Fin de l'entraînement
Browse files- README.md +3 -3
- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +609 -0
README.md
CHANGED
|
@@ -14,9 +14,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
-
- Loss: 0.
|
| 18 |
-
- Mse: 0.
|
| 19 |
-
- Mae: 0.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 0.0531
|
| 18 |
+
- Mse: 0.1291
|
| 19 |
+
- Mae: 0.3119
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 29.985507246376812,
|
| 3 |
+
"eval_loss": 0.05310577526688576,
|
| 4 |
+
"eval_mae": 0.3118866980075836,
|
| 5 |
+
"eval_mse": 0.12907913327217102,
|
| 6 |
+
"eval_runtime": 58.3393,
|
| 7 |
+
"eval_samples_per_second": 7.096,
|
| 8 |
+
"eval_steps_per_second": 0.891,
|
| 9 |
+
"total_flos": 0.0,
|
| 10 |
+
"train_loss": 0.11046674571006126,
|
| 11 |
+
"train_runtime": 98364.9673,
|
| 12 |
+
"train_samples_per_second": 0.504,
|
| 13 |
+
"train_steps_per_second": 0.016
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 29.985507246376812,
|
| 3 |
+
"eval_loss": 0.05310577526688576,
|
| 4 |
+
"eval_mae": 0.3118866980075836,
|
| 5 |
+
"eval_mse": 0.12907913327217102,
|
| 6 |
+
"eval_runtime": 58.3393,
|
| 7 |
+
"eval_samples_per_second": 7.096,
|
| 8 |
+
"eval_steps_per_second": 0.891
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 29.985507246376812,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.11046674571006126,
|
| 5 |
+
"train_runtime": 98364.9673,
|
| 6 |
+
"train_samples_per_second": 0.504,
|
| 7 |
+
"train_steps_per_second": 0.016
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,609 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 969,
|
| 3 |
+
"best_metric": 0.12907913327217102,
|
| 4 |
+
"best_model_checkpoint": "./vit_focus_full/checkpoint-969",
|
| 5 |
+
"epoch": 29.985507246376812,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1530,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.7729468599033816,
|
| 14 |
+
"grad_norm": 10.6144437789917,
|
| 15 |
+
"learning_rate": 4.872549019607843e-05,
|
| 16 |
+
"loss": 0.3146,
|
| 17 |
+
"step": 40
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.9855072463768116,
|
| 21 |
+
"eval_loss": 0.05953465402126312,
|
| 22 |
+
"eval_mae": 0.3265259861946106,
|
| 23 |
+
"eval_mse": 0.1403445601463318,
|
| 24 |
+
"eval_runtime": 57.1346,
|
| 25 |
+
"eval_samples_per_second": 7.246,
|
| 26 |
+
"eval_steps_per_second": 0.91,
|
| 27 |
+
"step": 51
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 1.5603864734299517,
|
| 31 |
+
"grad_norm": 5.5437469482421875,
|
| 32 |
+
"learning_rate": 4.741830065359477e-05,
|
| 33 |
+
"loss": 0.2488,
|
| 34 |
+
"step": 80
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 1.9855072463768115,
|
| 38 |
+
"eval_loss": 0.05661279708147049,
|
| 39 |
+
"eval_mae": 0.3253043293952942,
|
| 40 |
+
"eval_mse": 0.13950611650943756,
|
| 41 |
+
"eval_runtime": 56.4199,
|
| 42 |
+
"eval_samples_per_second": 7.338,
|
| 43 |
+
"eval_steps_per_second": 0.922,
|
| 44 |
+
"step": 102
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.3478260869565215,
|
| 48 |
+
"grad_norm": 4.464972972869873,
|
| 49 |
+
"learning_rate": 4.6111111111111115e-05,
|
| 50 |
+
"loss": 0.2278,
|
| 51 |
+
"step": 120
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 2.9855072463768115,
|
| 55 |
+
"eval_loss": 0.06113344058394432,
|
| 56 |
+
"eval_mae": 0.3287981450557709,
|
| 57 |
+
"eval_mse": 0.14261041581630707,
|
| 58 |
+
"eval_runtime": 56.9728,
|
| 59 |
+
"eval_samples_per_second": 7.267,
|
| 60 |
+
"eval_steps_per_second": 0.913,
|
| 61 |
+
"step": 153
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 3.135265700483092,
|
| 65 |
+
"grad_norm": 8.567197799682617,
|
| 66 |
+
"learning_rate": 4.480392156862745e-05,
|
| 67 |
+
"loss": 0.2148,
|
| 68 |
+
"step": 160
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 3.9082125603864735,
|
| 72 |
+
"grad_norm": 5.523195266723633,
|
| 73 |
+
"learning_rate": 4.3496732026143795e-05,
|
| 74 |
+
"loss": 0.206,
|
| 75 |
+
"step": 200
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 3.9855072463768115,
|
| 79 |
+
"eval_loss": 0.05355362221598625,
|
| 80 |
+
"eval_mae": 0.31797775626182556,
|
| 81 |
+
"eval_mse": 0.13227654993534088,
|
| 82 |
+
"eval_runtime": 56.8583,
|
| 83 |
+
"eval_samples_per_second": 7.281,
|
| 84 |
+
"eval_steps_per_second": 0.915,
|
| 85 |
+
"step": 204
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 4.695652173913043,
|
| 89 |
+
"grad_norm": 6.00140380859375,
|
| 90 |
+
"learning_rate": 4.218954248366013e-05,
|
| 91 |
+
"loss": 0.1902,
|
| 92 |
+
"step": 240
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 4.9855072463768115,
|
| 96 |
+
"eval_loss": 0.06186218187212944,
|
| 97 |
+
"eval_mae": 0.3270839750766754,
|
| 98 |
+
"eval_mse": 0.1410592794418335,
|
| 99 |
+
"eval_runtime": 56.1466,
|
| 100 |
+
"eval_samples_per_second": 7.374,
|
| 101 |
+
"eval_steps_per_second": 0.926,
|
| 102 |
+
"step": 255
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 5.483091787439614,
|
| 106 |
+
"grad_norm": 9.328702926635742,
|
| 107 |
+
"learning_rate": 4.0882352941176474e-05,
|
| 108 |
+
"loss": 0.187,
|
| 109 |
+
"step": 280
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 5.9855072463768115,
|
| 113 |
+
"eval_loss": 0.05080530419945717,
|
| 114 |
+
"eval_mae": 0.3168753385543823,
|
| 115 |
+
"eval_mse": 0.1319676637649536,
|
| 116 |
+
"eval_runtime": 57.4612,
|
| 117 |
+
"eval_samples_per_second": 7.205,
|
| 118 |
+
"eval_steps_per_second": 0.905,
|
| 119 |
+
"step": 306
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 6.270531400966184,
|
| 123 |
+
"grad_norm": 7.799366474151611,
|
| 124 |
+
"learning_rate": 3.957516339869281e-05,
|
| 125 |
+
"loss": 0.1757,
|
| 126 |
+
"step": 320
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 6.9855072463768115,
|
| 130 |
+
"eval_loss": 0.05371123179793358,
|
| 131 |
+
"eval_mae": 0.31825557351112366,
|
| 132 |
+
"eval_mse": 0.13387194275856018,
|
| 133 |
+
"eval_runtime": 57.3782,
|
| 134 |
+
"eval_samples_per_second": 7.215,
|
| 135 |
+
"eval_steps_per_second": 0.906,
|
| 136 |
+
"step": 357
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 7.057971014492754,
|
| 140 |
+
"grad_norm": 4.06664514541626,
|
| 141 |
+
"learning_rate": 3.8267973856209146e-05,
|
| 142 |
+
"loss": 0.1677,
|
| 143 |
+
"step": 360
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 7.830917874396135,
|
| 147 |
+
"grad_norm": 5.403101921081543,
|
| 148 |
+
"learning_rate": 3.6960784313725496e-05,
|
| 149 |
+
"loss": 0.1523,
|
| 150 |
+
"step": 400
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 7.9855072463768115,
|
| 154 |
+
"eval_loss": 0.055755238980054855,
|
| 155 |
+
"eval_mae": 0.31683334708213806,
|
| 156 |
+
"eval_mse": 0.13297995924949646,
|
| 157 |
+
"eval_runtime": 65.7904,
|
| 158 |
+
"eval_samples_per_second": 6.293,
|
| 159 |
+
"eval_steps_per_second": 0.79,
|
| 160 |
+
"step": 408
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 8.618357487922705,
|
| 164 |
+
"grad_norm": 6.7577948570251465,
|
| 165 |
+
"learning_rate": 3.565359477124183e-05,
|
| 166 |
+
"loss": 0.1528,
|
| 167 |
+
"step": 440
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 8.985507246376812,
|
| 171 |
+
"eval_loss": 0.05914789438247681,
|
| 172 |
+
"eval_mae": 0.3224806785583496,
|
| 173 |
+
"eval_mse": 0.1381232738494873,
|
| 174 |
+
"eval_runtime": 56.9831,
|
| 175 |
+
"eval_samples_per_second": 7.265,
|
| 176 |
+
"eval_steps_per_second": 0.913,
|
| 177 |
+
"step": 459
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 9.405797101449275,
|
| 181 |
+
"grad_norm": 4.654517650604248,
|
| 182 |
+
"learning_rate": 3.434640522875817e-05,
|
| 183 |
+
"loss": 0.1416,
|
| 184 |
+
"step": 480
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 9.985507246376812,
|
| 188 |
+
"eval_loss": 0.05355934053659439,
|
| 189 |
+
"eval_mae": 0.3197546601295471,
|
| 190 |
+
"eval_mse": 0.1352616846561432,
|
| 191 |
+
"eval_runtime": 57.4136,
|
| 192 |
+
"eval_samples_per_second": 7.211,
|
| 193 |
+
"eval_steps_per_second": 0.906,
|
| 194 |
+
"step": 510
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 10.193236714975846,
|
| 198 |
+
"grad_norm": 4.063232421875,
|
| 199 |
+
"learning_rate": 3.303921568627451e-05,
|
| 200 |
+
"loss": 0.1391,
|
| 201 |
+
"step": 520
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 10.966183574879228,
|
| 205 |
+
"grad_norm": 4.905858993530273,
|
| 206 |
+
"learning_rate": 3.173202614379085e-05,
|
| 207 |
+
"loss": 0.1298,
|
| 208 |
+
"step": 560
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 10.985507246376812,
|
| 212 |
+
"eval_loss": 0.05300646275281906,
|
| 213 |
+
"eval_mae": 0.3164079189300537,
|
| 214 |
+
"eval_mse": 0.132521390914917,
|
| 215 |
+
"eval_runtime": 58.995,
|
| 216 |
+
"eval_samples_per_second": 7.018,
|
| 217 |
+
"eval_steps_per_second": 0.881,
|
| 218 |
+
"step": 561
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 11.753623188405797,
|
| 222 |
+
"grad_norm": 4.643632411956787,
|
| 223 |
+
"learning_rate": 3.0424836601307187e-05,
|
| 224 |
+
"loss": 0.1161,
|
| 225 |
+
"step": 600
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 11.985507246376812,
|
| 229 |
+
"eval_loss": 0.0511205680668354,
|
| 230 |
+
"eval_mae": 0.315570205450058,
|
| 231 |
+
"eval_mse": 0.13146661221981049,
|
| 232 |
+
"eval_runtime": 57.0447,
|
| 233 |
+
"eval_samples_per_second": 7.257,
|
| 234 |
+
"eval_steps_per_second": 0.912,
|
| 235 |
+
"step": 612
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 12.541062801932368,
|
| 239 |
+
"grad_norm": 3.0849831104278564,
|
| 240 |
+
"learning_rate": 2.9117647058823534e-05,
|
| 241 |
+
"loss": 0.1085,
|
| 242 |
+
"step": 640
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 12.985507246376812,
|
| 246 |
+
"eval_loss": 0.05314180254936218,
|
| 247 |
+
"eval_mae": 0.32430657744407654,
|
| 248 |
+
"eval_mse": 0.13849547505378723,
|
| 249 |
+
"eval_runtime": 631.8234,
|
| 250 |
+
"eval_samples_per_second": 0.655,
|
| 251 |
+
"eval_steps_per_second": 0.082,
|
| 252 |
+
"step": 663
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 13.328502415458937,
|
| 256 |
+
"grad_norm": 5.586836338043213,
|
| 257 |
+
"learning_rate": 2.7810457516339873e-05,
|
| 258 |
+
"loss": 0.1028,
|
| 259 |
+
"step": 680
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 13.985507246376812,
|
| 263 |
+
"eval_loss": 0.05296429246664047,
|
| 264 |
+
"eval_mae": 0.31508708000183105,
|
| 265 |
+
"eval_mse": 0.1316087543964386,
|
| 266 |
+
"eval_runtime": 57.8458,
|
| 267 |
+
"eval_samples_per_second": 7.157,
|
| 268 |
+
"eval_steps_per_second": 0.899,
|
| 269 |
+
"step": 714
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 14.115942028985508,
|
| 273 |
+
"grad_norm": 3.5024545192718506,
|
| 274 |
+
"learning_rate": 2.650326797385621e-05,
|
| 275 |
+
"loss": 0.0974,
|
| 276 |
+
"step": 720
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 14.88888888888889,
|
| 280 |
+
"grad_norm": 3.7782580852508545,
|
| 281 |
+
"learning_rate": 2.519607843137255e-05,
|
| 282 |
+
"loss": 0.0891,
|
| 283 |
+
"step": 760
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 14.985507246376812,
|
| 287 |
+
"eval_loss": 0.0540492981672287,
|
| 288 |
+
"eval_mae": 0.31779569387435913,
|
| 289 |
+
"eval_mse": 0.1337898075580597,
|
| 290 |
+
"eval_runtime": 57.6717,
|
| 291 |
+
"eval_samples_per_second": 7.179,
|
| 292 |
+
"eval_steps_per_second": 0.902,
|
| 293 |
+
"step": 765
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 15.676328502415458,
|
| 297 |
+
"grad_norm": 3.615967035293579,
|
| 298 |
+
"learning_rate": 2.3888888888888892e-05,
|
| 299 |
+
"loss": 0.0878,
|
| 300 |
+
"step": 800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 15.985507246376812,
|
| 304 |
+
"eval_loss": 0.05357988178730011,
|
| 305 |
+
"eval_mae": 0.3177140951156616,
|
| 306 |
+
"eval_mse": 0.13350851833820343,
|
| 307 |
+
"eval_runtime": 57.5097,
|
| 308 |
+
"eval_samples_per_second": 7.199,
|
| 309 |
+
"eval_steps_per_second": 0.904,
|
| 310 |
+
"step": 816
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 16.463768115942027,
|
| 314 |
+
"grad_norm": 9.533724784851074,
|
| 315 |
+
"learning_rate": 2.258169934640523e-05,
|
| 316 |
+
"loss": 0.077,
|
| 317 |
+
"step": 840
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 16.985507246376812,
|
| 321 |
+
"eval_loss": 0.05338989570736885,
|
| 322 |
+
"eval_mae": 0.31321439146995544,
|
| 323 |
+
"eval_mse": 0.12988565862178802,
|
| 324 |
+
"eval_runtime": 58.1505,
|
| 325 |
+
"eval_samples_per_second": 7.119,
|
| 326 |
+
"eval_steps_per_second": 0.894,
|
| 327 |
+
"step": 867
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 17.2512077294686,
|
| 331 |
+
"grad_norm": 3.7093381881713867,
|
| 332 |
+
"learning_rate": 2.1274509803921568e-05,
|
| 333 |
+
"loss": 0.0769,
|
| 334 |
+
"step": 880
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 17.985507246376812,
|
| 338 |
+
"eval_loss": 0.0548611618578434,
|
| 339 |
+
"eval_mae": 0.3149418532848358,
|
| 340 |
+
"eval_mse": 0.1313086301088333,
|
| 341 |
+
"eval_runtime": 56.4832,
|
| 342 |
+
"eval_samples_per_second": 7.33,
|
| 343 |
+
"eval_steps_per_second": 0.921,
|
| 344 |
+
"step": 918
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 18.03864734299517,
|
| 348 |
+
"grad_norm": 2.9852871894836426,
|
| 349 |
+
"learning_rate": 1.996732026143791e-05,
|
| 350 |
+
"loss": 0.0717,
|
| 351 |
+
"step": 920
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 18.81159420289855,
|
| 355 |
+
"grad_norm": 3.3752264976501465,
|
| 356 |
+
"learning_rate": 1.866013071895425e-05,
|
| 357 |
+
"loss": 0.0663,
|
| 358 |
+
"step": 960
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 18.985507246376812,
|
| 362 |
+
"eval_loss": 0.05310577526688576,
|
| 363 |
+
"eval_mae": 0.3118866980075836,
|
| 364 |
+
"eval_mse": 0.12907913327217102,
|
| 365 |
+
"eval_runtime": 58.2255,
|
| 366 |
+
"eval_samples_per_second": 7.11,
|
| 367 |
+
"eval_steps_per_second": 0.893,
|
| 368 |
+
"step": 969
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 19.59903381642512,
|
| 372 |
+
"grad_norm": 2.9139506816864014,
|
| 373 |
+
"learning_rate": 1.735294117647059e-05,
|
| 374 |
+
"loss": 0.064,
|
| 375 |
+
"step": 1000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 19.985507246376812,
|
| 379 |
+
"eval_loss": 0.05400167778134346,
|
| 380 |
+
"eval_mae": 0.31967055797576904,
|
| 381 |
+
"eval_mse": 0.13520964980125427,
|
| 382 |
+
"eval_runtime": 58.0572,
|
| 383 |
+
"eval_samples_per_second": 7.131,
|
| 384 |
+
"eval_steps_per_second": 0.896,
|
| 385 |
+
"step": 1020
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 20.386473429951693,
|
| 389 |
+
"grad_norm": 3.1011509895324707,
|
| 390 |
+
"learning_rate": 1.604575163398693e-05,
|
| 391 |
+
"loss": 0.0608,
|
| 392 |
+
"step": 1040
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 20.985507246376812,
|
| 396 |
+
"eval_loss": 0.05348004400730133,
|
| 397 |
+
"eval_mae": 0.3179128170013428,
|
| 398 |
+
"eval_mse": 0.13336069881916046,
|
| 399 |
+
"eval_runtime": 56.8284,
|
| 400 |
+
"eval_samples_per_second": 7.285,
|
| 401 |
+
"eval_steps_per_second": 0.915,
|
| 402 |
+
"step": 1071
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 21.17391304347826,
|
| 406 |
+
"grad_norm": 2.4269816875457764,
|
| 407 |
+
"learning_rate": 1.473856209150327e-05,
|
| 408 |
+
"loss": 0.0558,
|
| 409 |
+
"step": 1080
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 21.946859903381643,
|
| 413 |
+
"grad_norm": 2.612093925476074,
|
| 414 |
+
"learning_rate": 1.3431372549019607e-05,
|
| 415 |
+
"loss": 0.0548,
|
| 416 |
+
"step": 1120
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 21.985507246376812,
|
| 420 |
+
"eval_loss": 0.052902594208717346,
|
| 421 |
+
"eval_mae": 0.3134055733680725,
|
| 422 |
+
"eval_mse": 0.129911869764328,
|
| 423 |
+
"eval_runtime": 57.5407,
|
| 424 |
+
"eval_samples_per_second": 7.195,
|
| 425 |
+
"eval_steps_per_second": 0.904,
|
| 426 |
+
"step": 1122
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 22.734299516908212,
|
| 430 |
+
"grad_norm": 1.7072349786758423,
|
| 431 |
+
"learning_rate": 1.2124183006535949e-05,
|
| 432 |
+
"loss": 0.0517,
|
| 433 |
+
"step": 1160
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 22.985507246376812,
|
| 437 |
+
"eval_loss": 0.05338846519589424,
|
| 438 |
+
"eval_mae": 0.31519371271133423,
|
| 439 |
+
"eval_mse": 0.13099054992198944,
|
| 440 |
+
"eval_runtime": 2988.6114,
|
| 441 |
+
"eval_samples_per_second": 0.139,
|
| 442 |
+
"eval_steps_per_second": 0.017,
|
| 443 |
+
"step": 1173
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 23.52173913043478,
|
| 447 |
+
"grad_norm": 2.942000389099121,
|
| 448 |
+
"learning_rate": 1.0816993464052288e-05,
|
| 449 |
+
"loss": 0.0498,
|
| 450 |
+
"step": 1200
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 23.985507246376812,
|
| 454 |
+
"eval_loss": 0.05435283109545708,
|
| 455 |
+
"eval_mae": 0.31506991386413574,
|
| 456 |
+
"eval_mse": 0.13137240707874298,
|
| 457 |
+
"eval_runtime": 158.629,
|
| 458 |
+
"eval_samples_per_second": 2.61,
|
| 459 |
+
"eval_steps_per_second": 0.328,
|
| 460 |
+
"step": 1224
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 24.309178743961354,
|
| 464 |
+
"grad_norm": 1.7872236967086792,
|
| 465 |
+
"learning_rate": 9.509803921568628e-06,
|
| 466 |
+
"loss": 0.047,
|
| 467 |
+
"step": 1240
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 24.985507246376812,
|
| 471 |
+
"eval_loss": 0.05310087278485298,
|
| 472 |
+
"eval_mae": 0.3145076036453247,
|
| 473 |
+
"eval_mse": 0.13092052936553955,
|
| 474 |
+
"eval_runtime": 59.2601,
|
| 475 |
+
"eval_samples_per_second": 6.986,
|
| 476 |
+
"eval_steps_per_second": 0.877,
|
| 477 |
+
"step": 1275
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 25.096618357487923,
|
| 481 |
+
"grad_norm": 1.8146392107009888,
|
| 482 |
+
"learning_rate": 8.202614379084967e-06,
|
| 483 |
+
"loss": 0.0467,
|
| 484 |
+
"step": 1280
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 25.869565217391305,
|
| 488 |
+
"grad_norm": 1.8770432472229004,
|
| 489 |
+
"learning_rate": 6.895424836601308e-06,
|
| 490 |
+
"loss": 0.0443,
|
| 491 |
+
"step": 1320
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 25.985507246376812,
|
| 495 |
+
"eval_loss": 0.053730811923742294,
|
| 496 |
+
"eval_mae": 0.31641700863838196,
|
| 497 |
+
"eval_mse": 0.1325235366821289,
|
| 498 |
+
"eval_runtime": 8331.3737,
|
| 499 |
+
"eval_samples_per_second": 0.05,
|
| 500 |
+
"eval_steps_per_second": 0.006,
|
| 501 |
+
"step": 1326
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 26.657004830917874,
|
| 505 |
+
"grad_norm": 2.1211466789245605,
|
| 506 |
+
"learning_rate": 5.588235294117647e-06,
|
| 507 |
+
"loss": 0.042,
|
| 508 |
+
"step": 1360
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 26.985507246376812,
|
| 512 |
+
"eval_loss": 0.05325399339199066,
|
| 513 |
+
"eval_mae": 0.31560125946998596,
|
| 514 |
+
"eval_mse": 0.13193772733211517,
|
| 515 |
+
"eval_runtime": 3946.2723,
|
| 516 |
+
"eval_samples_per_second": 0.105,
|
| 517 |
+
"eval_steps_per_second": 0.013,
|
| 518 |
+
"step": 1377
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 27.444444444444443,
|
| 522 |
+
"grad_norm": 1.9497586488723755,
|
| 523 |
+
"learning_rate": 4.281045751633987e-06,
|
| 524 |
+
"loss": 0.0397,
|
| 525 |
+
"step": 1400
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 27.985507246376812,
|
| 529 |
+
"eval_loss": 0.052952226251363754,
|
| 530 |
+
"eval_mae": 0.3155405521392822,
|
| 531 |
+
"eval_mse": 0.13170257210731506,
|
| 532 |
+
"eval_runtime": 58.8468,
|
| 533 |
+
"eval_samples_per_second": 7.035,
|
| 534 |
+
"eval_steps_per_second": 0.884,
|
| 535 |
+
"step": 1428
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 28.231884057971016,
|
| 539 |
+
"grad_norm": 5.6321330070495605,
|
| 540 |
+
"learning_rate": 2.9738562091503266e-06,
|
| 541 |
+
"loss": 0.0411,
|
| 542 |
+
"step": 1440
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 28.985507246376812,
|
| 546 |
+
"eval_loss": 0.05421222001314163,
|
| 547 |
+
"eval_mae": 0.31665799021720886,
|
| 548 |
+
"eval_mse": 0.13281531631946564,
|
| 549 |
+
"eval_runtime": 60.076,
|
| 550 |
+
"eval_samples_per_second": 6.891,
|
| 551 |
+
"eval_steps_per_second": 0.866,
|
| 552 |
+
"step": 1479
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 29.019323671497585,
|
| 556 |
+
"grad_norm": 1.5062155723571777,
|
| 557 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 558 |
+
"loss": 0.0385,
|
| 559 |
+
"step": 1480
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 29.792270531400966,
|
| 563 |
+
"grad_norm": 3.7781600952148438,
|
| 564 |
+
"learning_rate": 3.5947712418300653e-07,
|
| 565 |
+
"loss": 0.0382,
|
| 566 |
+
"step": 1520
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 29.985507246376812,
|
| 570 |
+
"eval_loss": 0.05334796383976936,
|
| 571 |
+
"eval_mae": 0.31658393144607544,
|
| 572 |
+
"eval_mse": 0.13268809020519257,
|
| 573 |
+
"eval_runtime": 61.4065,
|
| 574 |
+
"eval_samples_per_second": 6.742,
|
| 575 |
+
"eval_steps_per_second": 0.847,
|
| 576 |
+
"step": 1530
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 29.985507246376812,
|
| 580 |
+
"step": 1530,
|
| 581 |
+
"total_flos": 0.0,
|
| 582 |
+
"train_loss": 0.11046674571006126,
|
| 583 |
+
"train_runtime": 98364.9673,
|
| 584 |
+
"train_samples_per_second": 0.504,
|
| 585 |
+
"train_steps_per_second": 0.016
|
| 586 |
+
}
|
| 587 |
+
],
|
| 588 |
+
"logging_steps": 40,
|
| 589 |
+
"max_steps": 1530,
|
| 590 |
+
"num_input_tokens_seen": 0,
|
| 591 |
+
"num_train_epochs": 30,
|
| 592 |
+
"save_steps": 500,
|
| 593 |
+
"stateful_callbacks": {
|
| 594 |
+
"TrainerControl": {
|
| 595 |
+
"args": {
|
| 596 |
+
"should_epoch_stop": false,
|
| 597 |
+
"should_evaluate": false,
|
| 598 |
+
"should_log": false,
|
| 599 |
+
"should_save": true,
|
| 600 |
+
"should_training_stop": true
|
| 601 |
+
},
|
| 602 |
+
"attributes": {}
|
| 603 |
+
}
|
| 604 |
+
},
|
| 605 |
+
"total_flos": 0.0,
|
| 606 |
+
"train_batch_size": 8,
|
| 607 |
+
"trial_name": null,
|
| 608 |
+
"trial_params": null
|
| 609 |
+
}
|