Commit
•
2299fd2
1
Parent(s):
2a70f66
v3.5
Browse files- README.md +24 -11
- all_results.json +10 -10
- config.json +31 -12
- eval_results.json +6 -6
- preprocessor_config.json +10 -10
- pytorch_model.bin +2 -2
- train_results.json +5 -5
- trainer_state.json +512 -11
- training_args.bin +2 -2
README.md
CHANGED
|
@@ -23,47 +23,60 @@ model-index:
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
-
value: 0.
|
| 27 |
---
|
| 28 |
|
| 29 |
-
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 30 |
-
should probably proofread and complete it, then remove this comment. -->
|
| 31 |
|
| 32 |
# Cowboy Hat emoji 🤠 (Western)
|
| 33 |
|
| 34 |
-
This model is a fine-tuned version of [
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 0.
|
| 37 |
-
- Accuracy: 0.
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
| 41 |
When you want to know if an art is 🤠 or not 🤠.
|
| 42 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
## Intended uses & limitations
|
| 44 |
|
| 45 |
filter gelbooru data on 🤠 or not 🤠
|
| 46 |
|
| 47 |
## Training and evaluation data
|
| 48 |
|
| 49 |
-
Selected
|
| 50 |
|
| 51 |
## Training procedure
|
| 52 |
|
| 53 |
### Training hyperparameters
|
| 54 |
|
| 55 |
The following hyperparameters were used during training:
|
| 56 |
-
- learning_rate:
|
| 57 |
- train_batch_size: 8
|
| 58 |
- eval_batch_size: 8
|
| 59 |
-
- seed:
|
| 60 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 61 |
- lr_scheduler_type: linear
|
| 62 |
-
- num_epochs:
|
| 63 |
|
| 64 |
### Training results
|
| 65 |
|
| 66 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
### Framework versions
|
| 69 |
|
|
|
|
| 23 |
metrics:
|
| 24 |
- name: Accuracy
|
| 25 |
type: accuracy
|
| 26 |
+
value: 0.7777777777777778
|
| 27 |
---
|
| 28 |
|
|
|
|
|
|
|
| 29 |
|
| 30 |
# Cowboy Hat emoji 🤠 (Western)
|
| 31 |
|
| 32 |
+
This model is a fine-tuned version of [facebook/convnextv2-large-22k-384](https://huggingface.co/facebook/convnextv2-large-22k-384) on the imagefolder dataset.
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
+
- Loss: 0.4250
|
| 35 |
+
- Accuracy: 0.7778
|
| 36 |
|
| 37 |
## Model description
|
| 38 |
|
| 39 |
When you want to know if an art is 🤠 or not 🤠.
|
| 40 |
|
| 41 |
+
- Current iteration: v3.5 (Continuous Image Integration)
|
| 42 |
+
|
| 43 |
+
## Wait, why?
|
| 44 |
+
|
| 45 |
+
gelbooru contains a lot of images, however not all of them are in the same region as south eas asia. As such, to filter out such images we have created a classifier that in theory learns the differences between western (USA/Europe/etc.) and not western (Japan/China/SEA).
|
| 46 |
+
|
| 47 |
+
The definition of "Not Western" is limited to the the asian region (Japan, Korea, China, Taiwan, Thailand and the surroundign region). The author believes that the art is similar enough with the same "style" which he personally prefers over western art.
|
| 48 |
+
|
| 49 |
## Intended uses & limitations
|
| 50 |
|
| 51 |
filter gelbooru data on 🤠 or not 🤠
|
| 52 |
|
| 53 |
## Training and evaluation data
|
| 54 |
|
| 55 |
+
Selected 358 images of 🤠 and not 🤠.
|
| 56 |
|
| 57 |
## Training procedure
|
| 58 |
|
| 59 |
### Training hyperparameters
|
| 60 |
|
| 61 |
The following hyperparameters were used during training:
|
| 62 |
+
- learning_rate: 2e-05
|
| 63 |
- train_batch_size: 8
|
| 64 |
- eval_batch_size: 8
|
| 65 |
+
- seed: 802565
|
| 66 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 67 |
- lr_scheduler_type: linear
|
| 68 |
+
- num_epochs: 5.0
|
| 69 |
|
| 70 |
### Training results
|
| 71 |
|
| 72 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 73 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 74 |
+
| 0.7384 | 1.0 | 152 | 0.4268 | 0.7963 |
|
| 75 |
+
| 0.2888 | 2.0 | 304 | 0.4250 | 0.7778 |
|
| 76 |
+
| 0.2953 | 3.0 | 456 | 0.4250 | 0.7778 |
|
| 77 |
+
| 0.4914 | 4.0 | 608 | 0.4250 | 0.7778 |
|
| 78 |
+
| 0.4099 | 5.0 | 760 | 0.4250 | 0.7778 |
|
| 79 |
+
|
| 80 |
|
| 81 |
### Framework versions
|
| 82 |
|
all_results.json
CHANGED
|
@@ -1,12 +1,12 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"train_loss": 0.
|
| 9 |
-
"train_runtime":
|
| 10 |
-
"train_samples_per_second":
|
| 11 |
-
"train_steps_per_second": 1.
|
| 12 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.7777777777777778,
|
| 4 |
+
"eval_loss": 0.4249849319458008,
|
| 5 |
+
"eval_runtime": 5.6527,
|
| 6 |
+
"eval_samples_per_second": 9.553,
|
| 7 |
+
"eval_steps_per_second": 1.238,
|
| 8 |
+
"train_loss": 0.39741649894337905,
|
| 9 |
+
"train_runtime": 401.1562,
|
| 10 |
+
"train_samples_per_second": 3.789,
|
| 11 |
+
"train_steps_per_second": 1.895
|
| 12 |
}
|
config.json
CHANGED
|
@@ -1,33 +1,52 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"architectures": [
|
| 4 |
-
"
|
| 5 |
],
|
| 6 |
-
"
|
| 7 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
"finetuning_task": "image-classification",
|
| 9 |
"hidden_act": "gelu",
|
| 10 |
-
"
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
"id2label": {
|
| 13 |
"0": "not_western",
|
| 14 |
"1": "western"
|
| 15 |
},
|
| 16 |
"image_size": 224,
|
| 17 |
"initializer_range": 0.02,
|
| 18 |
-
"intermediate_size": 3072,
|
| 19 |
"label2id": {
|
| 20 |
"not_western": "0",
|
| 21 |
"western": "1"
|
| 22 |
},
|
| 23 |
"layer_norm_eps": 1e-12,
|
| 24 |
-
"model_type": "
|
| 25 |
-
"num_attention_heads": 12,
|
| 26 |
"num_channels": 3,
|
| 27 |
-
"
|
| 28 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
"problem_type": "single_label_classification",
|
| 30 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
"torch_dtype": "float32",
|
| 32 |
"transformers_version": "4.30.0.dev0"
|
| 33 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "facebook/convnextv2-large-22k-384",
|
| 3 |
"architectures": [
|
| 4 |
+
"ConvNextV2ForImageClassification"
|
| 5 |
],
|
| 6 |
+
"depths": [
|
| 7 |
+
3,
|
| 8 |
+
3,
|
| 9 |
+
27,
|
| 10 |
+
3
|
| 11 |
+
],
|
| 12 |
+
"drop_path_rate": 0.0,
|
| 13 |
"finetuning_task": "image-classification",
|
| 14 |
"hidden_act": "gelu",
|
| 15 |
+
"hidden_sizes": [
|
| 16 |
+
192,
|
| 17 |
+
384,
|
| 18 |
+
768,
|
| 19 |
+
1536
|
| 20 |
+
],
|
| 21 |
"id2label": {
|
| 22 |
"0": "not_western",
|
| 23 |
"1": "western"
|
| 24 |
},
|
| 25 |
"image_size": 224,
|
| 26 |
"initializer_range": 0.02,
|
|
|
|
| 27 |
"label2id": {
|
| 28 |
"not_western": "0",
|
| 29 |
"western": "1"
|
| 30 |
},
|
| 31 |
"layer_norm_eps": 1e-12,
|
| 32 |
+
"model_type": "convnextv2",
|
|
|
|
| 33 |
"num_channels": 3,
|
| 34 |
+
"num_stages": 4,
|
| 35 |
+
"out_features": [
|
| 36 |
+
"stage4"
|
| 37 |
+
],
|
| 38 |
+
"out_indices": [
|
| 39 |
+
4
|
| 40 |
+
],
|
| 41 |
+
"patch_size": 4,
|
| 42 |
"problem_type": "single_label_classification",
|
| 43 |
+
"stage_names": [
|
| 44 |
+
"stem",
|
| 45 |
+
"stage1",
|
| 46 |
+
"stage2",
|
| 47 |
+
"stage3",
|
| 48 |
+
"stage4"
|
| 49 |
+
],
|
| 50 |
"torch_dtype": "float32",
|
| 51 |
"transformers_version": "4.30.0.dev0"
|
| 52 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.7777777777777778,
|
| 4 |
+
"eval_loss": 0.4249849319458008,
|
| 5 |
+
"eval_runtime": 5.6527,
|
| 6 |
+
"eval_samples_per_second": 9.553,
|
| 7 |
+
"eval_steps_per_second": 1.238
|
| 8 |
}
|
preprocessor_config.json
CHANGED
|
@@ -1,22 +1,22 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"do_normalize": true,
|
| 3 |
"do_rescale": true,
|
| 4 |
"do_resize": true,
|
| 5 |
"image_mean": [
|
| 6 |
-
0.
|
| 7 |
-
0.
|
| 8 |
-
0.
|
| 9 |
],
|
| 10 |
-
"image_processor_type": "
|
| 11 |
"image_std": [
|
| 12 |
-
0.
|
| 13 |
-
0.
|
| 14 |
-
0.
|
| 15 |
],
|
| 16 |
-
"resample":
|
| 17 |
"rescale_factor": 0.00392156862745098,
|
| 18 |
"size": {
|
| 19 |
-
"
|
| 20 |
-
"width": 224
|
| 21 |
}
|
| 22 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"crop_pct": 0.875,
|
| 3 |
"do_normalize": true,
|
| 4 |
"do_rescale": true,
|
| 5 |
"do_resize": true,
|
| 6 |
"image_mean": [
|
| 7 |
+
0.485,
|
| 8 |
+
0.456,
|
| 9 |
+
0.406
|
| 10 |
],
|
| 11 |
+
"image_processor_type": "ConvNextImageProcessor",
|
| 12 |
"image_std": [
|
| 13 |
+
0.229,
|
| 14 |
+
0.224,
|
| 15 |
+
0.225
|
| 16 |
],
|
| 17 |
+
"resample": 3,
|
| 18 |
"rescale_factor": 0.00392156862745098,
|
| 19 |
"size": {
|
| 20 |
+
"shortest_edge": 384
|
|
|
|
| 21 |
}
|
| 22 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fba82cf2e625f4ef8e2d6128e772638a78a40848602ee5ed01861967aa0bc0a7
|
| 3 |
+
size 785824061
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second":
|
| 6 |
-
"train_steps_per_second": 1.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"train_loss": 0.39741649894337905,
|
| 4 |
+
"train_runtime": 401.1562,
|
| 5 |
+
"train_samples_per_second": 3.789,
|
| 6 |
+
"train_steps_per_second": 1.895
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,25 +1,526 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
{
|
| 11 |
"epoch": 3.0,
|
| 12 |
-
"
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
}
|
| 19 |
],
|
| 20 |
-
"max_steps":
|
| 21 |
-
"num_train_epochs":
|
| 22 |
-
"total_flos":
|
| 23 |
"trial_name": null,
|
| 24 |
"trial_params": null
|
| 25 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"global_step": 760,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.07,
|
| 12 |
+
"learning_rate": 1.8842105263157898e-05,
|
| 13 |
+
"loss": 0.8301,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.13,
|
| 18 |
+
"learning_rate": 1.7789473684210527e-05,
|
| 19 |
+
"loss": 0.7976,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.2,
|
| 24 |
+
"learning_rate": 1.673684210526316e-05,
|
| 25 |
+
"loss": 0.5757,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.26,
|
| 30 |
+
"learning_rate": 1.568421052631579e-05,
|
| 31 |
+
"loss": 0.4569,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.33,
|
| 36 |
+
"learning_rate": 1.4631578947368424e-05,
|
| 37 |
+
"loss": 0.7308,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.39,
|
| 42 |
+
"learning_rate": 1.3578947368421055e-05,
|
| 43 |
+
"loss": 0.6149,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.46,
|
| 48 |
+
"learning_rate": 1.2526315789473684e-05,
|
| 49 |
+
"loss": 0.666,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.53,
|
| 54 |
+
"learning_rate": 1.1473684210526317e-05,
|
| 55 |
+
"loss": 0.6656,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.59,
|
| 60 |
+
"learning_rate": 1.0421052631578948e-05,
|
| 61 |
+
"loss": 0.5629,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.66,
|
| 66 |
+
"learning_rate": 9.36842105263158e-06,
|
| 67 |
+
"loss": 0.8803,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.72,
|
| 72 |
+
"learning_rate": 8.315789473684212e-06,
|
| 73 |
+
"loss": 0.4404,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.79,
|
| 78 |
+
"learning_rate": 7.263157894736843e-06,
|
| 79 |
+
"loss": 0.3989,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.86,
|
| 84 |
+
"learning_rate": 6.2105263157894745e-06,
|
| 85 |
+
"loss": 0.5877,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.92,
|
| 90 |
+
"learning_rate": 5.157894736842106e-06,
|
| 91 |
+
"loss": 0.5024,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.99,
|
| 96 |
+
"learning_rate": 4.105263157894737e-06,
|
| 97 |
+
"loss": 0.7384,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.0,
|
| 102 |
+
"eval_accuracy": 0.7962962962962963,
|
| 103 |
+
"eval_loss": 0.42684096097946167,
|
| 104 |
+
"eval_runtime": 5.5879,
|
| 105 |
+
"eval_samples_per_second": 9.664,
|
| 106 |
+
"eval_steps_per_second": 1.253,
|
| 107 |
+
"step": 152
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.05,
|
| 111 |
+
"learning_rate": 3.052631578947369e-06,
|
| 112 |
+
"loss": 0.4284,
|
| 113 |
+
"step": 160
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 1.12,
|
| 117 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 118 |
+
"loss": 0.1814,
|
| 119 |
+
"step": 170
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.18,
|
| 123 |
+
"learning_rate": 9.473684210526317e-07,
|
| 124 |
+
"loss": 0.2693,
|
| 125 |
+
"step": 180
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 1.25,
|
| 129 |
+
"learning_rate": 0.0,
|
| 130 |
+
"loss": 0.3471,
|
| 131 |
+
"step": 190
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.32,
|
| 135 |
+
"learning_rate": 0.0,
|
| 136 |
+
"loss": 0.3491,
|
| 137 |
+
"step": 200
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.38,
|
| 141 |
+
"learning_rate": 0.0,
|
| 142 |
+
"loss": 0.2447,
|
| 143 |
+
"step": 210
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.45,
|
| 147 |
+
"learning_rate": 0.0,
|
| 148 |
+
"loss": 0.4893,
|
| 149 |
+
"step": 220
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.51,
|
| 153 |
+
"learning_rate": 0.0,
|
| 154 |
+
"loss": 0.2365,
|
| 155 |
+
"step": 230
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 1.58,
|
| 159 |
+
"learning_rate": 0.0,
|
| 160 |
+
"loss": 0.2919,
|
| 161 |
+
"step": 240
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.64,
|
| 165 |
+
"learning_rate": 0.0,
|
| 166 |
+
"loss": 0.4932,
|
| 167 |
+
"step": 250
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 1.71,
|
| 171 |
+
"learning_rate": 0.0,
|
| 172 |
+
"loss": 0.2127,
|
| 173 |
+
"step": 260
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 1.78,
|
| 177 |
+
"learning_rate": 0.0,
|
| 178 |
+
"loss": 0.2355,
|
| 179 |
+
"step": 270
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.84,
|
| 183 |
+
"learning_rate": 0.0,
|
| 184 |
+
"loss": 0.4058,
|
| 185 |
+
"step": 280
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.91,
|
| 189 |
+
"learning_rate": 0.0,
|
| 190 |
+
"loss": 0.4578,
|
| 191 |
+
"step": 290
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.97,
|
| 195 |
+
"learning_rate": 0.0,
|
| 196 |
+
"loss": 0.2888,
|
| 197 |
+
"step": 300
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 2.0,
|
| 201 |
+
"eval_accuracy": 0.7777777777777778,
|
| 202 |
+
"eval_loss": 0.4249849319458008,
|
| 203 |
+
"eval_runtime": 5.3688,
|
| 204 |
+
"eval_samples_per_second": 10.058,
|
| 205 |
+
"eval_steps_per_second": 1.304,
|
| 206 |
+
"step": 304
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.04,
|
| 210 |
+
"learning_rate": 0.0,
|
| 211 |
+
"loss": 0.5527,
|
| 212 |
+
"step": 310
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.11,
|
| 216 |
+
"learning_rate": 0.0,
|
| 217 |
+
"loss": 0.4907,
|
| 218 |
+
"step": 320
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.17,
|
| 222 |
+
"learning_rate": 0.0,
|
| 223 |
+
"loss": 0.6212,
|
| 224 |
+
"step": 330
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.24,
|
| 228 |
+
"learning_rate": 0.0,
|
| 229 |
+
"loss": 0.3302,
|
| 230 |
+
"step": 340
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.3,
|
| 234 |
+
"learning_rate": 0.0,
|
| 235 |
+
"loss": 0.4032,
|
| 236 |
+
"step": 350
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.37,
|
| 240 |
+
"learning_rate": 0.0,
|
| 241 |
+
"loss": 0.2037,
|
| 242 |
+
"step": 360
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.43,
|
| 246 |
+
"learning_rate": 0.0,
|
| 247 |
+
"loss": 0.2949,
|
| 248 |
+
"step": 370
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.5,
|
| 252 |
+
"learning_rate": 0.0,
|
| 253 |
+
"loss": 0.3068,
|
| 254 |
+
"step": 380
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.57,
|
| 258 |
+
"learning_rate": 0.0,
|
| 259 |
+
"loss": 0.2455,
|
| 260 |
+
"step": 390
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.63,
|
| 264 |
+
"learning_rate": 0.0,
|
| 265 |
+
"loss": 0.2118,
|
| 266 |
+
"step": 400
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.7,
|
| 270 |
+
"learning_rate": 0.0,
|
| 271 |
+
"loss": 0.5577,
|
| 272 |
+
"step": 410
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.76,
|
| 276 |
+
"learning_rate": 0.0,
|
| 277 |
+
"loss": 0.3432,
|
| 278 |
+
"step": 420
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.83,
|
| 282 |
+
"learning_rate": 0.0,
|
| 283 |
+
"loss": 0.2259,
|
| 284 |
+
"step": 430
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.89,
|
| 288 |
+
"learning_rate": 0.0,
|
| 289 |
+
"loss": 0.2736,
|
| 290 |
+
"step": 440
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.96,
|
| 294 |
+
"learning_rate": 0.0,
|
| 295 |
+
"loss": 0.2953,
|
| 296 |
+
"step": 450
|
| 297 |
+
},
|
| 298 |
{
|
| 299 |
"epoch": 3.0,
|
| 300 |
+
"eval_accuracy": 0.7777777777777778,
|
| 301 |
+
"eval_loss": 0.4249849319458008,
|
| 302 |
+
"eval_runtime": 5.4342,
|
| 303 |
+
"eval_samples_per_second": 9.937,
|
| 304 |
+
"eval_steps_per_second": 1.288,
|
| 305 |
+
"step": 456
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 3.03,
|
| 309 |
+
"learning_rate": 0.0,
|
| 310 |
+
"loss": 0.2847,
|
| 311 |
+
"step": 460
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 3.09,
|
| 315 |
+
"learning_rate": 0.0,
|
| 316 |
+
"loss": 0.2117,
|
| 317 |
+
"step": 470
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 3.16,
|
| 321 |
+
"learning_rate": 0.0,
|
| 322 |
+
"loss": 0.2116,
|
| 323 |
+
"step": 480
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.22,
|
| 327 |
+
"learning_rate": 0.0,
|
| 328 |
+
"loss": 0.3803,
|
| 329 |
+
"step": 490
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 3.29,
|
| 333 |
+
"learning_rate": 0.0,
|
| 334 |
+
"loss": 0.3206,
|
| 335 |
+
"step": 500
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 3.36,
|
| 339 |
+
"learning_rate": 0.0,
|
| 340 |
+
"loss": 0.4362,
|
| 341 |
+
"step": 510
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 3.42,
|
| 345 |
+
"learning_rate": 0.0,
|
| 346 |
+
"loss": 0.4031,
|
| 347 |
+
"step": 520
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 3.49,
|
| 351 |
+
"learning_rate": 0.0,
|
| 352 |
+
"loss": 0.2682,
|
| 353 |
+
"step": 530
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 3.55,
|
| 357 |
+
"learning_rate": 0.0,
|
| 358 |
+
"loss": 0.292,
|
| 359 |
+
"step": 540
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.62,
|
| 363 |
+
"learning_rate": 0.0,
|
| 364 |
+
"loss": 0.3007,
|
| 365 |
+
"step": 550
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.68,
|
| 369 |
+
"learning_rate": 0.0,
|
| 370 |
+
"loss": 0.5192,
|
| 371 |
+
"step": 560
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 3.75,
|
| 375 |
+
"learning_rate": 0.0,
|
| 376 |
+
"loss": 0.3809,
|
| 377 |
+
"step": 570
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 3.82,
|
| 381 |
+
"learning_rate": 0.0,
|
| 382 |
+
"loss": 0.3756,
|
| 383 |
+
"step": 580
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 3.88,
|
| 387 |
+
"learning_rate": 0.0,
|
| 388 |
+
"loss": 0.3174,
|
| 389 |
+
"step": 590
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 3.95,
|
| 393 |
+
"learning_rate": 0.0,
|
| 394 |
+
"loss": 0.4914,
|
| 395 |
+
"step": 600
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 4.0,
|
| 399 |
+
"eval_accuracy": 0.7777777777777778,
|
| 400 |
+
"eval_loss": 0.4249849319458008,
|
| 401 |
+
"eval_runtime": 5.3734,
|
| 402 |
+
"eval_samples_per_second": 10.05,
|
| 403 |
+
"eval_steps_per_second": 1.303,
|
| 404 |
+
"step": 608
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 4.01,
|
| 408 |
+
"learning_rate": 0.0,
|
| 409 |
+
"loss": 0.2904,
|
| 410 |
+
"step": 610
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 4.08,
|
| 414 |
+
"learning_rate": 0.0,
|
| 415 |
+
"loss": 0.433,
|
| 416 |
+
"step": 620
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.14,
|
| 420 |
+
"learning_rate": 0.0,
|
| 421 |
+
"loss": 0.1684,
|
| 422 |
+
"step": 630
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 4.21,
|
| 426 |
+
"learning_rate": 0.0,
|
| 427 |
+
"loss": 0.2903,
|
| 428 |
+
"step": 640
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 4.28,
|
| 432 |
+
"learning_rate": 0.0,
|
| 433 |
+
"loss": 0.3659,
|
| 434 |
+
"step": 650
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 4.34,
|
| 438 |
+
"learning_rate": 0.0,
|
| 439 |
+
"loss": 0.2125,
|
| 440 |
+
"step": 660
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 4.41,
|
| 444 |
+
"learning_rate": 0.0,
|
| 445 |
+
"loss": 0.3639,
|
| 446 |
+
"step": 670
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 4.47,
|
| 450 |
+
"learning_rate": 0.0,
|
| 451 |
+
"loss": 0.4092,
|
| 452 |
+
"step": 680
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 4.54,
|
| 456 |
+
"learning_rate": 0.0,
|
| 457 |
+
"loss": 0.3467,
|
| 458 |
+
"step": 690
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 4.61,
|
| 462 |
+
"learning_rate": 0.0,
|
| 463 |
+
"loss": 0.3013,
|
| 464 |
+
"step": 700
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.67,
|
| 468 |
+
"learning_rate": 0.0,
|
| 469 |
+
"loss": 0.2546,
|
| 470 |
+
"step": 710
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 4.74,
|
| 474 |
+
"learning_rate": 0.0,
|
| 475 |
+
"loss": 0.3741,
|
| 476 |
+
"step": 720
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 4.8,
|
| 480 |
+
"learning_rate": 0.0,
|
| 481 |
+
"loss": 0.394,
|
| 482 |
+
"step": 730
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.87,
|
| 486 |
+
"learning_rate": 0.0,
|
| 487 |
+
"loss": 0.2283,
|
| 488 |
+
"step": 740
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.93,
|
| 492 |
+
"learning_rate": 0.0,
|
| 493 |
+
"loss": 0.4307,
|
| 494 |
+
"step": 750
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 5.0,
|
| 498 |
+
"learning_rate": 0.0,
|
| 499 |
+
"loss": 0.4099,
|
| 500 |
+
"step": 760
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 5.0,
|
| 504 |
+
"eval_accuracy": 0.7777777777777778,
|
| 505 |
+
"eval_loss": 0.4249849319458008,
|
| 506 |
+
"eval_runtime": 5.8483,
|
| 507 |
+
"eval_samples_per_second": 9.233,
|
| 508 |
+
"eval_steps_per_second": 1.197,
|
| 509 |
+
"step": 760
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 5.0,
|
| 513 |
+
"step": 760,
|
| 514 |
+
"total_flos": 7.945332390564987e+17,
|
| 515 |
+
"train_loss": 0.39741649894337905,
|
| 516 |
+
"train_runtime": 401.1562,
|
| 517 |
+
"train_samples_per_second": 3.789,
|
| 518 |
+
"train_steps_per_second": 1.895
|
| 519 |
}
|
| 520 |
],
|
| 521 |
+
"max_steps": 760,
|
| 522 |
+
"num_train_epochs": 5,
|
| 523 |
+
"total_flos": 7.945332390564987e+17,
|
| 524 |
"trial_name": null,
|
| 525 |
"trial_params": null
|
| 526 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd3da6756d0e2aff4054a6a80b1de56d02c3602827fa3ba6063bd32a09111ae0
|
| 3 |
+
size 3835
|