End of training
Browse files- README.md +112 -0
- all_results.json +13 -0
- config.json +53 -0
- eval_results.json +8 -0
- model.safetensors +3 -0
- preprocessor_config.json +22 -0
- runs/Mar07_03-18-48_606d6922676f/events.out.tfevents.1709781530.606d6922676f.1016.0 +3 -0
- runs/Mar07_03-18-48_606d6922676f/events.out.tfevents.1709785268.606d6922676f.1016.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +3280 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,112 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
base_model: microsoft/resnet-50
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- accuracy
|
8 |
+
model-index:
|
9 |
+
- name: resnet-Alzheimer
|
10 |
+
results: []
|
11 |
+
---
|
12 |
+
|
13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
+
should probably proofread and complete it, then remove this comment. -->
|
15 |
+
|
16 |
+
# resnet-Alzheimer
|
17 |
+
|
18 |
+
This model is a fine-tuned version of [microsoft/resnet-50](https://huggingface.co/microsoft/resnet-50) on an unknown dataset.
|
19 |
+
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 0.0932
|
21 |
+
- Accuracy: 0.9795
|
22 |
+
|
23 |
+
## Model description
|
24 |
+
|
25 |
+
More information needed
|
26 |
+
|
27 |
+
## Intended uses & limitations
|
28 |
+
|
29 |
+
More information needed
|
30 |
+
|
31 |
+
## Training and evaluation data
|
32 |
+
|
33 |
+
More information needed
|
34 |
+
|
35 |
+
## Training procedure
|
36 |
+
|
37 |
+
### Training hyperparameters
|
38 |
+
|
39 |
+
The following hyperparameters were used during training:
|
40 |
+
- learning_rate: 0.002
|
41 |
+
- train_batch_size: 16
|
42 |
+
- eval_batch_size: 16
|
43 |
+
- seed: 42
|
44 |
+
- gradient_accumulation_steps: 4
|
45 |
+
- total_train_batch_size: 64
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: linear
|
48 |
+
- lr_scheduler_warmup_ratio: 0.1
|
49 |
+
- num_epochs: 50
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
54 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
55 |
+
| 1.0127 | 1.0 | 80 | 0.9888 | 0.5088 |
|
56 |
+
| 0.9345 | 2.0 | 160 | 0.9422 | 0.5303 |
|
57 |
+
| 0.8889 | 3.0 | 240 | 0.8724 | 0.5781 |
|
58 |
+
| 0.8843 | 4.0 | 320 | 0.8536 | 0.5889 |
|
59 |
+
| 0.8397 | 5.0 | 400 | 0.8354 | 0.6152 |
|
60 |
+
| 0.8624 | 6.0 | 480 | 0.9221 | 0.5381 |
|
61 |
+
| 0.7543 | 7.0 | 560 | 0.7568 | 0.6475 |
|
62 |
+
| 0.6993 | 8.0 | 640 | 0.8830 | 0.6133 |
|
63 |
+
| 0.7045 | 9.0 | 720 | 0.7373 | 0.6582 |
|
64 |
+
| 0.6557 | 10.0 | 800 | 0.6076 | 0.7451 |
|
65 |
+
| 0.5876 | 11.0 | 880 | 0.7281 | 0.6992 |
|
66 |
+
| 0.5732 | 12.0 | 960 | 0.5769 | 0.7510 |
|
67 |
+
| 0.4864 | 13.0 | 1040 | 0.4457 | 0.8311 |
|
68 |
+
| 0.5175 | 14.0 | 1120 | 0.5278 | 0.7842 |
|
69 |
+
| 0.4865 | 15.0 | 1200 | 0.4164 | 0.8379 |
|
70 |
+
| 0.4049 | 16.0 | 1280 | 0.4204 | 0.8301 |
|
71 |
+
| 0.4167 | 17.0 | 1360 | 0.4720 | 0.8281 |
|
72 |
+
| 0.36 | 18.0 | 1440 | 0.4660 | 0.8164 |
|
73 |
+
| 0.3195 | 19.0 | 1520 | 0.3064 | 0.8770 |
|
74 |
+
| 0.3652 | 20.0 | 1600 | 0.2571 | 0.9121 |
|
75 |
+
| 0.2794 | 21.0 | 1680 | 0.2450 | 0.9150 |
|
76 |
+
| 0.2704 | 22.0 | 1760 | 0.2391 | 0.9033 |
|
77 |
+
| 0.2612 | 23.0 | 1840 | 0.2352 | 0.9277 |
|
78 |
+
| 0.2425 | 24.0 | 1920 | 0.4720 | 0.8281 |
|
79 |
+
| 0.2567 | 25.0 | 2000 | 0.2296 | 0.9131 |
|
80 |
+
| 0.2302 | 26.0 | 2080 | 0.3067 | 0.8945 |
|
81 |
+
| 0.2358 | 27.0 | 2160 | 0.1776 | 0.9375 |
|
82 |
+
| 0.2173 | 28.0 | 2240 | 0.1596 | 0.9492 |
|
83 |
+
| 0.1798 | 29.0 | 2320 | 0.1548 | 0.9414 |
|
84 |
+
| 0.197 | 30.0 | 2400 | 0.1740 | 0.9570 |
|
85 |
+
| 0.1654 | 31.0 | 2480 | 0.1217 | 0.9668 |
|
86 |
+
| 0.1896 | 32.0 | 2560 | 0.2552 | 0.9258 |
|
87 |
+
| 0.1705 | 33.0 | 2640 | 0.1031 | 0.9727 |
|
88 |
+
| 0.1689 | 34.0 | 2720 | 0.1011 | 0.9688 |
|
89 |
+
| 0.1439 | 35.0 | 2800 | 0.1175 | 0.9648 |
|
90 |
+
| 0.1606 | 36.0 | 2880 | 0.1805 | 0.9443 |
|
91 |
+
| 0.1281 | 37.0 | 2960 | 0.1254 | 0.9678 |
|
92 |
+
| 0.1518 | 38.0 | 3040 | 0.1184 | 0.9648 |
|
93 |
+
| 0.1531 | 39.0 | 3120 | 0.0992 | 0.9736 |
|
94 |
+
| 0.132 | 40.0 | 3200 | 0.0920 | 0.9775 |
|
95 |
+
| 0.134 | 41.0 | 3280 | 0.1391 | 0.9639 |
|
96 |
+
| 0.1413 | 42.0 | 3360 | 0.1122 | 0.9717 |
|
97 |
+
| 0.1097 | 43.0 | 3440 | 0.1171 | 0.9678 |
|
98 |
+
| 0.1167 | 44.0 | 3520 | 0.1054 | 0.9766 |
|
99 |
+
| 0.1388 | 45.0 | 3600 | 0.0932 | 0.9795 |
|
100 |
+
| 0.1221 | 46.0 | 3680 | 0.0946 | 0.9766 |
|
101 |
+
| 0.1099 | 47.0 | 3760 | 0.1116 | 0.9756 |
|
102 |
+
| 0.1041 | 48.0 | 3840 | 0.1126 | 0.9746 |
|
103 |
+
| 0.1025 | 49.0 | 3920 | 0.1114 | 0.9756 |
|
104 |
+
| 0.0887 | 50.0 | 4000 | 0.1056 | 0.9756 |
|
105 |
+
|
106 |
+
|
107 |
+
### Framework versions
|
108 |
+
|
109 |
+
- Transformers 4.38.2
|
110 |
+
- Pytorch 2.1.0+cu121
|
111 |
+
- Datasets 2.18.0
|
112 |
+
- Tokenizers 0.15.2
|
all_results.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"eval_accuracy": 0.9794921875,
|
4 |
+
"eval_loss": 0.09323666244745255,
|
5 |
+
"eval_runtime": 4.9584,
|
6 |
+
"eval_samples_per_second": 206.517,
|
7 |
+
"eval_steps_per_second": 12.907,
|
8 |
+
"total_flos": 5.437210780237824e+18,
|
9 |
+
"train_loss": 0.3629864407479763,
|
10 |
+
"train_runtime": 3465.6999,
|
11 |
+
"train_samples_per_second": 73.867,
|
12 |
+
"train_steps_per_second": 1.154
|
13 |
+
}
|
config.json
ADDED
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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_bottleneck": false,
|
13 |
+
"downsample_in_first_stage": false,
|
14 |
+
"embedding_size": 64,
|
15 |
+
"hidden_act": "relu",
|
16 |
+
"hidden_sizes": [
|
17 |
+
256,
|
18 |
+
512,
|
19 |
+
1024,
|
20 |
+
2048
|
21 |
+
],
|
22 |
+
"id2label": {
|
23 |
+
"0": "Mild_Demented",
|
24 |
+
"1": "Moderate_Demented",
|
25 |
+
"2": "Non_Demented",
|
26 |
+
"3": "Very_Mild_Demented"
|
27 |
+
},
|
28 |
+
"label2id": {
|
29 |
+
"Mild_Demented": "0",
|
30 |
+
"Moderate_Demented": "1",
|
31 |
+
"Non_Demented": "2",
|
32 |
+
"Very_Mild_Demented": "3"
|
33 |
+
},
|
34 |
+
"layer_type": "bottleneck",
|
35 |
+
"model_type": "resnet",
|
36 |
+
"num_channels": 3,
|
37 |
+
"out_features": [
|
38 |
+
"stage4"
|
39 |
+
],
|
40 |
+
"out_indices": [
|
41 |
+
4
|
42 |
+
],
|
43 |
+
"problem_type": "single_label_classification",
|
44 |
+
"stage_names": [
|
45 |
+
"stem",
|
46 |
+
"stage1",
|
47 |
+
"stage2",
|
48 |
+
"stage3",
|
49 |
+
"stage4"
|
50 |
+
],
|
51 |
+
"torch_dtype": "float32",
|
52 |
+
"transformers_version": "4.38.2"
|
53 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"eval_accuracy": 0.9794921875,
|
4 |
+
"eval_loss": 0.09323666244745255,
|
5 |
+
"eval_runtime": 4.9584,
|
6 |
+
"eval_samples_per_second": 206.517,
|
7 |
+
"eval_steps_per_second": 12.907
|
8 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:029a8ae5f126fd073b1ab5d8557516e85c13d2f89151fe30dd1b908285b38fa0
|
3 |
+
size 94319344
|
preprocessor_config.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"crop_pct": 0.875,
|
3 |
+
"do_normalize": true,
|
4 |
+
"do_rescale": true,
|
5 |
+
"do_resize": true,
|
6 |
+
"image_mean": [
|
7 |
+
0.485,
|
8 |
+
0.456,
|
9 |
+
0.406
|
10 |
+
],
|
11 |
+
"image_processor_type": "ConvNextImageProcessor",
|
12 |
+
"image_std": [
|
13 |
+
0.229,
|
14 |
+
0.224,
|
15 |
+
0.225
|
16 |
+
],
|
17 |
+
"resample": 3,
|
18 |
+
"rescale_factor": 0.00392156862745098,
|
19 |
+
"size": {
|
20 |
+
"shortest_edge": 224
|
21 |
+
}
|
22 |
+
}
|
runs/Mar07_03-18-48_606d6922676f/events.out.tfevents.1709781530.606d6922676f.1016.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5790612438d06e2ed0cceaee3ef25c441d28e09f2b9c07c8d451e313069f40b7
|
3 |
+
size 105676
|
runs/Mar07_03-18-48_606d6922676f/events.out.tfevents.1709785268.606d6922676f.1016.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cc2935e6bd8a6623dced1d3dbbb6b041793ca5563cab7ba253ded482f2c1c2cc
|
3 |
+
size 411
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 50.0,
|
3 |
+
"total_flos": 5.437210780237824e+18,
|
4 |
+
"train_loss": 0.3629864407479763,
|
5 |
+
"train_runtime": 3465.6999,
|
6 |
+
"train_samples_per_second": 73.867,
|
7 |
+
"train_steps_per_second": 1.154
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,3280 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.9794921875,
|
3 |
+
"best_model_checkpoint": "resnet-Alzheimer/checkpoint-3600",
|
4 |
+
"epoch": 50.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 4000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.12,
|
13 |
+
"grad_norm": 2.1754038333892822,
|
14 |
+
"learning_rate": 5e-05,
|
15 |
+
"loss": 1.3812,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.25,
|
20 |
+
"grad_norm": 3.91097092628479,
|
21 |
+
"learning_rate": 0.0001,
|
22 |
+
"loss": 1.34,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.38,
|
27 |
+
"grad_norm": 4.134509563446045,
|
28 |
+
"learning_rate": 0.00015,
|
29 |
+
"loss": 1.2635,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.5,
|
34 |
+
"grad_norm": 6.377187252044678,
|
35 |
+
"learning_rate": 0.0002,
|
36 |
+
"loss": 1.1824,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.62,
|
41 |
+
"grad_norm": 7.713193416595459,
|
42 |
+
"learning_rate": 0.00025,
|
43 |
+
"loss": 1.1226,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.75,
|
48 |
+
"grad_norm": 3.569382429122925,
|
49 |
+
"learning_rate": 0.0003,
|
50 |
+
"loss": 1.0308,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.88,
|
55 |
+
"grad_norm": 5.650737285614014,
|
56 |
+
"learning_rate": 0.00035,
|
57 |
+
"loss": 1.0115,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 1.0,
|
62 |
+
"grad_norm": 5.970870494842529,
|
63 |
+
"learning_rate": 0.0004,
|
64 |
+
"loss": 1.0127,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 1.0,
|
69 |
+
"eval_accuracy": 0.5087890625,
|
70 |
+
"eval_loss": 0.9888483285903931,
|
71 |
+
"eval_runtime": 5.4153,
|
72 |
+
"eval_samples_per_second": 189.093,
|
73 |
+
"eval_steps_per_second": 11.818,
|
74 |
+
"step": 80
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 1.12,
|
78 |
+
"grad_norm": 2.92154860496521,
|
79 |
+
"learning_rate": 0.00045000000000000004,
|
80 |
+
"loss": 1.0175,
|
81 |
+
"step": 90
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 1.25,
|
85 |
+
"grad_norm": 4.131512641906738,
|
86 |
+
"learning_rate": 0.0005,
|
87 |
+
"loss": 0.918,
|
88 |
+
"step": 100
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 1.38,
|
92 |
+
"grad_norm": 2.9838943481445312,
|
93 |
+
"learning_rate": 0.00055,
|
94 |
+
"loss": 0.9613,
|
95 |
+
"step": 110
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"epoch": 1.5,
|
99 |
+
"grad_norm": 2.4230854511260986,
|
100 |
+
"learning_rate": 0.0006,
|
101 |
+
"loss": 0.9188,
|
102 |
+
"step": 120
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"epoch": 1.62,
|
106 |
+
"grad_norm": 3.774385690689087,
|
107 |
+
"learning_rate": 0.0006500000000000001,
|
108 |
+
"loss": 0.9245,
|
109 |
+
"step": 130
|
110 |
+
},
|
111 |
+
{
|
112 |
+
"epoch": 1.75,
|
113 |
+
"grad_norm": 3.175440549850464,
|
114 |
+
"learning_rate": 0.0007,
|
115 |
+
"loss": 0.8713,
|
116 |
+
"step": 140
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 1.88,
|
120 |
+
"grad_norm": 2.7985305786132812,
|
121 |
+
"learning_rate": 0.00075,
|
122 |
+
"loss": 0.9264,
|
123 |
+
"step": 150
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 2.0,
|
127 |
+
"grad_norm": 1.5870747566223145,
|
128 |
+
"learning_rate": 0.0008,
|
129 |
+
"loss": 0.9345,
|
130 |
+
"step": 160
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"epoch": 2.0,
|
134 |
+
"eval_accuracy": 0.5302734375,
|
135 |
+
"eval_loss": 0.942151665687561,
|
136 |
+
"eval_runtime": 5.1954,
|
137 |
+
"eval_samples_per_second": 197.099,
|
138 |
+
"eval_steps_per_second": 12.319,
|
139 |
+
"step": 160
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"epoch": 2.12,
|
143 |
+
"grad_norm": 1.3990691900253296,
|
144 |
+
"learning_rate": 0.00085,
|
145 |
+
"loss": 0.8851,
|
146 |
+
"step": 170
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 2.25,
|
150 |
+
"grad_norm": 1.4066977500915527,
|
151 |
+
"learning_rate": 0.0009000000000000001,
|
152 |
+
"loss": 0.9081,
|
153 |
+
"step": 180
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 2.38,
|
157 |
+
"grad_norm": 1.5531185865402222,
|
158 |
+
"learning_rate": 0.00095,
|
159 |
+
"loss": 0.8888,
|
160 |
+
"step": 190
|
161 |
+
},
|
162 |
+
{
|
163 |
+
"epoch": 2.5,
|
164 |
+
"grad_norm": 1.960204839706421,
|
165 |
+
"learning_rate": 0.001,
|
166 |
+
"loss": 0.9206,
|
167 |
+
"step": 200
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 2.62,
|
171 |
+
"grad_norm": 1.0218795537948608,
|
172 |
+
"learning_rate": 0.0010500000000000002,
|
173 |
+
"loss": 0.8735,
|
174 |
+
"step": 210
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 2.75,
|
178 |
+
"grad_norm": 1.9216176271438599,
|
179 |
+
"learning_rate": 0.0011,
|
180 |
+
"loss": 0.924,
|
181 |
+
"step": 220
|
182 |
+
},
|
183 |
+
{
|
184 |
+
"epoch": 2.88,
|
185 |
+
"grad_norm": 1.9017541408538818,
|
186 |
+
"learning_rate": 0.00115,
|
187 |
+
"loss": 0.9327,
|
188 |
+
"step": 230
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 3.0,
|
192 |
+
"grad_norm": 1.556686282157898,
|
193 |
+
"learning_rate": 0.0012,
|
194 |
+
"loss": 0.8889,
|
195 |
+
"step": 240
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 3.0,
|
199 |
+
"eval_accuracy": 0.578125,
|
200 |
+
"eval_loss": 0.8723889589309692,
|
201 |
+
"eval_runtime": 4.9225,
|
202 |
+
"eval_samples_per_second": 208.026,
|
203 |
+
"eval_steps_per_second": 13.002,
|
204 |
+
"step": 240
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 3.12,
|
208 |
+
"grad_norm": 0.9459726810455322,
|
209 |
+
"learning_rate": 0.00125,
|
210 |
+
"loss": 0.8643,
|
211 |
+
"step": 250
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"epoch": 3.25,
|
215 |
+
"grad_norm": 0.749912679195404,
|
216 |
+
"learning_rate": 0.0013000000000000002,
|
217 |
+
"loss": 0.8978,
|
218 |
+
"step": 260
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 3.38,
|
222 |
+
"grad_norm": 0.9228126406669617,
|
223 |
+
"learning_rate": 0.00135,
|
224 |
+
"loss": 0.8838,
|
225 |
+
"step": 270
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 3.5,
|
229 |
+
"grad_norm": 1.0743939876556396,
|
230 |
+
"learning_rate": 0.0014,
|
231 |
+
"loss": 0.8868,
|
232 |
+
"step": 280
|
233 |
+
},
|
234 |
+
{
|
235 |
+
"epoch": 3.62,
|
236 |
+
"grad_norm": 0.997053325176239,
|
237 |
+
"learning_rate": 0.00145,
|
238 |
+
"loss": 0.8632,
|
239 |
+
"step": 290
|
240 |
+
},
|
241 |
+
{
|
242 |
+
"epoch": 3.75,
|
243 |
+
"grad_norm": 0.5891302227973938,
|
244 |
+
"learning_rate": 0.0015,
|
245 |
+
"loss": 0.8501,
|
246 |
+
"step": 300
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 3.88,
|
250 |
+
"grad_norm": 1.0819345712661743,
|
251 |
+
"learning_rate": 0.0015500000000000002,
|
252 |
+
"loss": 0.884,
|
253 |
+
"step": 310
|
254 |
+
},
|
255 |
+
{
|
256 |
+
"epoch": 4.0,
|
257 |
+
"grad_norm": 0.5622245073318481,
|
258 |
+
"learning_rate": 0.0016,
|
259 |
+
"loss": 0.8843,
|
260 |
+
"step": 320
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 4.0,
|
264 |
+
"eval_accuracy": 0.5888671875,
|
265 |
+
"eval_loss": 0.8535706996917725,
|
266 |
+
"eval_runtime": 6.5338,
|
267 |
+
"eval_samples_per_second": 156.725,
|
268 |
+
"eval_steps_per_second": 9.795,
|
269 |
+
"step": 320
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 4.12,
|
273 |
+
"grad_norm": 0.7301604747772217,
|
274 |
+
"learning_rate": 0.00165,
|
275 |
+
"loss": 0.8241,
|
276 |
+
"step": 330
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 4.25,
|
280 |
+
"grad_norm": 1.222732424736023,
|
281 |
+
"learning_rate": 0.0017,
|
282 |
+
"loss": 0.8929,
|
283 |
+
"step": 340
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 4.38,
|
287 |
+
"grad_norm": 0.8520879745483398,
|
288 |
+
"learning_rate": 0.00175,
|
289 |
+
"loss": 0.8726,
|
290 |
+
"step": 350
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 4.5,
|
294 |
+
"grad_norm": 0.6151734590530396,
|
295 |
+
"learning_rate": 0.0018000000000000002,
|
296 |
+
"loss": 0.8851,
|
297 |
+
"step": 360
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 4.62,
|
301 |
+
"grad_norm": 0.6786526441574097,
|
302 |
+
"learning_rate": 0.00185,
|
303 |
+
"loss": 0.8544,
|
304 |
+
"step": 370
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 4.75,
|
308 |
+
"grad_norm": 0.8025469779968262,
|
309 |
+
"learning_rate": 0.0019,
|
310 |
+
"loss": 0.8432,
|
311 |
+
"step": 380
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 4.88,
|
315 |
+
"grad_norm": 1.0158729553222656,
|
316 |
+
"learning_rate": 0.00195,
|
317 |
+
"loss": 0.8261,
|
318 |
+
"step": 390
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 5.0,
|
322 |
+
"grad_norm": 0.7815405130386353,
|
323 |
+
"learning_rate": 0.002,
|
324 |
+
"loss": 0.8397,
|
325 |
+
"step": 400
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 5.0,
|
329 |
+
"eval_accuracy": 0.615234375,
|
330 |
+
"eval_loss": 0.8353910446166992,
|
331 |
+
"eval_runtime": 5.3022,
|
332 |
+
"eval_samples_per_second": 193.129,
|
333 |
+
"eval_steps_per_second": 12.071,
|
334 |
+
"step": 400
|
335 |
+
},
|
336 |
+
{
|
337 |
+
"epoch": 5.12,
|
338 |
+
"grad_norm": 0.9535025954246521,
|
339 |
+
"learning_rate": 0.0019944444444444445,
|
340 |
+
"loss": 0.8636,
|
341 |
+
"step": 410
|
342 |
+
},
|
343 |
+
{
|
344 |
+
"epoch": 5.25,
|
345 |
+
"grad_norm": 0.43990448117256165,
|
346 |
+
"learning_rate": 0.001988888888888889,
|
347 |
+
"loss": 0.7931,
|
348 |
+
"step": 420
|
349 |
+
},
|
350 |
+
{
|
351 |
+
"epoch": 5.38,
|
352 |
+
"grad_norm": 0.6062633395195007,
|
353 |
+
"learning_rate": 0.0019833333333333335,
|
354 |
+
"loss": 0.8345,
|
355 |
+
"step": 430
|
356 |
+
},
|
357 |
+
{
|
358 |
+
"epoch": 5.5,
|
359 |
+
"grad_norm": 0.6349042057991028,
|
360 |
+
"learning_rate": 0.001977777777777778,
|
361 |
+
"loss": 0.8593,
|
362 |
+
"step": 440
|
363 |
+
},
|
364 |
+
{
|
365 |
+
"epoch": 5.62,
|
366 |
+
"grad_norm": 0.6786915063858032,
|
367 |
+
"learning_rate": 0.0019722222222222224,
|
368 |
+
"loss": 0.8435,
|
369 |
+
"step": 450
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 5.75,
|
373 |
+
"grad_norm": 0.7090786695480347,
|
374 |
+
"learning_rate": 0.0019666666666666665,
|
375 |
+
"loss": 0.8008,
|
376 |
+
"step": 460
|
377 |
+
},
|
378 |
+
{
|
379 |
+
"epoch": 5.88,
|
380 |
+
"grad_norm": 0.6538481712341309,
|
381 |
+
"learning_rate": 0.001961111111111111,
|
382 |
+
"loss": 0.763,
|
383 |
+
"step": 470
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 6.0,
|
387 |
+
"grad_norm": 0.4316461682319641,
|
388 |
+
"learning_rate": 0.0019555555555555554,
|
389 |
+
"loss": 0.8624,
|
390 |
+
"step": 480
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 6.0,
|
394 |
+
"eval_accuracy": 0.5380859375,
|
395 |
+
"eval_loss": 0.9221189022064209,
|
396 |
+
"eval_runtime": 4.9173,
|
397 |
+
"eval_samples_per_second": 208.245,
|
398 |
+
"eval_steps_per_second": 13.015,
|
399 |
+
"step": 480
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 6.12,
|
403 |
+
"grad_norm": 0.5696819424629211,
|
404 |
+
"learning_rate": 0.00195,
|
405 |
+
"loss": 0.8065,
|
406 |
+
"step": 490
|
407 |
+
},
|
408 |
+
{
|
409 |
+
"epoch": 6.25,
|
410 |
+
"grad_norm": 0.6260728240013123,
|
411 |
+
"learning_rate": 0.0019444444444444444,
|
412 |
+
"loss": 0.7873,
|
413 |
+
"step": 500
|
414 |
+
},
|
415 |
+
{
|
416 |
+
"epoch": 6.38,
|
417 |
+
"grad_norm": 0.6295855045318604,
|
418 |
+
"learning_rate": 0.0019388888888888889,
|
419 |
+
"loss": 0.7802,
|
420 |
+
"step": 510
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"epoch": 6.5,
|
424 |
+
"grad_norm": 0.6074417233467102,
|
425 |
+
"learning_rate": 0.0019333333333333333,
|
426 |
+
"loss": 0.7907,
|
427 |
+
"step": 520
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"epoch": 6.62,
|
431 |
+
"grad_norm": 0.6099679470062256,
|
432 |
+
"learning_rate": 0.0019277777777777778,
|
433 |
+
"loss": 0.7391,
|
434 |
+
"step": 530
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 6.75,
|
438 |
+
"grad_norm": 0.9349565505981445,
|
439 |
+
"learning_rate": 0.0019222222222222223,
|
440 |
+
"loss": 0.7749,
|
441 |
+
"step": 540
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 6.88,
|
445 |
+
"grad_norm": 0.6923946142196655,
|
446 |
+
"learning_rate": 0.0019166666666666668,
|
447 |
+
"loss": 0.8232,
|
448 |
+
"step": 550
|
449 |
+
},
|
450 |
+
{
|
451 |
+
"epoch": 7.0,
|
452 |
+
"grad_norm": 0.5967056751251221,
|
453 |
+
"learning_rate": 0.0019111111111111113,
|
454 |
+
"loss": 0.7543,
|
455 |
+
"step": 560
|
456 |
+
},
|
457 |
+
{
|
458 |
+
"epoch": 7.0,
|
459 |
+
"eval_accuracy": 0.6474609375,
|
460 |
+
"eval_loss": 0.7568103671073914,
|
461 |
+
"eval_runtime": 5.3014,
|
462 |
+
"eval_samples_per_second": 193.156,
|
463 |
+
"eval_steps_per_second": 12.072,
|
464 |
+
"step": 560
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 7.12,
|
468 |
+
"grad_norm": 0.7925052642822266,
|
469 |
+
"learning_rate": 0.0019055555555555555,
|
470 |
+
"loss": 0.7086,
|
471 |
+
"step": 570
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 7.25,
|
475 |
+
"grad_norm": 0.715761125087738,
|
476 |
+
"learning_rate": 0.0019,
|
477 |
+
"loss": 0.7901,
|
478 |
+
"step": 580
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 7.38,
|
482 |
+
"grad_norm": 0.6602711081504822,
|
483 |
+
"learning_rate": 0.0018944444444444445,
|
484 |
+
"loss": 0.7375,
|
485 |
+
"step": 590
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 7.5,
|
489 |
+
"grad_norm": 0.5104066729545593,
|
490 |
+
"learning_rate": 0.001888888888888889,
|
491 |
+
"loss": 0.7805,
|
492 |
+
"step": 600
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 7.62,
|
496 |
+
"grad_norm": 0.6333702802658081,
|
497 |
+
"learning_rate": 0.0018833333333333334,
|
498 |
+
"loss": 0.7017,
|
499 |
+
"step": 610
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 7.75,
|
503 |
+
"grad_norm": 0.5703239440917969,
|
504 |
+
"learning_rate": 0.001877777777777778,
|
505 |
+
"loss": 0.7086,
|
506 |
+
"step": 620
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 7.88,
|
510 |
+
"grad_norm": 0.8939486742019653,
|
511 |
+
"learning_rate": 0.0018722222222222222,
|
512 |
+
"loss": 0.7399,
|
513 |
+
"step": 630
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 8.0,
|
517 |
+
"grad_norm": 0.6808524131774902,
|
518 |
+
"learning_rate": 0.0018666666666666666,
|
519 |
+
"loss": 0.6993,
|
520 |
+
"step": 640
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 8.0,
|
524 |
+
"eval_accuracy": 0.61328125,
|
525 |
+
"eval_loss": 0.8830391764640808,
|
526 |
+
"eval_runtime": 4.9073,
|
527 |
+
"eval_samples_per_second": 208.667,
|
528 |
+
"eval_steps_per_second": 13.042,
|
529 |
+
"step": 640
|
530 |
+
},
|
531 |
+
{
|
532 |
+
"epoch": 8.12,
|
533 |
+
"grad_norm": 0.7670312523841858,
|
534 |
+
"learning_rate": 0.0018611111111111111,
|
535 |
+
"loss": 0.7304,
|
536 |
+
"step": 650
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 8.25,
|
540 |
+
"grad_norm": 0.518883466720581,
|
541 |
+
"learning_rate": 0.0018555555555555556,
|
542 |
+
"loss": 0.6759,
|
543 |
+
"step": 660
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 8.38,
|
547 |
+
"grad_norm": 0.6331384778022766,
|
548 |
+
"learning_rate": 0.00185,
|
549 |
+
"loss": 0.7323,
|
550 |
+
"step": 670
|
551 |
+
},
|
552 |
+
{
|
553 |
+
"epoch": 8.5,
|
554 |
+
"grad_norm": 0.5934571027755737,
|
555 |
+
"learning_rate": 0.0018444444444444446,
|
556 |
+
"loss": 0.7109,
|
557 |
+
"step": 680
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 8.62,
|
561 |
+
"grad_norm": 0.5555841326713562,
|
562 |
+
"learning_rate": 0.0018388888888888888,
|
563 |
+
"loss": 0.7361,
|
564 |
+
"step": 690
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 8.75,
|
568 |
+
"grad_norm": 0.45028582215309143,
|
569 |
+
"learning_rate": 0.0018333333333333333,
|
570 |
+
"loss": 0.7209,
|
571 |
+
"step": 700
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 8.88,
|
575 |
+
"grad_norm": 0.4313984811306,
|
576 |
+
"learning_rate": 0.0018277777777777778,
|
577 |
+
"loss": 0.692,
|
578 |
+
"step": 710
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 9.0,
|
582 |
+
"grad_norm": 0.6221916675567627,
|
583 |
+
"learning_rate": 0.0018222222222222223,
|
584 |
+
"loss": 0.7045,
|
585 |
+
"step": 720
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 9.0,
|
589 |
+
"eval_accuracy": 0.658203125,
|
590 |
+
"eval_loss": 0.7372878789901733,
|
591 |
+
"eval_runtime": 4.9218,
|
592 |
+
"eval_samples_per_second": 208.053,
|
593 |
+
"eval_steps_per_second": 13.003,
|
594 |
+
"step": 720
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"epoch": 9.12,
|
598 |
+
"grad_norm": 0.9794626832008362,
|
599 |
+
"learning_rate": 0.0018166666666666667,
|
600 |
+
"loss": 0.641,
|
601 |
+
"step": 730
|
602 |
+
},
|
603 |
+
{
|
604 |
+
"epoch": 9.25,
|
605 |
+
"grad_norm": 0.8530990481376648,
|
606 |
+
"learning_rate": 0.0018111111111111112,
|
607 |
+
"loss": 0.6616,
|
608 |
+
"step": 740
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 9.38,
|
612 |
+
"grad_norm": 0.5696712136268616,
|
613 |
+
"learning_rate": 0.0018055555555555557,
|
614 |
+
"loss": 0.6685,
|
615 |
+
"step": 750
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 9.5,
|
619 |
+
"grad_norm": 0.6695945858955383,
|
620 |
+
"learning_rate": 0.0018000000000000002,
|
621 |
+
"loss": 0.6079,
|
622 |
+
"step": 760
|
623 |
+
},
|
624 |
+
{
|
625 |
+
"epoch": 9.62,
|
626 |
+
"grad_norm": 0.9470874667167664,
|
627 |
+
"learning_rate": 0.0017944444444444446,
|
628 |
+
"loss": 0.6362,
|
629 |
+
"step": 770
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 9.75,
|
633 |
+
"grad_norm": 1.0435755252838135,
|
634 |
+
"learning_rate": 0.001788888888888889,
|
635 |
+
"loss": 0.7036,
|
636 |
+
"step": 780
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"epoch": 9.88,
|
640 |
+
"grad_norm": 0.4934737980365753,
|
641 |
+
"learning_rate": 0.0017833333333333334,
|
642 |
+
"loss": 0.6955,
|
643 |
+
"step": 790
|
644 |
+
},
|
645 |
+
{
|
646 |
+
"epoch": 10.0,
|
647 |
+
"grad_norm": 0.49625110626220703,
|
648 |
+
"learning_rate": 0.0017777777777777776,
|
649 |
+
"loss": 0.6557,
|
650 |
+
"step": 800
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 10.0,
|
654 |
+
"eval_accuracy": 0.7451171875,
|
655 |
+
"eval_loss": 0.6075544357299805,
|
656 |
+
"eval_runtime": 5.2806,
|
657 |
+
"eval_samples_per_second": 193.918,
|
658 |
+
"eval_steps_per_second": 12.12,
|
659 |
+
"step": 800
|
660 |
+
},
|
661 |
+
{
|
662 |
+
"epoch": 10.12,
|
663 |
+
"grad_norm": 0.49739229679107666,
|
664 |
+
"learning_rate": 0.0017722222222222221,
|
665 |
+
"loss": 0.605,
|
666 |
+
"step": 810
|
667 |
+
},
|
668 |
+
{
|
669 |
+
"epoch": 10.25,
|
670 |
+
"grad_norm": 0.6317277550697327,
|
671 |
+
"learning_rate": 0.0017666666666666666,
|
672 |
+
"loss": 0.5332,
|
673 |
+
"step": 820
|
674 |
+
},
|
675 |
+
{
|
676 |
+
"epoch": 10.38,
|
677 |
+
"grad_norm": 0.756879985332489,
|
678 |
+
"learning_rate": 0.001761111111111111,
|
679 |
+
"loss": 0.5619,
|
680 |
+
"step": 830
|
681 |
+
},
|
682 |
+
{
|
683 |
+
"epoch": 10.5,
|
684 |
+
"grad_norm": 0.6143298149108887,
|
685 |
+
"learning_rate": 0.0017555555555555556,
|
686 |
+
"loss": 0.601,
|
687 |
+
"step": 840
|
688 |
+
},
|
689 |
+
{
|
690 |
+
"epoch": 10.62,
|
691 |
+
"grad_norm": 0.7249147891998291,
|
692 |
+
"learning_rate": 0.00175,
|
693 |
+
"loss": 0.5935,
|
694 |
+
"step": 850
|
695 |
+
},
|
696 |
+
{
|
697 |
+
"epoch": 10.75,
|
698 |
+
"grad_norm": 0.4532654285430908,
|
699 |
+
"learning_rate": 0.0017444444444444445,
|
700 |
+
"loss": 0.5988,
|
701 |
+
"step": 860
|
702 |
+
},
|
703 |
+
{
|
704 |
+
"epoch": 10.88,
|
705 |
+
"grad_norm": 0.5738415718078613,
|
706 |
+
"learning_rate": 0.001738888888888889,
|
707 |
+
"loss": 0.6634,
|
708 |
+
"step": 870
|
709 |
+
},
|
710 |
+
{
|
711 |
+
"epoch": 11.0,
|
712 |
+
"grad_norm": 0.5514868497848511,
|
713 |
+
"learning_rate": 0.0017333333333333335,
|
714 |
+
"loss": 0.5876,
|
715 |
+
"step": 880
|
716 |
+
},
|
717 |
+
{
|
718 |
+
"epoch": 11.0,
|
719 |
+
"eval_accuracy": 0.69921875,
|
720 |
+
"eval_loss": 0.7281272411346436,
|
721 |
+
"eval_runtime": 4.8994,
|
722 |
+
"eval_samples_per_second": 209.004,
|
723 |
+
"eval_steps_per_second": 13.063,
|
724 |
+
"step": 880
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 11.12,
|
728 |
+
"grad_norm": 0.7158863544464111,
|
729 |
+
"learning_rate": 0.001727777777777778,
|
730 |
+
"loss": 0.606,
|
731 |
+
"step": 890
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 11.25,
|
735 |
+
"grad_norm": 0.7355363368988037,
|
736 |
+
"learning_rate": 0.0017222222222222224,
|
737 |
+
"loss": 0.5923,
|
738 |
+
"step": 900
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 11.38,
|
742 |
+
"grad_norm": 0.7794367671012878,
|
743 |
+
"learning_rate": 0.0017166666666666667,
|
744 |
+
"loss": 0.5935,
|
745 |
+
"step": 910
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 11.5,
|
749 |
+
"grad_norm": 0.9755826592445374,
|
750 |
+
"learning_rate": 0.0017111111111111112,
|
751 |
+
"loss": 0.644,
|
752 |
+
"step": 920
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 11.62,
|
756 |
+
"grad_norm": 0.6257722973823547,
|
757 |
+
"learning_rate": 0.0017055555555555554,
|
758 |
+
"loss": 0.617,
|
759 |
+
"step": 930
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 11.75,
|
763 |
+
"grad_norm": 0.8550503253936768,
|
764 |
+
"learning_rate": 0.0017,
|
765 |
+
"loss": 0.5854,
|
766 |
+
"step": 940
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 11.88,
|
770 |
+
"grad_norm": 0.7347137928009033,
|
771 |
+
"learning_rate": 0.0016944444444444444,
|
772 |
+
"loss": 0.6358,
|
773 |
+
"step": 950
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 12.0,
|
777 |
+
"grad_norm": 0.7867416739463806,
|
778 |
+
"learning_rate": 0.0016888888888888889,
|
779 |
+
"loss": 0.5732,
|
780 |
+
"step": 960
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 12.0,
|
784 |
+
"eval_accuracy": 0.7509765625,
|
785 |
+
"eval_loss": 0.5769097208976746,
|
786 |
+
"eval_runtime": 5.275,
|
787 |
+
"eval_samples_per_second": 194.122,
|
788 |
+
"eval_steps_per_second": 12.133,
|
789 |
+
"step": 960
|
790 |
+
},
|
791 |
+
{
|
792 |
+
"epoch": 12.12,
|
793 |
+
"grad_norm": 0.6022630333900452,
|
794 |
+
"learning_rate": 0.0016833333333333333,
|
795 |
+
"loss": 0.5643,
|
796 |
+
"step": 970
|
797 |
+
},
|
798 |
+
{
|
799 |
+
"epoch": 12.25,
|
800 |
+
"grad_norm": 0.599958062171936,
|
801 |
+
"learning_rate": 0.0016777777777777778,
|
802 |
+
"loss": 0.5438,
|
803 |
+
"step": 980
|
804 |
+
},
|
805 |
+
{
|
806 |
+
"epoch": 12.38,
|
807 |
+
"grad_norm": 0.6484814286231995,
|
808 |
+
"learning_rate": 0.0016722222222222223,
|
809 |
+
"loss": 0.5208,
|
810 |
+
"step": 990
|
811 |
+
},
|
812 |
+
{
|
813 |
+
"epoch": 12.5,
|
814 |
+
"grad_norm": 0.8167735934257507,
|
815 |
+
"learning_rate": 0.0016666666666666668,
|
816 |
+
"loss": 0.5369,
|
817 |
+
"step": 1000
|
818 |
+
},
|
819 |
+
{
|
820 |
+
"epoch": 12.62,
|
821 |
+
"grad_norm": 0.49088793992996216,
|
822 |
+
"learning_rate": 0.0016611111111111113,
|
823 |
+
"loss": 0.4803,
|
824 |
+
"step": 1010
|
825 |
+
},
|
826 |
+
{
|
827 |
+
"epoch": 12.75,
|
828 |
+
"grad_norm": 0.6817615628242493,
|
829 |
+
"learning_rate": 0.0016555555555555555,
|
830 |
+
"loss": 0.5102,
|
831 |
+
"step": 1020
|
832 |
+
},
|
833 |
+
{
|
834 |
+
"epoch": 12.88,
|
835 |
+
"grad_norm": 0.8656439781188965,
|
836 |
+
"learning_rate": 0.00165,
|
837 |
+
"loss": 0.5287,
|
838 |
+
"step": 1030
|
839 |
+
},
|
840 |
+
{
|
841 |
+
"epoch": 13.0,
|
842 |
+
"grad_norm": 0.5195401310920715,
|
843 |
+
"learning_rate": 0.0016444444444444445,
|
844 |
+
"loss": 0.4864,
|
845 |
+
"step": 1040
|
846 |
+
},
|
847 |
+
{
|
848 |
+
"epoch": 13.0,
|
849 |
+
"eval_accuracy": 0.8310546875,
|
850 |
+
"eval_loss": 0.445728600025177,
|
851 |
+
"eval_runtime": 4.9054,
|
852 |
+
"eval_samples_per_second": 208.751,
|
853 |
+
"eval_steps_per_second": 13.047,
|
854 |
+
"step": 1040
|
855 |
+
},
|
856 |
+
{
|
857 |
+
"epoch": 13.12,
|
858 |
+
"grad_norm": 0.7564366459846497,
|
859 |
+
"learning_rate": 0.001638888888888889,
|
860 |
+
"loss": 0.4715,
|
861 |
+
"step": 1050
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 13.25,
|
865 |
+
"grad_norm": 0.6976212859153748,
|
866 |
+
"learning_rate": 0.0016333333333333334,
|
867 |
+
"loss": 0.471,
|
868 |
+
"step": 1060
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"epoch": 13.38,
|
872 |
+
"grad_norm": 0.7652568817138672,
|
873 |
+
"learning_rate": 0.001627777777777778,
|
874 |
+
"loss": 0.4821,
|
875 |
+
"step": 1070
|
876 |
+
},
|
877 |
+
{
|
878 |
+
"epoch": 13.5,
|
879 |
+
"grad_norm": 0.7834269404411316,
|
880 |
+
"learning_rate": 0.0016222222222222222,
|
881 |
+
"loss": 0.5091,
|
882 |
+
"step": 1080
|
883 |
+
},
|
884 |
+
{
|
885 |
+
"epoch": 13.62,
|
886 |
+
"grad_norm": 0.8186032176017761,
|
887 |
+
"learning_rate": 0.0016166666666666666,
|
888 |
+
"loss": 0.4611,
|
889 |
+
"step": 1090
|
890 |
+
},
|
891 |
+
{
|
892 |
+
"epoch": 13.75,
|
893 |
+
"grad_norm": 0.7720199227333069,
|
894 |
+
"learning_rate": 0.0016111111111111111,
|
895 |
+
"loss": 0.5397,
|
896 |
+
"step": 1100
|
897 |
+
},
|
898 |
+
{
|
899 |
+
"epoch": 13.88,
|
900 |
+
"grad_norm": 0.6797453165054321,
|
901 |
+
"learning_rate": 0.0016055555555555556,
|
902 |
+
"loss": 0.5135,
|
903 |
+
"step": 1110
|
904 |
+
},
|
905 |
+
{
|
906 |
+
"epoch": 14.0,
|
907 |
+
"grad_norm": 0.726184606552124,
|
908 |
+
"learning_rate": 0.0016,
|
909 |
+
"loss": 0.5175,
|
910 |
+
"step": 1120
|
911 |
+
},
|
912 |
+
{
|
913 |
+
"epoch": 14.0,
|
914 |
+
"eval_accuracy": 0.7841796875,
|
915 |
+
"eval_loss": 0.5278125405311584,
|
916 |
+
"eval_runtime": 4.8906,
|
917 |
+
"eval_samples_per_second": 209.383,
|
918 |
+
"eval_steps_per_second": 13.086,
|
919 |
+
"step": 1120
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 14.12,
|
923 |
+
"grad_norm": 0.6777172088623047,
|
924 |
+
"learning_rate": 0.0015944444444444446,
|
925 |
+
"loss": 0.4831,
|
926 |
+
"step": 1130
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 14.25,
|
930 |
+
"grad_norm": 0.6228752732276917,
|
931 |
+
"learning_rate": 0.0015888888888888888,
|
932 |
+
"loss": 0.4657,
|
933 |
+
"step": 1140
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 14.38,
|
937 |
+
"grad_norm": 0.7296370267868042,
|
938 |
+
"learning_rate": 0.0015833333333333333,
|
939 |
+
"loss": 0.5084,
|
940 |
+
"step": 1150
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 14.5,
|
944 |
+
"grad_norm": 0.7809439897537231,
|
945 |
+
"learning_rate": 0.0015777777777777778,
|
946 |
+
"loss": 0.4749,
|
947 |
+
"step": 1160
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 14.62,
|
951 |
+
"grad_norm": 0.4627506136894226,
|
952 |
+
"learning_rate": 0.0015722222222222223,
|
953 |
+
"loss": 0.4157,
|
954 |
+
"step": 1170
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 14.75,
|
958 |
+
"grad_norm": 0.465811163187027,
|
959 |
+
"learning_rate": 0.0015666666666666667,
|
960 |
+
"loss": 0.4192,
|
961 |
+
"step": 1180
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 14.88,
|
965 |
+
"grad_norm": 0.636384129524231,
|
966 |
+
"learning_rate": 0.0015611111111111112,
|
967 |
+
"loss": 0.4627,
|
968 |
+
"step": 1190
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 15.0,
|
972 |
+
"grad_norm": 0.8339561223983765,
|
973 |
+
"learning_rate": 0.0015555555555555557,
|
974 |
+
"loss": 0.4865,
|
975 |
+
"step": 1200
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 15.0,
|
979 |
+
"eval_accuracy": 0.837890625,
|
980 |
+
"eval_loss": 0.4163576364517212,
|
981 |
+
"eval_runtime": 5.3394,
|
982 |
+
"eval_samples_per_second": 191.782,
|
983 |
+
"eval_steps_per_second": 11.986,
|
984 |
+
"step": 1200
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 15.12,
|
988 |
+
"grad_norm": 0.5218497514724731,
|
989 |
+
"learning_rate": 0.0015500000000000002,
|
990 |
+
"loss": 0.4253,
|
991 |
+
"step": 1210
|
992 |
+
},
|
993 |
+
{
|
994 |
+
"epoch": 15.25,
|
995 |
+
"grad_norm": 0.6273193359375,
|
996 |
+
"learning_rate": 0.0015444444444444446,
|
997 |
+
"loss": 0.4474,
|
998 |
+
"step": 1220
|
999 |
+
},
|
1000 |
+
{
|
1001 |
+
"epoch": 15.38,
|
1002 |
+
"grad_norm": 0.6019622087478638,
|
1003 |
+
"learning_rate": 0.001538888888888889,
|
1004 |
+
"loss": 0.4008,
|
1005 |
+
"step": 1230
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 15.5,
|
1009 |
+
"grad_norm": 0.7020573616027832,
|
1010 |
+
"learning_rate": 0.0015333333333333334,
|
1011 |
+
"loss": 0.3768,
|
1012 |
+
"step": 1240
|
1013 |
+
},
|
1014 |
+
{
|
1015 |
+
"epoch": 15.62,
|
1016 |
+
"grad_norm": 0.577691376209259,
|
1017 |
+
"learning_rate": 0.0015277777777777776,
|
1018 |
+
"loss": 0.4108,
|
1019 |
+
"step": 1250
|
1020 |
+
},
|
1021 |
+
{
|
1022 |
+
"epoch": 15.75,
|
1023 |
+
"grad_norm": 0.8489026427268982,
|
1024 |
+
"learning_rate": 0.0015222222222222221,
|
1025 |
+
"loss": 0.3994,
|
1026 |
+
"step": 1260
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"epoch": 15.88,
|
1030 |
+
"grad_norm": 0.42233309149742126,
|
1031 |
+
"learning_rate": 0.0015166666666666666,
|
1032 |
+
"loss": 0.4292,
|
1033 |
+
"step": 1270
|
1034 |
+
},
|
1035 |
+
{
|
1036 |
+
"epoch": 16.0,
|
1037 |
+
"grad_norm": 0.48867735266685486,
|
1038 |
+
"learning_rate": 0.001511111111111111,
|
1039 |
+
"loss": 0.4049,
|
1040 |
+
"step": 1280
|
1041 |
+
},
|
1042 |
+
{
|
1043 |
+
"epoch": 16.0,
|
1044 |
+
"eval_accuracy": 0.830078125,
|
1045 |
+
"eval_loss": 0.4204105734825134,
|
1046 |
+
"eval_runtime": 4.8855,
|
1047 |
+
"eval_samples_per_second": 209.602,
|
1048 |
+
"eval_steps_per_second": 13.1,
|
1049 |
+
"step": 1280
|
1050 |
+
},
|
1051 |
+
{
|
1052 |
+
"epoch": 16.12,
|
1053 |
+
"grad_norm": 0.6492818593978882,
|
1054 |
+
"learning_rate": 0.0015055555555555556,
|
1055 |
+
"loss": 0.3885,
|
1056 |
+
"step": 1290
|
1057 |
+
},
|
1058 |
+
{
|
1059 |
+
"epoch": 16.25,
|
1060 |
+
"grad_norm": 0.4546281397342682,
|
1061 |
+
"learning_rate": 0.0015,
|
1062 |
+
"loss": 0.4096,
|
1063 |
+
"step": 1300
|
1064 |
+
},
|
1065 |
+
{
|
1066 |
+
"epoch": 16.38,
|
1067 |
+
"grad_norm": 0.6827344298362732,
|
1068 |
+
"learning_rate": 0.0014944444444444445,
|
1069 |
+
"loss": 0.3618,
|
1070 |
+
"step": 1310
|
1071 |
+
},
|
1072 |
+
{
|
1073 |
+
"epoch": 16.5,
|
1074 |
+
"grad_norm": 0.454326868057251,
|
1075 |
+
"learning_rate": 0.001488888888888889,
|
1076 |
+
"loss": 0.3863,
|
1077 |
+
"step": 1320
|
1078 |
+
},
|
1079 |
+
{
|
1080 |
+
"epoch": 16.62,
|
1081 |
+
"grad_norm": 0.6911420226097107,
|
1082 |
+
"learning_rate": 0.0014833333333333335,
|
1083 |
+
"loss": 0.4264,
|
1084 |
+
"step": 1330
|
1085 |
+
},
|
1086 |
+
{
|
1087 |
+
"epoch": 16.75,
|
1088 |
+
"grad_norm": 0.6122339367866516,
|
1089 |
+
"learning_rate": 0.001477777777777778,
|
1090 |
+
"loss": 0.4205,
|
1091 |
+
"step": 1340
|
1092 |
+
},
|
1093 |
+
{
|
1094 |
+
"epoch": 16.88,
|
1095 |
+
"grad_norm": 0.5123728513717651,
|
1096 |
+
"learning_rate": 0.0014722222222222224,
|
1097 |
+
"loss": 0.4419,
|
1098 |
+
"step": 1350
|
1099 |
+
},
|
1100 |
+
{
|
1101 |
+
"epoch": 17.0,
|
1102 |
+
"grad_norm": 1.0908498764038086,
|
1103 |
+
"learning_rate": 0.0014666666666666667,
|
1104 |
+
"loss": 0.4167,
|
1105 |
+
"step": 1360
|
1106 |
+
},
|
1107 |
+
{
|
1108 |
+
"epoch": 17.0,
|
1109 |
+
"eval_accuracy": 0.828125,
|
1110 |
+
"eval_loss": 0.47203314304351807,
|
1111 |
+
"eval_runtime": 5.3495,
|
1112 |
+
"eval_samples_per_second": 191.421,
|
1113 |
+
"eval_steps_per_second": 11.964,
|
1114 |
+
"step": 1360
|
1115 |
+
},
|
1116 |
+
{
|
1117 |
+
"epoch": 17.12,
|
1118 |
+
"grad_norm": 0.42975571751594543,
|
1119 |
+
"learning_rate": 0.0014611111111111112,
|
1120 |
+
"loss": 0.4006,
|
1121 |
+
"step": 1370
|
1122 |
+
},
|
1123 |
+
{
|
1124 |
+
"epoch": 17.25,
|
1125 |
+
"grad_norm": 0.6392154693603516,
|
1126 |
+
"learning_rate": 0.0014555555555555554,
|
1127 |
+
"loss": 0.3581,
|
1128 |
+
"step": 1380
|
1129 |
+
},
|
1130 |
+
{
|
1131 |
+
"epoch": 17.38,
|
1132 |
+
"grad_norm": 0.6548070907592773,
|
1133 |
+
"learning_rate": 0.00145,
|
1134 |
+
"loss": 0.3672,
|
1135 |
+
"step": 1390
|
1136 |
+
},
|
1137 |
+
{
|
1138 |
+
"epoch": 17.5,
|
1139 |
+
"grad_norm": 0.6939528584480286,
|
1140 |
+
"learning_rate": 0.0014444444444444444,
|
1141 |
+
"loss": 0.3514,
|
1142 |
+
"step": 1400
|
1143 |
+
},
|
1144 |
+
{
|
1145 |
+
"epoch": 17.62,
|
1146 |
+
"grad_norm": 0.6098494529724121,
|
1147 |
+
"learning_rate": 0.0014388888888888889,
|
1148 |
+
"loss": 0.3835,
|
1149 |
+
"step": 1410
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 17.75,
|
1153 |
+
"grad_norm": 0.5356572866439819,
|
1154 |
+
"learning_rate": 0.0014333333333333333,
|
1155 |
+
"loss": 0.3326,
|
1156 |
+
"step": 1420
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"epoch": 17.88,
|
1160 |
+
"grad_norm": 0.6472760438919067,
|
1161 |
+
"learning_rate": 0.0014277777777777778,
|
1162 |
+
"loss": 0.3829,
|
1163 |
+
"step": 1430
|
1164 |
+
},
|
1165 |
+
{
|
1166 |
+
"epoch": 18.0,
|
1167 |
+
"grad_norm": 0.67198646068573,
|
1168 |
+
"learning_rate": 0.0014222222222222223,
|
1169 |
+
"loss": 0.36,
|
1170 |
+
"step": 1440
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"epoch": 18.0,
|
1174 |
+
"eval_accuracy": 0.81640625,
|
1175 |
+
"eval_loss": 0.4660454988479614,
|
1176 |
+
"eval_runtime": 4.9124,
|
1177 |
+
"eval_samples_per_second": 208.451,
|
1178 |
+
"eval_steps_per_second": 13.028,
|
1179 |
+
"step": 1440
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 18.12,
|
1183 |
+
"grad_norm": 0.4594449996948242,
|
1184 |
+
"learning_rate": 0.0014166666666666668,
|
1185 |
+
"loss": 0.3549,
|
1186 |
+
"step": 1450
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 18.25,
|
1190 |
+
"grad_norm": 0.4456086754798889,
|
1191 |
+
"learning_rate": 0.0014111111111111112,
|
1192 |
+
"loss": 0.2899,
|
1193 |
+
"step": 1460
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 18.38,
|
1197 |
+
"grad_norm": 0.724087119102478,
|
1198 |
+
"learning_rate": 0.0014055555555555555,
|
1199 |
+
"loss": 0.2976,
|
1200 |
+
"step": 1470
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 18.5,
|
1204 |
+
"grad_norm": 0.8099024891853333,
|
1205 |
+
"learning_rate": 0.0014,
|
1206 |
+
"loss": 0.3706,
|
1207 |
+
"step": 1480
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 18.62,
|
1211 |
+
"grad_norm": 0.6271733641624451,
|
1212 |
+
"learning_rate": 0.0013944444444444445,
|
1213 |
+
"loss": 0.3591,
|
1214 |
+
"step": 1490
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 18.75,
|
1218 |
+
"grad_norm": 0.5864254236221313,
|
1219 |
+
"learning_rate": 0.001388888888888889,
|
1220 |
+
"loss": 0.3184,
|
1221 |
+
"step": 1500
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 18.88,
|
1225 |
+
"grad_norm": 0.4915286898612976,
|
1226 |
+
"learning_rate": 0.0013833333333333334,
|
1227 |
+
"loss": 0.301,
|
1228 |
+
"step": 1510
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 19.0,
|
1232 |
+
"grad_norm": 0.6932692527770996,
|
1233 |
+
"learning_rate": 0.001377777777777778,
|
1234 |
+
"loss": 0.3195,
|
1235 |
+
"step": 1520
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 19.0,
|
1239 |
+
"eval_accuracy": 0.876953125,
|
1240 |
+
"eval_loss": 0.306354820728302,
|
1241 |
+
"eval_runtime": 5.2563,
|
1242 |
+
"eval_samples_per_second": 194.815,
|
1243 |
+
"eval_steps_per_second": 12.176,
|
1244 |
+
"step": 1520
|
1245 |
+
},
|
1246 |
+
{
|
1247 |
+
"epoch": 19.12,
|
1248 |
+
"grad_norm": 0.5778792500495911,
|
1249 |
+
"learning_rate": 0.0013722222222222222,
|
1250 |
+
"loss": 0.3493,
|
1251 |
+
"step": 1530
|
1252 |
+
},
|
1253 |
+
{
|
1254 |
+
"epoch": 19.25,
|
1255 |
+
"grad_norm": 0.951936662197113,
|
1256 |
+
"learning_rate": 0.0013666666666666666,
|
1257 |
+
"loss": 0.3305,
|
1258 |
+
"step": 1540
|
1259 |
+
},
|
1260 |
+
{
|
1261 |
+
"epoch": 19.38,
|
1262 |
+
"grad_norm": 0.6778426170349121,
|
1263 |
+
"learning_rate": 0.0013611111111111111,
|
1264 |
+
"loss": 0.32,
|
1265 |
+
"step": 1550
|
1266 |
+
},
|
1267 |
+
{
|
1268 |
+
"epoch": 19.5,
|
1269 |
+
"grad_norm": 0.6356533765792847,
|
1270 |
+
"learning_rate": 0.0013555555555555556,
|
1271 |
+
"loss": 0.2889,
|
1272 |
+
"step": 1560
|
1273 |
+
},
|
1274 |
+
{
|
1275 |
+
"epoch": 19.62,
|
1276 |
+
"grad_norm": 0.6476128697395325,
|
1277 |
+
"learning_rate": 0.00135,
|
1278 |
+
"loss": 0.2907,
|
1279 |
+
"step": 1570
|
1280 |
+
},
|
1281 |
+
{
|
1282 |
+
"epoch": 19.75,
|
1283 |
+
"grad_norm": 0.4664938151836395,
|
1284 |
+
"learning_rate": 0.0013444444444444445,
|
1285 |
+
"loss": 0.3261,
|
1286 |
+
"step": 1580
|
1287 |
+
},
|
1288 |
+
{
|
1289 |
+
"epoch": 19.88,
|
1290 |
+
"grad_norm": 1.06290602684021,
|
1291 |
+
"learning_rate": 0.0013388888888888888,
|
1292 |
+
"loss": 0.3365,
|
1293 |
+
"step": 1590
|
1294 |
+
},
|
1295 |
+
{
|
1296 |
+
"epoch": 20.0,
|
1297 |
+
"grad_norm": 0.5365467667579651,
|
1298 |
+
"learning_rate": 0.0013333333333333333,
|
1299 |
+
"loss": 0.3652,
|
1300 |
+
"step": 1600
|
1301 |
+
},
|
1302 |
+
{
|
1303 |
+
"epoch": 20.0,
|
1304 |
+
"eval_accuracy": 0.912109375,
|
1305 |
+
"eval_loss": 0.25709766149520874,
|
1306 |
+
"eval_runtime": 4.9952,
|
1307 |
+
"eval_samples_per_second": 204.995,
|
1308 |
+
"eval_steps_per_second": 12.812,
|
1309 |
+
"step": 1600
|
1310 |
+
},
|
1311 |
+
{
|
1312 |
+
"epoch": 20.12,
|
1313 |
+
"grad_norm": 0.5051919221878052,
|
1314 |
+
"learning_rate": 0.0013277777777777778,
|
1315 |
+
"loss": 0.3147,
|
1316 |
+
"step": 1610
|
1317 |
+
},
|
1318 |
+
{
|
1319 |
+
"epoch": 20.25,
|
1320 |
+
"grad_norm": 0.5098996162414551,
|
1321 |
+
"learning_rate": 0.0013222222222222222,
|
1322 |
+
"loss": 0.3085,
|
1323 |
+
"step": 1620
|
1324 |
+
},
|
1325 |
+
{
|
1326 |
+
"epoch": 20.38,
|
1327 |
+
"grad_norm": 0.5585361123085022,
|
1328 |
+
"learning_rate": 0.0013166666666666667,
|
1329 |
+
"loss": 0.3679,
|
1330 |
+
"step": 1630
|
1331 |
+
},
|
1332 |
+
{
|
1333 |
+
"epoch": 20.5,
|
1334 |
+
"grad_norm": 0.38560378551483154,
|
1335 |
+
"learning_rate": 0.0013111111111111112,
|
1336 |
+
"loss": 0.2987,
|
1337 |
+
"step": 1640
|
1338 |
+
},
|
1339 |
+
{
|
1340 |
+
"epoch": 20.62,
|
1341 |
+
"grad_norm": 0.3209057152271271,
|
1342 |
+
"learning_rate": 0.0013055555555555557,
|
1343 |
+
"loss": 0.2792,
|
1344 |
+
"step": 1650
|
1345 |
+
},
|
1346 |
+
{
|
1347 |
+
"epoch": 20.75,
|
1348 |
+
"grad_norm": 0.6471489667892456,
|
1349 |
+
"learning_rate": 0.0013000000000000002,
|
1350 |
+
"loss": 0.2755,
|
1351 |
+
"step": 1660
|
1352 |
+
},
|
1353 |
+
{
|
1354 |
+
"epoch": 20.88,
|
1355 |
+
"grad_norm": 0.8814804553985596,
|
1356 |
+
"learning_rate": 0.0012944444444444446,
|
1357 |
+
"loss": 0.2993,
|
1358 |
+
"step": 1670
|
1359 |
+
},
|
1360 |
+
{
|
1361 |
+
"epoch": 21.0,
|
1362 |
+
"grad_norm": 0.5392754673957825,
|
1363 |
+
"learning_rate": 0.001288888888888889,
|
1364 |
+
"loss": 0.2794,
|
1365 |
+
"step": 1680
|
1366 |
+
},
|
1367 |
+
{
|
1368 |
+
"epoch": 21.0,
|
1369 |
+
"eval_accuracy": 0.9150390625,
|
1370 |
+
"eval_loss": 0.24504294991493225,
|
1371 |
+
"eval_runtime": 4.8909,
|
1372 |
+
"eval_samples_per_second": 209.37,
|
1373 |
+
"eval_steps_per_second": 13.086,
|
1374 |
+
"step": 1680
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 21.12,
|
1378 |
+
"grad_norm": 0.6234158873558044,
|
1379 |
+
"learning_rate": 0.0012833333333333334,
|
1380 |
+
"loss": 0.2926,
|
1381 |
+
"step": 1690
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 21.25,
|
1385 |
+
"grad_norm": 0.4284802973270416,
|
1386 |
+
"learning_rate": 0.0012777777777777776,
|
1387 |
+
"loss": 0.2803,
|
1388 |
+
"step": 1700
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 21.38,
|
1392 |
+
"grad_norm": 0.688140869140625,
|
1393 |
+
"learning_rate": 0.0012722222222222221,
|
1394 |
+
"loss": 0.2799,
|
1395 |
+
"step": 1710
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 21.5,
|
1399 |
+
"grad_norm": 0.8576880097389221,
|
1400 |
+
"learning_rate": 0.0012666666666666666,
|
1401 |
+
"loss": 0.2868,
|
1402 |
+
"step": 1720
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 21.62,
|
1406 |
+
"grad_norm": 0.6299762725830078,
|
1407 |
+
"learning_rate": 0.001261111111111111,
|
1408 |
+
"loss": 0.2971,
|
1409 |
+
"step": 1730
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 21.75,
|
1413 |
+
"grad_norm": 0.7093678116798401,
|
1414 |
+
"learning_rate": 0.0012555555555555555,
|
1415 |
+
"loss": 0.2905,
|
1416 |
+
"step": 1740
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 21.88,
|
1420 |
+
"grad_norm": 0.4271737039089203,
|
1421 |
+
"learning_rate": 0.00125,
|
1422 |
+
"loss": 0.3336,
|
1423 |
+
"step": 1750
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 22.0,
|
1427 |
+
"grad_norm": 0.6771571040153503,
|
1428 |
+
"learning_rate": 0.0012444444444444445,
|
1429 |
+
"loss": 0.2704,
|
1430 |
+
"step": 1760
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 22.0,
|
1434 |
+
"eval_accuracy": 0.9033203125,
|
1435 |
+
"eval_loss": 0.23907524347305298,
|
1436 |
+
"eval_runtime": 5.3054,
|
1437 |
+
"eval_samples_per_second": 193.012,
|
1438 |
+
"eval_steps_per_second": 12.063,
|
1439 |
+
"step": 1760
|
1440 |
+
},
|
1441 |
+
{
|
1442 |
+
"epoch": 22.12,
|
1443 |
+
"grad_norm": 0.44859397411346436,
|
1444 |
+
"learning_rate": 0.001238888888888889,
|
1445 |
+
"loss": 0.28,
|
1446 |
+
"step": 1770
|
1447 |
+
},
|
1448 |
+
{
|
1449 |
+
"epoch": 22.25,
|
1450 |
+
"grad_norm": 0.5617765784263611,
|
1451 |
+
"learning_rate": 0.0012333333333333335,
|
1452 |
+
"loss": 0.3093,
|
1453 |
+
"step": 1780
|
1454 |
+
},
|
1455 |
+
{
|
1456 |
+
"epoch": 22.38,
|
1457 |
+
"grad_norm": 0.6634913682937622,
|
1458 |
+
"learning_rate": 0.001227777777777778,
|
1459 |
+
"loss": 0.2417,
|
1460 |
+
"step": 1790
|
1461 |
+
},
|
1462 |
+
{
|
1463 |
+
"epoch": 22.5,
|
1464 |
+
"grad_norm": 0.670782744884491,
|
1465 |
+
"learning_rate": 0.0012222222222222224,
|
1466 |
+
"loss": 0.2932,
|
1467 |
+
"step": 1800
|
1468 |
+
},
|
1469 |
+
{
|
1470 |
+
"epoch": 22.62,
|
1471 |
+
"grad_norm": 0.6564796566963196,
|
1472 |
+
"learning_rate": 0.0012166666666666667,
|
1473 |
+
"loss": 0.3042,
|
1474 |
+
"step": 1810
|
1475 |
+
},
|
1476 |
+
{
|
1477 |
+
"epoch": 22.75,
|
1478 |
+
"grad_norm": 0.34089842438697815,
|
1479 |
+
"learning_rate": 0.0012111111111111112,
|
1480 |
+
"loss": 0.2925,
|
1481 |
+
"step": 1820
|
1482 |
+
},
|
1483 |
+
{
|
1484 |
+
"epoch": 22.88,
|
1485 |
+
"grad_norm": 0.5612368583679199,
|
1486 |
+
"learning_rate": 0.0012055555555555554,
|
1487 |
+
"loss": 0.2559,
|
1488 |
+
"step": 1830
|
1489 |
+
},
|
1490 |
+
{
|
1491 |
+
"epoch": 23.0,
|
1492 |
+
"grad_norm": 0.624458909034729,
|
1493 |
+
"learning_rate": 0.0012,
|
1494 |
+
"loss": 0.2612,
|
1495 |
+
"step": 1840
|
1496 |
+
},
|
1497 |
+
{
|
1498 |
+
"epoch": 23.0,
|
1499 |
+
"eval_accuracy": 0.927734375,
|
1500 |
+
"eval_loss": 0.23524078726768494,
|
1501 |
+
"eval_runtime": 4.902,
|
1502 |
+
"eval_samples_per_second": 208.893,
|
1503 |
+
"eval_steps_per_second": 13.056,
|
1504 |
+
"step": 1840
|
1505 |
+
},
|
1506 |
+
{
|
1507 |
+
"epoch": 23.12,
|
1508 |
+
"grad_norm": 0.6820557117462158,
|
1509 |
+
"learning_rate": 0.0011944444444444444,
|
1510 |
+
"loss": 0.2282,
|
1511 |
+
"step": 1850
|
1512 |
+
},
|
1513 |
+
{
|
1514 |
+
"epoch": 23.25,
|
1515 |
+
"grad_norm": 0.5979276895523071,
|
1516 |
+
"learning_rate": 0.0011888888888888889,
|
1517 |
+
"loss": 0.2569,
|
1518 |
+
"step": 1860
|
1519 |
+
},
|
1520 |
+
{
|
1521 |
+
"epoch": 23.38,
|
1522 |
+
"grad_norm": 0.5427021384239197,
|
1523 |
+
"learning_rate": 0.0011833333333333333,
|
1524 |
+
"loss": 0.2724,
|
1525 |
+
"step": 1870
|
1526 |
+
},
|
1527 |
+
{
|
1528 |
+
"epoch": 23.5,
|
1529 |
+
"grad_norm": 0.4382477104663849,
|
1530 |
+
"learning_rate": 0.0011777777777777778,
|
1531 |
+
"loss": 0.2616,
|
1532 |
+
"step": 1880
|
1533 |
+
},
|
1534 |
+
{
|
1535 |
+
"epoch": 23.62,
|
1536 |
+
"grad_norm": 0.6240445375442505,
|
1537 |
+
"learning_rate": 0.0011722222222222223,
|
1538 |
+
"loss": 0.2636,
|
1539 |
+
"step": 1890
|
1540 |
+
},
|
1541 |
+
{
|
1542 |
+
"epoch": 23.75,
|
1543 |
+
"grad_norm": 0.7440346479415894,
|
1544 |
+
"learning_rate": 0.0011666666666666668,
|
1545 |
+
"loss": 0.2913,
|
1546 |
+
"step": 1900
|
1547 |
+
},
|
1548 |
+
{
|
1549 |
+
"epoch": 23.88,
|
1550 |
+
"grad_norm": 0.4682701826095581,
|
1551 |
+
"learning_rate": 0.0011611111111111112,
|
1552 |
+
"loss": 0.2499,
|
1553 |
+
"step": 1910
|
1554 |
+
},
|
1555 |
+
{
|
1556 |
+
"epoch": 24.0,
|
1557 |
+
"grad_norm": 0.5112751722335815,
|
1558 |
+
"learning_rate": 0.0011555555555555555,
|
1559 |
+
"loss": 0.2425,
|
1560 |
+
"step": 1920
|
1561 |
+
},
|
1562 |
+
{
|
1563 |
+
"epoch": 24.0,
|
1564 |
+
"eval_accuracy": 0.828125,
|
1565 |
+
"eval_loss": 0.4720377027988434,
|
1566 |
+
"eval_runtime": 5.3156,
|
1567 |
+
"eval_samples_per_second": 192.639,
|
1568 |
+
"eval_steps_per_second": 12.04,
|
1569 |
+
"step": 1920
|
1570 |
+
},
|
1571 |
+
{
|
1572 |
+
"epoch": 24.12,
|
1573 |
+
"grad_norm": 0.765444278717041,
|
1574 |
+
"learning_rate": 0.00115,
|
1575 |
+
"loss": 0.2736,
|
1576 |
+
"step": 1930
|
1577 |
+
},
|
1578 |
+
{
|
1579 |
+
"epoch": 24.25,
|
1580 |
+
"grad_norm": 0.380066841840744,
|
1581 |
+
"learning_rate": 0.0011444444444444445,
|
1582 |
+
"loss": 0.2357,
|
1583 |
+
"step": 1940
|
1584 |
+
},
|
1585 |
+
{
|
1586 |
+
"epoch": 24.38,
|
1587 |
+
"grad_norm": 0.43320003151893616,
|
1588 |
+
"learning_rate": 0.001138888888888889,
|
1589 |
+
"loss": 0.2518,
|
1590 |
+
"step": 1950
|
1591 |
+
},
|
1592 |
+
{
|
1593 |
+
"epoch": 24.5,
|
1594 |
+
"grad_norm": 0.5003307461738586,
|
1595 |
+
"learning_rate": 0.0011333333333333334,
|
1596 |
+
"loss": 0.2898,
|
1597 |
+
"step": 1960
|
1598 |
+
},
|
1599 |
+
{
|
1600 |
+
"epoch": 24.62,
|
1601 |
+
"grad_norm": 0.41153478622436523,
|
1602 |
+
"learning_rate": 0.001127777777777778,
|
1603 |
+
"loss": 0.2209,
|
1604 |
+
"step": 1970
|
1605 |
+
},
|
1606 |
+
{
|
1607 |
+
"epoch": 24.75,
|
1608 |
+
"grad_norm": 0.41805940866470337,
|
1609 |
+
"learning_rate": 0.0011222222222222222,
|
1610 |
+
"loss": 0.235,
|
1611 |
+
"step": 1980
|
1612 |
+
},
|
1613 |
+
{
|
1614 |
+
"epoch": 24.88,
|
1615 |
+
"grad_norm": 0.5226410627365112,
|
1616 |
+
"learning_rate": 0.0011166666666666666,
|
1617 |
+
"loss": 0.2349,
|
1618 |
+
"step": 1990
|
1619 |
+
},
|
1620 |
+
{
|
1621 |
+
"epoch": 25.0,
|
1622 |
+
"grad_norm": 0.3767559826374054,
|
1623 |
+
"learning_rate": 0.0011111111111111111,
|
1624 |
+
"loss": 0.2567,
|
1625 |
+
"step": 2000
|
1626 |
+
},
|
1627 |
+
{
|
1628 |
+
"epoch": 25.0,
|
1629 |
+
"eval_accuracy": 0.9130859375,
|
1630 |
+
"eval_loss": 0.22960150241851807,
|
1631 |
+
"eval_runtime": 4.887,
|
1632 |
+
"eval_samples_per_second": 209.535,
|
1633 |
+
"eval_steps_per_second": 13.096,
|
1634 |
+
"step": 2000
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 25.12,
|
1638 |
+
"grad_norm": 0.6860052943229675,
|
1639 |
+
"learning_rate": 0.0011055555555555556,
|
1640 |
+
"loss": 0.2426,
|
1641 |
+
"step": 2010
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 25.25,
|
1645 |
+
"grad_norm": 0.3876688778400421,
|
1646 |
+
"learning_rate": 0.0011,
|
1647 |
+
"loss": 0.2243,
|
1648 |
+
"step": 2020
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 25.38,
|
1652 |
+
"grad_norm": 0.3251183032989502,
|
1653 |
+
"learning_rate": 0.0010944444444444445,
|
1654 |
+
"loss": 0.234,
|
1655 |
+
"step": 2030
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 25.5,
|
1659 |
+
"grad_norm": 0.5538493990898132,
|
1660 |
+
"learning_rate": 0.0010888888888888888,
|
1661 |
+
"loss": 0.2547,
|
1662 |
+
"step": 2040
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 25.62,
|
1666 |
+
"grad_norm": 0.6539644598960876,
|
1667 |
+
"learning_rate": 0.0010833333333333333,
|
1668 |
+
"loss": 0.2382,
|
1669 |
+
"step": 2050
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 25.75,
|
1673 |
+
"grad_norm": 0.6687932014465332,
|
1674 |
+
"learning_rate": 0.0010777777777777778,
|
1675 |
+
"loss": 0.2254,
|
1676 |
+
"step": 2060
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 25.88,
|
1680 |
+
"grad_norm": 0.6210919618606567,
|
1681 |
+
"learning_rate": 0.0010722222222222222,
|
1682 |
+
"loss": 0.2356,
|
1683 |
+
"step": 2070
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 26.0,
|
1687 |
+
"grad_norm": 0.5525135397911072,
|
1688 |
+
"learning_rate": 0.0010666666666666667,
|
1689 |
+
"loss": 0.2302,
|
1690 |
+
"step": 2080
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 26.0,
|
1694 |
+
"eval_accuracy": 0.89453125,
|
1695 |
+
"eval_loss": 0.30673664808273315,
|
1696 |
+
"eval_runtime": 4.9576,
|
1697 |
+
"eval_samples_per_second": 206.552,
|
1698 |
+
"eval_steps_per_second": 12.909,
|
1699 |
+
"step": 2080
|
1700 |
+
},
|
1701 |
+
{
|
1702 |
+
"epoch": 26.12,
|
1703 |
+
"grad_norm": 0.5014208555221558,
|
1704 |
+
"learning_rate": 0.0010611111111111112,
|
1705 |
+
"loss": 0.2403,
|
1706 |
+
"step": 2090
|
1707 |
+
},
|
1708 |
+
{
|
1709 |
+
"epoch": 26.25,
|
1710 |
+
"grad_norm": 0.6093131303787231,
|
1711 |
+
"learning_rate": 0.0010555555555555557,
|
1712 |
+
"loss": 0.2356,
|
1713 |
+
"step": 2100
|
1714 |
+
},
|
1715 |
+
{
|
1716 |
+
"epoch": 26.38,
|
1717 |
+
"grad_norm": 0.3627248704433441,
|
1718 |
+
"learning_rate": 0.0010500000000000002,
|
1719 |
+
"loss": 0.2509,
|
1720 |
+
"step": 2110
|
1721 |
+
},
|
1722 |
+
{
|
1723 |
+
"epoch": 26.5,
|
1724 |
+
"grad_norm": 0.4119124114513397,
|
1725 |
+
"learning_rate": 0.0010444444444444446,
|
1726 |
+
"loss": 0.1915,
|
1727 |
+
"step": 2120
|
1728 |
+
},
|
1729 |
+
{
|
1730 |
+
"epoch": 26.62,
|
1731 |
+
"grad_norm": 0.5565811395645142,
|
1732 |
+
"learning_rate": 0.0010388888888888889,
|
1733 |
+
"loss": 0.191,
|
1734 |
+
"step": 2130
|
1735 |
+
},
|
1736 |
+
{
|
1737 |
+
"epoch": 26.75,
|
1738 |
+
"grad_norm": 0.44097578525543213,
|
1739 |
+
"learning_rate": 0.0010333333333333334,
|
1740 |
+
"loss": 0.2353,
|
1741 |
+
"step": 2140
|
1742 |
+
},
|
1743 |
+
{
|
1744 |
+
"epoch": 26.88,
|
1745 |
+
"grad_norm": 0.4542636275291443,
|
1746 |
+
"learning_rate": 0.0010277777777777776,
|
1747 |
+
"loss": 0.2144,
|
1748 |
+
"step": 2150
|
1749 |
+
},
|
1750 |
+
{
|
1751 |
+
"epoch": 27.0,
|
1752 |
+
"grad_norm": 0.4763772785663605,
|
1753 |
+
"learning_rate": 0.0010222222222222221,
|
1754 |
+
"loss": 0.2358,
|
1755 |
+
"step": 2160
|
1756 |
+
},
|
1757 |
+
{
|
1758 |
+
"epoch": 27.0,
|
1759 |
+
"eval_accuracy": 0.9375,
|
1760 |
+
"eval_loss": 0.17758239805698395,
|
1761 |
+
"eval_runtime": 5.3185,
|
1762 |
+
"eval_samples_per_second": 192.534,
|
1763 |
+
"eval_steps_per_second": 12.033,
|
1764 |
+
"step": 2160
|
1765 |
+
},
|
1766 |
+
{
|
1767 |
+
"epoch": 27.12,
|
1768 |
+
"grad_norm": 0.7219308614730835,
|
1769 |
+
"learning_rate": 0.0010166666666666666,
|
1770 |
+
"loss": 0.238,
|
1771 |
+
"step": 2170
|
1772 |
+
},
|
1773 |
+
{
|
1774 |
+
"epoch": 27.25,
|
1775 |
+
"grad_norm": 0.7707520127296448,
|
1776 |
+
"learning_rate": 0.001011111111111111,
|
1777 |
+
"loss": 0.1863,
|
1778 |
+
"step": 2180
|
1779 |
+
},
|
1780 |
+
{
|
1781 |
+
"epoch": 27.38,
|
1782 |
+
"grad_norm": 0.6878935098648071,
|
1783 |
+
"learning_rate": 0.0010055555555555555,
|
1784 |
+
"loss": 0.2493,
|
1785 |
+
"step": 2190
|
1786 |
+
},
|
1787 |
+
{
|
1788 |
+
"epoch": 27.5,
|
1789 |
+
"grad_norm": 0.5451861619949341,
|
1790 |
+
"learning_rate": 0.001,
|
1791 |
+
"loss": 0.2374,
|
1792 |
+
"step": 2200
|
1793 |
+
},
|
1794 |
+
{
|
1795 |
+
"epoch": 27.62,
|
1796 |
+
"grad_norm": 0.39642319083213806,
|
1797 |
+
"learning_rate": 0.0009944444444444445,
|
1798 |
+
"loss": 0.2382,
|
1799 |
+
"step": 2210
|
1800 |
+
},
|
1801 |
+
{
|
1802 |
+
"epoch": 27.75,
|
1803 |
+
"grad_norm": 0.4122956097126007,
|
1804 |
+
"learning_rate": 0.000988888888888889,
|
1805 |
+
"loss": 0.2176,
|
1806 |
+
"step": 2220
|
1807 |
+
},
|
1808 |
+
{
|
1809 |
+
"epoch": 27.88,
|
1810 |
+
"grad_norm": 0.6155421733856201,
|
1811 |
+
"learning_rate": 0.0009833333333333332,
|
1812 |
+
"loss": 0.2128,
|
1813 |
+
"step": 2230
|
1814 |
+
},
|
1815 |
+
{
|
1816 |
+
"epoch": 28.0,
|
1817 |
+
"grad_norm": 0.7283052206039429,
|
1818 |
+
"learning_rate": 0.0009777777777777777,
|
1819 |
+
"loss": 0.2173,
|
1820 |
+
"step": 2240
|
1821 |
+
},
|
1822 |
+
{
|
1823 |
+
"epoch": 28.0,
|
1824 |
+
"eval_accuracy": 0.94921875,
|
1825 |
+
"eval_loss": 0.15962785482406616,
|
1826 |
+
"eval_runtime": 4.9115,
|
1827 |
+
"eval_samples_per_second": 208.489,
|
1828 |
+
"eval_steps_per_second": 13.031,
|
1829 |
+
"step": 2240
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 28.12,
|
1833 |
+
"grad_norm": 0.39027243852615356,
|
1834 |
+
"learning_rate": 0.0009722222222222222,
|
1835 |
+
"loss": 0.1979,
|
1836 |
+
"step": 2250
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 28.25,
|
1840 |
+
"grad_norm": 0.5258718729019165,
|
1841 |
+
"learning_rate": 0.0009666666666666667,
|
1842 |
+
"loss": 0.1447,
|
1843 |
+
"step": 2260
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 28.38,
|
1847 |
+
"grad_norm": 0.6615960001945496,
|
1848 |
+
"learning_rate": 0.0009611111111111112,
|
1849 |
+
"loss": 0.2403,
|
1850 |
+
"step": 2270
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 28.5,
|
1854 |
+
"grad_norm": 0.4044310748577118,
|
1855 |
+
"learning_rate": 0.0009555555555555556,
|
1856 |
+
"loss": 0.1981,
|
1857 |
+
"step": 2280
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 28.62,
|
1861 |
+
"grad_norm": 0.2666930556297302,
|
1862 |
+
"learning_rate": 0.00095,
|
1863 |
+
"loss": 0.2108,
|
1864 |
+
"step": 2290
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 28.75,
|
1868 |
+
"grad_norm": 0.5612334609031677,
|
1869 |
+
"learning_rate": 0.0009444444444444445,
|
1870 |
+
"loss": 0.1783,
|
1871 |
+
"step": 2300
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 28.88,
|
1875 |
+
"grad_norm": 0.48420026898384094,
|
1876 |
+
"learning_rate": 0.000938888888888889,
|
1877 |
+
"loss": 0.1848,
|
1878 |
+
"step": 2310
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 29.0,
|
1882 |
+
"grad_norm": 0.5850337743759155,
|
1883 |
+
"learning_rate": 0.0009333333333333333,
|
1884 |
+
"loss": 0.1798,
|
1885 |
+
"step": 2320
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 29.0,
|
1889 |
+
"eval_accuracy": 0.94140625,
|
1890 |
+
"eval_loss": 0.1548241674900055,
|
1891 |
+
"eval_runtime": 5.32,
|
1892 |
+
"eval_samples_per_second": 192.483,
|
1893 |
+
"eval_steps_per_second": 12.03,
|
1894 |
+
"step": 2320
|
1895 |
+
},
|
1896 |
+
{
|
1897 |
+
"epoch": 29.12,
|
1898 |
+
"grad_norm": 0.5059901475906372,
|
1899 |
+
"learning_rate": 0.0009277777777777778,
|
1900 |
+
"loss": 0.1954,
|
1901 |
+
"step": 2330
|
1902 |
+
},
|
1903 |
+
{
|
1904 |
+
"epoch": 29.25,
|
1905 |
+
"grad_norm": 0.22623513638973236,
|
1906 |
+
"learning_rate": 0.0009222222222222223,
|
1907 |
+
"loss": 0.1604,
|
1908 |
+
"step": 2340
|
1909 |
+
},
|
1910 |
+
{
|
1911 |
+
"epoch": 29.38,
|
1912 |
+
"grad_norm": 0.2330830693244934,
|
1913 |
+
"learning_rate": 0.0009166666666666666,
|
1914 |
+
"loss": 0.2125,
|
1915 |
+
"step": 2350
|
1916 |
+
},
|
1917 |
+
{
|
1918 |
+
"epoch": 29.5,
|
1919 |
+
"grad_norm": 0.4784901440143585,
|
1920 |
+
"learning_rate": 0.0009111111111111111,
|
1921 |
+
"loss": 0.1823,
|
1922 |
+
"step": 2360
|
1923 |
+
},
|
1924 |
+
{
|
1925 |
+
"epoch": 29.62,
|
1926 |
+
"grad_norm": 0.6156973242759705,
|
1927 |
+
"learning_rate": 0.0009055555555555556,
|
1928 |
+
"loss": 0.2289,
|
1929 |
+
"step": 2370
|
1930 |
+
},
|
1931 |
+
{
|
1932 |
+
"epoch": 29.75,
|
1933 |
+
"grad_norm": 0.4373360872268677,
|
1934 |
+
"learning_rate": 0.0009000000000000001,
|
1935 |
+
"loss": 0.2127,
|
1936 |
+
"step": 2380
|
1937 |
+
},
|
1938 |
+
{
|
1939 |
+
"epoch": 29.88,
|
1940 |
+
"grad_norm": 0.501115083694458,
|
1941 |
+
"learning_rate": 0.0008944444444444445,
|
1942 |
+
"loss": 0.2359,
|
1943 |
+
"step": 2390
|
1944 |
+
},
|
1945 |
+
{
|
1946 |
+
"epoch": 30.0,
|
1947 |
+
"grad_norm": 0.411662757396698,
|
1948 |
+
"learning_rate": 0.0008888888888888888,
|
1949 |
+
"loss": 0.197,
|
1950 |
+
"step": 2400
|
1951 |
+
},
|
1952 |
+
{
|
1953 |
+
"epoch": 30.0,
|
1954 |
+
"eval_accuracy": 0.95703125,
|
1955 |
+
"eval_loss": 0.17402663826942444,
|
1956 |
+
"eval_runtime": 4.919,
|
1957 |
+
"eval_samples_per_second": 208.172,
|
1958 |
+
"eval_steps_per_second": 13.011,
|
1959 |
+
"step": 2400
|
1960 |
+
},
|
1961 |
+
{
|
1962 |
+
"epoch": 30.12,
|
1963 |
+
"grad_norm": 0.45976510643959045,
|
1964 |
+
"learning_rate": 0.0008833333333333333,
|
1965 |
+
"loss": 0.1608,
|
1966 |
+
"step": 2410
|
1967 |
+
},
|
1968 |
+
{
|
1969 |
+
"epoch": 30.25,
|
1970 |
+
"grad_norm": 0.3243074417114258,
|
1971 |
+
"learning_rate": 0.0008777777777777778,
|
1972 |
+
"loss": 0.1742,
|
1973 |
+
"step": 2420
|
1974 |
+
},
|
1975 |
+
{
|
1976 |
+
"epoch": 30.38,
|
1977 |
+
"grad_norm": 0.5205725431442261,
|
1978 |
+
"learning_rate": 0.0008722222222222223,
|
1979 |
+
"loss": 0.1718,
|
1980 |
+
"step": 2430
|
1981 |
+
},
|
1982 |
+
{
|
1983 |
+
"epoch": 30.5,
|
1984 |
+
"grad_norm": 0.3976719081401825,
|
1985 |
+
"learning_rate": 0.0008666666666666667,
|
1986 |
+
"loss": 0.2247,
|
1987 |
+
"step": 2440
|
1988 |
+
},
|
1989 |
+
{
|
1990 |
+
"epoch": 30.62,
|
1991 |
+
"grad_norm": 0.2859196662902832,
|
1992 |
+
"learning_rate": 0.0008611111111111112,
|
1993 |
+
"loss": 0.1884,
|
1994 |
+
"step": 2450
|
1995 |
+
},
|
1996 |
+
{
|
1997 |
+
"epoch": 30.75,
|
1998 |
+
"grad_norm": 0.5310297012329102,
|
1999 |
+
"learning_rate": 0.0008555555555555556,
|
2000 |
+
"loss": 0.1672,
|
2001 |
+
"step": 2460
|
2002 |
+
},
|
2003 |
+
{
|
2004 |
+
"epoch": 30.88,
|
2005 |
+
"grad_norm": 0.5172590613365173,
|
2006 |
+
"learning_rate": 0.00085,
|
2007 |
+
"loss": 0.1828,
|
2008 |
+
"step": 2470
|
2009 |
+
},
|
2010 |
+
{
|
2011 |
+
"epoch": 31.0,
|
2012 |
+
"grad_norm": 0.6098745465278625,
|
2013 |
+
"learning_rate": 0.0008444444444444444,
|
2014 |
+
"loss": 0.1654,
|
2015 |
+
"step": 2480
|
2016 |
+
},
|
2017 |
+
{
|
2018 |
+
"epoch": 31.0,
|
2019 |
+
"eval_accuracy": 0.966796875,
|
2020 |
+
"eval_loss": 0.12167137861251831,
|
2021 |
+
"eval_runtime": 4.9956,
|
2022 |
+
"eval_samples_per_second": 204.98,
|
2023 |
+
"eval_steps_per_second": 12.811,
|
2024 |
+
"step": 2480
|
2025 |
+
},
|
2026 |
+
{
|
2027 |
+
"epoch": 31.12,
|
2028 |
+
"grad_norm": 0.3343498706817627,
|
2029 |
+
"learning_rate": 0.0008388888888888889,
|
2030 |
+
"loss": 0.1784,
|
2031 |
+
"step": 2490
|
2032 |
+
},
|
2033 |
+
{
|
2034 |
+
"epoch": 31.25,
|
2035 |
+
"grad_norm": 0.3938640058040619,
|
2036 |
+
"learning_rate": 0.0008333333333333334,
|
2037 |
+
"loss": 0.1697,
|
2038 |
+
"step": 2500
|
2039 |
+
},
|
2040 |
+
{
|
2041 |
+
"epoch": 31.38,
|
2042 |
+
"grad_norm": 0.41868484020233154,
|
2043 |
+
"learning_rate": 0.0008277777777777778,
|
2044 |
+
"loss": 0.2263,
|
2045 |
+
"step": 2510
|
2046 |
+
},
|
2047 |
+
{
|
2048 |
+
"epoch": 31.5,
|
2049 |
+
"grad_norm": 0.4363801181316376,
|
2050 |
+
"learning_rate": 0.0008222222222222222,
|
2051 |
+
"loss": 0.1762,
|
2052 |
+
"step": 2520
|
2053 |
+
},
|
2054 |
+
{
|
2055 |
+
"epoch": 31.62,
|
2056 |
+
"grad_norm": 0.5088948607444763,
|
2057 |
+
"learning_rate": 0.0008166666666666667,
|
2058 |
+
"loss": 0.1711,
|
2059 |
+
"step": 2530
|
2060 |
+
},
|
2061 |
+
{
|
2062 |
+
"epoch": 31.75,
|
2063 |
+
"grad_norm": 0.5423977375030518,
|
2064 |
+
"learning_rate": 0.0008111111111111111,
|
2065 |
+
"loss": 0.1675,
|
2066 |
+
"step": 2540
|
2067 |
+
},
|
2068 |
+
{
|
2069 |
+
"epoch": 31.88,
|
2070 |
+
"grad_norm": 0.431382954120636,
|
2071 |
+
"learning_rate": 0.0008055555555555556,
|
2072 |
+
"loss": 0.2216,
|
2073 |
+
"step": 2550
|
2074 |
+
},
|
2075 |
+
{
|
2076 |
+
"epoch": 32.0,
|
2077 |
+
"grad_norm": 0.4037337303161621,
|
2078 |
+
"learning_rate": 0.0008,
|
2079 |
+
"loss": 0.1896,
|
2080 |
+
"step": 2560
|
2081 |
+
},
|
2082 |
+
{
|
2083 |
+
"epoch": 32.0,
|
2084 |
+
"eval_accuracy": 0.92578125,
|
2085 |
+
"eval_loss": 0.2552070617675781,
|
2086 |
+
"eval_runtime": 5.2019,
|
2087 |
+
"eval_samples_per_second": 196.853,
|
2088 |
+
"eval_steps_per_second": 12.303,
|
2089 |
+
"step": 2560
|
2090 |
+
},
|
2091 |
+
{
|
2092 |
+
"epoch": 32.12,
|
2093 |
+
"grad_norm": 0.6025939583778381,
|
2094 |
+
"learning_rate": 0.0007944444444444444,
|
2095 |
+
"loss": 0.1926,
|
2096 |
+
"step": 2570
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"epoch": 32.25,
|
2100 |
+
"grad_norm": 0.7205588221549988,
|
2101 |
+
"learning_rate": 0.0007888888888888889,
|
2102 |
+
"loss": 0.1755,
|
2103 |
+
"step": 2580
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 32.38,
|
2107 |
+
"grad_norm": 0.3841509222984314,
|
2108 |
+
"learning_rate": 0.0007833333333333334,
|
2109 |
+
"loss": 0.1696,
|
2110 |
+
"step": 2590
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"epoch": 32.5,
|
2114 |
+
"grad_norm": 0.5659075975418091,
|
2115 |
+
"learning_rate": 0.0007777777777777778,
|
2116 |
+
"loss": 0.133,
|
2117 |
+
"step": 2600
|
2118 |
+
},
|
2119 |
+
{
|
2120 |
+
"epoch": 32.62,
|
2121 |
+
"grad_norm": 0.7011501789093018,
|
2122 |
+
"learning_rate": 0.0007722222222222223,
|
2123 |
+
"loss": 0.2069,
|
2124 |
+
"step": 2610
|
2125 |
+
},
|
2126 |
+
{
|
2127 |
+
"epoch": 32.75,
|
2128 |
+
"grad_norm": 0.5933576822280884,
|
2129 |
+
"learning_rate": 0.0007666666666666667,
|
2130 |
+
"loss": 0.1799,
|
2131 |
+
"step": 2620
|
2132 |
+
},
|
2133 |
+
{
|
2134 |
+
"epoch": 32.88,
|
2135 |
+
"grad_norm": 0.636463463306427,
|
2136 |
+
"learning_rate": 0.0007611111111111111,
|
2137 |
+
"loss": 0.1884,
|
2138 |
+
"step": 2630
|
2139 |
+
},
|
2140 |
+
{
|
2141 |
+
"epoch": 33.0,
|
2142 |
+
"grad_norm": 0.36000609397888184,
|
2143 |
+
"learning_rate": 0.0007555555555555555,
|
2144 |
+
"loss": 0.1705,
|
2145 |
+
"step": 2640
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 33.0,
|
2149 |
+
"eval_accuracy": 0.97265625,
|
2150 |
+
"eval_loss": 0.10305143892765045,
|
2151 |
+
"eval_runtime": 4.8746,
|
2152 |
+
"eval_samples_per_second": 210.07,
|
2153 |
+
"eval_steps_per_second": 13.129,
|
2154 |
+
"step": 2640
|
2155 |
+
},
|
2156 |
+
{
|
2157 |
+
"epoch": 33.12,
|
2158 |
+
"grad_norm": 0.25941601395606995,
|
2159 |
+
"learning_rate": 0.00075,
|
2160 |
+
"loss": 0.143,
|
2161 |
+
"step": 2650
|
2162 |
+
},
|
2163 |
+
{
|
2164 |
+
"epoch": 33.25,
|
2165 |
+
"grad_norm": 0.6486319899559021,
|
2166 |
+
"learning_rate": 0.0007444444444444445,
|
2167 |
+
"loss": 0.1819,
|
2168 |
+
"step": 2660
|
2169 |
+
},
|
2170 |
+
{
|
2171 |
+
"epoch": 33.38,
|
2172 |
+
"grad_norm": 0.34492290019989014,
|
2173 |
+
"learning_rate": 0.000738888888888889,
|
2174 |
+
"loss": 0.1877,
|
2175 |
+
"step": 2670
|
2176 |
+
},
|
2177 |
+
{
|
2178 |
+
"epoch": 33.5,
|
2179 |
+
"grad_norm": 0.5475990176200867,
|
2180 |
+
"learning_rate": 0.0007333333333333333,
|
2181 |
+
"loss": 0.1586,
|
2182 |
+
"step": 2680
|
2183 |
+
},
|
2184 |
+
{
|
2185 |
+
"epoch": 33.62,
|
2186 |
+
"grad_norm": 0.231631800532341,
|
2187 |
+
"learning_rate": 0.0007277777777777777,
|
2188 |
+
"loss": 0.145,
|
2189 |
+
"step": 2690
|
2190 |
+
},
|
2191 |
+
{
|
2192 |
+
"epoch": 33.75,
|
2193 |
+
"grad_norm": 0.6208530068397522,
|
2194 |
+
"learning_rate": 0.0007222222222222222,
|
2195 |
+
"loss": 0.2015,
|
2196 |
+
"step": 2700
|
2197 |
+
},
|
2198 |
+
{
|
2199 |
+
"epoch": 33.88,
|
2200 |
+
"grad_norm": 0.7229673862457275,
|
2201 |
+
"learning_rate": 0.0007166666666666667,
|
2202 |
+
"loss": 0.1814,
|
2203 |
+
"step": 2710
|
2204 |
+
},
|
2205 |
+
{
|
2206 |
+
"epoch": 34.0,
|
2207 |
+
"grad_norm": 0.38056522607803345,
|
2208 |
+
"learning_rate": 0.0007111111111111111,
|
2209 |
+
"loss": 0.1689,
|
2210 |
+
"step": 2720
|
2211 |
+
},
|
2212 |
+
{
|
2213 |
+
"epoch": 34.0,
|
2214 |
+
"eval_accuracy": 0.96875,
|
2215 |
+
"eval_loss": 0.10111749172210693,
|
2216 |
+
"eval_runtime": 5.2922,
|
2217 |
+
"eval_samples_per_second": 193.491,
|
2218 |
+
"eval_steps_per_second": 12.093,
|
2219 |
+
"step": 2720
|
2220 |
+
},
|
2221 |
+
{
|
2222 |
+
"epoch": 34.12,
|
2223 |
+
"grad_norm": 0.5405479669570923,
|
2224 |
+
"learning_rate": 0.0007055555555555556,
|
2225 |
+
"loss": 0.16,
|
2226 |
+
"step": 2730
|
2227 |
+
},
|
2228 |
+
{
|
2229 |
+
"epoch": 34.25,
|
2230 |
+
"grad_norm": 0.5781314373016357,
|
2231 |
+
"learning_rate": 0.0007,
|
2232 |
+
"loss": 0.1598,
|
2233 |
+
"step": 2740
|
2234 |
+
},
|
2235 |
+
{
|
2236 |
+
"epoch": 34.38,
|
2237 |
+
"grad_norm": 0.33385559916496277,
|
2238 |
+
"learning_rate": 0.0006944444444444445,
|
2239 |
+
"loss": 0.1747,
|
2240 |
+
"step": 2750
|
2241 |
+
},
|
2242 |
+
{
|
2243 |
+
"epoch": 34.5,
|
2244 |
+
"grad_norm": 0.36587977409362793,
|
2245 |
+
"learning_rate": 0.000688888888888889,
|
2246 |
+
"loss": 0.1376,
|
2247 |
+
"step": 2760
|
2248 |
+
},
|
2249 |
+
{
|
2250 |
+
"epoch": 34.62,
|
2251 |
+
"grad_norm": 0.3459375202655792,
|
2252 |
+
"learning_rate": 0.0006833333333333333,
|
2253 |
+
"loss": 0.1297,
|
2254 |
+
"step": 2770
|
2255 |
+
},
|
2256 |
+
{
|
2257 |
+
"epoch": 34.75,
|
2258 |
+
"grad_norm": 0.5182803273200989,
|
2259 |
+
"learning_rate": 0.0006777777777777778,
|
2260 |
+
"loss": 0.1747,
|
2261 |
+
"step": 2780
|
2262 |
+
},
|
2263 |
+
{
|
2264 |
+
"epoch": 34.88,
|
2265 |
+
"grad_norm": 0.39014366269111633,
|
2266 |
+
"learning_rate": 0.0006722222222222223,
|
2267 |
+
"loss": 0.169,
|
2268 |
+
"step": 2790
|
2269 |
+
},
|
2270 |
+
{
|
2271 |
+
"epoch": 35.0,
|
2272 |
+
"grad_norm": 0.4516375660896301,
|
2273 |
+
"learning_rate": 0.0006666666666666666,
|
2274 |
+
"loss": 0.1439,
|
2275 |
+
"step": 2800
|
2276 |
+
},
|
2277 |
+
{
|
2278 |
+
"epoch": 35.0,
|
2279 |
+
"eval_accuracy": 0.96484375,
|
2280 |
+
"eval_loss": 0.11748197674751282,
|
2281 |
+
"eval_runtime": 4.909,
|
2282 |
+
"eval_samples_per_second": 208.595,
|
2283 |
+
"eval_steps_per_second": 13.037,
|
2284 |
+
"step": 2800
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 35.12,
|
2288 |
+
"grad_norm": 0.47782474756240845,
|
2289 |
+
"learning_rate": 0.0006611111111111111,
|
2290 |
+
"loss": 0.1417,
|
2291 |
+
"step": 2810
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 35.25,
|
2295 |
+
"grad_norm": 0.11640643328428268,
|
2296 |
+
"learning_rate": 0.0006555555555555556,
|
2297 |
+
"loss": 0.1226,
|
2298 |
+
"step": 2820
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 35.38,
|
2302 |
+
"grad_norm": 0.4363173544406891,
|
2303 |
+
"learning_rate": 0.0006500000000000001,
|
2304 |
+
"loss": 0.14,
|
2305 |
+
"step": 2830
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 35.5,
|
2309 |
+
"grad_norm": 0.6676026582717896,
|
2310 |
+
"learning_rate": 0.0006444444444444444,
|
2311 |
+
"loss": 0.1548,
|
2312 |
+
"step": 2840
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 35.62,
|
2316 |
+
"grad_norm": 0.4940982162952423,
|
2317 |
+
"learning_rate": 0.0006388888888888888,
|
2318 |
+
"loss": 0.1554,
|
2319 |
+
"step": 2850
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 35.75,
|
2323 |
+
"grad_norm": 0.6478282809257507,
|
2324 |
+
"learning_rate": 0.0006333333333333333,
|
2325 |
+
"loss": 0.1641,
|
2326 |
+
"step": 2860
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 35.88,
|
2330 |
+
"grad_norm": 0.6007707715034485,
|
2331 |
+
"learning_rate": 0.0006277777777777778,
|
2332 |
+
"loss": 0.1484,
|
2333 |
+
"step": 2870
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 36.0,
|
2337 |
+
"grad_norm": 0.4945576786994934,
|
2338 |
+
"learning_rate": 0.0006222222222222223,
|
2339 |
+
"loss": 0.1606,
|
2340 |
+
"step": 2880
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 36.0,
|
2344 |
+
"eval_accuracy": 0.9443359375,
|
2345 |
+
"eval_loss": 0.18046385049819946,
|
2346 |
+
"eval_runtime": 5.0626,
|
2347 |
+
"eval_samples_per_second": 202.266,
|
2348 |
+
"eval_steps_per_second": 12.642,
|
2349 |
+
"step": 2880
|
2350 |
+
},
|
2351 |
+
{
|
2352 |
+
"epoch": 36.12,
|
2353 |
+
"grad_norm": 0.4033058285713196,
|
2354 |
+
"learning_rate": 0.0006166666666666667,
|
2355 |
+
"loss": 0.1372,
|
2356 |
+
"step": 2890
|
2357 |
+
},
|
2358 |
+
{
|
2359 |
+
"epoch": 36.25,
|
2360 |
+
"grad_norm": 0.30507412552833557,
|
2361 |
+
"learning_rate": 0.0006111111111111112,
|
2362 |
+
"loss": 0.1549,
|
2363 |
+
"step": 2900
|
2364 |
+
},
|
2365 |
+
{
|
2366 |
+
"epoch": 36.38,
|
2367 |
+
"grad_norm": 0.3899296820163727,
|
2368 |
+
"learning_rate": 0.0006055555555555556,
|
2369 |
+
"loss": 0.1667,
|
2370 |
+
"step": 2910
|
2371 |
+
},
|
2372 |
+
{
|
2373 |
+
"epoch": 36.5,
|
2374 |
+
"grad_norm": 0.44058963656425476,
|
2375 |
+
"learning_rate": 0.0006,
|
2376 |
+
"loss": 0.1712,
|
2377 |
+
"step": 2920
|
2378 |
+
},
|
2379 |
+
{
|
2380 |
+
"epoch": 36.62,
|
2381 |
+
"grad_norm": 0.4805178642272949,
|
2382 |
+
"learning_rate": 0.0005944444444444444,
|
2383 |
+
"loss": 0.1805,
|
2384 |
+
"step": 2930
|
2385 |
+
},
|
2386 |
+
{
|
2387 |
+
"epoch": 36.75,
|
2388 |
+
"grad_norm": 0.37880581617355347,
|
2389 |
+
"learning_rate": 0.0005888888888888889,
|
2390 |
+
"loss": 0.1411,
|
2391 |
+
"step": 2940
|
2392 |
+
},
|
2393 |
+
{
|
2394 |
+
"epoch": 36.88,
|
2395 |
+
"grad_norm": 0.4263412654399872,
|
2396 |
+
"learning_rate": 0.0005833333333333334,
|
2397 |
+
"loss": 0.1714,
|
2398 |
+
"step": 2950
|
2399 |
+
},
|
2400 |
+
{
|
2401 |
+
"epoch": 37.0,
|
2402 |
+
"grad_norm": 0.2723836898803711,
|
2403 |
+
"learning_rate": 0.0005777777777777778,
|
2404 |
+
"loss": 0.1281,
|
2405 |
+
"step": 2960
|
2406 |
+
},
|
2407 |
+
{
|
2408 |
+
"epoch": 37.0,
|
2409 |
+
"eval_accuracy": 0.9677734375,
|
2410 |
+
"eval_loss": 0.1253870278596878,
|
2411 |
+
"eval_runtime": 5.1275,
|
2412 |
+
"eval_samples_per_second": 199.709,
|
2413 |
+
"eval_steps_per_second": 12.482,
|
2414 |
+
"step": 2960
|
2415 |
+
},
|
2416 |
+
{
|
2417 |
+
"epoch": 37.12,
|
2418 |
+
"grad_norm": 0.4946765601634979,
|
2419 |
+
"learning_rate": 0.0005722222222222222,
|
2420 |
+
"loss": 0.1059,
|
2421 |
+
"step": 2970
|
2422 |
+
},
|
2423 |
+
{
|
2424 |
+
"epoch": 37.25,
|
2425 |
+
"grad_norm": 0.4709372818470001,
|
2426 |
+
"learning_rate": 0.0005666666666666667,
|
2427 |
+
"loss": 0.1321,
|
2428 |
+
"step": 2980
|
2429 |
+
},
|
2430 |
+
{
|
2431 |
+
"epoch": 37.38,
|
2432 |
+
"grad_norm": 0.36459285020828247,
|
2433 |
+
"learning_rate": 0.0005611111111111111,
|
2434 |
+
"loss": 0.1351,
|
2435 |
+
"step": 2990
|
2436 |
+
},
|
2437 |
+
{
|
2438 |
+
"epoch": 37.5,
|
2439 |
+
"grad_norm": 0.4145031273365021,
|
2440 |
+
"learning_rate": 0.0005555555555555556,
|
2441 |
+
"loss": 0.1545,
|
2442 |
+
"step": 3000
|
2443 |
+
},
|
2444 |
+
{
|
2445 |
+
"epoch": 37.62,
|
2446 |
+
"grad_norm": 0.5457221865653992,
|
2447 |
+
"learning_rate": 0.00055,
|
2448 |
+
"loss": 0.1424,
|
2449 |
+
"step": 3010
|
2450 |
+
},
|
2451 |
+
{
|
2452 |
+
"epoch": 37.75,
|
2453 |
+
"grad_norm": 0.5123695731163025,
|
2454 |
+
"learning_rate": 0.0005444444444444444,
|
2455 |
+
"loss": 0.1508,
|
2456 |
+
"step": 3020
|
2457 |
+
},
|
2458 |
+
{
|
2459 |
+
"epoch": 37.88,
|
2460 |
+
"grad_norm": 0.29368171095848083,
|
2461 |
+
"learning_rate": 0.0005388888888888889,
|
2462 |
+
"loss": 0.1438,
|
2463 |
+
"step": 3030
|
2464 |
+
},
|
2465 |
+
{
|
2466 |
+
"epoch": 38.0,
|
2467 |
+
"grad_norm": 0.6859858632087708,
|
2468 |
+
"learning_rate": 0.0005333333333333334,
|
2469 |
+
"loss": 0.1518,
|
2470 |
+
"step": 3040
|
2471 |
+
},
|
2472 |
+
{
|
2473 |
+
"epoch": 38.0,
|
2474 |
+
"eval_accuracy": 0.96484375,
|
2475 |
+
"eval_loss": 0.11837992072105408,
|
2476 |
+
"eval_runtime": 4.9042,
|
2477 |
+
"eval_samples_per_second": 208.8,
|
2478 |
+
"eval_steps_per_second": 13.05,
|
2479 |
+
"step": 3040
|
2480 |
+
},
|
2481 |
+
{
|
2482 |
+
"epoch": 38.12,
|
2483 |
+
"grad_norm": 0.3859548270702362,
|
2484 |
+
"learning_rate": 0.0005277777777777778,
|
2485 |
+
"loss": 0.1455,
|
2486 |
+
"step": 3050
|
2487 |
+
},
|
2488 |
+
{
|
2489 |
+
"epoch": 38.25,
|
2490 |
+
"grad_norm": 0.21001270413398743,
|
2491 |
+
"learning_rate": 0.0005222222222222223,
|
2492 |
+
"loss": 0.13,
|
2493 |
+
"step": 3060
|
2494 |
+
},
|
2495 |
+
{
|
2496 |
+
"epoch": 38.38,
|
2497 |
+
"grad_norm": 0.4814240038394928,
|
2498 |
+
"learning_rate": 0.0005166666666666667,
|
2499 |
+
"loss": 0.1291,
|
2500 |
+
"step": 3070
|
2501 |
+
},
|
2502 |
+
{
|
2503 |
+
"epoch": 38.5,
|
2504 |
+
"grad_norm": 0.4478558301925659,
|
2505 |
+
"learning_rate": 0.0005111111111111111,
|
2506 |
+
"loss": 0.1293,
|
2507 |
+
"step": 3080
|
2508 |
+
},
|
2509 |
+
{
|
2510 |
+
"epoch": 38.62,
|
2511 |
+
"grad_norm": 0.4811321496963501,
|
2512 |
+
"learning_rate": 0.0005055555555555555,
|
2513 |
+
"loss": 0.1231,
|
2514 |
+
"step": 3090
|
2515 |
+
},
|
2516 |
+
{
|
2517 |
+
"epoch": 38.75,
|
2518 |
+
"grad_norm": 0.2841961085796356,
|
2519 |
+
"learning_rate": 0.0005,
|
2520 |
+
"loss": 0.166,
|
2521 |
+
"step": 3100
|
2522 |
+
},
|
2523 |
+
{
|
2524 |
+
"epoch": 38.88,
|
2525 |
+
"grad_norm": 0.5479158759117126,
|
2526 |
+
"learning_rate": 0.0004944444444444445,
|
2527 |
+
"loss": 0.1044,
|
2528 |
+
"step": 3110
|
2529 |
+
},
|
2530 |
+
{
|
2531 |
+
"epoch": 39.0,
|
2532 |
+
"grad_norm": 0.37449321150779724,
|
2533 |
+
"learning_rate": 0.0004888888888888889,
|
2534 |
+
"loss": 0.1531,
|
2535 |
+
"step": 3120
|
2536 |
+
},
|
2537 |
+
{
|
2538 |
+
"epoch": 39.0,
|
2539 |
+
"eval_accuracy": 0.9736328125,
|
2540 |
+
"eval_loss": 0.09921471774578094,
|
2541 |
+
"eval_runtime": 5.3451,
|
2542 |
+
"eval_samples_per_second": 191.577,
|
2543 |
+
"eval_steps_per_second": 11.974,
|
2544 |
+
"step": 3120
|
2545 |
+
},
|
2546 |
+
{
|
2547 |
+
"epoch": 39.12,
|
2548 |
+
"grad_norm": 0.5961503386497498,
|
2549 |
+
"learning_rate": 0.00048333333333333334,
|
2550 |
+
"loss": 0.1321,
|
2551 |
+
"step": 3130
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"epoch": 39.25,
|
2555 |
+
"grad_norm": 0.3140615224838257,
|
2556 |
+
"learning_rate": 0.0004777777777777778,
|
2557 |
+
"loss": 0.1192,
|
2558 |
+
"step": 3140
|
2559 |
+
},
|
2560 |
+
{
|
2561 |
+
"epoch": 39.38,
|
2562 |
+
"grad_norm": 0.8949409127235413,
|
2563 |
+
"learning_rate": 0.00047222222222222224,
|
2564 |
+
"loss": 0.122,
|
2565 |
+
"step": 3150
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 39.5,
|
2569 |
+
"grad_norm": 0.21187840402126312,
|
2570 |
+
"learning_rate": 0.00046666666666666666,
|
2571 |
+
"loss": 0.1341,
|
2572 |
+
"step": 3160
|
2573 |
+
},
|
2574 |
+
{
|
2575 |
+
"epoch": 39.62,
|
2576 |
+
"grad_norm": 0.6364386081695557,
|
2577 |
+
"learning_rate": 0.00046111111111111114,
|
2578 |
+
"loss": 0.1327,
|
2579 |
+
"step": 3170
|
2580 |
+
},
|
2581 |
+
{
|
2582 |
+
"epoch": 39.75,
|
2583 |
+
"grad_norm": 0.2257820963859558,
|
2584 |
+
"learning_rate": 0.00045555555555555556,
|
2585 |
+
"loss": 0.1101,
|
2586 |
+
"step": 3180
|
2587 |
+
},
|
2588 |
+
{
|
2589 |
+
"epoch": 39.88,
|
2590 |
+
"grad_norm": 0.373692125082016,
|
2591 |
+
"learning_rate": 0.00045000000000000004,
|
2592 |
+
"loss": 0.1293,
|
2593 |
+
"step": 3190
|
2594 |
+
},
|
2595 |
+
{
|
2596 |
+
"epoch": 40.0,
|
2597 |
+
"grad_norm": 0.48990318179130554,
|
2598 |
+
"learning_rate": 0.0004444444444444444,
|
2599 |
+
"loss": 0.132,
|
2600 |
+
"step": 3200
|
2601 |
+
},
|
2602 |
+
{
|
2603 |
+
"epoch": 40.0,
|
2604 |
+
"eval_accuracy": 0.9775390625,
|
2605 |
+
"eval_loss": 0.09202806651592255,
|
2606 |
+
"eval_runtime": 4.9155,
|
2607 |
+
"eval_samples_per_second": 208.319,
|
2608 |
+
"eval_steps_per_second": 13.02,
|
2609 |
+
"step": 3200
|
2610 |
+
},
|
2611 |
+
{
|
2612 |
+
"epoch": 40.12,
|
2613 |
+
"grad_norm": 0.571524441242218,
|
2614 |
+
"learning_rate": 0.0004388888888888889,
|
2615 |
+
"loss": 0.1167,
|
2616 |
+
"step": 3210
|
2617 |
+
},
|
2618 |
+
{
|
2619 |
+
"epoch": 40.25,
|
2620 |
+
"grad_norm": 0.5896998643875122,
|
2621 |
+
"learning_rate": 0.00043333333333333337,
|
2622 |
+
"loss": 0.151,
|
2623 |
+
"step": 3220
|
2624 |
+
},
|
2625 |
+
{
|
2626 |
+
"epoch": 40.38,
|
2627 |
+
"grad_norm": 0.44366732239723206,
|
2628 |
+
"learning_rate": 0.0004277777777777778,
|
2629 |
+
"loss": 0.1422,
|
2630 |
+
"step": 3230
|
2631 |
+
},
|
2632 |
+
{
|
2633 |
+
"epoch": 40.5,
|
2634 |
+
"grad_norm": 0.314609169960022,
|
2635 |
+
"learning_rate": 0.0004222222222222222,
|
2636 |
+
"loss": 0.1253,
|
2637 |
+
"step": 3240
|
2638 |
+
},
|
2639 |
+
{
|
2640 |
+
"epoch": 40.62,
|
2641 |
+
"grad_norm": 0.3513747453689575,
|
2642 |
+
"learning_rate": 0.0004166666666666667,
|
2643 |
+
"loss": 0.119,
|
2644 |
+
"step": 3250
|
2645 |
+
},
|
2646 |
+
{
|
2647 |
+
"epoch": 40.75,
|
2648 |
+
"grad_norm": 0.3717803359031677,
|
2649 |
+
"learning_rate": 0.0004111111111111111,
|
2650 |
+
"loss": 0.1156,
|
2651 |
+
"step": 3260
|
2652 |
+
},
|
2653 |
+
{
|
2654 |
+
"epoch": 40.88,
|
2655 |
+
"grad_norm": 0.22342754900455475,
|
2656 |
+
"learning_rate": 0.00040555555555555554,
|
2657 |
+
"loss": 0.1372,
|
2658 |
+
"step": 3270
|
2659 |
+
},
|
2660 |
+
{
|
2661 |
+
"epoch": 41.0,
|
2662 |
+
"grad_norm": 0.41738444566726685,
|
2663 |
+
"learning_rate": 0.0004,
|
2664 |
+
"loss": 0.134,
|
2665 |
+
"step": 3280
|
2666 |
+
},
|
2667 |
+
{
|
2668 |
+
"epoch": 41.0,
|
2669 |
+
"eval_accuracy": 0.9638671875,
|
2670 |
+
"eval_loss": 0.13908132910728455,
|
2671 |
+
"eval_runtime": 5.2165,
|
2672 |
+
"eval_samples_per_second": 196.299,
|
2673 |
+
"eval_steps_per_second": 12.269,
|
2674 |
+
"step": 3280
|
2675 |
+
},
|
2676 |
+
{
|
2677 |
+
"epoch": 41.12,
|
2678 |
+
"grad_norm": 0.41814348101615906,
|
2679 |
+
"learning_rate": 0.00039444444444444444,
|
2680 |
+
"loss": 0.1279,
|
2681 |
+
"step": 3290
|
2682 |
+
},
|
2683 |
+
{
|
2684 |
+
"epoch": 41.25,
|
2685 |
+
"grad_norm": 0.9678131937980652,
|
2686 |
+
"learning_rate": 0.0003888888888888889,
|
2687 |
+
"loss": 0.1398,
|
2688 |
+
"step": 3300
|
2689 |
+
},
|
2690 |
+
{
|
2691 |
+
"epoch": 41.38,
|
2692 |
+
"grad_norm": 0.6725767850875854,
|
2693 |
+
"learning_rate": 0.00038333333333333334,
|
2694 |
+
"loss": 0.1492,
|
2695 |
+
"step": 3310
|
2696 |
+
},
|
2697 |
+
{
|
2698 |
+
"epoch": 41.5,
|
2699 |
+
"grad_norm": 0.31534790992736816,
|
2700 |
+
"learning_rate": 0.00037777777777777777,
|
2701 |
+
"loss": 0.1119,
|
2702 |
+
"step": 3320
|
2703 |
+
},
|
2704 |
+
{
|
2705 |
+
"epoch": 41.62,
|
2706 |
+
"grad_norm": 0.632583737373352,
|
2707 |
+
"learning_rate": 0.00037222222222222225,
|
2708 |
+
"loss": 0.1131,
|
2709 |
+
"step": 3330
|
2710 |
+
},
|
2711 |
+
{
|
2712 |
+
"epoch": 41.75,
|
2713 |
+
"grad_norm": 0.6746741533279419,
|
2714 |
+
"learning_rate": 0.00036666666666666667,
|
2715 |
+
"loss": 0.1351,
|
2716 |
+
"step": 3340
|
2717 |
+
},
|
2718 |
+
{
|
2719 |
+
"epoch": 41.88,
|
2720 |
+
"grad_norm": 0.3400849997997284,
|
2721 |
+
"learning_rate": 0.0003611111111111111,
|
2722 |
+
"loss": 0.0815,
|
2723 |
+
"step": 3350
|
2724 |
+
},
|
2725 |
+
{
|
2726 |
+
"epoch": 42.0,
|
2727 |
+
"grad_norm": 0.5605281591415405,
|
2728 |
+
"learning_rate": 0.00035555555555555557,
|
2729 |
+
"loss": 0.1413,
|
2730 |
+
"step": 3360
|
2731 |
+
},
|
2732 |
+
{
|
2733 |
+
"epoch": 42.0,
|
2734 |
+
"eval_accuracy": 0.9716796875,
|
2735 |
+
"eval_loss": 0.11220287531614304,
|
2736 |
+
"eval_runtime": 5.0927,
|
2737 |
+
"eval_samples_per_second": 201.072,
|
2738 |
+
"eval_steps_per_second": 12.567,
|
2739 |
+
"step": 3360
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 42.12,
|
2743 |
+
"grad_norm": 0.5148097276687622,
|
2744 |
+
"learning_rate": 0.00035,
|
2745 |
+
"loss": 0.125,
|
2746 |
+
"step": 3370
|
2747 |
+
},
|
2748 |
+
{
|
2749 |
+
"epoch": 42.25,
|
2750 |
+
"grad_norm": 0.38650012016296387,
|
2751 |
+
"learning_rate": 0.0003444444444444445,
|
2752 |
+
"loss": 0.1209,
|
2753 |
+
"step": 3380
|
2754 |
+
},
|
2755 |
+
{
|
2756 |
+
"epoch": 42.38,
|
2757 |
+
"grad_norm": 0.3292187750339508,
|
2758 |
+
"learning_rate": 0.0003388888888888889,
|
2759 |
+
"loss": 0.1236,
|
2760 |
+
"step": 3390
|
2761 |
+
},
|
2762 |
+
{
|
2763 |
+
"epoch": 42.5,
|
2764 |
+
"grad_norm": 0.20681746304035187,
|
2765 |
+
"learning_rate": 0.0003333333333333333,
|
2766 |
+
"loss": 0.0973,
|
2767 |
+
"step": 3400
|
2768 |
+
},
|
2769 |
+
{
|
2770 |
+
"epoch": 42.62,
|
2771 |
+
"grad_norm": 0.33743348717689514,
|
2772 |
+
"learning_rate": 0.0003277777777777778,
|
2773 |
+
"loss": 0.1208,
|
2774 |
+
"step": 3410
|
2775 |
+
},
|
2776 |
+
{
|
2777 |
+
"epoch": 42.75,
|
2778 |
+
"grad_norm": 0.34158453345298767,
|
2779 |
+
"learning_rate": 0.0003222222222222222,
|
2780 |
+
"loss": 0.11,
|
2781 |
+
"step": 3420
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 42.88,
|
2785 |
+
"grad_norm": 0.5730062127113342,
|
2786 |
+
"learning_rate": 0.00031666666666666665,
|
2787 |
+
"loss": 0.1292,
|
2788 |
+
"step": 3430
|
2789 |
+
},
|
2790 |
+
{
|
2791 |
+
"epoch": 43.0,
|
2792 |
+
"grad_norm": 0.44954267144203186,
|
2793 |
+
"learning_rate": 0.0003111111111111111,
|
2794 |
+
"loss": 0.1097,
|
2795 |
+
"step": 3440
|
2796 |
+
},
|
2797 |
+
{
|
2798 |
+
"epoch": 43.0,
|
2799 |
+
"eval_accuracy": 0.9677734375,
|
2800 |
+
"eval_loss": 0.11706902086734772,
|
2801 |
+
"eval_runtime": 4.9202,
|
2802 |
+
"eval_samples_per_second": 208.12,
|
2803 |
+
"eval_steps_per_second": 13.007,
|
2804 |
+
"step": 3440
|
2805 |
+
},
|
2806 |
+
{
|
2807 |
+
"epoch": 43.12,
|
2808 |
+
"grad_norm": 0.25731635093688965,
|
2809 |
+
"learning_rate": 0.0003055555555555556,
|
2810 |
+
"loss": 0.1161,
|
2811 |
+
"step": 3450
|
2812 |
+
},
|
2813 |
+
{
|
2814 |
+
"epoch": 43.25,
|
2815 |
+
"grad_norm": 0.5329569578170776,
|
2816 |
+
"learning_rate": 0.0003,
|
2817 |
+
"loss": 0.1507,
|
2818 |
+
"step": 3460
|
2819 |
+
},
|
2820 |
+
{
|
2821 |
+
"epoch": 43.38,
|
2822 |
+
"grad_norm": 0.3034692704677582,
|
2823 |
+
"learning_rate": 0.00029444444444444445,
|
2824 |
+
"loss": 0.1447,
|
2825 |
+
"step": 3470
|
2826 |
+
},
|
2827 |
+
{
|
2828 |
+
"epoch": 43.5,
|
2829 |
+
"grad_norm": 0.5483482480049133,
|
2830 |
+
"learning_rate": 0.0002888888888888889,
|
2831 |
+
"loss": 0.1323,
|
2832 |
+
"step": 3480
|
2833 |
+
},
|
2834 |
+
{
|
2835 |
+
"epoch": 43.62,
|
2836 |
+
"grad_norm": 0.279697984457016,
|
2837 |
+
"learning_rate": 0.00028333333333333335,
|
2838 |
+
"loss": 0.1,
|
2839 |
+
"step": 3490
|
2840 |
+
},
|
2841 |
+
{
|
2842 |
+
"epoch": 43.75,
|
2843 |
+
"grad_norm": 0.5593113303184509,
|
2844 |
+
"learning_rate": 0.0002777777777777778,
|
2845 |
+
"loss": 0.1169,
|
2846 |
+
"step": 3500
|
2847 |
+
},
|
2848 |
+
{
|
2849 |
+
"epoch": 43.88,
|
2850 |
+
"grad_norm": 0.621919572353363,
|
2851 |
+
"learning_rate": 0.0002722222222222222,
|
2852 |
+
"loss": 0.1119,
|
2853 |
+
"step": 3510
|
2854 |
+
},
|
2855 |
+
{
|
2856 |
+
"epoch": 44.0,
|
2857 |
+
"grad_norm": 0.37898024916648865,
|
2858 |
+
"learning_rate": 0.0002666666666666667,
|
2859 |
+
"loss": 0.1167,
|
2860 |
+
"step": 3520
|
2861 |
+
},
|
2862 |
+
{
|
2863 |
+
"epoch": 44.0,
|
2864 |
+
"eval_accuracy": 0.9765625,
|
2865 |
+
"eval_loss": 0.10542036592960358,
|
2866 |
+
"eval_runtime": 5.3473,
|
2867 |
+
"eval_samples_per_second": 191.5,
|
2868 |
+
"eval_steps_per_second": 11.969,
|
2869 |
+
"step": 3520
|
2870 |
+
},
|
2871 |
+
{
|
2872 |
+
"epoch": 44.12,
|
2873 |
+
"grad_norm": 0.40025296807289124,
|
2874 |
+
"learning_rate": 0.00026111111111111116,
|
2875 |
+
"loss": 0.1107,
|
2876 |
+
"step": 3530
|
2877 |
+
},
|
2878 |
+
{
|
2879 |
+
"epoch": 44.25,
|
2880 |
+
"grad_norm": 0.19010861217975616,
|
2881 |
+
"learning_rate": 0.00025555555555555553,
|
2882 |
+
"loss": 0.1008,
|
2883 |
+
"step": 3540
|
2884 |
+
},
|
2885 |
+
{
|
2886 |
+
"epoch": 44.38,
|
2887 |
+
"grad_norm": 0.33224934339523315,
|
2888 |
+
"learning_rate": 0.00025,
|
2889 |
+
"loss": 0.1355,
|
2890 |
+
"step": 3550
|
2891 |
+
},
|
2892 |
+
{
|
2893 |
+
"epoch": 44.5,
|
2894 |
+
"grad_norm": 0.4298325181007385,
|
2895 |
+
"learning_rate": 0.00024444444444444443,
|
2896 |
+
"loss": 0.106,
|
2897 |
+
"step": 3560
|
2898 |
+
},
|
2899 |
+
{
|
2900 |
+
"epoch": 44.62,
|
2901 |
+
"grad_norm": 0.4320330023765564,
|
2902 |
+
"learning_rate": 0.0002388888888888889,
|
2903 |
+
"loss": 0.1053,
|
2904 |
+
"step": 3570
|
2905 |
+
},
|
2906 |
+
{
|
2907 |
+
"epoch": 44.75,
|
2908 |
+
"grad_norm": 0.1121302917599678,
|
2909 |
+
"learning_rate": 0.00023333333333333333,
|
2910 |
+
"loss": 0.0845,
|
2911 |
+
"step": 3580
|
2912 |
+
},
|
2913 |
+
{
|
2914 |
+
"epoch": 44.88,
|
2915 |
+
"grad_norm": 0.3021819293498993,
|
2916 |
+
"learning_rate": 0.00022777777777777778,
|
2917 |
+
"loss": 0.1222,
|
2918 |
+
"step": 3590
|
2919 |
+
},
|
2920 |
+
{
|
2921 |
+
"epoch": 45.0,
|
2922 |
+
"grad_norm": 0.7353653311729431,
|
2923 |
+
"learning_rate": 0.0002222222222222222,
|
2924 |
+
"loss": 0.1388,
|
2925 |
+
"step": 3600
|
2926 |
+
},
|
2927 |
+
{
|
2928 |
+
"epoch": 45.0,
|
2929 |
+
"eval_accuracy": 0.9794921875,
|
2930 |
+
"eval_loss": 0.09323666244745255,
|
2931 |
+
"eval_runtime": 4.9368,
|
2932 |
+
"eval_samples_per_second": 207.422,
|
2933 |
+
"eval_steps_per_second": 12.964,
|
2934 |
+
"step": 3600
|
2935 |
+
},
|
2936 |
+
{
|
2937 |
+
"epoch": 45.12,
|
2938 |
+
"grad_norm": 0.5964930057525635,
|
2939 |
+
"learning_rate": 0.00021666666666666668,
|
2940 |
+
"loss": 0.1201,
|
2941 |
+
"step": 3610
|
2942 |
+
},
|
2943 |
+
{
|
2944 |
+
"epoch": 45.25,
|
2945 |
+
"grad_norm": 0.17329342663288116,
|
2946 |
+
"learning_rate": 0.0002111111111111111,
|
2947 |
+
"loss": 0.0905,
|
2948 |
+
"step": 3620
|
2949 |
+
},
|
2950 |
+
{
|
2951 |
+
"epoch": 45.38,
|
2952 |
+
"grad_norm": 0.5378609299659729,
|
2953 |
+
"learning_rate": 0.00020555555555555556,
|
2954 |
+
"loss": 0.0981,
|
2955 |
+
"step": 3630
|
2956 |
+
},
|
2957 |
+
{
|
2958 |
+
"epoch": 45.5,
|
2959 |
+
"grad_norm": 0.3457593619823456,
|
2960 |
+
"learning_rate": 0.0002,
|
2961 |
+
"loss": 0.1116,
|
2962 |
+
"step": 3640
|
2963 |
+
},
|
2964 |
+
{
|
2965 |
+
"epoch": 45.62,
|
2966 |
+
"grad_norm": 0.5954685211181641,
|
2967 |
+
"learning_rate": 0.00019444444444444446,
|
2968 |
+
"loss": 0.1037,
|
2969 |
+
"step": 3650
|
2970 |
+
},
|
2971 |
+
{
|
2972 |
+
"epoch": 45.75,
|
2973 |
+
"grad_norm": 0.1786712259054184,
|
2974 |
+
"learning_rate": 0.00018888888888888888,
|
2975 |
+
"loss": 0.0978,
|
2976 |
+
"step": 3660
|
2977 |
+
},
|
2978 |
+
{
|
2979 |
+
"epoch": 45.88,
|
2980 |
+
"grad_norm": 0.25224894285202026,
|
2981 |
+
"learning_rate": 0.00018333333333333334,
|
2982 |
+
"loss": 0.1089,
|
2983 |
+
"step": 3670
|
2984 |
+
},
|
2985 |
+
{
|
2986 |
+
"epoch": 46.0,
|
2987 |
+
"grad_norm": 0.33607247471809387,
|
2988 |
+
"learning_rate": 0.00017777777777777779,
|
2989 |
+
"loss": 0.1221,
|
2990 |
+
"step": 3680
|
2991 |
+
},
|
2992 |
+
{
|
2993 |
+
"epoch": 46.0,
|
2994 |
+
"eval_accuracy": 0.9765625,
|
2995 |
+
"eval_loss": 0.09462323784828186,
|
2996 |
+
"eval_runtime": 5.2287,
|
2997 |
+
"eval_samples_per_second": 195.844,
|
2998 |
+
"eval_steps_per_second": 12.24,
|
2999 |
+
"step": 3680
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 46.12,
|
3003 |
+
"grad_norm": 0.34634700417518616,
|
3004 |
+
"learning_rate": 0.00017222222222222224,
|
3005 |
+
"loss": 0.1243,
|
3006 |
+
"step": 3690
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 46.25,
|
3010 |
+
"grad_norm": 0.5061681866645813,
|
3011 |
+
"learning_rate": 0.00016666666666666666,
|
3012 |
+
"loss": 0.1115,
|
3013 |
+
"step": 3700
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 46.38,
|
3017 |
+
"grad_norm": 0.2837713658809662,
|
3018 |
+
"learning_rate": 0.0001611111111111111,
|
3019 |
+
"loss": 0.1008,
|
3020 |
+
"step": 3710
|
3021 |
+
},
|
3022 |
+
{
|
3023 |
+
"epoch": 46.5,
|
3024 |
+
"grad_norm": 0.2688066363334656,
|
3025 |
+
"learning_rate": 0.00015555555555555556,
|
3026 |
+
"loss": 0.1058,
|
3027 |
+
"step": 3720
|
3028 |
+
},
|
3029 |
+
{
|
3030 |
+
"epoch": 46.62,
|
3031 |
+
"grad_norm": 0.32675421237945557,
|
3032 |
+
"learning_rate": 0.00015,
|
3033 |
+
"loss": 0.0897,
|
3034 |
+
"step": 3730
|
3035 |
+
},
|
3036 |
+
{
|
3037 |
+
"epoch": 46.75,
|
3038 |
+
"grad_norm": 0.6959260702133179,
|
3039 |
+
"learning_rate": 0.00014444444444444444,
|
3040 |
+
"loss": 0.1182,
|
3041 |
+
"step": 3740
|
3042 |
+
},
|
3043 |
+
{
|
3044 |
+
"epoch": 46.88,
|
3045 |
+
"grad_norm": 0.3018099069595337,
|
3046 |
+
"learning_rate": 0.0001388888888888889,
|
3047 |
+
"loss": 0.1013,
|
3048 |
+
"step": 3750
|
3049 |
+
},
|
3050 |
+
{
|
3051 |
+
"epoch": 47.0,
|
3052 |
+
"grad_norm": 0.6018778085708618,
|
3053 |
+
"learning_rate": 0.00013333333333333334,
|
3054 |
+
"loss": 0.1099,
|
3055 |
+
"step": 3760
|
3056 |
+
},
|
3057 |
+
{
|
3058 |
+
"epoch": 47.0,
|
3059 |
+
"eval_accuracy": 0.9755859375,
|
3060 |
+
"eval_loss": 0.1115545928478241,
|
3061 |
+
"eval_runtime": 5.0799,
|
3062 |
+
"eval_samples_per_second": 201.581,
|
3063 |
+
"eval_steps_per_second": 12.599,
|
3064 |
+
"step": 3760
|
3065 |
+
},
|
3066 |
+
{
|
3067 |
+
"epoch": 47.12,
|
3068 |
+
"grad_norm": 0.42199546098709106,
|
3069 |
+
"learning_rate": 0.00012777777777777776,
|
3070 |
+
"loss": 0.1073,
|
3071 |
+
"step": 3770
|
3072 |
+
},
|
3073 |
+
{
|
3074 |
+
"epoch": 47.25,
|
3075 |
+
"grad_norm": 0.6451756358146667,
|
3076 |
+
"learning_rate": 0.00012222222222222221,
|
3077 |
+
"loss": 0.099,
|
3078 |
+
"step": 3780
|
3079 |
+
},
|
3080 |
+
{
|
3081 |
+
"epoch": 47.38,
|
3082 |
+
"grad_norm": 0.4935210943222046,
|
3083 |
+
"learning_rate": 0.00011666666666666667,
|
3084 |
+
"loss": 0.1077,
|
3085 |
+
"step": 3790
|
3086 |
+
},
|
3087 |
+
{
|
3088 |
+
"epoch": 47.5,
|
3089 |
+
"grad_norm": 0.2563684582710266,
|
3090 |
+
"learning_rate": 0.0001111111111111111,
|
3091 |
+
"loss": 0.0907,
|
3092 |
+
"step": 3800
|
3093 |
+
},
|
3094 |
+
{
|
3095 |
+
"epoch": 47.62,
|
3096 |
+
"grad_norm": 0.3351310193538666,
|
3097 |
+
"learning_rate": 0.00010555555555555555,
|
3098 |
+
"loss": 0.1059,
|
3099 |
+
"step": 3810
|
3100 |
+
},
|
3101 |
+
{
|
3102 |
+
"epoch": 47.75,
|
3103 |
+
"grad_norm": 0.39526107907295227,
|
3104 |
+
"learning_rate": 0.0001,
|
3105 |
+
"loss": 0.0868,
|
3106 |
+
"step": 3820
|
3107 |
+
},
|
3108 |
+
{
|
3109 |
+
"epoch": 47.88,
|
3110 |
+
"grad_norm": 0.4634101390838623,
|
3111 |
+
"learning_rate": 9.444444444444444e-05,
|
3112 |
+
"loss": 0.1098,
|
3113 |
+
"step": 3830
|
3114 |
+
},
|
3115 |
+
{
|
3116 |
+
"epoch": 48.0,
|
3117 |
+
"grad_norm": 0.5983624458312988,
|
3118 |
+
"learning_rate": 8.888888888888889e-05,
|
3119 |
+
"loss": 0.1041,
|
3120 |
+
"step": 3840
|
3121 |
+
},
|
3122 |
+
{
|
3123 |
+
"epoch": 48.0,
|
3124 |
+
"eval_accuracy": 0.974609375,
|
3125 |
+
"eval_loss": 0.11264081299304962,
|
3126 |
+
"eval_runtime": 4.9279,
|
3127 |
+
"eval_samples_per_second": 207.795,
|
3128 |
+
"eval_steps_per_second": 12.987,
|
3129 |
+
"step": 3840
|
3130 |
+
},
|
3131 |
+
{
|
3132 |
+
"epoch": 48.12,
|
3133 |
+
"grad_norm": 0.4093017578125,
|
3134 |
+
"learning_rate": 8.333333333333333e-05,
|
3135 |
+
"loss": 0.1134,
|
3136 |
+
"step": 3850
|
3137 |
+
},
|
3138 |
+
{
|
3139 |
+
"epoch": 48.25,
|
3140 |
+
"grad_norm": 0.6668171286582947,
|
3141 |
+
"learning_rate": 7.777777777777778e-05,
|
3142 |
+
"loss": 0.0948,
|
3143 |
+
"step": 3860
|
3144 |
+
},
|
3145 |
+
{
|
3146 |
+
"epoch": 48.38,
|
3147 |
+
"grad_norm": 0.24066688120365143,
|
3148 |
+
"learning_rate": 7.222222222222222e-05,
|
3149 |
+
"loss": 0.0958,
|
3150 |
+
"step": 3870
|
3151 |
+
},
|
3152 |
+
{
|
3153 |
+
"epoch": 48.5,
|
3154 |
+
"grad_norm": 0.2770562469959259,
|
3155 |
+
"learning_rate": 6.666666666666667e-05,
|
3156 |
+
"loss": 0.1021,
|
3157 |
+
"step": 3880
|
3158 |
+
},
|
3159 |
+
{
|
3160 |
+
"epoch": 48.62,
|
3161 |
+
"grad_norm": 0.45978790521621704,
|
3162 |
+
"learning_rate": 6.111111111111111e-05,
|
3163 |
+
"loss": 0.1084,
|
3164 |
+
"step": 3890
|
3165 |
+
},
|
3166 |
+
{
|
3167 |
+
"epoch": 48.75,
|
3168 |
+
"grad_norm": 0.594672441482544,
|
3169 |
+
"learning_rate": 5.555555555555555e-05,
|
3170 |
+
"loss": 0.1373,
|
3171 |
+
"step": 3900
|
3172 |
+
},
|
3173 |
+
{
|
3174 |
+
"epoch": 48.88,
|
3175 |
+
"grad_norm": 0.8167428374290466,
|
3176 |
+
"learning_rate": 5e-05,
|
3177 |
+
"loss": 0.1038,
|
3178 |
+
"step": 3910
|
3179 |
+
},
|
3180 |
+
{
|
3181 |
+
"epoch": 49.0,
|
3182 |
+
"grad_norm": 0.2987329661846161,
|
3183 |
+
"learning_rate": 4.4444444444444447e-05,
|
3184 |
+
"loss": 0.1025,
|
3185 |
+
"step": 3920
|
3186 |
+
},
|
3187 |
+
{
|
3188 |
+
"epoch": 49.0,
|
3189 |
+
"eval_accuracy": 0.9755859375,
|
3190 |
+
"eval_loss": 0.11138872057199478,
|
3191 |
+
"eval_runtime": 5.3184,
|
3192 |
+
"eval_samples_per_second": 192.54,
|
3193 |
+
"eval_steps_per_second": 12.034,
|
3194 |
+
"step": 3920
|
3195 |
+
},
|
3196 |
+
{
|
3197 |
+
"epoch": 49.12,
|
3198 |
+
"grad_norm": 0.3884102404117584,
|
3199 |
+
"learning_rate": 3.888888888888889e-05,
|
3200 |
+
"loss": 0.1018,
|
3201 |
+
"step": 3930
|
3202 |
+
},
|
3203 |
+
{
|
3204 |
+
"epoch": 49.25,
|
3205 |
+
"grad_norm": 0.2661769688129425,
|
3206 |
+
"learning_rate": 3.3333333333333335e-05,
|
3207 |
+
"loss": 0.1011,
|
3208 |
+
"step": 3940
|
3209 |
+
},
|
3210 |
+
{
|
3211 |
+
"epoch": 49.38,
|
3212 |
+
"grad_norm": 0.40820014476776123,
|
3213 |
+
"learning_rate": 2.7777777777777776e-05,
|
3214 |
+
"loss": 0.1488,
|
3215 |
+
"step": 3950
|
3216 |
+
},
|
3217 |
+
{
|
3218 |
+
"epoch": 49.5,
|
3219 |
+
"grad_norm": 0.46163231134414673,
|
3220 |
+
"learning_rate": 2.2222222222222223e-05,
|
3221 |
+
"loss": 0.1258,
|
3222 |
+
"step": 3960
|
3223 |
+
},
|
3224 |
+
{
|
3225 |
+
"epoch": 49.62,
|
3226 |
+
"grad_norm": 0.4315054416656494,
|
3227 |
+
"learning_rate": 1.6666666666666667e-05,
|
3228 |
+
"loss": 0.1018,
|
3229 |
+
"step": 3970
|
3230 |
+
},
|
3231 |
+
{
|
3232 |
+
"epoch": 49.75,
|
3233 |
+
"grad_norm": 0.2365369200706482,
|
3234 |
+
"learning_rate": 1.1111111111111112e-05,
|
3235 |
+
"loss": 0.0977,
|
3236 |
+
"step": 3980
|
3237 |
+
},
|
3238 |
+
{
|
3239 |
+
"epoch": 49.88,
|
3240 |
+
"grad_norm": 0.4910149574279785,
|
3241 |
+
"learning_rate": 5.555555555555556e-06,
|
3242 |
+
"loss": 0.1122,
|
3243 |
+
"step": 3990
|
3244 |
+
},
|
3245 |
+
{
|
3246 |
+
"epoch": 50.0,
|
3247 |
+
"grad_norm": 0.2623092234134674,
|
3248 |
+
"learning_rate": 0.0,
|
3249 |
+
"loss": 0.0887,
|
3250 |
+
"step": 4000
|
3251 |
+
},
|
3252 |
+
{
|
3253 |
+
"epoch": 50.0,
|
3254 |
+
"eval_accuracy": 0.9755859375,
|
3255 |
+
"eval_loss": 0.10555899888277054,
|
3256 |
+
"eval_runtime": 4.9229,
|
3257 |
+
"eval_samples_per_second": 208.005,
|
3258 |
+
"eval_steps_per_second": 13.0,
|
3259 |
+
"step": 4000
|
3260 |
+
},
|
3261 |
+
{
|
3262 |
+
"epoch": 50.0,
|
3263 |
+
"step": 4000,
|
3264 |
+
"total_flos": 5.437210780237824e+18,
|
3265 |
+
"train_loss": 0.3629864407479763,
|
3266 |
+
"train_runtime": 3465.6999,
|
3267 |
+
"train_samples_per_second": 73.867,
|
3268 |
+
"train_steps_per_second": 1.154
|
3269 |
+
}
|
3270 |
+
],
|
3271 |
+
"logging_steps": 10,
|
3272 |
+
"max_steps": 4000,
|
3273 |
+
"num_input_tokens_seen": 0,
|
3274 |
+
"num_train_epochs": 50,
|
3275 |
+
"save_steps": 500,
|
3276 |
+
"total_flos": 5.437210780237824e+18,
|
3277 |
+
"train_batch_size": 16,
|
3278 |
+
"trial_name": null,
|
3279 |
+
"trial_params": null
|
3280 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9ac0417085b4a2358ca8e0bc3a128e1354e847ed5e45152747b6de1925eaf689
|
3 |
+
size 4920
|