End of training
Browse files- README.md +2 -2
- all_results.json +11 -11
- eval_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +735 -66
README.md
CHANGED
|
@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Accuracy: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.2385
|
| 22 |
+
- Accuracy: 0.9328
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime": 2.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 4.
|
| 8 |
-
"total_flos": 2.
|
| 9 |
-
"train_loss":
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second": 72.
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 35.0,
|
| 3 |
+
"eval_accuracy": 0.9328358208955224,
|
| 4 |
+
"eval_loss": 0.23848846554756165,
|
| 5 |
+
"eval_runtime": 2.1668,
|
| 6 |
+
"eval_samples_per_second": 123.687,
|
| 7 |
+
"eval_steps_per_second": 4.154,
|
| 8 |
+
"total_flos": 2.0920841725722624e+18,
|
| 9 |
+
"train_loss": 0.6001271831361871,
|
| 10 |
+
"train_runtime": 1153.5131,
|
| 11 |
+
"train_samples_per_second": 72.942,
|
| 12 |
+
"train_steps_per_second": 0.576
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss":
|
| 5 |
-
"eval_runtime": 2.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 4.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 35.0,
|
| 3 |
+
"eval_accuracy": 0.9328358208955224,
|
| 4 |
+
"eval_loss": 0.23848846554756165,
|
| 5 |
+
"eval_runtime": 2.1668,
|
| 6 |
+
"eval_samples_per_second": 123.687,
|
| 7 |
+
"eval_steps_per_second": 4.154
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 2.
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second": 72.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 35.0,
|
| 3 |
+
"total_flos": 2.0920841725722624e+18,
|
| 4 |
+
"train_loss": 0.6001271831361871,
|
| 5 |
+
"train_runtime": 1153.5131,
|
| 6 |
+
"train_samples_per_second": 72.942,
|
| 7 |
+
"train_steps_per_second": 0.576
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,135 +1,804 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "/kaggle/working/swin-brain-tumor-type-classification/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.5263157894736842,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 1.0,
|
| 20 |
-
"eval_accuracy": 0.
|
| 21 |
-
"eval_loss": 2.
|
| 22 |
-
"eval_runtime": 2.
|
| 23 |
-
"eval_samples_per_second":
|
| 24 |
-
"eval_steps_per_second": 4.
|
| 25 |
"step": 19
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 1.0526315789473684,
|
| 29 |
-
"grad_norm": 5.
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss": 2.
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 1.5789473684210527,
|
| 36 |
-
"grad_norm":
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss":
|
| 39 |
"step": 30
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 2.0,
|
| 43 |
-
"eval_accuracy": 0.
|
| 44 |
-
"eval_loss":
|
| 45 |
-
"eval_runtime": 2.
|
| 46 |
-
"eval_samples_per_second": 126.
|
| 47 |
-
"eval_steps_per_second": 4.
|
| 48 |
"step": 38
|
| 49 |
},
|
| 50 |
{
|
| 51 |
"epoch": 2.1052631578947367,
|
| 52 |
-
"grad_norm":
|
| 53 |
-
"learning_rate":
|
| 54 |
-
"loss":
|
| 55 |
"step": 40
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 2.6315789473684212,
|
| 59 |
-
"grad_norm":
|
| 60 |
-
"learning_rate":
|
| 61 |
-
"loss": 1.
|
| 62 |
"step": 50
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 3.0,
|
| 66 |
-
"eval_accuracy": 0.
|
| 67 |
-
"eval_loss": 1.
|
| 68 |
-
"eval_runtime": 2.
|
| 69 |
-
"eval_samples_per_second":
|
| 70 |
-
"eval_steps_per_second": 4.
|
| 71 |
"step": 57
|
| 72 |
},
|
| 73 |
{
|
| 74 |
"epoch": 3.1578947368421053,
|
| 75 |
-
"grad_norm":
|
| 76 |
-
"learning_rate":
|
| 77 |
-
"loss": 1.
|
| 78 |
"step": 60
|
| 79 |
},
|
| 80 |
{
|
| 81 |
"epoch": 3.6842105263157894,
|
| 82 |
-
"grad_norm":
|
| 83 |
-
"learning_rate":
|
| 84 |
-
"loss": 1.
|
| 85 |
"step": 70
|
| 86 |
},
|
| 87 |
{
|
| 88 |
"epoch": 4.0,
|
| 89 |
-
"eval_accuracy": 0.
|
| 90 |
-
"eval_loss": 1.
|
| 91 |
-
"eval_runtime": 2.
|
| 92 |
-
"eval_samples_per_second":
|
| 93 |
-
"eval_steps_per_second":
|
| 94 |
"step": 76
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 4.2105263157894735,
|
| 98 |
-
"grad_norm":
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss": 1.
|
| 101 |
"step": 80
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 4.7368421052631575,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss": 1.
|
| 108 |
"step": 90
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 5.0,
|
| 112 |
-
"eval_accuracy": 0.
|
| 113 |
-
"eval_loss": 1.
|
| 114 |
-
"eval_runtime": 2.
|
| 115 |
-
"eval_samples_per_second":
|
| 116 |
-
"eval_steps_per_second":
|
| 117 |
"step": 95
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 5.
|
| 121 |
-
"
|
| 122 |
-
"
|
| 123 |
-
"
|
| 124 |
-
"
|
| 125 |
-
|
| 126 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 127 |
}
|
| 128 |
],
|
| 129 |
"logging_steps": 10,
|
| 130 |
-
"max_steps":
|
| 131 |
"num_input_tokens_seen": 0,
|
| 132 |
-
"num_train_epochs":
|
| 133 |
"save_steps": 500,
|
| 134 |
"stateful_callbacks": {
|
| 135 |
"CustomEarlyStoppingCallback": {
|
|
@@ -152,7 +821,7 @@
|
|
| 152 |
"attributes": {}
|
| 153 |
}
|
| 154 |
},
|
| 155 |
-
"total_flos": 2.
|
| 156 |
"train_batch_size": 32,
|
| 157 |
"trial_name": null,
|
| 158 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9328358208955224,
|
| 3 |
+
"best_model_checkpoint": "/kaggle/working/swin-brain-tumor-type-classification/checkpoint-627",
|
| 4 |
+
"epoch": 35.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 665,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.5263157894736842,
|
| 13 |
+
"grad_norm": 6.779318809509277,
|
| 14 |
+
"learning_rate": 7.4626865671641785e-06,
|
| 15 |
+
"loss": 2.6739,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.19776119402985073,
|
| 21 |
+
"eval_loss": 2.423863410949707,
|
| 22 |
+
"eval_runtime": 2.1298,
|
| 23 |
+
"eval_samples_per_second": 125.831,
|
| 24 |
+
"eval_steps_per_second": 4.226,
|
| 25 |
"step": 19
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 1.0526315789473684,
|
| 29 |
+
"grad_norm": 5.687614917755127,
|
| 30 |
+
"learning_rate": 1.4925373134328357e-05,
|
| 31 |
+
"loss": 2.4803,
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 1.5789473684210527,
|
| 36 |
+
"grad_norm": 7.469175338745117,
|
| 37 |
+
"learning_rate": 2.238805970149254e-05,
|
| 38 |
+
"loss": 2.2793,
|
| 39 |
"step": 30
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 2.0,
|
| 43 |
+
"eval_accuracy": 0.3619402985074627,
|
| 44 |
+
"eval_loss": 2.101121187210083,
|
| 45 |
+
"eval_runtime": 2.1211,
|
| 46 |
+
"eval_samples_per_second": 126.35,
|
| 47 |
+
"eval_steps_per_second": 4.243,
|
| 48 |
"step": 38
|
| 49 |
},
|
| 50 |
{
|
| 51 |
"epoch": 2.1052631578947367,
|
| 52 |
+
"grad_norm": 7.969486713409424,
|
| 53 |
+
"learning_rate": 2.9850746268656714e-05,
|
| 54 |
+
"loss": 2.1009,
|
| 55 |
"step": 40
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 2.6315789473684212,
|
| 59 |
+
"grad_norm": 11.056779861450195,
|
| 60 |
+
"learning_rate": 3.73134328358209e-05,
|
| 61 |
+
"loss": 1.9273,
|
| 62 |
"step": 50
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 3.0,
|
| 66 |
+
"eval_accuracy": 0.4962686567164179,
|
| 67 |
+
"eval_loss": 1.6876119375228882,
|
| 68 |
+
"eval_runtime": 2.1801,
|
| 69 |
+
"eval_samples_per_second": 122.928,
|
| 70 |
+
"eval_steps_per_second": 4.128,
|
| 71 |
"step": 57
|
| 72 |
},
|
| 73 |
{
|
| 74 |
"epoch": 3.1578947368421053,
|
| 75 |
+
"grad_norm": 7.5373334884643555,
|
| 76 |
+
"learning_rate": 4.477611940298508e-05,
|
| 77 |
+
"loss": 1.6252,
|
| 78 |
"step": 60
|
| 79 |
},
|
| 80 |
{
|
| 81 |
"epoch": 3.6842105263157894,
|
| 82 |
+
"grad_norm": 10.007360458374023,
|
| 83 |
+
"learning_rate": 4.974916387959867e-05,
|
| 84 |
+
"loss": 1.5549,
|
| 85 |
"step": 70
|
| 86 |
},
|
| 87 |
{
|
| 88 |
"epoch": 4.0,
|
| 89 |
+
"eval_accuracy": 0.5708955223880597,
|
| 90 |
+
"eval_loss": 1.3625893592834473,
|
| 91 |
+
"eval_runtime": 2.2749,
|
| 92 |
+
"eval_samples_per_second": 117.809,
|
| 93 |
+
"eval_steps_per_second": 3.956,
|
| 94 |
"step": 76
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 4.2105263157894735,
|
| 98 |
+
"grad_norm": 10.789587020874023,
|
| 99 |
+
"learning_rate": 4.891304347826087e-05,
|
| 100 |
+
"loss": 1.3971,
|
| 101 |
"step": 80
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 4.7368421052631575,
|
| 105 |
+
"grad_norm": 9.604104042053223,
|
| 106 |
+
"learning_rate": 4.8076923076923084e-05,
|
| 107 |
+
"loss": 1.2386,
|
| 108 |
"step": 90
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 5.0,
|
| 112 |
+
"eval_accuracy": 0.6268656716417911,
|
| 113 |
+
"eval_loss": 1.1268030405044556,
|
| 114 |
+
"eval_runtime": 2.5243,
|
| 115 |
+
"eval_samples_per_second": 106.167,
|
| 116 |
+
"eval_steps_per_second": 3.565,
|
| 117 |
"step": 95
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 5.2631578947368425,
|
| 121 |
+
"grad_norm": 12.851019859313965,
|
| 122 |
+
"learning_rate": 4.724080267558529e-05,
|
| 123 |
+
"loss": 1.1166,
|
| 124 |
+
"step": 100
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 5.7894736842105265,
|
| 128 |
+
"grad_norm": 11.019944190979004,
|
| 129 |
+
"learning_rate": 4.640468227424749e-05,
|
| 130 |
+
"loss": 1.1004,
|
| 131 |
+
"step": 110
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 6.0,
|
| 135 |
+
"eval_accuracy": 0.7014925373134329,
|
| 136 |
+
"eval_loss": 0.9444876909255981,
|
| 137 |
+
"eval_runtime": 2.1665,
|
| 138 |
+
"eval_samples_per_second": 123.7,
|
| 139 |
+
"eval_steps_per_second": 4.154,
|
| 140 |
+
"step": 114
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 6.315789473684211,
|
| 144 |
+
"grad_norm": 14.164745330810547,
|
| 145 |
+
"learning_rate": 4.55685618729097e-05,
|
| 146 |
+
"loss": 1.0225,
|
| 147 |
+
"step": 120
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 6.842105263157895,
|
| 151 |
+
"grad_norm": 9.702640533447266,
|
| 152 |
+
"learning_rate": 4.473244147157191e-05,
|
| 153 |
+
"loss": 0.9008,
|
| 154 |
+
"step": 130
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 7.0,
|
| 158 |
+
"eval_accuracy": 0.7313432835820896,
|
| 159 |
+
"eval_loss": 0.8700507879257202,
|
| 160 |
+
"eval_runtime": 2.1524,
|
| 161 |
+
"eval_samples_per_second": 124.514,
|
| 162 |
+
"eval_steps_per_second": 4.181,
|
| 163 |
+
"step": 133
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 7.368421052631579,
|
| 167 |
+
"grad_norm": 12.368742942810059,
|
| 168 |
+
"learning_rate": 4.3896321070234115e-05,
|
| 169 |
+
"loss": 0.8022,
|
| 170 |
+
"step": 140
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 7.894736842105263,
|
| 174 |
+
"grad_norm": 16.489498138427734,
|
| 175 |
+
"learning_rate": 4.306020066889632e-05,
|
| 176 |
+
"loss": 0.8023,
|
| 177 |
+
"step": 150
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 8.0,
|
| 181 |
+
"eval_accuracy": 0.7425373134328358,
|
| 182 |
+
"eval_loss": 0.7917194962501526,
|
| 183 |
+
"eval_runtime": 2.1698,
|
| 184 |
+
"eval_samples_per_second": 123.513,
|
| 185 |
+
"eval_steps_per_second": 4.148,
|
| 186 |
+
"step": 152
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 8.421052631578947,
|
| 190 |
+
"grad_norm": 11.644203186035156,
|
| 191 |
+
"learning_rate": 4.222408026755853e-05,
|
| 192 |
+
"loss": 0.8047,
|
| 193 |
+
"step": 160
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 8.947368421052632,
|
| 197 |
+
"grad_norm": 10.630133628845215,
|
| 198 |
+
"learning_rate": 4.1387959866220736e-05,
|
| 199 |
+
"loss": 0.6566,
|
| 200 |
+
"step": 170
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 9.0,
|
| 204 |
+
"eval_accuracy": 0.7611940298507462,
|
| 205 |
+
"eval_loss": 0.6989582777023315,
|
| 206 |
+
"eval_runtime": 2.1317,
|
| 207 |
+
"eval_samples_per_second": 125.721,
|
| 208 |
+
"eval_steps_per_second": 4.222,
|
| 209 |
+
"step": 171
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 9.473684210526315,
|
| 213 |
+
"grad_norm": 9.887234687805176,
|
| 214 |
+
"learning_rate": 4.055183946488295e-05,
|
| 215 |
+
"loss": 0.6759,
|
| 216 |
+
"step": 180
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 10.0,
|
| 220 |
+
"grad_norm": 16.37430191040039,
|
| 221 |
+
"learning_rate": 3.971571906354515e-05,
|
| 222 |
+
"loss": 0.6691,
|
| 223 |
+
"step": 190
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 10.0,
|
| 227 |
+
"eval_accuracy": 0.7985074626865671,
|
| 228 |
+
"eval_loss": 0.6203530430793762,
|
| 229 |
+
"eval_runtime": 2.1249,
|
| 230 |
+
"eval_samples_per_second": 126.122,
|
| 231 |
+
"eval_steps_per_second": 4.235,
|
| 232 |
+
"step": 190
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 10.526315789473685,
|
| 236 |
+
"grad_norm": 12.469463348388672,
|
| 237 |
+
"learning_rate": 3.8879598662207364e-05,
|
| 238 |
+
"loss": 0.5605,
|
| 239 |
+
"step": 200
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 11.0,
|
| 243 |
+
"eval_accuracy": 0.832089552238806,
|
| 244 |
+
"eval_loss": 0.5511359572410583,
|
| 245 |
+
"eval_runtime": 2.1778,
|
| 246 |
+
"eval_samples_per_second": 123.061,
|
| 247 |
+
"eval_steps_per_second": 4.133,
|
| 248 |
+
"step": 209
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 11.052631578947368,
|
| 252 |
+
"grad_norm": 12.518712043762207,
|
| 253 |
+
"learning_rate": 3.804347826086957e-05,
|
| 254 |
+
"loss": 0.5599,
|
| 255 |
+
"step": 210
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 11.578947368421053,
|
| 259 |
+
"grad_norm": 9.64068603515625,
|
| 260 |
+
"learning_rate": 3.7207357859531773e-05,
|
| 261 |
+
"loss": 0.5472,
|
| 262 |
+
"step": 220
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 12.0,
|
| 266 |
+
"eval_accuracy": 0.835820895522388,
|
| 267 |
+
"eval_loss": 0.49448007345199585,
|
| 268 |
+
"eval_runtime": 2.1908,
|
| 269 |
+
"eval_samples_per_second": 122.327,
|
| 270 |
+
"eval_steps_per_second": 4.108,
|
| 271 |
+
"step": 228
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 12.105263157894736,
|
| 275 |
+
"grad_norm": 8.733270645141602,
|
| 276 |
+
"learning_rate": 3.637123745819398e-05,
|
| 277 |
+
"loss": 0.4963,
|
| 278 |
+
"step": 230
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 12.631578947368421,
|
| 282 |
+
"grad_norm": 13.732227325439453,
|
| 283 |
+
"learning_rate": 3.553511705685619e-05,
|
| 284 |
+
"loss": 0.5098,
|
| 285 |
+
"step": 240
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 13.0,
|
| 289 |
+
"eval_accuracy": 0.8619402985074627,
|
| 290 |
+
"eval_loss": 0.4302270710468292,
|
| 291 |
+
"eval_runtime": 2.1349,
|
| 292 |
+
"eval_samples_per_second": 125.533,
|
| 293 |
+
"eval_steps_per_second": 4.216,
|
| 294 |
+
"step": 247
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 13.157894736842104,
|
| 298 |
+
"grad_norm": 8.828410148620605,
|
| 299 |
+
"learning_rate": 3.4698996655518395e-05,
|
| 300 |
+
"loss": 0.439,
|
| 301 |
+
"step": 250
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 13.68421052631579,
|
| 305 |
+
"grad_norm": 11.607008934020996,
|
| 306 |
+
"learning_rate": 3.3862876254180606e-05,
|
| 307 |
+
"loss": 0.4362,
|
| 308 |
+
"step": 260
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 14.0,
|
| 312 |
+
"eval_accuracy": 0.8843283582089553,
|
| 313 |
+
"eval_loss": 0.40266644954681396,
|
| 314 |
+
"eval_runtime": 2.121,
|
| 315 |
+
"eval_samples_per_second": 126.357,
|
| 316 |
+
"eval_steps_per_second": 4.243,
|
| 317 |
+
"step": 266
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 14.210526315789474,
|
| 321 |
+
"grad_norm": 7.250407695770264,
|
| 322 |
+
"learning_rate": 3.302675585284281e-05,
|
| 323 |
+
"loss": 0.4385,
|
| 324 |
+
"step": 270
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 14.736842105263158,
|
| 328 |
+
"grad_norm": 9.448441505432129,
|
| 329 |
+
"learning_rate": 3.2190635451505016e-05,
|
| 330 |
+
"loss": 0.416,
|
| 331 |
+
"step": 280
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 15.0,
|
| 335 |
+
"eval_accuracy": 0.8656716417910447,
|
| 336 |
+
"eval_loss": 0.39556798338890076,
|
| 337 |
+
"eval_runtime": 2.1456,
|
| 338 |
+
"eval_samples_per_second": 124.908,
|
| 339 |
+
"eval_steps_per_second": 4.195,
|
| 340 |
+
"step": 285
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 15.263157894736842,
|
| 344 |
+
"grad_norm": 7.7804274559021,
|
| 345 |
+
"learning_rate": 3.135451505016723e-05,
|
| 346 |
+
"loss": 0.3697,
|
| 347 |
+
"step": 290
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 15.789473684210526,
|
| 351 |
+
"grad_norm": 14.305925369262695,
|
| 352 |
+
"learning_rate": 3.051839464882943e-05,
|
| 353 |
+
"loss": 0.4095,
|
| 354 |
+
"step": 300
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 16.0,
|
| 358 |
+
"eval_accuracy": 0.8880597014925373,
|
| 359 |
+
"eval_loss": 0.3604812026023865,
|
| 360 |
+
"eval_runtime": 2.1246,
|
| 361 |
+
"eval_samples_per_second": 126.144,
|
| 362 |
+
"eval_steps_per_second": 4.236,
|
| 363 |
+
"step": 304
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 16.31578947368421,
|
| 367 |
+
"grad_norm": 11.360480308532715,
|
| 368 |
+
"learning_rate": 2.9682274247491644e-05,
|
| 369 |
+
"loss": 0.3991,
|
| 370 |
+
"step": 310
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 16.842105263157894,
|
| 374 |
+
"grad_norm": 11.096769332885742,
|
| 375 |
+
"learning_rate": 2.8846153846153845e-05,
|
| 376 |
+
"loss": 0.3577,
|
| 377 |
+
"step": 320
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 17.0,
|
| 381 |
+
"eval_accuracy": 0.8917910447761194,
|
| 382 |
+
"eval_loss": 0.3339410424232483,
|
| 383 |
+
"eval_runtime": 2.1207,
|
| 384 |
+
"eval_samples_per_second": 126.373,
|
| 385 |
+
"eval_steps_per_second": 4.244,
|
| 386 |
+
"step": 323
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 17.36842105263158,
|
| 390 |
+
"grad_norm": 9.879475593566895,
|
| 391 |
+
"learning_rate": 2.8010033444816054e-05,
|
| 392 |
+
"loss": 0.291,
|
| 393 |
+
"step": 330
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 17.894736842105264,
|
| 397 |
+
"grad_norm": 15.603042602539062,
|
| 398 |
+
"learning_rate": 2.7173913043478262e-05,
|
| 399 |
+
"loss": 0.3624,
|
| 400 |
+
"step": 340
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 18.0,
|
| 404 |
+
"eval_accuracy": 0.8694029850746269,
|
| 405 |
+
"eval_loss": 0.38826432824134827,
|
| 406 |
+
"eval_runtime": 2.2245,
|
| 407 |
+
"eval_samples_per_second": 120.479,
|
| 408 |
+
"eval_steps_per_second": 4.046,
|
| 409 |
+
"step": 342
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 18.42105263157895,
|
| 413 |
+
"grad_norm": 12.606857299804688,
|
| 414 |
+
"learning_rate": 2.633779264214047e-05,
|
| 415 |
+
"loss": 0.3399,
|
| 416 |
+
"step": 350
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 18.94736842105263,
|
| 420 |
+
"grad_norm": 12.286445617675781,
|
| 421 |
+
"learning_rate": 2.5501672240802675e-05,
|
| 422 |
+
"loss": 0.304,
|
| 423 |
+
"step": 360
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 19.0,
|
| 427 |
+
"eval_accuracy": 0.8768656716417911,
|
| 428 |
+
"eval_loss": 0.3496492803096771,
|
| 429 |
+
"eval_runtime": 2.1504,
|
| 430 |
+
"eval_samples_per_second": 124.629,
|
| 431 |
+
"eval_steps_per_second": 4.185,
|
| 432 |
+
"step": 361
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 19.473684210526315,
|
| 436 |
+
"grad_norm": 8.954345703125,
|
| 437 |
+
"learning_rate": 2.4665551839464883e-05,
|
| 438 |
+
"loss": 0.3156,
|
| 439 |
+
"step": 370
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 20.0,
|
| 443 |
+
"grad_norm": 30.505029678344727,
|
| 444 |
+
"learning_rate": 2.382943143812709e-05,
|
| 445 |
+
"loss": 0.2784,
|
| 446 |
+
"step": 380
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 20.0,
|
| 450 |
+
"eval_accuracy": 0.8805970149253731,
|
| 451 |
+
"eval_loss": 0.32751259207725525,
|
| 452 |
+
"eval_runtime": 2.1365,
|
| 453 |
+
"eval_samples_per_second": 125.441,
|
| 454 |
+
"eval_steps_per_second": 4.213,
|
| 455 |
+
"step": 380
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 20.526315789473685,
|
| 459 |
+
"grad_norm": 6.59121036529541,
|
| 460 |
+
"learning_rate": 2.29933110367893e-05,
|
| 461 |
+
"loss": 0.2763,
|
| 462 |
+
"step": 390
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 21.0,
|
| 466 |
+
"eval_accuracy": 0.8805970149253731,
|
| 467 |
+
"eval_loss": 0.3721315860748291,
|
| 468 |
+
"eval_runtime": 2.1701,
|
| 469 |
+
"eval_samples_per_second": 123.499,
|
| 470 |
+
"eval_steps_per_second": 4.147,
|
| 471 |
+
"step": 399
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 21.05263157894737,
|
| 475 |
+
"grad_norm": 6.725207328796387,
|
| 476 |
+
"learning_rate": 2.2157190635451504e-05,
|
| 477 |
+
"loss": 0.2903,
|
| 478 |
+
"step": 400
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 21.57894736842105,
|
| 482 |
+
"grad_norm": 12.423023223876953,
|
| 483 |
+
"learning_rate": 2.1321070234113713e-05,
|
| 484 |
+
"loss": 0.2824,
|
| 485 |
+
"step": 410
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 22.0,
|
| 489 |
+
"eval_accuracy": 0.8955223880597015,
|
| 490 |
+
"eval_loss": 0.31556689739227295,
|
| 491 |
+
"eval_runtime": 2.1623,
|
| 492 |
+
"eval_samples_per_second": 123.942,
|
| 493 |
+
"eval_steps_per_second": 4.162,
|
| 494 |
+
"step": 418
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 22.105263157894736,
|
| 498 |
+
"grad_norm": 7.236621856689453,
|
| 499 |
+
"learning_rate": 2.048494983277592e-05,
|
| 500 |
+
"loss": 0.2299,
|
| 501 |
+
"step": 420
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 22.63157894736842,
|
| 505 |
+
"grad_norm": 6.276068210601807,
|
| 506 |
+
"learning_rate": 1.964882943143813e-05,
|
| 507 |
+
"loss": 0.2453,
|
| 508 |
+
"step": 430
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 23.0,
|
| 512 |
+
"eval_accuracy": 0.8843283582089553,
|
| 513 |
+
"eval_loss": 0.3155025541782379,
|
| 514 |
+
"eval_runtime": 2.1435,
|
| 515 |
+
"eval_samples_per_second": 125.029,
|
| 516 |
+
"eval_steps_per_second": 4.199,
|
| 517 |
+
"step": 437
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 23.157894736842106,
|
| 521 |
+
"grad_norm": 6.001166820526123,
|
| 522 |
+
"learning_rate": 1.8812709030100337e-05,
|
| 523 |
+
"loss": 0.248,
|
| 524 |
+
"step": 440
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 23.68421052631579,
|
| 528 |
+
"grad_norm": 9.594138145446777,
|
| 529 |
+
"learning_rate": 1.7976588628762542e-05,
|
| 530 |
+
"loss": 0.2438,
|
| 531 |
+
"step": 450
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 24.0,
|
| 535 |
+
"eval_accuracy": 0.9029850746268657,
|
| 536 |
+
"eval_loss": 0.2927631437778473,
|
| 537 |
+
"eval_runtime": 2.2056,
|
| 538 |
+
"eval_samples_per_second": 121.509,
|
| 539 |
+
"eval_steps_per_second": 4.081,
|
| 540 |
+
"step": 456
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 24.210526315789473,
|
| 544 |
+
"grad_norm": 12.752880096435547,
|
| 545 |
+
"learning_rate": 1.714046822742475e-05,
|
| 546 |
+
"loss": 0.2472,
|
| 547 |
+
"step": 460
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 24.736842105263158,
|
| 551 |
+
"grad_norm": 5.313000202178955,
|
| 552 |
+
"learning_rate": 1.630434782608696e-05,
|
| 553 |
+
"loss": 0.2285,
|
| 554 |
+
"step": 470
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 25.0,
|
| 558 |
+
"eval_accuracy": 0.9216417910447762,
|
| 559 |
+
"eval_loss": 0.26671093702316284,
|
| 560 |
+
"eval_runtime": 2.2227,
|
| 561 |
+
"eval_samples_per_second": 120.573,
|
| 562 |
+
"eval_steps_per_second": 4.049,
|
| 563 |
+
"step": 475
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 25.263157894736842,
|
| 567 |
+
"grad_norm": 8.309535026550293,
|
| 568 |
+
"learning_rate": 1.5468227424749167e-05,
|
| 569 |
+
"loss": 0.2555,
|
| 570 |
+
"step": 480
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 25.789473684210527,
|
| 574 |
+
"grad_norm": 7.868557453155518,
|
| 575 |
+
"learning_rate": 1.4632107023411373e-05,
|
| 576 |
+
"loss": 0.2478,
|
| 577 |
+
"step": 490
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 26.0,
|
| 581 |
+
"eval_accuracy": 0.914179104477612,
|
| 582 |
+
"eval_loss": 0.2815816402435303,
|
| 583 |
+
"eval_runtime": 2.1383,
|
| 584 |
+
"eval_samples_per_second": 125.336,
|
| 585 |
+
"eval_steps_per_second": 4.209,
|
| 586 |
+
"step": 494
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 26.31578947368421,
|
| 590 |
+
"grad_norm": 6.559621334075928,
|
| 591 |
+
"learning_rate": 1.3795986622073578e-05,
|
| 592 |
+
"loss": 0.2601,
|
| 593 |
+
"step": 500
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 26.842105263157894,
|
| 597 |
+
"grad_norm": 7.318304538726807,
|
| 598 |
+
"learning_rate": 1.2959866220735786e-05,
|
| 599 |
+
"loss": 0.2242,
|
| 600 |
+
"step": 510
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 27.0,
|
| 604 |
+
"eval_accuracy": 0.8992537313432836,
|
| 605 |
+
"eval_loss": 0.2767724096775055,
|
| 606 |
+
"eval_runtime": 2.1721,
|
| 607 |
+
"eval_samples_per_second": 123.385,
|
| 608 |
+
"eval_steps_per_second": 4.144,
|
| 609 |
+
"step": 513
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 27.36842105263158,
|
| 613 |
+
"grad_norm": 9.232494354248047,
|
| 614 |
+
"learning_rate": 1.2123745819397994e-05,
|
| 615 |
+
"loss": 0.1988,
|
| 616 |
+
"step": 520
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 27.894736842105264,
|
| 620 |
+
"grad_norm": 6.716543197631836,
|
| 621 |
+
"learning_rate": 1.1287625418060201e-05,
|
| 622 |
+
"loss": 0.2,
|
| 623 |
+
"step": 530
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 28.0,
|
| 627 |
+
"eval_accuracy": 0.914179104477612,
|
| 628 |
+
"eval_loss": 0.2814846932888031,
|
| 629 |
+
"eval_runtime": 2.2098,
|
| 630 |
+
"eval_samples_per_second": 121.279,
|
| 631 |
+
"eval_steps_per_second": 4.073,
|
| 632 |
+
"step": 532
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 28.42105263157895,
|
| 636 |
+
"grad_norm": 9.893804550170898,
|
| 637 |
+
"learning_rate": 1.045150501672241e-05,
|
| 638 |
+
"loss": 0.2091,
|
| 639 |
+
"step": 540
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 28.94736842105263,
|
| 643 |
+
"grad_norm": 9.348865509033203,
|
| 644 |
+
"learning_rate": 9.615384615384616e-06,
|
| 645 |
+
"loss": 0.2076,
|
| 646 |
+
"step": 550
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 29.0,
|
| 650 |
+
"eval_accuracy": 0.9216417910447762,
|
| 651 |
+
"eval_loss": 0.24431268870830536,
|
| 652 |
+
"eval_runtime": 2.148,
|
| 653 |
+
"eval_samples_per_second": 124.765,
|
| 654 |
+
"eval_steps_per_second": 4.19,
|
| 655 |
+
"step": 551
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 29.473684210526315,
|
| 659 |
+
"grad_norm": 8.77604866027832,
|
| 660 |
+
"learning_rate": 8.779264214046822e-06,
|
| 661 |
+
"loss": 0.2026,
|
| 662 |
+
"step": 560
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 30.0,
|
| 666 |
+
"grad_norm": 14.116354942321777,
|
| 667 |
+
"learning_rate": 7.94314381270903e-06,
|
| 668 |
+
"loss": 0.1978,
|
| 669 |
+
"step": 570
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 30.0,
|
| 673 |
+
"eval_accuracy": 0.9216417910447762,
|
| 674 |
+
"eval_loss": 0.23809637129306793,
|
| 675 |
+
"eval_runtime": 2.1703,
|
| 676 |
+
"eval_samples_per_second": 123.486,
|
| 677 |
+
"eval_steps_per_second": 4.147,
|
| 678 |
+
"step": 570
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 30.526315789473685,
|
| 682 |
+
"grad_norm": 11.3234281539917,
|
| 683 |
+
"learning_rate": 7.107023411371237e-06,
|
| 684 |
+
"loss": 0.1821,
|
| 685 |
+
"step": 580
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 31.0,
|
| 689 |
+
"eval_accuracy": 0.9216417910447762,
|
| 690 |
+
"eval_loss": 0.25633105635643005,
|
| 691 |
+
"eval_runtime": 2.1804,
|
| 692 |
+
"eval_samples_per_second": 122.916,
|
| 693 |
+
"eval_steps_per_second": 4.128,
|
| 694 |
+
"step": 589
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 31.05263157894737,
|
| 698 |
+
"grad_norm": 7.66957950592041,
|
| 699 |
+
"learning_rate": 6.270903010033445e-06,
|
| 700 |
+
"loss": 0.2099,
|
| 701 |
+
"step": 590
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 31.57894736842105,
|
| 705 |
+
"grad_norm": 5.773667335510254,
|
| 706 |
+
"learning_rate": 5.4347826086956525e-06,
|
| 707 |
+
"loss": 0.1786,
|
| 708 |
+
"step": 600
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 32.0,
|
| 712 |
+
"eval_accuracy": 0.9253731343283582,
|
| 713 |
+
"eval_loss": 0.24489234387874603,
|
| 714 |
+
"eval_runtime": 2.2643,
|
| 715 |
+
"eval_samples_per_second": 118.359,
|
| 716 |
+
"eval_steps_per_second": 3.975,
|
| 717 |
+
"step": 608
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 32.10526315789474,
|
| 721 |
+
"grad_norm": 5.752429008483887,
|
| 722 |
+
"learning_rate": 4.598662207357859e-06,
|
| 723 |
+
"loss": 0.1785,
|
| 724 |
+
"step": 610
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 32.63157894736842,
|
| 728 |
+
"grad_norm": 9.643966674804688,
|
| 729 |
+
"learning_rate": 3.7625418060200673e-06,
|
| 730 |
+
"loss": 0.1809,
|
| 731 |
+
"step": 620
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 33.0,
|
| 735 |
+
"eval_accuracy": 0.9328358208955224,
|
| 736 |
+
"eval_loss": 0.23848846554756165,
|
| 737 |
+
"eval_runtime": 2.3814,
|
| 738 |
+
"eval_samples_per_second": 112.539,
|
| 739 |
+
"eval_steps_per_second": 3.779,
|
| 740 |
+
"step": 627
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 33.1578947368421,
|
| 744 |
+
"grad_norm": 8.57278823852539,
|
| 745 |
+
"learning_rate": 2.9264214046822746e-06,
|
| 746 |
+
"loss": 0.1825,
|
| 747 |
+
"step": 630
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 33.68421052631579,
|
| 751 |
+
"grad_norm": 6.0924296379089355,
|
| 752 |
+
"learning_rate": 2.0903010033444816e-06,
|
| 753 |
+
"loss": 0.1812,
|
| 754 |
+
"step": 640
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 34.0,
|
| 758 |
+
"eval_accuracy": 0.9291044776119403,
|
| 759 |
+
"eval_loss": 0.2448083907365799,
|
| 760 |
+
"eval_runtime": 2.3113,
|
| 761 |
+
"eval_samples_per_second": 115.954,
|
| 762 |
+
"eval_steps_per_second": 3.894,
|
| 763 |
+
"step": 646
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 34.21052631578947,
|
| 767 |
+
"grad_norm": 11.392114639282227,
|
| 768 |
+
"learning_rate": 1.254180602006689e-06,
|
| 769 |
+
"loss": 0.2006,
|
| 770 |
+
"step": 650
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 34.73684210526316,
|
| 774 |
+
"grad_norm": 7.01861047744751,
|
| 775 |
+
"learning_rate": 4.180602006688963e-07,
|
| 776 |
+
"loss": 0.1688,
|
| 777 |
+
"step": 660
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 35.0,
|
| 781 |
+
"eval_accuracy": 0.9291044776119403,
|
| 782 |
+
"eval_loss": 0.24462540447711945,
|
| 783 |
+
"eval_runtime": 2.2089,
|
| 784 |
+
"eval_samples_per_second": 121.33,
|
| 785 |
+
"eval_steps_per_second": 4.074,
|
| 786 |
+
"step": 665
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 35.0,
|
| 790 |
+
"step": 665,
|
| 791 |
+
"total_flos": 2.0920841725722624e+18,
|
| 792 |
+
"train_loss": 0.6001271831361871,
|
| 793 |
+
"train_runtime": 1153.5131,
|
| 794 |
+
"train_samples_per_second": 72.942,
|
| 795 |
+
"train_steps_per_second": 0.576
|
| 796 |
}
|
| 797 |
],
|
| 798 |
"logging_steps": 10,
|
| 799 |
+
"max_steps": 665,
|
| 800 |
"num_input_tokens_seen": 0,
|
| 801 |
+
"num_train_epochs": 35,
|
| 802 |
"save_steps": 500,
|
| 803 |
"stateful_callbacks": {
|
| 804 |
"CustomEarlyStoppingCallback": {
|
|
|
|
| 821 |
"attributes": {}
|
| 822 |
}
|
| 823 |
},
|
| 824 |
+
"total_flos": 2.0920841725722624e+18,
|
| 825 |
"train_batch_size": 32,
|
| 826 |
"trial_name": null,
|
| 827 |
"trial_params": null
|