dhritic99/vit-base-brain-alzheimer-detection
Browse files- README.md +71 -0
- all_results.json +13 -0
- config.json +36 -0
- eval_results.json +8 -0
- model.safetensors +3 -0
- preprocessor_config.json +22 -0
- train_results.json +8 -0
- trainer_state.json +855 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- accuracy
|
| 8 |
+
model-index:
|
| 9 |
+
- name: vit-base-brain-alzheimer-detection
|
| 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 |
+
# vit-base-brain-alzheimer-detection
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
|
| 19 |
+
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.2360
|
| 21 |
+
- Accuracy: 0.9523
|
| 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: 2e-05
|
| 41 |
+
- train_batch_size: 16
|
| 42 |
+
- eval_batch_size: 8
|
| 43 |
+
- seed: 42
|
| 44 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
+
- lr_scheduler_type: linear
|
| 46 |
+
- lr_scheduler_warmup_steps: 500
|
| 47 |
+
- num_epochs: 20
|
| 48 |
+
- mixed_precision_training: Native AMP
|
| 49 |
+
|
| 50 |
+
### Training results
|
| 51 |
+
|
| 52 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 53 |
+
|:-------------:|:-------:|:----:|:---------------:|:--------:|
|
| 54 |
+
| 0.7278 | 1.9531 | 500 | 0.7124 | 0.7051 |
|
| 55 |
+
| 0.3023 | 3.9062 | 1000 | 0.3776 | 0.8828 |
|
| 56 |
+
| 0.0997 | 5.8594 | 1500 | 0.2808 | 0.9131 |
|
| 57 |
+
| 0.0424 | 7.8125 | 2000 | 0.1914 | 0.9570 |
|
| 58 |
+
| 0.0108 | 9.7656 | 2500 | 0.4534 | 0.8945 |
|
| 59 |
+
| 0.0088 | 11.7188 | 3000 | 0.1554 | 0.9580 |
|
| 60 |
+
| 0.0051 | 13.6719 | 3500 | 0.1666 | 0.9590 |
|
| 61 |
+
| 0.0039 | 15.625 | 4000 | 0.1544 | 0.9648 |
|
| 62 |
+
| 0.0034 | 17.5781 | 4500 | 0.1575 | 0.9648 |
|
| 63 |
+
| 0.003 | 19.5312 | 5000 | 0.1592 | 0.9658 |
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
### Framework versions
|
| 67 |
+
|
| 68 |
+
- Transformers 4.42.4
|
| 69 |
+
- Pytorch 2.4.0+cu121
|
| 70 |
+
- Datasets 2.21.0
|
| 71 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.95234375,
|
| 4 |
+
"eval_loss": 0.23598822951316833,
|
| 5 |
+
"eval_runtime": 6.8172,
|
| 6 |
+
"eval_samples_per_second": 187.761,
|
| 7 |
+
"eval_steps_per_second": 23.47,
|
| 8 |
+
"total_flos": 6.348257582037074e+18,
|
| 9 |
+
"train_loss": 0.15641047188546509,
|
| 10 |
+
"train_runtime": 1165.0601,
|
| 11 |
+
"train_samples_per_second": 70.314,
|
| 12 |
+
"train_steps_per_second": 4.395
|
| 13 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-base-patch16-224-in21k",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1",
|
| 14 |
+
"2": "LABEL_2",
|
| 15 |
+
"3": "LABEL_3"
|
| 16 |
+
},
|
| 17 |
+
"image_size": 224,
|
| 18 |
+
"initializer_range": 0.02,
|
| 19 |
+
"intermediate_size": 3072,
|
| 20 |
+
"label2id": {
|
| 21 |
+
"LABEL_0": 0,
|
| 22 |
+
"LABEL_1": 1,
|
| 23 |
+
"LABEL_2": 2,
|
| 24 |
+
"LABEL_3": 3
|
| 25 |
+
},
|
| 26 |
+
"layer_norm_eps": 1e-12,
|
| 27 |
+
"model_type": "vit",
|
| 28 |
+
"num_attention_heads": 12,
|
| 29 |
+
"num_channels": 3,
|
| 30 |
+
"num_hidden_layers": 12,
|
| 31 |
+
"patch_size": 16,
|
| 32 |
+
"problem_type": "single_label_classification",
|
| 33 |
+
"qkv_bias": true,
|
| 34 |
+
"torch_dtype": "float32",
|
| 35 |
+
"transformers_version": "4.42.4"
|
| 36 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.95234375,
|
| 4 |
+
"eval_loss": 0.23598822951316833,
|
| 5 |
+
"eval_runtime": 6.8172,
|
| 6 |
+
"eval_samples_per_second": 187.761,
|
| 7 |
+
"eval_steps_per_second": 23.47
|
| 8 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c716579276bcb7a7f84a427b96742340b475820942e9f4e1a701200fbe83b55
|
| 3 |
+
size 343230128
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": true,
|
| 3 |
+
"do_rescale": true,
|
| 4 |
+
"do_resize": true,
|
| 5 |
+
"image_mean": [
|
| 6 |
+
0.5,
|
| 7 |
+
0.5,
|
| 8 |
+
0.5
|
| 9 |
+
],
|
| 10 |
+
"image_processor_type": "ViTImageProcessor",
|
| 11 |
+
"image_std": [
|
| 12 |
+
0.5,
|
| 13 |
+
0.5,
|
| 14 |
+
0.5
|
| 15 |
+
],
|
| 16 |
+
"resample": 2,
|
| 17 |
+
"rescale_factor": 0.00392156862745098,
|
| 18 |
+
"size": {
|
| 19 |
+
"height": 224,
|
| 20 |
+
"width": 224
|
| 21 |
+
}
|
| 22 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 6.348257582037074e+18,
|
| 4 |
+
"train_loss": 0.15641047188546509,
|
| 5 |
+
"train_runtime": 1165.0601,
|
| 6 |
+
"train_samples_per_second": 70.314,
|
| 7 |
+
"train_steps_per_second": 4.395
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,855 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.15439653396606445,
|
| 3 |
+
"best_model_checkpoint": "./vit-base-brain-alzheimer-detection/checkpoint-4000",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 5120,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1953125,
|
| 13 |
+
"grad_norm": 3.019329786300659,
|
| 14 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 15 |
+
"loss": 1.3778,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.390625,
|
| 20 |
+
"grad_norm": 2.534724712371826,
|
| 21 |
+
"learning_rate": 4.000000000000001e-06,
|
| 22 |
+
"loss": 1.1895,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.5859375,
|
| 27 |
+
"grad_norm": 2.774301052093506,
|
| 28 |
+
"learning_rate": 6e-06,
|
| 29 |
+
"loss": 1.0423,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.78125,
|
| 34 |
+
"grad_norm": 3.057741165161133,
|
| 35 |
+
"learning_rate": 8.000000000000001e-06,
|
| 36 |
+
"loss": 0.9859,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.9765625,
|
| 41 |
+
"grad_norm": 2.8204987049102783,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 0.9008,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.171875,
|
| 48 |
+
"grad_norm": 1.7199639081954956,
|
| 49 |
+
"learning_rate": 1.2e-05,
|
| 50 |
+
"loss": 0.8892,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.3671875,
|
| 55 |
+
"grad_norm": 3.138557195663452,
|
| 56 |
+
"learning_rate": 1.4e-05,
|
| 57 |
+
"loss": 0.8273,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.5625,
|
| 62 |
+
"grad_norm": 3.1744067668914795,
|
| 63 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 64 |
+
"loss": 0.7807,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.7578125,
|
| 69 |
+
"grad_norm": 5.47639799118042,
|
| 70 |
+
"learning_rate": 1.8e-05,
|
| 71 |
+
"loss": 0.7546,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.953125,
|
| 76 |
+
"grad_norm": 2.743036985397339,
|
| 77 |
+
"learning_rate": 2e-05,
|
| 78 |
+
"loss": 0.7278,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.953125,
|
| 83 |
+
"eval_accuracy": 0.705078125,
|
| 84 |
+
"eval_loss": 0.7124308347702026,
|
| 85 |
+
"eval_runtime": 5.7619,
|
| 86 |
+
"eval_samples_per_second": 177.72,
|
| 87 |
+
"eval_steps_per_second": 22.215,
|
| 88 |
+
"step": 500
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 2.1484375,
|
| 92 |
+
"grad_norm": 12.242274284362793,
|
| 93 |
+
"learning_rate": 1.9783549783549785e-05,
|
| 94 |
+
"loss": 0.6108,
|
| 95 |
+
"step": 550
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.34375,
|
| 99 |
+
"grad_norm": 9.961933135986328,
|
| 100 |
+
"learning_rate": 1.9567099567099568e-05,
|
| 101 |
+
"loss": 0.5815,
|
| 102 |
+
"step": 600
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 2.5390625,
|
| 106 |
+
"grad_norm": 8.35198974609375,
|
| 107 |
+
"learning_rate": 1.9350649350649354e-05,
|
| 108 |
+
"loss": 0.5702,
|
| 109 |
+
"step": 650
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 2.734375,
|
| 113 |
+
"grad_norm": 4.211214542388916,
|
| 114 |
+
"learning_rate": 1.9134199134199138e-05,
|
| 115 |
+
"loss": 0.4907,
|
| 116 |
+
"step": 700
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 2.9296875,
|
| 120 |
+
"grad_norm": 5.052712917327881,
|
| 121 |
+
"learning_rate": 1.891774891774892e-05,
|
| 122 |
+
"loss": 0.4993,
|
| 123 |
+
"step": 750
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 3.125,
|
| 127 |
+
"grad_norm": 2.4637234210968018,
|
| 128 |
+
"learning_rate": 1.8701298701298704e-05,
|
| 129 |
+
"loss": 0.3788,
|
| 130 |
+
"step": 800
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 3.3203125,
|
| 134 |
+
"grad_norm": 5.4455790519714355,
|
| 135 |
+
"learning_rate": 1.8484848484848487e-05,
|
| 136 |
+
"loss": 0.358,
|
| 137 |
+
"step": 850
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.515625,
|
| 141 |
+
"grad_norm": 13.753146171569824,
|
| 142 |
+
"learning_rate": 1.826839826839827e-05,
|
| 143 |
+
"loss": 0.273,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 3.7109375,
|
| 148 |
+
"grad_norm": 7.315534591674805,
|
| 149 |
+
"learning_rate": 1.8051948051948053e-05,
|
| 150 |
+
"loss": 0.2588,
|
| 151 |
+
"step": 950
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 3.90625,
|
| 155 |
+
"grad_norm": 15.05672836303711,
|
| 156 |
+
"learning_rate": 1.7835497835497836e-05,
|
| 157 |
+
"loss": 0.3023,
|
| 158 |
+
"step": 1000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 3.90625,
|
| 162 |
+
"eval_accuracy": 0.8828125,
|
| 163 |
+
"eval_loss": 0.3776271939277649,
|
| 164 |
+
"eval_runtime": 6.9418,
|
| 165 |
+
"eval_samples_per_second": 147.512,
|
| 166 |
+
"eval_steps_per_second": 18.439,
|
| 167 |
+
"step": 1000
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 4.1015625,
|
| 171 |
+
"grad_norm": 0.9234067797660828,
|
| 172 |
+
"learning_rate": 1.761904761904762e-05,
|
| 173 |
+
"loss": 0.2126,
|
| 174 |
+
"step": 1050
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 4.296875,
|
| 178 |
+
"grad_norm": 4.910510540008545,
|
| 179 |
+
"learning_rate": 1.7402597402597403e-05,
|
| 180 |
+
"loss": 0.1539,
|
| 181 |
+
"step": 1100
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 4.4921875,
|
| 185 |
+
"grad_norm": 6.401501655578613,
|
| 186 |
+
"learning_rate": 1.718614718614719e-05,
|
| 187 |
+
"loss": 0.2101,
|
| 188 |
+
"step": 1150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.6875,
|
| 192 |
+
"grad_norm": 1.9736899137496948,
|
| 193 |
+
"learning_rate": 1.6969696969696972e-05,
|
| 194 |
+
"loss": 0.1984,
|
| 195 |
+
"step": 1200
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 4.8828125,
|
| 199 |
+
"grad_norm": 8.743949890136719,
|
| 200 |
+
"learning_rate": 1.6753246753246756e-05,
|
| 201 |
+
"loss": 0.1249,
|
| 202 |
+
"step": 1250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 5.078125,
|
| 206 |
+
"grad_norm": 14.517168998718262,
|
| 207 |
+
"learning_rate": 1.653679653679654e-05,
|
| 208 |
+
"loss": 0.1279,
|
| 209 |
+
"step": 1300
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 5.2734375,
|
| 213 |
+
"grad_norm": 0.22373303771018982,
|
| 214 |
+
"learning_rate": 1.6320346320346322e-05,
|
| 215 |
+
"loss": 0.0986,
|
| 216 |
+
"step": 1350
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.46875,
|
| 220 |
+
"grad_norm": 0.18486973643302917,
|
| 221 |
+
"learning_rate": 1.6103896103896105e-05,
|
| 222 |
+
"loss": 0.0922,
|
| 223 |
+
"step": 1400
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 5.6640625,
|
| 227 |
+
"grad_norm": 0.4893491268157959,
|
| 228 |
+
"learning_rate": 1.5887445887445888e-05,
|
| 229 |
+
"loss": 0.0666,
|
| 230 |
+
"step": 1450
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.859375,
|
| 234 |
+
"grad_norm": 17.192604064941406,
|
| 235 |
+
"learning_rate": 1.567099567099567e-05,
|
| 236 |
+
"loss": 0.0997,
|
| 237 |
+
"step": 1500
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 5.859375,
|
| 241 |
+
"eval_accuracy": 0.9130859375,
|
| 242 |
+
"eval_loss": 0.2808457016944885,
|
| 243 |
+
"eval_runtime": 5.8114,
|
| 244 |
+
"eval_samples_per_second": 176.206,
|
| 245 |
+
"eval_steps_per_second": 22.026,
|
| 246 |
+
"step": 1500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 6.0546875,
|
| 250 |
+
"grad_norm": 0.13603505492210388,
|
| 251 |
+
"learning_rate": 1.5454545454545454e-05,
|
| 252 |
+
"loss": 0.0481,
|
| 253 |
+
"step": 1550
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 6.25,
|
| 257 |
+
"grad_norm": 44.868900299072266,
|
| 258 |
+
"learning_rate": 1.523809523809524e-05,
|
| 259 |
+
"loss": 0.0465,
|
| 260 |
+
"step": 1600
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 6.4453125,
|
| 264 |
+
"grad_norm": 3.908478021621704,
|
| 265 |
+
"learning_rate": 1.5021645021645024e-05,
|
| 266 |
+
"loss": 0.0485,
|
| 267 |
+
"step": 1650
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.640625,
|
| 271 |
+
"grad_norm": 0.24283438920974731,
|
| 272 |
+
"learning_rate": 1.4805194805194807e-05,
|
| 273 |
+
"loss": 0.0387,
|
| 274 |
+
"step": 1700
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 6.8359375,
|
| 278 |
+
"grad_norm": 21.719118118286133,
|
| 279 |
+
"learning_rate": 1.458874458874459e-05,
|
| 280 |
+
"loss": 0.0803,
|
| 281 |
+
"step": 1750
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 7.03125,
|
| 285 |
+
"grad_norm": 0.08735662698745728,
|
| 286 |
+
"learning_rate": 1.4372294372294374e-05,
|
| 287 |
+
"loss": 0.0266,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 7.2265625,
|
| 292 |
+
"grad_norm": 0.16133151948451996,
|
| 293 |
+
"learning_rate": 1.4155844155844157e-05,
|
| 294 |
+
"loss": 0.0252,
|
| 295 |
+
"step": 1850
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 7.421875,
|
| 299 |
+
"grad_norm": 0.12098684161901474,
|
| 300 |
+
"learning_rate": 1.3939393939393942e-05,
|
| 301 |
+
"loss": 0.0174,
|
| 302 |
+
"step": 1900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 7.6171875,
|
| 306 |
+
"grad_norm": 0.14388597011566162,
|
| 307 |
+
"learning_rate": 1.3722943722943725e-05,
|
| 308 |
+
"loss": 0.016,
|
| 309 |
+
"step": 1950
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 7.8125,
|
| 313 |
+
"grad_norm": 0.07232346385717392,
|
| 314 |
+
"learning_rate": 1.3506493506493508e-05,
|
| 315 |
+
"loss": 0.0424,
|
| 316 |
+
"step": 2000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 7.8125,
|
| 320 |
+
"eval_accuracy": 0.95703125,
|
| 321 |
+
"eval_loss": 0.1914406716823578,
|
| 322 |
+
"eval_runtime": 5.2231,
|
| 323 |
+
"eval_samples_per_second": 196.054,
|
| 324 |
+
"eval_steps_per_second": 24.507,
|
| 325 |
+
"step": 2000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 8.0078125,
|
| 329 |
+
"grad_norm": 0.06997977197170258,
|
| 330 |
+
"learning_rate": 1.3294372294372295e-05,
|
| 331 |
+
"loss": 0.031,
|
| 332 |
+
"step": 2050
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 8.203125,
|
| 336 |
+
"grad_norm": 0.6767904162406921,
|
| 337 |
+
"learning_rate": 1.3077922077922078e-05,
|
| 338 |
+
"loss": 0.016,
|
| 339 |
+
"step": 2100
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 8.3984375,
|
| 343 |
+
"grad_norm": 0.055935341864824295,
|
| 344 |
+
"learning_rate": 1.2861471861471862e-05,
|
| 345 |
+
"loss": 0.0128,
|
| 346 |
+
"step": 2150
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 8.59375,
|
| 350 |
+
"grad_norm": 0.05341877415776253,
|
| 351 |
+
"learning_rate": 1.2645021645021646e-05,
|
| 352 |
+
"loss": 0.0123,
|
| 353 |
+
"step": 2200
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 8.7890625,
|
| 357 |
+
"grad_norm": 0.051278505474328995,
|
| 358 |
+
"learning_rate": 1.242857142857143e-05,
|
| 359 |
+
"loss": 0.0243,
|
| 360 |
+
"step": 2250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 8.984375,
|
| 364 |
+
"grad_norm": 0.04905688017606735,
|
| 365 |
+
"learning_rate": 1.2212121212121213e-05,
|
| 366 |
+
"loss": 0.0116,
|
| 367 |
+
"step": 2300
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 9.1796875,
|
| 371 |
+
"grad_norm": 0.05154794827103615,
|
| 372 |
+
"learning_rate": 1.1995670995670996e-05,
|
| 373 |
+
"loss": 0.0104,
|
| 374 |
+
"step": 2350
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 9.375,
|
| 378 |
+
"grad_norm": 0.04436648637056351,
|
| 379 |
+
"learning_rate": 1.1779220779220779e-05,
|
| 380 |
+
"loss": 0.0096,
|
| 381 |
+
"step": 2400
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 9.5703125,
|
| 385 |
+
"grad_norm": 0.04592837765812874,
|
| 386 |
+
"learning_rate": 1.1562770562770564e-05,
|
| 387 |
+
"loss": 0.0096,
|
| 388 |
+
"step": 2450
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 9.765625,
|
| 392 |
+
"grad_norm": 0.5957643985748291,
|
| 393 |
+
"learning_rate": 1.1346320346320347e-05,
|
| 394 |
+
"loss": 0.0108,
|
| 395 |
+
"step": 2500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 9.765625,
|
| 399 |
+
"eval_accuracy": 0.89453125,
|
| 400 |
+
"eval_loss": 0.4534318447113037,
|
| 401 |
+
"eval_runtime": 5.8562,
|
| 402 |
+
"eval_samples_per_second": 174.857,
|
| 403 |
+
"eval_steps_per_second": 21.857,
|
| 404 |
+
"step": 2500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 9.9609375,
|
| 408 |
+
"grad_norm": 0.3423336446285248,
|
| 409 |
+
"learning_rate": 1.112987012987013e-05,
|
| 410 |
+
"loss": 0.0183,
|
| 411 |
+
"step": 2550
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 10.15625,
|
| 415 |
+
"grad_norm": 0.04016263037919998,
|
| 416 |
+
"learning_rate": 1.0913419913419913e-05,
|
| 417 |
+
"loss": 0.016,
|
| 418 |
+
"step": 2600
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 10.3515625,
|
| 422 |
+
"grad_norm": 0.04166042432188988,
|
| 423 |
+
"learning_rate": 1.0696969696969696e-05,
|
| 424 |
+
"loss": 0.0077,
|
| 425 |
+
"step": 2650
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 10.546875,
|
| 429 |
+
"grad_norm": 0.036698199808597565,
|
| 430 |
+
"learning_rate": 1.0480519480519481e-05,
|
| 431 |
+
"loss": 0.0077,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 10.7421875,
|
| 436 |
+
"grad_norm": 0.03544982895255089,
|
| 437 |
+
"learning_rate": 1.0264069264069264e-05,
|
| 438 |
+
"loss": 0.0076,
|
| 439 |
+
"step": 2750
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 10.9375,
|
| 443 |
+
"grad_norm": 0.033364105969667435,
|
| 444 |
+
"learning_rate": 1.0047619047619048e-05,
|
| 445 |
+
"loss": 0.0072,
|
| 446 |
+
"step": 2800
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 11.1328125,
|
| 450 |
+
"grad_norm": 0.034085117280483246,
|
| 451 |
+
"learning_rate": 9.831168831168832e-06,
|
| 452 |
+
"loss": 0.007,
|
| 453 |
+
"step": 2850
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 11.328125,
|
| 457 |
+
"grad_norm": 0.03140813857316971,
|
| 458 |
+
"learning_rate": 9.614718614718616e-06,
|
| 459 |
+
"loss": 0.0065,
|
| 460 |
+
"step": 2900
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 11.5234375,
|
| 464 |
+
"grad_norm": 0.035150136798620224,
|
| 465 |
+
"learning_rate": 9.398268398268399e-06,
|
| 466 |
+
"loss": 0.014,
|
| 467 |
+
"step": 2950
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 11.71875,
|
| 471 |
+
"grad_norm": 0.029343638569116592,
|
| 472 |
+
"learning_rate": 9.181818181818184e-06,
|
| 473 |
+
"loss": 0.0088,
|
| 474 |
+
"step": 3000
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 11.71875,
|
| 478 |
+
"eval_accuracy": 0.9580078125,
|
| 479 |
+
"eval_loss": 0.1553695797920227,
|
| 480 |
+
"eval_runtime": 5.2716,
|
| 481 |
+
"eval_samples_per_second": 194.25,
|
| 482 |
+
"eval_steps_per_second": 24.281,
|
| 483 |
+
"step": 3000
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 11.9140625,
|
| 487 |
+
"grad_norm": 0.02912583015859127,
|
| 488 |
+
"learning_rate": 8.965367965367967e-06,
|
| 489 |
+
"loss": 0.0063,
|
| 490 |
+
"step": 3050
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 12.109375,
|
| 494 |
+
"grad_norm": 0.02827904000878334,
|
| 495 |
+
"learning_rate": 8.74891774891775e-06,
|
| 496 |
+
"loss": 0.0067,
|
| 497 |
+
"step": 3100
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 12.3046875,
|
| 501 |
+
"grad_norm": 0.02774711512029171,
|
| 502 |
+
"learning_rate": 8.536796536796537e-06,
|
| 503 |
+
"loss": 0.0088,
|
| 504 |
+
"step": 3150
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 12.5,
|
| 508 |
+
"grad_norm": 0.027258064597845078,
|
| 509 |
+
"learning_rate": 8.32034632034632e-06,
|
| 510 |
+
"loss": 0.0055,
|
| 511 |
+
"step": 3200
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 12.6953125,
|
| 515 |
+
"grad_norm": 0.02587069943547249,
|
| 516 |
+
"learning_rate": 8.103896103896105e-06,
|
| 517 |
+
"loss": 0.0053,
|
| 518 |
+
"step": 3250
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 12.890625,
|
| 522 |
+
"grad_norm": 0.02641950733959675,
|
| 523 |
+
"learning_rate": 7.887445887445888e-06,
|
| 524 |
+
"loss": 0.0054,
|
| 525 |
+
"step": 3300
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 13.0859375,
|
| 529 |
+
"grad_norm": 0.028370605781674385,
|
| 530 |
+
"learning_rate": 7.670995670995672e-06,
|
| 531 |
+
"loss": 0.0052,
|
| 532 |
+
"step": 3350
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 13.28125,
|
| 536 |
+
"grad_norm": 0.02395276166498661,
|
| 537 |
+
"learning_rate": 7.454545454545456e-06,
|
| 538 |
+
"loss": 0.0049,
|
| 539 |
+
"step": 3400
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 13.4765625,
|
| 543 |
+
"grad_norm": 0.022906098514795303,
|
| 544 |
+
"learning_rate": 7.238095238095239e-06,
|
| 545 |
+
"loss": 0.0049,
|
| 546 |
+
"step": 3450
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 13.671875,
|
| 550 |
+
"grad_norm": 0.023122085258364677,
|
| 551 |
+
"learning_rate": 7.021645021645023e-06,
|
| 552 |
+
"loss": 0.0051,
|
| 553 |
+
"step": 3500
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 13.671875,
|
| 557 |
+
"eval_accuracy": 0.958984375,
|
| 558 |
+
"eval_loss": 0.16660287976264954,
|
| 559 |
+
"eval_runtime": 5.1728,
|
| 560 |
+
"eval_samples_per_second": 197.959,
|
| 561 |
+
"eval_steps_per_second": 24.745,
|
| 562 |
+
"step": 3500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 13.8671875,
|
| 566 |
+
"grad_norm": 0.022511709481477737,
|
| 567 |
+
"learning_rate": 6.805194805194806e-06,
|
| 568 |
+
"loss": 0.0047,
|
| 569 |
+
"step": 3550
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 14.0625,
|
| 573 |
+
"grad_norm": 0.06705322116613388,
|
| 574 |
+
"learning_rate": 6.588744588744589e-06,
|
| 575 |
+
"loss": 0.0061,
|
| 576 |
+
"step": 3600
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 14.2578125,
|
| 580 |
+
"grad_norm": 0.032343845814466476,
|
| 581 |
+
"learning_rate": 6.372294372294373e-06,
|
| 582 |
+
"loss": 0.0167,
|
| 583 |
+
"step": 3650
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 14.453125,
|
| 587 |
+
"grad_norm": 0.040268197655677795,
|
| 588 |
+
"learning_rate": 6.155844155844156e-06,
|
| 589 |
+
"loss": 0.0059,
|
| 590 |
+
"step": 3700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 14.6484375,
|
| 594 |
+
"grad_norm": 0.020453229546546936,
|
| 595 |
+
"learning_rate": 5.93939393939394e-06,
|
| 596 |
+
"loss": 0.0044,
|
| 597 |
+
"step": 3750
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 14.84375,
|
| 601 |
+
"grad_norm": 0.019762134179472923,
|
| 602 |
+
"learning_rate": 5.722943722943723e-06,
|
| 603 |
+
"loss": 0.0044,
|
| 604 |
+
"step": 3800
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 15.0390625,
|
| 608 |
+
"grad_norm": 0.020851831883192062,
|
| 609 |
+
"learning_rate": 5.5064935064935065e-06,
|
| 610 |
+
"loss": 0.0041,
|
| 611 |
+
"step": 3850
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 15.234375,
|
| 615 |
+
"grad_norm": 0.01954171247780323,
|
| 616 |
+
"learning_rate": 5.2900432900432905e-06,
|
| 617 |
+
"loss": 0.0041,
|
| 618 |
+
"step": 3900
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 15.4296875,
|
| 622 |
+
"grad_norm": 0.020014379173517227,
|
| 623 |
+
"learning_rate": 5.073593073593074e-06,
|
| 624 |
+
"loss": 0.0042,
|
| 625 |
+
"step": 3950
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 15.625,
|
| 629 |
+
"grad_norm": 0.01890181563794613,
|
| 630 |
+
"learning_rate": 4.857142857142858e-06,
|
| 631 |
+
"loss": 0.0039,
|
| 632 |
+
"step": 4000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 15.625,
|
| 636 |
+
"eval_accuracy": 0.96484375,
|
| 637 |
+
"eval_loss": 0.15439653396606445,
|
| 638 |
+
"eval_runtime": 5.142,
|
| 639 |
+
"eval_samples_per_second": 199.146,
|
| 640 |
+
"eval_steps_per_second": 24.893,
|
| 641 |
+
"step": 4000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 15.8203125,
|
| 645 |
+
"grad_norm": 0.017986547201871872,
|
| 646 |
+
"learning_rate": 4.640692640692641e-06,
|
| 647 |
+
"loss": 0.0038,
|
| 648 |
+
"step": 4050
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 16.015625,
|
| 652 |
+
"grad_norm": 0.02049080841243267,
|
| 653 |
+
"learning_rate": 4.424242424242425e-06,
|
| 654 |
+
"loss": 0.0037,
|
| 655 |
+
"step": 4100
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 16.2109375,
|
| 659 |
+
"grad_norm": 0.018345637246966362,
|
| 660 |
+
"learning_rate": 4.207792207792208e-06,
|
| 661 |
+
"loss": 0.0038,
|
| 662 |
+
"step": 4150
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 16.40625,
|
| 666 |
+
"grad_norm": 0.018377136439085007,
|
| 667 |
+
"learning_rate": 3.991341991341992e-06,
|
| 668 |
+
"loss": 0.0037,
|
| 669 |
+
"step": 4200
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 16.6015625,
|
| 673 |
+
"grad_norm": 0.01728520356118679,
|
| 674 |
+
"learning_rate": 3.774891774891775e-06,
|
| 675 |
+
"loss": 0.0037,
|
| 676 |
+
"step": 4250
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 16.796875,
|
| 680 |
+
"grad_norm": 0.017723916098475456,
|
| 681 |
+
"learning_rate": 3.558441558441559e-06,
|
| 682 |
+
"loss": 0.0035,
|
| 683 |
+
"step": 4300
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 16.9921875,
|
| 687 |
+
"grad_norm": 0.017535051330924034,
|
| 688 |
+
"learning_rate": 3.3419913419913426e-06,
|
| 689 |
+
"loss": 0.0036,
|
| 690 |
+
"step": 4350
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 17.1875,
|
| 694 |
+
"grad_norm": 0.01705988124012947,
|
| 695 |
+
"learning_rate": 3.1255411255411257e-06,
|
| 696 |
+
"loss": 0.0036,
|
| 697 |
+
"step": 4400
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 17.3828125,
|
| 701 |
+
"grad_norm": 0.01828162744641304,
|
| 702 |
+
"learning_rate": 2.9090909090909093e-06,
|
| 703 |
+
"loss": 0.0035,
|
| 704 |
+
"step": 4450
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 17.578125,
|
| 708 |
+
"grad_norm": 0.026976697146892548,
|
| 709 |
+
"learning_rate": 2.692640692640693e-06,
|
| 710 |
+
"loss": 0.0034,
|
| 711 |
+
"step": 4500
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 17.578125,
|
| 715 |
+
"eval_accuracy": 0.96484375,
|
| 716 |
+
"eval_loss": 0.15748074650764465,
|
| 717 |
+
"eval_runtime": 5.486,
|
| 718 |
+
"eval_samples_per_second": 186.655,
|
| 719 |
+
"eval_steps_per_second": 23.332,
|
| 720 |
+
"step": 4500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 17.7734375,
|
| 724 |
+
"grad_norm": 0.01693248189985752,
|
| 725 |
+
"learning_rate": 2.4761904761904764e-06,
|
| 726 |
+
"loss": 0.0033,
|
| 727 |
+
"step": 4550
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 17.96875,
|
| 731 |
+
"grad_norm": 0.016402654349803925,
|
| 732 |
+
"learning_rate": 2.25974025974026e-06,
|
| 733 |
+
"loss": 0.0034,
|
| 734 |
+
"step": 4600
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 18.1640625,
|
| 738 |
+
"grad_norm": 0.01663408987224102,
|
| 739 |
+
"learning_rate": 2.0432900432900435e-06,
|
| 740 |
+
"loss": 0.0033,
|
| 741 |
+
"step": 4650
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 18.359375,
|
| 745 |
+
"grad_norm": 0.02396741323173046,
|
| 746 |
+
"learning_rate": 1.8268398268398269e-06,
|
| 747 |
+
"loss": 0.0034,
|
| 748 |
+
"step": 4700
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 18.5546875,
|
| 752 |
+
"grad_norm": 0.016884848475456238,
|
| 753 |
+
"learning_rate": 1.6103896103896105e-06,
|
| 754 |
+
"loss": 0.0032,
|
| 755 |
+
"step": 4750
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 18.75,
|
| 759 |
+
"grad_norm": 0.016911398619413376,
|
| 760 |
+
"learning_rate": 1.3939393939393942e-06,
|
| 761 |
+
"loss": 0.0033,
|
| 762 |
+
"step": 4800
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 18.9453125,
|
| 766 |
+
"grad_norm": 0.01589149609208107,
|
| 767 |
+
"learning_rate": 1.1774891774891776e-06,
|
| 768 |
+
"loss": 0.0033,
|
| 769 |
+
"step": 4850
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 19.140625,
|
| 773 |
+
"grad_norm": 0.017035914584994316,
|
| 774 |
+
"learning_rate": 9.610389610389612e-07,
|
| 775 |
+
"loss": 0.0032,
|
| 776 |
+
"step": 4900
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 19.3359375,
|
| 780 |
+
"grad_norm": 0.01621602661907673,
|
| 781 |
+
"learning_rate": 7.445887445887446e-07,
|
| 782 |
+
"loss": 0.0033,
|
| 783 |
+
"step": 4950
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 19.53125,
|
| 787 |
+
"grad_norm": 0.015999143943190575,
|
| 788 |
+
"learning_rate": 5.281385281385282e-07,
|
| 789 |
+
"loss": 0.003,
|
| 790 |
+
"step": 5000
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 19.53125,
|
| 794 |
+
"eval_accuracy": 0.9658203125,
|
| 795 |
+
"eval_loss": 0.15924891829490662,
|
| 796 |
+
"eval_runtime": 5.5508,
|
| 797 |
+
"eval_samples_per_second": 184.477,
|
| 798 |
+
"eval_steps_per_second": 23.06,
|
| 799 |
+
"step": 5000
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 19.7265625,
|
| 803 |
+
"grad_norm": 0.014971093274652958,
|
| 804 |
+
"learning_rate": 3.116883116883117e-07,
|
| 805 |
+
"loss": 0.0033,
|
| 806 |
+
"step": 5050
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 19.921875,
|
| 810 |
+
"grad_norm": 0.015599260106682777,
|
| 811 |
+
"learning_rate": 9.523809523809525e-08,
|
| 812 |
+
"loss": 0.0031,
|
| 813 |
+
"step": 5100
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 20.0,
|
| 817 |
+
"step": 5120,
|
| 818 |
+
"total_flos": 6.348257582037074e+18,
|
| 819 |
+
"train_loss": 0.15641047188546509,
|
| 820 |
+
"train_runtime": 1165.0601,
|
| 821 |
+
"train_samples_per_second": 70.314,
|
| 822 |
+
"train_steps_per_second": 4.395
|
| 823 |
+
}
|
| 824 |
+
],
|
| 825 |
+
"logging_steps": 50,
|
| 826 |
+
"max_steps": 5120,
|
| 827 |
+
"num_input_tokens_seen": 0,
|
| 828 |
+
"num_train_epochs": 20,
|
| 829 |
+
"save_steps": 500,
|
| 830 |
+
"stateful_callbacks": {
|
| 831 |
+
"EarlyStoppingCallback": {
|
| 832 |
+
"args": {
|
| 833 |
+
"early_stopping_patience": 3,
|
| 834 |
+
"early_stopping_threshold": 0.0
|
| 835 |
+
},
|
| 836 |
+
"attributes": {
|
| 837 |
+
"early_stopping_patience_counter": 0
|
| 838 |
+
}
|
| 839 |
+
},
|
| 840 |
+
"TrainerControl": {
|
| 841 |
+
"args": {
|
| 842 |
+
"should_epoch_stop": false,
|
| 843 |
+
"should_evaluate": false,
|
| 844 |
+
"should_log": false,
|
| 845 |
+
"should_save": true,
|
| 846 |
+
"should_training_stop": true
|
| 847 |
+
},
|
| 848 |
+
"attributes": {}
|
| 849 |
+
}
|
| 850 |
+
},
|
| 851 |
+
"total_flos": 6.348257582037074e+18,
|
| 852 |
+
"train_batch_size": 16,
|
| 853 |
+
"trial_name": null,
|
| 854 |
+
"trial_params": null
|
| 855 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1fb96db67b3e0239848c4244752a8669a49b74668d944915c66370573f9236c
|
| 3 |
+
size 5112
|