Model save
Browse files- README.md +23 -23
- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +430 -0
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -15,8 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model was trained from scratch on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.8055
|
| 19 |
- Accuracy: 0.7597
|
|
|
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
@@ -51,28 +51,28 @@ The following hyperparameters were used during training:
|
|
| 51 |
|
| 52 |
### Training results
|
| 53 |
|
| 54 |
-
| Training Loss | Epoch | Step | Validation Loss |
|
| 55 |
-
|
| 56 |
-
| 0.5411 | 1.0 | 187 | 0.
|
| 57 |
-
| 0.362 | 2.0 | 375 | 0.
|
| 58 |
-
| 0.3047 | 3.0 | 562 | 0.
|
| 59 |
-
| 0.2469 | 4.0 | 750 | 0.
|
| 60 |
-
| 0.2022 | 5.0 | 937 | 0.
|
| 61 |
-
| 0.1681 | 6.0 | 1125 | 0.
|
| 62 |
-
| 0.1404 | 7.0 | 1312 | 0.
|
| 63 |
-
| 0.1197 | 8.0 | 1500 | 0.
|
| 64 |
-
| 0.1 | 9.0 | 1687 | 0.
|
| 65 |
-
| 0.089 | 10.0 | 1875 | 0.
|
| 66 |
-
| 0.0837 | 11.0 | 2062 | 0.
|
| 67 |
-
| 0.0788 | 12.0 | 2250 | 0.
|
| 68 |
-
| 0.0761 | 13.0 | 2437 | 0.
|
| 69 |
-
| 0.0749 | 14.0 | 2625 | 0.
|
| 70 |
-
| 0.0741 | 15.0 | 2812 | 0.
|
| 71 |
-
| 0.0726 | 16.0 | 3000 | 0.
|
| 72 |
-
| 0.0687 | 17.0 | 3187 | 0.
|
| 73 |
-
| 0.0682 | 18.0 | 3375 | 0.
|
| 74 |
-
| 0.0682 | 19.0 | 3562 | 0.
|
| 75 |
-
| 0.0682 | 19.95 | 3740 | 0.
|
| 76 |
|
| 77 |
|
| 78 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model was trained from scratch on an unknown dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
|
|
|
| 18 |
- Accuracy: 0.7597
|
| 19 |
+
- Loss: 0.8055
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 51 |
|
| 52 |
### Training results
|
| 53 |
|
| 54 |
+
| Training Loss | Epoch | Step | Accuracy | Validation Loss |
|
| 55 |
+
|:-------------:|:-----:|:----:|:--------:|:---------------:|
|
| 56 |
+
| 0.5411 | 1.0 | 187 | 0.7904 | 0.3938 |
|
| 57 |
+
| 0.362 | 2.0 | 375 | 0.7918 | 0.3804 |
|
| 58 |
+
| 0.3047 | 3.0 | 562 | 0.7891 | 0.3934 |
|
| 59 |
+
| 0.2469 | 4.0 | 750 | 0.7846 | 0.4226 |
|
| 60 |
+
| 0.2022 | 5.0 | 937 | 0.7803 | 0.4661 |
|
| 61 |
+
| 0.1681 | 6.0 | 1125 | 0.7761 | 0.5123 |
|
| 62 |
+
| 0.1404 | 7.0 | 1312 | 0.7721 | 0.5731 |
|
| 63 |
+
| 0.1197 | 8.0 | 1500 | 0.7701 | 0.6075 |
|
| 64 |
+
| 0.1 | 9.0 | 1687 | 0.7688 | 0.6317 |
|
| 65 |
+
| 0.089 | 10.0 | 1875 | 0.7664 | 0.6718 |
|
| 66 |
+
| 0.0837 | 11.0 | 2062 | 0.7653 | 0.6922 |
|
| 67 |
+
| 0.0788 | 12.0 | 2250 | 0.7632 | 0.7254 |
|
| 68 |
+
| 0.0761 | 13.0 | 2437 | 0.7629 | 0.7256 |
|
| 69 |
+
| 0.0749 | 14.0 | 2625 | 0.7621 | 0.7534 |
|
| 70 |
+
| 0.0741 | 15.0 | 2812 | 0.7620 | 0.7529 |
|
| 71 |
+
| 0.0726 | 16.0 | 3000 | 0.7611 | 0.7678 |
|
| 72 |
+
| 0.0687 | 17.0 | 3187 | 0.7610 | 0.7728 |
|
| 73 |
+
| 0.0682 | 18.0 | 3375 | 0.7603 | 0.7807 |
|
| 74 |
+
| 0.0682 | 19.0 | 3562 | 0.7610 | 0.7872 |
|
| 75 |
+
| 0.0682 | 19.95 | 3740 | 0.7597 | 0.8055 |
|
| 76 |
|
| 77 |
|
| 78 |
### Framework versions
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.95,
|
| 3 |
+
"eval_accuracy": 0.7597490196078431,
|
| 4 |
+
"eval_loss": 0.8054783940315247,
|
| 5 |
+
"eval_runtime": 6.7173,
|
| 6 |
+
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 74.434,
|
| 8 |
+
"eval_steps_per_second": 9.379,
|
| 9 |
+
"perplexity": 2.237766777062304,
|
| 10 |
+
"train_loss": 0.1472176224152672,
|
| 11 |
+
"train_runtime": 2965.7973,
|
| 12 |
+
"train_samples": 6000,
|
| 13 |
+
"train_samples_per_second": 40.461,
|
| 14 |
+
"train_steps_per_second": 1.261
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.95,
|
| 3 |
+
"eval_accuracy": 0.7597490196078431,
|
| 4 |
+
"eval_loss": 0.8054783940315247,
|
| 5 |
+
"eval_runtime": 6.7173,
|
| 6 |
+
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 74.434,
|
| 8 |
+
"eval_steps_per_second": 9.379,
|
| 9 |
+
"perplexity": 2.237766777062304
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.95,
|
| 3 |
+
"train_loss": 0.1472176224152672,
|
| 4 |
+
"train_runtime": 2965.7973,
|
| 5 |
+
"train_samples": 6000,
|
| 6 |
+
"train_samples_per_second": 40.461,
|
| 7 |
+
"train_steps_per_second": 1.261
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,430 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.946666666666665,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3740,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.53,
|
| 13 |
+
"learning_rate": 0.0001,
|
| 14 |
+
"loss": 0.5411,
|
| 15 |
+
"step": 100
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.0,
|
| 19 |
+
"eval_accuracy": 0.7903686274509804,
|
| 20 |
+
"eval_loss": 0.39375776052474976,
|
| 21 |
+
"eval_runtime": 7.7091,
|
| 22 |
+
"eval_samples_per_second": 64.858,
|
| 23 |
+
"eval_steps_per_second": 8.172,
|
| 24 |
+
"step": 187
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.07,
|
| 28 |
+
"learning_rate": 0.0001,
|
| 29 |
+
"loss": 0.4081,
|
| 30 |
+
"step": 200
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.6,
|
| 34 |
+
"learning_rate": 0.0001,
|
| 35 |
+
"loss": 0.362,
|
| 36 |
+
"step": 300
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 2.0,
|
| 40 |
+
"eval_accuracy": 0.7917960784313726,
|
| 41 |
+
"eval_loss": 0.3804325461387634,
|
| 42 |
+
"eval_runtime": 7.4651,
|
| 43 |
+
"eval_samples_per_second": 66.979,
|
| 44 |
+
"eval_steps_per_second": 8.439,
|
| 45 |
+
"step": 375
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.13,
|
| 49 |
+
"learning_rate": 0.0001,
|
| 50 |
+
"loss": 0.3376,
|
| 51 |
+
"step": 400
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 2.67,
|
| 55 |
+
"learning_rate": 0.0001,
|
| 56 |
+
"loss": 0.3047,
|
| 57 |
+
"step": 500
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.0,
|
| 61 |
+
"eval_accuracy": 0.7891450980392157,
|
| 62 |
+
"eval_loss": 0.39341700077056885,
|
| 63 |
+
"eval_runtime": 7.1191,
|
| 64 |
+
"eval_samples_per_second": 70.233,
|
| 65 |
+
"eval_steps_per_second": 8.849,
|
| 66 |
+
"step": 562
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 3.2,
|
| 70 |
+
"learning_rate": 0.0001,
|
| 71 |
+
"loss": 0.2848,
|
| 72 |
+
"step": 600
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 3.73,
|
| 76 |
+
"learning_rate": 0.0001,
|
| 77 |
+
"loss": 0.2469,
|
| 78 |
+
"step": 700
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 4.0,
|
| 82 |
+
"eval_accuracy": 0.7846274509803921,
|
| 83 |
+
"eval_loss": 0.42259183526039124,
|
| 84 |
+
"eval_runtime": 6.9941,
|
| 85 |
+
"eval_samples_per_second": 71.489,
|
| 86 |
+
"eval_steps_per_second": 9.008,
|
| 87 |
+
"step": 750
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 4.27,
|
| 91 |
+
"learning_rate": 0.0001,
|
| 92 |
+
"loss": 0.2215,
|
| 93 |
+
"step": 800
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 4.8,
|
| 97 |
+
"learning_rate": 0.0001,
|
| 98 |
+
"loss": 0.2022,
|
| 99 |
+
"step": 900
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 5.0,
|
| 103 |
+
"eval_accuracy": 0.7803058823529412,
|
| 104 |
+
"eval_loss": 0.4660645127296448,
|
| 105 |
+
"eval_runtime": 6.6511,
|
| 106 |
+
"eval_samples_per_second": 75.175,
|
| 107 |
+
"eval_steps_per_second": 9.472,
|
| 108 |
+
"step": 937
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 5.33,
|
| 112 |
+
"learning_rate": 0.0001,
|
| 113 |
+
"loss": 0.1783,
|
| 114 |
+
"step": 1000
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 5.87,
|
| 118 |
+
"learning_rate": 0.0001,
|
| 119 |
+
"loss": 0.1681,
|
| 120 |
+
"step": 1100
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 6.0,
|
| 124 |
+
"eval_accuracy": 0.7761333333333333,
|
| 125 |
+
"eval_loss": 0.5122596621513367,
|
| 126 |
+
"eval_runtime": 7.6517,
|
| 127 |
+
"eval_samples_per_second": 65.345,
|
| 128 |
+
"eval_steps_per_second": 8.233,
|
| 129 |
+
"step": 1125
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 6.4,
|
| 133 |
+
"learning_rate": 0.0001,
|
| 134 |
+
"loss": 0.1437,
|
| 135 |
+
"step": 1200
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 6.93,
|
| 139 |
+
"learning_rate": 0.0001,
|
| 140 |
+
"loss": 0.1404,
|
| 141 |
+
"step": 1300
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 7.0,
|
| 145 |
+
"eval_accuracy": 0.7720627450980392,
|
| 146 |
+
"eval_loss": 0.5730750560760498,
|
| 147 |
+
"eval_runtime": 7.0342,
|
| 148 |
+
"eval_samples_per_second": 71.081,
|
| 149 |
+
"eval_steps_per_second": 8.956,
|
| 150 |
+
"step": 1312
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 7.47,
|
| 154 |
+
"learning_rate": 0.0001,
|
| 155 |
+
"loss": 0.1179,
|
| 156 |
+
"step": 1400
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 8.0,
|
| 160 |
+
"learning_rate": 0.0001,
|
| 161 |
+
"loss": 0.1197,
|
| 162 |
+
"step": 1500
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 8.0,
|
| 166 |
+
"eval_accuracy": 0.770078431372549,
|
| 167 |
+
"eval_loss": 0.6074877977371216,
|
| 168 |
+
"eval_runtime": 6.9934,
|
| 169 |
+
"eval_samples_per_second": 71.496,
|
| 170 |
+
"eval_steps_per_second": 9.008,
|
| 171 |
+
"step": 1500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 8.53,
|
| 175 |
+
"learning_rate": 0.0001,
|
| 176 |
+
"loss": 0.1,
|
| 177 |
+
"step": 1600
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 9.0,
|
| 181 |
+
"eval_accuracy": 0.7688470588235294,
|
| 182 |
+
"eval_loss": 0.631741464138031,
|
| 183 |
+
"eval_runtime": 7.3864,
|
| 184 |
+
"eval_samples_per_second": 67.692,
|
| 185 |
+
"eval_steps_per_second": 8.529,
|
| 186 |
+
"step": 1687
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 9.07,
|
| 190 |
+
"learning_rate": 0.0001,
|
| 191 |
+
"loss": 0.1027,
|
| 192 |
+
"step": 1700
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 9.6,
|
| 196 |
+
"learning_rate": 0.0001,
|
| 197 |
+
"loss": 0.089,
|
| 198 |
+
"step": 1800
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 10.0,
|
| 202 |
+
"eval_accuracy": 0.7663764705882353,
|
| 203 |
+
"eval_loss": 0.6717921495437622,
|
| 204 |
+
"eval_runtime": 6.6214,
|
| 205 |
+
"eval_samples_per_second": 75.513,
|
| 206 |
+
"eval_steps_per_second": 9.515,
|
| 207 |
+
"step": 1875
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 10.13,
|
| 211 |
+
"learning_rate": 0.0001,
|
| 212 |
+
"loss": 0.0911,
|
| 213 |
+
"step": 1900
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 10.67,
|
| 217 |
+
"learning_rate": 0.0001,
|
| 218 |
+
"loss": 0.0837,
|
| 219 |
+
"step": 2000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 11.0,
|
| 223 |
+
"eval_accuracy": 0.7653333333333333,
|
| 224 |
+
"eval_loss": 0.6921772956848145,
|
| 225 |
+
"eval_runtime": 6.6402,
|
| 226 |
+
"eval_samples_per_second": 75.299,
|
| 227 |
+
"eval_steps_per_second": 9.488,
|
| 228 |
+
"step": 2062
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 11.2,
|
| 232 |
+
"learning_rate": 0.0001,
|
| 233 |
+
"loss": 0.0823,
|
| 234 |
+
"step": 2100
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 11.73,
|
| 238 |
+
"learning_rate": 0.0001,
|
| 239 |
+
"loss": 0.0788,
|
| 240 |
+
"step": 2200
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 12.0,
|
| 244 |
+
"eval_accuracy": 0.7631764705882353,
|
| 245 |
+
"eval_loss": 0.7253576517105103,
|
| 246 |
+
"eval_runtime": 7.6814,
|
| 247 |
+
"eval_samples_per_second": 65.092,
|
| 248 |
+
"eval_steps_per_second": 8.202,
|
| 249 |
+
"step": 2250
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 12.27,
|
| 253 |
+
"learning_rate": 0.0001,
|
| 254 |
+
"loss": 0.078,
|
| 255 |
+
"step": 2300
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 12.8,
|
| 259 |
+
"learning_rate": 0.0001,
|
| 260 |
+
"loss": 0.0761,
|
| 261 |
+
"step": 2400
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 13.0,
|
| 265 |
+
"eval_accuracy": 0.7628627450980392,
|
| 266 |
+
"eval_loss": 0.7256088256835938,
|
| 267 |
+
"eval_runtime": 7.6806,
|
| 268 |
+
"eval_samples_per_second": 65.099,
|
| 269 |
+
"eval_steps_per_second": 8.202,
|
| 270 |
+
"step": 2437
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 13.33,
|
| 274 |
+
"learning_rate": 0.0001,
|
| 275 |
+
"loss": 0.0743,
|
| 276 |
+
"step": 2500
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 13.87,
|
| 280 |
+
"learning_rate": 0.0001,
|
| 281 |
+
"loss": 0.0749,
|
| 282 |
+
"step": 2600
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 14.0,
|
| 286 |
+
"eval_accuracy": 0.7620627450980392,
|
| 287 |
+
"eval_loss": 0.7533740401268005,
|
| 288 |
+
"eval_runtime": 7.644,
|
| 289 |
+
"eval_samples_per_second": 65.41,
|
| 290 |
+
"eval_steps_per_second": 8.242,
|
| 291 |
+
"step": 2625
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 14.4,
|
| 295 |
+
"learning_rate": 0.0001,
|
| 296 |
+
"loss": 0.0719,
|
| 297 |
+
"step": 2700
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 14.93,
|
| 301 |
+
"learning_rate": 0.0001,
|
| 302 |
+
"loss": 0.0741,
|
| 303 |
+
"step": 2800
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 15.0,
|
| 307 |
+
"eval_accuracy": 0.7620078431372549,
|
| 308 |
+
"eval_loss": 0.7529163360595703,
|
| 309 |
+
"eval_runtime": 7.4208,
|
| 310 |
+
"eval_samples_per_second": 67.378,
|
| 311 |
+
"eval_steps_per_second": 8.49,
|
| 312 |
+
"step": 2812
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 15.47,
|
| 316 |
+
"learning_rate": 0.0001,
|
| 317 |
+
"loss": 0.0704,
|
| 318 |
+
"step": 2900
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 16.0,
|
| 322 |
+
"learning_rate": 0.0001,
|
| 323 |
+
"loss": 0.0726,
|
| 324 |
+
"step": 3000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 16.0,
|
| 328 |
+
"eval_accuracy": 0.7610745098039216,
|
| 329 |
+
"eval_loss": 0.7678206562995911,
|
| 330 |
+
"eval_runtime": 6.6228,
|
| 331 |
+
"eval_samples_per_second": 75.497,
|
| 332 |
+
"eval_steps_per_second": 9.513,
|
| 333 |
+
"step": 3000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 16.53,
|
| 337 |
+
"learning_rate": 0.0001,
|
| 338 |
+
"loss": 0.0687,
|
| 339 |
+
"step": 3100
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 17.0,
|
| 343 |
+
"eval_accuracy": 0.7610274509803922,
|
| 344 |
+
"eval_loss": 0.7728469371795654,
|
| 345 |
+
"eval_runtime": 7.7738,
|
| 346 |
+
"eval_samples_per_second": 64.319,
|
| 347 |
+
"eval_steps_per_second": 8.104,
|
| 348 |
+
"step": 3187
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 17.07,
|
| 352 |
+
"learning_rate": 0.0001,
|
| 353 |
+
"loss": 0.0706,
|
| 354 |
+
"step": 3200
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 17.6,
|
| 358 |
+
"learning_rate": 0.0001,
|
| 359 |
+
"loss": 0.0682,
|
| 360 |
+
"step": 3300
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 18.0,
|
| 364 |
+
"eval_accuracy": 0.7603058823529412,
|
| 365 |
+
"eval_loss": 0.7807328104972839,
|
| 366 |
+
"eval_runtime": 7.6425,
|
| 367 |
+
"eval_samples_per_second": 65.424,
|
| 368 |
+
"eval_steps_per_second": 8.243,
|
| 369 |
+
"step": 3375
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 18.13,
|
| 373 |
+
"learning_rate": 0.0001,
|
| 374 |
+
"loss": 0.069,
|
| 375 |
+
"step": 3400
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 18.67,
|
| 379 |
+
"learning_rate": 0.0001,
|
| 380 |
+
"loss": 0.0682,
|
| 381 |
+
"step": 3500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 19.0,
|
| 385 |
+
"eval_accuracy": 0.7609882352941176,
|
| 386 |
+
"eval_loss": 0.7872016429901123,
|
| 387 |
+
"eval_runtime": 7.0778,
|
| 388 |
+
"eval_samples_per_second": 70.644,
|
| 389 |
+
"eval_steps_per_second": 8.901,
|
| 390 |
+
"step": 3562
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 19.2,
|
| 394 |
+
"learning_rate": 0.0001,
|
| 395 |
+
"loss": 0.0682,
|
| 396 |
+
"step": 3600
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 19.73,
|
| 400 |
+
"learning_rate": 0.0001,
|
| 401 |
+
"loss": 0.0682,
|
| 402 |
+
"step": 3700
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 19.95,
|
| 406 |
+
"eval_accuracy": 0.7597490196078431,
|
| 407 |
+
"eval_loss": 0.8054783940315247,
|
| 408 |
+
"eval_runtime": 7.0912,
|
| 409 |
+
"eval_samples_per_second": 70.51,
|
| 410 |
+
"eval_steps_per_second": 8.884,
|
| 411 |
+
"step": 3740
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 19.95,
|
| 415 |
+
"step": 3740,
|
| 416 |
+
"total_flos": 8.667972359988183e+17,
|
| 417 |
+
"train_loss": 0.1472176224152672,
|
| 418 |
+
"train_runtime": 2965.7973,
|
| 419 |
+
"train_samples_per_second": 40.461,
|
| 420 |
+
"train_steps_per_second": 1.261
|
| 421 |
+
}
|
| 422 |
+
],
|
| 423 |
+
"logging_steps": 100,
|
| 424 |
+
"max_steps": 3740,
|
| 425 |
+
"num_train_epochs": 20,
|
| 426 |
+
"save_steps": 500,
|
| 427 |
+
"total_flos": 8.667972359988183e+17,
|
| 428 |
+
"trial_name": null,
|
| 429 |
+
"trial_params": null
|
| 430 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4728
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69229a474ae3be2188b1040a5433091508861cfca3d28c831620e7b24bd5c1d9
|
| 3 |
size 4728
|