Commit
•
60d4103
1
Parent(s):
439bbfd
End of training
Browse files- README.md +4 -4
- all_results.json +16 -0
- eval_results.json +11 -0
- train_results.json +8 -0
- trainer_state.json +591 -0
README.md
CHANGED
|
@@ -21,11 +21,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 21 |
|
| 22 |
This model is a fine-tuned version of [facebook/deit-base-patch16-224](https://huggingface.co/facebook/deit-base-patch16-224) on an unknown dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
-
- Loss: 0.
|
| 25 |
- Accuracy: 0.9622
|
| 26 |
-
- Precision: 0.
|
| 27 |
-
- Recall: 0.
|
| 28 |
-
- F1: 0.
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 21 |
|
| 22 |
This model is a fine-tuned version of [facebook/deit-base-patch16-224](https://huggingface.co/facebook/deit-base-patch16-224) on an unknown dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.0979
|
| 25 |
- Accuracy: 0.9622
|
| 26 |
+
- Precision: 0.9531
|
| 27 |
+
- Recall: 0.9562
|
| 28 |
+
- F1: 0.9546
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.88,
|
| 3 |
+
"eval_accuracy": 0.9621993127147767,
|
| 4 |
+
"eval_f1": 0.9546086216676121,
|
| 5 |
+
"eval_loss": 0.09792790561914444,
|
| 6 |
+
"eval_precision": 0.9530929801998389,
|
| 7 |
+
"eval_recall": 0.9561616937721433,
|
| 8 |
+
"eval_runtime": 6.2606,
|
| 9 |
+
"eval_samples_per_second": 92.963,
|
| 10 |
+
"eval_steps_per_second": 5.91,
|
| 11 |
+
"total_flos": 3.1439051980091965e+18,
|
| 12 |
+
"train_loss": 0.1791202223490155,
|
| 13 |
+
"train_runtime": 695.1527,
|
| 14 |
+
"train_samples_per_second": 58.649,
|
| 15 |
+
"train_steps_per_second": 0.906
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.88,
|
| 3 |
+
"eval_accuracy": 0.9621993127147767,
|
| 4 |
+
"eval_f1": 0.9546086216676121,
|
| 5 |
+
"eval_loss": 0.09792790561914444,
|
| 6 |
+
"eval_precision": 0.9530929801998389,
|
| 7 |
+
"eval_recall": 0.9561616937721433,
|
| 8 |
+
"eval_runtime": 6.2606,
|
| 9 |
+
"eval_samples_per_second": 92.963,
|
| 10 |
+
"eval_steps_per_second": 5.91
|
| 11 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.88,
|
| 3 |
+
"total_flos": 3.1439051980091965e+18,
|
| 4 |
+
"train_loss": 0.1791202223490155,
|
| 5 |
+
"train_runtime": 695.1527,
|
| 6 |
+
"train_samples_per_second": 58.649,
|
| 7 |
+
"train_steps_per_second": 0.906
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,591 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.975107296137339,
|
| 3 |
+
"best_model_checkpoint": "deit-base-patch16-224-finetuned-lora-medmnistv2/checkpoint-446",
|
| 4 |
+
"epoch": 9.882352941176471,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 630,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.16,
|
| 13 |
+
"grad_norm": 1.1300164461135864,
|
| 14 |
+
"learning_rate": 0.004920634920634921,
|
| 15 |
+
"loss": 0.4737,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.31,
|
| 20 |
+
"grad_norm": 0.7213650941848755,
|
| 21 |
+
"learning_rate": 0.004841269841269842,
|
| 22 |
+
"loss": 0.6061,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.47,
|
| 27 |
+
"grad_norm": 3.0624804496765137,
|
| 28 |
+
"learning_rate": 0.0047619047619047615,
|
| 29 |
+
"loss": 0.2754,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.63,
|
| 34 |
+
"grad_norm": 0.509834885597229,
|
| 35 |
+
"learning_rate": 0.004682539682539683,
|
| 36 |
+
"loss": 0.2462,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.78,
|
| 41 |
+
"grad_norm": 1.2371031045913696,
|
| 42 |
+
"learning_rate": 0.004603174603174603,
|
| 43 |
+
"loss": 0.205,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.94,
|
| 48 |
+
"grad_norm": 1.2106962203979492,
|
| 49 |
+
"learning_rate": 0.004523809523809524,
|
| 50 |
+
"loss": 0.2556,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.99,
|
| 55 |
+
"eval_accuracy": 0.9184549356223176,
|
| 56 |
+
"eval_f1": 0.8822492390220755,
|
| 57 |
+
"eval_loss": 0.20187422633171082,
|
| 58 |
+
"eval_precision": 0.9446022031823746,
|
| 59 |
+
"eval_recall": 0.8469420502475702,
|
| 60 |
+
"eval_runtime": 11.2928,
|
| 61 |
+
"eval_samples_per_second": 103.163,
|
| 62 |
+
"eval_steps_per_second": 6.464,
|
| 63 |
+
"step": 63
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.1,
|
| 67 |
+
"grad_norm": 2.5079874992370605,
|
| 68 |
+
"learning_rate": 0.0044444444444444444,
|
| 69 |
+
"loss": 0.282,
|
| 70 |
+
"step": 70
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.25,
|
| 74 |
+
"grad_norm": 0.6824792623519897,
|
| 75 |
+
"learning_rate": 0.004365079365079365,
|
| 76 |
+
"loss": 0.2291,
|
| 77 |
+
"step": 80
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.41,
|
| 81 |
+
"grad_norm": 1.294320821762085,
|
| 82 |
+
"learning_rate": 0.004285714285714286,
|
| 83 |
+
"loss": 0.208,
|
| 84 |
+
"step": 90
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.57,
|
| 88 |
+
"grad_norm": 2.1044976711273193,
|
| 89 |
+
"learning_rate": 0.004206349206349207,
|
| 90 |
+
"loss": 0.2288,
|
| 91 |
+
"step": 100
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.73,
|
| 95 |
+
"grad_norm": 0.8730837106704712,
|
| 96 |
+
"learning_rate": 0.0041269841269841265,
|
| 97 |
+
"loss": 0.3071,
|
| 98 |
+
"step": 110
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.88,
|
| 102 |
+
"grad_norm": 1.6219590902328491,
|
| 103 |
+
"learning_rate": 0.004047619047619048,
|
| 104 |
+
"loss": 0.2302,
|
| 105 |
+
"step": 120
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.99,
|
| 109 |
+
"eval_accuracy": 0.9613733905579399,
|
| 110 |
+
"eval_f1": 0.9513881612438695,
|
| 111 |
+
"eval_loss": 0.10982219129800797,
|
| 112 |
+
"eval_precision": 0.9396367521367521,
|
| 113 |
+
"eval_recall": 0.965355461825295,
|
| 114 |
+
"eval_runtime": 11.2742,
|
| 115 |
+
"eval_samples_per_second": 103.333,
|
| 116 |
+
"eval_steps_per_second": 6.475,
|
| 117 |
+
"step": 127
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 2.04,
|
| 121 |
+
"grad_norm": 1.7205462455749512,
|
| 122 |
+
"learning_rate": 0.003968253968253968,
|
| 123 |
+
"loss": 0.214,
|
| 124 |
+
"step": 130
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 2.2,
|
| 128 |
+
"grad_norm": 0.4880439341068268,
|
| 129 |
+
"learning_rate": 0.003888888888888889,
|
| 130 |
+
"loss": 0.2757,
|
| 131 |
+
"step": 140
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 2.35,
|
| 135 |
+
"grad_norm": 1.5420565605163574,
|
| 136 |
+
"learning_rate": 0.0038095238095238095,
|
| 137 |
+
"loss": 0.2314,
|
| 138 |
+
"step": 150
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.51,
|
| 142 |
+
"grad_norm": 0.872795820236206,
|
| 143 |
+
"learning_rate": 0.0037301587301587303,
|
| 144 |
+
"loss": 0.2689,
|
| 145 |
+
"step": 160
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 2.67,
|
| 149 |
+
"grad_norm": 1.3670737743377686,
|
| 150 |
+
"learning_rate": 0.0036507936507936506,
|
| 151 |
+
"loss": 0.1787,
|
| 152 |
+
"step": 170
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.82,
|
| 156 |
+
"grad_norm": 0.9697684049606323,
|
| 157 |
+
"learning_rate": 0.0035714285714285718,
|
| 158 |
+
"loss": 0.1962,
|
| 159 |
+
"step": 180
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.98,
|
| 163 |
+
"grad_norm": 0.6434329152107239,
|
| 164 |
+
"learning_rate": 0.003492063492063492,
|
| 165 |
+
"loss": 0.2258,
|
| 166 |
+
"step": 190
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 3.0,
|
| 170 |
+
"eval_accuracy": 0.9622317596566523,
|
| 171 |
+
"eval_f1": 0.9496135003577972,
|
| 172 |
+
"eval_loss": 0.11506476998329163,
|
| 173 |
+
"eval_precision": 0.9640835246121029,
|
| 174 |
+
"eval_recall": 0.9372077296900788,
|
| 175 |
+
"eval_runtime": 11.448,
|
| 176 |
+
"eval_samples_per_second": 101.764,
|
| 177 |
+
"eval_steps_per_second": 6.377,
|
| 178 |
+
"step": 191
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 3.14,
|
| 182 |
+
"grad_norm": 0.3758108913898468,
|
| 183 |
+
"learning_rate": 0.003412698412698413,
|
| 184 |
+
"loss": 0.2048,
|
| 185 |
+
"step": 200
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 3.29,
|
| 189 |
+
"grad_norm": 0.3552361726760864,
|
| 190 |
+
"learning_rate": 0.003333333333333333,
|
| 191 |
+
"loss": 0.1136,
|
| 192 |
+
"step": 210
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 3.45,
|
| 196 |
+
"grad_norm": 0.3793877959251404,
|
| 197 |
+
"learning_rate": 0.0032539682539682543,
|
| 198 |
+
"loss": 0.2048,
|
| 199 |
+
"step": 220
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 3.61,
|
| 203 |
+
"grad_norm": 0.6903932094573975,
|
| 204 |
+
"learning_rate": 0.0031746031746031746,
|
| 205 |
+
"loss": 0.2696,
|
| 206 |
+
"step": 230
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 3.76,
|
| 210 |
+
"grad_norm": 0.628604531288147,
|
| 211 |
+
"learning_rate": 0.0030952380952380953,
|
| 212 |
+
"loss": 0.1585,
|
| 213 |
+
"step": 240
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 3.92,
|
| 217 |
+
"grad_norm": 0.7250388860702515,
|
| 218 |
+
"learning_rate": 0.0030158730158730157,
|
| 219 |
+
"loss": 0.1465,
|
| 220 |
+
"step": 250
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 4.0,
|
| 224 |
+
"eval_accuracy": 0.9725321888412017,
|
| 225 |
+
"eval_f1": 0.9643166172765768,
|
| 226 |
+
"eval_loss": 0.07334637641906738,
|
| 227 |
+
"eval_precision": 0.9653162002256106,
|
| 228 |
+
"eval_recall": 0.9633286722904824,
|
| 229 |
+
"eval_runtime": 11.4298,
|
| 230 |
+
"eval_samples_per_second": 101.927,
|
| 231 |
+
"eval_steps_per_second": 6.387,
|
| 232 |
+
"step": 255
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.08,
|
| 236 |
+
"grad_norm": 0.4553186893463135,
|
| 237 |
+
"learning_rate": 0.002936507936507937,
|
| 238 |
+
"loss": 0.1609,
|
| 239 |
+
"step": 260
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.24,
|
| 243 |
+
"grad_norm": 0.3900231420993805,
|
| 244 |
+
"learning_rate": 0.002857142857142857,
|
| 245 |
+
"loss": 0.1517,
|
| 246 |
+
"step": 270
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.39,
|
| 250 |
+
"grad_norm": 0.45594316720962524,
|
| 251 |
+
"learning_rate": 0.002777777777777778,
|
| 252 |
+
"loss": 0.1705,
|
| 253 |
+
"step": 280
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 4.55,
|
| 257 |
+
"grad_norm": 0.46610358357429504,
|
| 258 |
+
"learning_rate": 0.002698412698412698,
|
| 259 |
+
"loss": 0.1451,
|
| 260 |
+
"step": 290
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 4.71,
|
| 264 |
+
"grad_norm": 0.557108461856842,
|
| 265 |
+
"learning_rate": 0.0026190476190476194,
|
| 266 |
+
"loss": 0.1582,
|
| 267 |
+
"step": 300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.86,
|
| 271 |
+
"grad_norm": 0.4486098885536194,
|
| 272 |
+
"learning_rate": 0.0025396825396825397,
|
| 273 |
+
"loss": 0.1763,
|
| 274 |
+
"step": 310
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 4.99,
|
| 278 |
+
"eval_accuracy": 0.9725321888412017,
|
| 279 |
+
"eval_f1": 0.9639271421162228,
|
| 280 |
+
"eval_loss": 0.07630692422389984,
|
| 281 |
+
"eval_precision": 0.9702764832336926,
|
| 282 |
+
"eval_recall": 0.9580085885445321,
|
| 283 |
+
"eval_runtime": 11.1034,
|
| 284 |
+
"eval_samples_per_second": 104.923,
|
| 285 |
+
"eval_steps_per_second": 6.575,
|
| 286 |
+
"step": 318
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 5.02,
|
| 290 |
+
"grad_norm": 0.5927975177764893,
|
| 291 |
+
"learning_rate": 0.0024603174603174604,
|
| 292 |
+
"loss": 0.1563,
|
| 293 |
+
"step": 320
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 5.18,
|
| 297 |
+
"grad_norm": 0.5728869438171387,
|
| 298 |
+
"learning_rate": 0.0023809523809523807,
|
| 299 |
+
"loss": 0.1666,
|
| 300 |
+
"step": 330
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 5.33,
|
| 304 |
+
"grad_norm": 0.34924712777137756,
|
| 305 |
+
"learning_rate": 0.0023015873015873015,
|
| 306 |
+
"loss": 0.1425,
|
| 307 |
+
"step": 340
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 5.49,
|
| 311 |
+
"grad_norm": 0.487810343503952,
|
| 312 |
+
"learning_rate": 0.0022222222222222222,
|
| 313 |
+
"loss": 0.1157,
|
| 314 |
+
"step": 350
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 5.65,
|
| 318 |
+
"grad_norm": 0.5470776557922363,
|
| 319 |
+
"learning_rate": 0.002142857142857143,
|
| 320 |
+
"loss": 0.1503,
|
| 321 |
+
"step": 360
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 5.8,
|
| 325 |
+
"grad_norm": 0.9323746562004089,
|
| 326 |
+
"learning_rate": 0.0020634920634920633,
|
| 327 |
+
"loss": 0.1813,
|
| 328 |
+
"step": 370
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 5.96,
|
| 332 |
+
"grad_norm": 0.4948749542236328,
|
| 333 |
+
"learning_rate": 0.001984126984126984,
|
| 334 |
+
"loss": 0.1627,
|
| 335 |
+
"step": 380
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 5.99,
|
| 339 |
+
"eval_accuracy": 0.9570815450643777,
|
| 340 |
+
"eval_f1": 0.9465564950602603,
|
| 341 |
+
"eval_loss": 0.10568992793560028,
|
| 342 |
+
"eval_precision": 0.9315479562643401,
|
| 343 |
+
"eval_recall": 0.965643911608289,
|
| 344 |
+
"eval_runtime": 11.3843,
|
| 345 |
+
"eval_samples_per_second": 102.334,
|
| 346 |
+
"eval_steps_per_second": 6.412,
|
| 347 |
+
"step": 382
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 6.12,
|
| 351 |
+
"grad_norm": 0.5047426819801331,
|
| 352 |
+
"learning_rate": 0.0019047619047619048,
|
| 353 |
+
"loss": 0.1335,
|
| 354 |
+
"step": 390
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 6.27,
|
| 358 |
+
"grad_norm": 0.3176202178001404,
|
| 359 |
+
"learning_rate": 0.0018253968253968253,
|
| 360 |
+
"loss": 0.0897,
|
| 361 |
+
"step": 400
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 6.43,
|
| 365 |
+
"grad_norm": 0.42543891072273254,
|
| 366 |
+
"learning_rate": 0.001746031746031746,
|
| 367 |
+
"loss": 0.1356,
|
| 368 |
+
"step": 410
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 6.59,
|
| 372 |
+
"grad_norm": 0.35515210032463074,
|
| 373 |
+
"learning_rate": 0.0016666666666666666,
|
| 374 |
+
"loss": 0.1324,
|
| 375 |
+
"step": 420
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 6.75,
|
| 379 |
+
"grad_norm": 0.900916337966919,
|
| 380 |
+
"learning_rate": 0.0015873015873015873,
|
| 381 |
+
"loss": 0.1435,
|
| 382 |
+
"step": 430
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 6.9,
|
| 386 |
+
"grad_norm": 0.6653035283088684,
|
| 387 |
+
"learning_rate": 0.0015079365079365078,
|
| 388 |
+
"loss": 0.1509,
|
| 389 |
+
"step": 440
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 7.0,
|
| 393 |
+
"eval_accuracy": 0.975107296137339,
|
| 394 |
+
"eval_f1": 0.9680345077825117,
|
| 395 |
+
"eval_loss": 0.07007680833339691,
|
| 396 |
+
"eval_precision": 0.9637800176993805,
|
| 397 |
+
"eval_recall": 0.9725189498135582,
|
| 398 |
+
"eval_runtime": 12.1771,
|
| 399 |
+
"eval_samples_per_second": 95.672,
|
| 400 |
+
"eval_steps_per_second": 5.995,
|
| 401 |
+
"step": 446
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 7.06,
|
| 405 |
+
"grad_norm": 0.31296506524086,
|
| 406 |
+
"learning_rate": 0.0014285714285714286,
|
| 407 |
+
"loss": 0.1322,
|
| 408 |
+
"step": 450
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.22,
|
| 412 |
+
"grad_norm": 0.6464645862579346,
|
| 413 |
+
"learning_rate": 0.001349206349206349,
|
| 414 |
+
"loss": 0.1457,
|
| 415 |
+
"step": 460
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.37,
|
| 419 |
+
"grad_norm": 1.0385708808898926,
|
| 420 |
+
"learning_rate": 0.0012698412698412698,
|
| 421 |
+
"loss": 0.1278,
|
| 422 |
+
"step": 470
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.53,
|
| 426 |
+
"grad_norm": 0.6095046401023865,
|
| 427 |
+
"learning_rate": 0.0011904761904761904,
|
| 428 |
+
"loss": 0.1168,
|
| 429 |
+
"step": 480
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.69,
|
| 433 |
+
"grad_norm": 0.7724207639694214,
|
| 434 |
+
"learning_rate": 0.0011111111111111111,
|
| 435 |
+
"loss": 0.1182,
|
| 436 |
+
"step": 490
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.84,
|
| 440 |
+
"grad_norm": 0.4225674569606781,
|
| 441 |
+
"learning_rate": 0.0010317460317460316,
|
| 442 |
+
"loss": 0.1325,
|
| 443 |
+
"step": 500
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 8.0,
|
| 447 |
+
"grad_norm": 0.47218218445777893,
|
| 448 |
+
"learning_rate": 0.0009523809523809524,
|
| 449 |
+
"loss": 0.1209,
|
| 450 |
+
"step": 510
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 8.0,
|
| 454 |
+
"eval_accuracy": 0.9570815450643777,
|
| 455 |
+
"eval_f1": 0.9465564950602603,
|
| 456 |
+
"eval_loss": 0.10465892404317856,
|
| 457 |
+
"eval_precision": 0.9315479562643401,
|
| 458 |
+
"eval_recall": 0.965643911608289,
|
| 459 |
+
"eval_runtime": 11.3686,
|
| 460 |
+
"eval_samples_per_second": 102.475,
|
| 461 |
+
"eval_steps_per_second": 6.421,
|
| 462 |
+
"step": 510
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 8.16,
|
| 466 |
+
"grad_norm": 0.296578973531723,
|
| 467 |
+
"learning_rate": 0.000873015873015873,
|
| 468 |
+
"loss": 0.1006,
|
| 469 |
+
"step": 520
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 8.31,
|
| 473 |
+
"grad_norm": 0.9204122424125671,
|
| 474 |
+
"learning_rate": 0.0007936507936507937,
|
| 475 |
+
"loss": 0.1202,
|
| 476 |
+
"step": 530
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 8.47,
|
| 480 |
+
"grad_norm": 0.5195232629776001,
|
| 481 |
+
"learning_rate": 0.0007142857142857143,
|
| 482 |
+
"loss": 0.1052,
|
| 483 |
+
"step": 540
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 8.63,
|
| 487 |
+
"grad_norm": 0.29583191871643066,
|
| 488 |
+
"learning_rate": 0.0006349206349206349,
|
| 489 |
+
"loss": 0.1285,
|
| 490 |
+
"step": 550
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.78,
|
| 494 |
+
"grad_norm": 0.4851154386997223,
|
| 495 |
+
"learning_rate": 0.0005555555555555556,
|
| 496 |
+
"loss": 0.1271,
|
| 497 |
+
"step": 560
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.94,
|
| 501 |
+
"grad_norm": 0.47467851638793945,
|
| 502 |
+
"learning_rate": 0.0004761904761904762,
|
| 503 |
+
"loss": 0.0961,
|
| 504 |
+
"step": 570
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.99,
|
| 508 |
+
"eval_accuracy": 0.9733905579399141,
|
| 509 |
+
"eval_f1": 0.966177521420591,
|
| 510 |
+
"eval_loss": 0.07205679267644882,
|
| 511 |
+
"eval_precision": 0.9577373642293522,
|
| 512 |
+
"eval_recall": 0.975613576624488,
|
| 513 |
+
"eval_runtime": 11.2412,
|
| 514 |
+
"eval_samples_per_second": 103.637,
|
| 515 |
+
"eval_steps_per_second": 6.494,
|
| 516 |
+
"step": 573
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 9.1,
|
| 520 |
+
"grad_norm": 0.29292458295822144,
|
| 521 |
+
"learning_rate": 0.0003968253968253968,
|
| 522 |
+
"loss": 0.096,
|
| 523 |
+
"step": 580
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 9.25,
|
| 527 |
+
"grad_norm": 0.513671338558197,
|
| 528 |
+
"learning_rate": 0.00031746031746031746,
|
| 529 |
+
"loss": 0.0871,
|
| 530 |
+
"step": 590
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 9.41,
|
| 534 |
+
"grad_norm": 0.8143641948699951,
|
| 535 |
+
"learning_rate": 0.0002380952380952381,
|
| 536 |
+
"loss": 0.1297,
|
| 537 |
+
"step": 600
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 9.57,
|
| 541 |
+
"grad_norm": 0.4731471836566925,
|
| 542 |
+
"learning_rate": 0.00015873015873015873,
|
| 543 |
+
"loss": 0.1175,
|
| 544 |
+
"step": 610
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 9.73,
|
| 548 |
+
"grad_norm": 0.30293089151382446,
|
| 549 |
+
"learning_rate": 7.936507936507937e-05,
|
| 550 |
+
"loss": 0.1166,
|
| 551 |
+
"step": 620
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 9.88,
|
| 555 |
+
"grad_norm": 0.3364422023296356,
|
| 556 |
+
"learning_rate": 0.0,
|
| 557 |
+
"loss": 0.1063,
|
| 558 |
+
"step": 630
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 9.88,
|
| 562 |
+
"eval_accuracy": 0.9622317596566523,
|
| 563 |
+
"eval_f1": 0.9526072485207102,
|
| 564 |
+
"eval_loss": 0.08854742348194122,
|
| 565 |
+
"eval_precision": 0.9398335630995462,
|
| 566 |
+
"eval_recall": 0.9680642154165903,
|
| 567 |
+
"eval_runtime": 11.4305,
|
| 568 |
+
"eval_samples_per_second": 101.92,
|
| 569 |
+
"eval_steps_per_second": 6.386,
|
| 570 |
+
"step": 630
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 9.88,
|
| 574 |
+
"step": 630,
|
| 575 |
+
"total_flos": 3.1439051980091965e+18,
|
| 576 |
+
"train_loss": 0.1791202223490155,
|
| 577 |
+
"train_runtime": 695.1527,
|
| 578 |
+
"train_samples_per_second": 58.649,
|
| 579 |
+
"train_steps_per_second": 0.906
|
| 580 |
+
}
|
| 581 |
+
],
|
| 582 |
+
"logging_steps": 10,
|
| 583 |
+
"max_steps": 630,
|
| 584 |
+
"num_input_tokens_seen": 0,
|
| 585 |
+
"num_train_epochs": 10,
|
| 586 |
+
"save_steps": 500,
|
| 587 |
+
"total_flos": 3.1439051980091965e+18,
|
| 588 |
+
"train_batch_size": 16,
|
| 589 |
+
"trial_name": null,
|
| 590 |
+
"trial_params": null
|
| 591 |
+
}
|