Hugging Face's logo

DHEIVER
/
Modelo-Avancado-de-Ultrassom-de-Mama

Image Classification
Transformers
PyTorch
convnextv2
Inference Endpoints
Model card Files Community
1
DHEIVER commited on
Commit
714962b
•
1 Parent(s): b22d4fd

Upload 10 files

Browse files
Files changed (9) hide show
  1. .gitignore +1 -0
  2. README.md +76 -0
  3. all_results.json +13 -0
  4. config.json +51 -0
  5. eval_results.json +8 -0
  6. preprocessor_config.json +22 -0
  7. train_results.json +8 -0
  8. trainer_state.json +3493 -0
  9. training_args.bin +3 -0
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
README.md ADDED
@@ -0,0 +1,76 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: facebook/convnextv2-large-1k-224
4
+ tags:
5
+ - generated_from_trainer
6
+ metrics:
7
+ - accuracy
8
+ model-index:
9
+ - name: convnextv2-large-1k-224-finetuned-BreastCancer-Classification-BreakHis-AH-60-20-20-Shuffled
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # convnextv2-large-1k-224-finetuned-BreastCancer-Classification-BreakHis-AH-60-20-20-Shuffled
17
+
18
+ This model is a fine-tuned version of [facebook/convnextv2-large-1k-224](https://huggingface.co/facebook/convnextv2-large-1k-224) on the None dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 0.0398
21
+ - Accuracy: 0.9882
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 5e-05
41
+ - train_batch_size: 16
42
+ - eval_batch_size: 16
43
+ - seed: 42
44
+ - gradient_accumulation_steps: 2
45
+ - total_train_batch_size: 32
46
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
+ - lr_scheduler_type: linear
48
+ - lr_scheduler_warmup_ratio: 0.9
49
+ - num_epochs: 14
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Accuracy | Validation Loss |
54
+ |:-------------:|:-----:|:----:|:--------:|:---------------:|
55
+ | 0.5059 | 1.0 | 199 | 0.9001 | 0.4826 |
56
+ | 0.2533 | 2.0 | 398 | 0.9515 | 0.2124 |
57
+ | 0.2358 | 3.0 | 597 | 0.9538 | 0.1543 |
58
+ | 0.2584 | 4.0 | 796 | 0.9642 | 0.1136 |
59
+ | 0.1085 | 5.0 | 995 | 0.9746 | 0.0891 |
60
+ | 0.1007 | 6.0 | 1194 | 0.9769 | 0.0725 |
61
+ | 0.1463 | 7.0 | 1393 | 0.9840 | 0.0541 |
62
+ | 0.3564 | 8.0 | 1592 | 0.9802 | 0.0880 |
63
+ | 0.0957 | 9.0 | 1791 | 0.9656 | 0.1375 |
64
+ | 0.1481 | 10.0 | 1990 | 0.0511 | 0.9873 |
65
+ | 0.1536 | 11.0 | 2189 | 0.0827 | 0.9713 |
66
+ | 0.0458 | 12.0 | 2388 | 0.0398 | 0.9882 |
67
+ | 0.4956 | 13.0 | 2587 | 0.3474 | 0.8643 |
68
+ | 0.0801 | 14.0 | 2786 | 0.0850 | 0.9797 |
69
+
70
+
71
+ ### Framework versions
72
+
73
+ - Transformers 4.31.0
74
+ - Pytorch 2.0.1+cu118
75
+ - Datasets 2.13.1
76
+ - Tokenizers 0.13.3
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 14.0,
3
+ "eval_accuracy": 0.9882186616399623,
4
+ "eval_loss": 0.03981461003422737,
5
+ "eval_runtime": 71.5873,
6
+ "eval_samples_per_second": 29.642,
7
+ "eval_steps_per_second": 1.858,
8
+ "total_flos": 1.58314339385421e+19,
9
+ "train_loss": 0.05687390304887928,
10
+ "train_runtime": 3560.0954,
11
+ "train_samples_per_second": 25.026,
12
+ "train_steps_per_second": 0.783
13
+ }
config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/convnextv2-large-1k-224",
3
+ "architectures": [
4
+ "ConvNextV2ForImageClassification"
5
+ ],
6
+ "depths": [
7
+ 3,
8
+ 3,
9
+ 27,
10
+ 3
11
+ ],
12
+ "drop_path_rate": 0.0,
13
+ "hidden_act": "gelu",
14
+ "hidden_sizes": [
15
+ 192,
16
+ 384,
17
+ 768,
18
+ 1536
19
+ ],
20
+ "id2label": {
21
+ "0": "benign",
22
+ "1": "malignant"
23
+ },
24
+ "image_size": 224,
25
+ "initializer_range": 0.02,
26
+ "label2id": {
27
+ "benign": 0,
28
+ "malignant": 1
29
+ },
30
+ "layer_norm_eps": 1e-12,
31
+ "model_type": "convnextv2",
32
+ "num_channels": 3,
33
+ "num_stages": 4,
34
+ "out_features": [
35
+ "stage4"
36
+ ],
37
+ "out_indices": [
38
+ 4
39
+ ],
40
+ "patch_size": 4,
41
+ "problem_type": "single_label_classification",
42
+ "stage_names": [
43
+ "stem",
44
+ "stage1",
45
+ "stage2",
46
+ "stage3",
47
+ "stage4"
48
+ ],
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.31.0"
51
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 14.0,
3
+ "eval_accuracy": 0.9882186616399623,
4
+ "eval_loss": 0.03981461003422737,
5
+ "eval_runtime": 71.5873,
6
+ "eval_samples_per_second": 29.642,
7
+ "eval_steps_per_second": 1.858
8
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_pct": 0.875,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "image_mean": [
7
+ 0.485,
8
+ 0.456,
9
+ 0.406
10
+ ],
11
+ "image_processor_type": "ConvNextImageProcessor",
12
+ "image_std": [
13
+ 0.229,
14
+ 0.224,
15
+ 0.225
16
+ ],
17
+ "resample": 3,
18
+ "rescale_factor": 0.00392156862745098,
19
+ "size": {
20
+ "shortest_edge": 224
21
+ }
22
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 14.0,
3
+ "total_flos": 1.58314339385421e+19,
4
+ "train_loss": 0.05687390304887928,
5
+ "train_runtime": 3560.0954,
6
+ "train_samples_per_second": 25.026,
7
+ "train_steps_per_second": 0.783
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3493 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9882186616399623,
3
+ "best_model_checkpoint": "convnextv2-large-1k-224-finetuned-BreastCancer-Classification-BreakHis-AH-60-20-20-Shuffled/checkpoint-2388",
4
+ "epoch": 14.0,
5
+ "global_step": 2786,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 9.96810207336523e-08,
13
+ "loss": 0.7048,
14
+ "step": 5
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 1.993620414673046e-07,
19
+ "loss": 0.6895,
20
+ "step": 10
21
+ },
22
+ {
23
+ "epoch": 0.08,
24
+ "learning_rate": 2.990430622009569e-07,
25
+ "loss": 0.6896,
26
+ "step": 15
27
+ },
28
+ {
29
+ "epoch": 0.1,
30
+ "learning_rate": 3.987240829346092e-07,
31
+ "loss": 0.6916,
32
+ "step": 20
33
+ },
34
+ {
35
+ "epoch": 0.13,
36
+ "learning_rate": 4.984051036682616e-07,
37
+ "loss": 0.6916,
38
+ "step": 25
39
+ },
40
+ {
41
+ "epoch": 0.15,
42
+ "learning_rate": 5.980861244019139e-07,
43
+ "loss": 0.6938,
44
+ "step": 30
45
+ },
46
+ {
47
+ "epoch": 0.18,
48
+ "learning_rate": 6.977671451355662e-07,
49
+ "loss": 0.6968,
50
+ "step": 35
51
+ },
52
+ {
53
+ "epoch": 0.2,
54
+ "learning_rate": 7.974481658692184e-07,
55
+ "loss": 0.6813,
56
+ "step": 40
57
+ },
58
+ {
59
+ "epoch": 0.23,
60
+ "learning_rate": 8.971291866028709e-07,
61
+ "loss": 0.6911,
62
+ "step": 45
63
+ },
64
+ {
65
+ "epoch": 0.25,
66
+ "learning_rate": 9.968102073365232e-07,
67
+ "loss": 0.6874,
68
+ "step": 50
69
+ },
70
+ {
71
+ "epoch": 0.28,
72
+ "learning_rate": 1.0964912280701754e-06,
73
+ "loss": 0.6877,
74
+ "step": 55
75
+ },
76
+ {
77
+ "epoch": 0.3,
78
+ "learning_rate": 1.1961722488038277e-06,
79
+ "loss": 0.6815,
80
+ "step": 60
81
+ },
82
+ {
83
+ "epoch": 0.33,
84
+ "learning_rate": 1.2958532695374803e-06,
85
+ "loss": 0.6833,
86
+ "step": 65
87
+ },
88
+ {
89
+ "epoch": 0.35,
90
+ "learning_rate": 1.3955342902711324e-06,
91
+ "loss": 0.6782,
92
+ "step": 70
93
+ },
94
+ {
95
+ "epoch": 0.38,
96
+ "learning_rate": 1.4952153110047847e-06,
97
+ "loss": 0.6742,
98
+ "step": 75
99
+ },
100
+ {
101
+ "epoch": 0.4,
102
+ "learning_rate": 1.5948963317384369e-06,
103
+ "loss": 0.6763,
104
+ "step": 80
105
+ },
106
+ {
107
+ "epoch": 0.43,
108
+ "learning_rate": 1.6945773524720894e-06,
109
+ "loss": 0.6771,
110
+ "step": 85
111
+ },
112
+ {
113
+ "epoch": 0.45,
114
+ "learning_rate": 1.7942583732057418e-06,
115
+ "loss": 0.6638,
116
+ "step": 90
117
+ },
118
+ {
119
+ "epoch": 0.48,
120
+ "learning_rate": 1.8939393939393941e-06,
121
+ "loss": 0.665,
122
+ "step": 95
123
+ },
124
+ {
125
+ "epoch": 0.5,
126
+ "learning_rate": 1.9936204146730465e-06,
127
+ "loss": 0.6631,
128
+ "step": 100
129
+ },
130
+ {
131
+ "epoch": 0.53,
132
+ "learning_rate": 2.093301435406699e-06,
133
+ "loss": 0.6522,
134
+ "step": 105
135
+ },
136
+ {
137
+ "epoch": 0.55,
138
+ "learning_rate": 2.1929824561403507e-06,
139
+ "loss": 0.6581,
140
+ "step": 110
141
+ },
142
+ {
143
+ "epoch": 0.58,
144
+ "learning_rate": 2.292663476874003e-06,
145
+ "loss": 0.6432,
146
+ "step": 115
147
+ },
148
+ {
149
+ "epoch": 0.6,
150
+ "learning_rate": 2.3923444976076554e-06,
151
+ "loss": 0.6576,
152
+ "step": 120
153
+ },
154
+ {
155
+ "epoch": 0.63,
156
+ "learning_rate": 2.492025518341308e-06,
157
+ "loss": 0.6368,
158
+ "step": 125
159
+ },
160
+ {
161
+ "epoch": 0.65,
162
+ "learning_rate": 2.5917065390749605e-06,
163
+ "loss": 0.6257,
164
+ "step": 130
165
+ },
166
+ {
167
+ "epoch": 0.68,
168
+ "learning_rate": 2.691387559808613e-06,
169
+ "loss": 0.624,
170
+ "step": 135
171
+ },
172
+ {
173
+ "epoch": 0.7,
174
+ "learning_rate": 2.7910685805422648e-06,
175
+ "loss": 0.6115,
176
+ "step": 140
177
+ },
178
+ {
179
+ "epoch": 0.73,
180
+ "learning_rate": 2.890749601275917e-06,
181
+ "loss": 0.6151,
182
+ "step": 145
183
+ },
184
+ {
185
+ "epoch": 0.75,
186
+ "learning_rate": 2.9904306220095695e-06,
187
+ "loss": 0.6045,
188
+ "step": 150
189
+ },
190
+ {
191
+ "epoch": 0.78,
192
+ "learning_rate": 3.090111642743222e-06,
193
+ "loss": 0.5888,
194
+ "step": 155
195
+ },
196
+ {
197
+ "epoch": 0.8,
198
+ "learning_rate": 3.1897926634768737e-06,
199
+ "loss": 0.5753,
200
+ "step": 160
201
+ },
202
+ {
203
+ "epoch": 0.83,
204
+ "learning_rate": 3.2894736842105265e-06,
205
+ "loss": 0.5607,
206
+ "step": 165
207
+ },
208
+ {
209
+ "epoch": 0.85,
210
+ "learning_rate": 3.389154704944179e-06,
211
+ "loss": 0.5578,
212
+ "step": 170
213
+ },
214
+ {
215
+ "epoch": 0.88,
216
+ "learning_rate": 3.488835725677831e-06,
217
+ "loss": 0.5461,
218
+ "step": 175
219
+ },
220
+ {
221
+ "epoch": 0.9,
222
+ "learning_rate": 3.5885167464114835e-06,
223
+ "loss": 0.5403,
224
+ "step": 180
225
+ },
226
+ {
227
+ "epoch": 0.93,
228
+ "learning_rate": 3.688197767145136e-06,
229
+ "loss": 0.5343,
230
+ "step": 185
231
+ },
232
+ {
233
+ "epoch": 0.95,
234
+ "learning_rate": 3.7878787878787882e-06,
235
+ "loss": 0.5206,
236
+ "step": 190
237
+ },
238
+ {
239
+ "epoch": 0.98,
240
+ "learning_rate": 3.88755980861244e-06,
241
+ "loss": 0.5059,
242
+ "step": 195
243
+ },
244
+ {
245
+ "epoch": 1.0,
246
+ "eval_accuracy": 0.9000942507068803,
247
+ "eval_loss": 0.4826333522796631,
248
+ "eval_runtime": 1914.453,
249
+ "eval_samples_per_second": 1.108,
250
+ "eval_steps_per_second": 0.069,
251
+ "step": 199
252
+ },
253
+ {
254
+ "epoch": 1.01,
255
+ "learning_rate": 3.987240829346093e-06,
256
+ "loss": 0.4989,
257
+ "step": 200
258
+ },
259
+ {
260
+ "epoch": 1.03,
261
+ "learning_rate": 4.086921850079746e-06,
262
+ "loss": 0.4963,
263
+ "step": 205
264
+ },
265
+ {
266
+ "epoch": 1.06,
267
+ "learning_rate": 4.186602870813398e-06,
268
+ "loss": 0.4905,
269
+ "step": 210
270
+ },
271
+ {
272
+ "epoch": 1.08,
273
+ "learning_rate": 4.2862838915470495e-06,
274
+ "loss": 0.4311,
275
+ "step": 215
276
+ },
277
+ {
278
+ "epoch": 1.11,
279
+ "learning_rate": 4.3859649122807014e-06,
280
+ "loss": 0.4353,
281
+ "step": 220
282
+ },
283
+ {
284
+ "epoch": 1.13,
285
+ "learning_rate": 4.485645933014354e-06,
286
+ "loss": 0.4574,
287
+ "step": 225
288
+ },
289
+ {
290
+ "epoch": 1.16,
291
+ "learning_rate": 4.585326953748006e-06,
292
+ "loss": 0.4508,
293
+ "step": 230
294
+ },
295
+ {
296
+ "epoch": 1.18,
297
+ "learning_rate": 4.685007974481659e-06,
298
+ "loss": 0.4541,
299
+ "step": 235
300
+ },
301
+ {
302
+ "epoch": 1.21,
303
+ "learning_rate": 4.784688995215311e-06,
304
+ "loss": 0.4089,
305
+ "step": 240
306
+ },
307
+ {
308
+ "epoch": 1.23,
309
+ "learning_rate": 4.8843700159489636e-06,
310
+ "loss": 0.4251,
311
+ "step": 245
312
+ },
313
+ {
314
+ "epoch": 1.26,
315
+ "learning_rate": 4.984051036682616e-06,
316
+ "loss": 0.4276,
317
+ "step": 250
318
+ },
319
+ {
320
+ "epoch": 1.28,
321
+ "learning_rate": 5.083732057416268e-06,
322
+ "loss": 0.4303,
323
+ "step": 255
324
+ },
325
+ {
326
+ "epoch": 1.31,
327
+ "learning_rate": 5.183413078149921e-06,
328
+ "loss": 0.4144,
329
+ "step": 260
330
+ },
331
+ {
332
+ "epoch": 1.33,
333
+ "learning_rate": 5.283094098883573e-06,
334
+ "loss": 0.3988,
335
+ "step": 265
336
+ },
337
+ {
338
+ "epoch": 1.36,
339
+ "learning_rate": 5.382775119617226e-06,
340
+ "loss": 0.3718,
341
+ "step": 270
342
+ },
343
+ {
344
+ "epoch": 1.38,
345
+ "learning_rate": 5.482456140350877e-06,
346
+ "loss": 0.3631,
347
+ "step": 275
348
+ },
349
+ {
350
+ "epoch": 1.41,
351
+ "learning_rate": 5.5821371610845296e-06,
352
+ "loss": 0.3524,
353
+ "step": 280
354
+ },
355
+ {
356
+ "epoch": 1.43,
357
+ "learning_rate": 5.681818181818182e-06,
358
+ "loss": 0.3498,
359
+ "step": 285
360
+ },
361
+ {
362
+ "epoch": 1.46,
363
+ "learning_rate": 5.781499202551834e-06,
364
+ "loss": 0.3501,
365
+ "step": 290
366
+ },
367
+ {
368
+ "epoch": 1.48,
369
+ "learning_rate": 5.881180223285487e-06,
370
+ "loss": 0.3949,
371
+ "step": 295
372
+ },
373
+ {
374
+ "epoch": 1.51,
375
+ "learning_rate": 5.980861244019139e-06,
376
+ "loss": 0.3294,
377
+ "step": 300
378
+ },
379
+ {
380
+ "epoch": 1.53,
381
+ "learning_rate": 6.080542264752792e-06,
382
+ "loss": 0.3177,
383
+ "step": 305
384
+ },
385
+ {
386
+ "epoch": 1.56,
387
+ "learning_rate": 6.180223285486444e-06,
388
+ "loss": 0.3063,
389
+ "step": 310
390
+ },
391
+ {
392
+ "epoch": 1.58,
393
+ "learning_rate": 6.2799043062200955e-06,
394
+ "loss": 0.3212,
395
+ "step": 315
396
+ },
397
+ {
398
+ "epoch": 1.61,
399
+ "learning_rate": 6.3795853269537475e-06,
400
+ "loss": 0.2946,
401
+ "step": 320
402
+ },
403
+ {
404
+ "epoch": 1.63,
405
+ "learning_rate": 6.4792663476874e-06,
406
+ "loss": 0.2695,
407
+ "step": 325
408
+ },
409
+ {
410
+ "epoch": 1.66,
411
+ "learning_rate": 6.578947368421053e-06,
412
+ "loss": 0.2869,
413
+ "step": 330
414
+ },
415
+ {
416
+ "epoch": 1.68,
417
+ "learning_rate": 6.678628389154705e-06,
418
+ "loss": 0.3035,
419
+ "step": 335
420
+ },
421
+ {
422
+ "epoch": 1.71,
423
+ "learning_rate": 6.778309409888358e-06,
424
+ "loss": 0.2946,
425
+ "step": 340
426
+ },
427
+ {
428
+ "epoch": 1.73,
429
+ "learning_rate": 6.87799043062201e-06,
430
+ "loss": 0.3297,
431
+ "step": 345
432
+ },
433
+ {
434
+ "epoch": 1.76,
435
+ "learning_rate": 6.977671451355662e-06,
436
+ "loss": 0.2265,
437
+ "step": 350
438
+ },
439
+ {
440
+ "epoch": 1.78,
441
+ "learning_rate": 7.077352472089314e-06,
442
+ "loss": 0.292,
443
+ "step": 355
444
+ },
445
+ {
446
+ "epoch": 1.81,
447
+ "learning_rate": 7.177033492822967e-06,
448
+ "loss": 0.3683,
449
+ "step": 360
450
+ },
451
+ {
452
+ "epoch": 1.83,
453
+ "learning_rate": 7.276714513556619e-06,
454
+ "loss": 0.286,
455
+ "step": 365
456
+ },
457
+ {
458
+ "epoch": 1.86,
459
+ "learning_rate": 7.376395534290272e-06,
460
+ "loss": 0.2633,
461
+ "step": 370
462
+ },
463
+ {
464
+ "epoch": 1.88,
465
+ "learning_rate": 7.476076555023924e-06,
466
+ "loss": 0.2371,
467
+ "step": 375
468
+ },
469
+ {
470
+ "epoch": 1.91,
471
+ "learning_rate": 7.5757575757575764e-06,
472
+ "loss": 0.2734,
473
+ "step": 380
474
+ },
475
+ {
476
+ "epoch": 1.93,
477
+ "learning_rate": 7.675438596491228e-06,
478
+ "loss": 0.2586,
479
+ "step": 385
480
+ },
481
+ {
482
+ "epoch": 1.96,
483
+ "learning_rate": 7.77511961722488e-06,
484
+ "loss": 0.245,
485
+ "step": 390
486
+ },
487
+ {
488
+ "epoch": 1.98,
489
+ "learning_rate": 7.874800637958534e-06,
490
+ "loss": 0.2533,
491
+ "step": 395
492
+ },
493
+ {
494
+ "epoch": 2.0,
495
+ "eval_accuracy": 0.9514608859566447,
496
+ "eval_loss": 0.2124292403459549,
497
+ "eval_runtime": 70.7108,
498
+ "eval_samples_per_second": 30.01,
499
+ "eval_steps_per_second": 1.881,
500
+ "step": 398
501
+ },
502
+ {
503
+ "epoch": 2.01,
504
+ "learning_rate": 7.974481658692186e-06,
505
+ "loss": 0.2251,
506
+ "step": 400
507
+ },
508
+ {
509
+ "epoch": 2.04,
510
+ "learning_rate": 8.074162679425838e-06,
511
+ "loss": 0.2319,
512
+ "step": 405
513
+ },
514
+ {
515
+ "epoch": 2.06,
516
+ "learning_rate": 8.173843700159491e-06,
517
+ "loss": 0.2642,
518
+ "step": 410
519
+ },
520
+ {
521
+ "epoch": 2.09,
522
+ "learning_rate": 8.273524720893143e-06,
523
+ "loss": 0.2336,
524
+ "step": 415
525
+ },
526
+ {
527
+ "epoch": 2.11,
528
+ "learning_rate": 8.373205741626795e-06,
529
+ "loss": 0.2232,
530
+ "step": 420
531
+ },
532
+ {
533
+ "epoch": 2.14,
534
+ "learning_rate": 8.472886762360447e-06,
535
+ "loss": 0.2451,
536
+ "step": 425
537
+ },
538
+ {
539
+ "epoch": 2.16,
540
+ "learning_rate": 8.572567783094099e-06,
541
+ "loss": 0.2953,
542
+ "step": 430
543
+ },
544
+ {
545
+ "epoch": 2.19,
546
+ "learning_rate": 8.672248803827751e-06,
547
+ "loss": 0.2671,
548
+ "step": 435
549
+ },
550
+ {
551
+ "epoch": 2.21,
552
+ "learning_rate": 8.771929824561403e-06,
553
+ "loss": 0.245,
554
+ "step": 440
555
+ },
556
+ {
557
+ "epoch": 2.24,
558
+ "learning_rate": 8.871610845295056e-06,
559
+ "loss": 0.2079,
560
+ "step": 445
561
+ },
562
+ {
563
+ "epoch": 2.26,
564
+ "learning_rate": 8.971291866028708e-06,
565
+ "loss": 0.1964,
566
+ "step": 450
567
+ },
568
+ {
569
+ "epoch": 2.29,
570
+ "learning_rate": 9.07097288676236e-06,
571
+ "loss": 0.1823,
572
+ "step": 455
573
+ },
574
+ {
575
+ "epoch": 2.31,
576
+ "learning_rate": 9.170653907496012e-06,
577
+ "loss": 0.2291,
578
+ "step": 460
579
+ },
580
+ {
581
+ "epoch": 2.34,
582
+ "learning_rate": 9.270334928229666e-06,
583
+ "loss": 0.2359,
584
+ "step": 465
585
+ },
586
+ {
587
+ "epoch": 2.36,
588
+ "learning_rate": 9.370015948963318e-06,
589
+ "loss": 0.2404,
590
+ "step": 470
591
+ },
592
+ {
593
+ "epoch": 2.39,
594
+ "learning_rate": 9.46969696969697e-06,
595
+ "loss": 0.2503,
596
+ "step": 475
597
+ },
598
+ {
599
+ "epoch": 2.41,
600
+ "learning_rate": 9.569377990430622e-06,
601
+ "loss": 0.2626,
602
+ "step": 480
603
+ },
604
+ {
605
+ "epoch": 2.44,
606
+ "learning_rate": 9.669059011164275e-06,
607
+ "loss": 0.2264,
608
+ "step": 485
609
+ },
610
+ {
611
+ "epoch": 2.46,
612
+ "learning_rate": 9.768740031897927e-06,
613
+ "loss": 0.1944,
614
+ "step": 490
615
+ },
616
+ {
617
+ "epoch": 2.49,
618
+ "learning_rate": 9.868421052631579e-06,
619
+ "loss": 0.2319,
620
+ "step": 495
621
+ },
622
+ {
623
+ "epoch": 2.51,
624
+ "learning_rate": 9.968102073365233e-06,
625
+ "loss": 0.273,
626
+ "step": 500
627
+ },
628
+ {
629
+ "epoch": 2.54,
630
+ "learning_rate": 1.0067783094098885e-05,
631
+ "loss": 0.1638,
632
+ "step": 505
633
+ },
634
+ {
635
+ "epoch": 2.56,
636
+ "learning_rate": 1.0167464114832537e-05,
637
+ "loss": 0.2288,
638
+ "step": 510
639
+ },
640
+ {
641
+ "epoch": 2.59,
642
+ "learning_rate": 1.0267145135566188e-05,
643
+ "loss": 0.1798,
644
+ "step": 515
645
+ },
646
+ {
647
+ "epoch": 2.61,
648
+ "learning_rate": 1.0366826156299842e-05,
649
+ "loss": 0.1836,
650
+ "step": 520
651
+ },
652
+ {
653
+ "epoch": 2.64,
654
+ "learning_rate": 1.0466507177033494e-05,
655
+ "loss": 0.3112,
656
+ "step": 525
657
+ },
658
+ {
659
+ "epoch": 2.66,
660
+ "learning_rate": 1.0566188197767146e-05,
661
+ "loss": 0.1693,
662
+ "step": 530
663
+ },
664
+ {
665
+ "epoch": 2.69,
666
+ "learning_rate": 1.0665869218500798e-05,
667
+ "loss": 0.2139,
668
+ "step": 535
669
+ },
670
+ {
671
+ "epoch": 2.71,
672
+ "learning_rate": 1.0765550239234451e-05,
673
+ "loss": 0.2522,
674
+ "step": 540
675
+ },
676
+ {
677
+ "epoch": 2.74,
678
+ "learning_rate": 1.0865231259968102e-05,
679
+ "loss": 0.2047,
680
+ "step": 545
681
+ },
682
+ {
683
+ "epoch": 2.76,
684
+ "learning_rate": 1.0964912280701754e-05,
685
+ "loss": 0.2797,
686
+ "step": 550
687
+ },
688
+ {
689
+ "epoch": 2.79,
690
+ "learning_rate": 1.1064593301435407e-05,
691
+ "loss": 0.1701,
692
+ "step": 555
693
+ },
694
+ {
695
+ "epoch": 2.81,
696
+ "learning_rate": 1.1164274322169059e-05,
697
+ "loss": 0.2006,
698
+ "step": 560
699
+ },
700
+ {
701
+ "epoch": 2.84,
702
+ "learning_rate": 1.1263955342902711e-05,
703
+ "loss": 0.2385,
704
+ "step": 565
705
+ },
706
+ {
707
+ "epoch": 2.86,
708
+ "learning_rate": 1.1363636363636365e-05,
709
+ "loss": 0.1963,
710
+ "step": 570
711
+ },
712
+ {
713
+ "epoch": 2.89,
714
+ "learning_rate": 1.1463317384370017e-05,
715
+ "loss": 0.1463,
716
+ "step": 575
717
+ },
718
+ {
719
+ "epoch": 2.91,
720
+ "learning_rate": 1.1562998405103668e-05,
721
+ "loss": 0.2504,
722
+ "step": 580
723
+ },
724
+ {
725
+ "epoch": 2.94,
726
+ "learning_rate": 1.166267942583732e-05,
727
+ "loss": 0.1663,
728
+ "step": 585
729
+ },
730
+ {
731
+ "epoch": 2.96,
732
+ "learning_rate": 1.1762360446570974e-05,
733
+ "loss": 0.1742,
734
+ "step": 590
735
+ },
736
+ {
737
+ "epoch": 2.99,
738
+ "learning_rate": 1.1862041467304626e-05,
739
+ "loss": 0.2358,
740
+ "step": 595
741
+ },
742
+ {
743
+ "epoch": 3.0,
744
+ "eval_accuracy": 0.9538171536286523,
745
+ "eval_loss": 0.15427447855472565,
746
+ "eval_runtime": 70.7075,
747
+ "eval_samples_per_second": 30.011,
748
+ "eval_steps_per_second": 1.881,
749
+ "step": 597
750
+ },
751
+ {
752
+ "epoch": 3.02,
753
+ "learning_rate": 1.1961722488038278e-05,
754
+ "loss": 0.173,
755
+ "step": 600
756
+ },
757
+ {
758
+ "epoch": 3.04,
759
+ "learning_rate": 1.206140350877193e-05,
760
+ "loss": 0.1766,
761
+ "step": 605
762
+ },
763
+ {
764
+ "epoch": 3.07,
765
+ "learning_rate": 1.2161084529505583e-05,
766
+ "loss": 0.182,
767
+ "step": 610
768
+ },
769
+ {
770
+ "epoch": 3.09,
771
+ "learning_rate": 1.2260765550239235e-05,
772
+ "loss": 0.2666,
773
+ "step": 615
774
+ },
775
+ {
776
+ "epoch": 3.12,
777
+ "learning_rate": 1.2360446570972887e-05,
778
+ "loss": 0.126,
779
+ "step": 620
780
+ },
781
+ {
782
+ "epoch": 3.14,
783
+ "learning_rate": 1.246012759170654e-05,
784
+ "loss": 0.1112,
785
+ "step": 625
786
+ },
787
+ {
788
+ "epoch": 3.17,
789
+ "learning_rate": 1.2559808612440191e-05,
790
+ "loss": 0.1267,
791
+ "step": 630
792
+ },
793
+ {
794
+ "epoch": 3.19,
795
+ "learning_rate": 1.2659489633173843e-05,
796
+ "loss": 0.2112,
797
+ "step": 635
798
+ },
799
+ {
800
+ "epoch": 3.22,
801
+ "learning_rate": 1.2759170653907495e-05,
802
+ "loss": 0.1375,
803
+ "step": 640
804
+ },
805
+ {
806
+ "epoch": 3.24,
807
+ "learning_rate": 1.2858851674641149e-05,
808
+ "loss": 0.2749,
809
+ "step": 645
810
+ },
811
+ {
812
+ "epoch": 3.27,
813
+ "learning_rate": 1.29585326953748e-05,
814
+ "loss": 0.2136,
815
+ "step": 650
816
+ },
817
+ {
818
+ "epoch": 3.29,
819
+ "learning_rate": 1.3058213716108452e-05,
820
+ "loss": 0.1766,
821
+ "step": 655
822
+ },
823
+ {
824
+ "epoch": 3.32,
825
+ "learning_rate": 1.3157894736842106e-05,
826
+ "loss": 0.1892,
827
+ "step": 660
828
+ },
829
+ {
830
+ "epoch": 3.34,
831
+ "learning_rate": 1.3257575757575758e-05,
832
+ "loss": 0.1417,
833
+ "step": 665
834
+ },
835
+ {
836
+ "epoch": 3.37,
837
+ "learning_rate": 1.335725677830941e-05,
838
+ "loss": 0.2264,
839
+ "step": 670
840
+ },
841
+ {
842
+ "epoch": 3.39,
843
+ "learning_rate": 1.3456937799043062e-05,
844
+ "loss": 0.1516,
845
+ "step": 675
846
+ },
847
+ {
848
+ "epoch": 3.42,
849
+ "learning_rate": 1.3556618819776715e-05,
850
+ "loss": 0.206,
851
+ "step": 680
852
+ },
853
+ {
854
+ "epoch": 3.44,
855
+ "learning_rate": 1.3656299840510367e-05,
856
+ "loss": 0.1868,
857
+ "step": 685
858
+ },
859
+ {
860
+ "epoch": 3.47,
861
+ "learning_rate": 1.375598086124402e-05,
862
+ "loss": 0.2041,
863
+ "step": 690
864
+ },
865
+ {
866
+ "epoch": 3.49,
867
+ "learning_rate": 1.3855661881977671e-05,
868
+ "loss": 0.1884,
869
+ "step": 695
870
+ },
871
+ {
872
+ "epoch": 3.52,
873
+ "learning_rate": 1.3955342902711325e-05,
874
+ "loss": 0.3096,
875
+ "step": 700
876
+ },
877
+ {
878
+ "epoch": 3.54,
879
+ "learning_rate": 1.4055023923444977e-05,
880
+ "loss": 0.1395,
881
+ "step": 705
882
+ },
883
+ {
884
+ "epoch": 3.57,
885
+ "learning_rate": 1.4154704944178629e-05,
886
+ "loss": 0.167,
887
+ "step": 710
888
+ },
889
+ {
890
+ "epoch": 3.59,
891
+ "learning_rate": 1.425438596491228e-05,
892
+ "loss": 0.1371,
893
+ "step": 715
894
+ },
895
+ {
896
+ "epoch": 3.62,
897
+ "learning_rate": 1.4354066985645934e-05,
898
+ "loss": 0.1149,
899
+ "step": 720
900
+ },
901
+ {
902
+ "epoch": 3.64,
903
+ "learning_rate": 1.4453748006379586e-05,
904
+ "loss": 0.1051,
905
+ "step": 725
906
+ },
907
+ {
908
+ "epoch": 3.67,
909
+ "learning_rate": 1.4553429027113238e-05,
910
+ "loss": 0.1705,
911
+ "step": 730
912
+ },
913
+ {
914
+ "epoch": 3.69,
915
+ "learning_rate": 1.4653110047846892e-05,
916
+ "loss": 0.1174,
917
+ "step": 735
918
+ },
919
+ {
920
+ "epoch": 3.72,
921
+ "learning_rate": 1.4752791068580543e-05,
922
+ "loss": 0.1356,
923
+ "step": 740
924
+ },
925
+ {
926
+ "epoch": 3.74,
927
+ "learning_rate": 1.4852472089314195e-05,
928
+ "loss": 0.3099,
929
+ "step": 745
930
+ },
931
+ {
932
+ "epoch": 3.77,
933
+ "learning_rate": 1.4952153110047847e-05,
934
+ "loss": 0.2023,
935
+ "step": 750
936
+ },
937
+ {
938
+ "epoch": 3.79,
939
+ "learning_rate": 1.5051834130781501e-05,
940
+ "loss": 0.1293,
941
+ "step": 755
942
+ },
943
+ {
944
+ "epoch": 3.82,
945
+ "learning_rate": 1.5151515151515153e-05,
946
+ "loss": 0.2193,
947
+ "step": 760
948
+ },
949
+ {
950
+ "epoch": 3.84,
951
+ "learning_rate": 1.5251196172248805e-05,
952
+ "loss": 0.1946,
953
+ "step": 765
954
+ },
955
+ {
956
+ "epoch": 3.87,
957
+ "learning_rate": 1.5350877192982457e-05,
958
+ "loss": 0.3177,
959
+ "step": 770
960
+ },
961
+ {
962
+ "epoch": 3.89,
963
+ "learning_rate": 1.545055821371611e-05,
964
+ "loss": 0.3677,
965
+ "step": 775
966
+ },
967
+ {
968
+ "epoch": 3.92,
969
+ "learning_rate": 1.555023923444976e-05,
970
+ "loss": 0.1895,
971
+ "step": 780
972
+ },
973
+ {
974
+ "epoch": 3.94,
975
+ "learning_rate": 1.5649920255183416e-05,
976
+ "loss": 0.2612,
977
+ "step": 785
978
+ },
979
+ {
980
+ "epoch": 3.97,
981
+ "learning_rate": 1.5749601275917068e-05,
982
+ "loss": 0.1511,
983
+ "step": 790
984
+ },
985
+ {
986
+ "epoch": 3.99,
987
+ "learning_rate": 1.584928229665072e-05,
988
+ "loss": 0.2584,
989
+ "step": 795
990
+ },
991
+ {
992
+ "epoch": 4.0,
993
+ "eval_accuracy": 0.9641847313854854,
994
+ "eval_loss": 0.11363683640956879,
995
+ "eval_runtime": 70.6659,
996
+ "eval_samples_per_second": 30.029,
997
+ "eval_steps_per_second": 1.882,
998
+ "step": 796
999
+ },
1000
+ {
1001
+ "epoch": 4.02,
1002
+ "learning_rate": 1.594896331738437e-05,
1003
+ "loss": 0.204,
1004
+ "step": 800
1005
+ },
1006
+ {
1007
+ "epoch": 4.05,
1008
+ "learning_rate": 1.6048644338118024e-05,
1009
+ "loss": 0.3212,
1010
+ "step": 805
1011
+ },
1012
+ {
1013
+ "epoch": 4.07,
1014
+ "learning_rate": 1.6148325358851675e-05,
1015
+ "loss": 0.2214,
1016
+ "step": 810
1017
+ },
1018
+ {
1019
+ "epoch": 4.1,
1020
+ "learning_rate": 1.6248006379585327e-05,
1021
+ "loss": 0.1679,
1022
+ "step": 815
1023
+ },
1024
+ {
1025
+ "epoch": 4.12,
1026
+ "learning_rate": 1.6347687400318983e-05,
1027
+ "loss": 0.1288,
1028
+ "step": 820
1029
+ },
1030
+ {
1031
+ "epoch": 4.15,
1032
+ "learning_rate": 1.6447368421052635e-05,
1033
+ "loss": 0.1769,
1034
+ "step": 825
1035
+ },
1036
+ {
1037
+ "epoch": 4.17,
1038
+ "learning_rate": 1.6547049441786287e-05,
1039
+ "loss": 0.1685,
1040
+ "step": 830
1041
+ },
1042
+ {
1043
+ "epoch": 4.2,
1044
+ "learning_rate": 1.664673046251994e-05,
1045
+ "loss": 0.147,
1046
+ "step": 835
1047
+ },
1048
+ {
1049
+ "epoch": 4.22,
1050
+ "learning_rate": 1.674641148325359e-05,
1051
+ "loss": 0.279,
1052
+ "step": 840
1053
+ },
1054
+ {
1055
+ "epoch": 4.25,
1056
+ "learning_rate": 1.684609250398724e-05,
1057
+ "loss": 0.3511,
1058
+ "step": 845
1059
+ },
1060
+ {
1061
+ "epoch": 4.27,
1062
+ "learning_rate": 1.6945773524720894e-05,
1063
+ "loss": 0.3512,
1064
+ "step": 850
1065
+ },
1066
+ {
1067
+ "epoch": 4.3,
1068
+ "learning_rate": 1.7045454545454546e-05,
1069
+ "loss": 0.1552,
1070
+ "step": 855
1071
+ },
1072
+ {
1073
+ "epoch": 4.32,
1074
+ "learning_rate": 1.7145135566188198e-05,
1075
+ "loss": 0.1517,
1076
+ "step": 860
1077
+ },
1078
+ {
1079
+ "epoch": 4.35,
1080
+ "learning_rate": 1.724481658692185e-05,
1081
+ "loss": 0.0997,
1082
+ "step": 865
1083
+ },
1084
+ {
1085
+ "epoch": 4.37,
1086
+ "learning_rate": 1.7344497607655502e-05,
1087
+ "loss": 0.1326,
1088
+ "step": 870
1089
+ },
1090
+ {
1091
+ "epoch": 4.4,
1092
+ "learning_rate": 1.7444178628389154e-05,
1093
+ "loss": 0.1793,
1094
+ "step": 875
1095
+ },
1096
+ {
1097
+ "epoch": 4.42,
1098
+ "learning_rate": 1.7543859649122806e-05,
1099
+ "loss": 0.1459,
1100
+ "step": 880
1101
+ },
1102
+ {
1103
+ "epoch": 4.45,
1104
+ "learning_rate": 1.7643540669856458e-05,
1105
+ "loss": 0.1259,
1106
+ "step": 885
1107
+ },
1108
+ {
1109
+ "epoch": 4.47,
1110
+ "learning_rate": 1.7743221690590113e-05,
1111
+ "loss": 0.1866,
1112
+ "step": 890
1113
+ },
1114
+ {
1115
+ "epoch": 4.5,
1116
+ "learning_rate": 1.7842902711323765e-05,
1117
+ "loss": 0.2281,
1118
+ "step": 895
1119
+ },
1120
+ {
1121
+ "epoch": 4.52,
1122
+ "learning_rate": 1.7942583732057417e-05,
1123
+ "loss": 0.1234,
1124
+ "step": 900
1125
+ },
1126
+ {
1127
+ "epoch": 4.55,
1128
+ "learning_rate": 1.804226475279107e-05,
1129
+ "loss": 0.1683,
1130
+ "step": 905
1131
+ },
1132
+ {
1133
+ "epoch": 4.57,
1134
+ "learning_rate": 1.814194577352472e-05,
1135
+ "loss": 0.1124,
1136
+ "step": 910
1137
+ },
1138
+ {
1139
+ "epoch": 4.6,
1140
+ "learning_rate": 1.8241626794258373e-05,
1141
+ "loss": 0.1016,
1142
+ "step": 915
1143
+ },
1144
+ {
1145
+ "epoch": 4.62,
1146
+ "learning_rate": 1.8341307814992024e-05,
1147
+ "loss": 0.1383,
1148
+ "step": 920
1149
+ },
1150
+ {
1151
+ "epoch": 4.65,
1152
+ "learning_rate": 1.844098883572568e-05,
1153
+ "loss": 0.2442,
1154
+ "step": 925
1155
+ },
1156
+ {
1157
+ "epoch": 4.67,
1158
+ "learning_rate": 1.8540669856459332e-05,
1159
+ "loss": 0.1057,
1160
+ "step": 930
1161
+ },
1162
+ {
1163
+ "epoch": 4.7,
1164
+ "learning_rate": 1.8640350877192984e-05,
1165
+ "loss": 0.1856,
1166
+ "step": 935
1167
+ },
1168
+ {
1169
+ "epoch": 4.72,
1170
+ "learning_rate": 1.8740031897926636e-05,
1171
+ "loss": 0.1272,
1172
+ "step": 940
1173
+ },
1174
+ {
1175
+ "epoch": 4.75,
1176
+ "learning_rate": 1.8839712918660287e-05,
1177
+ "loss": 0.1384,
1178
+ "step": 945
1179
+ },
1180
+ {
1181
+ "epoch": 4.77,
1182
+ "learning_rate": 1.893939393939394e-05,
1183
+ "loss": 0.2314,
1184
+ "step": 950
1185
+ },
1186
+ {
1187
+ "epoch": 4.8,
1188
+ "learning_rate": 1.903907496012759e-05,
1189
+ "loss": 0.1623,
1190
+ "step": 955
1191
+ },
1192
+ {
1193
+ "epoch": 4.82,
1194
+ "learning_rate": 1.9138755980861243e-05,
1195
+ "loss": 0.2701,
1196
+ "step": 960
1197
+ },
1198
+ {
1199
+ "epoch": 4.85,
1200
+ "learning_rate": 1.92384370015949e-05,
1201
+ "loss": 0.1866,
1202
+ "step": 965
1203
+ },
1204
+ {
1205
+ "epoch": 4.87,
1206
+ "learning_rate": 1.933811802232855e-05,
1207
+ "loss": 0.2126,
1208
+ "step": 970
1209
+ },
1210
+ {
1211
+ "epoch": 4.9,
1212
+ "learning_rate": 1.9437799043062202e-05,
1213
+ "loss": 0.1549,
1214
+ "step": 975
1215
+ },
1216
+ {
1217
+ "epoch": 4.92,
1218
+ "learning_rate": 1.9537480063795854e-05,
1219
+ "loss": 0.133,
1220
+ "step": 980
1221
+ },
1222
+ {
1223
+ "epoch": 4.95,
1224
+ "learning_rate": 1.9637161084529506e-05,
1225
+ "loss": 0.0753,
1226
+ "step": 985
1227
+ },
1228
+ {
1229
+ "epoch": 4.97,
1230
+ "learning_rate": 1.9736842105263158e-05,
1231
+ "loss": 0.0942,
1232
+ "step": 990
1233
+ },
1234
+ {
1235
+ "epoch": 5.0,
1236
+ "learning_rate": 1.983652312599681e-05,
1237
+ "loss": 0.1085,
1238
+ "step": 995
1239
+ },
1240
+ {
1241
+ "epoch": 5.0,
1242
+ "eval_accuracy": 0.9745523091423186,
1243
+ "eval_loss": 0.08914583176374435,
1244
+ "eval_runtime": 70.8532,
1245
+ "eval_samples_per_second": 29.949,
1246
+ "eval_steps_per_second": 1.877,
1247
+ "step": 995
1248
+ },
1249
+ {
1250
+ "epoch": 5.03,
1251
+ "learning_rate": 1.9936204146730465e-05,
1252
+ "loss": 0.1358,
1253
+ "step": 1000
1254
+ },
1255
+ {
1256
+ "epoch": 5.05,
1257
+ "learning_rate": 2.0035885167464117e-05,
1258
+ "loss": 0.1491,
1259
+ "step": 1005
1260
+ },
1261
+ {
1262
+ "epoch": 5.08,
1263
+ "learning_rate": 2.013556618819777e-05,
1264
+ "loss": 0.148,
1265
+ "step": 1010
1266
+ },
1267
+ {
1268
+ "epoch": 5.1,
1269
+ "learning_rate": 2.023524720893142e-05,
1270
+ "loss": 0.0951,
1271
+ "step": 1015
1272
+ },
1273
+ {
1274
+ "epoch": 5.13,
1275
+ "learning_rate": 2.0334928229665073e-05,
1276
+ "loss": 0.0913,
1277
+ "step": 1020
1278
+ },
1279
+ {
1280
+ "epoch": 5.15,
1281
+ "learning_rate": 2.0434609250398725e-05,
1282
+ "loss": 0.1532,
1283
+ "step": 1025
1284
+ },
1285
+ {
1286
+ "epoch": 5.18,
1287
+ "learning_rate": 2.0534290271132377e-05,
1288
+ "loss": 0.1276,
1289
+ "step": 1030
1290
+ },
1291
+ {
1292
+ "epoch": 5.2,
1293
+ "learning_rate": 2.0633971291866032e-05,
1294
+ "loss": 0.169,
1295
+ "step": 1035
1296
+ },
1297
+ {
1298
+ "epoch": 5.23,
1299
+ "learning_rate": 2.0733652312599684e-05,
1300
+ "loss": 0.1091,
1301
+ "step": 1040
1302
+ },
1303
+ {
1304
+ "epoch": 5.25,
1305
+ "learning_rate": 2.0833333333333336e-05,
1306
+ "loss": 0.0604,
1307
+ "step": 1045
1308
+ },
1309
+ {
1310
+ "epoch": 5.28,
1311
+ "learning_rate": 2.0933014354066988e-05,
1312
+ "loss": 0.2305,
1313
+ "step": 1050
1314
+ },
1315
+ {
1316
+ "epoch": 5.3,
1317
+ "learning_rate": 2.103269537480064e-05,
1318
+ "loss": 0.1552,
1319
+ "step": 1055
1320
+ },
1321
+ {
1322
+ "epoch": 5.33,
1323
+ "learning_rate": 2.1132376395534292e-05,
1324
+ "loss": 0.0952,
1325
+ "step": 1060
1326
+ },
1327
+ {
1328
+ "epoch": 5.35,
1329
+ "learning_rate": 2.1232057416267944e-05,
1330
+ "loss": 0.1021,
1331
+ "step": 1065
1332
+ },
1333
+ {
1334
+ "epoch": 5.38,
1335
+ "learning_rate": 2.1331738437001596e-05,
1336
+ "loss": 0.0702,
1337
+ "step": 1070
1338
+ },
1339
+ {
1340
+ "epoch": 5.4,
1341
+ "learning_rate": 2.143141945773525e-05,
1342
+ "loss": 0.1879,
1343
+ "step": 1075
1344
+ },
1345
+ {
1346
+ "epoch": 5.43,
1347
+ "learning_rate": 2.1531100478468903e-05,
1348
+ "loss": 0.1929,
1349
+ "step": 1080
1350
+ },
1351
+ {
1352
+ "epoch": 5.45,
1353
+ "learning_rate": 2.1630781499202555e-05,
1354
+ "loss": 0.1982,
1355
+ "step": 1085
1356
+ },
1357
+ {
1358
+ "epoch": 5.48,
1359
+ "learning_rate": 2.1730462519936203e-05,
1360
+ "loss": 0.1917,
1361
+ "step": 1090
1362
+ },
1363
+ {
1364
+ "epoch": 5.5,
1365
+ "learning_rate": 2.1830143540669855e-05,
1366
+ "loss": 0.3343,
1367
+ "step": 1095
1368
+ },
1369
+ {
1370
+ "epoch": 5.53,
1371
+ "learning_rate": 2.1929824561403507e-05,
1372
+ "loss": 0.1096,
1373
+ "step": 1100
1374
+ },
1375
+ {
1376
+ "epoch": 5.55,
1377
+ "learning_rate": 2.2029505582137162e-05,
1378
+ "loss": 0.1774,
1379
+ "step": 1105
1380
+ },
1381
+ {
1382
+ "epoch": 5.58,
1383
+ "learning_rate": 2.2129186602870814e-05,
1384
+ "loss": 0.1189,
1385
+ "step": 1110
1386
+ },
1387
+ {
1388
+ "epoch": 5.6,
1389
+ "learning_rate": 2.2228867623604466e-05,
1390
+ "loss": 0.1268,
1391
+ "step": 1115
1392
+ },
1393
+ {
1394
+ "epoch": 5.63,
1395
+ "learning_rate": 2.2328548644338118e-05,
1396
+ "loss": 0.0761,
1397
+ "step": 1120
1398
+ },
1399
+ {
1400
+ "epoch": 5.65,
1401
+ "learning_rate": 2.242822966507177e-05,
1402
+ "loss": 0.0904,
1403
+ "step": 1125
1404
+ },
1405
+ {
1406
+ "epoch": 5.68,
1407
+ "learning_rate": 2.2527910685805422e-05,
1408
+ "loss": 0.088,
1409
+ "step": 1130
1410
+ },
1411
+ {
1412
+ "epoch": 5.7,
1413
+ "learning_rate": 2.2627591706539074e-05,
1414
+ "loss": 0.0607,
1415
+ "step": 1135
1416
+ },
1417
+ {
1418
+ "epoch": 5.73,
1419
+ "learning_rate": 2.272727272727273e-05,
1420
+ "loss": 0.1923,
1421
+ "step": 1140
1422
+ },
1423
+ {
1424
+ "epoch": 5.75,
1425
+ "learning_rate": 2.282695374800638e-05,
1426
+ "loss": 0.1361,
1427
+ "step": 1145
1428
+ },
1429
+ {
1430
+ "epoch": 5.78,
1431
+ "learning_rate": 2.2926634768740033e-05,
1432
+ "loss": 0.1227,
1433
+ "step": 1150
1434
+ },
1435
+ {
1436
+ "epoch": 5.8,
1437
+ "learning_rate": 2.3026315789473685e-05,
1438
+ "loss": 0.1345,
1439
+ "step": 1155
1440
+ },
1441
+ {
1442
+ "epoch": 5.83,
1443
+ "learning_rate": 2.3125996810207337e-05,
1444
+ "loss": 0.179,
1445
+ "step": 1160
1446
+ },
1447
+ {
1448
+ "epoch": 5.85,
1449
+ "learning_rate": 2.322567783094099e-05,
1450
+ "loss": 0.1282,
1451
+ "step": 1165
1452
+ },
1453
+ {
1454
+ "epoch": 5.88,
1455
+ "learning_rate": 2.332535885167464e-05,
1456
+ "loss": 0.1503,
1457
+ "step": 1170
1458
+ },
1459
+ {
1460
+ "epoch": 5.9,
1461
+ "learning_rate": 2.3425039872408293e-05,
1462
+ "loss": 0.0902,
1463
+ "step": 1175
1464
+ },
1465
+ {
1466
+ "epoch": 5.93,
1467
+ "learning_rate": 2.3524720893141948e-05,
1468
+ "loss": 0.1091,
1469
+ "step": 1180
1470
+ },
1471
+ {
1472
+ "epoch": 5.95,
1473
+ "learning_rate": 2.36244019138756e-05,
1474
+ "loss": 0.0839,
1475
+ "step": 1185
1476
+ },
1477
+ {
1478
+ "epoch": 5.98,
1479
+ "learning_rate": 2.3724082934609252e-05,
1480
+ "loss": 0.1007,
1481
+ "step": 1190
1482
+ },
1483
+ {
1484
+ "epoch": 6.0,
1485
+ "eval_accuracy": 0.9769085768143261,
1486
+ "eval_loss": 0.07253596186637878,
1487
+ "eval_runtime": 70.8263,
1488
+ "eval_samples_per_second": 29.961,
1489
+ "eval_steps_per_second": 1.878,
1490
+ "step": 1194
1491
+ },
1492
+ {
1493
+ "epoch": 6.01,
1494
+ "learning_rate": 2.3823763955342904e-05,
1495
+ "loss": 0.0863,
1496
+ "step": 1195
1497
+ },
1498
+ {
1499
+ "epoch": 6.03,
1500
+ "learning_rate": 2.3923444976076556e-05,
1501
+ "loss": 0.088,
1502
+ "step": 1200
1503
+ },
1504
+ {
1505
+ "epoch": 6.06,
1506
+ "learning_rate": 2.4023125996810208e-05,
1507
+ "loss": 0.0649,
1508
+ "step": 1205
1509
+ },
1510
+ {
1511
+ "epoch": 6.08,
1512
+ "learning_rate": 2.412280701754386e-05,
1513
+ "loss": 0.0992,
1514
+ "step": 1210
1515
+ },
1516
+ {
1517
+ "epoch": 6.11,
1518
+ "learning_rate": 2.4222488038277515e-05,
1519
+ "loss": 0.223,
1520
+ "step": 1215
1521
+ },
1522
+ {
1523
+ "epoch": 6.13,
1524
+ "learning_rate": 2.4322169059011167e-05,
1525
+ "loss": 0.1472,
1526
+ "step": 1220
1527
+ },
1528
+ {
1529
+ "epoch": 6.16,
1530
+ "learning_rate": 2.442185007974482e-05,
1531
+ "loss": 0.1692,
1532
+ "step": 1225
1533
+ },
1534
+ {
1535
+ "epoch": 6.18,
1536
+ "learning_rate": 2.452153110047847e-05,
1537
+ "loss": 0.109,
1538
+ "step": 1230
1539
+ },
1540
+ {
1541
+ "epoch": 6.21,
1542
+ "learning_rate": 2.4621212121212123e-05,
1543
+ "loss": 0.0773,
1544
+ "step": 1235
1545
+ },
1546
+ {
1547
+ "epoch": 6.23,
1548
+ "learning_rate": 2.4720893141945774e-05,
1549
+ "loss": 0.0989,
1550
+ "step": 1240
1551
+ },
1552
+ {
1553
+ "epoch": 6.26,
1554
+ "learning_rate": 2.4820574162679426e-05,
1555
+ "loss": 0.0715,
1556
+ "step": 1245
1557
+ },
1558
+ {
1559
+ "epoch": 6.28,
1560
+ "learning_rate": 2.492025518341308e-05,
1561
+ "loss": 0.0389,
1562
+ "step": 1250
1563
+ },
1564
+ {
1565
+ "epoch": 6.31,
1566
+ "learning_rate": 2.5019936204146734e-05,
1567
+ "loss": 0.1711,
1568
+ "step": 1255
1569
+ },
1570
+ {
1571
+ "epoch": 6.33,
1572
+ "learning_rate": 2.5119617224880382e-05,
1573
+ "loss": 0.0931,
1574
+ "step": 1260
1575
+ },
1576
+ {
1577
+ "epoch": 6.36,
1578
+ "learning_rate": 2.5219298245614037e-05,
1579
+ "loss": 0.031,
1580
+ "step": 1265
1581
+ },
1582
+ {
1583
+ "epoch": 6.38,
1584
+ "learning_rate": 2.5318979266347686e-05,
1585
+ "loss": 0.1442,
1586
+ "step": 1270
1587
+ },
1588
+ {
1589
+ "epoch": 6.41,
1590
+ "learning_rate": 2.541866028708134e-05,
1591
+ "loss": 0.1046,
1592
+ "step": 1275
1593
+ },
1594
+ {
1595
+ "epoch": 6.43,
1596
+ "learning_rate": 2.551834130781499e-05,
1597
+ "loss": 0.0885,
1598
+ "step": 1280
1599
+ },
1600
+ {
1601
+ "epoch": 6.46,
1602
+ "learning_rate": 2.5618022328548645e-05,
1603
+ "loss": 0.1875,
1604
+ "step": 1285
1605
+ },
1606
+ {
1607
+ "epoch": 6.48,
1608
+ "learning_rate": 2.5717703349282297e-05,
1609
+ "loss": 0.0669,
1610
+ "step": 1290
1611
+ },
1612
+ {
1613
+ "epoch": 6.51,
1614
+ "learning_rate": 2.5817384370015952e-05,
1615
+ "loss": 0.1147,
1616
+ "step": 1295
1617
+ },
1618
+ {
1619
+ "epoch": 6.53,
1620
+ "learning_rate": 2.59170653907496e-05,
1621
+ "loss": 0.233,
1622
+ "step": 1300
1623
+ },
1624
+ {
1625
+ "epoch": 6.56,
1626
+ "learning_rate": 2.6016746411483256e-05,
1627
+ "loss": 0.2635,
1628
+ "step": 1305
1629
+ },
1630
+ {
1631
+ "epoch": 6.58,
1632
+ "learning_rate": 2.6116427432216905e-05,
1633
+ "loss": 0.1609,
1634
+ "step": 1310
1635
+ },
1636
+ {
1637
+ "epoch": 6.61,
1638
+ "learning_rate": 2.621610845295056e-05,
1639
+ "loss": 0.1306,
1640
+ "step": 1315
1641
+ },
1642
+ {
1643
+ "epoch": 6.63,
1644
+ "learning_rate": 2.6315789473684212e-05,
1645
+ "loss": 0.2663,
1646
+ "step": 1320
1647
+ },
1648
+ {
1649
+ "epoch": 6.66,
1650
+ "learning_rate": 2.6415470494417864e-05,
1651
+ "loss": 0.1074,
1652
+ "step": 1325
1653
+ },
1654
+ {
1655
+ "epoch": 6.68,
1656
+ "learning_rate": 2.6515151515151516e-05,
1657
+ "loss": 0.0561,
1658
+ "step": 1330
1659
+ },
1660
+ {
1661
+ "epoch": 6.71,
1662
+ "learning_rate": 2.661483253588517e-05,
1663
+ "loss": 0.043,
1664
+ "step": 1335
1665
+ },
1666
+ {
1667
+ "epoch": 6.73,
1668
+ "learning_rate": 2.671451355661882e-05,
1669
+ "loss": 0.1899,
1670
+ "step": 1340
1671
+ },
1672
+ {
1673
+ "epoch": 6.76,
1674
+ "learning_rate": 2.6814194577352475e-05,
1675
+ "loss": 0.0969,
1676
+ "step": 1345
1677
+ },
1678
+ {
1679
+ "epoch": 6.78,
1680
+ "learning_rate": 2.6913875598086123e-05,
1681
+ "loss": 0.2774,
1682
+ "step": 1350
1683
+ },
1684
+ {
1685
+ "epoch": 6.81,
1686
+ "learning_rate": 2.701355661881978e-05,
1687
+ "loss": 0.1214,
1688
+ "step": 1355
1689
+ },
1690
+ {
1691
+ "epoch": 6.83,
1692
+ "learning_rate": 2.711323763955343e-05,
1693
+ "loss": 0.2666,
1694
+ "step": 1360
1695
+ },
1696
+ {
1697
+ "epoch": 6.86,
1698
+ "learning_rate": 2.7212918660287086e-05,
1699
+ "loss": 0.1761,
1700
+ "step": 1365
1701
+ },
1702
+ {
1703
+ "epoch": 6.88,
1704
+ "learning_rate": 2.7312599681020735e-05,
1705
+ "loss": 0.0554,
1706
+ "step": 1370
1707
+ },
1708
+ {
1709
+ "epoch": 6.91,
1710
+ "learning_rate": 2.741228070175439e-05,
1711
+ "loss": 0.0588,
1712
+ "step": 1375
1713
+ },
1714
+ {
1715
+ "epoch": 6.93,
1716
+ "learning_rate": 2.751196172248804e-05,
1717
+ "loss": 0.0842,
1718
+ "step": 1380
1719
+ },
1720
+ {
1721
+ "epoch": 6.96,
1722
+ "learning_rate": 2.7611642743221694e-05,
1723
+ "loss": 0.1822,
1724
+ "step": 1385
1725
+ },
1726
+ {
1727
+ "epoch": 6.98,
1728
+ "learning_rate": 2.7711323763955342e-05,
1729
+ "loss": 0.1463,
1730
+ "step": 1390
1731
+ },
1732
+ {
1733
+ "epoch": 7.0,
1734
+ "eval_accuracy": 0.9839773798303487,
1735
+ "eval_loss": 0.05405208095908165,
1736
+ "eval_runtime": 70.9415,
1737
+ "eval_samples_per_second": 29.912,
1738
+ "eval_steps_per_second": 1.875,
1739
+ "step": 1393
1740
+ },
1741
+ {
1742
+ "epoch": 7.01,
1743
+ "learning_rate": 2.7811004784688998e-05,
1744
+ "loss": 0.0536,
1745
+ "step": 1395
1746
+ },
1747
+ {
1748
+ "epoch": 7.04,
1749
+ "learning_rate": 2.791068580542265e-05,
1750
+ "loss": 0.1287,
1751
+ "step": 1400
1752
+ },
1753
+ {
1754
+ "epoch": 7.06,
1755
+ "learning_rate": 2.8010366826156305e-05,
1756
+ "loss": 0.0914,
1757
+ "step": 1405
1758
+ },
1759
+ {
1760
+ "epoch": 7.09,
1761
+ "learning_rate": 2.8110047846889953e-05,
1762
+ "loss": 0.0428,
1763
+ "step": 1410
1764
+ },
1765
+ {
1766
+ "epoch": 7.11,
1767
+ "learning_rate": 2.820972886762361e-05,
1768
+ "loss": 0.213,
1769
+ "step": 1415
1770
+ },
1771
+ {
1772
+ "epoch": 7.14,
1773
+ "learning_rate": 2.8309409888357257e-05,
1774
+ "loss": 0.0962,
1775
+ "step": 1420
1776
+ },
1777
+ {
1778
+ "epoch": 7.16,
1779
+ "learning_rate": 2.8409090909090912e-05,
1780
+ "loss": 0.0824,
1781
+ "step": 1425
1782
+ },
1783
+ {
1784
+ "epoch": 7.19,
1785
+ "learning_rate": 2.850877192982456e-05,
1786
+ "loss": 0.1157,
1787
+ "step": 1430
1788
+ },
1789
+ {
1790
+ "epoch": 7.21,
1791
+ "learning_rate": 2.8608452950558216e-05,
1792
+ "loss": 0.1519,
1793
+ "step": 1435
1794
+ },
1795
+ {
1796
+ "epoch": 7.24,
1797
+ "learning_rate": 2.8708133971291868e-05,
1798
+ "loss": 0.2063,
1799
+ "step": 1440
1800
+ },
1801
+ {
1802
+ "epoch": 7.26,
1803
+ "learning_rate": 2.8807814992025524e-05,
1804
+ "loss": 0.1612,
1805
+ "step": 1445
1806
+ },
1807
+ {
1808
+ "epoch": 7.29,
1809
+ "learning_rate": 2.8907496012759172e-05,
1810
+ "loss": 0.0942,
1811
+ "step": 1450
1812
+ },
1813
+ {
1814
+ "epoch": 7.31,
1815
+ "learning_rate": 2.900717703349282e-05,
1816
+ "loss": 0.0405,
1817
+ "step": 1455
1818
+ },
1819
+ {
1820
+ "epoch": 7.34,
1821
+ "learning_rate": 2.9106858054226476e-05,
1822
+ "loss": 0.1686,
1823
+ "step": 1460
1824
+ },
1825
+ {
1826
+ "epoch": 7.36,
1827
+ "learning_rate": 2.9206539074960128e-05,
1828
+ "loss": 0.2167,
1829
+ "step": 1465
1830
+ },
1831
+ {
1832
+ "epoch": 7.39,
1833
+ "learning_rate": 2.9306220095693783e-05,
1834
+ "loss": 0.0604,
1835
+ "step": 1470
1836
+ },
1837
+ {
1838
+ "epoch": 7.41,
1839
+ "learning_rate": 2.940590111642743e-05,
1840
+ "loss": 0.1009,
1841
+ "step": 1475
1842
+ },
1843
+ {
1844
+ "epoch": 7.44,
1845
+ "learning_rate": 2.9505582137161087e-05,
1846
+ "loss": 0.1279,
1847
+ "step": 1480
1848
+ },
1849
+ {
1850
+ "epoch": 7.46,
1851
+ "learning_rate": 2.9605263157894735e-05,
1852
+ "loss": 0.1413,
1853
+ "step": 1485
1854
+ },
1855
+ {
1856
+ "epoch": 7.49,
1857
+ "learning_rate": 2.970494417862839e-05,
1858
+ "loss": 0.1586,
1859
+ "step": 1490
1860
+ },
1861
+ {
1862
+ "epoch": 7.51,
1863
+ "learning_rate": 2.980462519936204e-05,
1864
+ "loss": 0.1487,
1865
+ "step": 1495
1866
+ },
1867
+ {
1868
+ "epoch": 7.54,
1869
+ "learning_rate": 2.9904306220095695e-05,
1870
+ "loss": 0.0528,
1871
+ "step": 1500
1872
+ },
1873
+ {
1874
+ "epoch": 7.56,
1875
+ "learning_rate": 3.0003987240829347e-05,
1876
+ "loss": 0.1796,
1877
+ "step": 1505
1878
+ },
1879
+ {
1880
+ "epoch": 7.59,
1881
+ "learning_rate": 3.0103668261563002e-05,
1882
+ "loss": 0.1781,
1883
+ "step": 1510
1884
+ },
1885
+ {
1886
+ "epoch": 7.61,
1887
+ "learning_rate": 3.020334928229665e-05,
1888
+ "loss": 0.1745,
1889
+ "step": 1515
1890
+ },
1891
+ {
1892
+ "epoch": 7.64,
1893
+ "learning_rate": 3.0303030303030306e-05,
1894
+ "loss": 0.0827,
1895
+ "step": 1520
1896
+ },
1897
+ {
1898
+ "epoch": 7.66,
1899
+ "learning_rate": 3.0402711323763954e-05,
1900
+ "loss": 0.1123,
1901
+ "step": 1525
1902
+ },
1903
+ {
1904
+ "epoch": 7.69,
1905
+ "learning_rate": 3.050239234449761e-05,
1906
+ "loss": 0.0751,
1907
+ "step": 1530
1908
+ },
1909
+ {
1910
+ "epoch": 7.71,
1911
+ "learning_rate": 3.060207336523126e-05,
1912
+ "loss": 0.1033,
1913
+ "step": 1535
1914
+ },
1915
+ {
1916
+ "epoch": 7.74,
1917
+ "learning_rate": 3.0701754385964913e-05,
1918
+ "loss": 0.0861,
1919
+ "step": 1540
1920
+ },
1921
+ {
1922
+ "epoch": 7.76,
1923
+ "learning_rate": 3.080143540669856e-05,
1924
+ "loss": 0.0744,
1925
+ "step": 1545
1926
+ },
1927
+ {
1928
+ "epoch": 7.79,
1929
+ "learning_rate": 3.090111642743222e-05,
1930
+ "loss": 0.1511,
1931
+ "step": 1550
1932
+ },
1933
+ {
1934
+ "epoch": 7.81,
1935
+ "learning_rate": 3.100079744816587e-05,
1936
+ "loss": 0.1331,
1937
+ "step": 1555
1938
+ },
1939
+ {
1940
+ "epoch": 7.84,
1941
+ "learning_rate": 3.110047846889952e-05,
1942
+ "loss": 0.04,
1943
+ "step": 1560
1944
+ },
1945
+ {
1946
+ "epoch": 7.86,
1947
+ "learning_rate": 3.1200159489633176e-05,
1948
+ "loss": 0.0549,
1949
+ "step": 1565
1950
+ },
1951
+ {
1952
+ "epoch": 7.89,
1953
+ "learning_rate": 3.129984051036683e-05,
1954
+ "loss": 0.1301,
1955
+ "step": 1570
1956
+ },
1957
+ {
1958
+ "epoch": 7.91,
1959
+ "learning_rate": 3.139952153110048e-05,
1960
+ "loss": 0.0605,
1961
+ "step": 1575
1962
+ },
1963
+ {
1964
+ "epoch": 7.94,
1965
+ "learning_rate": 3.1499202551834136e-05,
1966
+ "loss": 0.1689,
1967
+ "step": 1580
1968
+ },
1969
+ {
1970
+ "epoch": 7.96,
1971
+ "learning_rate": 3.1598883572567784e-05,
1972
+ "loss": 0.4188,
1973
+ "step": 1585
1974
+ },
1975
+ {
1976
+ "epoch": 7.99,
1977
+ "learning_rate": 3.169856459330144e-05,
1978
+ "loss": 0.3564,
1979
+ "step": 1590
1980
+ },
1981
+ {
1982
+ "epoch": 8.0,
1983
+ "eval_accuracy": 0.9802073515551367,
1984
+ "eval_loss": 0.08797016739845276,
1985
+ "eval_runtime": 70.691,
1986
+ "eval_samples_per_second": 30.018,
1987
+ "eval_steps_per_second": 1.881,
1988
+ "step": 1592
1989
+ },
1990
+ {
1991
+ "epoch": 8.02,
1992
+ "learning_rate": 3.179824561403509e-05,
1993
+ "loss": 0.1236,
1994
+ "step": 1595
1995
+ },
1996
+ {
1997
+ "epoch": 8.04,
1998
+ "learning_rate": 3.189792663476874e-05,
1999
+ "loss": 0.0615,
2000
+ "step": 1600
2001
+ },
2002
+ {
2003
+ "epoch": 8.07,
2004
+ "learning_rate": 3.199760765550239e-05,
2005
+ "loss": 0.2747,
2006
+ "step": 1605
2007
+ },
2008
+ {
2009
+ "epoch": 8.09,
2010
+ "learning_rate": 3.209728867623605e-05,
2011
+ "loss": 0.1151,
2012
+ "step": 1610
2013
+ },
2014
+ {
2015
+ "epoch": 8.12,
2016
+ "learning_rate": 3.2196969696969696e-05,
2017
+ "loss": 0.1025,
2018
+ "step": 1615
2019
+ },
2020
+ {
2021
+ "epoch": 8.14,
2022
+ "learning_rate": 3.229665071770335e-05,
2023
+ "loss": 0.2233,
2024
+ "step": 1620
2025
+ },
2026
+ {
2027
+ "epoch": 8.17,
2028
+ "learning_rate": 3.2396331738437e-05,
2029
+ "loss": 0.0602,
2030
+ "step": 1625
2031
+ },
2032
+ {
2033
+ "epoch": 8.19,
2034
+ "learning_rate": 3.2496012759170655e-05,
2035
+ "loss": 0.0924,
2036
+ "step": 1630
2037
+ },
2038
+ {
2039
+ "epoch": 8.22,
2040
+ "learning_rate": 3.259569377990431e-05,
2041
+ "loss": 0.1035,
2042
+ "step": 1635
2043
+ },
2044
+ {
2045
+ "epoch": 8.24,
2046
+ "learning_rate": 3.2695374800637965e-05,
2047
+ "loss": 0.1069,
2048
+ "step": 1640
2049
+ },
2050
+ {
2051
+ "epoch": 8.27,
2052
+ "learning_rate": 3.2795055821371614e-05,
2053
+ "loss": 0.2335,
2054
+ "step": 1645
2055
+ },
2056
+ {
2057
+ "epoch": 8.29,
2058
+ "learning_rate": 3.289473684210527e-05,
2059
+ "loss": 0.1592,
2060
+ "step": 1650
2061
+ },
2062
+ {
2063
+ "epoch": 8.32,
2064
+ "learning_rate": 3.299441786283892e-05,
2065
+ "loss": 0.0557,
2066
+ "step": 1655
2067
+ },
2068
+ {
2069
+ "epoch": 8.34,
2070
+ "learning_rate": 3.309409888357257e-05,
2071
+ "loss": 0.0985,
2072
+ "step": 1660
2073
+ },
2074
+ {
2075
+ "epoch": 8.37,
2076
+ "learning_rate": 3.319377990430622e-05,
2077
+ "loss": 0.0447,
2078
+ "step": 1665
2079
+ },
2080
+ {
2081
+ "epoch": 8.39,
2082
+ "learning_rate": 3.329346092503988e-05,
2083
+ "loss": 0.2291,
2084
+ "step": 1670
2085
+ },
2086
+ {
2087
+ "epoch": 8.42,
2088
+ "learning_rate": 3.3393141945773525e-05,
2089
+ "loss": 0.1829,
2090
+ "step": 1675
2091
+ },
2092
+ {
2093
+ "epoch": 8.44,
2094
+ "learning_rate": 3.349282296650718e-05,
2095
+ "loss": 0.1253,
2096
+ "step": 1680
2097
+ },
2098
+ {
2099
+ "epoch": 8.47,
2100
+ "learning_rate": 3.359250398724083e-05,
2101
+ "loss": 0.0124,
2102
+ "step": 1685
2103
+ },
2104
+ {
2105
+ "epoch": 8.49,
2106
+ "learning_rate": 3.369218500797448e-05,
2107
+ "loss": 0.0723,
2108
+ "step": 1690
2109
+ },
2110
+ {
2111
+ "epoch": 8.52,
2112
+ "learning_rate": 3.379186602870813e-05,
2113
+ "loss": 0.1938,
2114
+ "step": 1695
2115
+ },
2116
+ {
2117
+ "epoch": 8.54,
2118
+ "learning_rate": 3.389154704944179e-05,
2119
+ "loss": 0.2344,
2120
+ "step": 1700
2121
+ },
2122
+ {
2123
+ "epoch": 8.57,
2124
+ "learning_rate": 3.3991228070175444e-05,
2125
+ "loss": 0.2282,
2126
+ "step": 1705
2127
+ },
2128
+ {
2129
+ "epoch": 8.59,
2130
+ "learning_rate": 3.409090909090909e-05,
2131
+ "loss": 0.1385,
2132
+ "step": 1710
2133
+ },
2134
+ {
2135
+ "epoch": 8.62,
2136
+ "learning_rate": 3.419059011164275e-05,
2137
+ "loss": 0.2466,
2138
+ "step": 1715
2139
+ },
2140
+ {
2141
+ "epoch": 8.64,
2142
+ "learning_rate": 3.4290271132376396e-05,
2143
+ "loss": 0.1208,
2144
+ "step": 1720
2145
+ },
2146
+ {
2147
+ "epoch": 8.67,
2148
+ "learning_rate": 3.438995215311005e-05,
2149
+ "loss": 0.091,
2150
+ "step": 1725
2151
+ },
2152
+ {
2153
+ "epoch": 8.69,
2154
+ "learning_rate": 3.44896331738437e-05,
2155
+ "loss": 0.0724,
2156
+ "step": 1730
2157
+ },
2158
+ {
2159
+ "epoch": 8.72,
2160
+ "learning_rate": 3.4589314194577355e-05,
2161
+ "loss": 0.1488,
2162
+ "step": 1735
2163
+ },
2164
+ {
2165
+ "epoch": 8.74,
2166
+ "learning_rate": 3.4688995215311004e-05,
2167
+ "loss": 0.0948,
2168
+ "step": 1740
2169
+ },
2170
+ {
2171
+ "epoch": 8.77,
2172
+ "learning_rate": 3.478867623604466e-05,
2173
+ "loss": 0.0663,
2174
+ "step": 1745
2175
+ },
2176
+ {
2177
+ "epoch": 8.79,
2178
+ "learning_rate": 3.488835725677831e-05,
2179
+ "loss": 0.0654,
2180
+ "step": 1750
2181
+ },
2182
+ {
2183
+ "epoch": 8.82,
2184
+ "learning_rate": 3.498803827751196e-05,
2185
+ "loss": 0.036,
2186
+ "step": 1755
2187
+ },
2188
+ {
2189
+ "epoch": 8.84,
2190
+ "learning_rate": 3.508771929824561e-05,
2191
+ "loss": 0.1179,
2192
+ "step": 1760
2193
+ },
2194
+ {
2195
+ "epoch": 8.87,
2196
+ "learning_rate": 3.518740031897927e-05,
2197
+ "loss": 0.1061,
2198
+ "step": 1765
2199
+ },
2200
+ {
2201
+ "epoch": 8.89,
2202
+ "learning_rate": 3.5287081339712915e-05,
2203
+ "loss": 0.0765,
2204
+ "step": 1770
2205
+ },
2206
+ {
2207
+ "epoch": 8.92,
2208
+ "learning_rate": 3.538676236044657e-05,
2209
+ "loss": 0.0466,
2210
+ "step": 1775
2211
+ },
2212
+ {
2213
+ "epoch": 8.94,
2214
+ "learning_rate": 3.5486443381180226e-05,
2215
+ "loss": 0.113,
2216
+ "step": 1780
2217
+ },
2218
+ {
2219
+ "epoch": 8.97,
2220
+ "learning_rate": 3.558612440191388e-05,
2221
+ "loss": 0.1163,
2222
+ "step": 1785
2223
+ },
2224
+ {
2225
+ "epoch": 8.99,
2226
+ "learning_rate": 3.568580542264753e-05,
2227
+ "loss": 0.0957,
2228
+ "step": 1790
2229
+ },
2230
+ {
2231
+ "epoch": 9.0,
2232
+ "eval_accuracy": 0.96559849198869,
2233
+ "eval_loss": 0.13746494054794312,
2234
+ "eval_runtime": 70.9091,
2235
+ "eval_samples_per_second": 29.926,
2236
+ "eval_steps_per_second": 1.876,
2237
+ "step": 1791
2238
+ },
2239
+ {
2240
+ "epoch": 9.02,
2241
+ "learning_rate": 3.5785486443381185e-05,
2242
+ "loss": 0.1195,
2243
+ "step": 1795
2244
+ },
2245
+ {
2246
+ "epoch": 9.05,
2247
+ "learning_rate": 3.5885167464114834e-05,
2248
+ "loss": 0.1754,
2249
+ "step": 1800
2250
+ },
2251
+ {
2252
+ "epoch": 9.07,
2253
+ "learning_rate": 3.598484848484849e-05,
2254
+ "loss": 0.1201,
2255
+ "step": 1805
2256
+ },
2257
+ {
2258
+ "epoch": 9.1,
2259
+ "learning_rate": 3.608452950558214e-05,
2260
+ "loss": 0.1016,
2261
+ "step": 1810
2262
+ },
2263
+ {
2264
+ "epoch": 9.12,
2265
+ "learning_rate": 3.618421052631579e-05,
2266
+ "loss": 0.2312,
2267
+ "step": 1815
2268
+ },
2269
+ {
2270
+ "epoch": 9.15,
2271
+ "learning_rate": 3.628389154704944e-05,
2272
+ "loss": 0.0625,
2273
+ "step": 1820
2274
+ },
2275
+ {
2276
+ "epoch": 9.17,
2277
+ "learning_rate": 3.6383572567783097e-05,
2278
+ "loss": 0.0684,
2279
+ "step": 1825
2280
+ },
2281
+ {
2282
+ "epoch": 9.2,
2283
+ "learning_rate": 3.6483253588516745e-05,
2284
+ "loss": 0.0445,
2285
+ "step": 1830
2286
+ },
2287
+ {
2288
+ "epoch": 9.22,
2289
+ "learning_rate": 3.65829346092504e-05,
2290
+ "loss": 0.0355,
2291
+ "step": 1835
2292
+ },
2293
+ {
2294
+ "epoch": 9.25,
2295
+ "learning_rate": 3.668261562998405e-05,
2296
+ "loss": 0.216,
2297
+ "step": 1840
2298
+ },
2299
+ {
2300
+ "epoch": 9.27,
2301
+ "learning_rate": 3.6782296650717704e-05,
2302
+ "loss": 0.1667,
2303
+ "step": 1845
2304
+ },
2305
+ {
2306
+ "epoch": 9.3,
2307
+ "learning_rate": 3.688197767145136e-05,
2308
+ "loss": 0.1845,
2309
+ "step": 1850
2310
+ },
2311
+ {
2312
+ "epoch": 9.32,
2313
+ "learning_rate": 3.6981658692185015e-05,
2314
+ "loss": 0.0561,
2315
+ "step": 1855
2316
+ },
2317
+ {
2318
+ "epoch": 9.35,
2319
+ "learning_rate": 3.7081339712918663e-05,
2320
+ "loss": 0.2596,
2321
+ "step": 1860
2322
+ },
2323
+ {
2324
+ "epoch": 9.37,
2325
+ "learning_rate": 3.718102073365232e-05,
2326
+ "loss": 0.3591,
2327
+ "step": 1865
2328
+ },
2329
+ {
2330
+ "epoch": 9.4,
2331
+ "learning_rate": 3.728070175438597e-05,
2332
+ "loss": 0.1171,
2333
+ "step": 1870
2334
+ },
2335
+ {
2336
+ "epoch": 9.42,
2337
+ "learning_rate": 3.738038277511962e-05,
2338
+ "loss": 0.0377,
2339
+ "step": 1875
2340
+ },
2341
+ {
2342
+ "epoch": 9.45,
2343
+ "learning_rate": 3.748006379585327e-05,
2344
+ "loss": 0.1234,
2345
+ "step": 1880
2346
+ },
2347
+ {
2348
+ "epoch": 9.47,
2349
+ "learning_rate": 3.7579744816586926e-05,
2350
+ "loss": 0.1602,
2351
+ "step": 1885
2352
+ },
2353
+ {
2354
+ "epoch": 9.5,
2355
+ "learning_rate": 3.7679425837320575e-05,
2356
+ "loss": 0.0583,
2357
+ "step": 1890
2358
+ },
2359
+ {
2360
+ "epoch": 9.52,
2361
+ "learning_rate": 3.777910685805423e-05,
2362
+ "loss": 0.1777,
2363
+ "step": 1895
2364
+ },
2365
+ {
2366
+ "epoch": 9.55,
2367
+ "learning_rate": 3.787878787878788e-05,
2368
+ "loss": 0.1032,
2369
+ "step": 1900
2370
+ },
2371
+ {
2372
+ "epoch": 9.57,
2373
+ "learning_rate": 3.7978468899521534e-05,
2374
+ "loss": 0.1959,
2375
+ "step": 1905
2376
+ },
2377
+ {
2378
+ "epoch": 9.6,
2379
+ "learning_rate": 3.807814992025518e-05,
2380
+ "loss": 0.1256,
2381
+ "step": 1910
2382
+ },
2383
+ {
2384
+ "epoch": 9.62,
2385
+ "learning_rate": 3.817783094098884e-05,
2386
+ "loss": 0.1693,
2387
+ "step": 1915
2388
+ },
2389
+ {
2390
+ "epoch": 9.65,
2391
+ "learning_rate": 3.8277511961722486e-05,
2392
+ "loss": 0.1142,
2393
+ "step": 1920
2394
+ },
2395
+ {
2396
+ "epoch": 9.67,
2397
+ "learning_rate": 3.837719298245614e-05,
2398
+ "loss": 0.1055,
2399
+ "step": 1925
2400
+ },
2401
+ {
2402
+ "epoch": 9.7,
2403
+ "learning_rate": 3.84768740031898e-05,
2404
+ "loss": 0.0944,
2405
+ "step": 1930
2406
+ },
2407
+ {
2408
+ "epoch": 9.72,
2409
+ "learning_rate": 3.8576555023923446e-05,
2410
+ "loss": 0.081,
2411
+ "step": 1935
2412
+ },
2413
+ {
2414
+ "epoch": 9.75,
2415
+ "learning_rate": 3.86762360446571e-05,
2416
+ "loss": 0.1426,
2417
+ "step": 1940
2418
+ },
2419
+ {
2420
+ "epoch": 9.77,
2421
+ "learning_rate": 3.877591706539075e-05,
2422
+ "loss": 0.1158,
2423
+ "step": 1945
2424
+ },
2425
+ {
2426
+ "epoch": 9.8,
2427
+ "learning_rate": 3.8875598086124405e-05,
2428
+ "loss": 0.0763,
2429
+ "step": 1950
2430
+ },
2431
+ {
2432
+ "epoch": 9.82,
2433
+ "learning_rate": 3.897527910685805e-05,
2434
+ "loss": 0.0932,
2435
+ "step": 1955
2436
+ },
2437
+ {
2438
+ "epoch": 9.85,
2439
+ "learning_rate": 3.907496012759171e-05,
2440
+ "loss": 0.0565,
2441
+ "step": 1960
2442
+ },
2443
+ {
2444
+ "epoch": 9.87,
2445
+ "learning_rate": 3.917464114832536e-05,
2446
+ "loss": 0.0578,
2447
+ "step": 1965
2448
+ },
2449
+ {
2450
+ "epoch": 9.9,
2451
+ "learning_rate": 3.927432216905901e-05,
2452
+ "loss": 0.0449,
2453
+ "step": 1970
2454
+ },
2455
+ {
2456
+ "epoch": 9.92,
2457
+ "learning_rate": 3.937400318979266e-05,
2458
+ "loss": 0.0792,
2459
+ "step": 1975
2460
+ },
2461
+ {
2462
+ "epoch": 9.95,
2463
+ "learning_rate": 3.9473684210526316e-05,
2464
+ "loss": 0.0463,
2465
+ "step": 1980
2466
+ },
2467
+ {
2468
+ "epoch": 9.97,
2469
+ "learning_rate": 3.9573365231259965e-05,
2470
+ "loss": 0.2275,
2471
+ "step": 1985
2472
+ },
2473
+ {
2474
+ "epoch": 10.0,
2475
+ "learning_rate": 3.967304625199362e-05,
2476
+ "loss": 0.1481,
2477
+ "step": 1990
2478
+ },
2479
+ {
2480
+ "epoch": 10.0,
2481
+ "eval_accuracy": 0.9872761545711592,
2482
+ "eval_loss": 0.051094409078359604,
2483
+ "eval_runtime": 70.5765,
2484
+ "eval_samples_per_second": 30.067,
2485
+ "eval_steps_per_second": 1.884,
2486
+ "step": 1990
2487
+ },
2488
+ {
2489
+ "epoch": 10.03,
2490
+ "learning_rate": 3.9772727272727275e-05,
2491
+ "loss": 0.0831,
2492
+ "step": 1995
2493
+ },
2494
+ {
2495
+ "epoch": 10.05,
2496
+ "learning_rate": 3.987240829346093e-05,
2497
+ "loss": 0.1587,
2498
+ "step": 2000
2499
+ },
2500
+ {
2501
+ "epoch": 10.08,
2502
+ "learning_rate": 3.997208931419458e-05,
2503
+ "loss": 0.0135,
2504
+ "step": 2005
2505
+ },
2506
+ {
2507
+ "epoch": 10.1,
2508
+ "learning_rate": 4.0071770334928235e-05,
2509
+ "loss": 0.1272,
2510
+ "step": 2010
2511
+ },
2512
+ {
2513
+ "epoch": 10.13,
2514
+ "learning_rate": 4.017145135566188e-05,
2515
+ "loss": 0.0556,
2516
+ "step": 2015
2517
+ },
2518
+ {
2519
+ "epoch": 10.15,
2520
+ "learning_rate": 4.027113237639554e-05,
2521
+ "loss": 0.1079,
2522
+ "step": 2020
2523
+ },
2524
+ {
2525
+ "epoch": 10.18,
2526
+ "learning_rate": 4.037081339712919e-05,
2527
+ "loss": 0.1884,
2528
+ "step": 2025
2529
+ },
2530
+ {
2531
+ "epoch": 10.2,
2532
+ "learning_rate": 4.047049441786284e-05,
2533
+ "loss": 0.0644,
2534
+ "step": 2030
2535
+ },
2536
+ {
2537
+ "epoch": 10.23,
2538
+ "learning_rate": 4.057017543859649e-05,
2539
+ "loss": 0.073,
2540
+ "step": 2035
2541
+ },
2542
+ {
2543
+ "epoch": 10.25,
2544
+ "learning_rate": 4.0669856459330146e-05,
2545
+ "loss": 0.0946,
2546
+ "step": 2040
2547
+ },
2548
+ {
2549
+ "epoch": 10.28,
2550
+ "learning_rate": 4.0769537480063795e-05,
2551
+ "loss": 0.0888,
2552
+ "step": 2045
2553
+ },
2554
+ {
2555
+ "epoch": 10.3,
2556
+ "learning_rate": 4.086921850079745e-05,
2557
+ "loss": 0.0212,
2558
+ "step": 2050
2559
+ },
2560
+ {
2561
+ "epoch": 10.33,
2562
+ "learning_rate": 4.09688995215311e-05,
2563
+ "loss": 0.0737,
2564
+ "step": 2055
2565
+ },
2566
+ {
2567
+ "epoch": 10.35,
2568
+ "learning_rate": 4.1068580542264754e-05,
2569
+ "loss": 0.0286,
2570
+ "step": 2060
2571
+ },
2572
+ {
2573
+ "epoch": 10.38,
2574
+ "learning_rate": 4.116826156299841e-05,
2575
+ "loss": 0.0866,
2576
+ "step": 2065
2577
+ },
2578
+ {
2579
+ "epoch": 10.4,
2580
+ "learning_rate": 4.1267942583732064e-05,
2581
+ "loss": 0.1419,
2582
+ "step": 2070
2583
+ },
2584
+ {
2585
+ "epoch": 10.43,
2586
+ "learning_rate": 4.136762360446571e-05,
2587
+ "loss": 0.1377,
2588
+ "step": 2075
2589
+ },
2590
+ {
2591
+ "epoch": 10.45,
2592
+ "learning_rate": 4.146730462519937e-05,
2593
+ "loss": 0.0808,
2594
+ "step": 2080
2595
+ },
2596
+ {
2597
+ "epoch": 10.48,
2598
+ "learning_rate": 4.156698564593302e-05,
2599
+ "loss": 0.0938,
2600
+ "step": 2085
2601
+ },
2602
+ {
2603
+ "epoch": 10.5,
2604
+ "learning_rate": 4.166666666666667e-05,
2605
+ "loss": 0.1383,
2606
+ "step": 2090
2607
+ },
2608
+ {
2609
+ "epoch": 10.53,
2610
+ "learning_rate": 4.176634768740032e-05,
2611
+ "loss": 0.0873,
2612
+ "step": 2095
2613
+ },
2614
+ {
2615
+ "epoch": 10.55,
2616
+ "learning_rate": 4.1866028708133976e-05,
2617
+ "loss": 0.0451,
2618
+ "step": 2100
2619
+ },
2620
+ {
2621
+ "epoch": 10.58,
2622
+ "learning_rate": 4.1965709728867624e-05,
2623
+ "loss": 0.0879,
2624
+ "step": 2105
2625
+ },
2626
+ {
2627
+ "epoch": 10.6,
2628
+ "learning_rate": 4.206539074960128e-05,
2629
+ "loss": 0.0927,
2630
+ "step": 2110
2631
+ },
2632
+ {
2633
+ "epoch": 10.63,
2634
+ "learning_rate": 4.216507177033493e-05,
2635
+ "loss": 0.0236,
2636
+ "step": 2115
2637
+ },
2638
+ {
2639
+ "epoch": 10.65,
2640
+ "learning_rate": 4.2264752791068584e-05,
2641
+ "loss": 0.0729,
2642
+ "step": 2120
2643
+ },
2644
+ {
2645
+ "epoch": 10.68,
2646
+ "learning_rate": 4.236443381180223e-05,
2647
+ "loss": 0.1569,
2648
+ "step": 2125
2649
+ },
2650
+ {
2651
+ "epoch": 10.7,
2652
+ "learning_rate": 4.246411483253589e-05,
2653
+ "loss": 0.2289,
2654
+ "step": 2130
2655
+ },
2656
+ {
2657
+ "epoch": 10.73,
2658
+ "learning_rate": 4.2563795853269536e-05,
2659
+ "loss": 0.3105,
2660
+ "step": 2135
2661
+ },
2662
+ {
2663
+ "epoch": 10.75,
2664
+ "learning_rate": 4.266347687400319e-05,
2665
+ "loss": 0.1285,
2666
+ "step": 2140
2667
+ },
2668
+ {
2669
+ "epoch": 10.78,
2670
+ "learning_rate": 4.2763157894736847e-05,
2671
+ "loss": 0.1607,
2672
+ "step": 2145
2673
+ },
2674
+ {
2675
+ "epoch": 10.8,
2676
+ "learning_rate": 4.28628389154705e-05,
2677
+ "loss": 0.1358,
2678
+ "step": 2150
2679
+ },
2680
+ {
2681
+ "epoch": 10.83,
2682
+ "learning_rate": 4.296251993620415e-05,
2683
+ "loss": 0.1026,
2684
+ "step": 2155
2685
+ },
2686
+ {
2687
+ "epoch": 10.85,
2688
+ "learning_rate": 4.3062200956937806e-05,
2689
+ "loss": 0.0586,
2690
+ "step": 2160
2691
+ },
2692
+ {
2693
+ "epoch": 10.88,
2694
+ "learning_rate": 4.3161881977671454e-05,
2695
+ "loss": 0.0435,
2696
+ "step": 2165
2697
+ },
2698
+ {
2699
+ "epoch": 10.9,
2700
+ "learning_rate": 4.326156299840511e-05,
2701
+ "loss": 0.2147,
2702
+ "step": 2170
2703
+ },
2704
+ {
2705
+ "epoch": 10.93,
2706
+ "learning_rate": 4.336124401913876e-05,
2707
+ "loss": 0.162,
2708
+ "step": 2175
2709
+ },
2710
+ {
2711
+ "epoch": 10.95,
2712
+ "learning_rate": 4.346092503987241e-05,
2713
+ "loss": 0.2294,
2714
+ "step": 2180
2715
+ },
2716
+ {
2717
+ "epoch": 10.98,
2718
+ "learning_rate": 4.356060606060606e-05,
2719
+ "loss": 0.1536,
2720
+ "step": 2185
2721
+ },
2722
+ {
2723
+ "epoch": 11.0,
2724
+ "eval_accuracy": 0.971253534401508,
2725
+ "eval_loss": 0.08268017321825027,
2726
+ "eval_runtime": 70.5616,
2727
+ "eval_samples_per_second": 30.073,
2728
+ "eval_steps_per_second": 1.885,
2729
+ "step": 2189
2730
+ },
2731
+ {
2732
+ "epoch": 11.01,
2733
+ "learning_rate": 4.366028708133971e-05,
2734
+ "loss": 0.131,
2735
+ "step": 2190
2736
+ },
2737
+ {
2738
+ "epoch": 11.03,
2739
+ "learning_rate": 4.3759968102073366e-05,
2740
+ "loss": 0.0911,
2741
+ "step": 2195
2742
+ },
2743
+ {
2744
+ "epoch": 11.06,
2745
+ "learning_rate": 4.3859649122807014e-05,
2746
+ "loss": 0.0458,
2747
+ "step": 2200
2748
+ },
2749
+ {
2750
+ "epoch": 11.08,
2751
+ "learning_rate": 4.395933014354067e-05,
2752
+ "loss": 0.1948,
2753
+ "step": 2205
2754
+ },
2755
+ {
2756
+ "epoch": 11.11,
2757
+ "learning_rate": 4.4059011164274325e-05,
2758
+ "loss": 0.0882,
2759
+ "step": 2210
2760
+ },
2761
+ {
2762
+ "epoch": 11.13,
2763
+ "learning_rate": 4.415869218500798e-05,
2764
+ "loss": 0.0359,
2765
+ "step": 2215
2766
+ },
2767
+ {
2768
+ "epoch": 11.16,
2769
+ "learning_rate": 4.425837320574163e-05,
2770
+ "loss": 0.0299,
2771
+ "step": 2220
2772
+ },
2773
+ {
2774
+ "epoch": 11.18,
2775
+ "learning_rate": 4.4358054226475284e-05,
2776
+ "loss": 0.1302,
2777
+ "step": 2225
2778
+ },
2779
+ {
2780
+ "epoch": 11.21,
2781
+ "learning_rate": 4.445773524720893e-05,
2782
+ "loss": 0.0896,
2783
+ "step": 2230
2784
+ },
2785
+ {
2786
+ "epoch": 11.23,
2787
+ "learning_rate": 4.455741626794259e-05,
2788
+ "loss": 0.1446,
2789
+ "step": 2235
2790
+ },
2791
+ {
2792
+ "epoch": 11.26,
2793
+ "learning_rate": 4.4657097288676236e-05,
2794
+ "loss": 0.0728,
2795
+ "step": 2240
2796
+ },
2797
+ {
2798
+ "epoch": 11.28,
2799
+ "learning_rate": 4.475677830940989e-05,
2800
+ "loss": 0.0577,
2801
+ "step": 2245
2802
+ },
2803
+ {
2804
+ "epoch": 11.31,
2805
+ "learning_rate": 4.485645933014354e-05,
2806
+ "loss": 0.0469,
2807
+ "step": 2250
2808
+ },
2809
+ {
2810
+ "epoch": 11.33,
2811
+ "learning_rate": 4.4956140350877196e-05,
2812
+ "loss": 0.1188,
2813
+ "step": 2255
2814
+ },
2815
+ {
2816
+ "epoch": 11.36,
2817
+ "learning_rate": 4.5055821371610844e-05,
2818
+ "loss": 0.0319,
2819
+ "step": 2260
2820
+ },
2821
+ {
2822
+ "epoch": 11.38,
2823
+ "learning_rate": 4.51555023923445e-05,
2824
+ "loss": 0.0795,
2825
+ "step": 2265
2826
+ },
2827
+ {
2828
+ "epoch": 11.41,
2829
+ "learning_rate": 4.525518341307815e-05,
2830
+ "loss": 0.0812,
2831
+ "step": 2270
2832
+ },
2833
+ {
2834
+ "epoch": 11.43,
2835
+ "learning_rate": 4.53548644338118e-05,
2836
+ "loss": 0.0215,
2837
+ "step": 2275
2838
+ },
2839
+ {
2840
+ "epoch": 11.46,
2841
+ "learning_rate": 4.545454545454546e-05,
2842
+ "loss": 0.0851,
2843
+ "step": 2280
2844
+ },
2845
+ {
2846
+ "epoch": 11.48,
2847
+ "learning_rate": 4.555422647527911e-05,
2848
+ "loss": 0.0818,
2849
+ "step": 2285
2850
+ },
2851
+ {
2852
+ "epoch": 11.51,
2853
+ "learning_rate": 4.565390749601276e-05,
2854
+ "loss": 0.0753,
2855
+ "step": 2290
2856
+ },
2857
+ {
2858
+ "epoch": 11.53,
2859
+ "learning_rate": 4.575358851674642e-05,
2860
+ "loss": 0.0488,
2861
+ "step": 2295
2862
+ },
2863
+ {
2864
+ "epoch": 11.56,
2865
+ "learning_rate": 4.5853269537480066e-05,
2866
+ "loss": 0.1423,
2867
+ "step": 2300
2868
+ },
2869
+ {
2870
+ "epoch": 11.58,
2871
+ "learning_rate": 4.595295055821372e-05,
2872
+ "loss": 0.0852,
2873
+ "step": 2305
2874
+ },
2875
+ {
2876
+ "epoch": 11.61,
2877
+ "learning_rate": 4.605263157894737e-05,
2878
+ "loss": 0.1462,
2879
+ "step": 2310
2880
+ },
2881
+ {
2882
+ "epoch": 11.63,
2883
+ "learning_rate": 4.6152312599681025e-05,
2884
+ "loss": 0.1058,
2885
+ "step": 2315
2886
+ },
2887
+ {
2888
+ "epoch": 11.66,
2889
+ "learning_rate": 4.6251993620414674e-05,
2890
+ "loss": 0.0474,
2891
+ "step": 2320
2892
+ },
2893
+ {
2894
+ "epoch": 11.68,
2895
+ "learning_rate": 4.635167464114833e-05,
2896
+ "loss": 0.0254,
2897
+ "step": 2325
2898
+ },
2899
+ {
2900
+ "epoch": 11.71,
2901
+ "learning_rate": 4.645135566188198e-05,
2902
+ "loss": 0.2376,
2903
+ "step": 2330
2904
+ },
2905
+ {
2906
+ "epoch": 11.73,
2907
+ "learning_rate": 4.655103668261563e-05,
2908
+ "loss": 0.1412,
2909
+ "step": 2335
2910
+ },
2911
+ {
2912
+ "epoch": 11.76,
2913
+ "learning_rate": 4.665071770334928e-05,
2914
+ "loss": 0.1011,
2915
+ "step": 2340
2916
+ },
2917
+ {
2918
+ "epoch": 11.78,
2919
+ "learning_rate": 4.675039872408294e-05,
2920
+ "loss": 0.1552,
2921
+ "step": 2345
2922
+ },
2923
+ {
2924
+ "epoch": 11.81,
2925
+ "learning_rate": 4.6850079744816585e-05,
2926
+ "loss": 0.0418,
2927
+ "step": 2350
2928
+ },
2929
+ {
2930
+ "epoch": 11.83,
2931
+ "learning_rate": 4.694976076555024e-05,
2932
+ "loss": 0.1005,
2933
+ "step": 2355
2934
+ },
2935
+ {
2936
+ "epoch": 11.86,
2937
+ "learning_rate": 4.7049441786283896e-05,
2938
+ "loss": 0.274,
2939
+ "step": 2360
2940
+ },
2941
+ {
2942
+ "epoch": 11.88,
2943
+ "learning_rate": 4.714912280701755e-05,
2944
+ "loss": 0.137,
2945
+ "step": 2365
2946
+ },
2947
+ {
2948
+ "epoch": 11.91,
2949
+ "learning_rate": 4.72488038277512e-05,
2950
+ "loss": 0.122,
2951
+ "step": 2370
2952
+ },
2953
+ {
2954
+ "epoch": 11.93,
2955
+ "learning_rate": 4.7348484848484855e-05,
2956
+ "loss": 0.0886,
2957
+ "step": 2375
2958
+ },
2959
+ {
2960
+ "epoch": 11.96,
2961
+ "learning_rate": 4.7448165869218504e-05,
2962
+ "loss": 0.1217,
2963
+ "step": 2380
2964
+ },
2965
+ {
2966
+ "epoch": 11.98,
2967
+ "learning_rate": 4.754784688995216e-05,
2968
+ "loss": 0.0458,
2969
+ "step": 2385
2970
+ },
2971
+ {
2972
+ "epoch": 12.0,
2973
+ "eval_accuracy": 0.9882186616399623,
2974
+ "eval_loss": 0.03981461003422737,
2975
+ "eval_runtime": 70.8311,
2976
+ "eval_samples_per_second": 29.959,
2977
+ "eval_steps_per_second": 1.878,
2978
+ "step": 2388
2979
+ },
2980
+ {
2981
+ "epoch": 12.01,
2982
+ "learning_rate": 4.764752791068581e-05,
2983
+ "loss": 0.0762,
2984
+ "step": 2390
2985
+ },
2986
+ {
2987
+ "epoch": 12.04,
2988
+ "learning_rate": 4.774720893141946e-05,
2989
+ "loss": 0.1132,
2990
+ "step": 2395
2991
+ },
2992
+ {
2993
+ "epoch": 12.06,
2994
+ "learning_rate": 4.784688995215311e-05,
2995
+ "loss": 0.1384,
2996
+ "step": 2400
2997
+ },
2998
+ {
2999
+ "epoch": 12.09,
3000
+ "learning_rate": 4.794657097288677e-05,
3001
+ "loss": 0.1139,
3002
+ "step": 2405
3003
+ },
3004
+ {
3005
+ "epoch": 12.11,
3006
+ "learning_rate": 4.8046251993620415e-05,
3007
+ "loss": 0.0659,
3008
+ "step": 2410
3009
+ },
3010
+ {
3011
+ "epoch": 12.14,
3012
+ "learning_rate": 4.8145933014354064e-05,
3013
+ "loss": 0.175,
3014
+ "step": 2415
3015
+ },
3016
+ {
3017
+ "epoch": 12.16,
3018
+ "learning_rate": 4.824561403508772e-05,
3019
+ "loss": 0.14,
3020
+ "step": 2420
3021
+ },
3022
+ {
3023
+ "epoch": 12.19,
3024
+ "learning_rate": 4.8345295055821374e-05,
3025
+ "loss": 0.0635,
3026
+ "step": 2425
3027
+ },
3028
+ {
3029
+ "epoch": 12.21,
3030
+ "learning_rate": 4.844497607655503e-05,
3031
+ "loss": 0.101,
3032
+ "step": 2430
3033
+ },
3034
+ {
3035
+ "epoch": 12.24,
3036
+ "learning_rate": 4.854465709728868e-05,
3037
+ "loss": 0.1797,
3038
+ "step": 2435
3039
+ },
3040
+ {
3041
+ "epoch": 12.26,
3042
+ "learning_rate": 4.8644338118022334e-05,
3043
+ "loss": 0.1087,
3044
+ "step": 2440
3045
+ },
3046
+ {
3047
+ "epoch": 12.29,
3048
+ "learning_rate": 4.874401913875598e-05,
3049
+ "loss": 0.0672,
3050
+ "step": 2445
3051
+ },
3052
+ {
3053
+ "epoch": 12.31,
3054
+ "learning_rate": 4.884370015948964e-05,
3055
+ "loss": 0.0629,
3056
+ "step": 2450
3057
+ },
3058
+ {
3059
+ "epoch": 12.34,
3060
+ "learning_rate": 4.8943381180223286e-05,
3061
+ "loss": 0.0354,
3062
+ "step": 2455
3063
+ },
3064
+ {
3065
+ "epoch": 12.36,
3066
+ "learning_rate": 4.904306220095694e-05,
3067
+ "loss": 0.0454,
3068
+ "step": 2460
3069
+ },
3070
+ {
3071
+ "epoch": 12.39,
3072
+ "learning_rate": 4.914274322169059e-05,
3073
+ "loss": 0.0085,
3074
+ "step": 2465
3075
+ },
3076
+ {
3077
+ "epoch": 12.41,
3078
+ "learning_rate": 4.9242424242424245e-05,
3079
+ "loss": 0.0789,
3080
+ "step": 2470
3081
+ },
3082
+ {
3083
+ "epoch": 12.44,
3084
+ "learning_rate": 4.9342105263157894e-05,
3085
+ "loss": 0.0784,
3086
+ "step": 2475
3087
+ },
3088
+ {
3089
+ "epoch": 12.46,
3090
+ "learning_rate": 4.944178628389155e-05,
3091
+ "loss": 0.1032,
3092
+ "step": 2480
3093
+ },
3094
+ {
3095
+ "epoch": 12.49,
3096
+ "learning_rate": 4.95414673046252e-05,
3097
+ "loss": 0.2055,
3098
+ "step": 2485
3099
+ },
3100
+ {
3101
+ "epoch": 12.51,
3102
+ "learning_rate": 4.964114832535885e-05,
3103
+ "loss": 0.1465,
3104
+ "step": 2490
3105
+ },
3106
+ {
3107
+ "epoch": 12.54,
3108
+ "learning_rate": 4.97408293460925e-05,
3109
+ "loss": 0.1464,
3110
+ "step": 2495
3111
+ },
3112
+ {
3113
+ "epoch": 12.56,
3114
+ "learning_rate": 4.984051036682616e-05,
3115
+ "loss": 0.0842,
3116
+ "step": 2500
3117
+ },
3118
+ {
3119
+ "epoch": 12.59,
3120
+ "learning_rate": 4.994019138755981e-05,
3121
+ "loss": 0.0296,
3122
+ "step": 2505
3123
+ },
3124
+ {
3125
+ "epoch": 12.61,
3126
+ "learning_rate": 4.964028776978418e-05,
3127
+ "loss": 0.0736,
3128
+ "step": 2510
3129
+ },
3130
+ {
3131
+ "epoch": 12.64,
3132
+ "learning_rate": 4.874100719424461e-05,
3133
+ "loss": 0.2944,
3134
+ "step": 2515
3135
+ },
3136
+ {
3137
+ "epoch": 12.66,
3138
+ "learning_rate": 4.784172661870504e-05,
3139
+ "loss": 0.6141,
3140
+ "step": 2520
3141
+ },
3142
+ {
3143
+ "epoch": 12.69,
3144
+ "learning_rate": 4.6942446043165475e-05,
3145
+ "loss": 0.7013,
3146
+ "step": 2525
3147
+ },
3148
+ {
3149
+ "epoch": 12.71,
3150
+ "learning_rate": 4.60431654676259e-05,
3151
+ "loss": 0.5952,
3152
+ "step": 2530
3153
+ },
3154
+ {
3155
+ "epoch": 12.74,
3156
+ "learning_rate": 4.5143884892086334e-05,
3157
+ "loss": 0.6078,
3158
+ "step": 2535
3159
+ },
3160
+ {
3161
+ "epoch": 12.76,
3162
+ "learning_rate": 4.424460431654677e-05,
3163
+ "loss": 0.6268,
3164
+ "step": 2540
3165
+ },
3166
+ {
3167
+ "epoch": 12.79,
3168
+ "learning_rate": 4.33453237410072e-05,
3169
+ "loss": 0.6536,
3170
+ "step": 2545
3171
+ },
3172
+ {
3173
+ "epoch": 12.81,
3174
+ "learning_rate": 4.244604316546763e-05,
3175
+ "loss": 0.5829,
3176
+ "step": 2550
3177
+ },
3178
+ {
3179
+ "epoch": 12.84,
3180
+ "learning_rate": 4.154676258992806e-05,
3181
+ "loss": 0.4209,
3182
+ "step": 2555
3183
+ },
3184
+ {
3185
+ "epoch": 12.86,
3186
+ "learning_rate": 4.064748201438849e-05,
3187
+ "loss": 0.448,
3188
+ "step": 2560
3189
+ },
3190
+ {
3191
+ "epoch": 12.89,
3192
+ "learning_rate": 3.9748201438848925e-05,
3193
+ "loss": 0.5086,
3194
+ "step": 2565
3195
+ },
3196
+ {
3197
+ "epoch": 12.91,
3198
+ "learning_rate": 3.884892086330936e-05,
3199
+ "loss": 0.5025,
3200
+ "step": 2570
3201
+ },
3202
+ {
3203
+ "epoch": 12.94,
3204
+ "learning_rate": 3.794964028776979e-05,
3205
+ "loss": 0.484,
3206
+ "step": 2575
3207
+ },
3208
+ {
3209
+ "epoch": 12.96,
3210
+ "learning_rate": 3.7050359712230217e-05,
3211
+ "loss": 0.3999,
3212
+ "step": 2580
3213
+ },
3214
+ {
3215
+ "epoch": 12.99,
3216
+ "learning_rate": 3.615107913669065e-05,
3217
+ "loss": 0.4956,
3218
+ "step": 2585
3219
+ },
3220
+ {
3221
+ "epoch": 13.0,
3222
+ "eval_accuracy": 0.8642789820923656,
3223
+ "eval_loss": 0.3473888337612152,
3224
+ "eval_runtime": 70.415,
3225
+ "eval_samples_per_second": 30.136,
3226
+ "eval_steps_per_second": 1.889,
3227
+ "step": 2587
3228
+ },
3229
+ {
3230
+ "epoch": 13.02,
3231
+ "learning_rate": 3.5251798561151075e-05,
3232
+ "loss": 0.5233,
3233
+ "step": 2590
3234
+ },
3235
+ {
3236
+ "epoch": 13.04,
3237
+ "learning_rate": 3.435251798561151e-05,
3238
+ "loss": 0.3388,
3239
+ "step": 2595
3240
+ },
3241
+ {
3242
+ "epoch": 13.07,
3243
+ "learning_rate": 3.345323741007194e-05,
3244
+ "loss": 0.4695,
3245
+ "step": 2600
3246
+ },
3247
+ {
3248
+ "epoch": 13.09,
3249
+ "learning_rate": 3.2553956834532374e-05,
3250
+ "loss": 0.4958,
3251
+ "step": 2605
3252
+ },
3253
+ {
3254
+ "epoch": 13.12,
3255
+ "learning_rate": 3.165467625899281e-05,
3256
+ "loss": 0.3863,
3257
+ "step": 2610
3258
+ },
3259
+ {
3260
+ "epoch": 13.14,
3261
+ "learning_rate": 3.075539568345324e-05,
3262
+ "loss": 0.2537,
3263
+ "step": 2615
3264
+ },
3265
+ {
3266
+ "epoch": 13.17,
3267
+ "learning_rate": 2.985611510791367e-05,
3268
+ "loss": 0.3639,
3269
+ "step": 2620
3270
+ },
3271
+ {
3272
+ "epoch": 13.19,
3273
+ "learning_rate": 2.89568345323741e-05,
3274
+ "loss": 0.2409,
3275
+ "step": 2625
3276
+ },
3277
+ {
3278
+ "epoch": 13.22,
3279
+ "learning_rate": 2.805755395683453e-05,
3280
+ "loss": 0.2559,
3281
+ "step": 2630
3282
+ },
3283
+ {
3284
+ "epoch": 13.24,
3285
+ "learning_rate": 2.7158273381294964e-05,
3286
+ "loss": 0.2671,
3287
+ "step": 2635
3288
+ },
3289
+ {
3290
+ "epoch": 13.27,
3291
+ "learning_rate": 2.6258992805755394e-05,
3292
+ "loss": 0.3325,
3293
+ "step": 2640
3294
+ },
3295
+ {
3296
+ "epoch": 13.29,
3297
+ "learning_rate": 2.5359712230215827e-05,
3298
+ "loss": 0.158,
3299
+ "step": 2645
3300
+ },
3301
+ {
3302
+ "epoch": 13.32,
3303
+ "learning_rate": 2.446043165467626e-05,
3304
+ "loss": 0.2713,
3305
+ "step": 2650
3306
+ },
3307
+ {
3308
+ "epoch": 13.34,
3309
+ "learning_rate": 2.3561151079136692e-05,
3310
+ "loss": 0.1543,
3311
+ "step": 2655
3312
+ },
3313
+ {
3314
+ "epoch": 13.37,
3315
+ "learning_rate": 2.2661870503597125e-05,
3316
+ "loss": 0.0994,
3317
+ "step": 2660
3318
+ },
3319
+ {
3320
+ "epoch": 13.39,
3321
+ "learning_rate": 2.1762589928057555e-05,
3322
+ "loss": 0.1104,
3323
+ "step": 2665
3324
+ },
3325
+ {
3326
+ "epoch": 13.42,
3327
+ "learning_rate": 2.0863309352517988e-05,
3328
+ "loss": 0.1673,
3329
+ "step": 2670
3330
+ },
3331
+ {
3332
+ "epoch": 13.44,
3333
+ "learning_rate": 1.996402877697842e-05,
3334
+ "loss": 0.1764,
3335
+ "step": 2675
3336
+ },
3337
+ {
3338
+ "epoch": 13.47,
3339
+ "learning_rate": 1.906474820143885e-05,
3340
+ "loss": 0.2138,
3341
+ "step": 2680
3342
+ },
3343
+ {
3344
+ "epoch": 13.49,
3345
+ "learning_rate": 1.8165467625899283e-05,
3346
+ "loss": 0.1848,
3347
+ "step": 2685
3348
+ },
3349
+ {
3350
+ "epoch": 13.52,
3351
+ "learning_rate": 1.7266187050359716e-05,
3352
+ "loss": 0.152,
3353
+ "step": 2690
3354
+ },
3355
+ {
3356
+ "epoch": 13.54,
3357
+ "learning_rate": 1.6366906474820145e-05,
3358
+ "loss": 0.261,
3359
+ "step": 2695
3360
+ },
3361
+ {
3362
+ "epoch": 13.57,
3363
+ "learning_rate": 1.5467625899280578e-05,
3364
+ "loss": 0.1587,
3365
+ "step": 2700
3366
+ },
3367
+ {
3368
+ "epoch": 13.59,
3369
+ "learning_rate": 1.4568345323741009e-05,
3370
+ "loss": 0.1373,
3371
+ "step": 2705
3372
+ },
3373
+ {
3374
+ "epoch": 13.62,
3375
+ "learning_rate": 1.366906474820144e-05,
3376
+ "loss": 0.1088,
3377
+ "step": 2710
3378
+ },
3379
+ {
3380
+ "epoch": 13.64,
3381
+ "learning_rate": 1.2769784172661871e-05,
3382
+ "loss": 0.0728,
3383
+ "step": 2715
3384
+ },
3385
+ {
3386
+ "epoch": 13.67,
3387
+ "learning_rate": 1.1870503597122303e-05,
3388
+ "loss": 0.0996,
3389
+ "step": 2720
3390
+ },
3391
+ {
3392
+ "epoch": 13.69,
3393
+ "learning_rate": 1.0971223021582734e-05,
3394
+ "loss": 0.1201,
3395
+ "step": 2725
3396
+ },
3397
+ {
3398
+ "epoch": 13.72,
3399
+ "learning_rate": 1.0071942446043167e-05,
3400
+ "loss": 0.1087,
3401
+ "step": 2730
3402
+ },
3403
+ {
3404
+ "epoch": 13.74,
3405
+ "learning_rate": 9.172661870503598e-06,
3406
+ "loss": 0.1906,
3407
+ "step": 2735
3408
+ },
3409
+ {
3410
+ "epoch": 13.77,
3411
+ "learning_rate": 8.273381294964029e-06,
3412
+ "loss": 0.1458,
3413
+ "step": 2740
3414
+ },
3415
+ {
3416
+ "epoch": 13.79,
3417
+ "learning_rate": 7.374100719424461e-06,
3418
+ "loss": 0.2096,
3419
+ "step": 2745
3420
+ },
3421
+ {
3422
+ "epoch": 13.82,
3423
+ "learning_rate": 6.474820143884892e-06,
3424
+ "loss": 0.2497,
3425
+ "step": 2750
3426
+ },
3427
+ {
3428
+ "epoch": 13.84,
3429
+ "learning_rate": 5.575539568345324e-06,
3430
+ "loss": 0.167,
3431
+ "step": 2755
3432
+ },
3433
+ {
3434
+ "epoch": 13.87,
3435
+ "learning_rate": 4.676258992805756e-06,
3436
+ "loss": 0.0807,
3437
+ "step": 2760
3438
+ },
3439
+ {
3440
+ "epoch": 13.89,
3441
+ "learning_rate": 3.7769784172661873e-06,
3442
+ "loss": 0.0646,
3443
+ "step": 2765
3444
+ },
3445
+ {
3446
+ "epoch": 13.92,
3447
+ "learning_rate": 2.877697841726619e-06,
3448
+ "loss": 0.1084,
3449
+ "step": 2770
3450
+ },
3451
+ {
3452
+ "epoch": 13.94,
3453
+ "learning_rate": 1.9784172661870504e-06,
3454
+ "loss": 0.0622,
3455
+ "step": 2775
3456
+ },
3457
+ {
3458
+ "epoch": 13.97,
3459
+ "learning_rate": 1.0791366906474822e-06,
3460
+ "loss": 0.066,
3461
+ "step": 2780
3462
+ },
3463
+ {
3464
+ "epoch": 13.99,
3465
+ "learning_rate": 1.7985611510791368e-07,
3466
+ "loss": 0.0801,
3467
+ "step": 2785
3468
+ },
3469
+ {
3470
+ "epoch": 14.0,
3471
+ "eval_accuracy": 0.9797360980207351,
3472
+ "eval_loss": 0.0849713608622551,
3473
+ "eval_runtime": 70.6282,
3474
+ "eval_samples_per_second": 30.045,
3475
+ "eval_steps_per_second": 1.883,
3476
+ "step": 2786
3477
+ },
3478
+ {
3479
+ "epoch": 14.0,
3480
+ "step": 2786,
3481
+ "total_flos": 1.58314339385421e+19,
3482
+ "train_loss": 0.05687390304887928,
3483
+ "train_runtime": 3560.0954,
3484
+ "train_samples_per_second": 25.026,
3485
+ "train_steps_per_second": 0.783
3486
+ }
3487
+ ],
3488
+ "max_steps": 2786,
3489
+ "num_train_epochs": 14,
3490
+ "total_flos": 1.58314339385421e+19,
3491
+ "trial_name": null,
3492
+ "trial_params": null
3493
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8480db1d2f6d397135d2177cba0fc9a38d71d27f1b0dbe341834621521ae2a11
3
+ size 4155