 698 |
-
"epoch":
|
| 699 |
-
"learning_rate": 6.
|
| 700 |
-
"loss": 0.
|
| 701 |
-
"step":
|
| 702 |
-
},
|
| 703 |
-
{
|
| 704 |
-
"epoch": 17.41,
|
| 705 |
-
"learning_rate": 6.275963463463464e-05,
|
| 706 |
-
"loss": 0.0011,
|
| 707 |
-
"step": 46371
|
| 708 |
},
|
| 709 |
{
|
| 710 |
-
"epoch":
|
| 711 |
-
"learning_rate": 6.
|
| 712 |
-
"loss": 0.
|
| 713 |
-
"step":
|
| 714 |
},
|
| 715 |
{
|
| 716 |
-
"epoch":
|
| 717 |
-
"learning_rate": 6.
|
| 718 |
-
"loss": 0.
|
| 719 |
-
"step":
|
| 720 |
},
|
| 721 |
{
|
| 722 |
-
"epoch":
|
| 723 |
-
"
|
| 724 |
-
"
|
| 725 |
-
"
|
| 726 |
-
"eval_runtime": 0.908,
|
| 727 |
-
"eval_samples_per_second": 394.28,
|
| 728 |
-
"eval_steps_per_second": 26.432,
|
| 729 |
-
"step": 47952
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch":
|
| 733 |
-
"learning_rate": 6.
|
| 734 |
-
"loss": 0.
|
| 735 |
-
"step":
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch":
|
| 739 |
-
"learning_rate": 6.
|
| 740 |
-
"loss": 0.
|
| 741 |
-
"step":
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch":
|
| 745 |
-
"learning_rate":
|
| 746 |
-
"loss": 0.
|
| 747 |
-
"step":
|
| 748 |
},
|
| 749 |
{
|
| 750 |
-
"epoch":
|
| 751 |
-
"learning_rate":
|
| 752 |
-
"loss": 0.
|
| 753 |
-
"step":
|
| 754 |
},
|
| 755 |
{
|
| 756 |
-
"epoch":
|
| 757 |
-
"learning_rate":
|
| 758 |
-
"loss": 0.
|
| 759 |
-
"step":
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch":
|
| 763 |
-
"eval_loss": 0.
|
| 764 |
"eval_max_distance": 1,
|
| 765 |
"eval_mean_distance": 0,
|
| 766 |
-
"eval_runtime": 0.
|
| 767 |
-
"eval_samples_per_second":
|
| 768 |
-
"eval_steps_per_second":
|
| 769 |
-
"step":
|
| 770 |
},
|
| 771 |
{
|
| 772 |
-
"epoch":
|
| 773 |
-
"learning_rate":
|
| 774 |
-
"loss": 0.
|
| 775 |
-
"step":
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 19.21,
|
| 779 |
-
"learning_rate": 5.7757757757757755e-05,
|
| 780 |
-
"loss": 0.0009,
|
| 781 |
-
"step": 51168
|
| 782 |
},
|
| 783 |
{
|
| 784 |
-
"epoch":
|
| 785 |
-
"learning_rate":
|
| 786 |
-
"loss": 0.
|
| 787 |
-
"step":
|
| 788 |
},
|
| 789 |
{
|
| 790 |
-
"epoch":
|
| 791 |
-
"learning_rate": 5.
|
| 792 |
-
"loss": 0.
|
| 793 |
-
"step":
|
| 794 |
},
|
| 795 |
{
|
| 796 |
-
"epoch":
|
| 797 |
-
"learning_rate": 5.
|
| 798 |
-
"loss": 0.
|
| 799 |
-
"step":
|
| 800 |
},
|
| 801 |
{
|
| 802 |
-
"epoch":
|
| 803 |
-
"
|
| 804 |
-
"
|
| 805 |
-
"
|
| 806 |
-
"eval_runtime": 0.9074,
|
| 807 |
-
"eval_samples_per_second": 394.526,
|
| 808 |
-
"eval_steps_per_second": 26.449,
|
| 809 |
-
"step": 53280
|
| 810 |
},
|
| 811 |
{
|
| 812 |
-
"epoch":
|
| 813 |
-
"learning_rate": 5.
|
| 814 |
-
"loss": 0.
|
| 815 |
-
"step":
|
| 816 |
},
|
| 817 |
{
|
| 818 |
-
"epoch":
|
| 819 |
-
"learning_rate": 5.
|
| 820 |
-
"loss": 0.
|
| 821 |
-
"step":
|
| 822 |
},
|
| 823 |
{
|
| 824 |
-
"epoch":
|
| 825 |
-
"learning_rate": 5.
|
| 826 |
-
"loss": 0.
|
| 827 |
-
"step":
|
| 828 |
},
|
| 829 |
{
|
| 830 |
-
"epoch":
|
| 831 |
-
"learning_rate": 5.
|
| 832 |
-
"loss": 0.
|
| 833 |
-
"step":
|
| 834 |
},
|
| 835 |
{
|
| 836 |
-
"epoch":
|
| 837 |
-
"learning_rate": 5.
|
| 838 |
-
"loss": 0.
|
| 839 |
-
"step":
|
| 840 |
},
|
| 841 |
{
|
| 842 |
-
"epoch":
|
| 843 |
-
"eval_loss": 0.
|
| 844 |
"eval_max_distance": 1,
|
| 845 |
"eval_mean_distance": 0,
|
| 846 |
-
"eval_runtime": 0.
|
| 847 |
-
"eval_samples_per_second":
|
| 848 |
-
"eval_steps_per_second":
|
| 849 |
-
"step":
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 21.01,
|
| 853 |
-
"learning_rate": 5.275588088088088e-05,
|
| 854 |
-
"loss": 0.0005,
|
| 855 |
-
"step": 55965
|
| 856 |
},
|
| 857 |
{
|
| 858 |
-
"epoch":
|
| 859 |
-
"learning_rate": 5.
|
| 860 |
-
"loss": 0.
|
| 861 |
-
"step":
|
| 862 |
},
|
| 863 |
{
|
| 864 |
-
"epoch":
|
| 865 |
-
"learning_rate": 5.
|
| 866 |
-
"loss": 0.
|
| 867 |
-
"step":
|
| 868 |
},
|
| 869 |
{
|
| 870 |
-
"epoch":
|
| 871 |
-
"learning_rate": 5.
|
| 872 |
-
"loss": 0.
|
| 873 |
-
"step":
|
| 874 |
},
|
| 875 |
{
|
| 876 |
-
"epoch":
|
| 877 |
-
"learning_rate": 5.
|
| 878 |
-
"loss": 0.
|
| 879 |
-
"step":
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch":
|
| 883 |
-
"
|
| 884 |
-
"
|
| 885 |
-
"
|
| 886 |
-
"eval_runtime": 0.9114,
|
| 887 |
-
"eval_samples_per_second": 392.814,
|
| 888 |
-
"eval_steps_per_second": 26.334,
|
| 889 |
-
"step": 58608
|
| 890 |
},
|
| 891 |
{
|
| 892 |
-
"epoch":
|
| 893 |
-
"learning_rate":
|
| 894 |
-
"loss": 0.
|
| 895 |
-
"step":
|
| 896 |
},
|
| 897 |
{
|
| 898 |
-
"epoch":
|
| 899 |
-
"learning_rate":
|
| 900 |
-
"loss": 0.
|
| 901 |
-
"step":
|
| 902 |
},
|
| 903 |
{
|
| 904 |
-
"epoch":
|
| 905 |
-
"learning_rate":
|
| 906 |
-
"loss": 0.
|
| 907 |
-
"step":
|
| 908 |
},
|
| 909 |
{
|
| 910 |
-
"epoch":
|
| 911 |
-
"learning_rate":
|
| 912 |
-
"loss": 0.
|
| 913 |
-
"step":
|
| 914 |
},
|
| 915 |
{
|
| 916 |
-
"epoch":
|
| 917 |
-
"learning_rate":
|
| 918 |
-
"loss": 0.
|
| 919 |
-
"step":
|
| 920 |
},
|
| 921 |
{
|
| 922 |
-
"epoch":
|
| 923 |
-
"eval_loss": 0.
|
| 924 |
"eval_max_distance": 1,
|
| 925 |
"eval_mean_distance": 0,
|
| 926 |
-
"eval_runtime": 0.
|
| 927 |
-
"eval_samples_per_second":
|
| 928 |
-
"eval_steps_per_second":
|
| 929 |
-
"step":
|
| 930 |
},
|
| 931 |
{
|
| 932 |
-
"epoch":
|
| 933 |
-
"learning_rate": 4.
|
| 934 |
-
"loss": 0.
|
| 935 |
-
"step":
|
| 936 |
},
|
| 937 |
{
|
| 938 |
-
"epoch":
|
| 939 |
-
"learning_rate": 4.
|
| 940 |
-
"loss": 0.
|
| 941 |
-
"step":
|
| 942 |
},
|
| 943 |
{
|
| 944 |
-
"epoch":
|
| 945 |
-
"learning_rate": 4.
|
| 946 |
-
"loss": 0.
|
| 947 |
-
"step":
|
| 948 |
},
|
| 949 |
{
|
| 950 |
-
"epoch":
|
| 951 |
-
"learning_rate": 4.
|
| 952 |
-
"loss": 0.
|
| 953 |
-
"step":
|
| 954 |
},
|
| 955 |
{
|
| 956 |
-
"epoch":
|
| 957 |
-
"learning_rate": 4.
|
| 958 |
-
"loss": 0.
|
| 959 |
-
"step":
|
| 960 |
},
|
| 961 |
{
|
| 962 |
-
"epoch":
|
| 963 |
-
"
|
| 964 |
-
"
|
| 965 |
-
"
|
| 966 |
-
"eval_runtime": 0.9149,
|
| 967 |
-
"eval_samples_per_second": 391.282,
|
| 968 |
-
"eval_steps_per_second": 26.231,
|
| 969 |
-
"step": 63936
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch":
|
| 973 |
-
"learning_rate": 4.
|
| 974 |
-
"loss": 0.
|
| 975 |
-
"step":
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch":
|
| 979 |
-
"learning_rate": 4.
|
| 980 |
-
"loss": 0.
|
| 981 |
-
"step":
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch":
|
| 985 |
-
"learning_rate": 4.
|
| 986 |
-
"loss": 0.
|
| 987 |
-
"step":
|
| 988 |
},
|
| 989 |
{
|
| 990 |
-
"epoch":
|
| 991 |
-
"learning_rate": 4.
|
| 992 |
-
"loss": 0.
|
| 993 |
-
"step":
|
| 994 |
},
|
| 995 |
{
|
| 996 |
-
"epoch":
|
| 997 |
-
"
|
| 998 |
-
"
|
| 999 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
-
"epoch":
|
| 1003 |
-
"
|
| 1004 |
-
"
|
| 1005 |
-
"
|
| 1006 |
-
"eval_runtime": 0.9087,
|
| 1007 |
-
"eval_samples_per_second": 393.969,
|
| 1008 |
-
"eval_steps_per_second": 26.411,
|
| 1009 |
-
"step": 66600
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
-
"epoch":
|
| 1013 |
-
"learning_rate": 4.
|
| 1014 |
-
"loss": 0.
|
| 1015 |
-
"step":
|
| 1016 |
},
|
| 1017 |
{
|
| 1018 |
-
"epoch":
|
| 1019 |
-
"learning_rate": 4.
|
| 1020 |
-
"loss": 0.
|
| 1021 |
-
"step":
|
| 1022 |
},
|
| 1023 |
{
|
| 1024 |
-
"epoch":
|
| 1025 |
-
"learning_rate": 4.
|
| 1026 |
-
"loss": 0.
|
| 1027 |
-
"step":
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
-
"epoch":
|
| 1031 |
-
"learning_rate":
|
| 1032 |
"loss": 0.0001,
|
| 1033 |
-
"step":
|
| 1034 |
},
|
| 1035 |
{
|
| 1036 |
-
"epoch":
|
| 1037 |
-
"learning_rate":
|
| 1038 |
-
"loss": 0.
|
| 1039 |
-
"step":
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
-
"epoch":
|
| 1043 |
-
"
|
| 1044 |
-
"
|
| 1045 |
-
"
|
| 1046 |
-
"eval_runtime": 0.9209,
|
| 1047 |
-
"eval_samples_per_second": 388.755,
|
| 1048 |
-
"eval_steps_per_second": 26.062,
|
| 1049 |
-
"step": 69264
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
-
"epoch":
|
| 1053 |
-
"learning_rate":
|
| 1054 |
-
"loss": 0.
|
| 1055 |
-
"step":
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
-
"epoch":
|
| 1059 |
-
"learning_rate": 3.
|
| 1060 |
-
"loss": 0.
|
| 1061 |
-
"step":
|
| 1062 |
},
|
| 1063 |
{
|
| 1064 |
-
"epoch":
|
| 1065 |
-
"learning_rate": 3.
|
| 1066 |
-
"loss": 0.
|
| 1067 |
-
"step":
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
-
"epoch":
|
| 1071 |
-
"
|
| 1072 |
-
"
|
| 1073 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch":
|
| 1077 |
-
"learning_rate": 3.
|
| 1078 |
"loss": 0.0004,
|
| 1079 |
-
"step":
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
-
"epoch":
|
| 1083 |
-
"
|
| 1084 |
-
"
|
| 1085 |
-
"
|
| 1086 |
-
"eval_runtime": 0.9071,
|
| 1087 |
-
"eval_samples_per_second": 394.647,
|
| 1088 |
-
"eval_steps_per_second": 26.457,
|
| 1089 |
-
"step": 71928
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
-
"epoch":
|
| 1093 |
-
"learning_rate": 3.
|
| 1094 |
-
"loss": 0.
|
| 1095 |
-
"step":
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
-
"epoch":
|
| 1099 |
-
"learning_rate": 3.
|
| 1100 |
-
"loss": 0.
|
| 1101 |
-
"step":
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch":
|
| 1105 |
-
"learning_rate": 3.
|
| 1106 |
-
"loss": 0.
|
| 1107 |
-
"step":
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
-
"epoch":
|
| 1111 |
-
"learning_rate": 3.
|
| 1112 |
"loss": 0.0001,
|
| 1113 |
-
"step":
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
-
"epoch":
|
| 1117 |
-
"learning_rate": 3.
|
| 1118 |
-
"loss": 0.
|
| 1119 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
-
"epoch":
|
| 1123 |
-
"eval_loss": 0.
|
| 1124 |
"eval_max_distance": 1,
|
| 1125 |
"eval_mean_distance": 0,
|
| 1126 |
-
"eval_runtime": 0.
|
| 1127 |
-
"eval_samples_per_second":
|
| 1128 |
-
"eval_steps_per_second":
|
| 1129 |
-
"step":
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
-
"epoch":
|
| 1133 |
-
"learning_rate": 3.
|
| 1134 |
-
"loss": 0.
|
| 1135 |
-
"step":
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
-
"epoch":
|
| 1139 |
-
"learning_rate": 3.
|
| 1140 |
"loss": 0.0008,
|
| 1141 |
-
"step":
|
| 1142 |
},
|
| 1143 |
{
|
| 1144 |
-
"epoch":
|
| 1145 |
-
"learning_rate": 3.
|
| 1146 |
"loss": 0.0005,
|
| 1147 |
-
"step":
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
-
"epoch":
|
| 1151 |
-
"learning_rate": 3.
|
| 1152 |
-
"loss": 0.
|
| 1153 |
-
"step":
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 28.81,
|
| 1157 |
-
"learning_rate": 3.108108108108108e-05,
|
| 1158 |
-
"loss": 0.0002,
|
| 1159 |
-
"step": 76752
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
-
"epoch":
|
| 1163 |
-
"
|
| 1164 |
-
"
|
| 1165 |
-
"
|
| 1166 |
-
"eval_runtime": 0.9144,
|
| 1167 |
-
"eval_samples_per_second": 391.524,
|
| 1168 |
-
"eval_steps_per_second": 26.247,
|
| 1169 |
-
"step": 77256
|
| 1170 |
},
|
| 1171 |
{
|
| 1172 |
-
"epoch":
|
| 1173 |
-
"learning_rate":
|
| 1174 |
-
"loss": 0.
|
| 1175 |
-
"step":
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
-
"epoch":
|
| 1179 |
-
"learning_rate": 2.
|
| 1180 |
-
"loss": 0.
|
| 1181 |
-
"step":
|
| 1182 |
},
|
| 1183 |
{
|
| 1184 |
-
"epoch":
|
| 1185 |
-
"learning_rate": 2.
|
| 1186 |
-
"loss": 0.
|
| 1187 |
-
"step":
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
-
"epoch":
|
| 1191 |
-
"learning_rate": 2.
|
| 1192 |
-
"loss": 0.
|
| 1193 |
-
"step":
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
-
"epoch":
|
| 1197 |
-
"learning_rate": 2.
|
| 1198 |
-
"loss": 0.
|
| 1199 |
-
"step":
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
-
"epoch":
|
| 1203 |
-
"eval_loss": 0.
|
| 1204 |
"eval_max_distance": 1,
|
| 1205 |
"eval_mean_distance": 0,
|
| 1206 |
-
"eval_runtime": 0.
|
| 1207 |
-
"eval_samples_per_second":
|
| 1208 |
-
"eval_steps_per_second":
|
| 1209 |
-
"step":
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
-
"epoch":
|
| 1213 |
-
"learning_rate": 2.
|
| 1214 |
-
"loss": 0.
|
| 1215 |
-
"step":
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
-
"epoch":
|
| 1219 |
-
"learning_rate": 2.
|
| 1220 |
-
"loss": 0.
|
| 1221 |
-
"step":
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
-
"epoch":
|
| 1225 |
-
"learning_rate": 2.
|
| 1226 |
"loss": 0.0001,
|
| 1227 |
-
"step":
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
-
"epoch":
|
| 1231 |
-
"learning_rate": 2.
|
| 1232 |
-
"loss": 0.
|
| 1233 |
-
"step":
|
| 1234 |
},
|
| 1235 |
{
|
| 1236 |
-
"epoch":
|
| 1237 |
-
"learning_rate": 2.
|
| 1238 |
-
"loss": 0.
|
| 1239 |
-
"step":
|
| 1240 |
},
|
| 1241 |
{
|
| 1242 |
-
"epoch":
|
| 1243 |
-
"
|
| 1244 |
-
"
|
| 1245 |
-
"
|
| 1246 |
-
"eval_runtime": 0.9096,
|
| 1247 |
-
"eval_samples_per_second": 393.574,
|
| 1248 |
-
"eval_steps_per_second": 26.385,
|
| 1249 |
-
"step": 82584
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
-
"epoch":
|
| 1253 |
-
"learning_rate": 2.
|
| 1254 |
"loss": 0.0008,
|
| 1255 |
-
"step":
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 31.21,
|
| 1259 |
-
"learning_rate": 2.4411911911911913e-05,
|
| 1260 |
-
"loss": 0.0003,
|
| 1261 |
-
"step": 83148
|
| 1262 |
},
|
| 1263 |
{
|
| 1264 |
-
"epoch":
|
| 1265 |
-
"learning_rate": 2.
|
| 1266 |
-
"loss": 0.
|
| 1267 |
-
"step":
|
| 1268 |
},
|
| 1269 |
{
|
| 1270 |
-
"epoch":
|
| 1271 |
-
"learning_rate": 2.
|
| 1272 |
"loss": 0.0002,
|
| 1273 |
-
"step":
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
-
"epoch":
|
| 1277 |
-
"learning_rate": 2.
|
| 1278 |
-
"loss": 0.
|
| 1279 |
-
"step":
|
| 1280 |
},
|
| 1281 |
{
|
| 1282 |
-
"epoch":
|
| 1283 |
-
"eval_loss": 0.
|
| 1284 |
"eval_max_distance": 1,
|
| 1285 |
"eval_mean_distance": 0,
|
| 1286 |
-
"eval_runtime": 0.
|
| 1287 |
-
"eval_samples_per_second":
|
| 1288 |
-
"eval_steps_per_second":
|
| 1289 |
-
"step":
|
| 1290 |
},
|
| 1291 |
{
|
| 1292 |
-
"epoch":
|
| 1293 |
-
"learning_rate": 2.
|
| 1294 |
-
"loss": 0.
|
| 1295 |
-
"step":
|
| 1296 |
-
},
|
| 1297 |
-
{
|
| 1298 |
-
"epoch": 32.21,
|
| 1299 |
-
"learning_rate": 2.1633091424758092e-05,
|
| 1300 |
-
"loss": 0.001,
|
| 1301 |
-
"step": 85813
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
-
"epoch":
|
| 1305 |
-
"learning_rate": 2.
|
| 1306 |
-
"loss": 0.
|
| 1307 |
-
"step":
|
| 1308 |
},
|
| 1309 |
{
|
| 1310 |
-
"epoch":
|
| 1311 |
-
"learning_rate": 2.
|
| 1312 |
-
"loss": 0.
|
| 1313 |
-
"step":
|
| 1314 |
},
|
| 1315 |
{
|
| 1316 |
-
"epoch":
|
| 1317 |
-
"learning_rate":
|
| 1318 |
-
"loss": 0.
|
| 1319 |
-
"step":
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
-
"epoch":
|
| 1323 |
-
"
|
| 1324 |
-
"
|
| 1325 |
-
"
|
| 1326 |
-
"eval_runtime": 0.9089,
|
| 1327 |
-
"eval_samples_per_second": 393.87,
|
| 1328 |
-
"eval_steps_per_second": 26.405,
|
| 1329 |
-
"step": 87912
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
-
"epoch":
|
| 1333 |
-
"learning_rate": 1.
|
| 1334 |
-
"loss": 0.
|
| 1335 |
-
"step":
|
| 1336 |
},
|
| 1337 |
{
|
| 1338 |
-
"epoch":
|
| 1339 |
-
"learning_rate": 1.
|
| 1340 |
-
"loss": 0.
|
| 1341 |
-
"step":
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
-
"epoch":
|
| 1345 |
-
"learning_rate": 1.
|
| 1346 |
-
"loss": 0.
|
| 1347 |
-
"step":
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
-
"epoch":
|
| 1351 |
-
"learning_rate": 1.
|
| 1352 |
-
"loss": 0.
|
| 1353 |
-
"step":
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
-
"epoch":
|
| 1357 |
-
"learning_rate": 1.
|
| 1358 |
-
"loss": 0.
|
| 1359 |
-
"step":
|
| 1360 |
},
|
| 1361 |
{
|
| 1362 |
-
"epoch":
|
| 1363 |
-
"eval_loss": 0.
|
| 1364 |
"eval_max_distance": 1,
|
| 1365 |
"eval_mean_distance": 0,
|
| 1366 |
-
"eval_runtime": 0.
|
| 1367 |
-
"eval_samples_per_second":
|
| 1368 |
-
"eval_steps_per_second":
|
| 1369 |
-
"step":
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
-
"epoch":
|
| 1373 |
-
"learning_rate": 1.
|
| 1374 |
-
"loss": 0.
|
| 1375 |
-
"step":
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
-
"epoch":
|
| 1379 |
-
"learning_rate": 1.
|
| 1380 |
-
"loss": 0.
|
| 1381 |
-
"step":
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
-
"epoch":
|
| 1385 |
-
"learning_rate": 1.
|
| 1386 |
"loss": 0.0003,
|
| 1387 |
-
"step":
|
| 1388 |
},
|
| 1389 |
{
|
| 1390 |
-
"epoch":
|
| 1391 |
-
"learning_rate": 1.
|
| 1392 |
-
"loss": 0.
|
| 1393 |
-
"step":
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
-
"epoch":
|
| 1397 |
-
"learning_rate": 1.
|
| 1398 |
-
"loss": 0.
|
| 1399 |
-
"step":
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
-
"epoch":
|
| 1403 |
-
"
|
| 1404 |
-
"
|
| 1405 |
-
"
|
| 1406 |
-
"eval_runtime": 0.9087,
|
| 1407 |
-
"eval_samples_per_second": 393.956,
|
| 1408 |
-
"eval_steps_per_second": 26.41,
|
| 1409 |
-
"step": 93240
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
-
"epoch":
|
| 1413 |
-
"learning_rate": 1.
|
| 1414 |
-
"loss": 0.
|
| 1415 |
-
"step":
|
| 1416 |
},
|
| 1417 |
{
|
| 1418 |
-
"epoch":
|
| 1419 |
-
"learning_rate": 1.
|
| 1420 |
-
"loss": 0.
|
| 1421 |
-
"step":
|
| 1422 |
},
|
| 1423 |
{
|
| 1424 |
-
"epoch":
|
| 1425 |
-
"learning_rate": 1.
|
| 1426 |
-
"loss": 0.
|
| 1427 |
-
"step":
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
-
"epoch":
|
| 1431 |
-
"learning_rate": 1.
|
| 1432 |
"loss": 0.0007,
|
| 1433 |
-
"step":
|
| 1434 |
},
|
| 1435 |
{
|
| 1436 |
-
"epoch":
|
| 1437 |
-
"
|
| 1438 |
-
"loss": 0.0001,
|
| 1439 |
-
"step": 95407
|
| 1440 |
-
},
|
| 1441 |
-
{
|
| 1442 |
-
"epoch": 36.0,
|
| 1443 |
-
"eval_loss": 0.05670797452330589,
|
| 1444 |
"eval_max_distance": 1,
|
| 1445 |
"eval_mean_distance": 0,
|
| 1446 |
-
"eval_runtime": 0.
|
| 1447 |
-
"eval_samples_per_second":
|
| 1448 |
-
"eval_steps_per_second":
|
| 1449 |
-
"step":
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
-
"epoch":
|
| 1453 |
-
"learning_rate": 1.
|
| 1454 |
-
"loss": 0.0003,
|
| 1455 |
-
"step": 95940
|
| 1456 |
-
},
|
| 1457 |
-
{
|
| 1458 |
-
"epoch": 36.21,
|
| 1459 |
-
"learning_rate": 1.0517809476142811e-05,
|
| 1460 |
"loss": 0.0004,
|
| 1461 |
-
"step":
|
| 1462 |
},
|
| 1463 |
{
|
| 1464 |
-
"epoch":
|
| 1465 |
-
"learning_rate":
|
| 1466 |
"loss": 0.0002,
|
| 1467 |
-
"step":
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
-
"epoch":
|
| 1471 |
-
"learning_rate": 9.
|
| 1472 |
-
"loss": 0.
|
| 1473 |
-
"step":
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
-
"epoch":
|
| 1477 |
-
"learning_rate": 8.
|
| 1478 |
"loss": 0.0009,
|
| 1479 |
-
"step":
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
-
"epoch":
|
| 1483 |
-
"
|
| 1484 |
-
"
|
| 1485 |
-
"
|
| 1486 |
-
"eval_runtime": 0.9347,
|
| 1487 |
-
"eval_samples_per_second": 383.022,
|
| 1488 |
-
"eval_steps_per_second": 25.677,
|
| 1489 |
-
"step": 98568
|
| 1490 |
},
|
| 1491 |
{
|
| 1492 |
-
"epoch":
|
| 1493 |
-
"learning_rate":
|
| 1494 |
-
"loss": 0.
|
| 1495 |
-
"step":
|
| 1496 |
},
|
| 1497 |
{
|
| 1498 |
-
"epoch":
|
| 1499 |
-
"learning_rate": 7.
|
| 1500 |
-
"loss": 0.
|
| 1501 |
-
"step":
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
-
"epoch":
|
| 1505 |
-
"learning_rate":
|
| 1506 |
-
"loss": 0.
|
| 1507 |
-
"step":
|
| 1508 |
},
|
| 1509 |
{
|
| 1510 |
-
"epoch":
|
| 1511 |
-
"learning_rate": 6.
|
| 1512 |
-
"loss": 0.
|
| 1513 |
-
"step":
|
| 1514 |
},
|
| 1515 |
{
|
| 1516 |
-
"epoch":
|
| 1517 |
-
"learning_rate":
|
| 1518 |
-
"loss": 0.
|
| 1519 |
-
"step":
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
-
"epoch":
|
| 1523 |
-
"eval_loss": 0.
|
| 1524 |
"eval_max_distance": 1,
|
| 1525 |
"eval_mean_distance": 0,
|
| 1526 |
-
"eval_runtime": 0.
|
| 1527 |
-
"eval_samples_per_second":
|
| 1528 |
-
"eval_steps_per_second":
|
| 1529 |
-
"step":
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 38.01,
|
| 1533 |
-
"learning_rate": 5.515932599265933e-06,
|
| 1534 |
-
"loss": 0.0002,
|
| 1535 |
-
"step": 101270
|
| 1536 |
},
|
| 1537 |
{
|
| 1538 |
-
"epoch":
|
| 1539 |
-
"learning_rate":
|
| 1540 |
-
"loss": 0.
|
| 1541 |
-
"step":
|
| 1542 |
},
|
| 1543 |
{
|
| 1544 |
-
"epoch":
|
| 1545 |
-
"learning_rate": 4.
|
| 1546 |
-
"loss": 0.
|
| 1547 |
-
"step":
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
-
"epoch":
|
| 1551 |
-
"learning_rate": 3.
|
| 1552 |
-
"loss": 0.
|
| 1553 |
-
"step":
|
| 1554 |
},
|
| 1555 |
{
|
| 1556 |
-
"epoch":
|
| 1557 |
-
"learning_rate": 3.
|
| 1558 |
-
"loss": 0.
|
| 1559 |
-
"step":
|
| 1560 |
},
|
| 1561 |
{
|
| 1562 |
-
"epoch":
|
| 1563 |
-
"
|
| 1564 |
-
"
|
| 1565 |
-
"
|
| 1566 |
-
"eval_runtime": 0.9146,
|
| 1567 |
-
"eval_samples_per_second": 391.433,
|
| 1568 |
-
"eval_steps_per_second": 26.241,
|
| 1569 |
-
"step": 103896
|
| 1570 |
},
|
| 1571 |
{
|
| 1572 |
-
"epoch":
|
| 1573 |
-
"learning_rate": 2.
|
| 1574 |
-
"loss": 0.
|
| 1575 |
-
"step":
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
-
"epoch":
|
| 1579 |
-
"learning_rate":
|
| 1580 |
-
"loss": 0.
|
| 1581 |
-
"step":
|
| 1582 |
},
|
| 1583 |
{
|
| 1584 |
-
"epoch":
|
| 1585 |
-
"learning_rate": 1.
|
| 1586 |
-
"loss": 0.
|
| 1587 |
-
"step":
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
-
"epoch":
|
| 1591 |
-
"learning_rate":
|
| 1592 |
-
"loss": 0.
|
| 1593 |
-
"step":
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
-
"epoch":
|
| 1597 |
-
"learning_rate":
|
| 1598 |
-
"loss": 0.
|
| 1599 |
-
"step":
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
-
"epoch":
|
| 1603 |
-
"eval_loss": 0.
|
| 1604 |
"eval_max_distance": 1,
|
| 1605 |
"eval_mean_distance": 0,
|
| 1606 |
-
"eval_runtime": 0.
|
| 1607 |
-
"eval_samples_per_second":
|
| 1608 |
-
"eval_steps_per_second":
|
| 1609 |
-
"step":
|
| 1610 |
-
},
|
| 1611 |
-
{
|
| 1612 |
-
"epoch":
|
| 1613 |
-
"step":
|
| 1614 |
-
"total_flos":
|
| 1615 |
-
"train_loss": 0.
|
| 1616 |
-
"train_runtime":
|
| 1617 |
-
"train_samples_per_second":
|
| 1618 |
-
"train_steps_per_second":
|
| 1619 |
}
|
| 1620 |
],
|
| 1621 |
-
"logging_steps":
|
| 1622 |
-
"max_steps":
|
| 1623 |
-
"num_train_epochs":
|
| 1624 |
-
"save_steps":
|
| 1625 |
-
"total_flos":
|
| 1626 |
"trial_name": null,
|
| 1627 |
"trial_params": null
|
| 1628 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 21000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 4.761904761904762e-08,
|
| 14 |
+
"loss": 30.681,
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.1,
|
| 19 |
+
"learning_rate": 5e-06,
|
| 20 |
+
"loss": 29.2188,
|
| 21 |
+
"step": 105
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.2,
|
| 25 |
+
"learning_rate": 1e-05,
|
| 26 |
+
"loss": 27.9484,
|
| 27 |
+
"step": 210
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.3,
|
| 31 |
+
"learning_rate": 1.5e-05,
|
| 32 |
+
"loss": 25.4971,
|
| 33 |
+
"step": 315
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.4,
|
| 37 |
+
"learning_rate": 2e-05,
|
| 38 |
+
"loss": 21.1988,
|
| 39 |
+
"step": 420
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.5,
|
| 43 |
+
"learning_rate": 2.5e-05,
|
| 44 |
+
"loss": 15.2033,
|
| 45 |
+
"step": 525
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.6,
|
| 49 |
+
"learning_rate": 3e-05,
|
| 50 |
+
"loss": 7.4762,
|
| 51 |
+
"step": 630
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.7,
|
| 55 |
+
"learning_rate": 3.5e-05,
|
| 56 |
+
"loss": 1.8812,
|
| 57 |
+
"step": 735
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.8,
|
| 61 |
+
"learning_rate": 4e-05,
|
| 62 |
+
"loss": 0.5258,
|
| 63 |
+
"step": 840
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.9,
|
| 67 |
+
"learning_rate": 4.5e-05,
|
| 68 |
+
"loss": 0.347,
|
| 69 |
+
"step": 945
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 1.0,
|
| 73 |
+
"learning_rate": 5e-05,
|
| 74 |
+
"loss": 0.2779,
|
| 75 |
+
"step": 1050
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 1.0,
|
| 79 |
+
"eval_loss": 0.33707377314567566,
|
| 80 |
"eval_max_distance": 1,
|
| 81 |
"eval_mean_distance": 0,
|
| 82 |
+
"eval_runtime": 0.804,
|
| 83 |
+
"eval_samples_per_second": 436.567,
|
| 84 |
+
"eval_steps_per_second": 13.682,
|
| 85 |
+
"step": 1050
|
| 86 |
},
|
| 87 |
{
|
| 88 |
+
"epoch": 1.1,
|
| 89 |
+
"learning_rate": 5.500000000000001e-05,
|
| 90 |
+
"loss": 0.2162,
|
| 91 |
+
"step": 1155
|
| 92 |
},
|
| 93 |
{
|
| 94 |
+
"epoch": 1.2,
|
| 95 |
+
"learning_rate": 6e-05,
|
| 96 |
+
"loss": 0.1397,
|
| 97 |
+
"step": 1260
|
| 98 |
},
|
| 99 |
{
|
| 100 |
+
"epoch": 1.3,
|
| 101 |
+
"learning_rate": 6.500000000000001e-05,
|
| 102 |
+
"loss": 0.1035,
|
| 103 |
+
"step": 1365
|
| 104 |
},
|
| 105 |
{
|
| 106 |
+
"epoch": 1.4,
|
| 107 |
+
"learning_rate": 7e-05,
|
| 108 |
+
"loss": 0.0765,
|
| 109 |
+
"step": 1470
|
| 110 |
},
|
| 111 |
{
|
| 112 |
+
"epoch": 1.5,
|
| 113 |
+
"learning_rate": 7.500000000000001e-05,
|
| 114 |
+
"loss": 0.035,
|
| 115 |
+
"step": 1575
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 1.6,
|
| 119 |
+
"learning_rate": 8e-05,
|
| 120 |
+
"loss": 0.0481,
|
| 121 |
+
"step": 1680
|
|
|
|
|
|
|
|
|
|
|
|
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 1.7,
|
| 125 |
+
"learning_rate": 8.5e-05,
|
| 126 |
+
"loss": 0.0285,
|
| 127 |
+
"step": 1785
|
| 128 |
},
|
| 129 |
{
|
| 130 |
+
"epoch": 1.8,
|
| 131 |
+
"learning_rate": 9e-05,
|
| 132 |
+
"loss": 0.0305,
|
| 133 |
+
"step": 1890
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 134 |
},
|
| 135 |
{
|
| 136 |
+
"epoch": 1.9,
|
| 137 |
+
"learning_rate": 9.5e-05,
|
| 138 |
+
"loss": 0.034,
|
| 139 |
+
"step": 1995
|
| 140 |
},
|
| 141 |
{
|
| 142 |
+
"epoch": 2.0,
|
| 143 |
+
"learning_rate": 0.0001,
|
| 144 |
+
"loss": 0.0346,
|
| 145 |
+
"step": 2100
|
| 146 |
},
|
| 147 |
{
|
| 148 |
+
"epoch": 2.0,
|
| 149 |
+
"eval_loss": 0.01727573201060295,
|
| 150 |
"eval_max_distance": 1,
|
| 151 |
"eval_mean_distance": 0,
|
| 152 |
+
"eval_runtime": 0.9791,
|
| 153 |
+
"eval_samples_per_second": 358.485,
|
| 154 |
+
"eval_steps_per_second": 11.235,
|
| 155 |
+
"step": 2100
|
| 156 |
},
|
| 157 |
{
|
| 158 |
+
"epoch": 2.1,
|
| 159 |
+
"learning_rate": 9.944444444444446e-05,
|
| 160 |
+
"loss": 0.0147,
|
| 161 |
+
"step": 2205
|
| 162 |
},
|
| 163 |
{
|
| 164 |
+
"epoch": 2.2,
|
| 165 |
+
"learning_rate": 9.888888888888889e-05,
|
| 166 |
+
"loss": 0.0125,
|
| 167 |
+
"step": 2310
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 168 |
},
|
| 169 |
{
|
| 170 |
+
"epoch": 2.3,
|
| 171 |
+
"learning_rate": 9.833333333333333e-05,
|
| 172 |
+
"loss": 0.0139,
|
| 173 |
+
"step": 2415
|
| 174 |
},
|
| 175 |
{
|
| 176 |
+
"epoch": 2.4,
|
| 177 |
+
"learning_rate": 9.777777777777778e-05,
|
| 178 |
+
"loss": 0.0129,
|
| 179 |
+
"step": 2520
|
| 180 |
},
|
| 181 |
{
|
| 182 |
+
"epoch": 2.5,
|
| 183 |
+
"learning_rate": 9.722222222222223e-05,
|
| 184 |
+
"loss": 0.0086,
|
| 185 |
+
"step": 2625
|
|
|
|
|
|
|
|
|
|
|
|
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 2.6,
|
| 189 |
+
"learning_rate": 9.666666666666667e-05,
|
| 190 |
+
"loss": 0.0118,
|
| 191 |
+
"step": 2730
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 2.7,
|
| 195 |
+
"learning_rate": 9.611111111111112e-05,
|
| 196 |
+
"loss": 0.0198,
|
| 197 |
+
"step": 2835
|
| 198 |
},
|
| 199 |
{
|
| 200 |
+
"epoch": 2.8,
|
| 201 |
+
"learning_rate": 9.555555555555557e-05,
|
| 202 |
+
"loss": 0.0082,
|
| 203 |
+
"step": 2940
|
| 204 |
},
|
| 205 |
{
|
| 206 |
+
"epoch": 2.9,
|
| 207 |
+
"learning_rate": 9.5e-05,
|
| 208 |
+
"loss": 0.0074,
|
| 209 |
+
"step": 3045
|
| 210 |
},
|
| 211 |
{
|
| 212 |
+
"epoch": 3.0,
|
| 213 |
+
"learning_rate": 9.444444444444444e-05,
|
| 214 |
+
"loss": 0.0072,
|
| 215 |
+
"step": 3150
|
| 216 |
},
|
| 217 |
{
|
| 218 |
+
"epoch": 3.0,
|
| 219 |
+
"eval_loss": 0.007698288187384605,
|
| 220 |
"eval_max_distance": 1,
|
| 221 |
"eval_mean_distance": 0,
|
| 222 |
+
"eval_runtime": 0.773,
|
| 223 |
+
"eval_samples_per_second": 454.074,
|
| 224 |
+
"eval_steps_per_second": 14.23,
|
| 225 |
+
"step": 3150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 3.1,
|
| 229 |
+
"learning_rate": 9.388888888888889e-05,
|
| 230 |
+
"loss": 0.0146,
|
| 231 |
+
"step": 3255
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 3.2,
|
| 235 |
+
"learning_rate": 9.333333333333334e-05,
|
| 236 |
+
"loss": 0.0072,
|
| 237 |
+
"step": 3360
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 3.3,
|
| 241 |
+
"learning_rate": 9.277777777777778e-05,
|
| 242 |
+
"loss": 0.0065,
|
| 243 |
+
"step": 3465
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 3.4,
|
| 247 |
+
"learning_rate": 9.222222222222223e-05,
|
| 248 |
+
"loss": 0.0092,
|
| 249 |
+
"step": 3570
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 3.5,
|
| 253 |
+
"learning_rate": 9.166666666666667e-05,
|
| 254 |
+
"loss": 0.0065,
|
| 255 |
+
"step": 3675
|
|
|
|
|
|
|
|
|
|
|
|
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 3.6,
|
| 259 |
+
"learning_rate": 9.111111111111112e-05,
|
| 260 |
+
"loss": 0.0083,
|
| 261 |
+
"step": 3780
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 3.7,
|
| 265 |
+
"learning_rate": 9.055555555555556e-05,
|
| 266 |
+
"loss": 0.009,
|
| 267 |
+
"step": 3885
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 3.8,
|
| 271 |
+
"learning_rate": 9e-05,
|
| 272 |
+
"loss": 0.0027,
|
| 273 |
+
"step": 3990
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 3.9,
|
| 277 |
+
"learning_rate": 8.944444444444446e-05,
|
| 278 |
+
"loss": 0.0073,
|
| 279 |
+
"step": 4095
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 4.0,
|
| 283 |
+
"learning_rate": 8.888888888888889e-05,
|
| 284 |
+
"loss": 0.0057,
|
| 285 |
+
"step": 4200
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 4.0,
|
| 289 |
+
"eval_loss": 0.009315615519881248,
|
| 290 |
"eval_max_distance": 1,
|
| 291 |
"eval_mean_distance": 0,
|
| 292 |
+
"eval_runtime": 0.7737,
|
| 293 |
+
"eval_samples_per_second": 453.665,
|
| 294 |
+
"eval_steps_per_second": 14.217,
|
| 295 |
+
"step": 4200
|
| 296 |
},
|
| 297 |
{
|
| 298 |
+
"epoch": 4.1,
|
| 299 |
+
"learning_rate": 8.833333333333333e-05,
|
| 300 |
+
"loss": 0.0072,
|
| 301 |
+
"step": 4305
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 302 |
},
|
| 303 |
{
|
| 304 |
+
"epoch": 4.2,
|
| 305 |
+
"learning_rate": 8.777777777777778e-05,
|
| 306 |
+
"loss": 0.0072,
|
| 307 |
+
"step": 4410
|
| 308 |
},
|
| 309 |
{
|
| 310 |
+
"epoch": 4.3,
|
| 311 |
+
"learning_rate": 8.722222222222223e-05,
|
| 312 |
+
"loss": 0.0109,
|
| 313 |
+
"step": 4515
|
| 314 |
},
|
| 315 |
{
|
| 316 |
+
"epoch": 4.4,
|
| 317 |
+
"learning_rate": 8.666666666666667e-05,
|
| 318 |
+
"loss": 0.01,
|
| 319 |
+
"step": 4620
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 4.5,
|
| 323 |
+
"learning_rate": 8.611111111111112e-05,
|
| 324 |
+
"loss": 0.0013,
|
| 325 |
+
"step": 4725
|
|
|
|
|
|
|
|
|
|
|
|
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 4.6,
|
| 329 |
+
"learning_rate": 8.555555555555556e-05,
|
| 330 |
+
"loss": 0.017,
|
| 331 |
+
"step": 4830
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 4.7,
|
| 335 |
+
"learning_rate": 8.5e-05,
|
| 336 |
+
"loss": 0.0061,
|
| 337 |
+
"step": 4935
|
| 338 |
},
|
| 339 |
{
|
| 340 |
+
"epoch": 4.8,
|
| 341 |
+
"learning_rate": 8.444444444444444e-05,
|
| 342 |
+
"loss": 0.0078,
|
| 343 |
+
"step": 5040
|
| 344 |
},
|
| 345 |
{
|
| 346 |
+
"epoch": 4.9,
|
| 347 |
+
"learning_rate": 8.38888888888889e-05,
|
| 348 |
+
"loss": 0.0028,
|
| 349 |
+
"step": 5145
|
| 350 |
},
|
| 351 |
{
|
| 352 |
+
"epoch": 5.0,
|
| 353 |
+
"learning_rate": 8.333333333333334e-05,
|
| 354 |
+
"loss": 0.005,
|
| 355 |
+
"step": 5250
|
| 356 |
},
|
| 357 |
{
|
| 358 |
+
"epoch": 5.0,
|
| 359 |
+
"eval_loss": 0.005270855501294136,
|
| 360 |
"eval_max_distance": 1,
|
| 361 |
"eval_mean_distance": 0,
|
| 362 |
+
"eval_runtime": 0.584,
|
| 363 |
+
"eval_samples_per_second": 601.029,
|
| 364 |
+
"eval_steps_per_second": 18.836,
|
| 365 |
+
"step": 5250
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 366 |
},
|
| 367 |
{
|
| 368 |
+
"epoch": 5.1,
|
| 369 |
+
"learning_rate": 8.277777777777778e-05,
|
| 370 |
+
"loss": 0.006,
|
| 371 |
+
"step": 5355
|
| 372 |
},
|
| 373 |
{
|
| 374 |
+
"epoch": 5.2,
|
| 375 |
+
"learning_rate": 8.222222222222222e-05,
|
| 376 |
+
"loss": 0.0068,
|
| 377 |
+
"step": 5460
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 378 |
},
|
| 379 |
{
|
| 380 |
+
"epoch": 5.3,
|
| 381 |
+
"learning_rate": 8.166666666666667e-05,
|
| 382 |
+
"loss": 0.0029,
|
| 383 |
+
"step": 5565
|
|
|
|
|
|
|
|
|
|
|
|
|
| 384 |
},
|
| 385 |
{
|
| 386 |
+
"epoch": 5.4,
|
| 387 |
+
"learning_rate": 8.111111111111112e-05,
|
| 388 |
"loss": 0.0016,
|
| 389 |
+
"step": 5670
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 390 |
},
|
| 391 |
{
|
| 392 |
+
"epoch": 5.5,
|
| 393 |
+
"learning_rate": 8.055555555555556e-05,
|
| 394 |
+
"loss": 0.0055,
|
| 395 |
+
"step": 5775
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 5.6,
|
| 399 |
+
"learning_rate": 8e-05,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 400 |
"loss": 0.0029,
|
| 401 |
+
"step": 5880
|
| 402 |
},
|
| 403 |
{
|
| 404 |
+
"epoch": 5.7,
|
| 405 |
+
"learning_rate": 7.944444444444444e-05,
|
| 406 |
+
"loss": 0.006,
|
| 407 |
+
"step": 5985
|
| 408 |
},
|
| 409 |
{
|
| 410 |
+
"epoch": 5.8,
|
| 411 |
+
"learning_rate": 7.88888888888889e-05,
|
| 412 |
+
"loss": 0.0028,
|
| 413 |
+
"step": 6090
|
| 414 |
},
|
| 415 |
{
|
| 416 |
+
"epoch": 5.9,
|
| 417 |
+
"learning_rate": 7.833333333333333e-05,
|
| 418 |
+
"loss": 0.0067,
|
| 419 |
+
"step": 6195
|
| 420 |
},
|
| 421 |
{
|
| 422 |
+
"epoch": 6.0,
|
| 423 |
+
"learning_rate": 7.777777777777778e-05,
|
| 424 |
+
"loss": 0.002,
|
| 425 |
+
"step": 6300
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 6.0,
|
| 429 |
+
"eval_loss": 0.005636307876557112,
|
| 430 |
"eval_max_distance": 1,
|
| 431 |
"eval_mean_distance": 0,
|
| 432 |
+
"eval_runtime": 0.5895,
|
| 433 |
+
"eval_samples_per_second": 595.407,
|
| 434 |
+
"eval_steps_per_second": 18.659,
|
| 435 |
+
"step": 6300
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 6.1,
|
| 439 |
+
"learning_rate": 7.722222222222223e-05,
|
| 440 |
+
"loss": 0.0003,
|
| 441 |
+
"step": 6405
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 6.2,
|
| 445 |
+
"learning_rate": 7.666666666666667e-05,
|
| 446 |
+
"loss": 0.002,
|
| 447 |
+
"step": 6510
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 6.3,
|
| 451 |
+
"learning_rate": 7.61111111111111e-05,
|
| 452 |
+
"loss": 0.0011,
|
| 453 |
+
"step": 6615
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 6.4,
|
| 457 |
+
"learning_rate": 7.555555555555556e-05,
|
| 458 |
+
"loss": 0.0028,
|
| 459 |
+
"step": 6720
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 6.5,
|
| 463 |
+
"learning_rate": 7.500000000000001e-05,
|
| 464 |
+
"loss": 0.0062,
|
| 465 |
+
"step": 6825
|
|
|
|
|
|
|
|
|
|
|
|
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 6.6,
|
| 469 |
+
"learning_rate": 7.444444444444444e-05,
|
| 470 |
+
"loss": 0.0006,
|
| 471 |
+
"step": 6930
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 6.7,
|
| 475 |
+
"learning_rate": 7.38888888888889e-05,
|
| 476 |
+
"loss": 0.003,
|
| 477 |
+
"step": 7035
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 6.8,
|
| 481 |
+
"learning_rate": 7.333333333333333e-05,
|
| 482 |
+
"loss": 0.0048,
|
| 483 |
+
"step": 7140
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 6.9,
|
| 487 |
+
"learning_rate": 7.277777777777778e-05,
|
| 488 |
+
"loss": 0.0038,
|
| 489 |
+
"step": 7245
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 7.0,
|
| 493 |
+
"learning_rate": 7.222222222222222e-05,
|
| 494 |
+
"loss": 0.002,
|
| 495 |
+
"step": 7350
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 7.0,
|
| 499 |
+
"eval_loss": 0.009145626798272133,
|
| 500 |
"eval_max_distance": 1,
|
| 501 |
"eval_mean_distance": 0,
|
| 502 |
+
"eval_runtime": 0.605,
|
| 503 |
+
"eval_samples_per_second": 580.167,
|
| 504 |
+
"eval_steps_per_second": 18.182,
|
| 505 |
+
"step": 7350
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 506 |
},
|
| 507 |
{
|
| 508 |
+
"epoch": 7.1,
|
| 509 |
+
"learning_rate": 7.166666666666667e-05,
|
| 510 |
+
"loss": 0.005,
|
| 511 |
+
"step": 7455
|
| 512 |
},
|
| 513 |
{
|
| 514 |
+
"epoch": 7.2,
|
| 515 |
+
"learning_rate": 7.111111111111112e-05,
|
| 516 |
+
"loss": 0.0018,
|
| 517 |
+
"step": 7560
|
| 518 |
},
|
| 519 |
{
|
| 520 |
+
"epoch": 7.3,
|
| 521 |
+
"learning_rate": 7.055555555555556e-05,
|
| 522 |
+
"loss": 0.0031,
|
| 523 |
+
"step": 7665
|
| 524 |
},
|
| 525 |
{
|
| 526 |
+
"epoch": 7.4,
|
| 527 |
+
"learning_rate": 7e-05,
|
| 528 |
+
"loss": 0.0027,
|
| 529 |
+
"step": 7770
|
| 530 |
},
|
| 531 |
{
|
| 532 |
+
"epoch": 7.5,
|
| 533 |
+
"learning_rate": 6.944444444444444e-05,
|
| 534 |
+
"loss": 0.0011,
|
| 535 |
+
"step": 7875
|
|
|
|
|
|
|
|
|
|
|
|
|
| 536 |
},
|
| 537 |
{
|
| 538 |
+
"epoch": 7.6,
|
| 539 |
+
"learning_rate": 6.88888888888889e-05,
|
| 540 |
+
"loss": 0.0032,
|
| 541 |
+
"step": 7980
|
| 542 |
},
|
| 543 |
{
|
| 544 |
+
"epoch": 7.7,
|
| 545 |
+
"learning_rate": 6.833333333333333e-05,
|
| 546 |
+
"loss": 0.0018,
|
| 547 |
+
"step": 8085
|
| 548 |
},
|
| 549 |
{
|
| 550 |
+
"epoch": 7.8,
|
| 551 |
+
"learning_rate": 6.777777777777778e-05,
|
| 552 |
+
"loss": 0.0032,
|
| 553 |
+
"step": 8190
|
| 554 |
},
|
| 555 |
{
|
| 556 |
+
"epoch": 7.9,
|
| 557 |
+
"learning_rate": 6.722222222222223e-05,
|
| 558 |
+
"loss": 0.0045,
|
| 559 |
+
"step": 8295
|
| 560 |
},
|
| 561 |
{
|
| 562 |
+
"epoch": 8.0,
|
| 563 |
+
"learning_rate": 6.666666666666667e-05,
|
| 564 |
+
"loss": 0.0018,
|
| 565 |
+
"step": 8400
|
| 566 |
},
|
| 567 |
{
|
| 568 |
+
"epoch": 8.0,
|
| 569 |
+
"eval_loss": 0.0010060666827484965,
|
| 570 |
+
"eval_max_distance": 0,
|
| 571 |
"eval_mean_distance": 0,
|
| 572 |
+
"eval_runtime": 0.618,
|
| 573 |
+
"eval_samples_per_second": 567.961,
|
| 574 |
+
"eval_steps_per_second": 17.799,
|
| 575 |
+
"step": 8400
|
| 576 |
},
|
| 577 |
{
|
| 578 |
+
"epoch": 8.1,
|
| 579 |
+
"learning_rate": 6.611111111111111e-05,
|
| 580 |
+
"loss": 0.003,
|
| 581 |
+
"step": 8505
|
| 582 |
},
|
| 583 |
{
|
| 584 |
+
"epoch": 8.2,
|
| 585 |
+
"learning_rate": 6.555555555555556e-05,
|
| 586 |
+
"loss": 0.0023,
|
| 587 |
+
"step": 8610
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 588 |
},
|
| 589 |
{
|
| 590 |
+
"epoch": 8.3,
|
| 591 |
+
"learning_rate": 6.500000000000001e-05,
|
| 592 |
+
"loss": 0.0028,
|
| 593 |
+
"step": 8715
|
| 594 |
},
|
| 595 |
{
|
| 596 |
+
"epoch": 8.4,
|
| 597 |
+
"learning_rate": 6.444444444444446e-05,
|
| 598 |
+
"loss": 0.002,
|
| 599 |
+
"step": 8820
|
| 600 |
},
|
| 601 |
{
|
| 602 |
+
"epoch": 8.5,
|
| 603 |
+
"learning_rate": 6.388888888888888e-05,
|
| 604 |
+
"loss": 0.0012,
|
| 605 |
+
"step": 8925
|
|
|
|
|
|
|
|
|
|
|
|
|
| 606 |
},
|
| 607 |
{
|
| 608 |
+
"epoch": 8.6,
|
| 609 |
+
"learning_rate": 6.333333333333333e-05,
|
| 610 |
+
"loss": 0.0013,
|
| 611 |
+
"step": 9030
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 8.7,
|
| 615 |
+
"learning_rate": 6.277777777777778e-05,
|
| 616 |
+
"loss": 0.0011,
|
| 617 |
+
"step": 9135
|
| 618 |
},
|
| 619 |
{
|
| 620 |
+
"epoch": 8.8,
|
| 621 |
+
"learning_rate": 6.222222222222222e-05,
|
| 622 |
+
"loss": 0.0008,
|
| 623 |
+
"step": 9240
|
| 624 |
},
|
| 625 |
{
|
| 626 |
+
"epoch": 8.9,
|
| 627 |
+
"learning_rate": 6.166666666666667e-05,
|
| 628 |
+
"loss": 0.0029,
|
| 629 |
+
"step": 9345
|
| 630 |
},
|
| 631 |
{
|
| 632 |
+
"epoch": 9.0,
|
| 633 |
+
"learning_rate": 6.111111111111112e-05,
|
| 634 |
+
"loss": 0.0044,
|
| 635 |
+
"step": 9450
|
| 636 |
},
|
| 637 |
{
|
| 638 |
+
"epoch": 9.0,
|
| 639 |
+
"eval_loss": 0.004300011787563562,
|
| 640 |
"eval_max_distance": 1,
|
| 641 |
"eval_mean_distance": 0,
|
| 642 |
+
"eval_runtime": 0.7083,
|
| 643 |
+
"eval_samples_per_second": 495.526,
|
| 644 |
+
"eval_steps_per_second": 15.529,
|
| 645 |
+
"step": 9450
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 9.1,
|
| 649 |
+
"learning_rate": 6.055555555555555e-05,
|
| 650 |
+
"loss": 0.0012,
|
| 651 |
+
"step": 9555
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 9.2,
|
| 655 |
+
"learning_rate": 6e-05,
|
| 656 |
+
"loss": 0.002,
|
| 657 |
+
"step": 9660
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 9.3,
|
| 661 |
+
"learning_rate": 5.9444444444444445e-05,
|
| 662 |
+
"loss": 0.0017,
|
| 663 |
+
"step": 9765
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 9.4,
|
| 667 |
+
"learning_rate": 5.8888888888888896e-05,
|
| 668 |
+
"loss": 0.0002,
|
| 669 |
+
"step": 9870
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 9.5,
|
| 673 |
+
"learning_rate": 5.833333333333334e-05,
|
| 674 |
+
"loss": 0.0015,
|
| 675 |
+
"step": 9975
|
|
|
|
|
|
|
|
|
|
|
|
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 9.6,
|
| 679 |
+
"learning_rate": 5.7777777777777776e-05,
|
| 680 |
+
"loss": 0.003,
|
| 681 |
+
"step": 10080
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 9.7,
|
| 685 |
+
"learning_rate": 5.722222222222222e-05,
|
| 686 |
+
"loss": 0.0011,
|
| 687 |
+
"step": 10185
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 9.8,
|
| 691 |
+
"learning_rate": 5.666666666666667e-05,
|
| 692 |
+
"loss": 0.0018,
|
| 693 |
+
"step": 10290
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 9.9,
|
| 697 |
+
"learning_rate": 5.6111111111111114e-05,
|
| 698 |
+
"loss": 0.0003,
|
| 699 |
+
"step": 10395
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 10.0,
|
| 703 |
+
"learning_rate": 5.555555555555556e-05,
|
| 704 |
+
"loss": 0.0024,
|
| 705 |
+
"step": 10500
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 10.0,
|
| 709 |
+
"eval_loss": 0.004753963556140661,
|
| 710 |
"eval_max_distance": 1,
|
| 711 |
"eval_mean_distance": 0,
|
| 712 |
+
"eval_runtime": 0.593,
|
| 713 |
+
"eval_samples_per_second": 591.906,
|
| 714 |
+
"eval_steps_per_second": 18.55,
|
| 715 |
+
"step": 10500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 716 |
},
|
| 717 |
{
|
| 718 |
+
"epoch": 10.1,
|
| 719 |
+
"learning_rate": 5.500000000000001e-05,
|
| 720 |
+
"loss": 0.0004,
|
| 721 |
+
"step": 10605
|
| 722 |
},
|
| 723 |
{
|
| 724 |
+
"epoch": 10.2,
|
| 725 |
+
"learning_rate": 5.4444444444444446e-05,
|
| 726 |
+
"loss": 0.0022,
|
| 727 |
+
"step": 10710
|
| 728 |
},
|
| 729 |
{
|
| 730 |
+
"epoch": 10.3,
|
| 731 |
+
"learning_rate": 5.388888888888889e-05,
|
| 732 |
+
"loss": 0.0016,
|
| 733 |
+
"step": 10815
|
| 734 |
},
|
| 735 |
{
|
| 736 |
+
"epoch": 10.4,
|
| 737 |
+
"learning_rate": 5.333333333333333e-05,
|
| 738 |
+
"loss": 0.0008,
|
| 739 |
+
"step": 10920
|
| 740 |
},
|
| 741 |
{
|
| 742 |
+
"epoch": 10.5,
|
| 743 |
+
"learning_rate": 5.2777777777777784e-05,
|
| 744 |
+
"loss": 0.0002,
|
| 745 |
+
"step": 11025
|
|
|
|
|
|
|
|
|
|
|
|
|
| 746 |
},
|
| 747 |
{
|
| 748 |
+
"epoch": 10.6,
|
| 749 |
+
"learning_rate": 5.222222222222223e-05,
|
| 750 |
+
"loss": 0.0023,
|
| 751 |
+
"step": 11130
|
| 752 |
},
|
| 753 |
{
|
| 754 |
+
"epoch": 10.7,
|
| 755 |
+
"learning_rate": 5.166666666666667e-05,
|
| 756 |
+
"loss": 0.0009,
|
| 757 |
+
"step": 11235
|
| 758 |
},
|
| 759 |
{
|
| 760 |
+
"epoch": 10.8,
|
| 761 |
+
"learning_rate": 5.111111111111111e-05,
|
| 762 |
+
"loss": 0.0036,
|
| 763 |
+
"step": 11340
|
| 764 |
},
|
| 765 |
{
|
| 766 |
+
"epoch": 10.9,
|
| 767 |
+
"learning_rate": 5.055555555555556e-05,
|
| 768 |
+
"loss": 0.0049,
|
| 769 |
+
"step": 11445
|
| 770 |
},
|
| 771 |
{
|
| 772 |
+
"epoch": 11.0,
|
| 773 |
+
"learning_rate": 5e-05,
|
| 774 |
+
"loss": 0.0032,
|
| 775 |
+
"step": 11550
|
| 776 |
},
|
| 777 |
{
|
| 778 |
+
"epoch": 11.0,
|
| 779 |
+
"eval_loss": 0.0022743879817426205,
|
| 780 |
"eval_max_distance": 1,
|
| 781 |
"eval_mean_distance": 0,
|
| 782 |
+
"eval_runtime": 0.656,
|
| 783 |
+
"eval_samples_per_second": 535.061,
|
| 784 |
+
"eval_steps_per_second": 16.768,
|
| 785 |
+
"step": 11550
|
| 786 |
},
|
| 787 |
{
|
| 788 |
+
"epoch": 11.1,
|
| 789 |
+
"learning_rate": 4.9444444444444446e-05,
|
| 790 |
+
"loss": 0.0036,
|
| 791 |
+
"step": 11655
|
| 792 |
},
|
| 793 |
{
|
| 794 |
+
"epoch": 11.2,
|
| 795 |
+
"learning_rate": 4.888888888888889e-05,
|
| 796 |
+
"loss": 0.0014,
|
| 797 |
+
"step": 11760
|
| 798 |
},
|
| 799 |
{
|
| 800 |
+
"epoch": 11.3,
|
| 801 |
+
"learning_rate": 4.8333333333333334e-05,
|
| 802 |
+
"loss": 0.0022,
|
| 803 |
+
"step": 11865
|
| 804 |
},
|
| 805 |
{
|
| 806 |
+
"epoch": 11.4,
|
| 807 |
+
"learning_rate": 4.7777777777777784e-05,
|
| 808 |
+
"loss": 0.0034,
|
| 809 |
+
"step": 11970
|
| 810 |
},
|
| 811 |
{
|
| 812 |
+
"epoch": 11.5,
|
| 813 |
+
"learning_rate": 4.722222222222222e-05,
|
| 814 |
+
"loss": 0.0009,
|
| 815 |
+
"step": 12075
|
| 816 |
},
|
| 817 |
{
|
| 818 |
+
"epoch": 11.6,
|
| 819 |
+
"learning_rate": 4.666666666666667e-05,
|
| 820 |
+
"loss": 0.0012,
|
| 821 |
+
"step": 12180
|
|
|
|
|
|
|
|
|
|
|
|
|
| 822 |
},
|
| 823 |
{
|
| 824 |
+
"epoch": 11.7,
|
| 825 |
+
"learning_rate": 4.6111111111111115e-05,
|
| 826 |
+
"loss": 0.0012,
|
| 827 |
+
"step": 12285
|
| 828 |
},
|
| 829 |
{
|
| 830 |
+
"epoch": 11.8,
|
| 831 |
+
"learning_rate": 4.555555555555556e-05,
|
| 832 |
+
"loss": 0.0023,
|
| 833 |
+
"step": 12390
|
| 834 |
},
|
| 835 |
{
|
| 836 |
+
"epoch": 11.9,
|
| 837 |
+
"learning_rate": 4.5e-05,
|
| 838 |
+
"loss": 0.0003,
|
| 839 |
+
"step": 12495
|
| 840 |
},
|
| 841 |
{
|
| 842 |
+
"epoch": 12.0,
|
| 843 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 844 |
+
"loss": 0.0028,
|
| 845 |
+
"step": 12600
|
| 846 |
},
|
| 847 |
{
|
| 848 |
+
"epoch": 12.0,
|
| 849 |
+
"eval_loss": 0.00027711206348612905,
|
| 850 |
+
"eval_max_distance": 0,
|
| 851 |
+
"eval_mean_distance": 0,
|
| 852 |
+
"eval_runtime": 0.5911,
|
| 853 |
+
"eval_samples_per_second": 593.792,
|
| 854 |
+
"eval_steps_per_second": 18.609,
|
| 855 |
+
"step": 12600
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 12.1,
|
| 859 |
+
"learning_rate": 4.388888888888889e-05,
|
| 860 |
+
"loss": 0.001,
|
| 861 |
+
"step": 12705
|
|
|
|
|
|
|
|
|
|
|
|
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 12.2,
|
| 865 |
+
"learning_rate": 4.3333333333333334e-05,
|
| 866 |
+
"loss": 0.0011,
|
| 867 |
+
"step": 12810
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 12.3,
|
| 871 |
+
"learning_rate": 4.277777777777778e-05,
|
| 872 |
+
"loss": 0.0028,
|
| 873 |
+
"step": 12915
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 12.4,
|
| 877 |
+
"learning_rate": 4.222222222222222e-05,
|
| 878 |
+
"loss": 0.0009,
|
| 879 |
+
"step": 13020
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 12.5,
|
| 883 |
+
"learning_rate": 4.166666666666667e-05,
|
| 884 |
"loss": 0.0001,
|
| 885 |
+
"step": 13125
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 12.6,
|
| 889 |
+
"learning_rate": 4.111111111111111e-05,
|
| 890 |
+
"loss": 0.0005,
|
| 891 |
+
"step": 13230
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 12.7,
|
| 895 |
+
"learning_rate": 4.055555555555556e-05,
|
| 896 |
+
"loss": 0.003,
|
| 897 |
+
"step": 13335
|
|
|
|
|
|
|
|
|
|
|
|
|
| 898 |
},
|
| 899 |
{
|
| 900 |
+
"epoch": 12.8,
|
| 901 |
+
"learning_rate": 4e-05,
|
| 902 |
+
"loss": 0.0,
|
| 903 |
+
"step": 13440
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 12.9,
|
| 907 |
+
"learning_rate": 3.944444444444445e-05,
|
| 908 |
+
"loss": 0.0038,
|
| 909 |
+
"step": 13545
|
| 910 |
},
|
| 911 |
{
|
| 912 |
+
"epoch": 13.0,
|
| 913 |
+
"learning_rate": 3.888888888888889e-05,
|
| 914 |
+
"loss": 0.0005,
|
| 915 |
+
"step": 13650
|
| 916 |
},
|
| 917 |
{
|
| 918 |
+
"epoch": 13.0,
|
| 919 |
+
"eval_loss": 0.0007963060052134097,
|
| 920 |
+
"eval_max_distance": 0,
|
| 921 |
+
"eval_mean_distance": 0,
|
| 922 |
+
"eval_runtime": 0.66,
|
| 923 |
+
"eval_samples_per_second": 531.819,
|
| 924 |
+
"eval_steps_per_second": 16.667,
|
| 925 |
+
"step": 13650
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 13.1,
|
| 929 |
+
"learning_rate": 3.8333333333333334e-05,
|
| 930 |
+
"loss": 0.0009,
|
| 931 |
+
"step": 13755
|
| 932 |
},
|
| 933 |
{
|
| 934 |
+
"epoch": 13.2,
|
| 935 |
+
"learning_rate": 3.777777777777778e-05,
|
| 936 |
"loss": 0.0004,
|
| 937 |
+
"step": 13860
|
| 938 |
},
|
| 939 |
{
|
| 940 |
+
"epoch": 13.3,
|
| 941 |
+
"learning_rate": 3.722222222222222e-05,
|
| 942 |
+
"loss": 0.0085,
|
| 943 |
+
"step": 13965
|
|
|
|
|
|
|
|
|
|
|
|
|
| 944 |
},
|
| 945 |
{
|
| 946 |
+
"epoch": 13.4,
|
| 947 |
+
"learning_rate": 3.6666666666666666e-05,
|
| 948 |
+
"loss": 0.0028,
|
| 949 |
+
"step": 14070
|
| 950 |
},
|
| 951 |
{
|
| 952 |
+
"epoch": 13.5,
|
| 953 |
+
"learning_rate": 3.611111111111111e-05,
|
| 954 |
+
"loss": 0.0019,
|
| 955 |
+
"step": 14175
|
| 956 |
},
|
| 957 |
{
|
| 958 |
+
"epoch": 13.6,
|
| 959 |
+
"learning_rate": 3.555555555555556e-05,
|
| 960 |
+
"loss": 0.0008,
|
| 961 |
+
"step": 14280
|
| 962 |
},
|
| 963 |
{
|
| 964 |
+
"epoch": 13.7,
|
| 965 |
+
"learning_rate": 3.5e-05,
|
| 966 |
"loss": 0.0001,
|
| 967 |
+
"step": 14385
|
| 968 |
},
|
| 969 |
{
|
| 970 |
+
"epoch": 13.8,
|
| 971 |
+
"learning_rate": 3.444444444444445e-05,
|
| 972 |
+
"loss": 0.0014,
|
| 973 |
+
"step": 14490
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 13.9,
|
| 977 |
+
"learning_rate": 3.388888888888889e-05,
|
| 978 |
+
"loss": 0.002,
|
| 979 |
+
"step": 14595
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 14.0,
|
| 983 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 984 |
+
"loss": 0.0026,
|
| 985 |
+
"step": 14700
|
| 986 |
},
|
| 987 |
{
|
| 988 |
+
"epoch": 14.0,
|
| 989 |
+
"eval_loss": 0.0011878299992531538,
|
| 990 |
"eval_max_distance": 1,
|
| 991 |
"eval_mean_distance": 0,
|
| 992 |
+
"eval_runtime": 0.599,
|
| 993 |
+
"eval_samples_per_second": 585.977,
|
| 994 |
+
"eval_steps_per_second": 18.364,
|
| 995 |
+
"step": 14700
|
| 996 |
},
|
| 997 |
{
|
| 998 |
+
"epoch": 14.1,
|
| 999 |
+
"learning_rate": 3.277777777777778e-05,
|
| 1000 |
+
"loss": 0.0016,
|
| 1001 |
+
"step": 14805
|
| 1002 |
},
|
| 1003 |
{
|
| 1004 |
+
"epoch": 14.2,
|
| 1005 |
+
"learning_rate": 3.222222222222223e-05,
|
| 1006 |
"loss": 0.0008,
|
| 1007 |
+
"step": 14910
|
| 1008 |
},
|
| 1009 |
{
|
| 1010 |
+
"epoch": 14.3,
|
| 1011 |
+
"learning_rate": 3.1666666666666666e-05,
|
| 1012 |
"loss": 0.0005,
|
| 1013 |
+
"step": 15015
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
+
"epoch": 14.4,
|
| 1017 |
+
"learning_rate": 3.111111111111111e-05,
|
| 1018 |
+
"loss": 0.0005,
|
| 1019 |
+
"step": 15120
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1020 |
},
|
| 1021 |
{
|
| 1022 |
+
"epoch": 14.5,
|
| 1023 |
+
"learning_rate": 3.055555555555556e-05,
|
| 1024 |
+
"loss": 0.001,
|
| 1025 |
+
"step": 15225
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
+
"epoch": 14.6,
|
| 1029 |
+
"learning_rate": 3e-05,
|
| 1030 |
+
"loss": 0.002,
|
| 1031 |
+
"step": 15330
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
+
"epoch": 14.7,
|
| 1035 |
+
"learning_rate": 2.9444444444444448e-05,
|
| 1036 |
+
"loss": 0.0001,
|
| 1037 |
+
"step": 15435
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
+
"epoch": 14.8,
|
| 1041 |
+
"learning_rate": 2.8888888888888888e-05,
|
| 1042 |
+
"loss": 0.0,
|
| 1043 |
+
"step": 15540
|
| 1044 |
},
|
| 1045 |
{
|
| 1046 |
+
"epoch": 14.9,
|
| 1047 |
+
"learning_rate": 2.8333333333333335e-05,
|
| 1048 |
+
"loss": 0.0,
|
| 1049 |
+
"step": 15645
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
+
"epoch": 15.0,
|
| 1053 |
+
"learning_rate": 2.777777777777778e-05,
|
| 1054 |
+
"loss": 0.0011,
|
| 1055 |
+
"step": 15750
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
+
"epoch": 15.0,
|
| 1059 |
+
"eval_loss": 0.017123280093073845,
|
| 1060 |
"eval_max_distance": 1,
|
| 1061 |
"eval_mean_distance": 0,
|
| 1062 |
+
"eval_runtime": 0.6147,
|
| 1063 |
+
"eval_samples_per_second": 570.998,
|
| 1064 |
+
"eval_steps_per_second": 17.895,
|
| 1065 |
+
"step": 15750
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 15.1,
|
| 1069 |
+
"learning_rate": 2.7222222222222223e-05,
|
| 1070 |
+
"loss": 0.0001,
|
| 1071 |
+
"step": 15855
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 15.2,
|
| 1075 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 1076 |
+
"loss": 0.0031,
|
| 1077 |
+
"step": 15960
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
+
"epoch": 15.3,
|
| 1081 |
+
"learning_rate": 2.6111111111111114e-05,
|
| 1082 |
"loss": 0.0001,
|
| 1083 |
+
"step": 16065
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
+
"epoch": 15.4,
|
| 1087 |
+
"learning_rate": 2.5555555555555554e-05,
|
| 1088 |
+
"loss": 0.0016,
|
| 1089 |
+
"step": 16170
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
+
"epoch": 15.5,
|
| 1093 |
+
"learning_rate": 2.5e-05,
|
| 1094 |
+
"loss": 0.0006,
|
| 1095 |
+
"step": 16275
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
+
"epoch": 15.6,
|
| 1099 |
+
"learning_rate": 2.4444444444444445e-05,
|
| 1100 |
+
"loss": 0.0002,
|
| 1101 |
+
"step": 16380
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 15.7,
|
| 1105 |
+
"learning_rate": 2.3888888888888892e-05,
|
| 1106 |
"loss": 0.0008,
|
| 1107 |
+
"step": 16485
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
+
"epoch": 15.8,
|
| 1111 |
+
"learning_rate": 2.3333333333333336e-05,
|
| 1112 |
+
"loss": 0.0016,
|
| 1113 |
+
"step": 16590
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
+
"epoch": 15.9,
|
| 1117 |
+
"learning_rate": 2.277777777777778e-05,
|
| 1118 |
"loss": 0.0002,
|
| 1119 |
+
"step": 16695
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
+
"epoch": 16.0,
|
| 1123 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 1124 |
+
"loss": 0.0,
|
| 1125 |
+
"step": 16800
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
+
"epoch": 16.0,
|
| 1129 |
+
"eval_loss": 0.011970149353146553,
|
| 1130 |
"eval_max_distance": 1,
|
| 1131 |
"eval_mean_distance": 0,
|
| 1132 |
+
"eval_runtime": 0.591,
|
| 1133 |
+
"eval_samples_per_second": 593.899,
|
| 1134 |
+
"eval_steps_per_second": 18.612,
|
| 1135 |
+
"step": 16800
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
+
"epoch": 16.1,
|
| 1139 |
+
"learning_rate": 2.1666666666666667e-05,
|
| 1140 |
+
"loss": 0.0009,
|
| 1141 |
+
"step": 16905
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1142 |
},
|
| 1143 |
{
|
| 1144 |
+
"epoch": 16.2,
|
| 1145 |
+
"learning_rate": 2.111111111111111e-05,
|
| 1146 |
+
"loss": 0.0009,
|
| 1147 |
+
"step": 17010
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
+
"epoch": 16.3,
|
| 1151 |
+
"learning_rate": 2.0555555555555555e-05,
|
| 1152 |
+
"loss": 0.0004,
|
| 1153 |
+
"step": 17115
|
| 1154 |
},
|
| 1155 |
{
|
| 1156 |
+
"epoch": 16.4,
|
| 1157 |
+
"learning_rate": 2e-05,
|
| 1158 |
+
"loss": 0.0001,
|
| 1159 |
+
"step": 17220
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
+
"epoch": 16.5,
|
| 1163 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 1164 |
+
"loss": 0.0032,
|
| 1165 |
+
"step": 17325
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
+
"epoch": 16.6,
|
| 1169 |
+
"learning_rate": 1.888888888888889e-05,
|
| 1170 |
+
"loss": 0.0,
|
| 1171 |
+
"step": 17430
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
+
"epoch": 16.7,
|
| 1175 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 1176 |
+
"loss": 0.0002,
|
| 1177 |
+
"step": 17535
|
| 1178 |
},
|
| 1179 |
{
|
| 1180 |
+
"epoch": 16.8,
|
| 1181 |
+
"learning_rate": 1.777777777777778e-05,
|
| 1182 |
+
"loss": 0.0,
|
| 1183 |
+
"step": 17640
|
| 1184 |
},
|
| 1185 |
{
|
| 1186 |
+
"epoch": 16.9,
|
| 1187 |
+
"learning_rate": 1.7222222222222224e-05,
|
| 1188 |
+
"loss": 0.001,
|
| 1189 |
+
"step": 17745
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
+
"epoch": 17.0,
|
| 1193 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1194 |
+
"loss": 0.0001,
|
| 1195 |
+
"step": 17850
|
| 1196 |
},
|
| 1197 |
{
|
| 1198 |
+
"epoch": 17.0,
|
| 1199 |
+
"eval_loss": 0.003116948762908578,
|
| 1200 |
"eval_max_distance": 1,
|
| 1201 |
"eval_mean_distance": 0,
|
| 1202 |
+
"eval_runtime": 0.6156,
|
| 1203 |
+
"eval_samples_per_second": 570.218,
|
| 1204 |
+
"eval_steps_per_second": 17.87,
|
| 1205 |
+
"step": 17850
|
| 1206 |
},
|
| 1207 |
{
|
| 1208 |
+
"epoch": 17.1,
|
| 1209 |
+
"learning_rate": 1.6111111111111115e-05,
|
| 1210 |
+
"loss": 0.0011,
|
| 1211 |
+
"step": 17955
|
| 1212 |
},
|
| 1213 |
{
|
| 1214 |
+
"epoch": 17.2,
|
| 1215 |
+
"learning_rate": 1.5555555555555555e-05,
|
| 1216 |
+
"loss": 0.0004,
|
| 1217 |
+
"step": 18060
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
+
"epoch": 17.3,
|
| 1221 |
+
"learning_rate": 1.5e-05,
|
| 1222 |
"loss": 0.0003,
|
| 1223 |
+
"step": 18165
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
+
"epoch": 17.4,
|
| 1227 |
+
"learning_rate": 1.4444444444444444e-05,
|
| 1228 |
+
"loss": 0.0001,
|
| 1229 |
+
"step": 18270
|
| 1230 |
},
|
| 1231 |
{
|
| 1232 |
+
"epoch": 17.5,
|
| 1233 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1234 |
+
"loss": 0.0001,
|
| 1235 |
+
"step": 18375
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
+
"epoch": 17.6,
|
| 1239 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 1240 |
+
"loss": 0.0004,
|
| 1241 |
+
"step": 18480
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
+
"epoch": 17.7,
|
| 1245 |
+
"learning_rate": 1.2777777777777777e-05,
|
| 1246 |
+
"loss": 0.0007,
|
| 1247 |
+
"step": 18585
|
| 1248 |
},
|
| 1249 |
{
|
| 1250 |
+
"epoch": 17.8,
|
| 1251 |
+
"learning_rate": 1.2222222222222222e-05,
|
| 1252 |
+
"loss": 0.0,
|
| 1253 |
+
"step": 18690
|
| 1254 |
},
|
| 1255 |
{
|
| 1256 |
+
"epoch": 17.9,
|
| 1257 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 1258 |
+
"loss": 0.0,
|
| 1259 |
+
"step": 18795
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
+
"epoch": 18.0,
|
| 1263 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 1264 |
"loss": 0.0007,
|
| 1265 |
+
"step": 18900
|
| 1266 |
},
|
| 1267 |
{
|
| 1268 |
+
"epoch": 18.0,
|
| 1269 |
+
"eval_loss": 0.016535792499780655,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1270 |
"eval_max_distance": 1,
|
| 1271 |
"eval_mean_distance": 0,
|
| 1272 |
+
"eval_runtime": 0.5881,
|
| 1273 |
+
"eval_samples_per_second": 596.809,
|
| 1274 |
+
"eval_steps_per_second": 18.703,
|
| 1275 |
+
"step": 18900
|
| 1276 |
},
|
| 1277 |
{
|
| 1278 |
+
"epoch": 18.1,
|
| 1279 |
+
"learning_rate": 1.0555555555555555e-05,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1280 |
"loss": 0.0004,
|
| 1281 |
+
"step": 19005
|
| 1282 |
},
|
| 1283 |
{
|
| 1284 |
+
"epoch": 18.2,
|
| 1285 |
+
"learning_rate": 1e-05,
|
| 1286 |
"loss": 0.0002,
|
| 1287 |
+
"step": 19110
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
+
"epoch": 18.3,
|
| 1291 |
+
"learning_rate": 9.444444444444445e-06,
|
| 1292 |
+
"loss": 0.0001,
|
| 1293 |
+
"step": 19215
|
| 1294 |
},
|
| 1295 |
{
|
| 1296 |
+
"epoch": 18.4,
|
| 1297 |
+
"learning_rate": 8.88888888888889e-06,
|
| 1298 |
"loss": 0.0009,
|
| 1299 |
+
"step": 19320
|
| 1300 |
},
|
| 1301 |
{
|
| 1302 |
+
"epoch": 18.5,
|
| 1303 |
+
"learning_rate": 8.333333333333334e-06,
|
| 1304 |
+
"loss": 0.0149,
|
| 1305 |
+
"step": 19425
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
+
"epoch": 18.6,
|
| 1309 |
+
"learning_rate": 7.777777777777777e-06,
|
| 1310 |
+
"loss": 0.0003,
|
| 1311 |
+
"step": 19530
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
+
"epoch": 18.7,
|
| 1315 |
+
"learning_rate": 7.222222222222222e-06,
|
| 1316 |
+
"loss": 0.0,
|
| 1317 |
+
"step": 19635
|
| 1318 |
},
|
| 1319 |
{
|
| 1320 |
+
"epoch": 18.8,
|
| 1321 |
+
"learning_rate": 6.666666666666667e-06,
|
| 1322 |
+
"loss": 0.0,
|
| 1323 |
+
"step": 19740
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
+
"epoch": 18.9,
|
| 1327 |
+
"learning_rate": 6.111111111111111e-06,
|
| 1328 |
+
"loss": 0.0,
|
| 1329 |
+
"step": 19845
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
+
"epoch": 19.0,
|
| 1333 |
+
"learning_rate": 5.555555555555556e-06,
|
| 1334 |
+
"loss": 0.001,
|
| 1335 |
+
"step": 19950
|
| 1336 |
},
|
| 1337 |
{
|
| 1338 |
+
"epoch": 19.0,
|
| 1339 |
+
"eval_loss": 0.01935943029820919,
|
| 1340 |
"eval_max_distance": 1,
|
| 1341 |
"eval_mean_distance": 0,
|
| 1342 |
+
"eval_runtime": 0.6346,
|
| 1343 |
+
"eval_samples_per_second": 553.109,
|
| 1344 |
+
"eval_steps_per_second": 17.334,
|
| 1345 |
+
"step": 19950
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1346 |
},
|
| 1347 |
{
|
| 1348 |
+
"epoch": 19.1,
|
| 1349 |
+
"learning_rate": 5e-06,
|
| 1350 |
+
"loss": 0.001,
|
| 1351 |
+
"step": 20055
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
+
"epoch": 19.2,
|
| 1355 |
+
"learning_rate": 4.444444444444445e-06,
|
| 1356 |
+
"loss": 0.0,
|
| 1357 |
+
"step": 20160
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
+
"epoch": 19.3,
|
| 1361 |
+
"learning_rate": 3.888888888888889e-06,
|
| 1362 |
+
"loss": 0.0004,
|
| 1363 |
+
"step": 20265
|
| 1364 |
},
|
| 1365 |
{
|
| 1366 |
+
"epoch": 19.4,
|
| 1367 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 1368 |
+
"loss": 0.0,
|
| 1369 |
+
"step": 20370
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
+
"epoch": 19.5,
|
| 1373 |
+
"learning_rate": 2.777777777777778e-06,
|
| 1374 |
+
"loss": 0.0011,
|
| 1375 |
+
"step": 20475
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
+
"epoch": 19.6,
|
| 1379 |
+
"learning_rate": 2.2222222222222225e-06,
|
| 1380 |
+
"loss": 0.0011,
|
| 1381 |
+
"step": 20580
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
+
"epoch": 19.7,
|
| 1385 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 1386 |
+
"loss": 0.0,
|
| 1387 |
+
"step": 20685
|
| 1388 |
},
|
| 1389 |
{
|
| 1390 |
+
"epoch": 19.8,
|
| 1391 |
+
"learning_rate": 1.1111111111111112e-06,
|
| 1392 |
+
"loss": 0.0003,
|
| 1393 |
+
"step": 20790
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
+
"epoch": 19.9,
|
| 1397 |
+
"learning_rate": 5.555555555555556e-07,
|
| 1398 |
+
"loss": 0.0004,
|
| 1399 |
+
"step": 20895
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
+
"epoch": 20.0,
|
| 1403 |
+
"learning_rate": 0.0,
|
| 1404 |
+
"loss": 0.0001,
|
| 1405 |
+
"step": 21000
|
| 1406 |
},
|
| 1407 |
{
|
| 1408 |
+
"epoch": 20.0,
|
| 1409 |
+
"eval_loss": 0.010230864398181438,
|
| 1410 |
"eval_max_distance": 1,
|
| 1411 |
"eval_mean_distance": 0,
|
| 1412 |
+
"eval_runtime": 0.597,
|
| 1413 |
+
"eval_samples_per_second": 587.939,
|
| 1414 |
+
"eval_steps_per_second": 18.425,
|
| 1415 |
+
"step": 21000
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 20.0,
|
| 1419 |
+
"step": 21000,
|
| 1420 |
+
"total_flos": 3.345253371675648e+16,
|
| 1421 |
+
"train_loss": 0.6543119361943335,
|
| 1422 |
+
"train_runtime": 2482.634,
|
| 1423 |
+
"train_samples_per_second": 270.592,
|
| 1424 |
+
"train_steps_per_second": 8.459
|
| 1425 |
}
|
| 1426 |
],
|
| 1427 |
+
"logging_steps": 105,
|
| 1428 |
+
"max_steps": 21000,
|
| 1429 |
+
"num_train_epochs": 20,
|
| 1430 |
+
"save_steps": 210,
|
| 1431 |
+
"total_flos": 3.345253371675648e+16,
|
| 1432 |
"trial_name": null,
|
| 1433 |
"trial_params": null
|
| 1434 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4664
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a608a335d69abf0fd9d81359bad97ed11bb291ac004c10e1a3220cf8140d433
|
| 3 |
size 4664
|