End of training
Browse files- README.md +5 -5
- all_results.json +16 -0
- eval_results.json +11 -0
- runs/Mar16_21-02-20_774ccf98b3fe/events.out.tfevents.1742164700.774ccf98b3fe.5066.32 +3 -0
- train_results.json +8 -0
- trainer_state.json +536 -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 [nvidia/mit-b4](https://huggingface.co/nvidia/mit-b4) on an unknown dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
-
- Loss: 0.
|
| 25 |
-
- Accuracy: 0.
|
| 26 |
-
- F1: 0.
|
| 27 |
-
- Precision: 0.
|
| 28 |
-
- Recall: 0.
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 21 |
|
| 22 |
This model is a fine-tuned version of [nvidia/mit-b4](https://huggingface.co/nvidia/mit-b4) on an unknown dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.1162
|
| 25 |
+
- Accuracy: 0.9701
|
| 26 |
+
- F1: 0.9701
|
| 27 |
+
- Precision: 0.9701
|
| 28 |
+
- Recall: 0.9701
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 11.930232558139535,
|
| 3 |
+
"eval_accuracy": 0.9701492537313433,
|
| 4 |
+
"eval_f1": 0.9701426609443169,
|
| 5 |
+
"eval_loss": 0.11616706103086472,
|
| 6 |
+
"eval_precision": 0.9701377402873191,
|
| 7 |
+
"eval_recall": 0.9701492537313433,
|
| 8 |
+
"eval_runtime": 57.5155,
|
| 9 |
+
"eval_samples_per_second": 38.442,
|
| 10 |
+
"eval_steps_per_second": 4.816,
|
| 11 |
+
"total_flos": 1.7667354394198278e+19,
|
| 12 |
+
"train_loss": 0.17970547851321614,
|
| 13 |
+
"train_runtime": 5653.5088,
|
| 14 |
+
"train_samples_per_second": 10.948,
|
| 15 |
+
"train_steps_per_second": 0.342
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 11.930232558139535,
|
| 3 |
+
"eval_accuracy": 0.9701492537313433,
|
| 4 |
+
"eval_f1": 0.9701426609443169,
|
| 5 |
+
"eval_loss": 0.11616706103086472,
|
| 6 |
+
"eval_precision": 0.9701377402873191,
|
| 7 |
+
"eval_recall": 0.9701492537313433,
|
| 8 |
+
"eval_runtime": 57.5155,
|
| 9 |
+
"eval_samples_per_second": 38.442,
|
| 10 |
+
"eval_steps_per_second": 4.816
|
| 11 |
+
}
|
runs/Mar16_21-02-20_774ccf98b3fe/events.out.tfevents.1742164700.774ccf98b3fe.5066.32
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96206c640804f39e8f0a908bdba8a5eacbc1e61fea77c9a2a62f625aba56daa9
|
| 3 |
+
size 560
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 11.930232558139535,
|
| 3 |
+
"total_flos": 1.7667354394198278e+19,
|
| 4 |
+
"train_loss": 0.17970547851321614,
|
| 5 |
+
"train_runtime": 5653.5088,
|
| 6 |
+
"train_samples_per_second": 10.948,
|
| 7 |
+
"train_steps_per_second": 0.342
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,536 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9701426609443169,
|
| 3 |
+
"best_model_checkpoint": "mit-b4-finetuned-stroke-binary/checkpoint-1700",
|
| 4 |
+
"epoch": 11.930232558139535,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 1932,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.31007751937984496,
|
| 13 |
+
"grad_norm": 2.3132071495056152,
|
| 14 |
+
"learning_rate": 5.154639175257732e-06,
|
| 15 |
+
"loss": 0.6722,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.6201550387596899,
|
| 20 |
+
"grad_norm": 6.47659969329834,
|
| 21 |
+
"learning_rate": 1.0309278350515464e-05,
|
| 22 |
+
"loss": 0.5714,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.6201550387596899,
|
| 27 |
+
"eval_accuracy": 0.7878787878787878,
|
| 28 |
+
"eval_f1": 0.7800219254126745,
|
| 29 |
+
"eval_loss": 0.477566123008728,
|
| 30 |
+
"eval_precision": 0.7900169125975578,
|
| 31 |
+
"eval_recall": 0.7878787878787878,
|
| 32 |
+
"eval_runtime": 56.589,
|
| 33 |
+
"eval_samples_per_second": 39.071,
|
| 34 |
+
"eval_steps_per_second": 4.895,
|
| 35 |
+
"step": 100
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.9302325581395349,
|
| 39 |
+
"grad_norm": 4.08937406539917,
|
| 40 |
+
"learning_rate": 1.5360824742268042e-05,
|
| 41 |
+
"loss": 0.4471,
|
| 42 |
+
"step": 150
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.235658914728682,
|
| 46 |
+
"grad_norm": 8.325615882873535,
|
| 47 |
+
"learning_rate": 1.9999738610404825e-05,
|
| 48 |
+
"loss": 0.3897,
|
| 49 |
+
"step": 200
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.235658914728682,
|
| 53 |
+
"eval_accuracy": 0.8715513342379014,
|
| 54 |
+
"eval_f1": 0.8704450429260716,
|
| 55 |
+
"eval_loss": 0.3238992393016815,
|
| 56 |
+
"eval_precision": 0.8710997928702509,
|
| 57 |
+
"eval_recall": 0.8715513342379014,
|
| 58 |
+
"eval_runtime": 56.7955,
|
| 59 |
+
"eval_samples_per_second": 38.929,
|
| 60 |
+
"eval_steps_per_second": 4.877,
|
| 61 |
+
"step": 200
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 1.5457364341085271,
|
| 65 |
+
"grad_norm": 8.041431427001953,
|
| 66 |
+
"learning_rate": 1.9952399350448247e-05,
|
| 67 |
+
"loss": 0.3177,
|
| 68 |
+
"step": 250
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.8558139534883722,
|
| 72 |
+
"grad_norm": 7.284913063049316,
|
| 73 |
+
"learning_rate": 1.9823819633544185e-05,
|
| 74 |
+
"loss": 0.2951,
|
| 75 |
+
"step": 300
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 1.8558139534883722,
|
| 79 |
+
"eval_accuracy": 0.8765264586160109,
|
| 80 |
+
"eval_f1": 0.8723990420158877,
|
| 81 |
+
"eval_loss": 0.31197357177734375,
|
| 82 |
+
"eval_precision": 0.885806076698763,
|
| 83 |
+
"eval_recall": 0.8765264586160109,
|
| 84 |
+
"eval_runtime": 56.5628,
|
| 85 |
+
"eval_samples_per_second": 39.089,
|
| 86 |
+
"eval_steps_per_second": 4.897,
|
| 87 |
+
"step": 300
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 2.1612403100775195,
|
| 91 |
+
"grad_norm": 3.927116870880127,
|
| 92 |
+
"learning_rate": 1.9615049043274207e-05,
|
| 93 |
+
"loss": 0.2786,
|
| 94 |
+
"step": 350
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 2.471317829457364,
|
| 98 |
+
"grad_norm": 13.424771308898926,
|
| 99 |
+
"learning_rate": 1.932779175343134e-05,
|
| 100 |
+
"loss": 0.23,
|
| 101 |
+
"step": 400
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.471317829457364,
|
| 105 |
+
"eval_accuracy": 0.9280868385345997,
|
| 106 |
+
"eval_f1": 0.9271058471970156,
|
| 107 |
+
"eval_loss": 0.1993584781885147,
|
| 108 |
+
"eval_precision": 0.9303851930689407,
|
| 109 |
+
"eval_recall": 0.9280868385345997,
|
| 110 |
+
"eval_runtime": 56.4712,
|
| 111 |
+
"eval_samples_per_second": 39.153,
|
| 112 |
+
"eval_steps_per_second": 4.905,
|
| 113 |
+
"step": 400
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 2.781395348837209,
|
| 117 |
+
"grad_norm": 11.889705657958984,
|
| 118 |
+
"learning_rate": 1.8964392617017013e-05,
|
| 119 |
+
"loss": 0.2424,
|
| 120 |
+
"step": 450
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 3.0868217054263565,
|
| 124 |
+
"grad_norm": 14.436951637268066,
|
| 125 |
+
"learning_rate": 1.8527818025436662e-05,
|
| 126 |
+
"loss": 0.2135,
|
| 127 |
+
"step": 500
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 3.0868217054263565,
|
| 131 |
+
"eval_accuracy": 0.9280868385345997,
|
| 132 |
+
"eval_f1": 0.9266634060111499,
|
| 133 |
+
"eval_loss": 0.2157154679298401,
|
| 134 |
+
"eval_precision": 0.9332581537403823,
|
| 135 |
+
"eval_recall": 0.9280868385345997,
|
| 136 |
+
"eval_runtime": 56.599,
|
| 137 |
+
"eval_samples_per_second": 39.064,
|
| 138 |
+
"eval_steps_per_second": 4.894,
|
| 139 |
+
"step": 500
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 3.3968992248062015,
|
| 143 |
+
"grad_norm": 16.79993438720703,
|
| 144 |
+
"learning_rate": 1.802163169413846e-05,
|
| 145 |
+
"loss": 0.1746,
|
| 146 |
+
"step": 550
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.7069767441860466,
|
| 150 |
+
"grad_norm": 14.388134002685547,
|
| 151 |
+
"learning_rate": 1.7449965572354675e-05,
|
| 152 |
+
"loss": 0.2106,
|
| 153 |
+
"step": 600
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 3.7069767441860466,
|
| 157 |
+
"eval_accuracy": 0.9380370872908186,
|
| 158 |
+
"eval_f1": 0.9382102726921231,
|
| 159 |
+
"eval_loss": 0.18085584044456482,
|
| 160 |
+
"eval_precision": 0.9386737275368563,
|
| 161 |
+
"eval_recall": 0.9380370872908186,
|
| 162 |
+
"eval_runtime": 56.6285,
|
| 163 |
+
"eval_samples_per_second": 39.044,
|
| 164 |
+
"eval_steps_per_second": 4.892,
|
| 165 |
+
"step": 600
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 4.0124031007751935,
|
| 169 |
+
"grad_norm": 14.440871238708496,
|
| 170 |
+
"learning_rate": 1.68174861144065e-05,
|
| 171 |
+
"loss": 0.1802,
|
| 172 |
+
"step": 650
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 4.322480620155039,
|
| 176 |
+
"grad_norm": 6.46437931060791,
|
| 177 |
+
"learning_rate": 1.612935618789643e-05,
|
| 178 |
+
"loss": 0.1576,
|
| 179 |
+
"step": 700
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 4.322480620155039,
|
| 183 |
+
"eval_accuracy": 0.9402985074626866,
|
| 184 |
+
"eval_f1": 0.9403501047780813,
|
| 185 |
+
"eval_loss": 0.1628771722316742,
|
| 186 |
+
"eval_precision": 0.9404274341623059,
|
| 187 |
+
"eval_recall": 0.9402985074626866,
|
| 188 |
+
"eval_runtime": 56.5976,
|
| 189 |
+
"eval_samples_per_second": 39.065,
|
| 190 |
+
"eval_steps_per_second": 4.894,
|
| 191 |
+
"step": 700
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 4.632558139534884,
|
| 195 |
+
"grad_norm": 12.626237869262695,
|
| 196 |
+
"learning_rate": 1.5391192929727884e-05,
|
| 197 |
+
"loss": 0.1752,
|
| 198 |
+
"step": 750
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 4.942635658914728,
|
| 202 |
+
"grad_norm": 20.061859130859375,
|
| 203 |
+
"learning_rate": 1.460902189396916e-05,
|
| 204 |
+
"loss": 0.1434,
|
| 205 |
+
"step": 800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 4.942635658914728,
|
| 209 |
+
"eval_accuracy": 0.9543193125282677,
|
| 210 |
+
"eval_f1": 0.9541868361381168,
|
| 211 |
+
"eval_loss": 0.1526043862104416,
|
| 212 |
+
"eval_precision": 0.9543134613523963,
|
| 213 |
+
"eval_recall": 0.9543193125282677,
|
| 214 |
+
"eval_runtime": 56.4759,
|
| 215 |
+
"eval_samples_per_second": 39.149,
|
| 216 |
+
"eval_steps_per_second": 4.905,
|
| 217 |
+
"step": 800
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 5.248062015503876,
|
| 221 |
+
"grad_norm": 11.535951614379883,
|
| 222 |
+
"learning_rate": 1.3789227865848282e-05,
|
| 223 |
+
"loss": 0.1601,
|
| 224 |
+
"step": 850
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 5.558139534883721,
|
| 228 |
+
"grad_norm": 9.65259838104248,
|
| 229 |
+
"learning_rate": 1.2938502743379212e-05,
|
| 230 |
+
"loss": 0.1391,
|
| 231 |
+
"step": 900
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 5.558139534883721,
|
| 235 |
+
"eval_accuracy": 0.9574853007688828,
|
| 236 |
+
"eval_f1": 0.9574664394664735,
|
| 237 |
+
"eval_loss": 0.1268271952867508,
|
| 238 |
+
"eval_precision": 0.9574541746977356,
|
| 239 |
+
"eval_recall": 0.9574853007688828,
|
| 240 |
+
"eval_runtime": 56.6517,
|
| 241 |
+
"eval_samples_per_second": 39.028,
|
| 242 |
+
"eval_steps_per_second": 4.89,
|
| 243 |
+
"step": 900
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 5.868217054263566,
|
| 247 |
+
"grad_norm": 7.883506774902344,
|
| 248 |
+
"learning_rate": 1.2063790912056577e-05,
|
| 249 |
+
"loss": 0.1605,
|
| 250 |
+
"step": 950
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 6.173643410852713,
|
| 254 |
+
"grad_norm": 9.489595413208008,
|
| 255 |
+
"learning_rate": 1.1172232558519983e-05,
|
| 256 |
+
"loss": 0.1048,
|
| 257 |
+
"step": 1000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 6.173643410852713,
|
| 261 |
+
"eval_accuracy": 0.9556761646313885,
|
| 262 |
+
"eval_f1": 0.9555092018481103,
|
| 263 |
+
"eval_loss": 0.14889651536941528,
|
| 264 |
+
"eval_precision": 0.9557688265871488,
|
| 265 |
+
"eval_recall": 0.9556761646313885,
|
| 266 |
+
"eval_runtime": 56.5428,
|
| 267 |
+
"eval_samples_per_second": 39.103,
|
| 268 |
+
"eval_steps_per_second": 4.899,
|
| 269 |
+
"step": 1000
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 6.4837209302325585,
|
| 273 |
+
"grad_norm": 5.59963846206665,
|
| 274 |
+
"learning_rate": 1.0271105385912779e-05,
|
| 275 |
+
"loss": 0.116,
|
| 276 |
+
"step": 1050
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 6.793798449612403,
|
| 280 |
+
"grad_norm": 4.873887538909912,
|
| 281 |
+
"learning_rate": 9.367765206707174e-06,
|
| 282 |
+
"loss": 0.1271,
|
| 283 |
+
"step": 1100
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 6.793798449612403,
|
| 287 |
+
"eval_accuracy": 0.9570330167345092,
|
| 288 |
+
"eval_f1": 0.9566191699282054,
|
| 289 |
+
"eval_loss": 0.14482761919498444,
|
| 290 |
+
"eval_precision": 0.9586051623091093,
|
| 291 |
+
"eval_recall": 0.9570330167345092,
|
| 292 |
+
"eval_runtime": 56.8397,
|
| 293 |
+
"eval_samples_per_second": 38.899,
|
| 294 |
+
"eval_steps_per_second": 4.873,
|
| 295 |
+
"step": 1100
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 7.09922480620155,
|
| 299 |
+
"grad_norm": 4.528378963470459,
|
| 300 |
+
"learning_rate": 8.469585897930557e-06,
|
| 301 |
+
"loss": 0.1333,
|
| 302 |
+
"step": 1150
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.409302325581395,
|
| 306 |
+
"grad_norm": 1.5636117458343506,
|
| 307 |
+
"learning_rate": 7.583899208932648e-06,
|
| 308 |
+
"loss": 0.091,
|
| 309 |
+
"step": 1200
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 7.409302325581395,
|
| 313 |
+
"eval_accuracy": 0.9570330167345092,
|
| 314 |
+
"eval_f1": 0.9567056534394107,
|
| 315 |
+
"eval_loss": 0.14507929980754852,
|
| 316 |
+
"eval_precision": 0.9579537259191305,
|
| 317 |
+
"eval_recall": 0.9570330167345092,
|
| 318 |
+
"eval_runtime": 56.7,
|
| 319 |
+
"eval_samples_per_second": 38.995,
|
| 320 |
+
"eval_steps_per_second": 4.885,
|
| 321 |
+
"step": 1200
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 7.7193798449612405,
|
| 325 |
+
"grad_norm": 5.480973720550537,
|
| 326 |
+
"learning_rate": 6.7179349130367235e-06,
|
| 327 |
+
"loss": 0.1089,
|
| 328 |
+
"step": 1250
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 8.024806201550387,
|
| 332 |
+
"grad_norm": 6.6158223152160645,
|
| 333 |
+
"learning_rate": 5.878761791611129e-06,
|
| 334 |
+
"loss": 0.1159,
|
| 335 |
+
"step": 1300
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 8.024806201550387,
|
| 339 |
+
"eval_accuracy": 0.9629127091813658,
|
| 340 |
+
"eval_f1": 0.9626761639787419,
|
| 341 |
+
"eval_loss": 0.1205127015709877,
|
| 342 |
+
"eval_precision": 0.9635871559570208,
|
| 343 |
+
"eval_recall": 0.9629127091813658,
|
| 344 |
+
"eval_runtime": 56.6004,
|
| 345 |
+
"eval_samples_per_second": 39.063,
|
| 346 |
+
"eval_steps_per_second": 4.894,
|
| 347 |
+
"step": 1300
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 8.334883720930232,
|
| 351 |
+
"grad_norm": 5.2470316886901855,
|
| 352 |
+
"learning_rate": 5.073229932302277e-06,
|
| 353 |
+
"loss": 0.0886,
|
| 354 |
+
"step": 1350
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 8.644961240310078,
|
| 358 |
+
"grad_norm": 6.15119743347168,
|
| 359 |
+
"learning_rate": 4.307914812442993e-06,
|
| 360 |
+
"loss": 0.1151,
|
| 361 |
+
"step": 1400
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 8.644961240310078,
|
| 365 |
+
"eval_accuracy": 0.9665309814563546,
|
| 366 |
+
"eval_f1": 0.9664379980687814,
|
| 367 |
+
"eval_loss": 0.11242391169071198,
|
| 368 |
+
"eval_precision": 0.9665871538113867,
|
| 369 |
+
"eval_recall": 0.9665309814563546,
|
| 370 |
+
"eval_runtime": 56.6713,
|
| 371 |
+
"eval_samples_per_second": 39.014,
|
| 372 |
+
"eval_steps_per_second": 4.888,
|
| 373 |
+
"step": 1400
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 8.955038759689922,
|
| 377 |
+
"grad_norm": 12.04592227935791,
|
| 378 |
+
"learning_rate": 3.589063624077802e-06,
|
| 379 |
+
"loss": 0.0798,
|
| 380 |
+
"step": 1450
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.26046511627907,
|
| 384 |
+
"grad_norm": 13.314713478088379,
|
| 385 |
+
"learning_rate": 2.922544278748801e-06,
|
| 386 |
+
"loss": 0.0735,
|
| 387 |
+
"step": 1500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 9.26046511627907,
|
| 391 |
+
"eval_accuracy": 0.9642695612844867,
|
| 392 |
+
"eval_f1": 0.9641123145223969,
|
| 393 |
+
"eval_loss": 0.11749936640262604,
|
| 394 |
+
"eval_precision": 0.9645326112328965,
|
| 395 |
+
"eval_recall": 0.9642695612844867,
|
| 396 |
+
"eval_runtime": 56.8212,
|
| 397 |
+
"eval_samples_per_second": 38.912,
|
| 398 |
+
"eval_steps_per_second": 4.875,
|
| 399 |
+
"step": 1500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 9.570542635658915,
|
| 403 |
+
"grad_norm": 1.456084132194519,
|
| 404 |
+
"learning_rate": 2.3137975083109153e-06,
|
| 405 |
+
"loss": 0.0746,
|
| 406 |
+
"step": 1550
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 9.88062015503876,
|
| 410 |
+
"grad_norm": 2.1265344619750977,
|
| 411 |
+
"learning_rate": 1.7677924527729228e-06,
|
| 412 |
+
"loss": 0.0537,
|
| 413 |
+
"step": 1600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 9.88062015503876,
|
| 417 |
+
"eval_accuracy": 0.9678878335594754,
|
| 418 |
+
"eval_f1": 0.9678102645900477,
|
| 419 |
+
"eval_loss": 0.11535227298736572,
|
| 420 |
+
"eval_precision": 0.9679181198554704,
|
| 421 |
+
"eval_recall": 0.9678878335594754,
|
| 422 |
+
"eval_runtime": 56.5576,
|
| 423 |
+
"eval_samples_per_second": 39.093,
|
| 424 |
+
"eval_steps_per_second": 4.898,
|
| 425 |
+
"step": 1600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 10.186046511627907,
|
| 429 |
+
"grad_norm": 12.95783805847168,
|
| 430 |
+
"learning_rate": 1.2889860976963542e-06,
|
| 431 |
+
"loss": 0.0857,
|
| 432 |
+
"step": 1650
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 10.496124031007753,
|
| 436 |
+
"grad_norm": 3.627340078353882,
|
| 437 |
+
"learning_rate": 8.812868922607565e-07,
|
| 438 |
+
"loss": 0.0666,
|
| 439 |
+
"step": 1700
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 10.496124031007753,
|
| 443 |
+
"eval_accuracy": 0.9701492537313433,
|
| 444 |
+
"eval_f1": 0.9701426609443169,
|
| 445 |
+
"eval_loss": 0.11616706103086472,
|
| 446 |
+
"eval_precision": 0.9701377402873191,
|
| 447 |
+
"eval_recall": 0.9701492537313433,
|
| 448 |
+
"eval_runtime": 56.4987,
|
| 449 |
+
"eval_samples_per_second": 39.134,
|
| 450 |
+
"eval_steps_per_second": 4.903,
|
| 451 |
+
"step": 1700
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 10.806201550387597,
|
| 455 |
+
"grad_norm": 9.41781997680664,
|
| 456 |
+
"learning_rate": 5.480228449774882e-07,
|
| 457 |
+
"loss": 0.0722,
|
| 458 |
+
"step": 1750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 11.111627906976745,
|
| 462 |
+
"grad_norm": 12.687678337097168,
|
| 463 |
+
"learning_rate": 2.9626582353969756e-07,
|
| 464 |
+
"loss": 0.0732,
|
| 465 |
+
"step": 1800
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 11.111627906976745,
|
| 469 |
+
"eval_accuracy": 0.9678878335594754,
|
| 470 |
+
"eval_f1": 0.9678179490403084,
|
| 471 |
+
"eval_loss": 0.11334193497896194,
|
| 472 |
+
"eval_precision": 0.9678997125749722,
|
| 473 |
+
"eval_recall": 0.9678878335594754,
|
| 474 |
+
"eval_runtime": 56.7681,
|
| 475 |
+
"eval_samples_per_second": 38.948,
|
| 476 |
+
"eval_steps_per_second": 4.88,
|
| 477 |
+
"step": 1800
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 11.421705426356588,
|
| 481 |
+
"grad_norm": 13.630194664001465,
|
| 482 |
+
"learning_rate": 1.1780223451346994e-07,
|
| 483 |
+
"loss": 0.0868,
|
| 484 |
+
"step": 1850
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 11.731782945736434,
|
| 488 |
+
"grad_norm": 17.37832260131836,
|
| 489 |
+
"learning_rate": 2.0006053801937543e-08,
|
| 490 |
+
"loss": 0.0775,
|
| 491 |
+
"step": 1900
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 11.731782945736434,
|
| 495 |
+
"eval_accuracy": 0.968340117593849,
|
| 496 |
+
"eval_f1": 0.9682674370094175,
|
| 497 |
+
"eval_loss": 0.112978994846344,
|
| 498 |
+
"eval_precision": 0.9683629530510336,
|
| 499 |
+
"eval_recall": 0.968340117593849,
|
| 500 |
+
"eval_runtime": 56.6602,
|
| 501 |
+
"eval_samples_per_second": 39.022,
|
| 502 |
+
"eval_steps_per_second": 4.889,
|
| 503 |
+
"step": 1900
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 11.930232558139535,
|
| 507 |
+
"step": 1932,
|
| 508 |
+
"total_flos": 1.7667354394198278e+19,
|
| 509 |
+
"train_loss": 0.17970547851321614,
|
| 510 |
+
"train_runtime": 5653.5088,
|
| 511 |
+
"train_samples_per_second": 10.948,
|
| 512 |
+
"train_steps_per_second": 0.342
|
| 513 |
+
}
|
| 514 |
+
],
|
| 515 |
+
"logging_steps": 50,
|
| 516 |
+
"max_steps": 1932,
|
| 517 |
+
"num_input_tokens_seen": 0,
|
| 518 |
+
"num_train_epochs": 12,
|
| 519 |
+
"save_steps": 100,
|
| 520 |
+
"stateful_callbacks": {
|
| 521 |
+
"TrainerControl": {
|
| 522 |
+
"args": {
|
| 523 |
+
"should_epoch_stop": false,
|
| 524 |
+
"should_evaluate": false,
|
| 525 |
+
"should_log": false,
|
| 526 |
+
"should_save": true,
|
| 527 |
+
"should_training_stop": true
|
| 528 |
+
},
|
| 529 |
+
"attributes": {}
|
| 530 |
+
}
|
| 531 |
+
},
|
| 532 |
+
"total_flos": 1.7667354394198278e+19,
|
| 533 |
+
"train_batch_size": 8,
|
| 534 |
+
"trial_name": null,
|
| 535 |
+
"trial_params": null
|
| 536 |
+
}
|