Hugging Face's logo

crangana
/
trained-age

Image Classification
Transformers
PyTorch
resnet
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
1
crangana commited on
Commit
57a7089
•
1 Parent(s): 64b8034

End of training

Browse files
Files changed (8) hide show
  1. README.md +80 -0
  2. all_results.json +8 -0
  3. config.json +62 -0
  4. preprocessor_config.json +23 -0
  5. pytorch_model.bin +3 -0
  6. train_results.json +8 -0
  7. trainer_state.json +3682 -0
  8. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,80 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: microsoft/resnet-50
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - fair_face
8
+ metrics:
9
+ - accuracy
10
+ model-index:
11
+ - name: trained-age
12
+ results:
13
+ - task:
14
+ name: Image Classification
15
+ type: image-classification
16
+ dataset:
17
+ name: fair_face
18
+ type: fair_face
19
+ config: '0.25'
20
+ split: validation
21
+ args: '0.25'
22
+ metrics:
23
+ - name: Accuracy
24
+ type: accuracy
25
+ value: 0.5164323534781815
26
+ ---
27
+
28
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
+ should probably proofread and complete it, then remove this comment. -->
30
+
31
+ # trained-age
32
+
33
+ This model is a fine-tuned version of [microsoft/resnet-50](https://huggingface.co/microsoft/resnet-50) on the fair_face dataset.
34
+ It achieves the following results on the evaluation set:
35
+ - Loss: 1.1340
36
+ - Accuracy: 0.5164
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 0.0002
56
+ - train_batch_size: 16
57
+ - eval_batch_size: 8
58
+ - seed: 42
59
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
60
+ - lr_scheduler_type: linear
61
+ - num_epochs: 4
62
+
63
+ ### Training results
64
+
65
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
66
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
67
+ | 1.3347 | 0.18 | 1000 | 1.3819 | 0.4296 |
68
+ | 1.3071 | 0.37 | 2000 | 1.2799 | 0.4642 |
69
+ | 1.297 | 0.55 | 3000 | 1.2503 | 0.4721 |
70
+ | 1.3121 | 0.74 | 4000 | 1.1661 | 0.4995 |
71
+ | 1.1806 | 0.92 | 5000 | 1.1137 | 0.5240 |
72
+ | 1.0839 | 1.11 | 6000 | 1.1340 | 0.5164 |
73
+
74
+
75
+ ### Framework versions
76
+
77
+ - Transformers 4.34.0
78
+ - Pytorch 2.0.1+cu118
79
+ - Datasets 2.14.5
80
+ - Tokenizers 0.14.0
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.11,
3
+ "total_flos": 2.0396723395200123e+18,
4
+ "train_loss": 1.29178360859553,
5
+ "train_runtime": 1561.0386,
6
+ "train_samples_per_second": 222.273,
7
+ "train_steps_per_second": 13.893
8
+ }
config.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/resnet-50",
3
+ "architectures": [
4
+ "ResNetForImageClassification"
5
+ ],
6
+ "depths": [
7
+ 3,
8
+ 4,
9
+ 6,
10
+ 3
11
+ ],
12
+ "downsample_in_first_stage": false,
13
+ "embedding_size": 64,
14
+ "hidden_act": "relu",
15
+ "hidden_sizes": [
16
+ 256,
17
+ 512,
18
+ 1024,
19
+ 2048
20
+ ],
21
+ "id2label": {
22
+ "0": "0-2",
23
+ "1": "3-9",
24
+ "2": "10-19",
25
+ "3": "20-29",
26
+ "4": "30-39",
27
+ "5": "40-49",
28
+ "6": "50-59",
29
+ "7": "60-69",
30
+ "8": "more than 70"
31
+ },
32
+ "label2id": {
33
+ "0-2": "0",
34
+ "10-19": "2",
35
+ "20-29": "3",
36
+ "3-9": "1",
37
+ "30-39": "4",
38
+ "40-49": "5",
39
+ "50-59": "6",
40
+ "60-69": "7",
41
+ "more than 70": "8"
42
+ },
43
+ "layer_type": "bottleneck",
44
+ "model_type": "resnet",
45
+ "num_channels": 3,
46
+ "out_features": [
47
+ "stage4"
48
+ ],
49
+ "out_indices": [
50
+ 4
51
+ ],
52
+ "problem_type": "single_label_classification",
53
+ "stage_names": [
54
+ "stem",
55
+ "stage1",
56
+ "stage2",
57
+ "stage3",
58
+ "stage4"
59
+ ],
60
+ "torch_dtype": "float32",
61
+ "transformers_version": "4.34.0"
62
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_pct": 0.875,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "feature_extractor_type": "ConvNextFeatureExtractor",
7
+ "image_mean": [
8
+ 0.485,
9
+ 0.456,
10
+ 0.406
11
+ ],
12
+ "image_processor_type": "ConvNextFeatureExtractor",
13
+ "image_std": [
14
+ 0.229,
15
+ 0.224,
16
+ 0.225
17
+ ],
18
+ "resample": 3,
19
+ "rescale_factor": 0.00392156862745098,
20
+ "size": {
21
+ "shortest_edge": 224
22
+ }
23
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12f330c4f743a48bb10df3a3c9ff27994e1c1dbbcb88450bb85e6104b181f7af
3
+ size 94432333
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.11,
3
+ "total_flos": 2.0396723395200123e+18,
4
+ "train_loss": 1.29178360859553,
5
+ "train_runtime": 1561.0386,
6
+ "train_samples_per_second": 222.273,
7
+ "train_steps_per_second": 13.893
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3682 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.1137409210205078,
3
+ "best_model_checkpoint": "./trained-age/checkpoint-5000",
4
+ "epoch": 1.1066027296200664,
5
+ "eval_steps": 1000,
6
+ "global_step": 6000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 0.000199907783105865,
14
+ "loss": 2.1543,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 0.00019981556621172998,
20
+ "loss": 2.0322,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.01,
25
+ "learning_rate": 0.000199723349317595,
26
+ "loss": 1.9709,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 0.00019963113242345997,
32
+ "loss": 1.9224,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 0.00019953891552932499,
38
+ "loss": 1.9126,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 0.00019944669863518997,
44
+ "loss": 1.9034,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 0.00019935448174105498,
50
+ "loss": 1.8643,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 0.00019926226484691997,
56
+ "loss": 1.7807,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.02,
61
+ "learning_rate": 0.00019917004795278495,
62
+ "loss": 1.7977,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.02,
67
+ "learning_rate": 0.00019907783105864993,
68
+ "loss": 1.9388,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 0.00019898561416451495,
74
+ "loss": 1.9642,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 0.00019889339727037993,
80
+ "loss": 1.9164,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 0.00019880118037624494,
86
+ "loss": 1.8912,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.03,
91
+ "learning_rate": 0.00019870896348210993,
92
+ "loss": 1.7802,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.03,
97
+ "learning_rate": 0.00019861674658797494,
98
+ "loss": 1.7922,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.03,
103
+ "learning_rate": 0.00019852452969383992,
104
+ "loss": 1.8446,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.03,
109
+ "learning_rate": 0.0001984323127997049,
110
+ "loss": 1.8217,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 0.0001983400959055699,
116
+ "loss": 1.8406,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.04,
121
+ "learning_rate": 0.0001982478790114349,
122
+ "loss": 1.8379,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.04,
127
+ "learning_rate": 0.0001981556621172999,
128
+ "loss": 1.7822,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.04,
133
+ "learning_rate": 0.0001980634452231649,
134
+ "loss": 1.8287,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.04,
139
+ "learning_rate": 0.00019797122832902988,
140
+ "loss": 1.8191,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.04,
145
+ "learning_rate": 0.0001978790114348949,
146
+ "loss": 1.7958,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.04,
151
+ "learning_rate": 0.00019778679454075988,
152
+ "loss": 1.7708,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.05,
157
+ "learning_rate": 0.00019769457764662486,
158
+ "loss": 1.7059,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.05,
163
+ "learning_rate": 0.00019760236075248985,
164
+ "loss": 1.7921,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.05,
169
+ "learning_rate": 0.00019751014385835486,
170
+ "loss": 1.8458,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.05,
175
+ "learning_rate": 0.00019741792696421984,
176
+ "loss": 1.7092,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.05,
181
+ "learning_rate": 0.00019732571007008486,
182
+ "loss": 1.7765,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.06,
187
+ "learning_rate": 0.00019723349317594984,
188
+ "loss": 1.7186,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.06,
193
+ "learning_rate": 0.00019714127628181485,
194
+ "loss": 1.7108,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.06,
199
+ "learning_rate": 0.00019704905938767984,
200
+ "loss": 1.5981,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.06,
205
+ "learning_rate": 0.00019695684249354482,
206
+ "loss": 1.7238,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.06,
211
+ "learning_rate": 0.0001968646255994098,
212
+ "loss": 1.677,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.06,
217
+ "learning_rate": 0.00019677240870527482,
218
+ "loss": 1.6465,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.07,
223
+ "learning_rate": 0.0001966801918111398,
224
+ "loss": 1.6533,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.07,
229
+ "learning_rate": 0.0001965879749170048,
230
+ "loss": 1.5514,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.07,
235
+ "learning_rate": 0.0001964957580228698,
236
+ "loss": 1.5717,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.07,
241
+ "learning_rate": 0.0001964035411287348,
242
+ "loss": 1.6208,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.07,
247
+ "learning_rate": 0.0001963113242345998,
248
+ "loss": 1.7294,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.08,
253
+ "learning_rate": 0.00019621910734046478,
254
+ "loss": 1.6014,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.08,
259
+ "learning_rate": 0.00019612689044632976,
260
+ "loss": 1.7207,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.08,
265
+ "learning_rate": 0.00019603467355219477,
266
+ "loss": 1.6452,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.08,
271
+ "learning_rate": 0.00019594245665805976,
272
+ "loss": 1.5979,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.08,
277
+ "learning_rate": 0.00019585023976392477,
278
+ "loss": 1.5809,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.08,
283
+ "learning_rate": 0.00019575802286978975,
284
+ "loss": 1.5626,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.09,
289
+ "learning_rate": 0.00019566580597565476,
290
+ "loss": 1.4701,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.09,
295
+ "learning_rate": 0.00019557358908151975,
296
+ "loss": 1.6025,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.09,
301
+ "learning_rate": 0.00019548137218738473,
302
+ "loss": 1.4946,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.09,
307
+ "learning_rate": 0.00019538915529324972,
308
+ "loss": 1.6431,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.09,
313
+ "learning_rate": 0.00019529693839911473,
314
+ "loss": 1.5584,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.1,
319
+ "learning_rate": 0.00019520472150497971,
320
+ "loss": 1.6565,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.1,
325
+ "learning_rate": 0.00019511250461084473,
326
+ "loss": 1.5616,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.1,
331
+ "learning_rate": 0.0001950202877167097,
332
+ "loss": 1.5366,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.1,
337
+ "learning_rate": 0.00019492807082257472,
338
+ "loss": 1.6421,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.1,
343
+ "learning_rate": 0.0001948358539284397,
344
+ "loss": 1.5919,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.11,
349
+ "learning_rate": 0.0001947436370343047,
350
+ "loss": 1.594,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.11,
355
+ "learning_rate": 0.00019465142014016967,
356
+ "loss": 1.5947,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.11,
361
+ "learning_rate": 0.00019455920324603469,
362
+ "loss": 1.612,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.11,
367
+ "learning_rate": 0.00019446698635189967,
368
+ "loss": 1.4975,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.11,
373
+ "learning_rate": 0.00019437476945776468,
374
+ "loss": 1.5175,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.11,
379
+ "learning_rate": 0.00019428255256362967,
380
+ "loss": 1.3869,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.12,
385
+ "learning_rate": 0.00019419033566949468,
386
+ "loss": 1.494,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.12,
391
+ "learning_rate": 0.00019409811877535966,
392
+ "loss": 1.557,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.12,
397
+ "learning_rate": 0.00019400590188122465,
398
+ "loss": 1.4168,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.12,
403
+ "learning_rate": 0.00019391368498708963,
404
+ "loss": 1.4237,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.12,
409
+ "learning_rate": 0.00019382146809295462,
410
+ "loss": 1.4369,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.13,
415
+ "learning_rate": 0.00019372925119881963,
416
+ "loss": 1.399,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.13,
421
+ "learning_rate": 0.00019363703430468464,
422
+ "loss": 1.4145,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.13,
427
+ "learning_rate": 0.00019354481741054962,
428
+ "loss": 1.5005,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.13,
433
+ "learning_rate": 0.0001934526005164146,
434
+ "loss": 1.5179,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.13,
439
+ "learning_rate": 0.00019336038362227962,
440
+ "loss": 1.4201,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.13,
445
+ "learning_rate": 0.0001932681667281446,
446
+ "loss": 1.5546,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.14,
451
+ "learning_rate": 0.0001931759498340096,
452
+ "loss": 1.5961,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.14,
457
+ "learning_rate": 0.00019308373293987457,
458
+ "loss": 1.4922,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.14,
463
+ "learning_rate": 0.00019299151604573958,
464
+ "loss": 1.4039,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.14,
469
+ "learning_rate": 0.0001928992991516046,
470
+ "loss": 1.4541,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 0.14,
475
+ "learning_rate": 0.00019280708225746958,
476
+ "loss": 1.4385,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 0.15,
481
+ "learning_rate": 0.00019271486536333456,
482
+ "loss": 1.4652,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 0.15,
487
+ "learning_rate": 0.00019262264846919958,
488
+ "loss": 1.4903,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 0.15,
493
+ "learning_rate": 0.00019253043157506456,
494
+ "loss": 1.5028,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 0.15,
499
+ "learning_rate": 0.00019243821468092954,
500
+ "loss": 1.4647,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 0.15,
505
+ "learning_rate": 0.00019234599778679453,
506
+ "loss": 1.4421,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 0.15,
511
+ "learning_rate": 0.00019225378089265954,
512
+ "loss": 1.483,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 0.16,
517
+ "learning_rate": 0.00019216156399852455,
518
+ "loss": 1.5329,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 0.16,
523
+ "learning_rate": 0.00019206934710438954,
524
+ "loss": 1.4918,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 0.16,
529
+ "learning_rate": 0.00019197713021025452,
530
+ "loss": 1.4519,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 0.16,
535
+ "learning_rate": 0.00019188491331611953,
536
+ "loss": 1.2952,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 0.16,
541
+ "learning_rate": 0.00019179269642198452,
542
+ "loss": 1.4764,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 0.17,
547
+ "learning_rate": 0.0001917004795278495,
548
+ "loss": 1.419,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 0.17,
553
+ "learning_rate": 0.00019160826263371449,
554
+ "loss": 1.4126,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 0.17,
559
+ "learning_rate": 0.0001915160457395795,
560
+ "loss": 1.4495,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 0.17,
565
+ "learning_rate": 0.0001914238288454445,
566
+ "loss": 1.4355,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 0.17,
571
+ "learning_rate": 0.0001913316119513095,
572
+ "loss": 1.3425,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 0.18,
577
+ "learning_rate": 0.00019123939505717448,
578
+ "loss": 1.3806,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 0.18,
583
+ "learning_rate": 0.0001911471781630395,
584
+ "loss": 1.3833,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 0.18,
589
+ "learning_rate": 0.00019105496126890447,
590
+ "loss": 1.4533,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 0.18,
595
+ "learning_rate": 0.00019096274437476946,
596
+ "loss": 1.4369,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 0.18,
601
+ "learning_rate": 0.00019087052748063444,
602
+ "loss": 1.4233,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 0.18,
607
+ "learning_rate": 0.00019077831058649945,
608
+ "loss": 1.3347,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 0.18,
613
+ "eval_accuracy": 0.4296147526017893,
614
+ "eval_loss": 1.3819409608840942,
615
+ "eval_runtime": 98.0284,
616
+ "eval_samples_per_second": 111.743,
617
+ "eval_steps_per_second": 13.976,
618
+ "step": 1000
619
+ },
620
+ {
621
+ "epoch": 0.19,
622
+ "learning_rate": 0.00019068609369236446,
623
+ "loss": 1.3295,
624
+ "step": 1010
625
+ },
626
+ {
627
+ "epoch": 0.19,
628
+ "learning_rate": 0.00019059387679822945,
629
+ "loss": 1.4622,
630
+ "step": 1020
631
+ },
632
+ {
633
+ "epoch": 0.19,
634
+ "learning_rate": 0.00019050165990409443,
635
+ "loss": 1.3483,
636
+ "step": 1030
637
+ },
638
+ {
639
+ "epoch": 0.19,
640
+ "learning_rate": 0.00019040944300995944,
641
+ "loss": 1.4762,
642
+ "step": 1040
643
+ },
644
+ {
645
+ "epoch": 0.19,
646
+ "learning_rate": 0.00019031722611582443,
647
+ "loss": 1.462,
648
+ "step": 1050
649
+ },
650
+ {
651
+ "epoch": 0.2,
652
+ "learning_rate": 0.00019022500922168941,
653
+ "loss": 1.5086,
654
+ "step": 1060
655
+ },
656
+ {
657
+ "epoch": 0.2,
658
+ "learning_rate": 0.0001901327923275544,
659
+ "loss": 1.4992,
660
+ "step": 1070
661
+ },
662
+ {
663
+ "epoch": 0.2,
664
+ "learning_rate": 0.0001900405754334194,
665
+ "loss": 1.3563,
666
+ "step": 1080
667
+ },
668
+ {
669
+ "epoch": 0.2,
670
+ "learning_rate": 0.00018994835853928442,
671
+ "loss": 1.3295,
672
+ "step": 1090
673
+ },
674
+ {
675
+ "epoch": 0.2,
676
+ "learning_rate": 0.0001898561416451494,
677
+ "loss": 1.3227,
678
+ "step": 1100
679
+ },
680
+ {
681
+ "epoch": 0.2,
682
+ "learning_rate": 0.0001897639247510144,
683
+ "loss": 1.4465,
684
+ "step": 1110
685
+ },
686
+ {
687
+ "epoch": 0.21,
688
+ "learning_rate": 0.0001896717078568794,
689
+ "loss": 1.4442,
690
+ "step": 1120
691
+ },
692
+ {
693
+ "epoch": 0.21,
694
+ "learning_rate": 0.00018957949096274439,
695
+ "loss": 1.3804,
696
+ "step": 1130
697
+ },
698
+ {
699
+ "epoch": 0.21,
700
+ "learning_rate": 0.00018948727406860937,
701
+ "loss": 1.3868,
702
+ "step": 1140
703
+ },
704
+ {
705
+ "epoch": 0.21,
706
+ "learning_rate": 0.00018939505717447435,
707
+ "loss": 1.4905,
708
+ "step": 1150
709
+ },
710
+ {
711
+ "epoch": 0.21,
712
+ "learning_rate": 0.00018930284028033937,
713
+ "loss": 1.3099,
714
+ "step": 1160
715
+ },
716
+ {
717
+ "epoch": 0.22,
718
+ "learning_rate": 0.00018921062338620438,
719
+ "loss": 1.4865,
720
+ "step": 1170
721
+ },
722
+ {
723
+ "epoch": 0.22,
724
+ "learning_rate": 0.00018911840649206936,
725
+ "loss": 1.4155,
726
+ "step": 1180
727
+ },
728
+ {
729
+ "epoch": 0.22,
730
+ "learning_rate": 0.00018902618959793435,
731
+ "loss": 1.4095,
732
+ "step": 1190
733
+ },
734
+ {
735
+ "epoch": 0.22,
736
+ "learning_rate": 0.00018893397270379936,
737
+ "loss": 1.3217,
738
+ "step": 1200
739
+ },
740
+ {
741
+ "epoch": 0.22,
742
+ "learning_rate": 0.00018884175580966434,
743
+ "loss": 1.5028,
744
+ "step": 1210
745
+ },
746
+ {
747
+ "epoch": 0.23,
748
+ "learning_rate": 0.00018874953891552933,
749
+ "loss": 1.4024,
750
+ "step": 1220
751
+ },
752
+ {
753
+ "epoch": 0.23,
754
+ "learning_rate": 0.0001886573220213943,
755
+ "loss": 1.4699,
756
+ "step": 1230
757
+ },
758
+ {
759
+ "epoch": 0.23,
760
+ "learning_rate": 0.00018856510512725932,
761
+ "loss": 1.4158,
762
+ "step": 1240
763
+ },
764
+ {
765
+ "epoch": 0.23,
766
+ "learning_rate": 0.00018847288823312433,
767
+ "loss": 1.4024,
768
+ "step": 1250
769
+ },
770
+ {
771
+ "epoch": 0.23,
772
+ "learning_rate": 0.00018838067133898932,
773
+ "loss": 1.4017,
774
+ "step": 1260
775
+ },
776
+ {
777
+ "epoch": 0.23,
778
+ "learning_rate": 0.0001882884544448543,
779
+ "loss": 1.3641,
780
+ "step": 1270
781
+ },
782
+ {
783
+ "epoch": 0.24,
784
+ "learning_rate": 0.0001881962375507193,
785
+ "loss": 1.4008,
786
+ "step": 1280
787
+ },
788
+ {
789
+ "epoch": 0.24,
790
+ "learning_rate": 0.0001881040206565843,
791
+ "loss": 1.3428,
792
+ "step": 1290
793
+ },
794
+ {
795
+ "epoch": 0.24,
796
+ "learning_rate": 0.00018801180376244928,
797
+ "loss": 1.3389,
798
+ "step": 1300
799
+ },
800
+ {
801
+ "epoch": 0.24,
802
+ "learning_rate": 0.00018791958686831427,
803
+ "loss": 1.3816,
804
+ "step": 1310
805
+ },
806
+ {
807
+ "epoch": 0.24,
808
+ "learning_rate": 0.00018782736997417928,
809
+ "loss": 1.3573,
810
+ "step": 1320
811
+ },
812
+ {
813
+ "epoch": 0.25,
814
+ "learning_rate": 0.0001877351530800443,
815
+ "loss": 1.3648,
816
+ "step": 1330
817
+ },
818
+ {
819
+ "epoch": 0.25,
820
+ "learning_rate": 0.00018764293618590928,
821
+ "loss": 1.4559,
822
+ "step": 1340
823
+ },
824
+ {
825
+ "epoch": 0.25,
826
+ "learning_rate": 0.00018755071929177426,
827
+ "loss": 1.3925,
828
+ "step": 1350
829
+ },
830
+ {
831
+ "epoch": 0.25,
832
+ "learning_rate": 0.00018745850239763924,
833
+ "loss": 1.3426,
834
+ "step": 1360
835
+ },
836
+ {
837
+ "epoch": 0.25,
838
+ "learning_rate": 0.00018736628550350426,
839
+ "loss": 1.3808,
840
+ "step": 1370
841
+ },
842
+ {
843
+ "epoch": 0.25,
844
+ "learning_rate": 0.00018727406860936924,
845
+ "loss": 1.4028,
846
+ "step": 1380
847
+ },
848
+ {
849
+ "epoch": 0.26,
850
+ "learning_rate": 0.00018718185171523422,
851
+ "loss": 1.4092,
852
+ "step": 1390
853
+ },
854
+ {
855
+ "epoch": 0.26,
856
+ "learning_rate": 0.00018708963482109924,
857
+ "loss": 1.3941,
858
+ "step": 1400
859
+ },
860
+ {
861
+ "epoch": 0.26,
862
+ "learning_rate": 0.00018699741792696425,
863
+ "loss": 1.3246,
864
+ "step": 1410
865
+ },
866
+ {
867
+ "epoch": 0.26,
868
+ "learning_rate": 0.00018690520103282923,
869
+ "loss": 1.3828,
870
+ "step": 1420
871
+ },
872
+ {
873
+ "epoch": 0.26,
874
+ "learning_rate": 0.00018681298413869422,
875
+ "loss": 1.3388,
876
+ "step": 1430
877
+ },
878
+ {
879
+ "epoch": 0.27,
880
+ "learning_rate": 0.0001867207672445592,
881
+ "loss": 1.3951,
882
+ "step": 1440
883
+ },
884
+ {
885
+ "epoch": 0.27,
886
+ "learning_rate": 0.0001866285503504242,
887
+ "loss": 1.3484,
888
+ "step": 1450
889
+ },
890
+ {
891
+ "epoch": 0.27,
892
+ "learning_rate": 0.0001865363334562892,
893
+ "loss": 1.5975,
894
+ "step": 1460
895
+ },
896
+ {
897
+ "epoch": 0.27,
898
+ "learning_rate": 0.00018644411656215418,
899
+ "loss": 1.4879,
900
+ "step": 1470
901
+ },
902
+ {
903
+ "epoch": 0.27,
904
+ "learning_rate": 0.0001863518996680192,
905
+ "loss": 1.3751,
906
+ "step": 1480
907
+ },
908
+ {
909
+ "epoch": 0.27,
910
+ "learning_rate": 0.0001862596827738842,
911
+ "loss": 1.3561,
912
+ "step": 1490
913
+ },
914
+ {
915
+ "epoch": 0.28,
916
+ "learning_rate": 0.0001861674658797492,
917
+ "loss": 1.3283,
918
+ "step": 1500
919
+ },
920
+ {
921
+ "epoch": 0.28,
922
+ "learning_rate": 0.00018607524898561417,
923
+ "loss": 1.4457,
924
+ "step": 1510
925
+ },
926
+ {
927
+ "epoch": 0.28,
928
+ "learning_rate": 0.00018598303209147916,
929
+ "loss": 1.4338,
930
+ "step": 1520
931
+ },
932
+ {
933
+ "epoch": 0.28,
934
+ "learning_rate": 0.00018589081519734417,
935
+ "loss": 1.313,
936
+ "step": 1530
937
+ },
938
+ {
939
+ "epoch": 0.28,
940
+ "learning_rate": 0.00018579859830320915,
941
+ "loss": 1.4286,
942
+ "step": 1540
943
+ },
944
+ {
945
+ "epoch": 0.29,
946
+ "learning_rate": 0.00018570638140907414,
947
+ "loss": 1.3152,
948
+ "step": 1550
949
+ },
950
+ {
951
+ "epoch": 0.29,
952
+ "learning_rate": 0.00018561416451493915,
953
+ "loss": 1.3988,
954
+ "step": 1560
955
+ },
956
+ {
957
+ "epoch": 0.29,
958
+ "learning_rate": 0.00018552194762080416,
959
+ "loss": 1.3082,
960
+ "step": 1570
961
+ },
962
+ {
963
+ "epoch": 0.29,
964
+ "learning_rate": 0.00018542973072666915,
965
+ "loss": 1.2954,
966
+ "step": 1580
967
+ },
968
+ {
969
+ "epoch": 0.29,
970
+ "learning_rate": 0.00018533751383253413,
971
+ "loss": 1.2728,
972
+ "step": 1590
973
+ },
974
+ {
975
+ "epoch": 0.3,
976
+ "learning_rate": 0.00018524529693839911,
977
+ "loss": 1.3098,
978
+ "step": 1600
979
+ },
980
+ {
981
+ "epoch": 0.3,
982
+ "learning_rate": 0.00018515308004426413,
983
+ "loss": 1.3343,
984
+ "step": 1610
985
+ },
986
+ {
987
+ "epoch": 0.3,
988
+ "learning_rate": 0.0001850608631501291,
989
+ "loss": 1.3702,
990
+ "step": 1620
991
+ },
992
+ {
993
+ "epoch": 0.3,
994
+ "learning_rate": 0.0001849686462559941,
995
+ "loss": 1.3413,
996
+ "step": 1630
997
+ },
998
+ {
999
+ "epoch": 0.3,
1000
+ "learning_rate": 0.0001848764293618591,
1001
+ "loss": 1.3541,
1002
+ "step": 1640
1003
+ },
1004
+ {
1005
+ "epoch": 0.3,
1006
+ "learning_rate": 0.00018478421246772412,
1007
+ "loss": 1.3882,
1008
+ "step": 1650
1009
+ },
1010
+ {
1011
+ "epoch": 0.31,
1012
+ "learning_rate": 0.0001846919955735891,
1013
+ "loss": 1.338,
1014
+ "step": 1660
1015
+ },
1016
+ {
1017
+ "epoch": 0.31,
1018
+ "learning_rate": 0.00018459977867945409,
1019
+ "loss": 1.3311,
1020
+ "step": 1670
1021
+ },
1022
+ {
1023
+ "epoch": 0.31,
1024
+ "learning_rate": 0.00018450756178531907,
1025
+ "loss": 1.3407,
1026
+ "step": 1680
1027
+ },
1028
+ {
1029
+ "epoch": 0.31,
1030
+ "learning_rate": 0.00018441534489118408,
1031
+ "loss": 1.4169,
1032
+ "step": 1690
1033
+ },
1034
+ {
1035
+ "epoch": 0.31,
1036
+ "learning_rate": 0.00018432312799704907,
1037
+ "loss": 1.2783,
1038
+ "step": 1700
1039
+ },
1040
+ {
1041
+ "epoch": 0.32,
1042
+ "learning_rate": 0.00018423091110291405,
1043
+ "loss": 1.3291,
1044
+ "step": 1710
1045
+ },
1046
+ {
1047
+ "epoch": 0.32,
1048
+ "learning_rate": 0.00018413869420877906,
1049
+ "loss": 1.4276,
1050
+ "step": 1720
1051
+ },
1052
+ {
1053
+ "epoch": 0.32,
1054
+ "learning_rate": 0.00018404647731464407,
1055
+ "loss": 1.4504,
1056
+ "step": 1730
1057
+ },
1058
+ {
1059
+ "epoch": 0.32,
1060
+ "learning_rate": 0.00018395426042050906,
1061
+ "loss": 1.356,
1062
+ "step": 1740
1063
+ },
1064
+ {
1065
+ "epoch": 0.32,
1066
+ "learning_rate": 0.00018386204352637404,
1067
+ "loss": 1.3567,
1068
+ "step": 1750
1069
+ },
1070
+ {
1071
+ "epoch": 0.32,
1072
+ "learning_rate": 0.00018376982663223903,
1073
+ "loss": 1.2416,
1074
+ "step": 1760
1075
+ },
1076
+ {
1077
+ "epoch": 0.33,
1078
+ "learning_rate": 0.00018367760973810404,
1079
+ "loss": 1.2971,
1080
+ "step": 1770
1081
+ },
1082
+ {
1083
+ "epoch": 0.33,
1084
+ "learning_rate": 0.00018358539284396902,
1085
+ "loss": 1.2429,
1086
+ "step": 1780
1087
+ },
1088
+ {
1089
+ "epoch": 0.33,
1090
+ "learning_rate": 0.000183493175949834,
1091
+ "loss": 1.3004,
1092
+ "step": 1790
1093
+ },
1094
+ {
1095
+ "epoch": 0.33,
1096
+ "learning_rate": 0.00018340095905569902,
1097
+ "loss": 1.3765,
1098
+ "step": 1800
1099
+ },
1100
+ {
1101
+ "epoch": 0.33,
1102
+ "learning_rate": 0.00018330874216156403,
1103
+ "loss": 1.4921,
1104
+ "step": 1810
1105
+ },
1106
+ {
1107
+ "epoch": 0.34,
1108
+ "learning_rate": 0.00018321652526742901,
1109
+ "loss": 1.3344,
1110
+ "step": 1820
1111
+ },
1112
+ {
1113
+ "epoch": 0.34,
1114
+ "learning_rate": 0.000183124308373294,
1115
+ "loss": 1.4283,
1116
+ "step": 1830
1117
+ },
1118
+ {
1119
+ "epoch": 0.34,
1120
+ "learning_rate": 0.00018303209147915898,
1121
+ "loss": 1.2761,
1122
+ "step": 1840
1123
+ },
1124
+ {
1125
+ "epoch": 0.34,
1126
+ "learning_rate": 0.000182939874585024,
1127
+ "loss": 1.3461,
1128
+ "step": 1850
1129
+ },
1130
+ {
1131
+ "epoch": 0.34,
1132
+ "learning_rate": 0.00018284765769088898,
1133
+ "loss": 1.2124,
1134
+ "step": 1860
1135
+ },
1136
+ {
1137
+ "epoch": 0.34,
1138
+ "learning_rate": 0.00018275544079675396,
1139
+ "loss": 1.3249,
1140
+ "step": 1870
1141
+ },
1142
+ {
1143
+ "epoch": 0.35,
1144
+ "learning_rate": 0.00018266322390261898,
1145
+ "loss": 1.3656,
1146
+ "step": 1880
1147
+ },
1148
+ {
1149
+ "epoch": 0.35,
1150
+ "learning_rate": 0.000182571007008484,
1151
+ "loss": 1.3039,
1152
+ "step": 1890
1153
+ },
1154
+ {
1155
+ "epoch": 0.35,
1156
+ "learning_rate": 0.00018247879011434897,
1157
+ "loss": 1.3722,
1158
+ "step": 1900
1159
+ },
1160
+ {
1161
+ "epoch": 0.35,
1162
+ "learning_rate": 0.00018238657322021396,
1163
+ "loss": 1.1888,
1164
+ "step": 1910
1165
+ },
1166
+ {
1167
+ "epoch": 0.35,
1168
+ "learning_rate": 0.00018229435632607894,
1169
+ "loss": 1.3604,
1170
+ "step": 1920
1171
+ },
1172
+ {
1173
+ "epoch": 0.36,
1174
+ "learning_rate": 0.00018220213943194392,
1175
+ "loss": 1.4169,
1176
+ "step": 1930
1177
+ },
1178
+ {
1179
+ "epoch": 0.36,
1180
+ "learning_rate": 0.00018210992253780894,
1181
+ "loss": 1.2332,
1182
+ "step": 1940
1183
+ },
1184
+ {
1185
+ "epoch": 0.36,
1186
+ "learning_rate": 0.00018201770564367392,
1187
+ "loss": 1.2774,
1188
+ "step": 1950
1189
+ },
1190
+ {
1191
+ "epoch": 0.36,
1192
+ "learning_rate": 0.00018192548874953893,
1193
+ "loss": 1.4248,
1194
+ "step": 1960
1195
+ },
1196
+ {
1197
+ "epoch": 0.36,
1198
+ "learning_rate": 0.00018183327185540392,
1199
+ "loss": 1.3928,
1200
+ "step": 1970
1201
+ },
1202
+ {
1203
+ "epoch": 0.37,
1204
+ "learning_rate": 0.00018174105496126893,
1205
+ "loss": 1.2566,
1206
+ "step": 1980
1207
+ },
1208
+ {
1209
+ "epoch": 0.37,
1210
+ "learning_rate": 0.0001816488380671339,
1211
+ "loss": 1.2803,
1212
+ "step": 1990
1213
+ },
1214
+ {
1215
+ "epoch": 0.37,
1216
+ "learning_rate": 0.0001815566211729989,
1217
+ "loss": 1.3071,
1218
+ "step": 2000
1219
+ },
1220
+ {
1221
+ "epoch": 0.37,
1222
+ "eval_accuracy": 0.464213985758627,
1223
+ "eval_loss": 1.279906988143921,
1224
+ "eval_runtime": 97.8688,
1225
+ "eval_samples_per_second": 111.925,
1226
+ "eval_steps_per_second": 13.998,
1227
+ "step": 2000
1228
+ },
1229
+ {
1230
+ "epoch": 0.37,
1231
+ "learning_rate": 0.00018146440427886388,
1232
+ "loss": 1.3159,
1233
+ "step": 2010
1234
+ },
1235
+ {
1236
+ "epoch": 0.37,
1237
+ "learning_rate": 0.0001813721873847289,
1238
+ "loss": 1.4002,
1239
+ "step": 2020
1240
+ },
1241
+ {
1242
+ "epoch": 0.37,
1243
+ "learning_rate": 0.00018127997049059388,
1244
+ "loss": 1.2217,
1245
+ "step": 2030
1246
+ },
1247
+ {
1248
+ "epoch": 0.38,
1249
+ "learning_rate": 0.0001811877535964589,
1250
+ "loss": 1.2869,
1251
+ "step": 2040
1252
+ },
1253
+ {
1254
+ "epoch": 0.38,
1255
+ "learning_rate": 0.00018109553670232387,
1256
+ "loss": 1.2793,
1257
+ "step": 2050
1258
+ },
1259
+ {
1260
+ "epoch": 0.38,
1261
+ "learning_rate": 0.00018100331980818888,
1262
+ "loss": 1.3284,
1263
+ "step": 2060
1264
+ },
1265
+ {
1266
+ "epoch": 0.38,
1267
+ "learning_rate": 0.00018091110291405387,
1268
+ "loss": 1.1991,
1269
+ "step": 2070
1270
+ },
1271
+ {
1272
+ "epoch": 0.38,
1273
+ "learning_rate": 0.00018081888601991885,
1274
+ "loss": 1.1353,
1275
+ "step": 2080
1276
+ },
1277
+ {
1278
+ "epoch": 0.39,
1279
+ "learning_rate": 0.00018072666912578384,
1280
+ "loss": 1.3133,
1281
+ "step": 2090
1282
+ },
1283
+ {
1284
+ "epoch": 0.39,
1285
+ "learning_rate": 0.00018063445223164885,
1286
+ "loss": 1.3075,
1287
+ "step": 2100
1288
+ },
1289
+ {
1290
+ "epoch": 0.39,
1291
+ "learning_rate": 0.00018054223533751383,
1292
+ "loss": 1.4026,
1293
+ "step": 2110
1294
+ },
1295
+ {
1296
+ "epoch": 0.39,
1297
+ "learning_rate": 0.00018045001844337885,
1298
+ "loss": 1.3367,
1299
+ "step": 2120
1300
+ },
1301
+ {
1302
+ "epoch": 0.39,
1303
+ "learning_rate": 0.00018035780154924383,
1304
+ "loss": 1.3156,
1305
+ "step": 2130
1306
+ },
1307
+ {
1308
+ "epoch": 0.39,
1309
+ "learning_rate": 0.00018026558465510884,
1310
+ "loss": 1.2022,
1311
+ "step": 2140
1312
+ },
1313
+ {
1314
+ "epoch": 0.4,
1315
+ "learning_rate": 0.00018017336776097383,
1316
+ "loss": 1.3409,
1317
+ "step": 2150
1318
+ },
1319
+ {
1320
+ "epoch": 0.4,
1321
+ "learning_rate": 0.0001800811508668388,
1322
+ "loss": 1.2041,
1323
+ "step": 2160
1324
+ },
1325
+ {
1326
+ "epoch": 0.4,
1327
+ "learning_rate": 0.0001799889339727038,
1328
+ "loss": 1.3062,
1329
+ "step": 2170
1330
+ },
1331
+ {
1332
+ "epoch": 0.4,
1333
+ "learning_rate": 0.0001798967170785688,
1334
+ "loss": 1.2372,
1335
+ "step": 2180
1336
+ },
1337
+ {
1338
+ "epoch": 0.4,
1339
+ "learning_rate": 0.0001798045001844338,
1340
+ "loss": 1.2237,
1341
+ "step": 2190
1342
+ },
1343
+ {
1344
+ "epoch": 0.41,
1345
+ "learning_rate": 0.0001797122832902988,
1346
+ "loss": 1.2524,
1347
+ "step": 2200
1348
+ },
1349
+ {
1350
+ "epoch": 0.41,
1351
+ "learning_rate": 0.00017962006639616379,
1352
+ "loss": 1.4144,
1353
+ "step": 2210
1354
+ },
1355
+ {
1356
+ "epoch": 0.41,
1357
+ "learning_rate": 0.0001795278495020288,
1358
+ "loss": 1.1676,
1359
+ "step": 2220
1360
+ },
1361
+ {
1362
+ "epoch": 0.41,
1363
+ "learning_rate": 0.00017943563260789378,
1364
+ "loss": 1.2909,
1365
+ "step": 2230
1366
+ },
1367
+ {
1368
+ "epoch": 0.41,
1369
+ "learning_rate": 0.00017934341571375877,
1370
+ "loss": 1.372,
1371
+ "step": 2240
1372
+ },
1373
+ {
1374
+ "epoch": 0.41,
1375
+ "learning_rate": 0.00017925119881962375,
1376
+ "loss": 1.2902,
1377
+ "step": 2250
1378
+ },
1379
+ {
1380
+ "epoch": 0.42,
1381
+ "learning_rate": 0.00017915898192548876,
1382
+ "loss": 1.2131,
1383
+ "step": 2260
1384
+ },
1385
+ {
1386
+ "epoch": 0.42,
1387
+ "learning_rate": 0.00017906676503135375,
1388
+ "loss": 1.198,
1389
+ "step": 2270
1390
+ },
1391
+ {
1392
+ "epoch": 0.42,
1393
+ "learning_rate": 0.00017897454813721876,
1394
+ "loss": 1.2151,
1395
+ "step": 2280
1396
+ },
1397
+ {
1398
+ "epoch": 0.42,
1399
+ "learning_rate": 0.00017888233124308374,
1400
+ "loss": 1.3141,
1401
+ "step": 2290
1402
+ },
1403
+ {
1404
+ "epoch": 0.42,
1405
+ "learning_rate": 0.00017879011434894875,
1406
+ "loss": 1.3788,
1407
+ "step": 2300
1408
+ },
1409
+ {
1410
+ "epoch": 0.43,
1411
+ "learning_rate": 0.00017869789745481374,
1412
+ "loss": 1.2766,
1413
+ "step": 2310
1414
+ },
1415
+ {
1416
+ "epoch": 0.43,
1417
+ "learning_rate": 0.00017860568056067872,
1418
+ "loss": 1.2166,
1419
+ "step": 2320
1420
+ },
1421
+ {
1422
+ "epoch": 0.43,
1423
+ "learning_rate": 0.0001785134636665437,
1424
+ "loss": 1.3074,
1425
+ "step": 2330
1426
+ },
1427
+ {
1428
+ "epoch": 0.43,
1429
+ "learning_rate": 0.00017842124677240872,
1430
+ "loss": 1.2592,
1431
+ "step": 2340
1432
+ },
1433
+ {
1434
+ "epoch": 0.43,
1435
+ "learning_rate": 0.0001783290298782737,
1436
+ "loss": 1.2932,
1437
+ "step": 2350
1438
+ },
1439
+ {
1440
+ "epoch": 0.44,
1441
+ "learning_rate": 0.00017823681298413871,
1442
+ "loss": 1.2729,
1443
+ "step": 2360
1444
+ },
1445
+ {
1446
+ "epoch": 0.44,
1447
+ "learning_rate": 0.0001781445960900037,
1448
+ "loss": 1.256,
1449
+ "step": 2370
1450
+ },
1451
+ {
1452
+ "epoch": 0.44,
1453
+ "learning_rate": 0.0001780523791958687,
1454
+ "loss": 1.284,
1455
+ "step": 2380
1456
+ },
1457
+ {
1458
+ "epoch": 0.44,
1459
+ "learning_rate": 0.0001779601623017337,
1460
+ "loss": 1.2473,
1461
+ "step": 2390
1462
+ },
1463
+ {
1464
+ "epoch": 0.44,
1465
+ "learning_rate": 0.00017786794540759868,
1466
+ "loss": 1.3336,
1467
+ "step": 2400
1468
+ },
1469
+ {
1470
+ "epoch": 0.44,
1471
+ "learning_rate": 0.00017777572851346366,
1472
+ "loss": 1.3713,
1473
+ "step": 2410
1474
+ },
1475
+ {
1476
+ "epoch": 0.45,
1477
+ "learning_rate": 0.00017768351161932868,
1478
+ "loss": 1.3888,
1479
+ "step": 2420
1480
+ },
1481
+ {
1482
+ "epoch": 0.45,
1483
+ "learning_rate": 0.00017759129472519366,
1484
+ "loss": 1.2189,
1485
+ "step": 2430
1486
+ },
1487
+ {
1488
+ "epoch": 0.45,
1489
+ "learning_rate": 0.00017749907783105864,
1490
+ "loss": 1.3783,
1491
+ "step": 2440
1492
+ },
1493
+ {
1494
+ "epoch": 0.45,
1495
+ "learning_rate": 0.00017740686093692366,
1496
+ "loss": 1.2576,
1497
+ "step": 2450
1498
+ },
1499
+ {
1500
+ "epoch": 0.45,
1501
+ "learning_rate": 0.00017731464404278867,
1502
+ "loss": 1.2524,
1503
+ "step": 2460
1504
+ },
1505
+ {
1506
+ "epoch": 0.46,
1507
+ "learning_rate": 0.00017722242714865365,
1508
+ "loss": 1.1614,
1509
+ "step": 2470
1510
+ },
1511
+ {
1512
+ "epoch": 0.46,
1513
+ "learning_rate": 0.00017713021025451864,
1514
+ "loss": 1.2631,
1515
+ "step": 2480
1516
+ },
1517
+ {
1518
+ "epoch": 0.46,
1519
+ "learning_rate": 0.00017703799336038362,
1520
+ "loss": 1.2215,
1521
+ "step": 2490
1522
+ },
1523
+ {
1524
+ "epoch": 0.46,
1525
+ "learning_rate": 0.00017694577646624863,
1526
+ "loss": 1.266,
1527
+ "step": 2500
1528
+ },
1529
+ {
1530
+ "epoch": 0.46,
1531
+ "learning_rate": 0.00017685355957211362,
1532
+ "loss": 1.1883,
1533
+ "step": 2510
1534
+ },
1535
+ {
1536
+ "epoch": 0.46,
1537
+ "learning_rate": 0.0001767613426779786,
1538
+ "loss": 1.2202,
1539
+ "step": 2520
1540
+ },
1541
+ {
1542
+ "epoch": 0.47,
1543
+ "learning_rate": 0.0001766691257838436,
1544
+ "loss": 1.238,
1545
+ "step": 2530
1546
+ },
1547
+ {
1548
+ "epoch": 0.47,
1549
+ "learning_rate": 0.0001765769088897086,
1550
+ "loss": 1.2836,
1551
+ "step": 2540
1552
+ },
1553
+ {
1554
+ "epoch": 0.47,
1555
+ "learning_rate": 0.0001764846919955736,
1556
+ "loss": 1.2879,
1557
+ "step": 2550
1558
+ },
1559
+ {
1560
+ "epoch": 0.47,
1561
+ "learning_rate": 0.0001763924751014386,
1562
+ "loss": 1.17,
1563
+ "step": 2560
1564
+ },
1565
+ {
1566
+ "epoch": 0.47,
1567
+ "learning_rate": 0.00017630025820730358,
1568
+ "loss": 1.1675,
1569
+ "step": 2570
1570
+ },
1571
+ {
1572
+ "epoch": 0.48,
1573
+ "learning_rate": 0.00017620804131316856,
1574
+ "loss": 1.1862,
1575
+ "step": 2580
1576
+ },
1577
+ {
1578
+ "epoch": 0.48,
1579
+ "learning_rate": 0.00017611582441903357,
1580
+ "loss": 1.2787,
1581
+ "step": 2590
1582
+ },
1583
+ {
1584
+ "epoch": 0.48,
1585
+ "learning_rate": 0.00017602360752489856,
1586
+ "loss": 1.1561,
1587
+ "step": 2600
1588
+ },
1589
+ {
1590
+ "epoch": 0.48,
1591
+ "learning_rate": 0.00017593139063076357,
1592
+ "loss": 1.3317,
1593
+ "step": 2610
1594
+ },
1595
+ {
1596
+ "epoch": 0.48,
1597
+ "learning_rate": 0.00017583917373662855,
1598
+ "loss": 1.2562,
1599
+ "step": 2620
1600
+ },
1601
+ {
1602
+ "epoch": 0.49,
1603
+ "learning_rate": 0.00017574695684249356,
1604
+ "loss": 1.3731,
1605
+ "step": 2630
1606
+ },
1607
+ {
1608
+ "epoch": 0.49,
1609
+ "learning_rate": 0.00017565473994835855,
1610
+ "loss": 1.304,
1611
+ "step": 2640
1612
+ },
1613
+ {
1614
+ "epoch": 0.49,
1615
+ "learning_rate": 0.00017556252305422353,
1616
+ "loss": 1.2217,
1617
+ "step": 2650
1618
+ },
1619
+ {
1620
+ "epoch": 0.49,
1621
+ "learning_rate": 0.00017547030616008852,
1622
+ "loss": 1.2968,
1623
+ "step": 2660
1624
+ },
1625
+ {
1626
+ "epoch": 0.49,
1627
+ "learning_rate": 0.00017537808926595353,
1628
+ "loss": 1.2932,
1629
+ "step": 2670
1630
+ },
1631
+ {
1632
+ "epoch": 0.49,
1633
+ "learning_rate": 0.00017528587237181851,
1634
+ "loss": 1.3466,
1635
+ "step": 2680
1636
+ },
1637
+ {
1638
+ "epoch": 0.5,
1639
+ "learning_rate": 0.00017519365547768353,
1640
+ "loss": 1.272,
1641
+ "step": 2690
1642
+ },
1643
+ {
1644
+ "epoch": 0.5,
1645
+ "learning_rate": 0.0001751014385835485,
1646
+ "loss": 1.1447,
1647
+ "step": 2700
1648
+ },
1649
+ {
1650
+ "epoch": 0.5,
1651
+ "learning_rate": 0.00017500922168941352,
1652
+ "loss": 1.2865,
1653
+ "step": 2710
1654
+ },
1655
+ {
1656
+ "epoch": 0.5,
1657
+ "learning_rate": 0.0001749170047952785,
1658
+ "loss": 1.287,
1659
+ "step": 2720
1660
+ },
1661
+ {
1662
+ "epoch": 0.5,
1663
+ "learning_rate": 0.0001748247879011435,
1664
+ "loss": 1.2388,
1665
+ "step": 2730
1666
+ },
1667
+ {
1668
+ "epoch": 0.51,
1669
+ "learning_rate": 0.00017473257100700847,
1670
+ "loss": 1.2865,
1671
+ "step": 2740
1672
+ },
1673
+ {
1674
+ "epoch": 0.51,
1675
+ "learning_rate": 0.00017464035411287349,
1676
+ "loss": 1.2697,
1677
+ "step": 2750
1678
+ },
1679
+ {
1680
+ "epoch": 0.51,
1681
+ "learning_rate": 0.00017454813721873847,
1682
+ "loss": 1.2994,
1683
+ "step": 2760
1684
+ },
1685
+ {
1686
+ "epoch": 0.51,
1687
+ "learning_rate": 0.00017445592032460348,
1688
+ "loss": 1.3243,
1689
+ "step": 2770
1690
+ },
1691
+ {
1692
+ "epoch": 0.51,
1693
+ "learning_rate": 0.00017436370343046847,
1694
+ "loss": 1.1788,
1695
+ "step": 2780
1696
+ },
1697
+ {
1698
+ "epoch": 0.51,
1699
+ "learning_rate": 0.00017427148653633348,
1700
+ "loss": 1.3052,
1701
+ "step": 2790
1702
+ },
1703
+ {
1704
+ "epoch": 0.52,
1705
+ "learning_rate": 0.00017417926964219846,
1706
+ "loss": 1.2467,
1707
+ "step": 2800
1708
+ },
1709
+ {
1710
+ "epoch": 0.52,
1711
+ "learning_rate": 0.00017408705274806345,
1712
+ "loss": 1.2646,
1713
+ "step": 2810
1714
+ },
1715
+ {
1716
+ "epoch": 0.52,
1717
+ "learning_rate": 0.00017399483585392843,
1718
+ "loss": 1.3098,
1719
+ "step": 2820
1720
+ },
1721
+ {
1722
+ "epoch": 0.52,
1723
+ "learning_rate": 0.00017390261895979344,
1724
+ "loss": 1.3249,
1725
+ "step": 2830
1726
+ },
1727
+ {
1728
+ "epoch": 0.52,
1729
+ "learning_rate": 0.00017381040206565843,
1730
+ "loss": 1.3779,
1731
+ "step": 2840
1732
+ },
1733
+ {
1734
+ "epoch": 0.53,
1735
+ "learning_rate": 0.00017371818517152344,
1736
+ "loss": 1.3639,
1737
+ "step": 2850
1738
+ },
1739
+ {
1740
+ "epoch": 0.53,
1741
+ "learning_rate": 0.00017362596827738842,
1742
+ "loss": 1.2494,
1743
+ "step": 2860
1744
+ },
1745
+ {
1746
+ "epoch": 0.53,
1747
+ "learning_rate": 0.00017353375138325343,
1748
+ "loss": 1.2869,
1749
+ "step": 2870
1750
+ },
1751
+ {
1752
+ "epoch": 0.53,
1753
+ "learning_rate": 0.00017344153448911842,
1754
+ "loss": 1.2766,
1755
+ "step": 2880
1756
+ },
1757
+ {
1758
+ "epoch": 0.53,
1759
+ "learning_rate": 0.0001733493175949834,
1760
+ "loss": 1.1126,
1761
+ "step": 2890
1762
+ },
1763
+ {
1764
+ "epoch": 0.53,
1765
+ "learning_rate": 0.0001732571007008484,
1766
+ "loss": 1.1705,
1767
+ "step": 2900
1768
+ },
1769
+ {
1770
+ "epoch": 0.54,
1771
+ "learning_rate": 0.0001731648838067134,
1772
+ "loss": 1.2194,
1773
+ "step": 2910
1774
+ },
1775
+ {
1776
+ "epoch": 0.54,
1777
+ "learning_rate": 0.00017307266691257838,
1778
+ "loss": 1.2963,
1779
+ "step": 2920
1780
+ },
1781
+ {
1782
+ "epoch": 0.54,
1783
+ "learning_rate": 0.0001729804500184434,
1784
+ "loss": 1.2211,
1785
+ "step": 2930
1786
+ },
1787
+ {
1788
+ "epoch": 0.54,
1789
+ "learning_rate": 0.00017288823312430838,
1790
+ "loss": 1.1875,
1791
+ "step": 2940
1792
+ },
1793
+ {
1794
+ "epoch": 0.54,
1795
+ "learning_rate": 0.0001727960162301734,
1796
+ "loss": 1.3088,
1797
+ "step": 2950
1798
+ },
1799
+ {
1800
+ "epoch": 0.55,
1801
+ "learning_rate": 0.00017270379933603838,
1802
+ "loss": 1.3143,
1803
+ "step": 2960
1804
+ },
1805
+ {
1806
+ "epoch": 0.55,
1807
+ "learning_rate": 0.00017261158244190336,
1808
+ "loss": 1.2684,
1809
+ "step": 2970
1810
+ },
1811
+ {
1812
+ "epoch": 0.55,
1813
+ "learning_rate": 0.00017251936554776834,
1814
+ "loss": 1.1308,
1815
+ "step": 2980
1816
+ },
1817
+ {
1818
+ "epoch": 0.55,
1819
+ "learning_rate": 0.00017242714865363336,
1820
+ "loss": 1.2508,
1821
+ "step": 2990
1822
+ },
1823
+ {
1824
+ "epoch": 0.55,
1825
+ "learning_rate": 0.00017233493175949834,
1826
+ "loss": 1.297,
1827
+ "step": 3000
1828
+ },
1829
+ {
1830
+ "epoch": 0.55,
1831
+ "eval_accuracy": 0.4720649990870915,
1832
+ "eval_loss": 1.250298261642456,
1833
+ "eval_runtime": 98.4328,
1834
+ "eval_samples_per_second": 111.284,
1835
+ "eval_steps_per_second": 13.918,
1836
+ "step": 3000
1837
+ },
1838
+ {
1839
+ "epoch": 0.56,
1840
+ "learning_rate": 0.00017224271486536335,
1841
+ "loss": 1.2519,
1842
+ "step": 3010
1843
+ },
1844
+ {
1845
+ "epoch": 0.56,
1846
+ "learning_rate": 0.00017215049797122834,
1847
+ "loss": 1.1632,
1848
+ "step": 3020
1849
+ },
1850
+ {
1851
+ "epoch": 0.56,
1852
+ "learning_rate": 0.00017205828107709335,
1853
+ "loss": 1.2063,
1854
+ "step": 3030
1855
+ },
1856
+ {
1857
+ "epoch": 0.56,
1858
+ "learning_rate": 0.00017196606418295833,
1859
+ "loss": 1.2205,
1860
+ "step": 3040
1861
+ },
1862
+ {
1863
+ "epoch": 0.56,
1864
+ "learning_rate": 0.00017187384728882332,
1865
+ "loss": 1.2787,
1866
+ "step": 3050
1867
+ },
1868
+ {
1869
+ "epoch": 0.56,
1870
+ "learning_rate": 0.0001717816303946883,
1871
+ "loss": 1.1501,
1872
+ "step": 3060
1873
+ },
1874
+ {
1875
+ "epoch": 0.57,
1876
+ "learning_rate": 0.0001716894135005533,
1877
+ "loss": 1.4336,
1878
+ "step": 3070
1879
+ },
1880
+ {
1881
+ "epoch": 0.57,
1882
+ "learning_rate": 0.0001715971966064183,
1883
+ "loss": 1.228,
1884
+ "step": 3080
1885
+ },
1886
+ {
1887
+ "epoch": 0.57,
1888
+ "learning_rate": 0.0001715049797122833,
1889
+ "loss": 1.2311,
1890
+ "step": 3090
1891
+ },
1892
+ {
1893
+ "epoch": 0.57,
1894
+ "learning_rate": 0.0001714127628181483,
1895
+ "loss": 1.2151,
1896
+ "step": 3100
1897
+ },
1898
+ {
1899
+ "epoch": 0.57,
1900
+ "learning_rate": 0.0001713205459240133,
1901
+ "loss": 1.324,
1902
+ "step": 3110
1903
+ },
1904
+ {
1905
+ "epoch": 0.58,
1906
+ "learning_rate": 0.0001712283290298783,
1907
+ "loss": 1.2236,
1908
+ "step": 3120
1909
+ },
1910
+ {
1911
+ "epoch": 0.58,
1912
+ "learning_rate": 0.00017113611213574327,
1913
+ "loss": 1.1156,
1914
+ "step": 3130
1915
+ },
1916
+ {
1917
+ "epoch": 0.58,
1918
+ "learning_rate": 0.00017104389524160826,
1919
+ "loss": 1.1789,
1920
+ "step": 3140
1921
+ },
1922
+ {
1923
+ "epoch": 0.58,
1924
+ "learning_rate": 0.00017095167834747327,
1925
+ "loss": 1.1482,
1926
+ "step": 3150
1927
+ },
1928
+ {
1929
+ "epoch": 0.58,
1930
+ "learning_rate": 0.00017085946145333825,
1931
+ "loss": 1.2907,
1932
+ "step": 3160
1933
+ },
1934
+ {
1935
+ "epoch": 0.58,
1936
+ "learning_rate": 0.00017076724455920326,
1937
+ "loss": 1.255,
1938
+ "step": 3170
1939
+ },
1940
+ {
1941
+ "epoch": 0.59,
1942
+ "learning_rate": 0.00017067502766506825,
1943
+ "loss": 1.1669,
1944
+ "step": 3180
1945
+ },
1946
+ {
1947
+ "epoch": 0.59,
1948
+ "learning_rate": 0.00017058281077093323,
1949
+ "loss": 1.2531,
1950
+ "step": 3190
1951
+ },
1952
+ {
1953
+ "epoch": 0.59,
1954
+ "learning_rate": 0.00017049059387679825,
1955
+ "loss": 1.3122,
1956
+ "step": 3200
1957
+ },
1958
+ {
1959
+ "epoch": 0.59,
1960
+ "learning_rate": 0.00017039837698266323,
1961
+ "loss": 1.2077,
1962
+ "step": 3210
1963
+ },
1964
+ {
1965
+ "epoch": 0.59,
1966
+ "learning_rate": 0.00017030616008852821,
1967
+ "loss": 1.2112,
1968
+ "step": 3220
1969
+ },
1970
+ {
1971
+ "epoch": 0.6,
1972
+ "learning_rate": 0.0001702139431943932,
1973
+ "loss": 1.2232,
1974
+ "step": 3230
1975
+ },
1976
+ {
1977
+ "epoch": 0.6,
1978
+ "learning_rate": 0.0001701217263002582,
1979
+ "loss": 1.2612,
1980
+ "step": 3240
1981
+ },
1982
+ {
1983
+ "epoch": 0.6,
1984
+ "learning_rate": 0.00017002950940612322,
1985
+ "loss": 1.2819,
1986
+ "step": 3250
1987
+ },
1988
+ {
1989
+ "epoch": 0.6,
1990
+ "learning_rate": 0.0001699372925119882,
1991
+ "loss": 1.3206,
1992
+ "step": 3260
1993
+ },
1994
+ {
1995
+ "epoch": 0.6,
1996
+ "learning_rate": 0.0001698450756178532,
1997
+ "loss": 1.1989,
1998
+ "step": 3270
1999
+ },
2000
+ {
2001
+ "epoch": 0.6,
2002
+ "learning_rate": 0.0001697528587237182,
2003
+ "loss": 1.204,
2004
+ "step": 3280
2005
+ },
2006
+ {
2007
+ "epoch": 0.61,
2008
+ "learning_rate": 0.00016966064182958319,
2009
+ "loss": 1.1277,
2010
+ "step": 3290
2011
+ },
2012
+ {
2013
+ "epoch": 0.61,
2014
+ "learning_rate": 0.00016956842493544817,
2015
+ "loss": 1.1263,
2016
+ "step": 3300
2017
+ },
2018
+ {
2019
+ "epoch": 0.61,
2020
+ "learning_rate": 0.00016947620804131316,
2021
+ "loss": 1.1812,
2022
+ "step": 3310
2023
+ },
2024
+ {
2025
+ "epoch": 0.61,
2026
+ "learning_rate": 0.00016938399114717817,
2027
+ "loss": 1.2092,
2028
+ "step": 3320
2029
+ },
2030
+ {
2031
+ "epoch": 0.61,
2032
+ "learning_rate": 0.00016929177425304318,
2033
+ "loss": 1.1473,
2034
+ "step": 3330
2035
+ },
2036
+ {
2037
+ "epoch": 0.62,
2038
+ "learning_rate": 0.00016919955735890816,
2039
+ "loss": 1.3631,
2040
+ "step": 3340
2041
+ },
2042
+ {
2043
+ "epoch": 0.62,
2044
+ "learning_rate": 0.00016910734046477315,
2045
+ "loss": 1.2097,
2046
+ "step": 3350
2047
+ },
2048
+ {
2049
+ "epoch": 0.62,
2050
+ "learning_rate": 0.00016901512357063816,
2051
+ "loss": 1.1183,
2052
+ "step": 3360
2053
+ },
2054
+ {
2055
+ "epoch": 0.62,
2056
+ "learning_rate": 0.00016892290667650314,
2057
+ "loss": 1.1925,
2058
+ "step": 3370
2059
+ },
2060
+ {
2061
+ "epoch": 0.62,
2062
+ "learning_rate": 0.00016883068978236813,
2063
+ "loss": 1.2737,
2064
+ "step": 3380
2065
+ },
2066
+ {
2067
+ "epoch": 0.63,
2068
+ "learning_rate": 0.0001687384728882331,
2069
+ "loss": 1.0825,
2070
+ "step": 3390
2071
+ },
2072
+ {
2073
+ "epoch": 0.63,
2074
+ "learning_rate": 0.00016864625599409812,
2075
+ "loss": 1.2121,
2076
+ "step": 3400
2077
+ },
2078
+ {
2079
+ "epoch": 0.63,
2080
+ "learning_rate": 0.00016855403909996313,
2081
+ "loss": 1.167,
2082
+ "step": 3410
2083
+ },
2084
+ {
2085
+ "epoch": 0.63,
2086
+ "learning_rate": 0.00016846182220582812,
2087
+ "loss": 1.3436,
2088
+ "step": 3420
2089
+ },
2090
+ {
2091
+ "epoch": 0.63,
2092
+ "learning_rate": 0.0001683696053116931,
2093
+ "loss": 1.1877,
2094
+ "step": 3430
2095
+ },
2096
+ {
2097
+ "epoch": 0.63,
2098
+ "learning_rate": 0.00016827738841755812,
2099
+ "loss": 1.2584,
2100
+ "step": 3440
2101
+ },
2102
+ {
2103
+ "epoch": 0.64,
2104
+ "learning_rate": 0.0001681851715234231,
2105
+ "loss": 1.3245,
2106
+ "step": 3450
2107
+ },
2108
+ {
2109
+ "epoch": 0.64,
2110
+ "learning_rate": 0.00016809295462928808,
2111
+ "loss": 1.2784,
2112
+ "step": 3460
2113
+ },
2114
+ {
2115
+ "epoch": 0.64,
2116
+ "learning_rate": 0.00016800073773515307,
2117
+ "loss": 1.0824,
2118
+ "step": 3470
2119
+ },
2120
+ {
2121
+ "epoch": 0.64,
2122
+ "learning_rate": 0.00016790852084101808,
2123
+ "loss": 1.1694,
2124
+ "step": 3480
2125
+ },
2126
+ {
2127
+ "epoch": 0.64,
2128
+ "learning_rate": 0.0001678163039468831,
2129
+ "loss": 1.1642,
2130
+ "step": 3490
2131
+ },
2132
+ {
2133
+ "epoch": 0.65,
2134
+ "learning_rate": 0.00016772408705274808,
2135
+ "loss": 1.1797,
2136
+ "step": 3500
2137
+ },
2138
+ {
2139
+ "epoch": 0.65,
2140
+ "learning_rate": 0.00016763187015861306,
2141
+ "loss": 1.2161,
2142
+ "step": 3510
2143
+ },
2144
+ {
2145
+ "epoch": 0.65,
2146
+ "learning_rate": 0.00016753965326447807,
2147
+ "loss": 1.2212,
2148
+ "step": 3520
2149
+ },
2150
+ {
2151
+ "epoch": 0.65,
2152
+ "learning_rate": 0.00016744743637034306,
2153
+ "loss": 1.0782,
2154
+ "step": 3530
2155
+ },
2156
+ {
2157
+ "epoch": 0.65,
2158
+ "learning_rate": 0.00016735521947620804,
2159
+ "loss": 1.2315,
2160
+ "step": 3540
2161
+ },
2162
+ {
2163
+ "epoch": 0.65,
2164
+ "learning_rate": 0.00016726300258207302,
2165
+ "loss": 1.1335,
2166
+ "step": 3550
2167
+ },
2168
+ {
2169
+ "epoch": 0.66,
2170
+ "learning_rate": 0.00016717078568793804,
2171
+ "loss": 1.1268,
2172
+ "step": 3560
2173
+ },
2174
+ {
2175
+ "epoch": 0.66,
2176
+ "learning_rate": 0.00016707856879380305,
2177
+ "loss": 1.3353,
2178
+ "step": 3570
2179
+ },
2180
+ {
2181
+ "epoch": 0.66,
2182
+ "learning_rate": 0.00016698635189966803,
2183
+ "loss": 1.2586,
2184
+ "step": 3580
2185
+ },
2186
+ {
2187
+ "epoch": 0.66,
2188
+ "learning_rate": 0.00016689413500553302,
2189
+ "loss": 1.3272,
2190
+ "step": 3590
2191
+ },
2192
+ {
2193
+ "epoch": 0.66,
2194
+ "learning_rate": 0.00016680191811139803,
2195
+ "loss": 1.1169,
2196
+ "step": 3600
2197
+ },
2198
+ {
2199
+ "epoch": 0.67,
2200
+ "learning_rate": 0.000166709701217263,
2201
+ "loss": 1.1928,
2202
+ "step": 3610
2203
+ },
2204
+ {
2205
+ "epoch": 0.67,
2206
+ "learning_rate": 0.000166617484323128,
2207
+ "loss": 1.0965,
2208
+ "step": 3620
2209
+ },
2210
+ {
2211
+ "epoch": 0.67,
2212
+ "learning_rate": 0.00016652526742899298,
2213
+ "loss": 1.2023,
2214
+ "step": 3630
2215
+ },
2216
+ {
2217
+ "epoch": 0.67,
2218
+ "learning_rate": 0.000166433050534858,
2219
+ "loss": 1.239,
2220
+ "step": 3640
2221
+ },
2222
+ {
2223
+ "epoch": 0.67,
2224
+ "learning_rate": 0.000166340833640723,
2225
+ "loss": 1.2667,
2226
+ "step": 3650
2227
+ },
2228
+ {
2229
+ "epoch": 0.68,
2230
+ "learning_rate": 0.000166248616746588,
2231
+ "loss": 1.3098,
2232
+ "step": 3660
2233
+ },
2234
+ {
2235
+ "epoch": 0.68,
2236
+ "learning_rate": 0.00016615639985245297,
2237
+ "loss": 1.1174,
2238
+ "step": 3670
2239
+ },
2240
+ {
2241
+ "epoch": 0.68,
2242
+ "learning_rate": 0.00016606418295831798,
2243
+ "loss": 1.1965,
2244
+ "step": 3680
2245
+ },
2246
+ {
2247
+ "epoch": 0.68,
2248
+ "learning_rate": 0.00016597196606418297,
2249
+ "loss": 1.1481,
2250
+ "step": 3690
2251
+ },
2252
+ {
2253
+ "epoch": 0.68,
2254
+ "learning_rate": 0.00016587974917004795,
2255
+ "loss": 1.1209,
2256
+ "step": 3700
2257
+ },
2258
+ {
2259
+ "epoch": 0.68,
2260
+ "learning_rate": 0.00016578753227591294,
2261
+ "loss": 1.1346,
2262
+ "step": 3710
2263
+ },
2264
+ {
2265
+ "epoch": 0.69,
2266
+ "learning_rate": 0.00016569531538177795,
2267
+ "loss": 1.2816,
2268
+ "step": 3720
2269
+ },
2270
+ {
2271
+ "epoch": 0.69,
2272
+ "learning_rate": 0.00016560309848764296,
2273
+ "loss": 1.1948,
2274
+ "step": 3730
2275
+ },
2276
+ {
2277
+ "epoch": 0.69,
2278
+ "learning_rate": 0.00016551088159350795,
2279
+ "loss": 1.1452,
2280
+ "step": 3740
2281
+ },
2282
+ {
2283
+ "epoch": 0.69,
2284
+ "learning_rate": 0.00016541866469937293,
2285
+ "loss": 1.1751,
2286
+ "step": 3750
2287
+ },
2288
+ {
2289
+ "epoch": 0.69,
2290
+ "learning_rate": 0.00016532644780523794,
2291
+ "loss": 1.212,
2292
+ "step": 3760
2293
+ },
2294
+ {
2295
+ "epoch": 0.7,
2296
+ "learning_rate": 0.00016523423091110293,
2297
+ "loss": 1.1469,
2298
+ "step": 3770
2299
+ },
2300
+ {
2301
+ "epoch": 0.7,
2302
+ "learning_rate": 0.0001651420140169679,
2303
+ "loss": 1.2593,
2304
+ "step": 3780
2305
+ },
2306
+ {
2307
+ "epoch": 0.7,
2308
+ "learning_rate": 0.0001650497971228329,
2309
+ "loss": 1.1924,
2310
+ "step": 3790
2311
+ },
2312
+ {
2313
+ "epoch": 0.7,
2314
+ "learning_rate": 0.0001649575802286979,
2315
+ "loss": 1.1856,
2316
+ "step": 3800
2317
+ },
2318
+ {
2319
+ "epoch": 0.7,
2320
+ "learning_rate": 0.00016486536333456292,
2321
+ "loss": 1.2306,
2322
+ "step": 3810
2323
+ },
2324
+ {
2325
+ "epoch": 0.7,
2326
+ "learning_rate": 0.0001647731464404279,
2327
+ "loss": 1.2943,
2328
+ "step": 3820
2329
+ },
2330
+ {
2331
+ "epoch": 0.71,
2332
+ "learning_rate": 0.00016468092954629289,
2333
+ "loss": 1.1842,
2334
+ "step": 3830
2335
+ },
2336
+ {
2337
+ "epoch": 0.71,
2338
+ "learning_rate": 0.00016458871265215787,
2339
+ "loss": 1.322,
2340
+ "step": 3840
2341
+ },
2342
+ {
2343
+ "epoch": 0.71,
2344
+ "learning_rate": 0.00016449649575802288,
2345
+ "loss": 1.1888,
2346
+ "step": 3850
2347
+ },
2348
+ {
2349
+ "epoch": 0.71,
2350
+ "learning_rate": 0.00016440427886388787,
2351
+ "loss": 1.1909,
2352
+ "step": 3860
2353
+ },
2354
+ {
2355
+ "epoch": 0.71,
2356
+ "learning_rate": 0.00016431206196975285,
2357
+ "loss": 1.2228,
2358
+ "step": 3870
2359
+ },
2360
+ {
2361
+ "epoch": 0.72,
2362
+ "learning_rate": 0.00016421984507561786,
2363
+ "loss": 1.1169,
2364
+ "step": 3880
2365
+ },
2366
+ {
2367
+ "epoch": 0.72,
2368
+ "learning_rate": 0.00016412762818148287,
2369
+ "loss": 1.2129,
2370
+ "step": 3890
2371
+ },
2372
+ {
2373
+ "epoch": 0.72,
2374
+ "learning_rate": 0.00016403541128734786,
2375
+ "loss": 1.2108,
2376
+ "step": 3900
2377
+ },
2378
+ {
2379
+ "epoch": 0.72,
2380
+ "learning_rate": 0.00016394319439321284,
2381
+ "loss": 1.3192,
2382
+ "step": 3910
2383
+ },
2384
+ {
2385
+ "epoch": 0.72,
2386
+ "learning_rate": 0.00016385097749907783,
2387
+ "loss": 1.11,
2388
+ "step": 3920
2389
+ },
2390
+ {
2391
+ "epoch": 0.72,
2392
+ "learning_rate": 0.00016375876060494284,
2393
+ "loss": 1.1571,
2394
+ "step": 3930
2395
+ },
2396
+ {
2397
+ "epoch": 0.73,
2398
+ "learning_rate": 0.00016366654371080782,
2399
+ "loss": 1.2533,
2400
+ "step": 3940
2401
+ },
2402
+ {
2403
+ "epoch": 0.73,
2404
+ "learning_rate": 0.0001635743268166728,
2405
+ "loss": 1.0954,
2406
+ "step": 3950
2407
+ },
2408
+ {
2409
+ "epoch": 0.73,
2410
+ "learning_rate": 0.00016348210992253782,
2411
+ "loss": 1.1373,
2412
+ "step": 3960
2413
+ },
2414
+ {
2415
+ "epoch": 0.73,
2416
+ "learning_rate": 0.00016338989302840283,
2417
+ "loss": 1.1435,
2418
+ "step": 3970
2419
+ },
2420
+ {
2421
+ "epoch": 0.73,
2422
+ "learning_rate": 0.00016329767613426782,
2423
+ "loss": 1.1704,
2424
+ "step": 3980
2425
+ },
2426
+ {
2427
+ "epoch": 0.74,
2428
+ "learning_rate": 0.0001632054592401328,
2429
+ "loss": 1.0662,
2430
+ "step": 3990
2431
+ },
2432
+ {
2433
+ "epoch": 0.74,
2434
+ "learning_rate": 0.00016311324234599778,
2435
+ "loss": 1.3121,
2436
+ "step": 4000
2437
+ },
2438
+ {
2439
+ "epoch": 0.74,
2440
+ "eval_accuracy": 0.49945225488406064,
2441
+ "eval_loss": 1.1661431789398193,
2442
+ "eval_runtime": 98.5644,
2443
+ "eval_samples_per_second": 111.135,
2444
+ "eval_steps_per_second": 13.9,
2445
+ "step": 4000
2446
+ },
2447
+ {
2448
+ "epoch": 0.74,
2449
+ "learning_rate": 0.0001630210254518628,
2450
+ "loss": 1.2862,
2451
+ "step": 4010
2452
+ },
2453
+ {
2454
+ "epoch": 0.74,
2455
+ "learning_rate": 0.00016292880855772778,
2456
+ "loss": 1.2184,
2457
+ "step": 4020
2458
+ },
2459
+ {
2460
+ "epoch": 0.74,
2461
+ "learning_rate": 0.00016283659166359276,
2462
+ "loss": 1.1609,
2463
+ "step": 4030
2464
+ },
2465
+ {
2466
+ "epoch": 0.75,
2467
+ "learning_rate": 0.00016274437476945778,
2468
+ "loss": 1.1033,
2469
+ "step": 4040
2470
+ },
2471
+ {
2472
+ "epoch": 0.75,
2473
+ "learning_rate": 0.0001626521578753228,
2474
+ "loss": 1.1446,
2475
+ "step": 4050
2476
+ },
2477
+ {
2478
+ "epoch": 0.75,
2479
+ "learning_rate": 0.00016255994098118777,
2480
+ "loss": 1.2345,
2481
+ "step": 4060
2482
+ },
2483
+ {
2484
+ "epoch": 0.75,
2485
+ "learning_rate": 0.00016246772408705276,
2486
+ "loss": 1.1732,
2487
+ "step": 4070
2488
+ },
2489
+ {
2490
+ "epoch": 0.75,
2491
+ "learning_rate": 0.00016237550719291774,
2492
+ "loss": 1.2738,
2493
+ "step": 4080
2494
+ },
2495
+ {
2496
+ "epoch": 0.75,
2497
+ "learning_rate": 0.00016228329029878275,
2498
+ "loss": 1.2372,
2499
+ "step": 4090
2500
+ },
2501
+ {
2502
+ "epoch": 0.76,
2503
+ "learning_rate": 0.00016219107340464774,
2504
+ "loss": 1.142,
2505
+ "step": 4100
2506
+ },
2507
+ {
2508
+ "epoch": 0.76,
2509
+ "learning_rate": 0.00016209885651051272,
2510
+ "loss": 1.277,
2511
+ "step": 4110
2512
+ },
2513
+ {
2514
+ "epoch": 0.76,
2515
+ "learning_rate": 0.00016200663961637773,
2516
+ "loss": 1.1757,
2517
+ "step": 4120
2518
+ },
2519
+ {
2520
+ "epoch": 0.76,
2521
+ "learning_rate": 0.00016191442272224274,
2522
+ "loss": 1.239,
2523
+ "step": 4130
2524
+ },
2525
+ {
2526
+ "epoch": 0.76,
2527
+ "learning_rate": 0.00016182220582810773,
2528
+ "loss": 1.1598,
2529
+ "step": 4140
2530
+ },
2531
+ {
2532
+ "epoch": 0.77,
2533
+ "learning_rate": 0.0001617299889339727,
2534
+ "loss": 1.2175,
2535
+ "step": 4150
2536
+ },
2537
+ {
2538
+ "epoch": 0.77,
2539
+ "learning_rate": 0.0001616377720398377,
2540
+ "loss": 1.1578,
2541
+ "step": 4160
2542
+ },
2543
+ {
2544
+ "epoch": 0.77,
2545
+ "learning_rate": 0.0001615455551457027,
2546
+ "loss": 1.2052,
2547
+ "step": 4170
2548
+ },
2549
+ {
2550
+ "epoch": 0.77,
2551
+ "learning_rate": 0.0001614533382515677,
2552
+ "loss": 1.1283,
2553
+ "step": 4180
2554
+ },
2555
+ {
2556
+ "epoch": 0.77,
2557
+ "learning_rate": 0.00016136112135743268,
2558
+ "loss": 1.233,
2559
+ "step": 4190
2560
+ },
2561
+ {
2562
+ "epoch": 0.77,
2563
+ "learning_rate": 0.0001612689044632977,
2564
+ "loss": 1.1148,
2565
+ "step": 4200
2566
+ },
2567
+ {
2568
+ "epoch": 0.78,
2569
+ "learning_rate": 0.0001611766875691627,
2570
+ "loss": 1.1895,
2571
+ "step": 4210
2572
+ },
2573
+ {
2574
+ "epoch": 0.78,
2575
+ "learning_rate": 0.00016108447067502768,
2576
+ "loss": 1.3549,
2577
+ "step": 4220
2578
+ },
2579
+ {
2580
+ "epoch": 0.78,
2581
+ "learning_rate": 0.00016099225378089267,
2582
+ "loss": 1.2073,
2583
+ "step": 4230
2584
+ },
2585
+ {
2586
+ "epoch": 0.78,
2587
+ "learning_rate": 0.00016090003688675765,
2588
+ "loss": 1.194,
2589
+ "step": 4240
2590
+ },
2591
+ {
2592
+ "epoch": 0.78,
2593
+ "learning_rate": 0.00016080781999262267,
2594
+ "loss": 1.2183,
2595
+ "step": 4250
2596
+ },
2597
+ {
2598
+ "epoch": 0.79,
2599
+ "learning_rate": 0.00016071560309848765,
2600
+ "loss": 1.2088,
2601
+ "step": 4260
2602
+ },
2603
+ {
2604
+ "epoch": 0.79,
2605
+ "learning_rate": 0.00016062338620435263,
2606
+ "loss": 1.3126,
2607
+ "step": 4270
2608
+ },
2609
+ {
2610
+ "epoch": 0.79,
2611
+ "learning_rate": 0.00016053116931021765,
2612
+ "loss": 1.2154,
2613
+ "step": 4280
2614
+ },
2615
+ {
2616
+ "epoch": 0.79,
2617
+ "learning_rate": 0.00016043895241608266,
2618
+ "loss": 1.1869,
2619
+ "step": 4290
2620
+ },
2621
+ {
2622
+ "epoch": 0.79,
2623
+ "learning_rate": 0.00016034673552194764,
2624
+ "loss": 1.1124,
2625
+ "step": 4300
2626
+ },
2627
+ {
2628
+ "epoch": 0.79,
2629
+ "learning_rate": 0.00016025451862781263,
2630
+ "loss": 1.1789,
2631
+ "step": 4310
2632
+ },
2633
+ {
2634
+ "epoch": 0.8,
2635
+ "learning_rate": 0.0001601623017336776,
2636
+ "loss": 1.1254,
2637
+ "step": 4320
2638
+ },
2639
+ {
2640
+ "epoch": 0.8,
2641
+ "learning_rate": 0.00016007008483954262,
2642
+ "loss": 1.3232,
2643
+ "step": 4330
2644
+ },
2645
+ {
2646
+ "epoch": 0.8,
2647
+ "learning_rate": 0.0001599778679454076,
2648
+ "loss": 1.1419,
2649
+ "step": 4340
2650
+ },
2651
+ {
2652
+ "epoch": 0.8,
2653
+ "learning_rate": 0.0001598856510512726,
2654
+ "loss": 1.2861,
2655
+ "step": 4350
2656
+ },
2657
+ {
2658
+ "epoch": 0.8,
2659
+ "learning_rate": 0.0001597934341571376,
2660
+ "loss": 1.1492,
2661
+ "step": 4360
2662
+ },
2663
+ {
2664
+ "epoch": 0.81,
2665
+ "learning_rate": 0.0001597012172630026,
2666
+ "loss": 1.2304,
2667
+ "step": 4370
2668
+ },
2669
+ {
2670
+ "epoch": 0.81,
2671
+ "learning_rate": 0.0001596090003688676,
2672
+ "loss": 1.2289,
2673
+ "step": 4380
2674
+ },
2675
+ {
2676
+ "epoch": 0.81,
2677
+ "learning_rate": 0.00015951678347473258,
2678
+ "loss": 1.1479,
2679
+ "step": 4390
2680
+ },
2681
+ {
2682
+ "epoch": 0.81,
2683
+ "learning_rate": 0.00015942456658059757,
2684
+ "loss": 1.222,
2685
+ "step": 4400
2686
+ },
2687
+ {
2688
+ "epoch": 0.81,
2689
+ "learning_rate": 0.00015933234968646258,
2690
+ "loss": 1.105,
2691
+ "step": 4410
2692
+ },
2693
+ {
2694
+ "epoch": 0.82,
2695
+ "learning_rate": 0.00015924013279232756,
2696
+ "loss": 1.1094,
2697
+ "step": 4420
2698
+ },
2699
+ {
2700
+ "epoch": 0.82,
2701
+ "learning_rate": 0.00015914791589819255,
2702
+ "loss": 1.0859,
2703
+ "step": 4430
2704
+ },
2705
+ {
2706
+ "epoch": 0.82,
2707
+ "learning_rate": 0.00015905569900405756,
2708
+ "loss": 1.2148,
2709
+ "step": 4440
2710
+ },
2711
+ {
2712
+ "epoch": 0.82,
2713
+ "learning_rate": 0.00015896348210992254,
2714
+ "loss": 1.2015,
2715
+ "step": 4450
2716
+ },
2717
+ {
2718
+ "epoch": 0.82,
2719
+ "learning_rate": 0.00015887126521578755,
2720
+ "loss": 1.0869,
2721
+ "step": 4460
2722
+ },
2723
+ {
2724
+ "epoch": 0.82,
2725
+ "learning_rate": 0.00015877904832165254,
2726
+ "loss": 1.2852,
2727
+ "step": 4470
2728
+ },
2729
+ {
2730
+ "epoch": 0.83,
2731
+ "learning_rate": 0.00015868683142751752,
2732
+ "loss": 1.14,
2733
+ "step": 4480
2734
+ },
2735
+ {
2736
+ "epoch": 0.83,
2737
+ "learning_rate": 0.0001585946145333825,
2738
+ "loss": 1.2118,
2739
+ "step": 4490
2740
+ },
2741
+ {
2742
+ "epoch": 0.83,
2743
+ "learning_rate": 0.00015850239763924752,
2744
+ "loss": 1.1661,
2745
+ "step": 4500
2746
+ },
2747
+ {
2748
+ "epoch": 0.83,
2749
+ "learning_rate": 0.0001584101807451125,
2750
+ "loss": 1.07,
2751
+ "step": 4510
2752
+ },
2753
+ {
2754
+ "epoch": 0.83,
2755
+ "learning_rate": 0.00015831796385097752,
2756
+ "loss": 1.1214,
2757
+ "step": 4520
2758
+ },
2759
+ {
2760
+ "epoch": 0.84,
2761
+ "learning_rate": 0.0001582257469568425,
2762
+ "loss": 1.285,
2763
+ "step": 4530
2764
+ },
2765
+ {
2766
+ "epoch": 0.84,
2767
+ "learning_rate": 0.0001581335300627075,
2768
+ "loss": 1.1742,
2769
+ "step": 4540
2770
+ },
2771
+ {
2772
+ "epoch": 0.84,
2773
+ "learning_rate": 0.0001580413131685725,
2774
+ "loss": 1.1425,
2775
+ "step": 4550
2776
+ },
2777
+ {
2778
+ "epoch": 0.84,
2779
+ "learning_rate": 0.00015794909627443748,
2780
+ "loss": 1.0878,
2781
+ "step": 4560
2782
+ },
2783
+ {
2784
+ "epoch": 0.84,
2785
+ "learning_rate": 0.00015785687938030246,
2786
+ "loss": 1.2311,
2787
+ "step": 4570
2788
+ },
2789
+ {
2790
+ "epoch": 0.84,
2791
+ "learning_rate": 0.00015776466248616748,
2792
+ "loss": 1.1238,
2793
+ "step": 4580
2794
+ },
2795
+ {
2796
+ "epoch": 0.85,
2797
+ "learning_rate": 0.00015767244559203246,
2798
+ "loss": 1.0743,
2799
+ "step": 4590
2800
+ },
2801
+ {
2802
+ "epoch": 0.85,
2803
+ "learning_rate": 0.00015758022869789747,
2804
+ "loss": 1.0789,
2805
+ "step": 4600
2806
+ },
2807
+ {
2808
+ "epoch": 0.85,
2809
+ "learning_rate": 0.00015748801180376246,
2810
+ "loss": 1.1919,
2811
+ "step": 4610
2812
+ },
2813
+ {
2814
+ "epoch": 0.85,
2815
+ "learning_rate": 0.00015739579490962747,
2816
+ "loss": 1.1697,
2817
+ "step": 4620
2818
+ },
2819
+ {
2820
+ "epoch": 0.85,
2821
+ "learning_rate": 0.00015730357801549245,
2822
+ "loss": 1.2384,
2823
+ "step": 4630
2824
+ },
2825
+ {
2826
+ "epoch": 0.86,
2827
+ "learning_rate": 0.00015721136112135744,
2828
+ "loss": 1.1858,
2829
+ "step": 4640
2830
+ },
2831
+ {
2832
+ "epoch": 0.86,
2833
+ "learning_rate": 0.00015711914422722242,
2834
+ "loss": 1.1766,
2835
+ "step": 4650
2836
+ },
2837
+ {
2838
+ "epoch": 0.86,
2839
+ "learning_rate": 0.00015702692733308743,
2840
+ "loss": 1.0852,
2841
+ "step": 4660
2842
+ },
2843
+ {
2844
+ "epoch": 0.86,
2845
+ "learning_rate": 0.00015693471043895242,
2846
+ "loss": 1.1499,
2847
+ "step": 4670
2848
+ },
2849
+ {
2850
+ "epoch": 0.86,
2851
+ "learning_rate": 0.00015684249354481743,
2852
+ "loss": 1.2804,
2853
+ "step": 4680
2854
+ },
2855
+ {
2856
+ "epoch": 0.86,
2857
+ "learning_rate": 0.0001567502766506824,
2858
+ "loss": 1.2195,
2859
+ "step": 4690
2860
+ },
2861
+ {
2862
+ "epoch": 0.87,
2863
+ "learning_rate": 0.00015665805975654742,
2864
+ "loss": 1.1809,
2865
+ "step": 4700
2866
+ },
2867
+ {
2868
+ "epoch": 0.87,
2869
+ "learning_rate": 0.0001565658428624124,
2870
+ "loss": 1.2306,
2871
+ "step": 4710
2872
+ },
2873
+ {
2874
+ "epoch": 0.87,
2875
+ "learning_rate": 0.0001564736259682774,
2876
+ "loss": 1.1516,
2877
+ "step": 4720
2878
+ },
2879
+ {
2880
+ "epoch": 0.87,
2881
+ "learning_rate": 0.00015638140907414238,
2882
+ "loss": 1.1354,
2883
+ "step": 4730
2884
+ },
2885
+ {
2886
+ "epoch": 0.87,
2887
+ "learning_rate": 0.0001562891921800074,
2888
+ "loss": 1.1047,
2889
+ "step": 4740
2890
+ },
2891
+ {
2892
+ "epoch": 0.88,
2893
+ "learning_rate": 0.00015619697528587237,
2894
+ "loss": 1.1401,
2895
+ "step": 4750
2896
+ },
2897
+ {
2898
+ "epoch": 0.88,
2899
+ "learning_rate": 0.00015610475839173736,
2900
+ "loss": 1.1011,
2901
+ "step": 4760
2902
+ },
2903
+ {
2904
+ "epoch": 0.88,
2905
+ "learning_rate": 0.00015601254149760237,
2906
+ "loss": 1.0627,
2907
+ "step": 4770
2908
+ },
2909
+ {
2910
+ "epoch": 0.88,
2911
+ "learning_rate": 0.00015592032460346738,
2912
+ "loss": 1.2008,
2913
+ "step": 4780
2914
+ },
2915
+ {
2916
+ "epoch": 0.88,
2917
+ "learning_rate": 0.00015582810770933237,
2918
+ "loss": 1.0876,
2919
+ "step": 4790
2920
+ },
2921
+ {
2922
+ "epoch": 0.89,
2923
+ "learning_rate": 0.00015573589081519735,
2924
+ "loss": 1.2179,
2925
+ "step": 4800
2926
+ },
2927
+ {
2928
+ "epoch": 0.89,
2929
+ "learning_rate": 0.00015564367392106233,
2930
+ "loss": 1.1274,
2931
+ "step": 4810
2932
+ },
2933
+ {
2934
+ "epoch": 0.89,
2935
+ "learning_rate": 0.00015555145702692735,
2936
+ "loss": 1.2143,
2937
+ "step": 4820
2938
+ },
2939
+ {
2940
+ "epoch": 0.89,
2941
+ "learning_rate": 0.00015545924013279233,
2942
+ "loss": 1.264,
2943
+ "step": 4830
2944
+ },
2945
+ {
2946
+ "epoch": 0.89,
2947
+ "learning_rate": 0.00015536702323865731,
2948
+ "loss": 1.0355,
2949
+ "step": 4840
2950
+ },
2951
+ {
2952
+ "epoch": 0.89,
2953
+ "learning_rate": 0.00015527480634452233,
2954
+ "loss": 1.1883,
2955
+ "step": 4850
2956
+ },
2957
+ {
2958
+ "epoch": 0.9,
2959
+ "learning_rate": 0.00015518258945038734,
2960
+ "loss": 1.151,
2961
+ "step": 4860
2962
+ },
2963
+ {
2964
+ "epoch": 0.9,
2965
+ "learning_rate": 0.00015509037255625232,
2966
+ "loss": 1.1937,
2967
+ "step": 4870
2968
+ },
2969
+ {
2970
+ "epoch": 0.9,
2971
+ "learning_rate": 0.0001549981556621173,
2972
+ "loss": 1.1426,
2973
+ "step": 4880
2974
+ },
2975
+ {
2976
+ "epoch": 0.9,
2977
+ "learning_rate": 0.0001549059387679823,
2978
+ "loss": 1.1356,
2979
+ "step": 4890
2980
+ },
2981
+ {
2982
+ "epoch": 0.9,
2983
+ "learning_rate": 0.0001548137218738473,
2984
+ "loss": 1.0953,
2985
+ "step": 4900
2986
+ },
2987
+ {
2988
+ "epoch": 0.91,
2989
+ "learning_rate": 0.0001547215049797123,
2990
+ "loss": 1.2168,
2991
+ "step": 4910
2992
+ },
2993
+ {
2994
+ "epoch": 0.91,
2995
+ "learning_rate": 0.00015462928808557727,
2996
+ "loss": 1.2641,
2997
+ "step": 4920
2998
+ },
2999
+ {
3000
+ "epoch": 0.91,
3001
+ "learning_rate": 0.00015453707119144228,
3002
+ "loss": 0.9988,
3003
+ "step": 4930
3004
+ },
3005
+ {
3006
+ "epoch": 0.91,
3007
+ "learning_rate": 0.0001544448542973073,
3008
+ "loss": 1.2423,
3009
+ "step": 4940
3010
+ },
3011
+ {
3012
+ "epoch": 0.91,
3013
+ "learning_rate": 0.00015435263740317228,
3014
+ "loss": 1.1201,
3015
+ "step": 4950
3016
+ },
3017
+ {
3018
+ "epoch": 0.91,
3019
+ "learning_rate": 0.00015426042050903726,
3020
+ "loss": 1.1261,
3021
+ "step": 4960
3022
+ },
3023
+ {
3024
+ "epoch": 0.92,
3025
+ "learning_rate": 0.00015416820361490225,
3026
+ "loss": 1.1374,
3027
+ "step": 4970
3028
+ },
3029
+ {
3030
+ "epoch": 0.92,
3031
+ "learning_rate": 0.00015407598672076726,
3032
+ "loss": 0.9945,
3033
+ "step": 4980
3034
+ },
3035
+ {
3036
+ "epoch": 0.92,
3037
+ "learning_rate": 0.00015398376982663224,
3038
+ "loss": 1.1362,
3039
+ "step": 4990
3040
+ },
3041
+ {
3042
+ "epoch": 0.92,
3043
+ "learning_rate": 0.00015389155293249723,
3044
+ "loss": 1.1806,
3045
+ "step": 5000
3046
+ },
3047
+ {
3048
+ "epoch": 0.92,
3049
+ "eval_accuracy": 0.5240094942486763,
3050
+ "eval_loss": 1.1137409210205078,
3051
+ "eval_runtime": 98.12,
3052
+ "eval_samples_per_second": 111.639,
3053
+ "eval_steps_per_second": 13.963,
3054
+ "step": 5000
3055
+ },
3056
+ {
3057
+ "epoch": 0.92,
3058
+ "learning_rate": 0.00015379933603836224,
3059
+ "loss": 1.1386,
3060
+ "step": 5010
3061
+ },
3062
+ {
3063
+ "epoch": 0.93,
3064
+ "learning_rate": 0.00015370711914422725,
3065
+ "loss": 1.129,
3066
+ "step": 5020
3067
+ },
3068
+ {
3069
+ "epoch": 0.93,
3070
+ "learning_rate": 0.00015361490225009223,
3071
+ "loss": 1.1191,
3072
+ "step": 5030
3073
+ },
3074
+ {
3075
+ "epoch": 0.93,
3076
+ "learning_rate": 0.00015352268535595722,
3077
+ "loss": 1.2112,
3078
+ "step": 5040
3079
+ },
3080
+ {
3081
+ "epoch": 0.93,
3082
+ "learning_rate": 0.0001534304684618222,
3083
+ "loss": 1.0587,
3084
+ "step": 5050
3085
+ },
3086
+ {
3087
+ "epoch": 0.93,
3088
+ "learning_rate": 0.0001533382515676872,
3089
+ "loss": 1.2067,
3090
+ "step": 5060
3091
+ },
3092
+ {
3093
+ "epoch": 0.94,
3094
+ "learning_rate": 0.0001532460346735522,
3095
+ "loss": 1.1811,
3096
+ "step": 5070
3097
+ },
3098
+ {
3099
+ "epoch": 0.94,
3100
+ "learning_rate": 0.00015315381777941718,
3101
+ "loss": 1.1983,
3102
+ "step": 5080
3103
+ },
3104
+ {
3105
+ "epoch": 0.94,
3106
+ "learning_rate": 0.0001530616008852822,
3107
+ "loss": 1.2142,
3108
+ "step": 5090
3109
+ },
3110
+ {
3111
+ "epoch": 0.94,
3112
+ "learning_rate": 0.00015296938399114718,
3113
+ "loss": 1.1877,
3114
+ "step": 5100
3115
+ },
3116
+ {
3117
+ "epoch": 0.94,
3118
+ "learning_rate": 0.0001528771670970122,
3119
+ "loss": 1.1208,
3120
+ "step": 5110
3121
+ },
3122
+ {
3123
+ "epoch": 0.94,
3124
+ "learning_rate": 0.00015278495020287718,
3125
+ "loss": 1.2037,
3126
+ "step": 5120
3127
+ },
3128
+ {
3129
+ "epoch": 0.95,
3130
+ "learning_rate": 0.00015269273330874216,
3131
+ "loss": 1.0518,
3132
+ "step": 5130
3133
+ },
3134
+ {
3135
+ "epoch": 0.95,
3136
+ "learning_rate": 0.00015260051641460714,
3137
+ "loss": 1.1385,
3138
+ "step": 5140
3139
+ },
3140
+ {
3141
+ "epoch": 0.95,
3142
+ "learning_rate": 0.00015250829952047216,
3143
+ "loss": 1.2212,
3144
+ "step": 5150
3145
+ },
3146
+ {
3147
+ "epoch": 0.95,
3148
+ "learning_rate": 0.00015241608262633714,
3149
+ "loss": 1.2147,
3150
+ "step": 5160
3151
+ },
3152
+ {
3153
+ "epoch": 0.95,
3154
+ "learning_rate": 0.00015232386573220215,
3155
+ "loss": 1.0961,
3156
+ "step": 5170
3157
+ },
3158
+ {
3159
+ "epoch": 0.96,
3160
+ "learning_rate": 0.00015223164883806714,
3161
+ "loss": 1.1507,
3162
+ "step": 5180
3163
+ },
3164
+ {
3165
+ "epoch": 0.96,
3166
+ "learning_rate": 0.00015213943194393215,
3167
+ "loss": 1.2598,
3168
+ "step": 5190
3169
+ },
3170
+ {
3171
+ "epoch": 0.96,
3172
+ "learning_rate": 0.00015204721504979713,
3173
+ "loss": 1.1551,
3174
+ "step": 5200
3175
+ },
3176
+ {
3177
+ "epoch": 0.96,
3178
+ "learning_rate": 0.00015195499815566212,
3179
+ "loss": 1.3621,
3180
+ "step": 5210
3181
+ },
3182
+ {
3183
+ "epoch": 0.96,
3184
+ "learning_rate": 0.0001518627812615271,
3185
+ "loss": 1.1444,
3186
+ "step": 5220
3187
+ },
3188
+ {
3189
+ "epoch": 0.96,
3190
+ "learning_rate": 0.0001517705643673921,
3191
+ "loss": 1.1967,
3192
+ "step": 5230
3193
+ },
3194
+ {
3195
+ "epoch": 0.97,
3196
+ "learning_rate": 0.0001516783474732571,
3197
+ "loss": 1.1623,
3198
+ "step": 5240
3199
+ },
3200
+ {
3201
+ "epoch": 0.97,
3202
+ "learning_rate": 0.0001515861305791221,
3203
+ "loss": 1.1205,
3204
+ "step": 5250
3205
+ },
3206
+ {
3207
+ "epoch": 0.97,
3208
+ "learning_rate": 0.0001514939136849871,
3209
+ "loss": 1.0907,
3210
+ "step": 5260
3211
+ },
3212
+ {
3213
+ "epoch": 0.97,
3214
+ "learning_rate": 0.0001514016967908521,
3215
+ "loss": 1.0634,
3216
+ "step": 5270
3217
+ },
3218
+ {
3219
+ "epoch": 0.97,
3220
+ "learning_rate": 0.0001513094798967171,
3221
+ "loss": 1.1764,
3222
+ "step": 5280
3223
+ },
3224
+ {
3225
+ "epoch": 0.98,
3226
+ "learning_rate": 0.00015121726300258207,
3227
+ "loss": 1.169,
3228
+ "step": 5290
3229
+ },
3230
+ {
3231
+ "epoch": 0.98,
3232
+ "learning_rate": 0.00015112504610844706,
3233
+ "loss": 1.0883,
3234
+ "step": 5300
3235
+ },
3236
+ {
3237
+ "epoch": 0.98,
3238
+ "learning_rate": 0.00015103282921431207,
3239
+ "loss": 1.2949,
3240
+ "step": 5310
3241
+ },
3242
+ {
3243
+ "epoch": 0.98,
3244
+ "learning_rate": 0.00015094061232017705,
3245
+ "loss": 1.1286,
3246
+ "step": 5320
3247
+ },
3248
+ {
3249
+ "epoch": 0.98,
3250
+ "learning_rate": 0.00015084839542604207,
3251
+ "loss": 1.08,
3252
+ "step": 5330
3253
+ },
3254
+ {
3255
+ "epoch": 0.98,
3256
+ "learning_rate": 0.00015075617853190705,
3257
+ "loss": 1.146,
3258
+ "step": 5340
3259
+ },
3260
+ {
3261
+ "epoch": 0.99,
3262
+ "learning_rate": 0.00015066396163777206,
3263
+ "loss": 1.2019,
3264
+ "step": 5350
3265
+ },
3266
+ {
3267
+ "epoch": 0.99,
3268
+ "learning_rate": 0.00015057174474363705,
3269
+ "loss": 1.1424,
3270
+ "step": 5360
3271
+ },
3272
+ {
3273
+ "epoch": 0.99,
3274
+ "learning_rate": 0.00015047952784950203,
3275
+ "loss": 1.114,
3276
+ "step": 5370
3277
+ },
3278
+ {
3279
+ "epoch": 0.99,
3280
+ "learning_rate": 0.00015038731095536701,
3281
+ "loss": 1.1133,
3282
+ "step": 5380
3283
+ },
3284
+ {
3285
+ "epoch": 0.99,
3286
+ "learning_rate": 0.00015029509406123203,
3287
+ "loss": 1.1234,
3288
+ "step": 5390
3289
+ },
3290
+ {
3291
+ "epoch": 1.0,
3292
+ "learning_rate": 0.000150202877167097,
3293
+ "loss": 1.2356,
3294
+ "step": 5400
3295
+ },
3296
+ {
3297
+ "epoch": 1.0,
3298
+ "learning_rate": 0.00015011066027296202,
3299
+ "loss": 1.2222,
3300
+ "step": 5410
3301
+ },
3302
+ {
3303
+ "epoch": 1.0,
3304
+ "learning_rate": 0.000150018443378827,
3305
+ "loss": 1.1921,
3306
+ "step": 5420
3307
+ },
3308
+ {
3309
+ "epoch": 1.0,
3310
+ "learning_rate": 0.00014992622648469202,
3311
+ "loss": 1.1302,
3312
+ "step": 5430
3313
+ },
3314
+ {
3315
+ "epoch": 1.0,
3316
+ "learning_rate": 0.000149834009590557,
3317
+ "loss": 1.1424,
3318
+ "step": 5440
3319
+ },
3320
+ {
3321
+ "epoch": 1.01,
3322
+ "learning_rate": 0.000149741792696422,
3323
+ "loss": 1.1571,
3324
+ "step": 5450
3325
+ },
3326
+ {
3327
+ "epoch": 1.01,
3328
+ "learning_rate": 0.00014964957580228697,
3329
+ "loss": 1.0979,
3330
+ "step": 5460
3331
+ },
3332
+ {
3333
+ "epoch": 1.01,
3334
+ "learning_rate": 0.00014955735890815198,
3335
+ "loss": 1.0564,
3336
+ "step": 5470
3337
+ },
3338
+ {
3339
+ "epoch": 1.01,
3340
+ "learning_rate": 0.00014946514201401697,
3341
+ "loss": 1.007,
3342
+ "step": 5480
3343
+ },
3344
+ {
3345
+ "epoch": 1.01,
3346
+ "learning_rate": 0.00014937292511988198,
3347
+ "loss": 1.2216,
3348
+ "step": 5490
3349
+ },
3350
+ {
3351
+ "epoch": 1.01,
3352
+ "learning_rate": 0.00014928070822574696,
3353
+ "loss": 0.9655,
3354
+ "step": 5500
3355
+ },
3356
+ {
3357
+ "epoch": 1.02,
3358
+ "learning_rate": 0.00014918849133161197,
3359
+ "loss": 0.9971,
3360
+ "step": 5510
3361
+ },
3362
+ {
3363
+ "epoch": 1.02,
3364
+ "learning_rate": 0.00014909627443747696,
3365
+ "loss": 1.0476,
3366
+ "step": 5520
3367
+ },
3368
+ {
3369
+ "epoch": 1.02,
3370
+ "learning_rate": 0.00014900405754334194,
3371
+ "loss": 1.0946,
3372
+ "step": 5530
3373
+ },
3374
+ {
3375
+ "epoch": 1.02,
3376
+ "learning_rate": 0.00014891184064920693,
3377
+ "loss": 1.128,
3378
+ "step": 5540
3379
+ },
3380
+ {
3381
+ "epoch": 1.02,
3382
+ "learning_rate": 0.00014881962375507194,
3383
+ "loss": 1.124,
3384
+ "step": 5550
3385
+ },
3386
+ {
3387
+ "epoch": 1.03,
3388
+ "learning_rate": 0.00014872740686093692,
3389
+ "loss": 1.0277,
3390
+ "step": 5560
3391
+ },
3392
+ {
3393
+ "epoch": 1.03,
3394
+ "learning_rate": 0.00014863518996680193,
3395
+ "loss": 1.2334,
3396
+ "step": 5570
3397
+ },
3398
+ {
3399
+ "epoch": 1.03,
3400
+ "learning_rate": 0.00014854297307266692,
3401
+ "loss": 0.9921,
3402
+ "step": 5580
3403
+ },
3404
+ {
3405
+ "epoch": 1.03,
3406
+ "learning_rate": 0.00014845075617853193,
3407
+ "loss": 1.0135,
3408
+ "step": 5590
3409
+ },
3410
+ {
3411
+ "epoch": 1.03,
3412
+ "learning_rate": 0.00014835853928439692,
3413
+ "loss": 1.0203,
3414
+ "step": 5600
3415
+ },
3416
+ {
3417
+ "epoch": 1.03,
3418
+ "learning_rate": 0.0001482663223902619,
3419
+ "loss": 1.0778,
3420
+ "step": 5610
3421
+ },
3422
+ {
3423
+ "epoch": 1.04,
3424
+ "learning_rate": 0.00014817410549612688,
3425
+ "loss": 1.0985,
3426
+ "step": 5620
3427
+ },
3428
+ {
3429
+ "epoch": 1.04,
3430
+ "learning_rate": 0.0001480818886019919,
3431
+ "loss": 1.1,
3432
+ "step": 5630
3433
+ },
3434
+ {
3435
+ "epoch": 1.04,
3436
+ "learning_rate": 0.00014798967170785688,
3437
+ "loss": 1.1413,
3438
+ "step": 5640
3439
+ },
3440
+ {
3441
+ "epoch": 1.04,
3442
+ "learning_rate": 0.0001478974548137219,
3443
+ "loss": 1.0165,
3444
+ "step": 5650
3445
+ },
3446
+ {
3447
+ "epoch": 1.04,
3448
+ "learning_rate": 0.00014780523791958688,
3449
+ "loss": 1.0329,
3450
+ "step": 5660
3451
+ },
3452
+ {
3453
+ "epoch": 1.05,
3454
+ "learning_rate": 0.0001477130210254519,
3455
+ "loss": 1.0341,
3456
+ "step": 5670
3457
+ },
3458
+ {
3459
+ "epoch": 1.05,
3460
+ "learning_rate": 0.00014762080413131687,
3461
+ "loss": 0.9721,
3462
+ "step": 5680
3463
+ },
3464
+ {
3465
+ "epoch": 1.05,
3466
+ "learning_rate": 0.00014752858723718186,
3467
+ "loss": 1.1558,
3468
+ "step": 5690
3469
+ },
3470
+ {
3471
+ "epoch": 1.05,
3472
+ "learning_rate": 0.00014743637034304684,
3473
+ "loss": 1.0964,
3474
+ "step": 5700
3475
+ },
3476
+ {
3477
+ "epoch": 1.05,
3478
+ "learning_rate": 0.00014734415344891183,
3479
+ "loss": 0.9984,
3480
+ "step": 5710
3481
+ },
3482
+ {
3483
+ "epoch": 1.05,
3484
+ "learning_rate": 0.00014725193655477684,
3485
+ "loss": 1.1214,
3486
+ "step": 5720
3487
+ },
3488
+ {
3489
+ "epoch": 1.06,
3490
+ "learning_rate": 0.00014715971966064185,
3491
+ "loss": 1.0954,
3492
+ "step": 5730
3493
+ },
3494
+ {
3495
+ "epoch": 1.06,
3496
+ "learning_rate": 0.00014706750276650683,
3497
+ "loss": 1.0381,
3498
+ "step": 5740
3499
+ },
3500
+ {
3501
+ "epoch": 1.06,
3502
+ "learning_rate": 0.00014697528587237182,
3503
+ "loss": 1.0516,
3504
+ "step": 5750
3505
+ },
3506
+ {
3507
+ "epoch": 1.06,
3508
+ "learning_rate": 0.00014688306897823683,
3509
+ "loss": 1.1157,
3510
+ "step": 5760
3511
+ },
3512
+ {
3513
+ "epoch": 1.06,
3514
+ "learning_rate": 0.0001467908520841018,
3515
+ "loss": 0.9874,
3516
+ "step": 5770
3517
+ },
3518
+ {
3519
+ "epoch": 1.07,
3520
+ "learning_rate": 0.0001466986351899668,
3521
+ "loss": 1.0634,
3522
+ "step": 5780
3523
+ },
3524
+ {
3525
+ "epoch": 1.07,
3526
+ "learning_rate": 0.00014660641829583178,
3527
+ "loss": 1.1058,
3528
+ "step": 5790
3529
+ },
3530
+ {
3531
+ "epoch": 1.07,
3532
+ "learning_rate": 0.0001465142014016968,
3533
+ "loss": 1.0966,
3534
+ "step": 5800
3535
+ },
3536
+ {
3537
+ "epoch": 1.07,
3538
+ "learning_rate": 0.0001464219845075618,
3539
+ "loss": 1.0457,
3540
+ "step": 5810
3541
+ },
3542
+ {
3543
+ "epoch": 1.07,
3544
+ "learning_rate": 0.0001463297676134268,
3545
+ "loss": 1.1258,
3546
+ "step": 5820
3547
+ },
3548
+ {
3549
+ "epoch": 1.08,
3550
+ "learning_rate": 0.00014623755071929177,
3551
+ "loss": 1.1258,
3552
+ "step": 5830
3553
+ },
3554
+ {
3555
+ "epoch": 1.08,
3556
+ "learning_rate": 0.00014614533382515679,
3557
+ "loss": 1.2046,
3558
+ "step": 5840
3559
+ },
3560
+ {
3561
+ "epoch": 1.08,
3562
+ "learning_rate": 0.00014605311693102177,
3563
+ "loss": 1.0688,
3564
+ "step": 5850
3565
+ },
3566
+ {
3567
+ "epoch": 1.08,
3568
+ "learning_rate": 0.00014596090003688675,
3569
+ "loss": 0.9497,
3570
+ "step": 5860
3571
+ },
3572
+ {
3573
+ "epoch": 1.08,
3574
+ "learning_rate": 0.00014586868314275174,
3575
+ "loss": 1.1829,
3576
+ "step": 5870
3577
+ },
3578
+ {
3579
+ "epoch": 1.08,
3580
+ "learning_rate": 0.00014577646624861675,
3581
+ "loss": 1.1155,
3582
+ "step": 5880
3583
+ },
3584
+ {
3585
+ "epoch": 1.09,
3586
+ "learning_rate": 0.00014568424935448176,
3587
+ "loss": 1.1273,
3588
+ "step": 5890
3589
+ },
3590
+ {
3591
+ "epoch": 1.09,
3592
+ "learning_rate": 0.00014559203246034675,
3593
+ "loss": 1.011,
3594
+ "step": 5900
3595
+ },
3596
+ {
3597
+ "epoch": 1.09,
3598
+ "learning_rate": 0.00014549981556621173,
3599
+ "loss": 1.081,
3600
+ "step": 5910
3601
+ },
3602
+ {
3603
+ "epoch": 1.09,
3604
+ "learning_rate": 0.00014540759867207674,
3605
+ "loss": 1.041,
3606
+ "step": 5920
3607
+ },
3608
+ {
3609
+ "epoch": 1.09,
3610
+ "learning_rate": 0.00014531538177794173,
3611
+ "loss": 1.1286,
3612
+ "step": 5930
3613
+ },
3614
+ {
3615
+ "epoch": 1.1,
3616
+ "learning_rate": 0.0001452231648838067,
3617
+ "loss": 0.9914,
3618
+ "step": 5940
3619
+ },
3620
+ {
3621
+ "epoch": 1.1,
3622
+ "learning_rate": 0.0001451309479896717,
3623
+ "loss": 1.0105,
3624
+ "step": 5950
3625
+ },
3626
+ {
3627
+ "epoch": 1.1,
3628
+ "learning_rate": 0.0001450387310955367,
3629
+ "loss": 1.0517,
3630
+ "step": 5960
3631
+ },
3632
+ {
3633
+ "epoch": 1.1,
3634
+ "learning_rate": 0.00014494651420140172,
3635
+ "loss": 1.0245,
3636
+ "step": 5970
3637
+ },
3638
+ {
3639
+ "epoch": 1.1,
3640
+ "learning_rate": 0.0001448542973072667,
3641
+ "loss": 1.0713,
3642
+ "step": 5980
3643
+ },
3644
+ {
3645
+ "epoch": 1.1,
3646
+ "learning_rate": 0.0001447620804131317,
3647
+ "loss": 1.0638,
3648
+ "step": 5990
3649
+ },
3650
+ {
3651
+ "epoch": 1.11,
3652
+ "learning_rate": 0.0001446698635189967,
3653
+ "loss": 1.0839,
3654
+ "step": 6000
3655
+ },
3656
+ {
3657
+ "epoch": 1.11,
3658
+ "eval_accuracy": 0.5164323534781815,
3659
+ "eval_loss": 1.1340410709381104,
3660
+ "eval_runtime": 98.4525,
3661
+ "eval_samples_per_second": 111.262,
3662
+ "eval_steps_per_second": 13.915,
3663
+ "step": 6000
3664
+ },
3665
+ {
3666
+ "epoch": 1.11,
3667
+ "step": 6000,
3668
+ "total_flos": 2.0396723395200123e+18,
3669
+ "train_loss": 1.29178360859553,
3670
+ "train_runtime": 1561.0386,
3671
+ "train_samples_per_second": 222.273,
3672
+ "train_steps_per_second": 13.893
3673
+ }
3674
+ ],
3675
+ "logging_steps": 10,
3676
+ "max_steps": 21688,
3677
+ "num_train_epochs": 4,
3678
+ "save_steps": 1000,
3679
+ "total_flos": 2.0396723395200123e+18,
3680
+ "trial_name": null,
3681
+ "trial_params": null
3682
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1746d82fdd60852384979d6009f2a4915a439dc4c45129e71484536dece42636
3
+ size 4027