End of training
Browse files- README.md +33 -47
- all_results.json +5 -16
- train_results.json +5 -5
- trainer_state.json +189 -399
README.md
CHANGED
|
@@ -20,8 +20,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 20 |
|
| 21 |
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mbe dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
-
- Accuracy: 0.
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
@@ -49,56 +49,42 @@ The following hyperparameters were used during training:
|
|
| 49 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
- lr_scheduler_type: constant
|
| 51 |
- lr_scheduler_warmup_ratio: 0.03
|
| 52 |
-
-
|
| 53 |
|
| 54 |
### Training results
|
| 55 |
|
| 56 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 57 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 58 |
-
| 1.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
| 62 |
-
| 0.
|
| 63 |
-
| 0.586 | 0.4 | 60 | 0.
|
| 64 |
-
| 0.
|
| 65 |
-
| 0.
|
| 66 |
-
| 0.
|
| 67 |
-
| 0.
|
| 68 |
-
| 0.
|
| 69 |
-
| 0.
|
| 70 |
-
| 0.
|
| 71 |
-
| 0.
|
| 72 |
-
| 0.
|
| 73 |
-
| 0.
|
| 74 |
-
| 0.
|
| 75 |
-
| 0.
|
| 76 |
-
| 0.
|
| 77 |
-
| 0.
|
| 78 |
-
| 0.
|
| 79 |
-
| 0.
|
| 80 |
-
| 0.
|
| 81 |
-
| 0.
|
| 82 |
-
| 0.
|
| 83 |
-
| 0.
|
| 84 |
-
| 0.
|
| 85 |
-
| 0.
|
| 86 |
-
| 0.
|
| 87 |
-
| 0.
|
| 88 |
-
| 0.2304 | 2.07 | 310 | 0.8200 | 0.5362 |
|
| 89 |
-
| 0.1696 | 2.14 | 320 | 0.9087 | 0.5296 |
|
| 90 |
-
| 0.2255 | 2.2 | 330 | 0.7566 | 0.5362 |
|
| 91 |
-
| 0.1923 | 2.27 | 340 | 0.7020 | 0.5197 |
|
| 92 |
-
| 0.281 | 2.34 | 350 | 0.6653 | 0.5033 |
|
| 93 |
-
| 0.2311 | 2.4 | 360 | 0.6412 | 0.5132 |
|
| 94 |
-
| 0.1523 | 2.47 | 370 | 0.8846 | 0.5230 |
|
| 95 |
-
| 0.2451 | 2.54 | 380 | 0.9252 | 0.5164 |
|
| 96 |
-
| 0.2022 | 2.6 | 390 | 0.7422 | 0.5197 |
|
| 97 |
-
| 0.217 | 2.67 | 400 | 0.7558 | 0.5329 |
|
| 98 |
-
| 0.165 | 2.74 | 410 | 0.7846 | 0.5428 |
|
| 99 |
-
| 0.2025 | 2.8 | 420 | 0.7254 | 0.5230 |
|
| 100 |
-
| 0.2201 | 2.87 | 430 | 0.6531 | 0.5296 |
|
| 101 |
-
| 0.2037 | 2.94 | 440 | 0.7827 | 0.5493 |
|
| 102 |
|
| 103 |
|
| 104 |
### Framework versions
|
|
|
|
| 20 |
|
| 21 |
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the mbe dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.5616
|
| 24 |
+
- Accuracy: 0.5362
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
|
|
| 49 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
- lr_scheduler_type: constant
|
| 51 |
- lr_scheduler_warmup_ratio: 0.03
|
| 52 |
+
- training_steps: 300
|
| 53 |
|
| 54 |
### Training results
|
| 55 |
|
| 56 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 57 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 58 |
+
| 1.5245 | 0.07 | 10 | 0.6507 | 0.3355 |
|
| 59 |
+
| 0.6666 | 0.13 | 20 | 0.6464 | 0.3816 |
|
| 60 |
+
| 0.6527 | 0.2 | 30 | 0.6427 | 0.3684 |
|
| 61 |
+
| 0.6168 | 0.27 | 40 | 0.6321 | 0.3980 |
|
| 62 |
+
| 0.6584 | 0.33 | 50 | 0.6182 | 0.3914 |
|
| 63 |
+
| 0.586 | 0.4 | 60 | 0.6244 | 0.4145 |
|
| 64 |
+
| 0.5924 | 0.47 | 70 | 0.6034 | 0.4342 |
|
| 65 |
+
| 0.6069 | 0.53 | 80 | 0.6096 | 0.4375 |
|
| 66 |
+
| 0.5999 | 0.6 | 90 | 0.6096 | 0.4408 |
|
| 67 |
+
| 0.6206 | 0.67 | 100 | 0.6070 | 0.4572 |
|
| 68 |
+
| 0.5793 | 0.73 | 110 | 0.6016 | 0.4572 |
|
| 69 |
+
| 0.6208 | 0.8 | 120 | 0.5902 | 0.4605 |
|
| 70 |
+
| 0.5622 | 0.87 | 130 | 0.5775 | 0.4770 |
|
| 71 |
+
| 0.5502 | 0.93 | 140 | 0.5761 | 0.4671 |
|
| 72 |
+
| 0.5958 | 1.0 | 150 | 0.5606 | 0.4901 |
|
| 73 |
+
| 0.4558 | 1.07 | 160 | 0.5840 | 0.4737 |
|
| 74 |
+
| 0.4411 | 1.14 | 170 | 0.5631 | 0.4901 |
|
| 75 |
+
| 0.4144 | 1.2 | 180 | 0.5745 | 0.5 |
|
| 76 |
+
| 0.4647 | 1.27 | 190 | 0.5932 | 0.4605 |
|
| 77 |
+
| 0.4504 | 1.34 | 200 | 0.5799 | 0.5099 |
|
| 78 |
+
| 0.4299 | 1.4 | 210 | 0.6488 | 0.4934 |
|
| 79 |
+
| 0.425 | 1.47 | 220 | 0.5704 | 0.5132 |
|
| 80 |
+
| 0.4152 | 1.54 | 230 | 0.5582 | 0.5066 |
|
| 81 |
+
| 0.425 | 1.6 | 240 | 0.5489 | 0.5329 |
|
| 82 |
+
| 0.446 | 1.67 | 250 | 0.5479 | 0.5197 |
|
| 83 |
+
| 0.3908 | 1.74 | 260 | 0.5564 | 0.5164 |
|
| 84 |
+
| 0.443 | 1.8 | 270 | 0.5419 | 0.5033 |
|
| 85 |
+
| 0.4081 | 1.87 | 280 | 0.5948 | 0.5066 |
|
| 86 |
+
| 0.3944 | 1.94 | 290 | 0.5547 | 0.5395 |
|
| 87 |
+
| 0.4005 | 2.0 | 300 | 0.5616 | 0.5362 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
|
| 89 |
|
| 90 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,18 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 2.
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"eval_precision_macro": 0.7222302630120379,
|
| 8 |
-
"eval_precision_micro": 0.8141509433962264,
|
| 9 |
-
"eval_recall_macro": 0.7125706602249756,
|
| 10 |
-
"eval_recall_micro": 0.8141509433962264,
|
| 11 |
-
"eval_runtime": 19.7133,
|
| 12 |
-
"eval_samples_per_second": 25.009,
|
| 13 |
-
"eval_steps_per_second": 1.573,
|
| 14 |
-
"train_loss": 0.43748926956381573,
|
| 15 |
-
"train_runtime": 1886.9121,
|
| 16 |
-
"train_samples_per_second": 3.809,
|
| 17 |
-
"train_steps_per_second": 0.237
|
| 18 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 0.5479170862833659,
|
| 4 |
+
"train_runtime": 1462.1781,
|
| 5 |
+
"train_samples_per_second": 3.283,
|
| 6 |
+
"train_steps_per_second": 0.205
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
}
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 2.
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second": 3.
|
| 6 |
-
"train_steps_per_second": 0.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 0.5479170862833659,
|
| 4 |
+
"train_runtime": 1462.1781,
|
| 5 |
+
"train_samples_per_second": 3.283,
|
| 6 |
+
"train_steps_per_second": 0.205
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 2.
|
| 5 |
"eval_steps": 10,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -11,76 +11,76 @@
|
|
| 11 |
{
|
| 12 |
"epoch": 0.07,
|
| 13 |
"learning_rate": 3e-05,
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.07,
|
| 19 |
-
"eval_accuracy": 0.
|
| 20 |
-
"eval_loss": 0.
|
| 21 |
-
"eval_runtime": 16.
|
| 22 |
-
"eval_samples_per_second":
|
| 23 |
-
"eval_steps_per_second": 4.
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.13,
|
| 28 |
"learning_rate": 3e-05,
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 20
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.13,
|
| 34 |
-
"eval_accuracy": 0.
|
| 35 |
-
"eval_loss": 0.
|
| 36 |
-
"eval_runtime": 16.
|
| 37 |
-
"eval_samples_per_second":
|
| 38 |
-
"eval_steps_per_second": 4.
|
| 39 |
"step": 20
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.2,
|
| 43 |
"learning_rate": 3e-05,
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.2,
|
| 49 |
-
"eval_accuracy": 0.
|
| 50 |
-
"eval_loss": 0.
|
| 51 |
-
"eval_runtime": 16.
|
| 52 |
-
"eval_samples_per_second":
|
| 53 |
-
"eval_steps_per_second": 4.
|
| 54 |
"step": 30
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.27,
|
| 58 |
"learning_rate": 3e-05,
|
| 59 |
-
"loss": 0.
|
| 60 |
"step": 40
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 0.27,
|
| 64 |
-
"eval_accuracy": 0.
|
| 65 |
-
"eval_loss": 0.
|
| 66 |
-
"eval_runtime":
|
| 67 |
-
"eval_samples_per_second":
|
| 68 |
-
"eval_steps_per_second": 4.
|
| 69 |
"step": 40
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.33,
|
| 73 |
"learning_rate": 3e-05,
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 50
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.33,
|
| 79 |
-
"eval_accuracy": 0.
|
| 80 |
-
"eval_loss": 0.
|
| 81 |
-
"eval_runtime": 16.
|
| 82 |
-
"eval_samples_per_second":
|
| 83 |
-
"eval_steps_per_second": 4.
|
| 84 |
"step": 50
|
| 85 |
},
|
| 86 |
{
|
|
@@ -91,599 +91,389 @@
|
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"epoch": 0.4,
|
| 94 |
-
"eval_accuracy": 0.
|
| 95 |
-
"eval_loss": 0.
|
| 96 |
-
"eval_runtime": 16.
|
| 97 |
-
"eval_samples_per_second":
|
| 98 |
-
"eval_steps_per_second": 4.
|
| 99 |
"step": 60
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.47,
|
| 103 |
"learning_rate": 3e-05,
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 70
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.47,
|
| 109 |
-
"eval_accuracy": 0.
|
| 110 |
-
"eval_loss": 0.
|
| 111 |
-
"eval_runtime": 16.
|
| 112 |
-
"eval_samples_per_second":
|
| 113 |
-
"eval_steps_per_second": 4.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.53,
|
| 118 |
"learning_rate": 3e-05,
|
| 119 |
-
"loss": 0.
|
| 120 |
"step": 80
|
| 121 |
},
|
| 122 |
{
|
| 123 |
"epoch": 0.53,
|
| 124 |
-
"eval_accuracy": 0.
|
| 125 |
-
"eval_loss": 0.
|
| 126 |
-
"eval_runtime": 16.
|
| 127 |
-
"eval_samples_per_second":
|
| 128 |
-
"eval_steps_per_second": 4.
|
| 129 |
"step": 80
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.6,
|
| 133 |
"learning_rate": 3e-05,
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 90
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.6,
|
| 139 |
-
"eval_accuracy": 0.
|
| 140 |
-
"eval_loss": 0.
|
| 141 |
-
"eval_runtime": 16.
|
| 142 |
-
"eval_samples_per_second":
|
| 143 |
-
"eval_steps_per_second": 4.
|
| 144 |
"step": 90
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.67,
|
| 148 |
"learning_rate": 3e-05,
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 100
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.67,
|
| 154 |
-
"eval_accuracy": 0.
|
| 155 |
-
"eval_loss": 0.
|
| 156 |
-
"eval_runtime": 16.
|
| 157 |
-
"eval_samples_per_second":
|
| 158 |
-
"eval_steps_per_second": 4.
|
| 159 |
"step": 100
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.73,
|
| 163 |
"learning_rate": 3e-05,
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 110
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.73,
|
| 169 |
-
"eval_accuracy": 0.
|
| 170 |
-
"eval_loss": 0.
|
| 171 |
-
"eval_runtime": 16.
|
| 172 |
-
"eval_samples_per_second":
|
| 173 |
-
"eval_steps_per_second": 4.
|
| 174 |
"step": 110
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.8,
|
| 178 |
"learning_rate": 3e-05,
|
| 179 |
-
"loss": 0.
|
| 180 |
"step": 120
|
| 181 |
},
|
| 182 |
{
|
| 183 |
"epoch": 0.8,
|
| 184 |
-
"eval_accuracy": 0.
|
| 185 |
-
"eval_loss": 0.
|
| 186 |
-
"eval_runtime": 16.
|
| 187 |
-
"eval_samples_per_second":
|
| 188 |
-
"eval_steps_per_second": 4.
|
| 189 |
"step": 120
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.87,
|
| 193 |
"learning_rate": 3e-05,
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 130
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 0.87,
|
| 199 |
-
"eval_accuracy": 0.
|
| 200 |
-
"eval_loss": 0.
|
| 201 |
-
"eval_runtime": 16.
|
| 202 |
-
"eval_samples_per_second":
|
| 203 |
-
"eval_steps_per_second": 4.
|
| 204 |
"step": 130
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 0.93,
|
| 208 |
"learning_rate": 3e-05,
|
| 209 |
-
"loss": 0.
|
| 210 |
"step": 140
|
| 211 |
},
|
| 212 |
{
|
| 213 |
"epoch": 0.93,
|
| 214 |
-
"eval_accuracy": 0.
|
| 215 |
-
"eval_loss": 0.
|
| 216 |
-
"eval_runtime": 16.
|
| 217 |
-
"eval_samples_per_second":
|
| 218 |
-
"eval_steps_per_second": 4.
|
| 219 |
"step": 140
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 1.0,
|
| 223 |
"learning_rate": 3e-05,
|
| 224 |
-
"loss": 0.
|
| 225 |
"step": 150
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 1.0,
|
| 229 |
-
"eval_accuracy": 0.
|
| 230 |
-
"eval_loss": 0.
|
| 231 |
-
"eval_runtime": 16.
|
| 232 |
-
"eval_samples_per_second":
|
| 233 |
-
"eval_steps_per_second": 4.
|
| 234 |
"step": 150
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 1.07,
|
| 238 |
"learning_rate": 3e-05,
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 1.07,
|
| 244 |
-
"eval_accuracy": 0.
|
| 245 |
-
"eval_loss": 0.
|
| 246 |
-
"eval_runtime": 16.
|
| 247 |
-
"eval_samples_per_second":
|
| 248 |
-
"eval_steps_per_second": 4.
|
| 249 |
"step": 160
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 1.14,
|
| 253 |
"learning_rate": 3e-05,
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 170
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.14,
|
| 259 |
-
"eval_accuracy": 0.
|
| 260 |
-
"eval_loss": 0.
|
| 261 |
-
"eval_runtime": 16.
|
| 262 |
-
"eval_samples_per_second":
|
| 263 |
-
"eval_steps_per_second": 4.
|
| 264 |
"step": 170
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 1.2,
|
| 268 |
"learning_rate": 3e-05,
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 180
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 1.2,
|
| 274 |
-
"eval_accuracy": 0.
|
| 275 |
-
"eval_loss": 0.
|
| 276 |
-
"eval_runtime": 16.
|
| 277 |
-
"eval_samples_per_second":
|
| 278 |
-
"eval_steps_per_second": 4.
|
| 279 |
"step": 180
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 1.27,
|
| 283 |
"learning_rate": 3e-05,
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 190
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 1.27,
|
| 289 |
-
"eval_accuracy": 0.
|
| 290 |
-
"eval_loss": 0.
|
| 291 |
-
"eval_runtime": 16.
|
| 292 |
-
"eval_samples_per_second":
|
| 293 |
-
"eval_steps_per_second": 4.
|
| 294 |
"step": 190
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 1.34,
|
| 298 |
"learning_rate": 3e-05,
|
| 299 |
-
"loss": 0.
|
| 300 |
"step": 200
|
| 301 |
},
|
| 302 |
{
|
| 303 |
"epoch": 1.34,
|
| 304 |
"eval_accuracy": 0.5098684210526315,
|
| 305 |
-
"eval_loss": 0.
|
| 306 |
-
"eval_runtime": 16.
|
| 307 |
-
"eval_samples_per_second":
|
| 308 |
-
"eval_steps_per_second": 4.
|
| 309 |
"step": 200
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 1.4,
|
| 313 |
"learning_rate": 3e-05,
|
| 314 |
-
"loss": 0.
|
| 315 |
"step": 210
|
| 316 |
},
|
| 317 |
{
|
| 318 |
"epoch": 1.4,
|
| 319 |
-
"eval_accuracy": 0.
|
| 320 |
-
"eval_loss": 0.
|
| 321 |
-
"eval_runtime": 16.
|
| 322 |
-
"eval_samples_per_second":
|
| 323 |
-
"eval_steps_per_second": 4.
|
| 324 |
"step": 210
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 1.47,
|
| 328 |
"learning_rate": 3e-05,
|
| 329 |
-
"loss": 0.
|
| 330 |
"step": 220
|
| 331 |
},
|
| 332 |
{
|
| 333 |
"epoch": 1.47,
|
| 334 |
-
"eval_accuracy": 0.
|
| 335 |
-
"eval_loss": 0.
|
| 336 |
-
"eval_runtime": 16.
|
| 337 |
-
"eval_samples_per_second":
|
| 338 |
-
"eval_steps_per_second": 4.
|
| 339 |
"step": 220
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.54,
|
| 343 |
"learning_rate": 3e-05,
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 230
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 1.54,
|
| 349 |
-
"eval_accuracy": 0.
|
| 350 |
-
"eval_loss": 0.
|
| 351 |
-
"eval_runtime": 16.
|
| 352 |
-
"eval_samples_per_second":
|
| 353 |
-
"eval_steps_per_second": 4.
|
| 354 |
"step": 230
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 1.6,
|
| 358 |
"learning_rate": 3e-05,
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 240
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.6,
|
| 364 |
-
"eval_accuracy": 0.
|
| 365 |
-
"eval_loss": 0.
|
| 366 |
-
"eval_runtime": 16.
|
| 367 |
-
"eval_samples_per_second":
|
| 368 |
-
"eval_steps_per_second": 4.
|
| 369 |
"step": 240
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 1.67,
|
| 373 |
"learning_rate": 3e-05,
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 250
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 1.67,
|
| 379 |
-
"eval_accuracy": 0.
|
| 380 |
-
"eval_loss": 0.
|
| 381 |
-
"eval_runtime": 16.
|
| 382 |
-
"eval_samples_per_second":
|
| 383 |
-
"eval_steps_per_second": 4.
|
| 384 |
"step": 250
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 1.74,
|
| 388 |
"learning_rate": 3e-05,
|
| 389 |
-
"loss": 0.
|
| 390 |
"step": 260
|
| 391 |
},
|
| 392 |
{
|
| 393 |
"epoch": 1.74,
|
| 394 |
-
"eval_accuracy": 0.
|
| 395 |
-
"eval_loss": 0.
|
| 396 |
-
"eval_runtime": 16.
|
| 397 |
-
"eval_samples_per_second":
|
| 398 |
-
"eval_steps_per_second": 4.
|
| 399 |
"step": 260
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 1.8,
|
| 403 |
"learning_rate": 3e-05,
|
| 404 |
-
"loss": 0.
|
| 405 |
"step": 270
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 1.8,
|
| 409 |
-
"eval_accuracy": 0.
|
| 410 |
-
"eval_loss": 0.
|
| 411 |
-
"eval_runtime": 16.
|
| 412 |
-
"eval_samples_per_second":
|
| 413 |
-
"eval_steps_per_second": 4.
|
| 414 |
"step": 270
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 1.87,
|
| 418 |
"learning_rate": 3e-05,
|
| 419 |
-
"loss": 0.
|
| 420 |
"step": 280
|
| 421 |
},
|
| 422 |
{
|
| 423 |
"epoch": 1.87,
|
| 424 |
"eval_accuracy": 0.506578947368421,
|
| 425 |
-
"eval_loss": 0.
|
| 426 |
-
"eval_runtime": 16.
|
| 427 |
-
"eval_samples_per_second":
|
| 428 |
-
"eval_steps_per_second": 4.
|
| 429 |
"step": 280
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 1.94,
|
| 433 |
"learning_rate": 3e-05,
|
| 434 |
-
"loss": 0.
|
| 435 |
"step": 290
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 1.94,
|
| 439 |
-
"eval_accuracy": 0.
|
| 440 |
-
"eval_loss": 0.
|
| 441 |
-
"eval_runtime": 16.
|
| 442 |
-
"eval_samples_per_second":
|
| 443 |
-
"eval_steps_per_second": 4.
|
| 444 |
"step": 290
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 2.0,
|
| 448 |
"learning_rate": 3e-05,
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 300
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 2.0,
|
| 454 |
-
"eval_accuracy": 0.5164473684210527,
|
| 455 |
-
"eval_loss": 0.5669550895690918,
|
| 456 |
-
"eval_runtime": 16.3688,
|
| 457 |
-
"eval_samples_per_second": 18.572,
|
| 458 |
-
"eval_steps_per_second": 4.643,
|
| 459 |
-
"step": 300
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 2.07,
|
| 463 |
-
"learning_rate": 3e-05,
|
| 464 |
-
"loss": 0.2304,
|
| 465 |
-
"step": 310
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 2.07,
|
| 469 |
-
"eval_accuracy": 0.5361842105263158,
|
| 470 |
-
"eval_loss": 0.8199814558029175,
|
| 471 |
-
"eval_runtime": 16.3644,
|
| 472 |
-
"eval_samples_per_second": 18.577,
|
| 473 |
-
"eval_steps_per_second": 4.644,
|
| 474 |
-
"step": 310
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 2.14,
|
| 478 |
-
"learning_rate": 3e-05,
|
| 479 |
-
"loss": 0.1696,
|
| 480 |
-
"step": 320
|
| 481 |
-
},
|
| 482 |
-
{
|
| 483 |
-
"epoch": 2.14,
|
| 484 |
-
"eval_accuracy": 0.5296052631578947,
|
| 485 |
-
"eval_loss": 0.9087279438972473,
|
| 486 |
-
"eval_runtime": 16.362,
|
| 487 |
-
"eval_samples_per_second": 18.58,
|
| 488 |
-
"eval_steps_per_second": 4.645,
|
| 489 |
-
"step": 320
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 2.2,
|
| 493 |
-
"learning_rate": 3e-05,
|
| 494 |
-
"loss": 0.2255,
|
| 495 |
-
"step": 330
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 2.2,
|
| 499 |
"eval_accuracy": 0.5361842105263158,
|
| 500 |
-
"eval_loss": 0.
|
| 501 |
-
"eval_runtime": 16.
|
| 502 |
-
"eval_samples_per_second":
|
| 503 |
-
"eval_steps_per_second": 4.
|
| 504 |
-
"step":
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 2.27,
|
| 508 |
-
"learning_rate": 3e-05,
|
| 509 |
-
"loss": 0.1923,
|
| 510 |
-
"step": 340
|
| 511 |
-
},
|
| 512 |
-
{
|
| 513 |
-
"epoch": 2.27,
|
| 514 |
-
"eval_accuracy": 0.5197368421052632,
|
| 515 |
-
"eval_loss": 0.701989471912384,
|
| 516 |
-
"eval_runtime": 16.3692,
|
| 517 |
-
"eval_samples_per_second": 18.571,
|
| 518 |
-
"eval_steps_per_second": 4.643,
|
| 519 |
-
"step": 340
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 2.34,
|
| 523 |
-
"learning_rate": 3e-05,
|
| 524 |
-
"loss": 0.281,
|
| 525 |
-
"step": 350
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 2.34,
|
| 529 |
-
"eval_accuracy": 0.5032894736842105,
|
| 530 |
-
"eval_loss": 0.6653422117233276,
|
| 531 |
-
"eval_runtime": 16.3703,
|
| 532 |
-
"eval_samples_per_second": 18.57,
|
| 533 |
-
"eval_steps_per_second": 4.643,
|
| 534 |
-
"step": 350
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 2.4,
|
| 538 |
-
"learning_rate": 3e-05,
|
| 539 |
-
"loss": 0.2311,
|
| 540 |
-
"step": 360
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"epoch": 2.4,
|
| 544 |
-
"eval_accuracy": 0.5131578947368421,
|
| 545 |
-
"eval_loss": 0.6411683559417725,
|
| 546 |
-
"eval_runtime": 16.3702,
|
| 547 |
-
"eval_samples_per_second": 18.57,
|
| 548 |
-
"eval_steps_per_second": 4.643,
|
| 549 |
-
"step": 360
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 2.47,
|
| 553 |
-
"learning_rate": 3e-05,
|
| 554 |
-
"loss": 0.1523,
|
| 555 |
-
"step": 370
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 2.47,
|
| 559 |
-
"eval_accuracy": 0.5230263157894737,
|
| 560 |
-
"eval_loss": 0.8846220970153809,
|
| 561 |
-
"eval_runtime": 16.3728,
|
| 562 |
-
"eval_samples_per_second": 18.567,
|
| 563 |
-
"eval_steps_per_second": 4.642,
|
| 564 |
-
"step": 370
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 2.54,
|
| 568 |
-
"learning_rate": 3e-05,
|
| 569 |
-
"loss": 0.2451,
|
| 570 |
-
"step": 380
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 2.54,
|
| 574 |
-
"eval_accuracy": 0.5164473684210527,
|
| 575 |
-
"eval_loss": 0.9251819252967834,
|
| 576 |
-
"eval_runtime": 16.379,
|
| 577 |
-
"eval_samples_per_second": 18.56,
|
| 578 |
-
"eval_steps_per_second": 4.64,
|
| 579 |
-
"step": 380
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 2.6,
|
| 583 |
-
"learning_rate": 3e-05,
|
| 584 |
-
"loss": 0.2022,
|
| 585 |
-
"step": 390
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 2.6,
|
| 589 |
-
"eval_accuracy": 0.5197368421052632,
|
| 590 |
-
"eval_loss": 0.7422206401824951,
|
| 591 |
-
"eval_runtime": 16.3727,
|
| 592 |
-
"eval_samples_per_second": 18.567,
|
| 593 |
-
"eval_steps_per_second": 4.642,
|
| 594 |
-
"step": 390
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 2.67,
|
| 598 |
-
"learning_rate": 3e-05,
|
| 599 |
-
"loss": 0.217,
|
| 600 |
-
"step": 400
|
| 601 |
-
},
|
| 602 |
-
{
|
| 603 |
-
"epoch": 2.67,
|
| 604 |
-
"eval_accuracy": 0.5328947368421053,
|
| 605 |
-
"eval_loss": 0.7557851076126099,
|
| 606 |
-
"eval_runtime": 16.3638,
|
| 607 |
-
"eval_samples_per_second": 18.578,
|
| 608 |
-
"eval_steps_per_second": 4.644,
|
| 609 |
-
"step": 400
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 2.74,
|
| 613 |
-
"learning_rate": 3e-05,
|
| 614 |
-
"loss": 0.165,
|
| 615 |
-
"step": 410
|
| 616 |
-
},
|
| 617 |
-
{
|
| 618 |
-
"epoch": 2.74,
|
| 619 |
-
"eval_accuracy": 0.5427631578947368,
|
| 620 |
-
"eval_loss": 0.7846018075942993,
|
| 621 |
-
"eval_runtime": 16.3674,
|
| 622 |
-
"eval_samples_per_second": 18.574,
|
| 623 |
-
"eval_steps_per_second": 4.643,
|
| 624 |
-
"step": 410
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 2.8,
|
| 628 |
-
"learning_rate": 3e-05,
|
| 629 |
-
"loss": 0.2025,
|
| 630 |
-
"step": 420
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 2.8,
|
| 634 |
-
"eval_accuracy": 0.5230263157894737,
|
| 635 |
-
"eval_loss": 0.725389301776886,
|
| 636 |
-
"eval_runtime": 16.3775,
|
| 637 |
-
"eval_samples_per_second": 18.562,
|
| 638 |
-
"eval_steps_per_second": 4.641,
|
| 639 |
-
"step": 420
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 2.87,
|
| 643 |
-
"learning_rate": 3e-05,
|
| 644 |
-
"loss": 0.2201,
|
| 645 |
-
"step": 430
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 2.87,
|
| 649 |
-
"eval_accuracy": 0.5296052631578947,
|
| 650 |
-
"eval_loss": 0.6530900001525879,
|
| 651 |
-
"eval_runtime": 16.3694,
|
| 652 |
-
"eval_samples_per_second": 18.571,
|
| 653 |
-
"eval_steps_per_second": 4.643,
|
| 654 |
-
"step": 430
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch": 2.
|
| 658 |
-
"
|
| 659 |
-
"
|
| 660 |
-
"
|
| 661 |
-
|
| 662 |
-
|
| 663 |
-
"
|
| 664 |
-
"eval_accuracy": 0.5493421052631579,
|
| 665 |
-
"eval_loss": 0.7827179431915283,
|
| 666 |
-
"eval_runtime": 16.3682,
|
| 667 |
-
"eval_samples_per_second": 18.573,
|
| 668 |
-
"eval_steps_per_second": 4.643,
|
| 669 |
-
"step": 440
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 2.98,
|
| 673 |
-
"step": 447,
|
| 674 |
-
"total_flos": 1.09848374069035e+17,
|
| 675 |
-
"train_loss": 0.43748926956381573,
|
| 676 |
-
"train_runtime": 1886.9121,
|
| 677 |
-
"train_samples_per_second": 3.809,
|
| 678 |
-
"train_steps_per_second": 0.237
|
| 679 |
}
|
| 680 |
],
|
| 681 |
"logging_steps": 10,
|
| 682 |
-
"max_steps":
|
| 683 |
"num_input_tokens_seen": 0,
|
| 684 |
"num_train_epochs": 3,
|
| 685 |
-
"save_steps":
|
| 686 |
-
"total_flos":
|
| 687 |
"train_batch_size": 4,
|
| 688 |
"trial_name": null,
|
| 689 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.003338898163606,
|
| 5 |
"eval_steps": 10,
|
| 6 |
+
"global_step": 300,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.07,
|
| 13 |
"learning_rate": 3e-05,
|
| 14 |
+
"loss": 1.5245,
|
| 15 |
"step": 10
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.07,
|
| 19 |
+
"eval_accuracy": 0.3355263157894737,
|
| 20 |
+
"eval_loss": 0.6506821513175964,
|
| 21 |
+
"eval_runtime": 16.9243,
|
| 22 |
+
"eval_samples_per_second": 17.962,
|
| 23 |
+
"eval_steps_per_second": 4.491,
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.13,
|
| 28 |
"learning_rate": 3e-05,
|
| 29 |
+
"loss": 0.6666,
|
| 30 |
"step": 20
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.13,
|
| 34 |
+
"eval_accuracy": 0.3815789473684211,
|
| 35 |
+
"eval_loss": 0.6464425325393677,
|
| 36 |
+
"eval_runtime": 16.9195,
|
| 37 |
+
"eval_samples_per_second": 17.967,
|
| 38 |
+
"eval_steps_per_second": 4.492,
|
| 39 |
"step": 20
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.2,
|
| 43 |
"learning_rate": 3e-05,
|
| 44 |
+
"loss": 0.6527,
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.2,
|
| 49 |
+
"eval_accuracy": 0.3684210526315789,
|
| 50 |
+
"eval_loss": 0.6426967978477478,
|
| 51 |
+
"eval_runtime": 16.9282,
|
| 52 |
+
"eval_samples_per_second": 17.958,
|
| 53 |
+
"eval_steps_per_second": 4.49,
|
| 54 |
"step": 30
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.27,
|
| 58 |
"learning_rate": 3e-05,
|
| 59 |
+
"loss": 0.6168,
|
| 60 |
"step": 40
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 0.27,
|
| 64 |
+
"eval_accuracy": 0.3980263157894737,
|
| 65 |
+
"eval_loss": 0.6321499943733215,
|
| 66 |
+
"eval_runtime": 17.0092,
|
| 67 |
+
"eval_samples_per_second": 17.873,
|
| 68 |
+
"eval_steps_per_second": 4.468,
|
| 69 |
"step": 40
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.33,
|
| 73 |
"learning_rate": 3e-05,
|
| 74 |
+
"loss": 0.6584,
|
| 75 |
"step": 50
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 0.33,
|
| 79 |
+
"eval_accuracy": 0.39144736842105265,
|
| 80 |
+
"eval_loss": 0.6181844472885132,
|
| 81 |
+
"eval_runtime": 16.9419,
|
| 82 |
+
"eval_samples_per_second": 17.944,
|
| 83 |
+
"eval_steps_per_second": 4.486,
|
| 84 |
"step": 50
|
| 85 |
},
|
| 86 |
{
|
|
|
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"epoch": 0.4,
|
| 94 |
+
"eval_accuracy": 0.4144736842105263,
|
| 95 |
+
"eval_loss": 0.6244160532951355,
|
| 96 |
+
"eval_runtime": 16.9269,
|
| 97 |
+
"eval_samples_per_second": 17.96,
|
| 98 |
+
"eval_steps_per_second": 4.49,
|
| 99 |
"step": 60
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.47,
|
| 103 |
"learning_rate": 3e-05,
|
| 104 |
+
"loss": 0.5924,
|
| 105 |
"step": 70
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 0.47,
|
| 109 |
+
"eval_accuracy": 0.4342105263157895,
|
| 110 |
+
"eval_loss": 0.6033625602722168,
|
| 111 |
+
"eval_runtime": 16.9236,
|
| 112 |
+
"eval_samples_per_second": 17.963,
|
| 113 |
+
"eval_steps_per_second": 4.491,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.53,
|
| 118 |
"learning_rate": 3e-05,
|
| 119 |
+
"loss": 0.6069,
|
| 120 |
"step": 80
|
| 121 |
},
|
| 122 |
{
|
| 123 |
"epoch": 0.53,
|
| 124 |
+
"eval_accuracy": 0.4375,
|
| 125 |
+
"eval_loss": 0.6096391677856445,
|
| 126 |
+
"eval_runtime": 16.9238,
|
| 127 |
+
"eval_samples_per_second": 17.963,
|
| 128 |
+
"eval_steps_per_second": 4.491,
|
| 129 |
"step": 80
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.6,
|
| 133 |
"learning_rate": 3e-05,
|
| 134 |
+
"loss": 0.5999,
|
| 135 |
"step": 90
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.6,
|
| 139 |
+
"eval_accuracy": 0.4407894736842105,
|
| 140 |
+
"eval_loss": 0.6095999479293823,
|
| 141 |
+
"eval_runtime": 16.9342,
|
| 142 |
+
"eval_samples_per_second": 17.952,
|
| 143 |
+
"eval_steps_per_second": 4.488,
|
| 144 |
"step": 90
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.67,
|
| 148 |
"learning_rate": 3e-05,
|
| 149 |
+
"loss": 0.6206,
|
| 150 |
"step": 100
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.67,
|
| 154 |
+
"eval_accuracy": 0.45723684210526316,
|
| 155 |
+
"eval_loss": 0.607021152973175,
|
| 156 |
+
"eval_runtime": 16.9304,
|
| 157 |
+
"eval_samples_per_second": 17.956,
|
| 158 |
+
"eval_steps_per_second": 4.489,
|
| 159 |
"step": 100
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 0.73,
|
| 163 |
"learning_rate": 3e-05,
|
| 164 |
+
"loss": 0.5793,
|
| 165 |
"step": 110
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.73,
|
| 169 |
+
"eval_accuracy": 0.45723684210526316,
|
| 170 |
+
"eval_loss": 0.601601243019104,
|
| 171 |
+
"eval_runtime": 16.9375,
|
| 172 |
+
"eval_samples_per_second": 17.948,
|
| 173 |
+
"eval_steps_per_second": 4.487,
|
| 174 |
"step": 110
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 0.8,
|
| 178 |
"learning_rate": 3e-05,
|
| 179 |
+
"loss": 0.6208,
|
| 180 |
"step": 120
|
| 181 |
},
|
| 182 |
{
|
| 183 |
"epoch": 0.8,
|
| 184 |
+
"eval_accuracy": 0.4605263157894737,
|
| 185 |
+
"eval_loss": 0.5902404189109802,
|
| 186 |
+
"eval_runtime": 16.924,
|
| 187 |
+
"eval_samples_per_second": 17.963,
|
| 188 |
+
"eval_steps_per_second": 4.491,
|
| 189 |
"step": 120
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.87,
|
| 193 |
"learning_rate": 3e-05,
|
| 194 |
+
"loss": 0.5622,
|
| 195 |
"step": 130
|
| 196 |
},
|
| 197 |
{
|
| 198 |
"epoch": 0.87,
|
| 199 |
+
"eval_accuracy": 0.4769736842105263,
|
| 200 |
+
"eval_loss": 0.5775408744812012,
|
| 201 |
+
"eval_runtime": 16.9329,
|
| 202 |
+
"eval_samples_per_second": 17.953,
|
| 203 |
+
"eval_steps_per_second": 4.488,
|
| 204 |
"step": 130
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 0.93,
|
| 208 |
"learning_rate": 3e-05,
|
| 209 |
+
"loss": 0.5502,
|
| 210 |
"step": 140
|
| 211 |
},
|
| 212 |
{
|
| 213 |
"epoch": 0.93,
|
| 214 |
+
"eval_accuracy": 0.46710526315789475,
|
| 215 |
+
"eval_loss": 0.57607102394104,
|
| 216 |
+
"eval_runtime": 16.9226,
|
| 217 |
+
"eval_samples_per_second": 17.964,
|
| 218 |
+
"eval_steps_per_second": 4.491,
|
| 219 |
"step": 140
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 1.0,
|
| 223 |
"learning_rate": 3e-05,
|
| 224 |
+
"loss": 0.5958,
|
| 225 |
"step": 150
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 1.0,
|
| 229 |
+
"eval_accuracy": 0.4901315789473684,
|
| 230 |
+
"eval_loss": 0.5606401562690735,
|
| 231 |
+
"eval_runtime": 16.929,
|
| 232 |
+
"eval_samples_per_second": 17.957,
|
| 233 |
+
"eval_steps_per_second": 4.489,
|
| 234 |
"step": 150
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 1.07,
|
| 238 |
"learning_rate": 3e-05,
|
| 239 |
+
"loss": 0.4558,
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 1.07,
|
| 244 |
+
"eval_accuracy": 0.47368421052631576,
|
| 245 |
+
"eval_loss": 0.5839833617210388,
|
| 246 |
+
"eval_runtime": 16.9304,
|
| 247 |
+
"eval_samples_per_second": 17.956,
|
| 248 |
+
"eval_steps_per_second": 4.489,
|
| 249 |
"step": 160
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 1.14,
|
| 253 |
"learning_rate": 3e-05,
|
| 254 |
+
"loss": 0.4411,
|
| 255 |
"step": 170
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 1.14,
|
| 259 |
+
"eval_accuracy": 0.4901315789473684,
|
| 260 |
+
"eval_loss": 0.5631235837936401,
|
| 261 |
+
"eval_runtime": 16.9238,
|
| 262 |
+
"eval_samples_per_second": 17.963,
|
| 263 |
+
"eval_steps_per_second": 4.491,
|
| 264 |
"step": 170
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 1.2,
|
| 268 |
"learning_rate": 3e-05,
|
| 269 |
+
"loss": 0.4144,
|
| 270 |
"step": 180
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 1.2,
|
| 274 |
+
"eval_accuracy": 0.5,
|
| 275 |
+
"eval_loss": 0.5744868516921997,
|
| 276 |
+
"eval_runtime": 16.9382,
|
| 277 |
+
"eval_samples_per_second": 17.948,
|
| 278 |
+
"eval_steps_per_second": 4.487,
|
| 279 |
"step": 180
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 1.27,
|
| 283 |
"learning_rate": 3e-05,
|
| 284 |
+
"loss": 0.4647,
|
| 285 |
"step": 190
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 1.27,
|
| 289 |
+
"eval_accuracy": 0.4605263157894737,
|
| 290 |
+
"eval_loss": 0.593177080154419,
|
| 291 |
+
"eval_runtime": 16.932,
|
| 292 |
+
"eval_samples_per_second": 17.954,
|
| 293 |
+
"eval_steps_per_second": 4.489,
|
| 294 |
"step": 190
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 1.34,
|
| 298 |
"learning_rate": 3e-05,
|
| 299 |
+
"loss": 0.4504,
|
| 300 |
"step": 200
|
| 301 |
},
|
| 302 |
{
|
| 303 |
"epoch": 1.34,
|
| 304 |
"eval_accuracy": 0.5098684210526315,
|
| 305 |
+
"eval_loss": 0.5798581838607788,
|
| 306 |
+
"eval_runtime": 16.9337,
|
| 307 |
+
"eval_samples_per_second": 17.952,
|
| 308 |
+
"eval_steps_per_second": 4.488,
|
| 309 |
"step": 200
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 1.4,
|
| 313 |
"learning_rate": 3e-05,
|
| 314 |
+
"loss": 0.4299,
|
| 315 |
"step": 210
|
| 316 |
},
|
| 317 |
{
|
| 318 |
"epoch": 1.4,
|
| 319 |
+
"eval_accuracy": 0.4934210526315789,
|
| 320 |
+
"eval_loss": 0.64882493019104,
|
| 321 |
+
"eval_runtime": 16.9391,
|
| 322 |
+
"eval_samples_per_second": 17.947,
|
| 323 |
+
"eval_steps_per_second": 4.487,
|
| 324 |
"step": 210
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 1.47,
|
| 328 |
"learning_rate": 3e-05,
|
| 329 |
+
"loss": 0.425,
|
| 330 |
"step": 220
|
| 331 |
},
|
| 332 |
{
|
| 333 |
"epoch": 1.47,
|
| 334 |
+
"eval_accuracy": 0.5131578947368421,
|
| 335 |
+
"eval_loss": 0.5704348683357239,
|
| 336 |
+
"eval_runtime": 16.9325,
|
| 337 |
+
"eval_samples_per_second": 17.954,
|
| 338 |
+
"eval_steps_per_second": 4.488,
|
| 339 |
"step": 220
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.54,
|
| 343 |
"learning_rate": 3e-05,
|
| 344 |
+
"loss": 0.4152,
|
| 345 |
"step": 230
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 1.54,
|
| 349 |
+
"eval_accuracy": 0.506578947368421,
|
| 350 |
+
"eval_loss": 0.5582014322280884,
|
| 351 |
+
"eval_runtime": 16.9258,
|
| 352 |
+
"eval_samples_per_second": 17.961,
|
| 353 |
+
"eval_steps_per_second": 4.49,
|
| 354 |
"step": 230
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 1.6,
|
| 358 |
"learning_rate": 3e-05,
|
| 359 |
+
"loss": 0.425,
|
| 360 |
"step": 240
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.6,
|
| 364 |
+
"eval_accuracy": 0.5328947368421053,
|
| 365 |
+
"eval_loss": 0.5488855838775635,
|
| 366 |
+
"eval_runtime": 16.9288,
|
| 367 |
+
"eval_samples_per_second": 17.958,
|
| 368 |
+
"eval_steps_per_second": 4.489,
|
| 369 |
"step": 240
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 1.67,
|
| 373 |
"learning_rate": 3e-05,
|
| 374 |
+
"loss": 0.446,
|
| 375 |
"step": 250
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 1.67,
|
| 379 |
+
"eval_accuracy": 0.5197368421052632,
|
| 380 |
+
"eval_loss": 0.5479023456573486,
|
| 381 |
+
"eval_runtime": 16.9319,
|
| 382 |
+
"eval_samples_per_second": 17.954,
|
| 383 |
+
"eval_steps_per_second": 4.489,
|
| 384 |
"step": 250
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 1.74,
|
| 388 |
"learning_rate": 3e-05,
|
| 389 |
+
"loss": 0.3908,
|
| 390 |
"step": 260
|
| 391 |
},
|
| 392 |
{
|
| 393 |
"epoch": 1.74,
|
| 394 |
+
"eval_accuracy": 0.5164473684210527,
|
| 395 |
+
"eval_loss": 0.5564107894897461,
|
| 396 |
+
"eval_runtime": 16.9414,
|
| 397 |
+
"eval_samples_per_second": 17.944,
|
| 398 |
+
"eval_steps_per_second": 4.486,
|
| 399 |
"step": 260
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 1.8,
|
| 403 |
"learning_rate": 3e-05,
|
| 404 |
+
"loss": 0.443,
|
| 405 |
"step": 270
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 1.8,
|
| 409 |
+
"eval_accuracy": 0.5032894736842105,
|
| 410 |
+
"eval_loss": 0.5418796539306641,
|
| 411 |
+
"eval_runtime": 16.9208,
|
| 412 |
+
"eval_samples_per_second": 17.966,
|
| 413 |
+
"eval_steps_per_second": 4.492,
|
| 414 |
"step": 270
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 1.87,
|
| 418 |
"learning_rate": 3e-05,
|
| 419 |
+
"loss": 0.4081,
|
| 420 |
"step": 280
|
| 421 |
},
|
| 422 |
{
|
| 423 |
"epoch": 1.87,
|
| 424 |
"eval_accuracy": 0.506578947368421,
|
| 425 |
+
"eval_loss": 0.5948407053947449,
|
| 426 |
+
"eval_runtime": 16.9289,
|
| 427 |
+
"eval_samples_per_second": 17.957,
|
| 428 |
+
"eval_steps_per_second": 4.489,
|
| 429 |
"step": 280
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 1.94,
|
| 433 |
"learning_rate": 3e-05,
|
| 434 |
+
"loss": 0.3944,
|
| 435 |
"step": 290
|
| 436 |
},
|
| 437 |
{
|
| 438 |
"epoch": 1.94,
|
| 439 |
+
"eval_accuracy": 0.5394736842105263,
|
| 440 |
+
"eval_loss": 0.554680347442627,
|
| 441 |
+
"eval_runtime": 16.9311,
|
| 442 |
+
"eval_samples_per_second": 17.955,
|
| 443 |
+
"eval_steps_per_second": 4.489,
|
| 444 |
"step": 290
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 2.0,
|
| 448 |
"learning_rate": 3e-05,
|
| 449 |
+
"loss": 0.4005,
|
| 450 |
"step": 300
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 2.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 454 |
"eval_accuracy": 0.5361842105263158,
|
| 455 |
+
"eval_loss": 0.5615983009338379,
|
| 456 |
+
"eval_runtime": 16.9277,
|
| 457 |
+
"eval_samples_per_second": 17.959,
|
| 458 |
+
"eval_steps_per_second": 4.49,
|
| 459 |
+
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 2.0,
|
| 463 |
+
"step": 300,
|
| 464 |
+
"total_flos": 7.380159778455552e+16,
|
| 465 |
+
"train_loss": 0.5479170862833659,
|
| 466 |
+
"train_runtime": 1462.1781,
|
| 467 |
+
"train_samples_per_second": 3.283,
|
| 468 |
+
"train_steps_per_second": 0.205
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 469 |
}
|
| 470 |
],
|
| 471 |
"logging_steps": 10,
|
| 472 |
+
"max_steps": 300,
|
| 473 |
"num_input_tokens_seen": 0,
|
| 474 |
"num_train_epochs": 3,
|
| 475 |
+
"save_steps": 150,
|
| 476 |
+
"total_flos": 7.380159778455552e+16,
|
| 477 |
"train_batch_size": 4,
|
| 478 |
"trial_name": null,
|
| 479 |
"trial_params": null
|