Commit
•
4cbc492
1
Parent(s):
f2b71b2
Yayyy Lets Share hateful Memes!!!
Browse files- README.md +70 -0
- all_results.json +13 -0
- config.json +32 -0
- eval_results.json +8 -0
- model.safetensors +3 -0
- preprocessor_config.json +36 -0
- runs/Apr10_11-12-24_decc3680cd07/events.out.tfevents.1712747583.decc3680cd07.703.0 +3 -0
- runs/Apr10_11-12-24_decc3680cd07/events.out.tfevents.1712748254.decc3680cd07.703.1 +3 -0
- runs/Apr10_11-28-52_decc3680cd07/events.out.tfevents.1712748567.decc3680cd07.703.2 +3 -0
- runs/Apr10_11-28-52_decc3680cd07/events.out.tfevents.1712749961.decc3680cd07.703.3 +3 -0
- train_results.json +8 -0
- trainer_state.json +3077 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: google/vit-base-patch16-224
|
| 4 |
+
tags:
|
| 5 |
+
- image-classification
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
metrics:
|
| 8 |
+
- accuracy
|
| 9 |
+
model-index:
|
| 10 |
+
- name: vit-base-hate-meme
|
| 11 |
+
results: []
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# vit-base-hate-meme
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the emily49/hateful_memes_train_dev dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.7461
|
| 22 |
+
- Accuracy: 0.498
|
| 23 |
+
|
| 24 |
+
## Model description
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Intended uses & limitations
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training and evaluation data
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training procedure
|
| 37 |
+
|
| 38 |
+
### Training hyperparameters
|
| 39 |
+
|
| 40 |
+
The following hyperparameters were used during training:
|
| 41 |
+
- learning_rate: 0.0002
|
| 42 |
+
- train_batch_size: 16
|
| 43 |
+
- eval_batch_size: 32
|
| 44 |
+
- seed: 42
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 500
|
| 48 |
+
- num_epochs: 8
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 54 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 55 |
+
| 0.6483 | 1.0 | 532 | 0.7933 | 0.5 |
|
| 56 |
+
| 0.6203 | 2.0 | 1064 | 0.7461 | 0.498 |
|
| 57 |
+
| 0.5955 | 3.0 | 1596 | 0.7539 | 0.518 |
|
| 58 |
+
| 0.4796 | 4.0 | 2128 | 0.9387 | 0.536 |
|
| 59 |
+
| 0.425 | 5.0 | 2660 | 0.9824 | 0.538 |
|
| 60 |
+
| 0.2869 | 6.0 | 3192 | 1.6186 | 0.542 |
|
| 61 |
+
| 0.0337 | 7.0 | 3724 | 2.5670 | 0.538 |
|
| 62 |
+
| 0.0166 | 8.0 | 4256 | 2.8061 | 0.538 |
|
| 63 |
+
|
| 64 |
+
|
| 65 |
+
### Framework versions
|
| 66 |
+
|
| 67 |
+
- Transformers 4.39.3
|
| 68 |
+
- Pytorch 2.2.1+cu121
|
| 69 |
+
- Datasets 2.18.0
|
| 70 |
+
- Tokenizers 0.15.2
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 8.0,
|
| 3 |
+
"eval_accuracy": 0.498,
|
| 4 |
+
"eval_loss": 0.746105968952179,
|
| 5 |
+
"eval_runtime": 5.5949,
|
| 6 |
+
"eval_samples_per_second": 89.367,
|
| 7 |
+
"eval_steps_per_second": 2.86,
|
| 8 |
+
"total_flos": 5.269455293792256e+18,
|
| 9 |
+
"train_loss": 0.40350076897398923,
|
| 10 |
+
"train_runtime": 1386.2467,
|
| 11 |
+
"train_samples_per_second": 49.053,
|
| 12 |
+
"train_steps_per_second": 3.07
|
| 13 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-base-patch16-224",
|
| 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": 0,
|
| 13 |
+
"1": 1
|
| 14 |
+
},
|
| 15 |
+
"image_size": 224,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 3072,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"0": "0",
|
| 20 |
+
"1": "1"
|
| 21 |
+
},
|
| 22 |
+
"layer_norm_eps": 1e-12,
|
| 23 |
+
"model_type": "vit",
|
| 24 |
+
"num_attention_heads": 12,
|
| 25 |
+
"num_channels": 3,
|
| 26 |
+
"num_hidden_layers": 12,
|
| 27 |
+
"patch_size": 16,
|
| 28 |
+
"problem_type": "single_label_classification",
|
| 29 |
+
"qkv_bias": true,
|
| 30 |
+
"torch_dtype": "float32",
|
| 31 |
+
"transformers_version": "4.39.3"
|
| 32 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 8.0,
|
| 3 |
+
"eval_accuracy": 0.498,
|
| 4 |
+
"eval_loss": 0.746105968952179,
|
| 5 |
+
"eval_runtime": 5.5949,
|
| 6 |
+
"eval_samples_per_second": 89.367,
|
| 7 |
+
"eval_steps_per_second": 2.86
|
| 8 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8710ba864bae83388e22f3b5e1f0c884776b615f25c94f15cc3b917142ee7cd
|
| 3 |
+
size 343223968
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_valid_processor_keys": [
|
| 3 |
+
"images",
|
| 4 |
+
"do_resize",
|
| 5 |
+
"size",
|
| 6 |
+
"resample",
|
| 7 |
+
"do_rescale",
|
| 8 |
+
"rescale_factor",
|
| 9 |
+
"do_normalize",
|
| 10 |
+
"image_mean",
|
| 11 |
+
"image_std",
|
| 12 |
+
"return_tensors",
|
| 13 |
+
"data_format",
|
| 14 |
+
"input_data_format"
|
| 15 |
+
],
|
| 16 |
+
"do_normalize": true,
|
| 17 |
+
"do_rescale": true,
|
| 18 |
+
"do_resize": true,
|
| 19 |
+
"image_mean": [
|
| 20 |
+
0.5,
|
| 21 |
+
0.5,
|
| 22 |
+
0.5
|
| 23 |
+
],
|
| 24 |
+
"image_processor_type": "ViTImageProcessor",
|
| 25 |
+
"image_std": [
|
| 26 |
+
0.5,
|
| 27 |
+
0.5,
|
| 28 |
+
0.5
|
| 29 |
+
],
|
| 30 |
+
"resample": 2,
|
| 31 |
+
"rescale_factor": 0.00392156862745098,
|
| 32 |
+
"size": {
|
| 33 |
+
"height": 224,
|
| 34 |
+
"width": 224
|
| 35 |
+
}
|
| 36 |
+
}
|
runs/Apr10_11-12-24_decc3680cd07/events.out.tfevents.1712747583.decc3680cd07.703.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a011c4bbc10c6dbcabde2d519d8dd8e46e8fed5e7f235be6a03a4f1f9167c09
|
| 3 |
+
size 17683
|
runs/Apr10_11-12-24_decc3680cd07/events.out.tfevents.1712748254.decc3680cd07.703.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c5a5acc663006d4f4fb7a1e1628f3b8737def1c2dd1e8ef679c731785b7a7c2
|
| 3 |
+
size 411
|
runs/Apr10_11-28-52_decc3680cd07/events.out.tfevents.1712748567.decc3680cd07.703.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e4a2a9209bdac152cd3b814c4135eeb3a4a1ce69d566121293d43e83d339d0f
|
| 3 |
+
size 97231
|
runs/Apr10_11-28-52_decc3680cd07/events.out.tfevents.1712749961.decc3680cd07.703.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48fa56668bf4f481f9001ec1c3d4156d5ee4238f2e1bb620f9c97dc4a195f027
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 8.0,
|
| 3 |
+
"total_flos": 5.269455293792256e+18,
|
| 4 |
+
"train_loss": 0.40350076897398923,
|
| 5 |
+
"train_runtime": 1386.2467,
|
| 6 |
+
"train_samples_per_second": 49.053,
|
| 7 |
+
"train_steps_per_second": 3.07
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3077 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.746105968952179,
|
| 3 |
+
"best_model_checkpoint": "./vit-base-hate-meme/checkpoint-1064",
|
| 4 |
+
"epoch": 8.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4256,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 1.893226146697998,
|
| 14 |
+
"learning_rate": 3.6e-06,
|
| 15 |
+
"loss": 0.6079,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04,
|
| 20 |
+
"grad_norm": 1.8366069793701172,
|
| 21 |
+
"learning_rate": 7.2e-06,
|
| 22 |
+
"loss": 0.6672,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.06,
|
| 27 |
+
"grad_norm": 3.005919933319092,
|
| 28 |
+
"learning_rate": 1.1200000000000001e-05,
|
| 29 |
+
"loss": 0.6246,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08,
|
| 34 |
+
"grad_norm": 0.47935283184051514,
|
| 35 |
+
"learning_rate": 1.52e-05,
|
| 36 |
+
"loss": 0.6929,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09,
|
| 41 |
+
"grad_norm": 3.6089649200439453,
|
| 42 |
+
"learning_rate": 1.9200000000000003e-05,
|
| 43 |
+
"loss": 0.6635,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11,
|
| 48 |
+
"grad_norm": 1.0585061311721802,
|
| 49 |
+
"learning_rate": 2.32e-05,
|
| 50 |
+
"loss": 0.626,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.13,
|
| 55 |
+
"grad_norm": 0.16031716763973236,
|
| 56 |
+
"learning_rate": 2.7200000000000004e-05,
|
| 57 |
+
"loss": 0.6391,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.15,
|
| 62 |
+
"grad_norm": 0.742310106754303,
|
| 63 |
+
"learning_rate": 3.12e-05,
|
| 64 |
+
"loss": 0.6518,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.17,
|
| 69 |
+
"grad_norm": 1.0254610776901245,
|
| 70 |
+
"learning_rate": 3.52e-05,
|
| 71 |
+
"loss": 0.6408,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.19,
|
| 76 |
+
"grad_norm": 1.074569821357727,
|
| 77 |
+
"learning_rate": 3.9200000000000004e-05,
|
| 78 |
+
"loss": 0.6446,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.21,
|
| 83 |
+
"grad_norm": 0.3864551782608032,
|
| 84 |
+
"learning_rate": 4.32e-05,
|
| 85 |
+
"loss": 0.6591,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.23,
|
| 90 |
+
"grad_norm": 1.1516258716583252,
|
| 91 |
+
"learning_rate": 4.72e-05,
|
| 92 |
+
"loss": 0.6288,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.24,
|
| 97 |
+
"grad_norm": 1.2458890676498413,
|
| 98 |
+
"learning_rate": 5.1200000000000004e-05,
|
| 99 |
+
"loss": 0.6325,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.26,
|
| 104 |
+
"grad_norm": 0.7807619571685791,
|
| 105 |
+
"learning_rate": 5.520000000000001e-05,
|
| 106 |
+
"loss": 0.662,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.28,
|
| 111 |
+
"grad_norm": 4.871969699859619,
|
| 112 |
+
"learning_rate": 5.88e-05,
|
| 113 |
+
"loss": 0.6877,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3,
|
| 118 |
+
"grad_norm": 2.10189151763916,
|
| 119 |
+
"learning_rate": 6.280000000000001e-05,
|
| 120 |
+
"loss": 0.6344,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.32,
|
| 125 |
+
"grad_norm": 5.75629186630249,
|
| 126 |
+
"learning_rate": 6.680000000000001e-05,
|
| 127 |
+
"loss": 0.6487,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.34,
|
| 132 |
+
"grad_norm": 0.9313517212867737,
|
| 133 |
+
"learning_rate": 7.08e-05,
|
| 134 |
+
"loss": 0.5723,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.36,
|
| 139 |
+
"grad_norm": 3.9287912845611572,
|
| 140 |
+
"learning_rate": 7.48e-05,
|
| 141 |
+
"loss": 0.6724,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.38,
|
| 146 |
+
"grad_norm": 4.517734527587891,
|
| 147 |
+
"learning_rate": 7.88e-05,
|
| 148 |
+
"loss": 0.647,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.39,
|
| 153 |
+
"grad_norm": 1.1182656288146973,
|
| 154 |
+
"learning_rate": 8.28e-05,
|
| 155 |
+
"loss": 0.6423,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.41,
|
| 160 |
+
"grad_norm": 3.048739194869995,
|
| 161 |
+
"learning_rate": 8.680000000000001e-05,
|
| 162 |
+
"loss": 0.5778,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.43,
|
| 167 |
+
"grad_norm": 1.1923900842666626,
|
| 168 |
+
"learning_rate": 9.080000000000001e-05,
|
| 169 |
+
"loss": 0.6203,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45,
|
| 174 |
+
"grad_norm": 2.969618082046509,
|
| 175 |
+
"learning_rate": 9.48e-05,
|
| 176 |
+
"loss": 0.6514,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.47,
|
| 181 |
+
"grad_norm": 0.6900493502616882,
|
| 182 |
+
"learning_rate": 9.88e-05,
|
| 183 |
+
"loss": 0.6581,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.49,
|
| 188 |
+
"grad_norm": 2.1176040172576904,
|
| 189 |
+
"learning_rate": 0.0001028,
|
| 190 |
+
"loss": 0.6104,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.51,
|
| 195 |
+
"grad_norm": 1.029479742050171,
|
| 196 |
+
"learning_rate": 0.00010680000000000001,
|
| 197 |
+
"loss": 0.5896,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.53,
|
| 202 |
+
"grad_norm": 0.8468737602233887,
|
| 203 |
+
"learning_rate": 0.00011080000000000001,
|
| 204 |
+
"loss": 0.6067,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.55,
|
| 209 |
+
"grad_norm": 1.5066865682601929,
|
| 210 |
+
"learning_rate": 0.0001148,
|
| 211 |
+
"loss": 0.6,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.56,
|
| 216 |
+
"grad_norm": 0.6331378817558289,
|
| 217 |
+
"learning_rate": 0.0001188,
|
| 218 |
+
"loss": 0.6437,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.58,
|
| 223 |
+
"grad_norm": 2.0694966316223145,
|
| 224 |
+
"learning_rate": 0.0001228,
|
| 225 |
+
"loss": 0.6304,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6,
|
| 230 |
+
"grad_norm": 0.4095279574394226,
|
| 231 |
+
"learning_rate": 0.00012680000000000002,
|
| 232 |
+
"loss": 0.5867,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.62,
|
| 237 |
+
"grad_norm": 3.9158384799957275,
|
| 238 |
+
"learning_rate": 0.0001308,
|
| 239 |
+
"loss": 0.6397,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.64,
|
| 244 |
+
"grad_norm": 1.578360915184021,
|
| 245 |
+
"learning_rate": 0.00013480000000000002,
|
| 246 |
+
"loss": 0.6616,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.66,
|
| 251 |
+
"grad_norm": 0.3877038359642029,
|
| 252 |
+
"learning_rate": 0.00013879999999999999,
|
| 253 |
+
"loss": 0.6161,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.68,
|
| 258 |
+
"grad_norm": 1.0857969522476196,
|
| 259 |
+
"learning_rate": 0.0001428,
|
| 260 |
+
"loss": 0.6281,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7,
|
| 265 |
+
"grad_norm": 4.571174621582031,
|
| 266 |
+
"learning_rate": 0.00014680000000000002,
|
| 267 |
+
"loss": 0.6338,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.71,
|
| 272 |
+
"grad_norm": 1.5768461227416992,
|
| 273 |
+
"learning_rate": 0.0001508,
|
| 274 |
+
"loss": 0.6602,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.73,
|
| 279 |
+
"grad_norm": 1.675762414932251,
|
| 280 |
+
"learning_rate": 0.00015480000000000002,
|
| 281 |
+
"loss": 0.6649,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.75,
|
| 286 |
+
"grad_norm": 0.6675221920013428,
|
| 287 |
+
"learning_rate": 0.0001588,
|
| 288 |
+
"loss": 0.6342,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.77,
|
| 293 |
+
"grad_norm": 3.4393932819366455,
|
| 294 |
+
"learning_rate": 0.0001628,
|
| 295 |
+
"loss": 0.6678,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.79,
|
| 300 |
+
"grad_norm": 4.380407810211182,
|
| 301 |
+
"learning_rate": 0.0001668,
|
| 302 |
+
"loss": 0.6496,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.81,
|
| 307 |
+
"grad_norm": 1.5338869094848633,
|
| 308 |
+
"learning_rate": 0.0001708,
|
| 309 |
+
"loss": 0.6416,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.83,
|
| 314 |
+
"grad_norm": 4.133732318878174,
|
| 315 |
+
"learning_rate": 0.00017480000000000002,
|
| 316 |
+
"loss": 0.6142,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.85,
|
| 321 |
+
"grad_norm": 1.7674825191497803,
|
| 322 |
+
"learning_rate": 0.0001788,
|
| 323 |
+
"loss": 0.5776,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.86,
|
| 328 |
+
"grad_norm": 0.7282730937004089,
|
| 329 |
+
"learning_rate": 0.00018280000000000003,
|
| 330 |
+
"loss": 0.6721,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.88,
|
| 335 |
+
"grad_norm": 3.9304282665252686,
|
| 336 |
+
"learning_rate": 0.00018680000000000001,
|
| 337 |
+
"loss": 0.6354,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9,
|
| 342 |
+
"grad_norm": 2.655442476272583,
|
| 343 |
+
"learning_rate": 0.0001908,
|
| 344 |
+
"loss": 0.6835,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.92,
|
| 349 |
+
"grad_norm": 3.454685926437378,
|
| 350 |
+
"learning_rate": 0.0001948,
|
| 351 |
+
"loss": 0.6148,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.94,
|
| 356 |
+
"grad_norm": 2.250034809112549,
|
| 357 |
+
"learning_rate": 0.0001988,
|
| 358 |
+
"loss": 0.6026,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.96,
|
| 363 |
+
"grad_norm": 0.33308330178260803,
|
| 364 |
+
"learning_rate": 0.0001996272630457934,
|
| 365 |
+
"loss": 0.6638,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.98,
|
| 370 |
+
"grad_norm": 2.148402214050293,
|
| 371 |
+
"learning_rate": 0.00019909478168264112,
|
| 372 |
+
"loss": 0.6184,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0,
|
| 377 |
+
"grad_norm": 4.5404582023620605,
|
| 378 |
+
"learning_rate": 0.00019856230031948884,
|
| 379 |
+
"loss": 0.6483,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.0,
|
| 384 |
+
"eval_accuracy": 0.5,
|
| 385 |
+
"eval_loss": 0.7933283448219299,
|
| 386 |
+
"eval_runtime": 5.2926,
|
| 387 |
+
"eval_samples_per_second": 94.472,
|
| 388 |
+
"eval_steps_per_second": 3.023,
|
| 389 |
+
"step": 532
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.02,
|
| 393 |
+
"grad_norm": 0.8757955431938171,
|
| 394 |
+
"learning_rate": 0.00019802981895633654,
|
| 395 |
+
"loss": 0.56,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.03,
|
| 400 |
+
"grad_norm": 2.5476624965667725,
|
| 401 |
+
"learning_rate": 0.00019749733759318424,
|
| 402 |
+
"loss": 0.7584,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.05,
|
| 407 |
+
"grad_norm": 4.550638675689697,
|
| 408 |
+
"learning_rate": 0.00019696485623003196,
|
| 409 |
+
"loss": 0.6235,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.07,
|
| 414 |
+
"grad_norm": 5.46172571182251,
|
| 415 |
+
"learning_rate": 0.00019643237486687966,
|
| 416 |
+
"loss": 0.6473,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.09,
|
| 421 |
+
"grad_norm": 2.1486427783966064,
|
| 422 |
+
"learning_rate": 0.00019589989350372736,
|
| 423 |
+
"loss": 0.6259,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.11,
|
| 428 |
+
"grad_norm": 4.4022955894470215,
|
| 429 |
+
"learning_rate": 0.0001953674121405751,
|
| 430 |
+
"loss": 0.6031,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.13,
|
| 435 |
+
"grad_norm": 1.3872004747390747,
|
| 436 |
+
"learning_rate": 0.0001948349307774228,
|
| 437 |
+
"loss": 0.5789,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.15,
|
| 442 |
+
"grad_norm": 1.193093180656433,
|
| 443 |
+
"learning_rate": 0.0001943024494142705,
|
| 444 |
+
"loss": 0.6409,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.17,
|
| 449 |
+
"grad_norm": 4.4846415519714355,
|
| 450 |
+
"learning_rate": 0.0001937699680511182,
|
| 451 |
+
"loss": 0.6433,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.18,
|
| 456 |
+
"grad_norm": 1.3275984525680542,
|
| 457 |
+
"learning_rate": 0.00019323748668796594,
|
| 458 |
+
"loss": 0.6853,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.2,
|
| 463 |
+
"grad_norm": 0.5516121983528137,
|
| 464 |
+
"learning_rate": 0.00019270500532481363,
|
| 465 |
+
"loss": 0.6577,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.22,
|
| 470 |
+
"grad_norm": 0.6763603091239929,
|
| 471 |
+
"learning_rate": 0.00019217252396166133,
|
| 472 |
+
"loss": 0.6181,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.24,
|
| 477 |
+
"grad_norm": 0.7651398181915283,
|
| 478 |
+
"learning_rate": 0.00019164004259850906,
|
| 479 |
+
"loss": 0.6698,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.26,
|
| 484 |
+
"grad_norm": 6.038695335388184,
|
| 485 |
+
"learning_rate": 0.00019110756123535678,
|
| 486 |
+
"loss": 0.6425,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.28,
|
| 491 |
+
"grad_norm": 1.9138271808624268,
|
| 492 |
+
"learning_rate": 0.00019057507987220448,
|
| 493 |
+
"loss": 0.6428,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.3,
|
| 498 |
+
"grad_norm": 2.1620535850524902,
|
| 499 |
+
"learning_rate": 0.00019004259850905218,
|
| 500 |
+
"loss": 0.6643,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.32,
|
| 505 |
+
"grad_norm": 0.9087329506874084,
|
| 506 |
+
"learning_rate": 0.0001895101171458999,
|
| 507 |
+
"loss": 0.6042,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.33,
|
| 512 |
+
"grad_norm": 3.8417811393737793,
|
| 513 |
+
"learning_rate": 0.0001889776357827476,
|
| 514 |
+
"loss": 0.7502,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.35,
|
| 519 |
+
"grad_norm": 3.6006288528442383,
|
| 520 |
+
"learning_rate": 0.0001884451544195953,
|
| 521 |
+
"loss": 0.6141,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.37,
|
| 526 |
+
"grad_norm": 1.9083914756774902,
|
| 527 |
+
"learning_rate": 0.00018791267305644303,
|
| 528 |
+
"loss": 0.6221,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.39,
|
| 533 |
+
"grad_norm": 2.8606581687927246,
|
| 534 |
+
"learning_rate": 0.00018738019169329076,
|
| 535 |
+
"loss": 0.593,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.41,
|
| 540 |
+
"grad_norm": 1.3666915893554688,
|
| 541 |
+
"learning_rate": 0.00018684771033013845,
|
| 542 |
+
"loss": 0.6307,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.43,
|
| 547 |
+
"grad_norm": 0.7459832429885864,
|
| 548 |
+
"learning_rate": 0.00018631522896698615,
|
| 549 |
+
"loss": 0.6774,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.45,
|
| 554 |
+
"grad_norm": 0.8331255912780762,
|
| 555 |
+
"learning_rate": 0.00018578274760383388,
|
| 556 |
+
"loss": 0.6637,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.47,
|
| 561 |
+
"grad_norm": 4.606655120849609,
|
| 562 |
+
"learning_rate": 0.00018525026624068158,
|
| 563 |
+
"loss": 0.5936,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.48,
|
| 568 |
+
"grad_norm": 0.44328516721725464,
|
| 569 |
+
"learning_rate": 0.00018471778487752928,
|
| 570 |
+
"loss": 0.6676,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.5,
|
| 575 |
+
"grad_norm": 4.314159393310547,
|
| 576 |
+
"learning_rate": 0.00018418530351437703,
|
| 577 |
+
"loss": 0.7248,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.52,
|
| 582 |
+
"grad_norm": 1.6179046630859375,
|
| 583 |
+
"learning_rate": 0.00018365282215122473,
|
| 584 |
+
"loss": 0.673,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.54,
|
| 589 |
+
"grad_norm": 4.103793144226074,
|
| 590 |
+
"learning_rate": 0.00018312034078807243,
|
| 591 |
+
"loss": 0.6637,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.56,
|
| 596 |
+
"grad_norm": 0.7133169770240784,
|
| 597 |
+
"learning_rate": 0.00018258785942492015,
|
| 598 |
+
"loss": 0.6193,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.58,
|
| 603 |
+
"grad_norm": 1.252029299736023,
|
| 604 |
+
"learning_rate": 0.00018205537806176785,
|
| 605 |
+
"loss": 0.594,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.6,
|
| 610 |
+
"grad_norm": 0.786902904510498,
|
| 611 |
+
"learning_rate": 0.00018152289669861555,
|
| 612 |
+
"loss": 0.5598,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.62,
|
| 617 |
+
"grad_norm": 5.060324668884277,
|
| 618 |
+
"learning_rate": 0.00018099041533546325,
|
| 619 |
+
"loss": 0.6586,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.64,
|
| 624 |
+
"grad_norm": 2.4381039142608643,
|
| 625 |
+
"learning_rate": 0.000180457933972311,
|
| 626 |
+
"loss": 0.6487,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.65,
|
| 631 |
+
"grad_norm": 2.3903331756591797,
|
| 632 |
+
"learning_rate": 0.0001799254526091587,
|
| 633 |
+
"loss": 0.624,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.67,
|
| 638 |
+
"grad_norm": 1.2366740703582764,
|
| 639 |
+
"learning_rate": 0.0001793929712460064,
|
| 640 |
+
"loss": 0.5997,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.69,
|
| 645 |
+
"grad_norm": 3.3629353046417236,
|
| 646 |
+
"learning_rate": 0.00017886048988285412,
|
| 647 |
+
"loss": 0.5978,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.71,
|
| 652 |
+
"grad_norm": 2.4150805473327637,
|
| 653 |
+
"learning_rate": 0.00017832800851970182,
|
| 654 |
+
"loss": 0.6342,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.73,
|
| 659 |
+
"grad_norm": 0.2921897768974304,
|
| 660 |
+
"learning_rate": 0.00017779552715654952,
|
| 661 |
+
"loss": 0.655,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.75,
|
| 666 |
+
"grad_norm": 2.8717799186706543,
|
| 667 |
+
"learning_rate": 0.00017726304579339722,
|
| 668 |
+
"loss": 0.6383,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.77,
|
| 673 |
+
"grad_norm": 1.2297135591506958,
|
| 674 |
+
"learning_rate": 0.00017673056443024497,
|
| 675 |
+
"loss": 0.6154,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.79,
|
| 680 |
+
"grad_norm": 2.8782236576080322,
|
| 681 |
+
"learning_rate": 0.00017619808306709267,
|
| 682 |
+
"loss": 0.6469,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.8,
|
| 687 |
+
"grad_norm": 3.708266019821167,
|
| 688 |
+
"learning_rate": 0.00017566560170394037,
|
| 689 |
+
"loss": 0.6462,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.82,
|
| 694 |
+
"grad_norm": 0.30842864513397217,
|
| 695 |
+
"learning_rate": 0.0001751331203407881,
|
| 696 |
+
"loss": 0.5934,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.84,
|
| 701 |
+
"grad_norm": 0.9341667294502258,
|
| 702 |
+
"learning_rate": 0.0001746006389776358,
|
| 703 |
+
"loss": 0.5741,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.86,
|
| 708 |
+
"grad_norm": 1.7768970727920532,
|
| 709 |
+
"learning_rate": 0.0001740681576144835,
|
| 710 |
+
"loss": 0.6028,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.88,
|
| 715 |
+
"grad_norm": 4.887693881988525,
|
| 716 |
+
"learning_rate": 0.0001735356762513312,
|
| 717 |
+
"loss": 0.6171,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.9,
|
| 722 |
+
"grad_norm": 2.1910202503204346,
|
| 723 |
+
"learning_rate": 0.00017300319488817894,
|
| 724 |
+
"loss": 0.6039,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.92,
|
| 729 |
+
"grad_norm": 5.127533912658691,
|
| 730 |
+
"learning_rate": 0.00017247071352502664,
|
| 731 |
+
"loss": 0.6404,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.94,
|
| 736 |
+
"grad_norm": 4.148878574371338,
|
| 737 |
+
"learning_rate": 0.00017193823216187434,
|
| 738 |
+
"loss": 0.6367,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.95,
|
| 743 |
+
"grad_norm": 0.9742037057876587,
|
| 744 |
+
"learning_rate": 0.00017140575079872207,
|
| 745 |
+
"loss": 0.6539,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.97,
|
| 750 |
+
"grad_norm": 7.3390913009643555,
|
| 751 |
+
"learning_rate": 0.00017087326943556976,
|
| 752 |
+
"loss": 0.6369,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.99,
|
| 757 |
+
"grad_norm": 0.8897514343261719,
|
| 758 |
+
"learning_rate": 0.00017034078807241746,
|
| 759 |
+
"loss": 0.6203,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.0,
|
| 764 |
+
"eval_accuracy": 0.498,
|
| 765 |
+
"eval_loss": 0.746105968952179,
|
| 766 |
+
"eval_runtime": 5.0077,
|
| 767 |
+
"eval_samples_per_second": 99.846,
|
| 768 |
+
"eval_steps_per_second": 3.195,
|
| 769 |
+
"step": 1064
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.01,
|
| 773 |
+
"grad_norm": 5.566906452178955,
|
| 774 |
+
"learning_rate": 0.0001698083067092652,
|
| 775 |
+
"loss": 0.7104,
|
| 776 |
+
"step": 1070
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.03,
|
| 780 |
+
"grad_norm": 4.016073703765869,
|
| 781 |
+
"learning_rate": 0.00016927582534611291,
|
| 782 |
+
"loss": 0.6763,
|
| 783 |
+
"step": 1080
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 2.05,
|
| 787 |
+
"grad_norm": 0.48662686347961426,
|
| 788 |
+
"learning_rate": 0.0001687433439829606,
|
| 789 |
+
"loss": 0.6283,
|
| 790 |
+
"step": 1090
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.07,
|
| 794 |
+
"grad_norm": 3.80304217338562,
|
| 795 |
+
"learning_rate": 0.0001682108626198083,
|
| 796 |
+
"loss": 0.5875,
|
| 797 |
+
"step": 1100
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 2.09,
|
| 801 |
+
"grad_norm": 3.6678249835968018,
|
| 802 |
+
"learning_rate": 0.00016767838125665604,
|
| 803 |
+
"loss": 0.6383,
|
| 804 |
+
"step": 1110
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.11,
|
| 808 |
+
"grad_norm": 0.7282447814941406,
|
| 809 |
+
"learning_rate": 0.00016714589989350374,
|
| 810 |
+
"loss": 0.5975,
|
| 811 |
+
"step": 1120
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.12,
|
| 815 |
+
"grad_norm": 5.8518829345703125,
|
| 816 |
+
"learning_rate": 0.00016661341853035143,
|
| 817 |
+
"loss": 0.6223,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.14,
|
| 822 |
+
"grad_norm": 0.6626885533332825,
|
| 823 |
+
"learning_rate": 0.00016608093716719916,
|
| 824 |
+
"loss": 0.6059,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.16,
|
| 829 |
+
"grad_norm": 2.4838600158691406,
|
| 830 |
+
"learning_rate": 0.00016554845580404689,
|
| 831 |
+
"loss": 0.6218,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.18,
|
| 836 |
+
"grad_norm": 3.6144859790802,
|
| 837 |
+
"learning_rate": 0.00016501597444089458,
|
| 838 |
+
"loss": 0.601,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.2,
|
| 843 |
+
"grad_norm": 0.9720486998558044,
|
| 844 |
+
"learning_rate": 0.00016448349307774228,
|
| 845 |
+
"loss": 0.5471,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.22,
|
| 850 |
+
"grad_norm": 5.14749002456665,
|
| 851 |
+
"learning_rate": 0.00016395101171459,
|
| 852 |
+
"loss": 0.6341,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.24,
|
| 857 |
+
"grad_norm": 4.361828804016113,
|
| 858 |
+
"learning_rate": 0.0001634185303514377,
|
| 859 |
+
"loss": 0.5915,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.26,
|
| 864 |
+
"grad_norm": 1.6119446754455566,
|
| 865 |
+
"learning_rate": 0.0001628860489882854,
|
| 866 |
+
"loss": 0.658,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.27,
|
| 871 |
+
"grad_norm": 0.6508371233940125,
|
| 872 |
+
"learning_rate": 0.00016235356762513313,
|
| 873 |
+
"loss": 0.5626,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.29,
|
| 878 |
+
"grad_norm": 0.7702106237411499,
|
| 879 |
+
"learning_rate": 0.00016182108626198086,
|
| 880 |
+
"loss": 0.6118,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.31,
|
| 885 |
+
"grad_norm": 0.45979470014572144,
|
| 886 |
+
"learning_rate": 0.00016128860489882855,
|
| 887 |
+
"loss": 0.6097,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.33,
|
| 892 |
+
"grad_norm": 1.6143121719360352,
|
| 893 |
+
"learning_rate": 0.00016075612353567625,
|
| 894 |
+
"loss": 0.6675,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.35,
|
| 899 |
+
"grad_norm": 2.4905166625976562,
|
| 900 |
+
"learning_rate": 0.00016022364217252398,
|
| 901 |
+
"loss": 0.5858,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.37,
|
| 906 |
+
"grad_norm": 0.7425104379653931,
|
| 907 |
+
"learning_rate": 0.00015969116080937168,
|
| 908 |
+
"loss": 0.5901,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.39,
|
| 913 |
+
"grad_norm": 1.7030450105667114,
|
| 914 |
+
"learning_rate": 0.00015915867944621938,
|
| 915 |
+
"loss": 0.5834,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.41,
|
| 920 |
+
"grad_norm": 3.350250720977783,
|
| 921 |
+
"learning_rate": 0.0001586261980830671,
|
| 922 |
+
"loss": 0.554,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.42,
|
| 927 |
+
"grad_norm": 2.625653028488159,
|
| 928 |
+
"learning_rate": 0.00015809371671991483,
|
| 929 |
+
"loss": 0.5784,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.44,
|
| 934 |
+
"grad_norm": 1.9641324281692505,
|
| 935 |
+
"learning_rate": 0.00015756123535676253,
|
| 936 |
+
"loss": 0.6203,
|
| 937 |
+
"step": 1300
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.46,
|
| 941 |
+
"grad_norm": 1.912189245223999,
|
| 942 |
+
"learning_rate": 0.00015702875399361022,
|
| 943 |
+
"loss": 0.6344,
|
| 944 |
+
"step": 1310
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.48,
|
| 948 |
+
"grad_norm": 2.581179141998291,
|
| 949 |
+
"learning_rate": 0.00015649627263045795,
|
| 950 |
+
"loss": 0.6706,
|
| 951 |
+
"step": 1320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.5,
|
| 955 |
+
"grad_norm": 0.7273750901222229,
|
| 956 |
+
"learning_rate": 0.00015596379126730565,
|
| 957 |
+
"loss": 0.6587,
|
| 958 |
+
"step": 1330
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.52,
|
| 962 |
+
"grad_norm": 0.5495343804359436,
|
| 963 |
+
"learning_rate": 0.00015543130990415335,
|
| 964 |
+
"loss": 0.6005,
|
| 965 |
+
"step": 1340
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.54,
|
| 969 |
+
"grad_norm": 2.7823123931884766,
|
| 970 |
+
"learning_rate": 0.00015489882854100107,
|
| 971 |
+
"loss": 0.5477,
|
| 972 |
+
"step": 1350
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.56,
|
| 976 |
+
"grad_norm": 2.560255527496338,
|
| 977 |
+
"learning_rate": 0.0001543663471778488,
|
| 978 |
+
"loss": 0.6153,
|
| 979 |
+
"step": 1360
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.58,
|
| 983 |
+
"grad_norm": 4.055157661437988,
|
| 984 |
+
"learning_rate": 0.0001538338658146965,
|
| 985 |
+
"loss": 0.6093,
|
| 986 |
+
"step": 1370
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.59,
|
| 990 |
+
"grad_norm": 1.9480986595153809,
|
| 991 |
+
"learning_rate": 0.0001533013844515442,
|
| 992 |
+
"loss": 0.653,
|
| 993 |
+
"step": 1380
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.61,
|
| 997 |
+
"grad_norm": 0.6847740411758423,
|
| 998 |
+
"learning_rate": 0.00015276890308839192,
|
| 999 |
+
"loss": 0.5842,
|
| 1000 |
+
"step": 1390
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.63,
|
| 1004 |
+
"grad_norm": 5.935375690460205,
|
| 1005 |
+
"learning_rate": 0.00015223642172523962,
|
| 1006 |
+
"loss": 0.5934,
|
| 1007 |
+
"step": 1400
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.65,
|
| 1011 |
+
"grad_norm": 2.9190566539764404,
|
| 1012 |
+
"learning_rate": 0.00015170394036208732,
|
| 1013 |
+
"loss": 0.5746,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.67,
|
| 1018 |
+
"grad_norm": 1.8768565654754639,
|
| 1019 |
+
"learning_rate": 0.00015117145899893504,
|
| 1020 |
+
"loss": 0.5792,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.69,
|
| 1025 |
+
"grad_norm": 0.8437117338180542,
|
| 1026 |
+
"learning_rate": 0.00015063897763578277,
|
| 1027 |
+
"loss": 0.6118,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.71,
|
| 1032 |
+
"grad_norm": 2.211174488067627,
|
| 1033 |
+
"learning_rate": 0.00015010649627263047,
|
| 1034 |
+
"loss": 0.5693,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.73,
|
| 1039 |
+
"grad_norm": 4.7668609619140625,
|
| 1040 |
+
"learning_rate": 0.00014957401490947817,
|
| 1041 |
+
"loss": 0.6926,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.74,
|
| 1046 |
+
"grad_norm": 1.8787531852722168,
|
| 1047 |
+
"learning_rate": 0.0001490415335463259,
|
| 1048 |
+
"loss": 0.5893,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.76,
|
| 1053 |
+
"grad_norm": 3.1121323108673096,
|
| 1054 |
+
"learning_rate": 0.0001485090521831736,
|
| 1055 |
+
"loss": 0.6001,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.78,
|
| 1060 |
+
"grad_norm": 1.5263334512710571,
|
| 1061 |
+
"learning_rate": 0.0001479765708200213,
|
| 1062 |
+
"loss": 0.6538,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.8,
|
| 1067 |
+
"grad_norm": 1.4740121364593506,
|
| 1068 |
+
"learning_rate": 0.00014744408945686902,
|
| 1069 |
+
"loss": 0.633,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.82,
|
| 1074 |
+
"grad_norm": 2.5917043685913086,
|
| 1075 |
+
"learning_rate": 0.00014691160809371674,
|
| 1076 |
+
"loss": 0.619,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.84,
|
| 1081 |
+
"grad_norm": 1.5224074125289917,
|
| 1082 |
+
"learning_rate": 0.00014637912673056444,
|
| 1083 |
+
"loss": 0.5376,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.86,
|
| 1088 |
+
"grad_norm": 1.7697654962539673,
|
| 1089 |
+
"learning_rate": 0.00014584664536741214,
|
| 1090 |
+
"loss": 0.5838,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.88,
|
| 1095 |
+
"grad_norm": 4.307724475860596,
|
| 1096 |
+
"learning_rate": 0.00014531416400425986,
|
| 1097 |
+
"loss": 0.5861,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.89,
|
| 1102 |
+
"grad_norm": 2.4162418842315674,
|
| 1103 |
+
"learning_rate": 0.00014478168264110756,
|
| 1104 |
+
"loss": 0.7074,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.91,
|
| 1109 |
+
"grad_norm": 3.3249616622924805,
|
| 1110 |
+
"learning_rate": 0.00014424920127795526,
|
| 1111 |
+
"loss": 0.6251,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.93,
|
| 1116 |
+
"grad_norm": 3.486440420150757,
|
| 1117 |
+
"learning_rate": 0.000143716719914803,
|
| 1118 |
+
"loss": 0.6949,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.95,
|
| 1123 |
+
"grad_norm": 1.9070874452590942,
|
| 1124 |
+
"learning_rate": 0.0001431842385516507,
|
| 1125 |
+
"loss": 0.6384,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.97,
|
| 1130 |
+
"grad_norm": 0.9392743706703186,
|
| 1131 |
+
"learning_rate": 0.0001426517571884984,
|
| 1132 |
+
"loss": 0.6292,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.99,
|
| 1137 |
+
"grad_norm": 1.008888840675354,
|
| 1138 |
+
"learning_rate": 0.0001421192758253461,
|
| 1139 |
+
"loss": 0.5955,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 3.0,
|
| 1144 |
+
"eval_accuracy": 0.518,
|
| 1145 |
+
"eval_loss": 0.7538835406303406,
|
| 1146 |
+
"eval_runtime": 5.5893,
|
| 1147 |
+
"eval_samples_per_second": 89.456,
|
| 1148 |
+
"eval_steps_per_second": 2.863,
|
| 1149 |
+
"step": 1596
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 3.01,
|
| 1153 |
+
"grad_norm": 1.014275074005127,
|
| 1154 |
+
"learning_rate": 0.00014158679446219384,
|
| 1155 |
+
"loss": 0.6254,
|
| 1156 |
+
"step": 1600
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 3.03,
|
| 1160 |
+
"grad_norm": 3.639158248901367,
|
| 1161 |
+
"learning_rate": 0.00014105431309904153,
|
| 1162 |
+
"loss": 0.5718,
|
| 1163 |
+
"step": 1610
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 3.05,
|
| 1167 |
+
"grad_norm": 3.8647546768188477,
|
| 1168 |
+
"learning_rate": 0.00014052183173588923,
|
| 1169 |
+
"loss": 0.5798,
|
| 1170 |
+
"step": 1620
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 3.06,
|
| 1174 |
+
"grad_norm": 2.522571086883545,
|
| 1175 |
+
"learning_rate": 0.00013998935037273696,
|
| 1176 |
+
"loss": 0.596,
|
| 1177 |
+
"step": 1630
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 3.08,
|
| 1181 |
+
"grad_norm": 0.9586620330810547,
|
| 1182 |
+
"learning_rate": 0.00013945686900958468,
|
| 1183 |
+
"loss": 0.546,
|
| 1184 |
+
"step": 1640
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 3.1,
|
| 1188 |
+
"grad_norm": 1.2226234674453735,
|
| 1189 |
+
"learning_rate": 0.00013892438764643238,
|
| 1190 |
+
"loss": 0.5209,
|
| 1191 |
+
"step": 1650
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 3.12,
|
| 1195 |
+
"grad_norm": 1.6508853435516357,
|
| 1196 |
+
"learning_rate": 0.00013839190628328008,
|
| 1197 |
+
"loss": 0.5763,
|
| 1198 |
+
"step": 1660
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 3.14,
|
| 1202 |
+
"grad_norm": 15.09642505645752,
|
| 1203 |
+
"learning_rate": 0.0001378594249201278,
|
| 1204 |
+
"loss": 0.5614,
|
| 1205 |
+
"step": 1670
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 3.16,
|
| 1209 |
+
"grad_norm": 2.6017839908599854,
|
| 1210 |
+
"learning_rate": 0.0001373269435569755,
|
| 1211 |
+
"loss": 0.6123,
|
| 1212 |
+
"step": 1680
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 3.18,
|
| 1216 |
+
"grad_norm": 2.848982572555542,
|
| 1217 |
+
"learning_rate": 0.0001367944621938232,
|
| 1218 |
+
"loss": 0.5747,
|
| 1219 |
+
"step": 1690
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 3.2,
|
| 1223 |
+
"grad_norm": 2.5202741622924805,
|
| 1224 |
+
"learning_rate": 0.00013626198083067093,
|
| 1225 |
+
"loss": 0.5505,
|
| 1226 |
+
"step": 1700
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.21,
|
| 1230 |
+
"grad_norm": 1.2070810794830322,
|
| 1231 |
+
"learning_rate": 0.00013572949946751866,
|
| 1232 |
+
"loss": 0.4952,
|
| 1233 |
+
"step": 1710
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 3.23,
|
| 1237 |
+
"grad_norm": 1.999449610710144,
|
| 1238 |
+
"learning_rate": 0.00013519701810436635,
|
| 1239 |
+
"loss": 0.4874,
|
| 1240 |
+
"step": 1720
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 3.25,
|
| 1244 |
+
"grad_norm": 1.2801100015640259,
|
| 1245 |
+
"learning_rate": 0.00013466453674121405,
|
| 1246 |
+
"loss": 0.568,
|
| 1247 |
+
"step": 1730
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 3.27,
|
| 1251 |
+
"grad_norm": 1.8408716917037964,
|
| 1252 |
+
"learning_rate": 0.00013413205537806178,
|
| 1253 |
+
"loss": 0.5894,
|
| 1254 |
+
"step": 1740
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 3.29,
|
| 1258 |
+
"grad_norm": 2.0327749252319336,
|
| 1259 |
+
"learning_rate": 0.00013359957401490948,
|
| 1260 |
+
"loss": 0.5749,
|
| 1261 |
+
"step": 1750
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 3.31,
|
| 1265 |
+
"grad_norm": 1.4758570194244385,
|
| 1266 |
+
"learning_rate": 0.00013306709265175718,
|
| 1267 |
+
"loss": 0.5308,
|
| 1268 |
+
"step": 1760
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.33,
|
| 1272 |
+
"grad_norm": 1.3630428314208984,
|
| 1273 |
+
"learning_rate": 0.00013253461128860493,
|
| 1274 |
+
"loss": 0.5475,
|
| 1275 |
+
"step": 1770
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.35,
|
| 1279 |
+
"grad_norm": 0.8149157166481018,
|
| 1280 |
+
"learning_rate": 0.00013200212992545263,
|
| 1281 |
+
"loss": 0.5295,
|
| 1282 |
+
"step": 1780
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 3.36,
|
| 1286 |
+
"grad_norm": 5.804959297180176,
|
| 1287 |
+
"learning_rate": 0.00013146964856230033,
|
| 1288 |
+
"loss": 0.5317,
|
| 1289 |
+
"step": 1790
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 3.38,
|
| 1293 |
+
"grad_norm": 1.0680649280548096,
|
| 1294 |
+
"learning_rate": 0.00013093716719914802,
|
| 1295 |
+
"loss": 0.5561,
|
| 1296 |
+
"step": 1800
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 3.4,
|
| 1300 |
+
"grad_norm": 2.4818127155303955,
|
| 1301 |
+
"learning_rate": 0.00013040468583599575,
|
| 1302 |
+
"loss": 0.6085,
|
| 1303 |
+
"step": 1810
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 3.42,
|
| 1307 |
+
"grad_norm": 1.4538164138793945,
|
| 1308 |
+
"learning_rate": 0.00012987220447284345,
|
| 1309 |
+
"loss": 0.5614,
|
| 1310 |
+
"step": 1820
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 3.44,
|
| 1314 |
+
"grad_norm": 2.130237102508545,
|
| 1315 |
+
"learning_rate": 0.00012933972310969115,
|
| 1316 |
+
"loss": 0.5255,
|
| 1317 |
+
"step": 1830
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 3.46,
|
| 1321 |
+
"grad_norm": 4.228801250457764,
|
| 1322 |
+
"learning_rate": 0.0001288072417465389,
|
| 1323 |
+
"loss": 0.592,
|
| 1324 |
+
"step": 1840
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 3.48,
|
| 1328 |
+
"grad_norm": 1.1142246723175049,
|
| 1329 |
+
"learning_rate": 0.0001282747603833866,
|
| 1330 |
+
"loss": 0.5262,
|
| 1331 |
+
"step": 1850
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 3.5,
|
| 1335 |
+
"grad_norm": 2.6029281616210938,
|
| 1336 |
+
"learning_rate": 0.0001277422790202343,
|
| 1337 |
+
"loss": 0.541,
|
| 1338 |
+
"step": 1860
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 3.52,
|
| 1342 |
+
"grad_norm": 3.902634382247925,
|
| 1343 |
+
"learning_rate": 0.000127209797657082,
|
| 1344 |
+
"loss": 0.5725,
|
| 1345 |
+
"step": 1870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 3.53,
|
| 1349 |
+
"grad_norm": 5.952896595001221,
|
| 1350 |
+
"learning_rate": 0.00012667731629392972,
|
| 1351 |
+
"loss": 0.597,
|
| 1352 |
+
"step": 1880
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.55,
|
| 1356 |
+
"grad_norm": 1.5133850574493408,
|
| 1357 |
+
"learning_rate": 0.00012614483493077742,
|
| 1358 |
+
"loss": 0.4974,
|
| 1359 |
+
"step": 1890
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 3.57,
|
| 1363 |
+
"grad_norm": 1.7601635456085205,
|
| 1364 |
+
"learning_rate": 0.00012561235356762512,
|
| 1365 |
+
"loss": 0.5825,
|
| 1366 |
+
"step": 1900
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 3.59,
|
| 1370 |
+
"grad_norm": 2.166303873062134,
|
| 1371 |
+
"learning_rate": 0.00012507987220447287,
|
| 1372 |
+
"loss": 0.4884,
|
| 1373 |
+
"step": 1910
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 3.61,
|
| 1377 |
+
"grad_norm": 1.2939203977584839,
|
| 1378 |
+
"learning_rate": 0.00012454739084132057,
|
| 1379 |
+
"loss": 0.5576,
|
| 1380 |
+
"step": 1920
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 3.63,
|
| 1384 |
+
"grad_norm": 3.2266693115234375,
|
| 1385 |
+
"learning_rate": 0.00012401490947816827,
|
| 1386 |
+
"loss": 0.5598,
|
| 1387 |
+
"step": 1930
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 3.65,
|
| 1391 |
+
"grad_norm": 1.2390942573547363,
|
| 1392 |
+
"learning_rate": 0.00012348242811501597,
|
| 1393 |
+
"loss": 0.507,
|
| 1394 |
+
"step": 1940
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 3.67,
|
| 1398 |
+
"grad_norm": 1.1552023887634277,
|
| 1399 |
+
"learning_rate": 0.0001229499467518637,
|
| 1400 |
+
"loss": 0.5781,
|
| 1401 |
+
"step": 1950
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 3.68,
|
| 1405 |
+
"grad_norm": 1.372052788734436,
|
| 1406 |
+
"learning_rate": 0.0001224174653887114,
|
| 1407 |
+
"loss": 0.5315,
|
| 1408 |
+
"step": 1960
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 3.7,
|
| 1412 |
+
"grad_norm": 2.0448787212371826,
|
| 1413 |
+
"learning_rate": 0.0001218849840255591,
|
| 1414 |
+
"loss": 0.5402,
|
| 1415 |
+
"step": 1970
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 3.72,
|
| 1419 |
+
"grad_norm": 1.5772067308425903,
|
| 1420 |
+
"learning_rate": 0.00012135250266240683,
|
| 1421 |
+
"loss": 0.518,
|
| 1422 |
+
"step": 1980
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 3.74,
|
| 1426 |
+
"grad_norm": 2.118274450302124,
|
| 1427 |
+
"learning_rate": 0.00012082002129925454,
|
| 1428 |
+
"loss": 0.5661,
|
| 1429 |
+
"step": 1990
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 3.76,
|
| 1433 |
+
"grad_norm": 2.9228522777557373,
|
| 1434 |
+
"learning_rate": 0.00012028753993610224,
|
| 1435 |
+
"loss": 0.5419,
|
| 1436 |
+
"step": 2000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 3.78,
|
| 1440 |
+
"grad_norm": 3.01971697807312,
|
| 1441 |
+
"learning_rate": 0.00011975505857294995,
|
| 1442 |
+
"loss": 0.5731,
|
| 1443 |
+
"step": 2010
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 3.8,
|
| 1447 |
+
"grad_norm": 1.2747687101364136,
|
| 1448 |
+
"learning_rate": 0.00011922257720979766,
|
| 1449 |
+
"loss": 0.5621,
|
| 1450 |
+
"step": 2020
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 3.82,
|
| 1454 |
+
"grad_norm": 2.164107084274292,
|
| 1455 |
+
"learning_rate": 0.00011869009584664536,
|
| 1456 |
+
"loss": 0.5818,
|
| 1457 |
+
"step": 2030
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 3.83,
|
| 1461 |
+
"grad_norm": 1.2156935930252075,
|
| 1462 |
+
"learning_rate": 0.00011815761448349307,
|
| 1463 |
+
"loss": 0.561,
|
| 1464 |
+
"step": 2040
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 3.85,
|
| 1468 |
+
"grad_norm": 1.5158003568649292,
|
| 1469 |
+
"learning_rate": 0.0001176251331203408,
|
| 1470 |
+
"loss": 0.5255,
|
| 1471 |
+
"step": 2050
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 3.87,
|
| 1475 |
+
"grad_norm": 1.795130729675293,
|
| 1476 |
+
"learning_rate": 0.00011709265175718851,
|
| 1477 |
+
"loss": 0.5451,
|
| 1478 |
+
"step": 2060
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 3.89,
|
| 1482 |
+
"grad_norm": 2.8883111476898193,
|
| 1483 |
+
"learning_rate": 0.00011656017039403621,
|
| 1484 |
+
"loss": 0.539,
|
| 1485 |
+
"step": 2070
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 3.91,
|
| 1489 |
+
"grad_norm": 2.8420214653015137,
|
| 1490 |
+
"learning_rate": 0.00011602768903088392,
|
| 1491 |
+
"loss": 0.5655,
|
| 1492 |
+
"step": 2080
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 3.93,
|
| 1496 |
+
"grad_norm": 1.114376425743103,
|
| 1497 |
+
"learning_rate": 0.00011549520766773163,
|
| 1498 |
+
"loss": 0.4676,
|
| 1499 |
+
"step": 2090
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 3.95,
|
| 1503 |
+
"grad_norm": 1.6993155479431152,
|
| 1504 |
+
"learning_rate": 0.00011496272630457933,
|
| 1505 |
+
"loss": 0.5308,
|
| 1506 |
+
"step": 2100
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 3.97,
|
| 1510 |
+
"grad_norm": 1.809404969215393,
|
| 1511 |
+
"learning_rate": 0.00011443024494142705,
|
| 1512 |
+
"loss": 0.4535,
|
| 1513 |
+
"step": 2110
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 3.98,
|
| 1517 |
+
"grad_norm": 3.3252310752868652,
|
| 1518 |
+
"learning_rate": 0.00011389776357827477,
|
| 1519 |
+
"loss": 0.4796,
|
| 1520 |
+
"step": 2120
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 4.0,
|
| 1524 |
+
"eval_accuracy": 0.536,
|
| 1525 |
+
"eval_loss": 0.9386767745018005,
|
| 1526 |
+
"eval_runtime": 5.6156,
|
| 1527 |
+
"eval_samples_per_second": 89.037,
|
| 1528 |
+
"eval_steps_per_second": 2.849,
|
| 1529 |
+
"step": 2128
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 4.0,
|
| 1533 |
+
"grad_norm": 1.6969894170761108,
|
| 1534 |
+
"learning_rate": 0.00011336528221512248,
|
| 1535 |
+
"loss": 0.5216,
|
| 1536 |
+
"step": 2130
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 4.02,
|
| 1540 |
+
"grad_norm": 3.2442750930786133,
|
| 1541 |
+
"learning_rate": 0.00011283280085197018,
|
| 1542 |
+
"loss": 0.4074,
|
| 1543 |
+
"step": 2140
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 4.04,
|
| 1547 |
+
"grad_norm": 2.2770748138427734,
|
| 1548 |
+
"learning_rate": 0.0001123003194888179,
|
| 1549 |
+
"loss": 0.3409,
|
| 1550 |
+
"step": 2150
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.06,
|
| 1554 |
+
"grad_norm": 2.7340738773345947,
|
| 1555 |
+
"learning_rate": 0.0001117678381256656,
|
| 1556 |
+
"loss": 0.4351,
|
| 1557 |
+
"step": 2160
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 4.08,
|
| 1561 |
+
"grad_norm": 2.154231071472168,
|
| 1562 |
+
"learning_rate": 0.0001112353567625133,
|
| 1563 |
+
"loss": 0.4451,
|
| 1564 |
+
"step": 2170
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 4.1,
|
| 1568 |
+
"grad_norm": 2.5511295795440674,
|
| 1569 |
+
"learning_rate": 0.00011070287539936102,
|
| 1570 |
+
"loss": 0.4674,
|
| 1571 |
+
"step": 2180
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 4.12,
|
| 1575 |
+
"grad_norm": 1.8929212093353271,
|
| 1576 |
+
"learning_rate": 0.00011017039403620874,
|
| 1577 |
+
"loss": 0.4184,
|
| 1578 |
+
"step": 2190
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 4.14,
|
| 1582 |
+
"grad_norm": 1.6195924282073975,
|
| 1583 |
+
"learning_rate": 0.00010963791267305645,
|
| 1584 |
+
"loss": 0.3456,
|
| 1585 |
+
"step": 2200
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 4.15,
|
| 1589 |
+
"grad_norm": 2.4593372344970703,
|
| 1590 |
+
"learning_rate": 0.00010910543130990417,
|
| 1591 |
+
"loss": 0.3398,
|
| 1592 |
+
"step": 2210
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 4.17,
|
| 1596 |
+
"grad_norm": 4.4898457527160645,
|
| 1597 |
+
"learning_rate": 0.00010857294994675187,
|
| 1598 |
+
"loss": 0.3741,
|
| 1599 |
+
"step": 2220
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 4.19,
|
| 1603 |
+
"grad_norm": 3.513038158416748,
|
| 1604 |
+
"learning_rate": 0.00010804046858359958,
|
| 1605 |
+
"loss": 0.5035,
|
| 1606 |
+
"step": 2230
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 4.21,
|
| 1610 |
+
"grad_norm": 2.14664626121521,
|
| 1611 |
+
"learning_rate": 0.00010750798722044728,
|
| 1612 |
+
"loss": 0.3854,
|
| 1613 |
+
"step": 2240
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 4.23,
|
| 1617 |
+
"grad_norm": 1.8807299137115479,
|
| 1618 |
+
"learning_rate": 0.00010697550585729502,
|
| 1619 |
+
"loss": 0.3503,
|
| 1620 |
+
"step": 2250
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 4.25,
|
| 1624 |
+
"grad_norm": 3.9768781661987305,
|
| 1625 |
+
"learning_rate": 0.00010644302449414271,
|
| 1626 |
+
"loss": 0.4615,
|
| 1627 |
+
"step": 2260
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.27,
|
| 1631 |
+
"grad_norm": 4.187206745147705,
|
| 1632 |
+
"learning_rate": 0.00010591054313099043,
|
| 1633 |
+
"loss": 0.3798,
|
| 1634 |
+
"step": 2270
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.29,
|
| 1638 |
+
"grad_norm": 5.973431587219238,
|
| 1639 |
+
"learning_rate": 0.00010537806176783814,
|
| 1640 |
+
"loss": 0.4394,
|
| 1641 |
+
"step": 2280
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 4.3,
|
| 1645 |
+
"grad_norm": 2.4600276947021484,
|
| 1646 |
+
"learning_rate": 0.00010484558040468584,
|
| 1647 |
+
"loss": 0.4936,
|
| 1648 |
+
"step": 2290
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 4.32,
|
| 1652 |
+
"grad_norm": 2.6891305446624756,
|
| 1653 |
+
"learning_rate": 0.00010431309904153355,
|
| 1654 |
+
"loss": 0.4825,
|
| 1655 |
+
"step": 2300
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 4.34,
|
| 1659 |
+
"grad_norm": 3.481198787689209,
|
| 1660 |
+
"learning_rate": 0.00010378061767838125,
|
| 1661 |
+
"loss": 0.5209,
|
| 1662 |
+
"step": 2310
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 4.36,
|
| 1666 |
+
"grad_norm": 1.2965519428253174,
|
| 1667 |
+
"learning_rate": 0.0001033013844515442,
|
| 1668 |
+
"loss": 0.4865,
|
| 1669 |
+
"step": 2320
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 4.38,
|
| 1673 |
+
"grad_norm": 1.5025973320007324,
|
| 1674 |
+
"learning_rate": 0.00010276890308839191,
|
| 1675 |
+
"loss": 0.4439,
|
| 1676 |
+
"step": 2330
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.4,
|
| 1680 |
+
"grad_norm": 3.154419422149658,
|
| 1681 |
+
"learning_rate": 0.00010223642172523961,
|
| 1682 |
+
"loss": 0.4129,
|
| 1683 |
+
"step": 2340
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 4.42,
|
| 1687 |
+
"grad_norm": 3.1447813510894775,
|
| 1688 |
+
"learning_rate": 0.00010170394036208732,
|
| 1689 |
+
"loss": 0.4973,
|
| 1690 |
+
"step": 2350
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 4.44,
|
| 1694 |
+
"grad_norm": 2.3173677921295166,
|
| 1695 |
+
"learning_rate": 0.00010117145899893504,
|
| 1696 |
+
"loss": 0.4538,
|
| 1697 |
+
"step": 2360
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 4.45,
|
| 1701 |
+
"grad_norm": 1.7222325801849365,
|
| 1702 |
+
"learning_rate": 0.00010063897763578276,
|
| 1703 |
+
"loss": 0.4326,
|
| 1704 |
+
"step": 2370
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.47,
|
| 1708 |
+
"grad_norm": 1.9806593656539917,
|
| 1709 |
+
"learning_rate": 0.00010010649627263047,
|
| 1710 |
+
"loss": 0.4568,
|
| 1711 |
+
"step": 2380
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 4.49,
|
| 1715 |
+
"grad_norm": 1.6622909307479858,
|
| 1716 |
+
"learning_rate": 9.957401490947817e-05,
|
| 1717 |
+
"loss": 0.4504,
|
| 1718 |
+
"step": 2390
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.51,
|
| 1722 |
+
"grad_norm": 1.593268871307373,
|
| 1723 |
+
"learning_rate": 9.904153354632588e-05,
|
| 1724 |
+
"loss": 0.3772,
|
| 1725 |
+
"step": 2400
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 4.53,
|
| 1729 |
+
"grad_norm": 2.6867918968200684,
|
| 1730 |
+
"learning_rate": 9.850905218317358e-05,
|
| 1731 |
+
"loss": 0.4217,
|
| 1732 |
+
"step": 2410
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 4.55,
|
| 1736 |
+
"grad_norm": 2.0145435333251953,
|
| 1737 |
+
"learning_rate": 9.797657082002131e-05,
|
| 1738 |
+
"loss": 0.3476,
|
| 1739 |
+
"step": 2420
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 4.57,
|
| 1743 |
+
"grad_norm": 2.0304441452026367,
|
| 1744 |
+
"learning_rate": 9.744408945686902e-05,
|
| 1745 |
+
"loss": 0.3252,
|
| 1746 |
+
"step": 2430
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 4.59,
|
| 1750 |
+
"grad_norm": 4.424630165100098,
|
| 1751 |
+
"learning_rate": 9.691160809371672e-05,
|
| 1752 |
+
"loss": 0.5274,
|
| 1753 |
+
"step": 2440
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 4.61,
|
| 1757 |
+
"grad_norm": 1.801672101020813,
|
| 1758 |
+
"learning_rate": 9.637912673056444e-05,
|
| 1759 |
+
"loss": 0.4023,
|
| 1760 |
+
"step": 2450
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 4.62,
|
| 1764 |
+
"grad_norm": 2.912360906600952,
|
| 1765 |
+
"learning_rate": 9.584664536741214e-05,
|
| 1766 |
+
"loss": 0.3901,
|
| 1767 |
+
"step": 2460
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 4.64,
|
| 1771 |
+
"grad_norm": 2.676485538482666,
|
| 1772 |
+
"learning_rate": 9.531416400425986e-05,
|
| 1773 |
+
"loss": 0.4589,
|
| 1774 |
+
"step": 2470
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 4.66,
|
| 1778 |
+
"grad_norm": 2.1935267448425293,
|
| 1779 |
+
"learning_rate": 9.478168264110757e-05,
|
| 1780 |
+
"loss": 0.5248,
|
| 1781 |
+
"step": 2480
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 4.68,
|
| 1785 |
+
"grad_norm": 2.977781295776367,
|
| 1786 |
+
"learning_rate": 9.424920127795528e-05,
|
| 1787 |
+
"loss": 0.4278,
|
| 1788 |
+
"step": 2490
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 4.7,
|
| 1792 |
+
"grad_norm": 4.009756088256836,
|
| 1793 |
+
"learning_rate": 9.371671991480299e-05,
|
| 1794 |
+
"loss": 0.4132,
|
| 1795 |
+
"step": 2500
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 4.72,
|
| 1799 |
+
"grad_norm": 3.5155155658721924,
|
| 1800 |
+
"learning_rate": 9.318423855165069e-05,
|
| 1801 |
+
"loss": 0.3934,
|
| 1802 |
+
"step": 2510
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.74,
|
| 1806 |
+
"grad_norm": 2.687786817550659,
|
| 1807 |
+
"learning_rate": 9.265175718849842e-05,
|
| 1808 |
+
"loss": 0.3679,
|
| 1809 |
+
"step": 2520
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 4.76,
|
| 1813 |
+
"grad_norm": 2.9000587463378906,
|
| 1814 |
+
"learning_rate": 9.211927582534611e-05,
|
| 1815 |
+
"loss": 0.483,
|
| 1816 |
+
"step": 2530
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 4.77,
|
| 1820 |
+
"grad_norm": 2.172851085662842,
|
| 1821 |
+
"learning_rate": 9.158679446219383e-05,
|
| 1822 |
+
"loss": 0.3782,
|
| 1823 |
+
"step": 2540
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 4.79,
|
| 1827 |
+
"grad_norm": 4.356163024902344,
|
| 1828 |
+
"learning_rate": 9.105431309904154e-05,
|
| 1829 |
+
"loss": 0.4852,
|
| 1830 |
+
"step": 2550
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 4.81,
|
| 1834 |
+
"grad_norm": 3.4460835456848145,
|
| 1835 |
+
"learning_rate": 9.052183173588925e-05,
|
| 1836 |
+
"loss": 0.4019,
|
| 1837 |
+
"step": 2560
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 4.83,
|
| 1841 |
+
"grad_norm": 3.1064889430999756,
|
| 1842 |
+
"learning_rate": 8.998935037273696e-05,
|
| 1843 |
+
"loss": 0.397,
|
| 1844 |
+
"step": 2570
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.85,
|
| 1848 |
+
"grad_norm": 2.628739833831787,
|
| 1849 |
+
"learning_rate": 8.945686900958466e-05,
|
| 1850 |
+
"loss": 0.385,
|
| 1851 |
+
"step": 2580
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 4.87,
|
| 1855 |
+
"grad_norm": 1.598198413848877,
|
| 1856 |
+
"learning_rate": 8.892438764643239e-05,
|
| 1857 |
+
"loss": 0.2814,
|
| 1858 |
+
"step": 2590
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 4.89,
|
| 1862 |
+
"grad_norm": 4.517331123352051,
|
| 1863 |
+
"learning_rate": 8.839190628328009e-05,
|
| 1864 |
+
"loss": 0.3773,
|
| 1865 |
+
"step": 2600
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 4.91,
|
| 1869 |
+
"grad_norm": 3.0325510501861572,
|
| 1870 |
+
"learning_rate": 8.78594249201278e-05,
|
| 1871 |
+
"loss": 0.3999,
|
| 1872 |
+
"step": 2610
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 4.92,
|
| 1876 |
+
"grad_norm": 2.6737074851989746,
|
| 1877 |
+
"learning_rate": 8.732694355697551e-05,
|
| 1878 |
+
"loss": 0.4072,
|
| 1879 |
+
"step": 2620
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 4.94,
|
| 1883 |
+
"grad_norm": 2.7080023288726807,
|
| 1884 |
+
"learning_rate": 8.679446219382322e-05,
|
| 1885 |
+
"loss": 0.4635,
|
| 1886 |
+
"step": 2630
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.96,
|
| 1890 |
+
"grad_norm": 2.1088662147521973,
|
| 1891 |
+
"learning_rate": 8.626198083067093e-05,
|
| 1892 |
+
"loss": 0.4483,
|
| 1893 |
+
"step": 2640
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 4.98,
|
| 1897 |
+
"grad_norm": 1.194525957107544,
|
| 1898 |
+
"learning_rate": 8.572949946751863e-05,
|
| 1899 |
+
"loss": 0.4549,
|
| 1900 |
+
"step": 2650
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 5.0,
|
| 1904 |
+
"grad_norm": 12.147368431091309,
|
| 1905 |
+
"learning_rate": 8.519701810436636e-05,
|
| 1906 |
+
"loss": 0.425,
|
| 1907 |
+
"step": 2660
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 5.0,
|
| 1911 |
+
"eval_accuracy": 0.538,
|
| 1912 |
+
"eval_loss": 0.9824263453483582,
|
| 1913 |
+
"eval_runtime": 5.1851,
|
| 1914 |
+
"eval_samples_per_second": 96.431,
|
| 1915 |
+
"eval_steps_per_second": 3.086,
|
| 1916 |
+
"step": 2660
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 5.02,
|
| 1920 |
+
"grad_norm": 1.088959813117981,
|
| 1921 |
+
"learning_rate": 8.466453674121406e-05,
|
| 1922 |
+
"loss": 0.295,
|
| 1923 |
+
"step": 2670
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 5.04,
|
| 1927 |
+
"grad_norm": 1.6955100297927856,
|
| 1928 |
+
"learning_rate": 8.413205537806177e-05,
|
| 1929 |
+
"loss": 0.2041,
|
| 1930 |
+
"step": 2680
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 5.06,
|
| 1934 |
+
"grad_norm": 2.099268913269043,
|
| 1935 |
+
"learning_rate": 8.359957401490948e-05,
|
| 1936 |
+
"loss": 0.3164,
|
| 1937 |
+
"step": 2690
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 5.08,
|
| 1941 |
+
"grad_norm": 3.751316547393799,
|
| 1942 |
+
"learning_rate": 8.30670926517572e-05,
|
| 1943 |
+
"loss": 0.2167,
|
| 1944 |
+
"step": 2700
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 5.09,
|
| 1948 |
+
"grad_norm": 0.7041321396827698,
|
| 1949 |
+
"learning_rate": 8.25346112886049e-05,
|
| 1950 |
+
"loss": 0.3246,
|
| 1951 |
+
"step": 2710
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 5.11,
|
| 1955 |
+
"grad_norm": 1.6721878051757812,
|
| 1956 |
+
"learning_rate": 8.20021299254526e-05,
|
| 1957 |
+
"loss": 0.2365,
|
| 1958 |
+
"step": 2720
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 5.13,
|
| 1962 |
+
"grad_norm": 3.7711753845214844,
|
| 1963 |
+
"learning_rate": 8.146964856230033e-05,
|
| 1964 |
+
"loss": 0.22,
|
| 1965 |
+
"step": 2730
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 5.15,
|
| 1969 |
+
"grad_norm": 5.828824520111084,
|
| 1970 |
+
"learning_rate": 8.093716719914803e-05,
|
| 1971 |
+
"loss": 0.3559,
|
| 1972 |
+
"step": 2740
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 5.17,
|
| 1976 |
+
"grad_norm": 2.093899726867676,
|
| 1977 |
+
"learning_rate": 8.040468583599574e-05,
|
| 1978 |
+
"loss": 0.2061,
|
| 1979 |
+
"step": 2750
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 5.19,
|
| 1983 |
+
"grad_norm": 4.466729164123535,
|
| 1984 |
+
"learning_rate": 7.987220447284345e-05,
|
| 1985 |
+
"loss": 0.3012,
|
| 1986 |
+
"step": 2760
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 5.21,
|
| 1990 |
+
"grad_norm": 2.4526288509368896,
|
| 1991 |
+
"learning_rate": 7.933972310969116e-05,
|
| 1992 |
+
"loss": 0.2209,
|
| 1993 |
+
"step": 2770
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 5.23,
|
| 1997 |
+
"grad_norm": 4.241008281707764,
|
| 1998 |
+
"learning_rate": 7.880724174653888e-05,
|
| 1999 |
+
"loss": 0.2233,
|
| 2000 |
+
"step": 2780
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 5.24,
|
| 2004 |
+
"grad_norm": 2.828065872192383,
|
| 2005 |
+
"learning_rate": 7.827476038338658e-05,
|
| 2006 |
+
"loss": 0.1914,
|
| 2007 |
+
"step": 2790
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 5.26,
|
| 2011 |
+
"grad_norm": 4.489712715148926,
|
| 2012 |
+
"learning_rate": 7.77422790202343e-05,
|
| 2013 |
+
"loss": 0.2209,
|
| 2014 |
+
"step": 2800
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 5.28,
|
| 2018 |
+
"grad_norm": 7.091925621032715,
|
| 2019 |
+
"learning_rate": 7.7209797657082e-05,
|
| 2020 |
+
"loss": 0.2372,
|
| 2021 |
+
"step": 2810
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 5.3,
|
| 2025 |
+
"grad_norm": 3.1179749965667725,
|
| 2026 |
+
"learning_rate": 7.667731629392971e-05,
|
| 2027 |
+
"loss": 0.3655,
|
| 2028 |
+
"step": 2820
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 5.32,
|
| 2032 |
+
"grad_norm": 2.333217144012451,
|
| 2033 |
+
"learning_rate": 7.614483493077744e-05,
|
| 2034 |
+
"loss": 0.2942,
|
| 2035 |
+
"step": 2830
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 5.34,
|
| 2039 |
+
"grad_norm": 3.904360771179199,
|
| 2040 |
+
"learning_rate": 7.561235356762514e-05,
|
| 2041 |
+
"loss": 0.2113,
|
| 2042 |
+
"step": 2840
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 5.36,
|
| 2046 |
+
"grad_norm": 3.5696709156036377,
|
| 2047 |
+
"learning_rate": 7.507987220447285e-05,
|
| 2048 |
+
"loss": 0.1951,
|
| 2049 |
+
"step": 2850
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 5.38,
|
| 2053 |
+
"grad_norm": 2.9150383472442627,
|
| 2054 |
+
"learning_rate": 7.454739084132055e-05,
|
| 2055 |
+
"loss": 0.2142,
|
| 2056 |
+
"step": 2860
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 5.39,
|
| 2060 |
+
"grad_norm": 3.8467230796813965,
|
| 2061 |
+
"learning_rate": 7.401490947816827e-05,
|
| 2062 |
+
"loss": 0.2315,
|
| 2063 |
+
"step": 2870
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 5.41,
|
| 2067 |
+
"grad_norm": 2.5664279460906982,
|
| 2068 |
+
"learning_rate": 7.348242811501597e-05,
|
| 2069 |
+
"loss": 0.3081,
|
| 2070 |
+
"step": 2880
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 5.43,
|
| 2074 |
+
"grad_norm": 2.859978199005127,
|
| 2075 |
+
"learning_rate": 7.294994675186368e-05,
|
| 2076 |
+
"loss": 0.2426,
|
| 2077 |
+
"step": 2890
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 5.45,
|
| 2081 |
+
"grad_norm": 2.69465970993042,
|
| 2082 |
+
"learning_rate": 7.241746538871141e-05,
|
| 2083 |
+
"loss": 0.3018,
|
| 2084 |
+
"step": 2900
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 5.47,
|
| 2088 |
+
"grad_norm": 3.34781551361084,
|
| 2089 |
+
"learning_rate": 7.188498402555911e-05,
|
| 2090 |
+
"loss": 0.1438,
|
| 2091 |
+
"step": 2910
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 5.49,
|
| 2095 |
+
"grad_norm": 4.8066182136535645,
|
| 2096 |
+
"learning_rate": 7.135250266240682e-05,
|
| 2097 |
+
"loss": 0.224,
|
| 2098 |
+
"step": 2920
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 5.51,
|
| 2102 |
+
"grad_norm": 0.3521786630153656,
|
| 2103 |
+
"learning_rate": 7.082002129925452e-05,
|
| 2104 |
+
"loss": 0.2222,
|
| 2105 |
+
"step": 2930
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 5.53,
|
| 2109 |
+
"grad_norm": 2.7981643676757812,
|
| 2110 |
+
"learning_rate": 7.028753993610224e-05,
|
| 2111 |
+
"loss": 0.3419,
|
| 2112 |
+
"step": 2940
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 5.55,
|
| 2116 |
+
"grad_norm": 4.798647403717041,
|
| 2117 |
+
"learning_rate": 6.975505857294996e-05,
|
| 2118 |
+
"loss": 0.2757,
|
| 2119 |
+
"step": 2950
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 5.56,
|
| 2123 |
+
"grad_norm": 4.175738334655762,
|
| 2124 |
+
"learning_rate": 6.922257720979765e-05,
|
| 2125 |
+
"loss": 0.2388,
|
| 2126 |
+
"step": 2960
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 5.58,
|
| 2130 |
+
"grad_norm": 3.5461745262145996,
|
| 2131 |
+
"learning_rate": 6.869009584664538e-05,
|
| 2132 |
+
"loss": 0.1961,
|
| 2133 |
+
"step": 2970
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 5.6,
|
| 2137 |
+
"grad_norm": 2.714567184448242,
|
| 2138 |
+
"learning_rate": 6.815761448349308e-05,
|
| 2139 |
+
"loss": 0.2225,
|
| 2140 |
+
"step": 2980
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 5.62,
|
| 2144 |
+
"grad_norm": 4.761994361877441,
|
| 2145 |
+
"learning_rate": 6.762513312034079e-05,
|
| 2146 |
+
"loss": 0.23,
|
| 2147 |
+
"step": 2990
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 5.64,
|
| 2151 |
+
"grad_norm": 2.0303542613983154,
|
| 2152 |
+
"learning_rate": 6.70926517571885e-05,
|
| 2153 |
+
"loss": 0.1996,
|
| 2154 |
+
"step": 3000
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 5.66,
|
| 2158 |
+
"grad_norm": 3.595022678375244,
|
| 2159 |
+
"learning_rate": 6.656017039403621e-05,
|
| 2160 |
+
"loss": 0.2089,
|
| 2161 |
+
"step": 3010
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 5.68,
|
| 2165 |
+
"grad_norm": 4.19135856628418,
|
| 2166 |
+
"learning_rate": 6.602768903088393e-05,
|
| 2167 |
+
"loss": 0.2199,
|
| 2168 |
+
"step": 3020
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 5.7,
|
| 2172 |
+
"grad_norm": 6.599343299865723,
|
| 2173 |
+
"learning_rate": 6.549520766773163e-05,
|
| 2174 |
+
"loss": 0.2658,
|
| 2175 |
+
"step": 3030
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 5.71,
|
| 2179 |
+
"grad_norm": 4.5252685546875,
|
| 2180 |
+
"learning_rate": 6.496272630457935e-05,
|
| 2181 |
+
"loss": 0.1459,
|
| 2182 |
+
"step": 3040
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 5.73,
|
| 2186 |
+
"grad_norm": 2.3499250411987305,
|
| 2187 |
+
"learning_rate": 6.443024494142705e-05,
|
| 2188 |
+
"loss": 0.2447,
|
| 2189 |
+
"step": 3050
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 5.75,
|
| 2193 |
+
"grad_norm": 4.795304298400879,
|
| 2194 |
+
"learning_rate": 6.389776357827476e-05,
|
| 2195 |
+
"loss": 0.1987,
|
| 2196 |
+
"step": 3060
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 5.77,
|
| 2200 |
+
"grad_norm": 2.328916311264038,
|
| 2201 |
+
"learning_rate": 6.336528221512247e-05,
|
| 2202 |
+
"loss": 0.2369,
|
| 2203 |
+
"step": 3070
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 5.79,
|
| 2207 |
+
"grad_norm": 5.857803821563721,
|
| 2208 |
+
"learning_rate": 6.283280085197019e-05,
|
| 2209 |
+
"loss": 0.2668,
|
| 2210 |
+
"step": 3080
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 5.81,
|
| 2214 |
+
"grad_norm": 6.459561347961426,
|
| 2215 |
+
"learning_rate": 6.23003194888179e-05,
|
| 2216 |
+
"loss": 0.2438,
|
| 2217 |
+
"step": 3090
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 5.83,
|
| 2221 |
+
"grad_norm": 3.3628289699554443,
|
| 2222 |
+
"learning_rate": 6.17678381256656e-05,
|
| 2223 |
+
"loss": 0.2694,
|
| 2224 |
+
"step": 3100
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 5.85,
|
| 2228 |
+
"grad_norm": 3.036752223968506,
|
| 2229 |
+
"learning_rate": 6.123535676251332e-05,
|
| 2230 |
+
"loss": 0.3594,
|
| 2231 |
+
"step": 3110
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 5.86,
|
| 2235 |
+
"grad_norm": 1.0607565641403198,
|
| 2236 |
+
"learning_rate": 6.070287539936103e-05,
|
| 2237 |
+
"loss": 0.1804,
|
| 2238 |
+
"step": 3120
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 5.88,
|
| 2242 |
+
"grad_norm": 3.0828421115875244,
|
| 2243 |
+
"learning_rate": 6.017039403620873e-05,
|
| 2244 |
+
"loss": 0.2408,
|
| 2245 |
+
"step": 3130
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 5.9,
|
| 2249 |
+
"grad_norm": 6.958425045013428,
|
| 2250 |
+
"learning_rate": 5.963791267305645e-05,
|
| 2251 |
+
"loss": 0.2278,
|
| 2252 |
+
"step": 3140
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 5.92,
|
| 2256 |
+
"grad_norm": 6.2445526123046875,
|
| 2257 |
+
"learning_rate": 5.910543130990416e-05,
|
| 2258 |
+
"loss": 0.1687,
|
| 2259 |
+
"step": 3150
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 5.94,
|
| 2263 |
+
"grad_norm": 4.447784900665283,
|
| 2264 |
+
"learning_rate": 5.857294994675186e-05,
|
| 2265 |
+
"loss": 0.2019,
|
| 2266 |
+
"step": 3160
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 5.96,
|
| 2270 |
+
"grad_norm": 2.6770880222320557,
|
| 2271 |
+
"learning_rate": 5.8040468583599575e-05,
|
| 2272 |
+
"loss": 0.2857,
|
| 2273 |
+
"step": 3170
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 5.98,
|
| 2277 |
+
"grad_norm": 2.762964963912964,
|
| 2278 |
+
"learning_rate": 5.750798722044729e-05,
|
| 2279 |
+
"loss": 0.2092,
|
| 2280 |
+
"step": 3180
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 6.0,
|
| 2284 |
+
"grad_norm": 4.038713455200195,
|
| 2285 |
+
"learning_rate": 5.6975505857295e-05,
|
| 2286 |
+
"loss": 0.2869,
|
| 2287 |
+
"step": 3190
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 6.0,
|
| 2291 |
+
"eval_accuracy": 0.542,
|
| 2292 |
+
"eval_loss": 1.6185880899429321,
|
| 2293 |
+
"eval_runtime": 5.2709,
|
| 2294 |
+
"eval_samples_per_second": 94.861,
|
| 2295 |
+
"eval_steps_per_second": 3.036,
|
| 2296 |
+
"step": 3192
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 6.02,
|
| 2300 |
+
"grad_norm": 4.724943161010742,
|
| 2301 |
+
"learning_rate": 5.6443024494142705e-05,
|
| 2302 |
+
"loss": 0.1294,
|
| 2303 |
+
"step": 3200
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 6.03,
|
| 2307 |
+
"grad_norm": 0.55290687084198,
|
| 2308 |
+
"learning_rate": 5.5910543130990424e-05,
|
| 2309 |
+
"loss": 0.0765,
|
| 2310 |
+
"step": 3210
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 6.05,
|
| 2314 |
+
"grad_norm": 2.6764791011810303,
|
| 2315 |
+
"learning_rate": 5.537806176783813e-05,
|
| 2316 |
+
"loss": 0.0919,
|
| 2317 |
+
"step": 3220
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 6.07,
|
| 2321 |
+
"grad_norm": 0.35876473784446716,
|
| 2322 |
+
"learning_rate": 5.4845580404685834e-05,
|
| 2323 |
+
"loss": 0.1147,
|
| 2324 |
+
"step": 3230
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 6.09,
|
| 2328 |
+
"grad_norm": 7.607916355133057,
|
| 2329 |
+
"learning_rate": 5.4313099041533546e-05,
|
| 2330 |
+
"loss": 0.1175,
|
| 2331 |
+
"step": 3240
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 6.11,
|
| 2335 |
+
"grad_norm": 0.5861133337020874,
|
| 2336 |
+
"learning_rate": 5.3780617678381265e-05,
|
| 2337 |
+
"loss": 0.041,
|
| 2338 |
+
"step": 3250
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 6.13,
|
| 2342 |
+
"grad_norm": 7.388729095458984,
|
| 2343 |
+
"learning_rate": 5.324813631522897e-05,
|
| 2344 |
+
"loss": 0.0932,
|
| 2345 |
+
"step": 3260
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 6.15,
|
| 2349 |
+
"grad_norm": 0.1238517016172409,
|
| 2350 |
+
"learning_rate": 5.2715654952076676e-05,
|
| 2351 |
+
"loss": 0.1329,
|
| 2352 |
+
"step": 3270
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 6.17,
|
| 2356 |
+
"grad_norm": 2.567272424697876,
|
| 2357 |
+
"learning_rate": 5.2183173588924395e-05,
|
| 2358 |
+
"loss": 0.0433,
|
| 2359 |
+
"step": 3280
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 6.18,
|
| 2363 |
+
"grad_norm": 2.4135849475860596,
|
| 2364 |
+
"learning_rate": 5.16506922257721e-05,
|
| 2365 |
+
"loss": 0.068,
|
| 2366 |
+
"step": 3290
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 6.2,
|
| 2370 |
+
"grad_norm": 0.1513521671295166,
|
| 2371 |
+
"learning_rate": 5.1118210862619806e-05,
|
| 2372 |
+
"loss": 0.0358,
|
| 2373 |
+
"step": 3300
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 6.22,
|
| 2377 |
+
"grad_norm": 0.09611086547374725,
|
| 2378 |
+
"learning_rate": 5.058572949946752e-05,
|
| 2379 |
+
"loss": 0.1276,
|
| 2380 |
+
"step": 3310
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 6.24,
|
| 2384 |
+
"grad_norm": 0.34191715717315674,
|
| 2385 |
+
"learning_rate": 5.0053248136315237e-05,
|
| 2386 |
+
"loss": 0.0796,
|
| 2387 |
+
"step": 3320
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 6.26,
|
| 2391 |
+
"grad_norm": 0.03613116592168808,
|
| 2392 |
+
"learning_rate": 4.952076677316294e-05,
|
| 2393 |
+
"loss": 0.0595,
|
| 2394 |
+
"step": 3330
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 6.28,
|
| 2398 |
+
"grad_norm": 9.316259384155273,
|
| 2399 |
+
"learning_rate": 4.8988285410010654e-05,
|
| 2400 |
+
"loss": 0.0588,
|
| 2401 |
+
"step": 3340
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 6.3,
|
| 2405 |
+
"grad_norm": 9.937971115112305,
|
| 2406 |
+
"learning_rate": 4.845580404685836e-05,
|
| 2407 |
+
"loss": 0.1649,
|
| 2408 |
+
"step": 3350
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 6.32,
|
| 2412 |
+
"grad_norm": 8.708675384521484,
|
| 2413 |
+
"learning_rate": 4.792332268370607e-05,
|
| 2414 |
+
"loss": 0.2051,
|
| 2415 |
+
"step": 3360
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 6.33,
|
| 2419 |
+
"grad_norm": 1.0412534475326538,
|
| 2420 |
+
"learning_rate": 4.7390841320553784e-05,
|
| 2421 |
+
"loss": 0.117,
|
| 2422 |
+
"step": 3370
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 6.35,
|
| 2426 |
+
"grad_norm": 0.2932884395122528,
|
| 2427 |
+
"learning_rate": 4.6858359957401496e-05,
|
| 2428 |
+
"loss": 0.1319,
|
| 2429 |
+
"step": 3380
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 6.37,
|
| 2433 |
+
"grad_norm": 9.108260154724121,
|
| 2434 |
+
"learning_rate": 4.632587859424921e-05,
|
| 2435 |
+
"loss": 0.0514,
|
| 2436 |
+
"step": 3390
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 6.39,
|
| 2440 |
+
"grad_norm": 5.959402084350586,
|
| 2441 |
+
"learning_rate": 4.579339723109691e-05,
|
| 2442 |
+
"loss": 0.0938,
|
| 2443 |
+
"step": 3400
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 6.41,
|
| 2447 |
+
"grad_norm": 2.6586625576019287,
|
| 2448 |
+
"learning_rate": 4.5260915867944625e-05,
|
| 2449 |
+
"loss": 0.094,
|
| 2450 |
+
"step": 3410
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 6.43,
|
| 2454 |
+
"grad_norm": 0.6427993178367615,
|
| 2455 |
+
"learning_rate": 4.472843450479233e-05,
|
| 2456 |
+
"loss": 0.0934,
|
| 2457 |
+
"step": 3420
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 6.45,
|
| 2461 |
+
"grad_norm": 4.278239727020264,
|
| 2462 |
+
"learning_rate": 4.419595314164004e-05,
|
| 2463 |
+
"loss": 0.1062,
|
| 2464 |
+
"step": 3430
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 6.47,
|
| 2468 |
+
"grad_norm": 12.555079460144043,
|
| 2469 |
+
"learning_rate": 4.3663471778487755e-05,
|
| 2470 |
+
"loss": 0.0382,
|
| 2471 |
+
"step": 3440
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 6.48,
|
| 2475 |
+
"grad_norm": 7.286933422088623,
|
| 2476 |
+
"learning_rate": 4.313099041533547e-05,
|
| 2477 |
+
"loss": 0.07,
|
| 2478 |
+
"step": 3450
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 6.5,
|
| 2482 |
+
"grad_norm": 0.022136323153972626,
|
| 2483 |
+
"learning_rate": 4.259850905218318e-05,
|
| 2484 |
+
"loss": 0.0973,
|
| 2485 |
+
"step": 3460
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 6.52,
|
| 2489 |
+
"grad_norm": 13.952143669128418,
|
| 2490 |
+
"learning_rate": 4.2066027689030885e-05,
|
| 2491 |
+
"loss": 0.0956,
|
| 2492 |
+
"step": 3470
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 6.54,
|
| 2496 |
+
"grad_norm": 5.638768672943115,
|
| 2497 |
+
"learning_rate": 4.15335463258786e-05,
|
| 2498 |
+
"loss": 0.0594,
|
| 2499 |
+
"step": 3480
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 6.56,
|
| 2503 |
+
"grad_norm": 0.06895623356103897,
|
| 2504 |
+
"learning_rate": 4.10010649627263e-05,
|
| 2505 |
+
"loss": 0.1302,
|
| 2506 |
+
"step": 3490
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 6.58,
|
| 2510 |
+
"grad_norm": 0.5818095207214355,
|
| 2511 |
+
"learning_rate": 4.0468583599574014e-05,
|
| 2512 |
+
"loss": 0.1488,
|
| 2513 |
+
"step": 3500
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 6.6,
|
| 2517 |
+
"grad_norm": 0.13191522657871246,
|
| 2518 |
+
"learning_rate": 3.9989350372736954e-05,
|
| 2519 |
+
"loss": 0.0878,
|
| 2520 |
+
"step": 3510
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 6.62,
|
| 2524 |
+
"grad_norm": 18.15912628173828,
|
| 2525 |
+
"learning_rate": 3.9456869009584666e-05,
|
| 2526 |
+
"loss": 0.1311,
|
| 2527 |
+
"step": 3520
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 6.64,
|
| 2531 |
+
"grad_norm": 13.288291931152344,
|
| 2532 |
+
"learning_rate": 3.892438764643238e-05,
|
| 2533 |
+
"loss": 0.2676,
|
| 2534 |
+
"step": 3530
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 6.65,
|
| 2538 |
+
"grad_norm": 2.8228518962860107,
|
| 2539 |
+
"learning_rate": 3.839190628328009e-05,
|
| 2540 |
+
"loss": 0.0771,
|
| 2541 |
+
"step": 3540
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 6.67,
|
| 2545 |
+
"grad_norm": 0.11967118829488754,
|
| 2546 |
+
"learning_rate": 3.7859424920127795e-05,
|
| 2547 |
+
"loss": 0.047,
|
| 2548 |
+
"step": 3550
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 6.69,
|
| 2552 |
+
"grad_norm": 0.17968201637268066,
|
| 2553 |
+
"learning_rate": 3.732694355697551e-05,
|
| 2554 |
+
"loss": 0.0551,
|
| 2555 |
+
"step": 3560
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 6.71,
|
| 2559 |
+
"grad_norm": 3.643249273300171,
|
| 2560 |
+
"learning_rate": 3.679446219382322e-05,
|
| 2561 |
+
"loss": 0.0916,
|
| 2562 |
+
"step": 3570
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 6.73,
|
| 2566 |
+
"grad_norm": 1.9196984767913818,
|
| 2567 |
+
"learning_rate": 3.6261980830670925e-05,
|
| 2568 |
+
"loss": 0.0497,
|
| 2569 |
+
"step": 3580
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 6.75,
|
| 2573 |
+
"grad_norm": 14.608271598815918,
|
| 2574 |
+
"learning_rate": 3.572949946751864e-05,
|
| 2575 |
+
"loss": 0.2575,
|
| 2576 |
+
"step": 3590
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 6.77,
|
| 2580 |
+
"grad_norm": 0.6130942106246948,
|
| 2581 |
+
"learning_rate": 3.519701810436635e-05,
|
| 2582 |
+
"loss": 0.1304,
|
| 2583 |
+
"step": 3600
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 6.79,
|
| 2587 |
+
"grad_norm": 0.11332878470420837,
|
| 2588 |
+
"learning_rate": 3.466453674121406e-05,
|
| 2589 |
+
"loss": 0.0753,
|
| 2590 |
+
"step": 3610
|
| 2591 |
+
},
|
| 2592 |
+
{
|
| 2593 |
+
"epoch": 6.8,
|
| 2594 |
+
"grad_norm": 0.6603028178215027,
|
| 2595 |
+
"learning_rate": 3.413205537806177e-05,
|
| 2596 |
+
"loss": 0.0852,
|
| 2597 |
+
"step": 3620
|
| 2598 |
+
},
|
| 2599 |
+
{
|
| 2600 |
+
"epoch": 6.82,
|
| 2601 |
+
"grad_norm": 4.780795097351074,
|
| 2602 |
+
"learning_rate": 3.359957401490948e-05,
|
| 2603 |
+
"loss": 0.0702,
|
| 2604 |
+
"step": 3630
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 6.84,
|
| 2608 |
+
"grad_norm": 2.858586549758911,
|
| 2609 |
+
"learning_rate": 3.306709265175719e-05,
|
| 2610 |
+
"loss": 0.1232,
|
| 2611 |
+
"step": 3640
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 6.86,
|
| 2615 |
+
"grad_norm": 4.370979309082031,
|
| 2616 |
+
"learning_rate": 3.2534611288604896e-05,
|
| 2617 |
+
"loss": 0.1168,
|
| 2618 |
+
"step": 3650
|
| 2619 |
+
},
|
| 2620 |
+
{
|
| 2621 |
+
"epoch": 6.88,
|
| 2622 |
+
"grad_norm": 0.04837312549352646,
|
| 2623 |
+
"learning_rate": 3.2002129925452615e-05,
|
| 2624 |
+
"loss": 0.0296,
|
| 2625 |
+
"step": 3660
|
| 2626 |
+
},
|
| 2627 |
+
{
|
| 2628 |
+
"epoch": 6.9,
|
| 2629 |
+
"grad_norm": 0.31512418389320374,
|
| 2630 |
+
"learning_rate": 3.146964856230032e-05,
|
| 2631 |
+
"loss": 0.1051,
|
| 2632 |
+
"step": 3670
|
| 2633 |
+
},
|
| 2634 |
+
{
|
| 2635 |
+
"epoch": 6.92,
|
| 2636 |
+
"grad_norm": 3.6252450942993164,
|
| 2637 |
+
"learning_rate": 3.093716719914803e-05,
|
| 2638 |
+
"loss": 0.1216,
|
| 2639 |
+
"step": 3680
|
| 2640 |
+
},
|
| 2641 |
+
{
|
| 2642 |
+
"epoch": 6.94,
|
| 2643 |
+
"grad_norm": 0.5809699892997742,
|
| 2644 |
+
"learning_rate": 3.0404685835995745e-05,
|
| 2645 |
+
"loss": 0.0501,
|
| 2646 |
+
"step": 3690
|
| 2647 |
+
},
|
| 2648 |
+
{
|
| 2649 |
+
"epoch": 6.95,
|
| 2650 |
+
"grad_norm": 0.15533213317394257,
|
| 2651 |
+
"learning_rate": 2.987220447284345e-05,
|
| 2652 |
+
"loss": 0.0621,
|
| 2653 |
+
"step": 3700
|
| 2654 |
+
},
|
| 2655 |
+
{
|
| 2656 |
+
"epoch": 6.97,
|
| 2657 |
+
"grad_norm": 0.04270392283797264,
|
| 2658 |
+
"learning_rate": 2.9339723109691166e-05,
|
| 2659 |
+
"loss": 0.0353,
|
| 2660 |
+
"step": 3710
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 6.99,
|
| 2664 |
+
"grad_norm": 0.01716708578169346,
|
| 2665 |
+
"learning_rate": 2.880724174653887e-05,
|
| 2666 |
+
"loss": 0.0337,
|
| 2667 |
+
"step": 3720
|
| 2668 |
+
},
|
| 2669 |
+
{
|
| 2670 |
+
"epoch": 7.0,
|
| 2671 |
+
"eval_accuracy": 0.538,
|
| 2672 |
+
"eval_loss": 2.5670464038848877,
|
| 2673 |
+
"eval_runtime": 5.0132,
|
| 2674 |
+
"eval_samples_per_second": 99.737,
|
| 2675 |
+
"eval_steps_per_second": 3.192,
|
| 2676 |
+
"step": 3724
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 7.01,
|
| 2680 |
+
"grad_norm": 0.029269874095916748,
|
| 2681 |
+
"learning_rate": 2.8274760383386583e-05,
|
| 2682 |
+
"loss": 0.0997,
|
| 2683 |
+
"step": 3730
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 7.03,
|
| 2687 |
+
"grad_norm": 0.11788934469223022,
|
| 2688 |
+
"learning_rate": 2.7742279020234292e-05,
|
| 2689 |
+
"loss": 0.0082,
|
| 2690 |
+
"step": 3740
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 7.05,
|
| 2694 |
+
"grad_norm": 0.01990850828588009,
|
| 2695 |
+
"learning_rate": 2.7209797657082004e-05,
|
| 2696 |
+
"loss": 0.0251,
|
| 2697 |
+
"step": 3750
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 7.07,
|
| 2701 |
+
"grad_norm": 0.38145899772644043,
|
| 2702 |
+
"learning_rate": 2.6677316293929716e-05,
|
| 2703 |
+
"loss": 0.013,
|
| 2704 |
+
"step": 3760
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 7.09,
|
| 2708 |
+
"grad_norm": 2.5921037197113037,
|
| 2709 |
+
"learning_rate": 2.6144834930777425e-05,
|
| 2710 |
+
"loss": 0.0601,
|
| 2711 |
+
"step": 3770
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 7.11,
|
| 2715 |
+
"grad_norm": 0.03015085868537426,
|
| 2716 |
+
"learning_rate": 2.5612353567625137e-05,
|
| 2717 |
+
"loss": 0.0365,
|
| 2718 |
+
"step": 3780
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 7.12,
|
| 2722 |
+
"grad_norm": 9.706396102905273,
|
| 2723 |
+
"learning_rate": 2.5079872204472842e-05,
|
| 2724 |
+
"loss": 0.0722,
|
| 2725 |
+
"step": 3790
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 7.14,
|
| 2729 |
+
"grad_norm": 0.02435947395861149,
|
| 2730 |
+
"learning_rate": 2.4547390841320554e-05,
|
| 2731 |
+
"loss": 0.0082,
|
| 2732 |
+
"step": 3800
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 7.16,
|
| 2736 |
+
"grad_norm": 0.030334001407027245,
|
| 2737 |
+
"learning_rate": 2.4014909478168267e-05,
|
| 2738 |
+
"loss": 0.011,
|
| 2739 |
+
"step": 3810
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 7.18,
|
| 2743 |
+
"grad_norm": 3.0113189220428467,
|
| 2744 |
+
"learning_rate": 2.3482428115015975e-05,
|
| 2745 |
+
"loss": 0.0485,
|
| 2746 |
+
"step": 3820
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 7.2,
|
| 2750 |
+
"grad_norm": 0.021948257461190224,
|
| 2751 |
+
"learning_rate": 2.2949946751863684e-05,
|
| 2752 |
+
"loss": 0.0384,
|
| 2753 |
+
"step": 3830
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 7.22,
|
| 2757 |
+
"grad_norm": 0.02243644930422306,
|
| 2758 |
+
"learning_rate": 2.2417465388711396e-05,
|
| 2759 |
+
"loss": 0.0103,
|
| 2760 |
+
"step": 3840
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 7.24,
|
| 2764 |
+
"grad_norm": 0.03432391956448555,
|
| 2765 |
+
"learning_rate": 2.188498402555911e-05,
|
| 2766 |
+
"loss": 0.0436,
|
| 2767 |
+
"step": 3850
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 7.26,
|
| 2771 |
+
"grad_norm": 0.05325201898813248,
|
| 2772 |
+
"learning_rate": 2.1352502662406817e-05,
|
| 2773 |
+
"loss": 0.0519,
|
| 2774 |
+
"step": 3860
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 7.27,
|
| 2778 |
+
"grad_norm": 0.01565093919634819,
|
| 2779 |
+
"learning_rate": 2.0820021299254526e-05,
|
| 2780 |
+
"loss": 0.0225,
|
| 2781 |
+
"step": 3870
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 7.29,
|
| 2785 |
+
"grad_norm": 0.14941458404064178,
|
| 2786 |
+
"learning_rate": 2.0287539936102238e-05,
|
| 2787 |
+
"loss": 0.0014,
|
| 2788 |
+
"step": 3880
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 7.31,
|
| 2792 |
+
"grad_norm": 0.012375161051750183,
|
| 2793 |
+
"learning_rate": 1.9755058572949947e-05,
|
| 2794 |
+
"loss": 0.0155,
|
| 2795 |
+
"step": 3890
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 7.33,
|
| 2799 |
+
"grad_norm": 0.004966201260685921,
|
| 2800 |
+
"learning_rate": 1.9222577209797655e-05,
|
| 2801 |
+
"loss": 0.0131,
|
| 2802 |
+
"step": 3900
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 7.35,
|
| 2806 |
+
"grad_norm": 0.013580994680523872,
|
| 2807 |
+
"learning_rate": 1.869009584664537e-05,
|
| 2808 |
+
"loss": 0.0054,
|
| 2809 |
+
"step": 3910
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 7.37,
|
| 2813 |
+
"grad_norm": 0.006246180739253759,
|
| 2814 |
+
"learning_rate": 1.815761448349308e-05,
|
| 2815 |
+
"loss": 0.0953,
|
| 2816 |
+
"step": 3920
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 7.39,
|
| 2820 |
+
"grad_norm": 0.05573183670639992,
|
| 2821 |
+
"learning_rate": 1.762513312034079e-05,
|
| 2822 |
+
"loss": 0.0454,
|
| 2823 |
+
"step": 3930
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 7.41,
|
| 2827 |
+
"grad_norm": 0.023609144613146782,
|
| 2828 |
+
"learning_rate": 1.70926517571885e-05,
|
| 2829 |
+
"loss": 0.0313,
|
| 2830 |
+
"step": 3940
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 7.42,
|
| 2834 |
+
"grad_norm": 0.01226514670997858,
|
| 2835 |
+
"learning_rate": 1.656017039403621e-05,
|
| 2836 |
+
"loss": 0.0448,
|
| 2837 |
+
"step": 3950
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 7.44,
|
| 2841 |
+
"grad_norm": 20.82164764404297,
|
| 2842 |
+
"learning_rate": 1.6027689030883918e-05,
|
| 2843 |
+
"loss": 0.1159,
|
| 2844 |
+
"step": 3960
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 7.46,
|
| 2848 |
+
"grad_norm": 0.018167337402701378,
|
| 2849 |
+
"learning_rate": 1.549520766773163e-05,
|
| 2850 |
+
"loss": 0.0377,
|
| 2851 |
+
"step": 3970
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 7.48,
|
| 2855 |
+
"grad_norm": 0.031441397964954376,
|
| 2856 |
+
"learning_rate": 1.4962726304579342e-05,
|
| 2857 |
+
"loss": 0.0299,
|
| 2858 |
+
"step": 3980
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 7.5,
|
| 2862 |
+
"grad_norm": 0.2981891632080078,
|
| 2863 |
+
"learning_rate": 1.4430244941427051e-05,
|
| 2864 |
+
"loss": 0.0301,
|
| 2865 |
+
"step": 3990
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 7.52,
|
| 2869 |
+
"grad_norm": 0.052271194756031036,
|
| 2870 |
+
"learning_rate": 1.3897763578274761e-05,
|
| 2871 |
+
"loss": 0.0014,
|
| 2872 |
+
"step": 4000
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 7.54,
|
| 2876 |
+
"grad_norm": 0.07910721004009247,
|
| 2877 |
+
"learning_rate": 1.3365282215122472e-05,
|
| 2878 |
+
"loss": 0.0029,
|
| 2879 |
+
"step": 4010
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 7.56,
|
| 2883 |
+
"grad_norm": 1.7630151510238647,
|
| 2884 |
+
"learning_rate": 1.283280085197018e-05,
|
| 2885 |
+
"loss": 0.0421,
|
| 2886 |
+
"step": 4020
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 7.58,
|
| 2890 |
+
"grad_norm": 0.013812140561640263,
|
| 2891 |
+
"learning_rate": 1.2300319488817893e-05,
|
| 2892 |
+
"loss": 0.0221,
|
| 2893 |
+
"step": 4030
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 7.59,
|
| 2897 |
+
"grad_norm": 0.013484804891049862,
|
| 2898 |
+
"learning_rate": 1.1767838125665601e-05,
|
| 2899 |
+
"loss": 0.0011,
|
| 2900 |
+
"step": 4040
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 7.61,
|
| 2904 |
+
"grad_norm": 0.031141534447669983,
|
| 2905 |
+
"learning_rate": 1.1235356762513314e-05,
|
| 2906 |
+
"loss": 0.0963,
|
| 2907 |
+
"step": 4050
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 7.63,
|
| 2911 |
+
"grad_norm": 0.1125926822423935,
|
| 2912 |
+
"learning_rate": 1.0702875399361024e-05,
|
| 2913 |
+
"loss": 0.0206,
|
| 2914 |
+
"step": 4060
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 7.65,
|
| 2918 |
+
"grad_norm": 1.0204988718032837,
|
| 2919 |
+
"learning_rate": 1.0170394036208733e-05,
|
| 2920 |
+
"loss": 0.0019,
|
| 2921 |
+
"step": 4070
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 7.67,
|
| 2925 |
+
"grad_norm": 0.020539429038763046,
|
| 2926 |
+
"learning_rate": 9.637912673056443e-06,
|
| 2927 |
+
"loss": 0.0474,
|
| 2928 |
+
"step": 4080
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 7.69,
|
| 2932 |
+
"grad_norm": 0.02552681416273117,
|
| 2933 |
+
"learning_rate": 9.105431309904154e-06,
|
| 2934 |
+
"loss": 0.0263,
|
| 2935 |
+
"step": 4090
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 7.71,
|
| 2939 |
+
"grad_norm": 0.0711381658911705,
|
| 2940 |
+
"learning_rate": 8.572949946751864e-06,
|
| 2941 |
+
"loss": 0.0225,
|
| 2942 |
+
"step": 4100
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 7.73,
|
| 2946 |
+
"grad_norm": 0.031245287507772446,
|
| 2947 |
+
"learning_rate": 8.040468583599574e-06,
|
| 2948 |
+
"loss": 0.0012,
|
| 2949 |
+
"step": 4110
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 7.74,
|
| 2953 |
+
"grad_norm": 0.014275019988417625,
|
| 2954 |
+
"learning_rate": 7.507987220447285e-06,
|
| 2955 |
+
"loss": 0.0434,
|
| 2956 |
+
"step": 4120
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 7.76,
|
| 2960 |
+
"grad_norm": 0.060072433203458786,
|
| 2961 |
+
"learning_rate": 6.975505857294995e-06,
|
| 2962 |
+
"loss": 0.0846,
|
| 2963 |
+
"step": 4130
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 7.78,
|
| 2967 |
+
"grad_norm": 2.3613195419311523,
|
| 2968 |
+
"learning_rate": 6.443024494142705e-06,
|
| 2969 |
+
"loss": 0.0444,
|
| 2970 |
+
"step": 4140
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 7.8,
|
| 2974 |
+
"grad_norm": 0.03658520057797432,
|
| 2975 |
+
"learning_rate": 5.910543130990415e-06,
|
| 2976 |
+
"loss": 0.0022,
|
| 2977 |
+
"step": 4150
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 7.82,
|
| 2981 |
+
"grad_norm": 0.14258913695812225,
|
| 2982 |
+
"learning_rate": 5.378061767838126e-06,
|
| 2983 |
+
"loss": 0.0537,
|
| 2984 |
+
"step": 4160
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 7.84,
|
| 2988 |
+
"grad_norm": 0.017876233905553818,
|
| 2989 |
+
"learning_rate": 4.845580404685836e-06,
|
| 2990 |
+
"loss": 0.0654,
|
| 2991 |
+
"step": 4170
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 7.86,
|
| 2995 |
+
"grad_norm": 0.039430875331163406,
|
| 2996 |
+
"learning_rate": 4.3130990415335465e-06,
|
| 2997 |
+
"loss": 0.0241,
|
| 2998 |
+
"step": 4180
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 7.88,
|
| 3002 |
+
"grad_norm": 0.07222764194011688,
|
| 3003 |
+
"learning_rate": 3.780617678381257e-06,
|
| 3004 |
+
"loss": 0.0018,
|
| 3005 |
+
"step": 4190
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 7.89,
|
| 3009 |
+
"grad_norm": 0.2034127116203308,
|
| 3010 |
+
"learning_rate": 3.2481363152289674e-06,
|
| 3011 |
+
"loss": 0.0322,
|
| 3012 |
+
"step": 4200
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 7.91,
|
| 3016 |
+
"grad_norm": 32.39055633544922,
|
| 3017 |
+
"learning_rate": 2.7156549520766774e-06,
|
| 3018 |
+
"loss": 0.0278,
|
| 3019 |
+
"step": 4210
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 7.93,
|
| 3023 |
+
"grad_norm": 0.02537735551595688,
|
| 3024 |
+
"learning_rate": 2.183173588924388e-06,
|
| 3025 |
+
"loss": 0.0016,
|
| 3026 |
+
"step": 4220
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 7.95,
|
| 3030 |
+
"grad_norm": 5.338490009307861,
|
| 3031 |
+
"learning_rate": 1.650692225772098e-06,
|
| 3032 |
+
"loss": 0.0305,
|
| 3033 |
+
"step": 4230
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 7.97,
|
| 3037 |
+
"grad_norm": 0.04093256965279579,
|
| 3038 |
+
"learning_rate": 1.1182108626198083e-06,
|
| 3039 |
+
"loss": 0.0439,
|
| 3040 |
+
"step": 4240
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 7.99,
|
| 3044 |
+
"grad_norm": 0.01651330478489399,
|
| 3045 |
+
"learning_rate": 5.857294994675187e-07,
|
| 3046 |
+
"loss": 0.0166,
|
| 3047 |
+
"step": 4250
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 8.0,
|
| 3051 |
+
"eval_accuracy": 0.538,
|
| 3052 |
+
"eval_loss": 2.806075096130371,
|
| 3053 |
+
"eval_runtime": 5.1896,
|
| 3054 |
+
"eval_samples_per_second": 96.347,
|
| 3055 |
+
"eval_steps_per_second": 3.083,
|
| 3056 |
+
"step": 4256
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 8.0,
|
| 3060 |
+
"step": 4256,
|
| 3061 |
+
"total_flos": 5.269455293792256e+18,
|
| 3062 |
+
"train_loss": 0.40350076897398923,
|
| 3063 |
+
"train_runtime": 1386.2467,
|
| 3064 |
+
"train_samples_per_second": 49.053,
|
| 3065 |
+
"train_steps_per_second": 3.07
|
| 3066 |
+
}
|
| 3067 |
+
],
|
| 3068 |
+
"logging_steps": 10,
|
| 3069 |
+
"max_steps": 4256,
|
| 3070 |
+
"num_input_tokens_seen": 0,
|
| 3071 |
+
"num_train_epochs": 8,
|
| 3072 |
+
"save_steps": 500,
|
| 3073 |
+
"total_flos": 5.269455293792256e+18,
|
| 3074 |
+
"train_batch_size": 16,
|
| 3075 |
+
"trial_name": null,
|
| 3076 |
+
"trial_params": null
|
| 3077 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b19ba6bc19f5ba61d85c29b1ba40f80a59d96fadc9cd9b34010a076d641e735
|
| 3 |
+
size 4920
|