End of training
Browse files- README.md +2 -2
- all_results.json +8 -0
- confusion_matrix_kfold_fold_4.png +0 -0
- confusion_matrix_test_fold_4.png +0 -0
- confusion_matrix_trainfold_fold_4.png +0 -0
- test_results.json +8 -0
- trainer_state.json +771 -0
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 [google/vivit-b-16x2-kinetics400](https://huggingface.co/google/vivit-b-16x2-kinetics400) 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 [google/vivit-b-16x2-kinetics400](https://huggingface.co/google/vivit-b-16x2-kinetics400) on an unknown dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.8711
|
| 22 |
+
- Accuracy: 0.8096
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.099501661129569,
|
| 3 |
+
"eval_accuracy": 0.8095611285266457,
|
| 4 |
+
"eval_loss": 0.871106743812561,
|
| 5 |
+
"eval_runtime": 578.7185,
|
| 6 |
+
"eval_samples_per_second": 2.205,
|
| 7 |
+
"eval_steps_per_second": 1.102
|
| 8 |
+
}
|
confusion_matrix_kfold_fold_4.png
ADDED
|
confusion_matrix_test_fold_4.png
ADDED
|
confusion_matrix_trainfold_fold_4.png
ADDED
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.099501661129569,
|
| 3 |
+
"eval_accuracy": 0.8095611285266457,
|
| 4 |
+
"eval_loss": 0.871106743812561,
|
| 5 |
+
"eval_runtime": 578.7185,
|
| 6 |
+
"eval_samples_per_second": 2.205,
|
| 7 |
+
"eval_steps_per_second": 1.102
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,771 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9950186799501868,
|
| 3 |
+
"best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/10 fold timesformer/ViViT_default_fold__3__10_epoch_Aug_batch_2_4_BdSLW60/checkpoint-9030",
|
| 4 |
+
"epoch": 9.099501661129569,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 9030,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01107419712070875,
|
| 13 |
+
"grad_norm": 47.61359786987305,
|
| 14 |
+
"learning_rate": 5.537098560354375e-06,
|
| 15 |
+
"loss": 16.6093,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0221483942414175,
|
| 20 |
+
"grad_norm": 50.204322814941406,
|
| 21 |
+
"learning_rate": 1.107419712070875e-05,
|
| 22 |
+
"loss": 15.708,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03322259136212625,
|
| 27 |
+
"grad_norm": 54.761314392089844,
|
| 28 |
+
"learning_rate": 1.6611295681063124e-05,
|
| 29 |
+
"loss": 13.8199,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.044296788482835,
|
| 34 |
+
"grad_norm": 50.68623733520508,
|
| 35 |
+
"learning_rate": 2.21483942414175e-05,
|
| 36 |
+
"loss": 11.3308,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05537098560354374,
|
| 41 |
+
"grad_norm": 48.19834518432617,
|
| 42 |
+
"learning_rate": 2.7685492801771873e-05,
|
| 43 |
+
"loss": 8.5116,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0664451827242525,
|
| 48 |
+
"grad_norm": 42.175010681152344,
|
| 49 |
+
"learning_rate": 3.322259136212625e-05,
|
| 50 |
+
"loss": 5.9326,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07751937984496124,
|
| 55 |
+
"grad_norm": 35.635250091552734,
|
| 56 |
+
"learning_rate": 3.875968992248062e-05,
|
| 57 |
+
"loss": 3.9588,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08859357696567,
|
| 62 |
+
"grad_norm": 18.773351669311523,
|
| 63 |
+
"learning_rate": 4.4296788482835e-05,
|
| 64 |
+
"loss": 2.6783,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09966777408637874,
|
| 69 |
+
"grad_norm": 28.99567413330078,
|
| 70 |
+
"learning_rate": 4.983388704318937e-05,
|
| 71 |
+
"loss": 1.8734,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1,
|
| 76 |
+
"eval_accuracy": 0.933997509339975,
|
| 77 |
+
"eval_loss": 0.44073957204818726,
|
| 78 |
+
"eval_runtime": 409.4841,
|
| 79 |
+
"eval_samples_per_second": 1.961,
|
| 80 |
+
"eval_steps_per_second": 0.982,
|
| 81 |
+
"step": 903
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.010686600221484,
|
| 85 |
+
"grad_norm": 16.92263412475586,
|
| 86 |
+
"learning_rate": 4.940322382182848e-05,
|
| 87 |
+
"loss": 1.1598,
|
| 88 |
+
"step": 1000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.0217607973421927,
|
| 92 |
+
"grad_norm": 6.563269138336182,
|
| 93 |
+
"learning_rate": 4.878799064845577e-05,
|
| 94 |
+
"loss": 0.8537,
|
| 95 |
+
"step": 1100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.0328349944629014,
|
| 99 |
+
"grad_norm": 5.872159957885742,
|
| 100 |
+
"learning_rate": 4.8172757475083056e-05,
|
| 101 |
+
"loss": 0.6449,
|
| 102 |
+
"step": 1200
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0439091915836103,
|
| 106 |
+
"grad_norm": 32.95087432861328,
|
| 107 |
+
"learning_rate": 4.755752430171035e-05,
|
| 108 |
+
"loss": 0.6785,
|
| 109 |
+
"step": 1300
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.054983388704319,
|
| 113 |
+
"grad_norm": 6.226277828216553,
|
| 114 |
+
"learning_rate": 4.694229112833764e-05,
|
| 115 |
+
"loss": 0.4238,
|
| 116 |
+
"step": 1400
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.0660575858250276,
|
| 120 |
+
"grad_norm": 6.418692588806152,
|
| 121 |
+
"learning_rate": 4.6327057954964936e-05,
|
| 122 |
+
"loss": 0.3129,
|
| 123 |
+
"step": 1500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.0771317829457365,
|
| 127 |
+
"grad_norm": 2.4607443809509277,
|
| 128 |
+
"learning_rate": 4.571182478159223e-05,
|
| 129 |
+
"loss": 0.2705,
|
| 130 |
+
"step": 1600
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.0882059800664452,
|
| 134 |
+
"grad_norm": 0.37526658177375793,
|
| 135 |
+
"learning_rate": 4.5096591608219516e-05,
|
| 136 |
+
"loss": 0.2863,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.0992801771871539,
|
| 141 |
+
"grad_norm": 1.0517969131469727,
|
| 142 |
+
"learning_rate": 4.448135843484681e-05,
|
| 143 |
+
"loss": 0.1847,
|
| 144 |
+
"step": 1800
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.1000553709856036,
|
| 148 |
+
"eval_accuracy": 0.9850560398505604,
|
| 149 |
+
"eval_loss": 0.049893345683813095,
|
| 150 |
+
"eval_runtime": 399.0117,
|
| 151 |
+
"eval_samples_per_second": 2.012,
|
| 152 |
+
"eval_steps_per_second": 1.007,
|
| 153 |
+
"step": 1807
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.010299003322259,
|
| 157 |
+
"grad_norm": 0.2012716829776764,
|
| 158 |
+
"learning_rate": 4.38661252614741e-05,
|
| 159 |
+
"loss": 0.1895,
|
| 160 |
+
"step": 1900
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.021373200442968,
|
| 164 |
+
"grad_norm": 0.1978013664484024,
|
| 165 |
+
"learning_rate": 4.325089208810139e-05,
|
| 166 |
+
"loss": 0.0969,
|
| 167 |
+
"step": 2000
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.0324473975636765,
|
| 171 |
+
"grad_norm": 14.746320724487305,
|
| 172 |
+
"learning_rate": 4.263565891472868e-05,
|
| 173 |
+
"loss": 0.1451,
|
| 174 |
+
"step": 2100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.0435215946843854,
|
| 178 |
+
"grad_norm": 0.06208398565649986,
|
| 179 |
+
"learning_rate": 4.2020425741355975e-05,
|
| 180 |
+
"loss": 0.1256,
|
| 181 |
+
"step": 2200
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.0545957918050943,
|
| 185 |
+
"grad_norm": 56.54954147338867,
|
| 186 |
+
"learning_rate": 4.140519256798327e-05,
|
| 187 |
+
"loss": 0.1377,
|
| 188 |
+
"step": 2300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.065669988925803,
|
| 192 |
+
"grad_norm": 0.05553647503256798,
|
| 193 |
+
"learning_rate": 4.078995939461056e-05,
|
| 194 |
+
"loss": 0.1855,
|
| 195 |
+
"step": 2400
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.0767441860465117,
|
| 199 |
+
"grad_norm": 0.08767577260732651,
|
| 200 |
+
"learning_rate": 4.0174726221237855e-05,
|
| 201 |
+
"loss": 0.0228,
|
| 202 |
+
"step": 2500
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.0878183831672206,
|
| 206 |
+
"grad_norm": 0.2166462242603302,
|
| 207 |
+
"learning_rate": 3.955949304786514e-05,
|
| 208 |
+
"loss": 0.0732,
|
| 209 |
+
"step": 2600
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.098892580287929,
|
| 213 |
+
"grad_norm": 0.03969873860478401,
|
| 214 |
+
"learning_rate": 3.8944259874492434e-05,
|
| 215 |
+
"loss": 0.0493,
|
| 216 |
+
"step": 2700
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.1,
|
| 220 |
+
"eval_accuracy": 0.9900373599003736,
|
| 221 |
+
"eval_loss": 0.03785159811377525,
|
| 222 |
+
"eval_runtime": 390.7573,
|
| 223 |
+
"eval_samples_per_second": 2.055,
|
| 224 |
+
"eval_steps_per_second": 1.029,
|
| 225 |
+
"step": 2710
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.009911406423034,
|
| 229 |
+
"grad_norm": 0.08723873645067215,
|
| 230 |
+
"learning_rate": 3.832902670111973e-05,
|
| 231 |
+
"loss": 0.1001,
|
| 232 |
+
"step": 2800
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.020985603543743,
|
| 236 |
+
"grad_norm": 0.07349704205989838,
|
| 237 |
+
"learning_rate": 3.7713793527747014e-05,
|
| 238 |
+
"loss": 0.0485,
|
| 239 |
+
"step": 2900
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.032059800664452,
|
| 243 |
+
"grad_norm": 0.06810366362333298,
|
| 244 |
+
"learning_rate": 3.7098560354374314e-05,
|
| 245 |
+
"loss": 0.0791,
|
| 246 |
+
"step": 3000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.0431339977851604,
|
| 250 |
+
"grad_norm": 0.7501605153083801,
|
| 251 |
+
"learning_rate": 3.64833271810016e-05,
|
| 252 |
+
"loss": 0.0892,
|
| 253 |
+
"step": 3100
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.0542081949058693,
|
| 257 |
+
"grad_norm": 0.08259281516075134,
|
| 258 |
+
"learning_rate": 3.5868094007628894e-05,
|
| 259 |
+
"loss": 0.0307,
|
| 260 |
+
"step": 3200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.065282392026578,
|
| 264 |
+
"grad_norm": 0.21411548554897308,
|
| 265 |
+
"learning_rate": 3.525286083425619e-05,
|
| 266 |
+
"loss": 0.0163,
|
| 267 |
+
"step": 3300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.076356589147287,
|
| 271 |
+
"grad_norm": 11.443143844604492,
|
| 272 |
+
"learning_rate": 3.463762766088347e-05,
|
| 273 |
+
"loss": 0.0561,
|
| 274 |
+
"step": 3400
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.0874307862679955,
|
| 278 |
+
"grad_norm": 0.02381887286901474,
|
| 279 |
+
"learning_rate": 3.4022394487510767e-05,
|
| 280 |
+
"loss": 0.1438,
|
| 281 |
+
"step": 3500
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.0985049833887044,
|
| 285 |
+
"grad_norm": 0.014776869677007198,
|
| 286 |
+
"learning_rate": 3.340716131413806e-05,
|
| 287 |
+
"loss": 0.0795,
|
| 288 |
+
"step": 3600
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.1000553709856034,
|
| 292 |
+
"eval_accuracy": 0.9763387297633873,
|
| 293 |
+
"eval_loss": 0.07058216631412506,
|
| 294 |
+
"eval_runtime": 382.0214,
|
| 295 |
+
"eval_samples_per_second": 2.102,
|
| 296 |
+
"eval_steps_per_second": 1.052,
|
| 297 |
+
"step": 3614
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 4.0095238095238095,
|
| 301 |
+
"grad_norm": 0.027067458257079124,
|
| 302 |
+
"learning_rate": 3.2791928140765346e-05,
|
| 303 |
+
"loss": 0.0549,
|
| 304 |
+
"step": 3700
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.020598006644518,
|
| 308 |
+
"grad_norm": 0.025334712117910385,
|
| 309 |
+
"learning_rate": 3.2176694967392646e-05,
|
| 310 |
+
"loss": 0.0408,
|
| 311 |
+
"step": 3800
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.031672203765227,
|
| 315 |
+
"grad_norm": 0.017670810222625732,
|
| 316 |
+
"learning_rate": 3.156146179401994e-05,
|
| 317 |
+
"loss": 0.0517,
|
| 318 |
+
"step": 3900
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.042746400885936,
|
| 322 |
+
"grad_norm": 0.027048593387007713,
|
| 323 |
+
"learning_rate": 3.0946228620647226e-05,
|
| 324 |
+
"loss": 0.0055,
|
| 325 |
+
"step": 4000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.053820598006644,
|
| 329 |
+
"grad_norm": 1.3971487283706665,
|
| 330 |
+
"learning_rate": 3.033099544727452e-05,
|
| 331 |
+
"loss": 0.0976,
|
| 332 |
+
"step": 4100
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.064894795127353,
|
| 336 |
+
"grad_norm": 0.027830231934785843,
|
| 337 |
+
"learning_rate": 2.971576227390181e-05,
|
| 338 |
+
"loss": 0.0105,
|
| 339 |
+
"step": 4200
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.075968992248062,
|
| 343 |
+
"grad_norm": 1.100161075592041,
|
| 344 |
+
"learning_rate": 2.91005291005291e-05,
|
| 345 |
+
"loss": 0.025,
|
| 346 |
+
"step": 4300
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 4.087043189368771,
|
| 350 |
+
"grad_norm": 0.021534092724323273,
|
| 351 |
+
"learning_rate": 2.8485295927156392e-05,
|
| 352 |
+
"loss": 0.0895,
|
| 353 |
+
"step": 4400
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.09811738648948,
|
| 357 |
+
"grad_norm": 0.00850300956517458,
|
| 358 |
+
"learning_rate": 2.787006275378369e-05,
|
| 359 |
+
"loss": 0.0397,
|
| 360 |
+
"step": 4500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.1,
|
| 364 |
+
"eval_accuracy": 0.9925280199252802,
|
| 365 |
+
"eval_loss": 0.043671417981386185,
|
| 366 |
+
"eval_runtime": 382.2405,
|
| 367 |
+
"eval_samples_per_second": 2.101,
|
| 368 |
+
"eval_steps_per_second": 1.052,
|
| 369 |
+
"step": 4517
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 5.009136212624584,
|
| 373 |
+
"grad_norm": 0.009393773972988129,
|
| 374 |
+
"learning_rate": 2.725482958041098e-05,
|
| 375 |
+
"loss": 0.0385,
|
| 376 |
+
"step": 4600
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.020210409745293,
|
| 380 |
+
"grad_norm": 0.013099305331707,
|
| 381 |
+
"learning_rate": 2.6639596407038268e-05,
|
| 382 |
+
"loss": 0.023,
|
| 383 |
+
"step": 4700
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 5.031284606866002,
|
| 387 |
+
"grad_norm": 4.182851314544678,
|
| 388 |
+
"learning_rate": 2.602436323366556e-05,
|
| 389 |
+
"loss": 0.0013,
|
| 390 |
+
"step": 4800
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 5.042358803986711,
|
| 394 |
+
"grad_norm": 0.005996107589453459,
|
| 395 |
+
"learning_rate": 2.540913006029285e-05,
|
| 396 |
+
"loss": 0.0446,
|
| 397 |
+
"step": 4900
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 5.05343300110742,
|
| 401 |
+
"grad_norm": 0.11919166892766953,
|
| 402 |
+
"learning_rate": 2.479389688692014e-05,
|
| 403 |
+
"loss": 0.0485,
|
| 404 |
+
"step": 5000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 5.064507198228129,
|
| 408 |
+
"grad_norm": 0.0478375069797039,
|
| 409 |
+
"learning_rate": 2.4178663713547438e-05,
|
| 410 |
+
"loss": 0.0326,
|
| 411 |
+
"step": 5100
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 5.075581395348837,
|
| 415 |
+
"grad_norm": 0.00391443632543087,
|
| 416 |
+
"learning_rate": 2.3563430540174727e-05,
|
| 417 |
+
"loss": 0.0013,
|
| 418 |
+
"step": 5200
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 5.086655592469546,
|
| 422 |
+
"grad_norm": 0.0053817410953342915,
|
| 423 |
+
"learning_rate": 2.2948197366802017e-05,
|
| 424 |
+
"loss": 0.0106,
|
| 425 |
+
"step": 5300
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 5.097729789590255,
|
| 429 |
+
"grad_norm": 0.0058421180583536625,
|
| 430 |
+
"learning_rate": 2.233296419342931e-05,
|
| 431 |
+
"loss": 0.0008,
|
| 432 |
+
"step": 5400
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 5.100055370985603,
|
| 436 |
+
"eval_accuracy": 0.9912826899128269,
|
| 437 |
+
"eval_loss": 0.04847064986824989,
|
| 438 |
+
"eval_runtime": 379.9323,
|
| 439 |
+
"eval_samples_per_second": 2.114,
|
| 440 |
+
"eval_steps_per_second": 1.058,
|
| 441 |
+
"step": 5421
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 6.00874861572536,
|
| 445 |
+
"grad_norm": 0.007814995013177395,
|
| 446 |
+
"learning_rate": 2.1717731020056604e-05,
|
| 447 |
+
"loss": 0.0025,
|
| 448 |
+
"step": 5500
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.019822812846068,
|
| 452 |
+
"grad_norm": 0.00458895368501544,
|
| 453 |
+
"learning_rate": 2.1102497846683894e-05,
|
| 454 |
+
"loss": 0.1088,
|
| 455 |
+
"step": 5600
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 6.030897009966777,
|
| 459 |
+
"grad_norm": 0.0877087414264679,
|
| 460 |
+
"learning_rate": 2.0487264673311183e-05,
|
| 461 |
+
"loss": 0.0786,
|
| 462 |
+
"step": 5700
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.041971207087486,
|
| 466 |
+
"grad_norm": 0.006522389128804207,
|
| 467 |
+
"learning_rate": 1.987203149993848e-05,
|
| 468 |
+
"loss": 0.0321,
|
| 469 |
+
"step": 5800
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 6.053045404208195,
|
| 473 |
+
"grad_norm": 0.007930456660687923,
|
| 474 |
+
"learning_rate": 1.925679832656577e-05,
|
| 475 |
+
"loss": 0.0413,
|
| 476 |
+
"step": 5900
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 6.064119601328904,
|
| 480 |
+
"grad_norm": 0.003752330085262656,
|
| 481 |
+
"learning_rate": 1.864156515319306e-05,
|
| 482 |
+
"loss": 0.0005,
|
| 483 |
+
"step": 6000
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 6.075193798449613,
|
| 487 |
+
"grad_norm": 0.0031280298717319965,
|
| 488 |
+
"learning_rate": 1.8026331979820353e-05,
|
| 489 |
+
"loss": 0.0036,
|
| 490 |
+
"step": 6100
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 6.086267995570321,
|
| 494 |
+
"grad_norm": 0.003108856501057744,
|
| 495 |
+
"learning_rate": 1.7411098806447646e-05,
|
| 496 |
+
"loss": 0.0666,
|
| 497 |
+
"step": 6200
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 6.09734219269103,
|
| 501 |
+
"grad_norm": 0.005239582154899836,
|
| 502 |
+
"learning_rate": 1.6795865633074936e-05,
|
| 503 |
+
"loss": 0.0191,
|
| 504 |
+
"step": 6300
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 6.1,
|
| 508 |
+
"eval_accuracy": 0.9912826899128269,
|
| 509 |
+
"eval_loss": 0.03256941959261894,
|
| 510 |
+
"eval_runtime": 377.5854,
|
| 511 |
+
"eval_samples_per_second": 2.127,
|
| 512 |
+
"eval_steps_per_second": 1.065,
|
| 513 |
+
"step": 6324
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.008361018826135,
|
| 517 |
+
"grad_norm": 0.008326631039381027,
|
| 518 |
+
"learning_rate": 1.6180632459702226e-05,
|
| 519 |
+
"loss": 0.0006,
|
| 520 |
+
"step": 6400
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.019435215946844,
|
| 524 |
+
"grad_norm": 0.015157405287027359,
|
| 525 |
+
"learning_rate": 1.556539928632952e-05,
|
| 526 |
+
"loss": 0.0005,
|
| 527 |
+
"step": 6500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.030509413067553,
|
| 531 |
+
"grad_norm": 0.002715812996029854,
|
| 532 |
+
"learning_rate": 1.4950166112956812e-05,
|
| 533 |
+
"loss": 0.0003,
|
| 534 |
+
"step": 6600
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.041583610188261,
|
| 538 |
+
"grad_norm": 0.0072884103283286095,
|
| 539 |
+
"learning_rate": 1.4334932939584104e-05,
|
| 540 |
+
"loss": 0.0372,
|
| 541 |
+
"step": 6700
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.05265780730897,
|
| 545 |
+
"grad_norm": 0.0038811948616057634,
|
| 546 |
+
"learning_rate": 1.3719699766211393e-05,
|
| 547 |
+
"loss": 0.0259,
|
| 548 |
+
"step": 6800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.063732004429679,
|
| 552 |
+
"grad_norm": 0.008665332570672035,
|
| 553 |
+
"learning_rate": 1.3104466592838688e-05,
|
| 554 |
+
"loss": 0.0024,
|
| 555 |
+
"step": 6900
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.074806201550388,
|
| 559 |
+
"grad_norm": 0.002044640714302659,
|
| 560 |
+
"learning_rate": 1.2489233419465978e-05,
|
| 561 |
+
"loss": 0.0003,
|
| 562 |
+
"step": 7000
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.0858803986710965,
|
| 566 |
+
"grad_norm": 0.010357621125876904,
|
| 567 |
+
"learning_rate": 1.187400024609327e-05,
|
| 568 |
+
"loss": 0.0163,
|
| 569 |
+
"step": 7100
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.096954595791805,
|
| 573 |
+
"grad_norm": 0.0026492071337997913,
|
| 574 |
+
"learning_rate": 1.1258767072720563e-05,
|
| 575 |
+
"loss": 0.0513,
|
| 576 |
+
"step": 7200
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 7.100055370985603,
|
| 580 |
+
"eval_accuracy": 0.9925280199252802,
|
| 581 |
+
"eval_loss": 0.036151375621557236,
|
| 582 |
+
"eval_runtime": 383.0674,
|
| 583 |
+
"eval_samples_per_second": 2.096,
|
| 584 |
+
"eval_steps_per_second": 1.049,
|
| 585 |
+
"step": 7228
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 8.007973421926911,
|
| 589 |
+
"grad_norm": 0.001502937520854175,
|
| 590 |
+
"learning_rate": 1.0643533899347853e-05,
|
| 591 |
+
"loss": 0.0007,
|
| 592 |
+
"step": 7300
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 8.019047619047619,
|
| 596 |
+
"grad_norm": 0.001738361781463027,
|
| 597 |
+
"learning_rate": 1.0028300725975146e-05,
|
| 598 |
+
"loss": 0.0002,
|
| 599 |
+
"step": 7400
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 8.030121816168327,
|
| 603 |
+
"grad_norm": 0.002816119696944952,
|
| 604 |
+
"learning_rate": 9.413067552602436e-06,
|
| 605 |
+
"loss": 0.0035,
|
| 606 |
+
"step": 7500
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 8.041196013289037,
|
| 610 |
+
"grad_norm": 0.014468424953520298,
|
| 611 |
+
"learning_rate": 8.797834379229729e-06,
|
| 612 |
+
"loss": 0.0005,
|
| 613 |
+
"step": 7600
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 8.052270210409745,
|
| 617 |
+
"grad_norm": 0.004878249950706959,
|
| 618 |
+
"learning_rate": 8.18260120585702e-06,
|
| 619 |
+
"loss": 0.0002,
|
| 620 |
+
"step": 7700
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 8.063344407530455,
|
| 624 |
+
"grad_norm": 0.005755834747105837,
|
| 625 |
+
"learning_rate": 7.567368032484312e-06,
|
| 626 |
+
"loss": 0.0267,
|
| 627 |
+
"step": 7800
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 8.074418604651163,
|
| 631 |
+
"grad_norm": 0.0021511847153306007,
|
| 632 |
+
"learning_rate": 6.952134859111603e-06,
|
| 633 |
+
"loss": 0.0229,
|
| 634 |
+
"step": 7900
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 8.085492801771872,
|
| 638 |
+
"grad_norm": 0.00909422431141138,
|
| 639 |
+
"learning_rate": 6.336901685738895e-06,
|
| 640 |
+
"loss": 0.0002,
|
| 641 |
+
"step": 8000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 8.09656699889258,
|
| 645 |
+
"grad_norm": 0.0021005752496421337,
|
| 646 |
+
"learning_rate": 5.7216685123661875e-06,
|
| 647 |
+
"loss": 0.0009,
|
| 648 |
+
"step": 8100
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 8.1,
|
| 652 |
+
"eval_accuracy": 0.9937733499377335,
|
| 653 |
+
"eval_loss": 0.04350635036826134,
|
| 654 |
+
"eval_runtime": 392.0186,
|
| 655 |
+
"eval_samples_per_second": 2.048,
|
| 656 |
+
"eval_steps_per_second": 1.025,
|
| 657 |
+
"step": 8131
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 9.007585825027686,
|
| 661 |
+
"grad_norm": 0.0032245826441794634,
|
| 662 |
+
"learning_rate": 5.106435338993479e-06,
|
| 663 |
+
"loss": 0.002,
|
| 664 |
+
"step": 8200
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 9.018660022148394,
|
| 668 |
+
"grad_norm": 0.003902060678228736,
|
| 669 |
+
"learning_rate": 4.4912021656207705e-06,
|
| 670 |
+
"loss": 0.0006,
|
| 671 |
+
"step": 8300
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 9.029734219269104,
|
| 675 |
+
"grad_norm": 0.010045597329735756,
|
| 676 |
+
"learning_rate": 3.875968992248062e-06,
|
| 677 |
+
"loss": 0.0264,
|
| 678 |
+
"step": 8400
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 9.040808416389812,
|
| 682 |
+
"grad_norm": 0.0016225556610152125,
|
| 683 |
+
"learning_rate": 3.260735818875354e-06,
|
| 684 |
+
"loss": 0.0297,
|
| 685 |
+
"step": 8500
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 9.05188261351052,
|
| 689 |
+
"grad_norm": 0.00418157409876585,
|
| 690 |
+
"learning_rate": 2.6455026455026455e-06,
|
| 691 |
+
"loss": 0.0003,
|
| 692 |
+
"step": 8600
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 9.06295681063123,
|
| 696 |
+
"grad_norm": 0.001903184805996716,
|
| 697 |
+
"learning_rate": 2.0302694721299375e-06,
|
| 698 |
+
"loss": 0.02,
|
| 699 |
+
"step": 8700
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 9.074031007751937,
|
| 703 |
+
"grad_norm": 7.255136013031006,
|
| 704 |
+
"learning_rate": 1.415036298757229e-06,
|
| 705 |
+
"loss": 0.0016,
|
| 706 |
+
"step": 8800
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 9.085105204872647,
|
| 710 |
+
"grad_norm": 0.002232867991551757,
|
| 711 |
+
"learning_rate": 7.998031253845208e-07,
|
| 712 |
+
"loss": 0.0179,
|
| 713 |
+
"step": 8900
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 9.096179401993355,
|
| 717 |
+
"grad_norm": 0.0014163912273943424,
|
| 718 |
+
"learning_rate": 1.8456995201181249e-07,
|
| 719 |
+
"loss": 0.0022,
|
| 720 |
+
"step": 9000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 9.099501661129569,
|
| 724 |
+
"eval_accuracy": 0.9950186799501868,
|
| 725 |
+
"eval_loss": 0.02610412798821926,
|
| 726 |
+
"eval_runtime": 379.405,
|
| 727 |
+
"eval_samples_per_second": 2.116,
|
| 728 |
+
"eval_steps_per_second": 1.06,
|
| 729 |
+
"step": 9030
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 9.099501661129569,
|
| 733 |
+
"step": 9030,
|
| 734 |
+
"total_flos": 1.851754249265873e+20,
|
| 735 |
+
"train_loss": 0.9763831017361511,
|
| 736 |
+
"train_runtime": 54804.8494,
|
| 737 |
+
"train_samples_per_second": 1.318,
|
| 738 |
+
"train_steps_per_second": 0.165
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 9.099501661129569,
|
| 742 |
+
"eval_accuracy": 0.8095611285266457,
|
| 743 |
+
"eval_loss": 0.871106743812561,
|
| 744 |
+
"eval_runtime": 578.7185,
|
| 745 |
+
"eval_samples_per_second": 2.205,
|
| 746 |
+
"eval_steps_per_second": 1.102,
|
| 747 |
+
"step": 9030
|
| 748 |
+
}
|
| 749 |
+
],
|
| 750 |
+
"logging_steps": 100,
|
| 751 |
+
"max_steps": 9030,
|
| 752 |
+
"num_input_tokens_seen": 0,
|
| 753 |
+
"num_train_epochs": 9223372036854775807,
|
| 754 |
+
"save_steps": 500,
|
| 755 |
+
"stateful_callbacks": {
|
| 756 |
+
"TrainerControl": {
|
| 757 |
+
"args": {
|
| 758 |
+
"should_epoch_stop": false,
|
| 759 |
+
"should_evaluate": false,
|
| 760 |
+
"should_log": false,
|
| 761 |
+
"should_save": true,
|
| 762 |
+
"should_training_stop": true
|
| 763 |
+
},
|
| 764 |
+
"attributes": {}
|
| 765 |
+
}
|
| 766 |
+
},
|
| 767 |
+
"total_flos": 1.851754249265873e+20,
|
| 768 |
+
"train_batch_size": 2,
|
| 769 |
+
"trial_name": null,
|
| 770 |
+
"trial_params": null
|
| 771 |
+
}
|