End of training
Browse files- README.md +2 -2
- all_results.json +16 -0
- eval_results.json +10 -0
- train_results.json +9 -0
- trainer_state.json +817 -0
README.md
CHANGED
|
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/CodeGPT-small-java](https://huggingface.co/microsoft/CodeGPT-small-java) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 1.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [microsoft/CodeGPT-small-java](https://huggingface.co/microsoft/CodeGPT-small-java) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 1.2841
|
| 21 |
+
- Accuracy: 0.7565
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.7564910391618287,
|
| 4 |
+
"eval_loss": 1.2840659618377686,
|
| 5 |
+
"eval_runtime": 84.9049,
|
| 6 |
+
"eval_samples": 4924,
|
| 7 |
+
"eval_samples_per_second": 57.994,
|
| 8 |
+
"eval_steps_per_second": 2.426,
|
| 9 |
+
"perplexity": 3.6112932956782884,
|
| 10 |
+
"total_flos": 8.315357626368e+16,
|
| 11 |
+
"train_loss": 0.615904873986939,
|
| 12 |
+
"train_runtime": 4775.4033,
|
| 13 |
+
"train_samples": 31824,
|
| 14 |
+
"train_samples_per_second": 33.321,
|
| 15 |
+
"train_steps_per_second": 1.042
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.7564910391618287,
|
| 4 |
+
"eval_loss": 1.2840659618377686,
|
| 5 |
+
"eval_runtime": 84.9049,
|
| 6 |
+
"eval_samples": 4924,
|
| 7 |
+
"eval_samples_per_second": 57.994,
|
| 8 |
+
"eval_steps_per_second": 2.426,
|
| 9 |
+
"perplexity": 3.6112932956782884
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 8.315357626368e+16,
|
| 4 |
+
"train_loss": 0.615904873986939,
|
| 5 |
+
"train_runtime": 4775.4033,
|
| 6 |
+
"train_samples": 31824,
|
| 7 |
+
"train_samples_per_second": 33.321,
|
| 8 |
+
"train_steps_per_second": 1.042
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,817 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3500,
|
| 3 |
+
"best_metric": 1.2840659618377686,
|
| 4 |
+
"best_model_checkpoint": "/home/hyang45/se_git_repo/final_addtional_exp_mia_trans/save/new/microsoft_CodeGPT-small-java/1/checkpoint-3500",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4975,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.05027652086475616,
|
| 14 |
+
"grad_norm": 0.38967880606651306,
|
| 15 |
+
"learning_rate": 3.9200000000000004e-05,
|
| 16 |
+
"loss": 1.4062,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.10055304172951232,
|
| 21 |
+
"grad_norm": 0.25512877106666565,
|
| 22 |
+
"learning_rate": 7.92e-05,
|
| 23 |
+
"loss": 0.9927,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.15082956259426847,
|
| 28 |
+
"grad_norm": 0.2808470129966736,
|
| 29 |
+
"learning_rate": 7.998005947565195e-05,
|
| 30 |
+
"loss": 0.8156,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.20110608345902464,
|
| 35 |
+
"grad_norm": 0.2150183469057083,
|
| 36 |
+
"learning_rate": 7.991862264348792e-05,
|
| 37 |
+
"loss": 0.7848,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.2513826043237808,
|
| 42 |
+
"grad_norm": 0.2027038335800171,
|
| 43 |
+
"learning_rate": 7.981574497732115e-05,
|
| 44 |
+
"loss": 0.764,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.30165912518853694,
|
| 49 |
+
"grad_norm": 0.23373721539974213,
|
| 50 |
+
"learning_rate": 7.967153327783776e-05,
|
| 51 |
+
"loss": 0.7302,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.35193564605329314,
|
| 56 |
+
"grad_norm": 0.18324315547943115,
|
| 57 |
+
"learning_rate": 7.948613725594216e-05,
|
| 58 |
+
"loss": 0.7287,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.4022121669180493,
|
| 63 |
+
"grad_norm": 0.21186716854572296,
|
| 64 |
+
"learning_rate": 7.925974937733732e-05,
|
| 65 |
+
"loss": 0.7011,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.45248868778280543,
|
| 70 |
+
"grad_norm": 0.3749410808086395,
|
| 71 |
+
"learning_rate": 7.899260466271975e-05,
|
| 72 |
+
"loss": 0.7029,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5027652086475616,
|
| 77 |
+
"grad_norm": 0.1961645632982254,
|
| 78 |
+
"learning_rate": 7.868498044379681e-05,
|
| 79 |
+
"loss": 0.6967,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.5027652086475616,
|
| 84 |
+
"eval_accuracy": 0.7520552872875925,
|
| 85 |
+
"eval_loss": 1.3162076473236084,
|
| 86 |
+
"eval_runtime": 85.0292,
|
| 87 |
+
"eval_samples_per_second": 57.91,
|
| 88 |
+
"eval_steps_per_second": 2.423,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.5530417295123178,
|
| 93 |
+
"grad_norm": 0.17420420050621033,
|
| 94 |
+
"learning_rate": 7.833719607537954e-05,
|
| 95 |
+
"loss": 0.6851,
|
| 96 |
+
"step": 550
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.6033182503770739,
|
| 100 |
+
"grad_norm": 0.2493220567703247,
|
| 101 |
+
"learning_rate": 7.794961260384997e-05,
|
| 102 |
+
"loss": 0.671,
|
| 103 |
+
"step": 600
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.6535947712418301,
|
| 107 |
+
"grad_norm": 0.20172835886478424,
|
| 108 |
+
"learning_rate": 7.752263239234692e-05,
|
| 109 |
+
"loss": 0.662,
|
| 110 |
+
"step": 650
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.7038712921065863,
|
| 114 |
+
"grad_norm": 0.25561460852622986,
|
| 115 |
+
"learning_rate": 7.705669870305975e-05,
|
| 116 |
+
"loss": 0.6763,
|
| 117 |
+
"step": 700
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.7541478129713424,
|
| 121 |
+
"grad_norm": 0.19745612144470215,
|
| 122 |
+
"learning_rate": 7.655229523706317e-05,
|
| 123 |
+
"loss": 0.6545,
|
| 124 |
+
"step": 750
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.8044243338360986,
|
| 128 |
+
"grad_norm": 0.1649468094110489,
|
| 129 |
+
"learning_rate": 7.600994563217132e-05,
|
| 130 |
+
"loss": 0.6463,
|
| 131 |
+
"step": 800
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.8547008547008547,
|
| 135 |
+
"grad_norm": 0.20483672618865967,
|
| 136 |
+
"learning_rate": 7.543021291933214e-05,
|
| 137 |
+
"loss": 0.6625,
|
| 138 |
+
"step": 850
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.9049773755656109,
|
| 142 |
+
"grad_norm": 0.1755620539188385,
|
| 143 |
+
"learning_rate": 7.481369893812647e-05,
|
| 144 |
+
"loss": 0.6629,
|
| 145 |
+
"step": 900
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.9552538964303671,
|
| 149 |
+
"grad_norm": 0.1850636750459671,
|
| 150 |
+
"learning_rate": 7.416104371197853e-05,
|
| 151 |
+
"loss": 0.6329,
|
| 152 |
+
"step": 950
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.0050276520864756,
|
| 156 |
+
"grad_norm": 0.17517416179180145,
|
| 157 |
+
"learning_rate": 7.347292478372667e-05,
|
| 158 |
+
"loss": 0.6481,
|
| 159 |
+
"step": 1000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.0050276520864756,
|
| 163 |
+
"eval_accuracy": 0.7550630780433458,
|
| 164 |
+
"eval_loss": 1.3002119064331055,
|
| 165 |
+
"eval_runtime": 91.3368,
|
| 166 |
+
"eval_samples_per_second": 53.91,
|
| 167 |
+
"eval_steps_per_second": 2.255,
|
| 168 |
+
"step": 1000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.0553041729512318,
|
| 172 |
+
"grad_norm": 0.14569388329982758,
|
| 173 |
+
"learning_rate": 7.275005651224378e-05,
|
| 174 |
+
"loss": 0.6276,
|
| 175 |
+
"step": 1050
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.105580693815988,
|
| 179 |
+
"grad_norm": 0.1603846698999405,
|
| 180 |
+
"learning_rate": 7.199318933083797e-05,
|
| 181 |
+
"loss": 0.6346,
|
| 182 |
+
"step": 1100
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.155857214680744,
|
| 186 |
+
"grad_norm": 0.20617897808551788,
|
| 187 |
+
"learning_rate": 7.120310896820298e-05,
|
| 188 |
+
"loss": 0.6191,
|
| 189 |
+
"step": 1150
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.2061337355455002,
|
| 193 |
+
"grad_norm": 0.2762608230113983,
|
| 194 |
+
"learning_rate": 7.038063563272751e-05,
|
| 195 |
+
"loss": 0.63,
|
| 196 |
+
"step": 1200
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.2564102564102564,
|
| 200 |
+
"grad_norm": 0.13912968337535858,
|
| 201 |
+
"learning_rate": 6.952662316100984e-05,
|
| 202 |
+
"loss": 0.6265,
|
| 203 |
+
"step": 1250
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.3066867772750126,
|
| 207 |
+
"grad_norm": 0.27741900086402893,
|
| 208 |
+
"learning_rate": 6.864195813146212e-05,
|
| 209 |
+
"loss": 0.6204,
|
| 210 |
+
"step": 1300
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.3569632981397688,
|
| 214 |
+
"grad_norm": 0.15955646336078644,
|
| 215 |
+
"learning_rate": 6.772755894392405e-05,
|
| 216 |
+
"loss": 0.6077,
|
| 217 |
+
"step": 1350
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.4072398190045248,
|
| 221 |
+
"grad_norm": 0.15095743536949158,
|
| 222 |
+
"learning_rate": 6.678437486624186e-05,
|
| 223 |
+
"loss": 0.6361,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.457516339869281,
|
| 228 |
+
"grad_norm": 0.14820362627506256,
|
| 229 |
+
"learning_rate": 6.58133850488023e-05,
|
| 230 |
+
"loss": 0.6286,
|
| 231 |
+
"step": 1450
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.5077928607340372,
|
| 235 |
+
"grad_norm": 0.16185027360916138,
|
| 236 |
+
"learning_rate": 6.481559750804423e-05,
|
| 237 |
+
"loss": 0.6091,
|
| 238 |
+
"step": 1500
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.5077928607340372,
|
| 242 |
+
"eval_accuracy": 0.7550170211853606,
|
| 243 |
+
"eval_loss": 1.2964880466461182,
|
| 244 |
+
"eval_runtime": 85.294,
|
| 245 |
+
"eval_samples_per_second": 57.73,
|
| 246 |
+
"eval_steps_per_second": 2.415,
|
| 247 |
+
"step": 1500
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.5580693815987934,
|
| 251 |
+
"grad_norm": 0.16229036450386047,
|
| 252 |
+
"learning_rate": 6.379204808000379e-05,
|
| 253 |
+
"loss": 0.6024,
|
| 254 |
+
"step": 1550
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.6083459024635496,
|
| 258 |
+
"grad_norm": 0.14892268180847168,
|
| 259 |
+
"learning_rate": 6.274379934497889e-05,
|
| 260 |
+
"loss": 0.605,
|
| 261 |
+
"step": 1600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.6586224233283056,
|
| 265 |
+
"grad_norm": 0.15874284505844116,
|
| 266 |
+
"learning_rate": 6.167193952442972e-05,
|
| 267 |
+
"loss": 0.5853,
|
| 268 |
+
"step": 1650
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.708898944193062,
|
| 272 |
+
"grad_norm": 0.16238285601139069,
|
| 273 |
+
"learning_rate": 6.057758135126026e-05,
|
| 274 |
+
"loss": 0.6054,
|
| 275 |
+
"step": 1700
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.759175465057818,
|
| 279 |
+
"grad_norm": 0.16057848930358887,
|
| 280 |
+
"learning_rate": 5.946186091465375e-05,
|
| 281 |
+
"loss": 0.6252,
|
| 282 |
+
"step": 1750
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.8094519859225742,
|
| 286 |
+
"grad_norm": 0.2041744887828827,
|
| 287 |
+
"learning_rate": 5.8325936480661065e-05,
|
| 288 |
+
"loss": 0.5942,
|
| 289 |
+
"step": 1800
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.8597285067873304,
|
| 293 |
+
"grad_norm": 0.23550090193748474,
|
| 294 |
+
"learning_rate": 5.717098728976677e-05,
|
| 295 |
+
"loss": 0.6056,
|
| 296 |
+
"step": 1850
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.9100050276520864,
|
| 300 |
+
"grad_norm": 0.2757298946380615,
|
| 301 |
+
"learning_rate": 5.599821233268077e-05,
|
| 302 |
+
"loss": 0.6023,
|
| 303 |
+
"step": 1900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.9602815485168428,
|
| 307 |
+
"grad_norm": 0.1491357684135437,
|
| 308 |
+
"learning_rate": 5.4808829105626623e-05,
|
| 309 |
+
"loss": 0.6097,
|
| 310 |
+
"step": 1950
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 2.010055304172951,
|
| 314 |
+
"grad_norm": 0.21005308628082275,
|
| 315 |
+
"learning_rate": 5.360407234641879e-05,
|
| 316 |
+
"loss": 0.6091,
|
| 317 |
+
"step": 2000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 2.010055304172951,
|
| 321 |
+
"eval_accuracy": 0.7556497074198393,
|
| 322 |
+
"eval_loss": 1.289443016052246,
|
| 323 |
+
"eval_runtime": 85.2093,
|
| 324 |
+
"eval_samples_per_second": 57.787,
|
| 325 |
+
"eval_steps_per_second": 2.418,
|
| 326 |
+
"step": 2000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.060331825037707,
|
| 330 |
+
"grad_norm": 0.15782801806926727,
|
| 331 |
+
"learning_rate": 5.238519275264064e-05,
|
| 332 |
+
"loss": 0.5902,
|
| 333 |
+
"step": 2050
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.1106083459024636,
|
| 337 |
+
"grad_norm": 0.1624271422624588,
|
| 338 |
+
"learning_rate": 5.1153455683254215e-05,
|
| 339 |
+
"loss": 0.5942,
|
| 340 |
+
"step": 2100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.1608848667672196,
|
| 344 |
+
"grad_norm": 0.16978579759597778,
|
| 345 |
+
"learning_rate": 4.9910139844989574e-05,
|
| 346 |
+
"loss": 0.5845,
|
| 347 |
+
"step": 2150
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.211161387631976,
|
| 351 |
+
"grad_norm": 0.14723461866378784,
|
| 352 |
+
"learning_rate": 4.865653596487719e-05,
|
| 353 |
+
"loss": 0.5954,
|
| 354 |
+
"step": 2200
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.261437908496732,
|
| 358 |
+
"grad_norm": 0.16163045167922974,
|
| 359 |
+
"learning_rate": 4.739394545030192e-05,
|
| 360 |
+
"loss": 0.5821,
|
| 361 |
+
"step": 2250
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.311714429361488,
|
| 365 |
+
"grad_norm": 0.22584794461727142,
|
| 366 |
+
"learning_rate": 4.6123679037969156e-05,
|
| 367 |
+
"loss": 0.5811,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.3619909502262444,
|
| 372 |
+
"grad_norm": 0.16639545559883118,
|
| 373 |
+
"learning_rate": 4.484705543318603e-05,
|
| 374 |
+
"loss": 0.5923,
|
| 375 |
+
"step": 2350
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.4122674710910004,
|
| 379 |
+
"grad_norm": 0.14507165551185608,
|
| 380 |
+
"learning_rate": 4.356539994087004e-05,
|
| 381 |
+
"loss": 0.5758,
|
| 382 |
+
"step": 2400
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.462543991955757,
|
| 386 |
+
"grad_norm": 0.17524458467960358,
|
| 387 |
+
"learning_rate": 4.228004308970651e-05,
|
| 388 |
+
"loss": 0.5793,
|
| 389 |
+
"step": 2450
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.5128205128205128,
|
| 393 |
+
"grad_norm": 0.20955400168895721,
|
| 394 |
+
"learning_rate": 4.099231925088303e-05,
|
| 395 |
+
"loss": 0.5805,
|
| 396 |
+
"step": 2500
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.5128205128205128,
|
| 400 |
+
"eval_accuracy": 0.7555808206538009,
|
| 401 |
+
"eval_loss": 1.2918362617492676,
|
| 402 |
+
"eval_runtime": 84.9046,
|
| 403 |
+
"eval_samples_per_second": 57.994,
|
| 404 |
+
"eval_steps_per_second": 2.426,
|
| 405 |
+
"step": 2500
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 2.5630970336852688,
|
| 409 |
+
"grad_norm": 0.15553168952465057,
|
| 410 |
+
"learning_rate": 3.970356525283485e-05,
|
| 411 |
+
"loss": 0.5902,
|
| 412 |
+
"step": 2550
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 2.613373554550025,
|
| 416 |
+
"grad_norm": 0.24212880432605743,
|
| 417 |
+
"learning_rate": 3.841511899343937e-05,
|
| 418 |
+
"loss": 0.5896,
|
| 419 |
+
"step": 2600
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.663650075414781,
|
| 423 |
+
"grad_norm": 0.14325997233390808,
|
| 424 |
+
"learning_rate": 3.712831805110037e-05,
|
| 425 |
+
"loss": 0.5821,
|
| 426 |
+
"step": 2650
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.7139265962795376,
|
| 430 |
+
"grad_norm": 0.16522951424121857,
|
| 431 |
+
"learning_rate": 3.584449829616395e-05,
|
| 432 |
+
"loss": 0.5842,
|
| 433 |
+
"step": 2700
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.7642031171442936,
|
| 437 |
+
"grad_norm": 0.12420392781496048,
|
| 438 |
+
"learning_rate": 3.456499250410759e-05,
|
| 439 |
+
"loss": 0.5844,
|
| 440 |
+
"step": 2750
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.8144796380090495,
|
| 444 |
+
"grad_norm": 0.17248398065567017,
|
| 445 |
+
"learning_rate": 3.3291128971942105e-05,
|
| 446 |
+
"loss": 0.587,
|
| 447 |
+
"step": 2800
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.864756158873806,
|
| 451 |
+
"grad_norm": 0.14520220458507538,
|
| 452 |
+
"learning_rate": 3.2024230139262936e-05,
|
| 453 |
+
"loss": 0.5837,
|
| 454 |
+
"step": 2850
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.915032679738562,
|
| 458 |
+
"grad_norm": 0.14959931373596191,
|
| 459 |
+
"learning_rate": 3.076561121538211e-05,
|
| 460 |
+
"loss": 0.5827,
|
| 461 |
+
"step": 2900
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.9653092006033184,
|
| 465 |
+
"grad_norm": 0.2161024659872055,
|
| 466 |
+
"learning_rate": 2.9516578813966154e-05,
|
| 467 |
+
"loss": 0.5717,
|
| 468 |
+
"step": 2950
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 3.015082956259427,
|
| 472 |
+
"grad_norm": 0.14117109775543213,
|
| 473 |
+
"learning_rate": 2.827842959659771e-05,
|
| 474 |
+
"loss": 0.557,
|
| 475 |
+
"step": 3000
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 3.015082956259427,
|
| 479 |
+
"eval_accuracy": 0.7559059979528521,
|
| 480 |
+
"eval_loss": 1.2862486839294434,
|
| 481 |
+
"eval_runtime": 85.3029,
|
| 482 |
+
"eval_samples_per_second": 57.724,
|
| 483 |
+
"eval_steps_per_second": 2.415,
|
| 484 |
+
"step": 3000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 3.065359477124183,
|
| 488 |
+
"grad_norm": 0.180222749710083,
|
| 489 |
+
"learning_rate": 2.7052448926668305e-05,
|
| 490 |
+
"loss": 0.5686,
|
| 491 |
+
"step": 3050
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.115635997988939,
|
| 495 |
+
"grad_norm": 0.2005385011434555,
|
| 496 |
+
"learning_rate": 2.5839909535000498e-05,
|
| 497 |
+
"loss": 0.5682,
|
| 498 |
+
"step": 3100
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.165912518853695,
|
| 502 |
+
"grad_norm": 0.18873438239097595,
|
| 503 |
+
"learning_rate": 2.4642070198584236e-05,
|
| 504 |
+
"loss": 0.5816,
|
| 505 |
+
"step": 3150
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 3.2161890397184516,
|
| 509 |
+
"grad_norm": 0.18745166063308716,
|
| 510 |
+
"learning_rate": 2.3460174433798927e-05,
|
| 511 |
+
"loss": 0.5739,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.2664655605832076,
|
| 516 |
+
"grad_norm": 0.17067401111125946,
|
| 517 |
+
"learning_rate": 2.2295449205478362e-05,
|
| 518 |
+
"loss": 0.5864,
|
| 519 |
+
"step": 3250
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.3167420814479636,
|
| 523 |
+
"grad_norm": 0.16623741388320923,
|
| 524 |
+
"learning_rate": 2.114910365315819e-05,
|
| 525 |
+
"loss": 0.5795,
|
| 526 |
+
"step": 3300
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.36701860231272,
|
| 530 |
+
"grad_norm": 0.15764974057674408,
|
| 531 |
+
"learning_rate": 2.0022327835828362e-05,
|
| 532 |
+
"loss": 0.5704,
|
| 533 |
+
"step": 3350
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 3.417295123177476,
|
| 537 |
+
"grad_norm": 0.14753426611423492,
|
| 538 |
+
"learning_rate": 1.891629149649378e-05,
|
| 539 |
+
"loss": 0.5633,
|
| 540 |
+
"step": 3400
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 3.4675716440422324,
|
| 544 |
+
"grad_norm": 0.3436726927757263,
|
| 545 |
+
"learning_rate": 1.7832142847825763e-05,
|
| 546 |
+
"loss": 0.5753,
|
| 547 |
+
"step": 3450
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 3.5178481649069884,
|
| 551 |
+
"grad_norm": 0.1702040284872055,
|
| 552 |
+
"learning_rate": 1.677100738016461e-05,
|
| 553 |
+
"loss": 0.5513,
|
| 554 |
+
"step": 3500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.5178481649069884,
|
| 558 |
+
"eval_accuracy": 0.7564910391618287,
|
| 559 |
+
"eval_loss": 1.2840659618377686,
|
| 560 |
+
"eval_runtime": 85.0647,
|
| 561 |
+
"eval_samples_per_second": 57.885,
|
| 562 |
+
"eval_steps_per_second": 2.422,
|
| 563 |
+
"step": 3500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.5681246857717444,
|
| 567 |
+
"grad_norm": 0.18295463919639587,
|
| 568 |
+
"learning_rate": 1.573398669311116e-05,
|
| 569 |
+
"loss": 0.5823,
|
| 570 |
+
"step": 3550
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.6184012066365008,
|
| 574 |
+
"grad_norm": 0.17615950107574463,
|
| 575 |
+
"learning_rate": 1.472215735192e-05,
|
| 576 |
+
"loss": 0.5645,
|
| 577 |
+
"step": 3600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 3.6686777275012568,
|
| 581 |
+
"grad_norm": 0.2151349037885666,
|
| 582 |
+
"learning_rate": 1.373656976988175e-05,
|
| 583 |
+
"loss": 0.5725,
|
| 584 |
+
"step": 3650
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.718954248366013,
|
| 588 |
+
"grad_norm": 0.13829338550567627,
|
| 589 |
+
"learning_rate": 1.2778247117854439e-05,
|
| 590 |
+
"loss": 0.5576,
|
| 591 |
+
"step": 3700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 3.769230769230769,
|
| 595 |
+
"grad_norm": 0.16056522727012634,
|
| 596 |
+
"learning_rate": 1.1848184262076306e-05,
|
| 597 |
+
"loss": 0.5768,
|
| 598 |
+
"step": 3750
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 3.819507290095525,
|
| 602 |
+
"grad_norm": 0.18423378467559814,
|
| 603 |
+
"learning_rate": 1.0947346731362357e-05,
|
| 604 |
+
"loss": 0.5681,
|
| 605 |
+
"step": 3800
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 3.8697838109602816,
|
| 609 |
+
"grad_norm": 0.15498067438602448,
|
| 610 |
+
"learning_rate": 1.0076669714757261e-05,
|
| 611 |
+
"loss": 0.5507,
|
| 612 |
+
"step": 3850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 3.9200603318250375,
|
| 616 |
+
"grad_norm": 0.15752053260803223,
|
| 617 |
+
"learning_rate": 9.237057090684773e-06,
|
| 618 |
+
"loss": 0.5641,
|
| 619 |
+
"step": 3900
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 3.970336852689794,
|
| 623 |
+
"grad_norm": 0.14952270686626434,
|
| 624 |
+
"learning_rate": 8.42938048860198e-06,
|
| 625 |
+
"loss": 0.5704,
|
| 626 |
+
"step": 3950
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.020110608345902,
|
| 630 |
+
"grad_norm": 0.13633842766284943,
|
| 631 |
+
"learning_rate": 7.654478384131975e-06,
|
| 632 |
+
"loss": 0.5468,
|
| 633 |
+
"step": 4000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.020110608345902,
|
| 637 |
+
"eval_accuracy": 0.7562369323591513,
|
| 638 |
+
"eval_loss": 1.2848825454711914,
|
| 639 |
+
"eval_runtime": 85.0878,
|
| 640 |
+
"eval_samples_per_second": 57.87,
|
| 641 |
+
"eval_steps_per_second": 2.421,
|
| 642 |
+
"step": 4000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 4.070387129210658,
|
| 646 |
+
"grad_norm": 0.1506165862083435,
|
| 647 |
+
"learning_rate": 6.913155228614838e-06,
|
| 648 |
+
"loss": 0.5674,
|
| 649 |
+
"step": 4050
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 4.120663650075414,
|
| 653 |
+
"grad_norm": 0.15265598893165588,
|
| 654 |
+
"learning_rate": 6.206180613980266e-06,
|
| 655 |
+
"loss": 0.5625,
|
| 656 |
+
"step": 4100
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 4.170940170940171,
|
| 660 |
+
"grad_norm": 0.13488344848155975,
|
| 661 |
+
"learning_rate": 5.534288473808831e-06,
|
| 662 |
+
"loss": 0.5624,
|
| 663 |
+
"step": 4150
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 4.221216691804927,
|
| 667 |
+
"grad_norm": 0.13856124877929688,
|
| 668 |
+
"learning_rate": 4.898176321411478e-06,
|
| 669 |
+
"loss": 0.5628,
|
| 670 |
+
"step": 4200
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 4.271493212669683,
|
| 674 |
+
"grad_norm": 0.18089675903320312,
|
| 675 |
+
"learning_rate": 4.2985045257180415e-06,
|
| 676 |
+
"loss": 0.5731,
|
| 677 |
+
"step": 4250
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 4.321769733534439,
|
| 681 |
+
"grad_norm": 0.14551880955696106,
|
| 682 |
+
"learning_rate": 3.735895625726493e-06,
|
| 683 |
+
"loss": 0.5562,
|
| 684 |
+
"step": 4300
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 4.372046254399196,
|
| 688 |
+
"grad_norm": 0.13140977919101715,
|
| 689 |
+
"learning_rate": 3.210933684224804e-06,
|
| 690 |
+
"loss": 0.5729,
|
| 691 |
+
"step": 4350
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 4.422322775263952,
|
| 695 |
+
"grad_norm": 0.16188330948352814,
|
| 696 |
+
"learning_rate": 2.7241636814561514e-06,
|
| 697 |
+
"loss": 0.5633,
|
| 698 |
+
"step": 4400
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.472599296128708,
|
| 702 |
+
"grad_norm": 0.14776812493801117,
|
| 703 |
+
"learning_rate": 2.2760909493570127e-06,
|
| 704 |
+
"loss": 0.5677,
|
| 705 |
+
"step": 4450
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 4.522875816993464,
|
| 709 |
+
"grad_norm": 0.15746714174747467,
|
| 710 |
+
"learning_rate": 1.8671806469554976e-06,
|
| 711 |
+
"loss": 0.5513,
|
| 712 |
+
"step": 4500
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 4.522875816993464,
|
| 716 |
+
"eval_accuracy": 0.7559050053481541,
|
| 717 |
+
"eval_loss": 1.2894773483276367,
|
| 718 |
+
"eval_runtime": 85.2306,
|
| 719 |
+
"eval_samples_per_second": 57.773,
|
| 720 |
+
"eval_steps_per_second": 2.417,
|
| 721 |
+
"step": 4500
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 4.57315233785822,
|
| 725 |
+
"grad_norm": 0.1549079418182373,
|
| 726 |
+
"learning_rate": 1.497857277474455e-06,
|
| 727 |
+
"loss": 0.5727,
|
| 728 |
+
"step": 4550
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 4.623428858722976,
|
| 732 |
+
"grad_norm": 0.1668921858072281,
|
| 733 |
+
"learning_rate": 1.1685042476407404e-06,
|
| 734 |
+
"loss": 0.5641,
|
| 735 |
+
"step": 4600
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 4.673705379587733,
|
| 739 |
+
"grad_norm": 0.134994775056839,
|
| 740 |
+
"learning_rate": 8.794634696580551e-07,
|
| 741 |
+
"loss": 0.5597,
|
| 742 |
+
"step": 4650
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 4.723981900452489,
|
| 746 |
+
"grad_norm": 0.1448977142572403,
|
| 747 |
+
"learning_rate": 6.310350062567194e-07,
|
| 748 |
+
"loss": 0.5672,
|
| 749 |
+
"step": 4700
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 4.774258421317245,
|
| 753 |
+
"grad_norm": 0.16990165412425995,
|
| 754 |
+
"learning_rate": 4.2347675918863463e-07,
|
| 755 |
+
"loss": 0.5669,
|
| 756 |
+
"step": 4750
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 4.824534942182001,
|
| 760 |
+
"grad_norm": 0.16192808747291565,
|
| 761 |
+
"learning_rate": 2.5700420149102233e-07,
|
| 762 |
+
"loss": 0.5778,
|
| 763 |
+
"step": 4800
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 4.874811463046758,
|
| 767 |
+
"grad_norm": 0.154259592294693,
|
| 768 |
+
"learning_rate": 1.3179015379677585e-07,
|
| 769 |
+
"loss": 0.5774,
|
| 770 |
+
"step": 4850
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 4.925087983911514,
|
| 774 |
+
"grad_norm": 0.16326895356178284,
|
| 775 |
+
"learning_rate": 4.796460492368926e-08,
|
| 776 |
+
"loss": 0.5586,
|
| 777 |
+
"step": 4900
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 4.97536450477627,
|
| 781 |
+
"grad_norm": 0.18583731353282928,
|
| 782 |
+
"learning_rate": 5.614576928785198e-09,
|
| 783 |
+
"loss": 0.5682,
|
| 784 |
+
"step": 4950
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 5.0,
|
| 788 |
+
"step": 4975,
|
| 789 |
+
"total_flos": 8.315357626368e+16,
|
| 790 |
+
"train_loss": 0.615904873986939,
|
| 791 |
+
"train_runtime": 4775.4033,
|
| 792 |
+
"train_samples_per_second": 33.321,
|
| 793 |
+
"train_steps_per_second": 1.042
|
| 794 |
+
}
|
| 795 |
+
],
|
| 796 |
+
"logging_steps": 50,
|
| 797 |
+
"max_steps": 4975,
|
| 798 |
+
"num_input_tokens_seen": 0,
|
| 799 |
+
"num_train_epochs": 5,
|
| 800 |
+
"save_steps": 500,
|
| 801 |
+
"stateful_callbacks": {
|
| 802 |
+
"TrainerControl": {
|
| 803 |
+
"args": {
|
| 804 |
+
"should_epoch_stop": false,
|
| 805 |
+
"should_evaluate": false,
|
| 806 |
+
"should_log": false,
|
| 807 |
+
"should_save": true,
|
| 808 |
+
"should_training_stop": true
|
| 809 |
+
},
|
| 810 |
+
"attributes": {}
|
| 811 |
+
}
|
| 812 |
+
},
|
| 813 |
+
"total_flos": 8.315357626368e+16,
|
| 814 |
+
"train_batch_size": 8,
|
| 815 |
+
"trial_name": null,
|
| 816 |
+
"trial_params": null
|
| 817 |
+
}
|