Commit
•
c3d9125
1
Parent(s):
a23add9
End of training
Browse files- README.md +1 -1
- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Apr19_02-04-45_c63135b2894f/events.out.tfevents.1713493957.c63135b2894f.1649.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +540 -0
README.md
CHANGED
|
@@ -17,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 1.
|
| 21 |
- Accuracy: 0.5667
|
| 22 |
|
| 23 |
## Model description
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 1.1112
|
| 21 |
- Accuracy: 0.5667
|
| 22 |
|
| 23 |
## Model description
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 37.94,
|
| 3 |
+
"eval_accuracy": 0.5666666666666667,
|
| 4 |
+
"eval_loss": 1.111220121383667,
|
| 5 |
+
"eval_runtime": 1.5578,
|
| 6 |
+
"eval_samples_per_second": 38.515,
|
| 7 |
+
"eval_steps_per_second": 1.284,
|
| 8 |
+
"total_flos": 2.864620236542755e+18,
|
| 9 |
+
"train_loss": 1.2232356217442726,
|
| 10 |
+
"train_runtime": 1646.5284,
|
| 11 |
+
"train_samples_per_second": 24.845,
|
| 12 |
+
"train_steps_per_second": 0.179
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 37.94,
|
| 3 |
+
"eval_accuracy": 0.5666666666666667,
|
| 4 |
+
"eval_loss": 1.111220121383667,
|
| 5 |
+
"eval_runtime": 1.5578,
|
| 6 |
+
"eval_samples_per_second": 38.515,
|
| 7 |
+
"eval_steps_per_second": 1.284
|
| 8 |
+
}
|
runs/Apr19_02-04-45_c63135b2894f/events.out.tfevents.1713493957.c63135b2894f.1649.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e368deab499d6ffda07802f24c2d1b22c3526480afbd62c693219fab68b5225
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 37.94,
|
| 3 |
+
"total_flos": 2.864620236542755e+18,
|
| 4 |
+
"train_loss": 1.2232356217442726,
|
| 5 |
+
"train_runtime": 1646.5284,
|
| 6 |
+
"train_samples_per_second": 24.845,
|
| 7 |
+
"train_steps_per_second": 0.179
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,540 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5666666666666667,
|
| 3 |
+
"best_model_checkpoint": "vit-base-patch16-224-dmae-va-U5-42C/checkpoint-232",
|
| 4 |
+
"epoch": 37.935483870967744,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 294,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.9,
|
| 13 |
+
"eval_accuracy": 0.13333333333333333,
|
| 14 |
+
"eval_loss": 1.4545705318450928,
|
| 15 |
+
"eval_runtime": 2.7807,
|
| 16 |
+
"eval_samples_per_second": 21.578,
|
| 17 |
+
"eval_steps_per_second": 0.719,
|
| 18 |
+
"step": 7
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 1.55,
|
| 22 |
+
"grad_norm": 5.473387241363525,
|
| 23 |
+
"learning_rate": 4e-07,
|
| 24 |
+
"loss": 1.5342,
|
| 25 |
+
"step": 12
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.94,
|
| 29 |
+
"eval_accuracy": 0.13333333333333333,
|
| 30 |
+
"eval_loss": 1.4379044771194458,
|
| 31 |
+
"eval_runtime": 1.5117,
|
| 32 |
+
"eval_samples_per_second": 39.69,
|
| 33 |
+
"eval_steps_per_second": 1.323,
|
| 34 |
+
"step": 15
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.97,
|
| 38 |
+
"eval_accuracy": 0.16666666666666666,
|
| 39 |
+
"eval_loss": 1.411478042602539,
|
| 40 |
+
"eval_runtime": 1.4724,
|
| 41 |
+
"eval_samples_per_second": 40.749,
|
| 42 |
+
"eval_steps_per_second": 1.358,
|
| 43 |
+
"step": 23
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 3.1,
|
| 47 |
+
"grad_norm": 6.9020280838012695,
|
| 48 |
+
"learning_rate": 8e-07,
|
| 49 |
+
"loss": 1.5331,
|
| 50 |
+
"step": 24
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 4.0,
|
| 54 |
+
"eval_accuracy": 0.2,
|
| 55 |
+
"eval_loss": 1.3786956071853638,
|
| 56 |
+
"eval_runtime": 1.4845,
|
| 57 |
+
"eval_samples_per_second": 40.418,
|
| 58 |
+
"eval_steps_per_second": 1.347,
|
| 59 |
+
"step": 31
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 4.65,
|
| 63 |
+
"grad_norm": 4.692047119140625,
|
| 64 |
+
"learning_rate": 9.772727272727273e-07,
|
| 65 |
+
"loss": 1.4639,
|
| 66 |
+
"step": 36
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 4.9,
|
| 70 |
+
"eval_accuracy": 0.2833333333333333,
|
| 71 |
+
"eval_loss": 1.351299524307251,
|
| 72 |
+
"eval_runtime": 1.5497,
|
| 73 |
+
"eval_samples_per_second": 38.718,
|
| 74 |
+
"eval_steps_per_second": 1.291,
|
| 75 |
+
"step": 38
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 5.94,
|
| 79 |
+
"eval_accuracy": 0.3333333333333333,
|
| 80 |
+
"eval_loss": 1.3290389776229858,
|
| 81 |
+
"eval_runtime": 1.48,
|
| 82 |
+
"eval_samples_per_second": 40.54,
|
| 83 |
+
"eval_steps_per_second": 1.351,
|
| 84 |
+
"step": 46
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 6.19,
|
| 88 |
+
"grad_norm": 3.2680552005767822,
|
| 89 |
+
"learning_rate": 9.318181818181817e-07,
|
| 90 |
+
"loss": 1.4056,
|
| 91 |
+
"step": 48
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 6.97,
|
| 95 |
+
"eval_accuracy": 0.38333333333333336,
|
| 96 |
+
"eval_loss": 1.3113869428634644,
|
| 97 |
+
"eval_runtime": 1.5135,
|
| 98 |
+
"eval_samples_per_second": 39.644,
|
| 99 |
+
"eval_steps_per_second": 1.321,
|
| 100 |
+
"step": 54
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 7.74,
|
| 104 |
+
"grad_norm": 2.349689245223999,
|
| 105 |
+
"learning_rate": 8.863636363636363e-07,
|
| 106 |
+
"loss": 1.3679,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 8.0,
|
| 111 |
+
"eval_accuracy": 0.43333333333333335,
|
| 112 |
+
"eval_loss": 1.2940715551376343,
|
| 113 |
+
"eval_runtime": 1.9694,
|
| 114 |
+
"eval_samples_per_second": 30.465,
|
| 115 |
+
"eval_steps_per_second": 1.016,
|
| 116 |
+
"step": 62
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 8.9,
|
| 120 |
+
"eval_accuracy": 0.4666666666666667,
|
| 121 |
+
"eval_loss": 1.2827337980270386,
|
| 122 |
+
"eval_runtime": 1.9571,
|
| 123 |
+
"eval_samples_per_second": 30.658,
|
| 124 |
+
"eval_steps_per_second": 1.022,
|
| 125 |
+
"step": 69
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 9.29,
|
| 129 |
+
"grad_norm": 2.8183329105377197,
|
| 130 |
+
"learning_rate": 8.409090909090909e-07,
|
| 131 |
+
"loss": 1.3387,
|
| 132 |
+
"step": 72
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 9.94,
|
| 136 |
+
"eval_accuracy": 0.5,
|
| 137 |
+
"eval_loss": 1.2678014039993286,
|
| 138 |
+
"eval_runtime": 1.5047,
|
| 139 |
+
"eval_samples_per_second": 39.874,
|
| 140 |
+
"eval_steps_per_second": 1.329,
|
| 141 |
+
"step": 77
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 10.84,
|
| 145 |
+
"grad_norm": 2.4520814418792725,
|
| 146 |
+
"learning_rate": 7.954545454545454e-07,
|
| 147 |
+
"loss": 1.2992,
|
| 148 |
+
"step": 84
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 10.97,
|
| 152 |
+
"eval_accuracy": 0.4666666666666667,
|
| 153 |
+
"eval_loss": 1.2557296752929688,
|
| 154 |
+
"eval_runtime": 1.8046,
|
| 155 |
+
"eval_samples_per_second": 33.248,
|
| 156 |
+
"eval_steps_per_second": 1.108,
|
| 157 |
+
"step": 85
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 12.0,
|
| 161 |
+
"eval_accuracy": 0.4666666666666667,
|
| 162 |
+
"eval_loss": 1.2453906536102295,
|
| 163 |
+
"eval_runtime": 1.4943,
|
| 164 |
+
"eval_samples_per_second": 40.152,
|
| 165 |
+
"eval_steps_per_second": 1.338,
|
| 166 |
+
"step": 93
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 12.39,
|
| 170 |
+
"grad_norm": 2.357367992401123,
|
| 171 |
+
"learning_rate": 7.5e-07,
|
| 172 |
+
"loss": 1.2797,
|
| 173 |
+
"step": 96
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 12.9,
|
| 177 |
+
"eval_accuracy": 0.48333333333333334,
|
| 178 |
+
"eval_loss": 1.234529972076416,
|
| 179 |
+
"eval_runtime": 1.5563,
|
| 180 |
+
"eval_samples_per_second": 38.554,
|
| 181 |
+
"eval_steps_per_second": 1.285,
|
| 182 |
+
"step": 100
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 13.94,
|
| 186 |
+
"grad_norm": 2.453970432281494,
|
| 187 |
+
"learning_rate": 7.045454545454545e-07,
|
| 188 |
+
"loss": 1.2507,
|
| 189 |
+
"step": 108
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 13.94,
|
| 193 |
+
"eval_accuracy": 0.48333333333333334,
|
| 194 |
+
"eval_loss": 1.221469759941101,
|
| 195 |
+
"eval_runtime": 1.4885,
|
| 196 |
+
"eval_samples_per_second": 40.308,
|
| 197 |
+
"eval_steps_per_second": 1.344,
|
| 198 |
+
"step": 108
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 14.97,
|
| 202 |
+
"eval_accuracy": 0.5,
|
| 203 |
+
"eval_loss": 1.2108979225158691,
|
| 204 |
+
"eval_runtime": 1.4973,
|
| 205 |
+
"eval_samples_per_second": 40.072,
|
| 206 |
+
"eval_steps_per_second": 1.336,
|
| 207 |
+
"step": 116
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 15.48,
|
| 211 |
+
"grad_norm": 2.9230761528015137,
|
| 212 |
+
"learning_rate": 6.59090909090909e-07,
|
| 213 |
+
"loss": 1.2337,
|
| 214 |
+
"step": 120
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 16.0,
|
| 218 |
+
"eval_accuracy": 0.5,
|
| 219 |
+
"eval_loss": 1.2004987001419067,
|
| 220 |
+
"eval_runtime": 1.7915,
|
| 221 |
+
"eval_samples_per_second": 33.492,
|
| 222 |
+
"eval_steps_per_second": 1.116,
|
| 223 |
+
"step": 124
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 16.9,
|
| 227 |
+
"eval_accuracy": 0.5,
|
| 228 |
+
"eval_loss": 1.1903845071792603,
|
| 229 |
+
"eval_runtime": 1.4887,
|
| 230 |
+
"eval_samples_per_second": 40.304,
|
| 231 |
+
"eval_steps_per_second": 1.343,
|
| 232 |
+
"step": 131
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 17.03,
|
| 236 |
+
"grad_norm": 2.8677401542663574,
|
| 237 |
+
"learning_rate": 6.136363636363636e-07,
|
| 238 |
+
"loss": 1.2076,
|
| 239 |
+
"step": 132
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 17.94,
|
| 243 |
+
"eval_accuracy": 0.5166666666666667,
|
| 244 |
+
"eval_loss": 1.1796098947525024,
|
| 245 |
+
"eval_runtime": 1.4929,
|
| 246 |
+
"eval_samples_per_second": 40.191,
|
| 247 |
+
"eval_steps_per_second": 1.34,
|
| 248 |
+
"step": 139
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 18.58,
|
| 252 |
+
"grad_norm": 2.4694573879241943,
|
| 253 |
+
"learning_rate": 5.681818181818182e-07,
|
| 254 |
+
"loss": 1.1968,
|
| 255 |
+
"step": 144
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 18.97,
|
| 259 |
+
"eval_accuracy": 0.5333333333333333,
|
| 260 |
+
"eval_loss": 1.1699223518371582,
|
| 261 |
+
"eval_runtime": 1.4962,
|
| 262 |
+
"eval_samples_per_second": 40.102,
|
| 263 |
+
"eval_steps_per_second": 1.337,
|
| 264 |
+
"step": 147
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 20.0,
|
| 268 |
+
"eval_accuracy": 0.5333333333333333,
|
| 269 |
+
"eval_loss": 1.1609662771224976,
|
| 270 |
+
"eval_runtime": 2.0442,
|
| 271 |
+
"eval_samples_per_second": 29.351,
|
| 272 |
+
"eval_steps_per_second": 0.978,
|
| 273 |
+
"step": 155
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 20.13,
|
| 277 |
+
"grad_norm": 2.54495906829834,
|
| 278 |
+
"learning_rate": 5.227272727272727e-07,
|
| 279 |
+
"loss": 1.171,
|
| 280 |
+
"step": 156
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 20.9,
|
| 284 |
+
"eval_accuracy": 0.5333333333333333,
|
| 285 |
+
"eval_loss": 1.1543792486190796,
|
| 286 |
+
"eval_runtime": 1.4771,
|
| 287 |
+
"eval_samples_per_second": 40.621,
|
| 288 |
+
"eval_steps_per_second": 1.354,
|
| 289 |
+
"step": 162
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 21.68,
|
| 293 |
+
"grad_norm": 2.391249418258667,
|
| 294 |
+
"learning_rate": 4.772727272727273e-07,
|
| 295 |
+
"loss": 1.1572,
|
| 296 |
+
"step": 168
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 21.94,
|
| 300 |
+
"eval_accuracy": 0.5333333333333333,
|
| 301 |
+
"eval_loss": 1.1475883722305298,
|
| 302 |
+
"eval_runtime": 1.4966,
|
| 303 |
+
"eval_samples_per_second": 40.091,
|
| 304 |
+
"eval_steps_per_second": 1.336,
|
| 305 |
+
"step": 170
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 22.97,
|
| 309 |
+
"eval_accuracy": 0.5333333333333333,
|
| 310 |
+
"eval_loss": 1.1411240100860596,
|
| 311 |
+
"eval_runtime": 1.5033,
|
| 312 |
+
"eval_samples_per_second": 39.912,
|
| 313 |
+
"eval_steps_per_second": 1.33,
|
| 314 |
+
"step": 178
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 23.23,
|
| 318 |
+
"grad_norm": 2.524062395095825,
|
| 319 |
+
"learning_rate": 4.318181818181818e-07,
|
| 320 |
+
"loss": 1.1383,
|
| 321 |
+
"step": 180
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 24.0,
|
| 325 |
+
"eval_accuracy": 0.5333333333333333,
|
| 326 |
+
"eval_loss": 1.135024905204773,
|
| 327 |
+
"eval_runtime": 1.6434,
|
| 328 |
+
"eval_samples_per_second": 36.51,
|
| 329 |
+
"eval_steps_per_second": 1.217,
|
| 330 |
+
"step": 186
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 24.77,
|
| 334 |
+
"grad_norm": 2.7990708351135254,
|
| 335 |
+
"learning_rate": 3.8636363636363636e-07,
|
| 336 |
+
"loss": 1.14,
|
| 337 |
+
"step": 192
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 24.9,
|
| 341 |
+
"eval_accuracy": 0.5333333333333333,
|
| 342 |
+
"eval_loss": 1.1297781467437744,
|
| 343 |
+
"eval_runtime": 1.4951,
|
| 344 |
+
"eval_samples_per_second": 40.131,
|
| 345 |
+
"eval_steps_per_second": 1.338,
|
| 346 |
+
"step": 193
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 25.94,
|
| 350 |
+
"eval_accuracy": 0.55,
|
| 351 |
+
"eval_loss": 1.1256134510040283,
|
| 352 |
+
"eval_runtime": 1.4962,
|
| 353 |
+
"eval_samples_per_second": 40.102,
|
| 354 |
+
"eval_steps_per_second": 1.337,
|
| 355 |
+
"step": 201
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 26.32,
|
| 359 |
+
"grad_norm": 2.7583816051483154,
|
| 360 |
+
"learning_rate": 3.4090909090909085e-07,
|
| 361 |
+
"loss": 1.1114,
|
| 362 |
+
"step": 204
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 26.97,
|
| 366 |
+
"eval_accuracy": 0.55,
|
| 367 |
+
"eval_loss": 1.1212241649627686,
|
| 368 |
+
"eval_runtime": 1.6451,
|
| 369 |
+
"eval_samples_per_second": 36.472,
|
| 370 |
+
"eval_steps_per_second": 1.216,
|
| 371 |
+
"step": 209
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 27.87,
|
| 375 |
+
"grad_norm": 2.4490418434143066,
|
| 376 |
+
"learning_rate": 2.9545454545454545e-07,
|
| 377 |
+
"loss": 1.1094,
|
| 378 |
+
"step": 216
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 28.0,
|
| 382 |
+
"eval_accuracy": 0.55,
|
| 383 |
+
"eval_loss": 1.1173356771469116,
|
| 384 |
+
"eval_runtime": 1.5074,
|
| 385 |
+
"eval_samples_per_second": 39.803,
|
| 386 |
+
"eval_steps_per_second": 1.327,
|
| 387 |
+
"step": 217
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 28.9,
|
| 391 |
+
"eval_accuracy": 0.55,
|
| 392 |
+
"eval_loss": 1.1143361330032349,
|
| 393 |
+
"eval_runtime": 1.5142,
|
| 394 |
+
"eval_samples_per_second": 39.625,
|
| 395 |
+
"eval_steps_per_second": 1.321,
|
| 396 |
+
"step": 224
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 29.42,
|
| 400 |
+
"grad_norm": 2.93764066696167,
|
| 401 |
+
"learning_rate": 2.5e-07,
|
| 402 |
+
"loss": 1.0872,
|
| 403 |
+
"step": 228
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 29.94,
|
| 407 |
+
"eval_accuracy": 0.5666666666666667,
|
| 408 |
+
"eval_loss": 1.111220121383667,
|
| 409 |
+
"eval_runtime": 1.4968,
|
| 410 |
+
"eval_samples_per_second": 40.085,
|
| 411 |
+
"eval_steps_per_second": 1.336,
|
| 412 |
+
"step": 232
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 30.97,
|
| 416 |
+
"grad_norm": 2.4049625396728516,
|
| 417 |
+
"learning_rate": 2.0454545454545456e-07,
|
| 418 |
+
"loss": 1.0941,
|
| 419 |
+
"step": 240
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 30.97,
|
| 423 |
+
"eval_accuracy": 0.5666666666666667,
|
| 424 |
+
"eval_loss": 1.1078248023986816,
|
| 425 |
+
"eval_runtime": 1.9337,
|
| 426 |
+
"eval_samples_per_second": 31.028,
|
| 427 |
+
"eval_steps_per_second": 1.034,
|
| 428 |
+
"step": 240
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 32.0,
|
| 432 |
+
"eval_accuracy": 0.5666666666666667,
|
| 433 |
+
"eval_loss": 1.1053519248962402,
|
| 434 |
+
"eval_runtime": 1.483,
|
| 435 |
+
"eval_samples_per_second": 40.459,
|
| 436 |
+
"eval_steps_per_second": 1.349,
|
| 437 |
+
"step": 248
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 32.52,
|
| 441 |
+
"grad_norm": 2.444944381713867,
|
| 442 |
+
"learning_rate": 1.5909090909090907e-07,
|
| 443 |
+
"loss": 1.0882,
|
| 444 |
+
"step": 252
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 32.9,
|
| 448 |
+
"eval_accuracy": 0.5666666666666667,
|
| 449 |
+
"eval_loss": 1.1032441854476929,
|
| 450 |
+
"eval_runtime": 1.5278,
|
| 451 |
+
"eval_samples_per_second": 39.272,
|
| 452 |
+
"eval_steps_per_second": 1.309,
|
| 453 |
+
"step": 255
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 33.94,
|
| 457 |
+
"eval_accuracy": 0.5666666666666667,
|
| 458 |
+
"eval_loss": 1.101210355758667,
|
| 459 |
+
"eval_runtime": 1.4836,
|
| 460 |
+
"eval_samples_per_second": 40.443,
|
| 461 |
+
"eval_steps_per_second": 1.348,
|
| 462 |
+
"step": 263
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 34.06,
|
| 466 |
+
"grad_norm": 2.511270523071289,
|
| 467 |
+
"learning_rate": 1.1363636363636363e-07,
|
| 468 |
+
"loss": 1.0685,
|
| 469 |
+
"step": 264
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 34.97,
|
| 473 |
+
"eval_accuracy": 0.5666666666666667,
|
| 474 |
+
"eval_loss": 1.099798560142517,
|
| 475 |
+
"eval_runtime": 2.009,
|
| 476 |
+
"eval_samples_per_second": 29.866,
|
| 477 |
+
"eval_steps_per_second": 0.996,
|
| 478 |
+
"step": 271
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 35.61,
|
| 482 |
+
"grad_norm": 2.6111106872558594,
|
| 483 |
+
"learning_rate": 6.818181818181817e-08,
|
| 484 |
+
"loss": 1.0775,
|
| 485 |
+
"step": 276
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 36.0,
|
| 489 |
+
"eval_accuracy": 0.5666666666666667,
|
| 490 |
+
"eval_loss": 1.0987868309020996,
|
| 491 |
+
"eval_runtime": 1.5072,
|
| 492 |
+
"eval_samples_per_second": 39.81,
|
| 493 |
+
"eval_steps_per_second": 1.327,
|
| 494 |
+
"step": 279
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 36.9,
|
| 498 |
+
"eval_accuracy": 0.5666666666666667,
|
| 499 |
+
"eval_loss": 1.098325252532959,
|
| 500 |
+
"eval_runtime": 1.4717,
|
| 501 |
+
"eval_samples_per_second": 40.77,
|
| 502 |
+
"eval_steps_per_second": 1.359,
|
| 503 |
+
"step": 286
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 37.16,
|
| 507 |
+
"grad_norm": 2.345568895339966,
|
| 508 |
+
"learning_rate": 2.2727272727272725e-08,
|
| 509 |
+
"loss": 1.0817,
|
| 510 |
+
"step": 288
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 37.94,
|
| 514 |
+
"eval_accuracy": 0.5666666666666667,
|
| 515 |
+
"eval_loss": 1.0981043577194214,
|
| 516 |
+
"eval_runtime": 1.576,
|
| 517 |
+
"eval_samples_per_second": 38.072,
|
| 518 |
+
"eval_steps_per_second": 1.269,
|
| 519 |
+
"step": 294
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 37.94,
|
| 523 |
+
"step": 294,
|
| 524 |
+
"total_flos": 2.864620236542755e+18,
|
| 525 |
+
"train_loss": 1.2232356217442726,
|
| 526 |
+
"train_runtime": 1646.5284,
|
| 527 |
+
"train_samples_per_second": 24.845,
|
| 528 |
+
"train_steps_per_second": 0.179
|
| 529 |
+
}
|
| 530 |
+
],
|
| 531 |
+
"logging_steps": 12,
|
| 532 |
+
"max_steps": 294,
|
| 533 |
+
"num_input_tokens_seen": 0,
|
| 534 |
+
"num_train_epochs": 42,
|
| 535 |
+
"save_steps": 500,
|
| 536 |
+
"total_flos": 2.864620236542755e+18,
|
| 537 |
+
"train_batch_size": 32,
|
| 538 |
+
"trial_name": null,
|
| 539 |
+
"trial_params": null
|
| 540 |
+
}
|