Commit
·
e294472
1
Parent(s):
900d962
Add evaluation results and output files to model card
Browse files- README.md +36 -0
- checkpoint-140/config.json +24 -0
- checkpoint-140/model.safetensors +3 -0
- checkpoint-140/optimizer.pt +3 -0
- checkpoint-140/preprocessor_config.json +23 -0
- checkpoint-140/rng_state.pth +3 -0
- checkpoint-140/scaler.pt +3 -0
- checkpoint-140/scheduler.pt +3 -0
- checkpoint-140/trainer_state.json +896 -0
- checkpoint-140/training_args.bin +3 -0
- checkpoint-60/config.json +24 -0
- checkpoint-60/model.safetensors +3 -0
- checkpoint-60/optimizer.pt +3 -0
- checkpoint-60/preprocessor_config.json +23 -0
- checkpoint-60/rng_state.pth +3 -0
- checkpoint-60/scaler.pt +3 -0
- checkpoint-60/scheduler.pt +3 -0
- checkpoint-60/trainer_state.json +414 -0
- checkpoint-60/training_args.bin +3 -0
- config.json +24 -0
- model.safetensors +3 -0
- pr_curve.png +0 -0
- preprocessor_config.json +23 -0
- roc_curve.png +0 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: code
|
| 3 |
+
tags:
|
| 4 |
+
- binary-classification
|
| 5 |
+
- model
|
| 6 |
+
- evaluation
|
| 7 |
+
metrics:
|
| 8 |
+
- average_precision: 0.94
|
| 9 |
+
- roc_auc: 0.77
|
| 10 |
+
- best threshold according to F1: 0.35
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Binary Classification Model
|
| 14 |
+
|
| 15 |
+
## Evaluation Results
|
| 16 |
+
|
| 17 |
+
**Average Precision:** 0.94
|
| 18 |
+
**ROC AUC:** 0.77
|
| 19 |
+
**best threshold according to F1: 0.35
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
## Visualizations
|
| 23 |
+
|
| 24 |
+
### Precision-Recall Curve
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
### ROC Curve
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
## Output Files and Directories
|
| 31 |
+
|
| 32 |
+
- 📂 `checkpoint-171/`
|
| 33 |
+
- `config.json`
|
| 34 |
+
- `model.safetensors`
|
| 35 |
+
- `preprocessor_config.json`
|
| 36 |
+
- `training_args.bin`
|
checkpoint-140/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-large-patch16-224-in21k",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 1024,
|
| 11 |
+
"image_size": 256,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 4096,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"model_type": "vit",
|
| 16 |
+
"num_attention_heads": 16,
|
| 17 |
+
"num_channels": 3,
|
| 18 |
+
"num_hidden_layers": 24,
|
| 19 |
+
"patch_size": 16,
|
| 20 |
+
"problem_type": "single_label_classification",
|
| 21 |
+
"qkv_bias": true,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.49.0"
|
| 24 |
+
}
|
checkpoint-140/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:411bc377ee23010cd104a7db7aa3b35593fbc4aec8b99a0de7cf6d3aebe2e6d5
|
| 3 |
+
size 1213507024
|
checkpoint-140/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fdb21d4fe924730bbaf2207322d323e644de556d92c10b5d68d054ca74f42ee3
|
| 3 |
+
size 2427249066
|
checkpoint-140/preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": false,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTFeatureExtractor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"resample": 2,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-140/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2f115a5be9979771da3a2b09f730d7ed2191451f3782c4df1f5bc84d6de6f11
|
| 3 |
+
size 14244
|
checkpoint-140/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7049f1e74bf3476ebf57f8bbfe08ec3a0311fb5d49b48b0e4a8fd2c8f05762f
|
| 3 |
+
size 988
|
checkpoint-140/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:66c6a214ae787ddac4978d5c5b90944353a182391d33432e9f22fb5f397b5439
|
| 3 |
+
size 1064
|
checkpoint-140/trainer_state.json
ADDED
|
@@ -0,0 +1,896 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9830508474576272,
|
| 3 |
+
"best_model_checkpoint": "/tmp/logs/GenetikaPlus/clf_soma_model_v1.0.1/checkpoint-60",
|
| 4 |
+
"epoch": 35.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 140,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2857142857142857,
|
| 13 |
+
"grad_norm": 3.202357053756714,
|
| 14 |
+
"learning_rate": 8.333333333333333e-05,
|
| 15 |
+
"loss": 0.6786,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.8571428571428571,
|
| 20 |
+
"grad_norm": 4.824620246887207,
|
| 21 |
+
"learning_rate": 0.00025,
|
| 22 |
+
"loss": 0.7317,
|
| 23 |
+
"step": 3
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_accuracy": 0.9577464788732394,
|
| 28 |
+
"eval_best_threshold": 0.3037576377391815,
|
| 29 |
+
"eval_f1": 0.9783132530120482,
|
| 30 |
+
"eval_loss": 0.6482536792755127,
|
| 31 |
+
"eval_pr_auc": 0.9798696623714835,
|
| 32 |
+
"eval_precision": 0.9575471698113207,
|
| 33 |
+
"eval_recall": 1.0,
|
| 34 |
+
"eval_roc_auc": 0.7300492610837439,
|
| 35 |
+
"eval_runtime": 3.6503,
|
| 36 |
+
"eval_samples_per_second": 58.351,
|
| 37 |
+
"eval_steps_per_second": 1.096,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.5714285714285714,
|
| 42 |
+
"grad_norm": 8.443906784057617,
|
| 43 |
+
"learning_rate": 0.0005,
|
| 44 |
+
"loss": 0.9287,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.0,
|
| 49 |
+
"eval_accuracy": 0.9577464788732394,
|
| 50 |
+
"eval_best_threshold": 0.09376094490289688,
|
| 51 |
+
"eval_f1": 0.9783132530120482,
|
| 52 |
+
"eval_loss": 1.2041555643081665,
|
| 53 |
+
"eval_pr_auc": 0.9823049604214897,
|
| 54 |
+
"eval_precision": 0.9575471698113207,
|
| 55 |
+
"eval_recall": 1.0,
|
| 56 |
+
"eval_roc_auc": 0.7192118226600985,
|
| 57 |
+
"eval_runtime": 3.8529,
|
| 58 |
+
"eval_samples_per_second": 55.283,
|
| 59 |
+
"eval_steps_per_second": 1.038,
|
| 60 |
+
"step": 8
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 2.2857142857142856,
|
| 64 |
+
"grad_norm": 7.837808609008789,
|
| 65 |
+
"learning_rate": 0.0004999685319184688,
|
| 66 |
+
"loss": 1.0591,
|
| 67 |
+
"step": 9
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 3.0,
|
| 71 |
+
"grad_norm": 0.33617040514945984,
|
| 72 |
+
"learning_rate": 0.0004998741355957963,
|
| 73 |
+
"loss": 0.6429,
|
| 74 |
+
"step": 12
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 3.0,
|
| 78 |
+
"eval_accuracy": 0.9530516431924883,
|
| 79 |
+
"eval_best_threshold": 0.11066323518753052,
|
| 80 |
+
"eval_f1": 0.9759615384615384,
|
| 81 |
+
"eval_loss": 0.6775969862937927,
|
| 82 |
+
"eval_pr_auc": 0.9570451344092105,
|
| 83 |
+
"eval_precision": 0.9530516431924883,
|
| 84 |
+
"eval_recall": 1.0,
|
| 85 |
+
"eval_roc_auc": 0.4647783251231527,
|
| 86 |
+
"eval_runtime": 4.3875,
|
| 87 |
+
"eval_samples_per_second": 48.547,
|
| 88 |
+
"eval_steps_per_second": 0.912,
|
| 89 |
+
"step": 12
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 3.857142857142857,
|
| 93 |
+
"grad_norm": 2.0180256366729736,
|
| 94 |
+
"learning_rate": 0.000499716834795752,
|
| 95 |
+
"loss": 0.7296,
|
| 96 |
+
"step": 15
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 4.0,
|
| 100 |
+
"eval_accuracy": 0.9577464788732394,
|
| 101 |
+
"eval_best_threshold": 0.488298624753952,
|
| 102 |
+
"eval_f1": 0.9783132530120482,
|
| 103 |
+
"eval_loss": 0.6953739523887634,
|
| 104 |
+
"eval_pr_auc": 0.9387681415619215,
|
| 105 |
+
"eval_precision": 0.9575471698113207,
|
| 106 |
+
"eval_recall": 1.0,
|
| 107 |
+
"eval_roc_auc": 0.3980295566502463,
|
| 108 |
+
"eval_runtime": 3.9434,
|
| 109 |
+
"eval_samples_per_second": 54.014,
|
| 110 |
+
"eval_steps_per_second": 1.014,
|
| 111 |
+
"step": 16
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 4.571428571428571,
|
| 115 |
+
"grad_norm": 0.6065838932991028,
|
| 116 |
+
"learning_rate": 0.0004994966691179711,
|
| 117 |
+
"loss": 0.5691,
|
| 118 |
+
"step": 18
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 5.0,
|
| 122 |
+
"eval_accuracy": 0.9577464788732394,
|
| 123 |
+
"eval_best_threshold": 0.4559290409088135,
|
| 124 |
+
"eval_f1": 0.9783132530120482,
|
| 125 |
+
"eval_loss": 0.6963727474212646,
|
| 126 |
+
"eval_pr_auc": 0.9437271902417405,
|
| 127 |
+
"eval_precision": 0.9575471698113207,
|
| 128 |
+
"eval_recall": 1.0,
|
| 129 |
+
"eval_roc_auc": 0.42167487684729066,
|
| 130 |
+
"eval_runtime": 4.451,
|
| 131 |
+
"eval_samples_per_second": 47.854,
|
| 132 |
+
"eval_steps_per_second": 0.899,
|
| 133 |
+
"step": 20
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 5.285714285714286,
|
| 137 |
+
"grad_norm": 0.7243921160697937,
|
| 138 |
+
"learning_rate": 0.0004992136939879857,
|
| 139 |
+
"loss": 0.5782,
|
| 140 |
+
"step": 21
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 6.0,
|
| 144 |
+
"grad_norm": 0.2275872528553009,
|
| 145 |
+
"learning_rate": 0.0004988679806432712,
|
| 146 |
+
"loss": 0.5628,
|
| 147 |
+
"step": 24
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 6.0,
|
| 151 |
+
"eval_accuracy": 0.9624413145539906,
|
| 152 |
+
"eval_best_threshold": 0.40434572100639343,
|
| 153 |
+
"eval_f1": 0.9806763285024155,
|
| 154 |
+
"eval_loss": 0.6847261786460876,
|
| 155 |
+
"eval_pr_auc": 0.9548760246480454,
|
| 156 |
+
"eval_precision": 0.9620853080568721,
|
| 157 |
+
"eval_recall": 1.0,
|
| 158 |
+
"eval_roc_auc": 0.5586206896551724,
|
| 159 |
+
"eval_runtime": 4.6725,
|
| 160 |
+
"eval_samples_per_second": 45.586,
|
| 161 |
+
"eval_steps_per_second": 0.856,
|
| 162 |
+
"step": 24
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 6.857142857142857,
|
| 166 |
+
"grad_norm": 1.0703548192977905,
|
| 167 |
+
"learning_rate": 0.0004984596161153135,
|
| 168 |
+
"loss": 0.6532,
|
| 169 |
+
"step": 27
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 7.0,
|
| 173 |
+
"eval_accuracy": 0.9624413145539906,
|
| 174 |
+
"eval_best_threshold": 0.3936326503753662,
|
| 175 |
+
"eval_f1": 0.9805825242718447,
|
| 176 |
+
"eval_loss": 0.6566980481147766,
|
| 177 |
+
"eval_pr_auc": 0.9620782247005666,
|
| 178 |
+
"eval_precision": 0.9665071770334929,
|
| 179 |
+
"eval_recall": 0.9950738916256158,
|
| 180 |
+
"eval_roc_auc": 0.5571428571428572,
|
| 181 |
+
"eval_runtime": 4.3448,
|
| 182 |
+
"eval_samples_per_second": 49.024,
|
| 183 |
+
"eval_steps_per_second": 0.921,
|
| 184 |
+
"step": 28
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 7.571428571428571,
|
| 188 |
+
"grad_norm": 2.041855573654175,
|
| 189 |
+
"learning_rate": 0.0004979887032076989,
|
| 190 |
+
"loss": 0.5593,
|
| 191 |
+
"step": 30
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 8.0,
|
| 195 |
+
"eval_accuracy": 0.9624413145539906,
|
| 196 |
+
"eval_best_threshold": 0.2755194306373596,
|
| 197 |
+
"eval_f1": 0.9806763285024155,
|
| 198 |
+
"eval_loss": 0.6331985592842102,
|
| 199 |
+
"eval_pr_auc": 0.9751121096020542,
|
| 200 |
+
"eval_precision": 0.9620853080568721,
|
| 201 |
+
"eval_recall": 1.0,
|
| 202 |
+
"eval_roc_auc": 0.6586206896551724,
|
| 203 |
+
"eval_runtime": 4.7804,
|
| 204 |
+
"eval_samples_per_second": 44.557,
|
| 205 |
+
"eval_steps_per_second": 0.837,
|
| 206 |
+
"step": 32
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 8.285714285714286,
|
| 210 |
+
"grad_norm": 0.8734021186828613,
|
| 211 |
+
"learning_rate": 0.0004974553604702333,
|
| 212 |
+
"loss": 0.5002,
|
| 213 |
+
"step": 33
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 9.0,
|
| 217 |
+
"grad_norm": Infinity,
|
| 218 |
+
"learning_rate": 0.0004970651810649666,
|
| 219 |
+
"loss": 0.5346,
|
| 220 |
+
"step": 36
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 9.0,
|
| 224 |
+
"eval_accuracy": 0.9577464788732394,
|
| 225 |
+
"eval_best_threshold": 0.14823205769062042,
|
| 226 |
+
"eval_f1": 0.9783132530120482,
|
| 227 |
+
"eval_loss": 0.6035900712013245,
|
| 228 |
+
"eval_pr_auc": 0.9819011692535515,
|
| 229 |
+
"eval_precision": 0.9575471698113207,
|
| 230 |
+
"eval_recall": 1.0,
|
| 231 |
+
"eval_roc_auc": 0.7339901477832512,
|
| 232 |
+
"eval_runtime": 4.0039,
|
| 233 |
+
"eval_samples_per_second": 53.198,
|
| 234 |
+
"eval_steps_per_second": 0.999,
|
| 235 |
+
"step": 36
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 9.857142857142858,
|
| 239 |
+
"grad_norm": 1.1991403102874756,
|
| 240 |
+
"learning_rate": 0.0004964280947263676,
|
| 241 |
+
"loss": 0.5955,
|
| 242 |
+
"step": 39
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 10.0,
|
| 246 |
+
"eval_accuracy": 0.9624413145539906,
|
| 247 |
+
"eval_best_threshold": 0.5386950373649597,
|
| 248 |
+
"eval_f1": 0.9805825242718447,
|
| 249 |
+
"eval_loss": 0.6214358806610107,
|
| 250 |
+
"eval_pr_auc": 0.9624581599102329,
|
| 251 |
+
"eval_precision": 0.9665071770334929,
|
| 252 |
+
"eval_recall": 0.9950738916256158,
|
| 253 |
+
"eval_roc_auc": 0.6002463054187192,
|
| 254 |
+
"eval_runtime": 4.5609,
|
| 255 |
+
"eval_samples_per_second": 46.701,
|
| 256 |
+
"eval_steps_per_second": 0.877,
|
| 257 |
+
"step": 40
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 10.571428571428571,
|
| 261 |
+
"grad_norm": 2.6938207149505615,
|
| 262 |
+
"learning_rate": 0.0004957289714327572,
|
| 263 |
+
"loss": 0.4496,
|
| 264 |
+
"step": 42
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 11.0,
|
| 268 |
+
"eval_accuracy": 0.9530516431924883,
|
| 269 |
+
"eval_best_threshold": 0.04157886654138565,
|
| 270 |
+
"eval_f1": 0.9759615384615384,
|
| 271 |
+
"eval_loss": 0.8543573617935181,
|
| 272 |
+
"eval_pr_auc": 0.9879501355451374,
|
| 273 |
+
"eval_precision": 0.9530516431924883,
|
| 274 |
+
"eval_recall": 1.0,
|
| 275 |
+
"eval_roc_auc": 0.7901477832512316,
|
| 276 |
+
"eval_runtime": 4.6335,
|
| 277 |
+
"eval_samples_per_second": 45.969,
|
| 278 |
+
"eval_steps_per_second": 0.863,
|
| 279 |
+
"step": 44
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 11.285714285714286,
|
| 283 |
+
"grad_norm": 3.554097890853882,
|
| 284 |
+
"learning_rate": 0.0004949679871846857,
|
| 285 |
+
"loss": 0.6118,
|
| 286 |
+
"step": 45
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 12.0,
|
| 290 |
+
"grad_norm": 2.869584083557129,
|
| 291 |
+
"learning_rate": 0.0004941453335558681,
|
| 292 |
+
"loss": 0.6877,
|
| 293 |
+
"step": 48
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 12.0,
|
| 297 |
+
"eval_accuracy": 0.9530516431924883,
|
| 298 |
+
"eval_best_threshold": 0.17342814803123474,
|
| 299 |
+
"eval_f1": 0.9759615384615384,
|
| 300 |
+
"eval_loss": 0.5579107403755188,
|
| 301 |
+
"eval_pr_auc": 0.9873731974164195,
|
| 302 |
+
"eval_precision": 0.9530516431924883,
|
| 303 |
+
"eval_recall": 1.0,
|
| 304 |
+
"eval_roc_auc": 0.7891625615763547,
|
| 305 |
+
"eval_runtime": 4.1544,
|
| 306 |
+
"eval_samples_per_second": 51.271,
|
| 307 |
+
"eval_steps_per_second": 0.963,
|
| 308 |
+
"step": 48
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 12.857142857142858,
|
| 312 |
+
"grad_norm": 1.1457290649414062,
|
| 313 |
+
"learning_rate": 0.0004932612176449559,
|
| 314 |
+
"loss": 0.5374,
|
| 315 |
+
"step": 51
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 13.0,
|
| 319 |
+
"eval_accuracy": 0.9624413145539906,
|
| 320 |
+
"eval_best_threshold": 0.5808722376823425,
|
| 321 |
+
"eval_f1": 0.9806763285024155,
|
| 322 |
+
"eval_loss": 0.5983880758285522,
|
| 323 |
+
"eval_pr_auc": 0.9837750859728142,
|
| 324 |
+
"eval_precision": 0.9620853080568721,
|
| 325 |
+
"eval_recall": 1.0,
|
| 326 |
+
"eval_roc_auc": 0.7578817733990147,
|
| 327 |
+
"eval_runtime": 4.6879,
|
| 328 |
+
"eval_samples_per_second": 45.436,
|
| 329 |
+
"eval_steps_per_second": 0.853,
|
| 330 |
+
"step": 52
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 13.571428571428571,
|
| 334 |
+
"grad_norm": 2.2151918411254883,
|
| 335 |
+
"learning_rate": 0.000492315862023402,
|
| 336 |
+
"loss": 0.5329,
|
| 337 |
+
"step": 54
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 14.0,
|
| 341 |
+
"eval_accuracy": 0.9624413145539906,
|
| 342 |
+
"eval_best_threshold": 0.1792430430650711,
|
| 343 |
+
"eval_f1": 0.9806763285024155,
|
| 344 |
+
"eval_loss": 0.5941116213798523,
|
| 345 |
+
"eval_pr_auc": 0.987113937851267,
|
| 346 |
+
"eval_precision": 0.9620853080568721,
|
| 347 |
+
"eval_recall": 1.0,
|
| 348 |
+
"eval_roc_auc": 0.7970443349753695,
|
| 349 |
+
"eval_runtime": 4.3199,
|
| 350 |
+
"eval_samples_per_second": 49.306,
|
| 351 |
+
"eval_steps_per_second": 0.926,
|
| 352 |
+
"step": 56
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 14.285714285714286,
|
| 356 |
+
"grad_norm": 2.5402119159698486,
|
| 357 |
+
"learning_rate": 0.0004913095046794281,
|
| 358 |
+
"loss": 0.4644,
|
| 359 |
+
"step": 57
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 15.0,
|
| 363 |
+
"grad_norm": 2.0666849613189697,
|
| 364 |
+
"learning_rate": 0.0004902423989581143,
|
| 365 |
+
"loss": 0.5655,
|
| 366 |
+
"step": 60
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 15.0,
|
| 370 |
+
"eval_accuracy": 0.9671361502347418,
|
| 371 |
+
"eval_best_threshold": 0.09999631345272064,
|
| 372 |
+
"eval_f1": 0.9830508474576272,
|
| 373 |
+
"eval_loss": 0.6284329891204834,
|
| 374 |
+
"eval_pr_auc": 0.9873903178253213,
|
| 375 |
+
"eval_precision": 0.9666666666666667,
|
| 376 |
+
"eval_recall": 1.0,
|
| 377 |
+
"eval_roc_auc": 0.8014778325123153,
|
| 378 |
+
"eval_runtime": 3.9163,
|
| 379 |
+
"eval_samples_per_second": 54.388,
|
| 380 |
+
"eval_steps_per_second": 1.021,
|
| 381 |
+
"step": 60
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 15.857142857142858,
|
| 385 |
+
"grad_norm": 0.3262127935886383,
|
| 386 |
+
"learning_rate": 0.000489114813497619,
|
| 387 |
+
"loss": 0.5623,
|
| 388 |
+
"step": 63
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 16.0,
|
| 392 |
+
"eval_accuracy": 0.9624413145539906,
|
| 393 |
+
"eval_best_threshold": 0.40762078762054443,
|
| 394 |
+
"eval_f1": 0.9806763285024155,
|
| 395 |
+
"eval_loss": 0.5445640683174133,
|
| 396 |
+
"eval_pr_auc": 0.9843443641688615,
|
| 397 |
+
"eval_precision": 0.9620853080568721,
|
| 398 |
+
"eval_recall": 1.0,
|
| 399 |
+
"eval_roc_auc": 0.7625615763546798,
|
| 400 |
+
"eval_runtime": 4.4164,
|
| 401 |
+
"eval_samples_per_second": 48.229,
|
| 402 |
+
"eval_steps_per_second": 0.906,
|
| 403 |
+
"step": 64
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 16.571428571428573,
|
| 407 |
+
"grad_norm": 1.1104872226715088,
|
| 408 |
+
"learning_rate": 0.000487927032161552,
|
| 409 |
+
"loss": 0.4934,
|
| 410 |
+
"step": 66
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 17.0,
|
| 414 |
+
"eval_accuracy": 0.9624413145539906,
|
| 415 |
+
"eval_best_threshold": 0.08254930377006531,
|
| 416 |
+
"eval_f1": 0.9806763285024155,
|
| 417 |
+
"eval_loss": 0.5435951352119446,
|
| 418 |
+
"eval_pr_auc": 0.9867802837753047,
|
| 419 |
+
"eval_precision": 0.9620853080568721,
|
| 420 |
+
"eval_recall": 1.0,
|
| 421 |
+
"eval_roc_auc": 0.7965517241379311,
|
| 422 |
+
"eval_runtime": 4.8194,
|
| 423 |
+
"eval_samples_per_second": 44.197,
|
| 424 |
+
"eval_steps_per_second": 0.83,
|
| 425 |
+
"step": 68
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 17.285714285714285,
|
| 429 |
+
"grad_norm": 1.3775012493133545,
|
| 430 |
+
"learning_rate": 0.00048667935396751264,
|
| 431 |
+
"loss": 0.4147,
|
| 432 |
+
"step": 69
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 18.0,
|
| 436 |
+
"grad_norm": 0.987482488155365,
|
| 437 |
+
"learning_rate": 0.0004853720930118139,
|
| 438 |
+
"loss": 0.4838,
|
| 439 |
+
"step": 72
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 18.0,
|
| 443 |
+
"eval_accuracy": 0.9624413145539906,
|
| 444 |
+
"eval_best_threshold": 0.2583727538585663,
|
| 445 |
+
"eval_f1": 0.9806763285024155,
|
| 446 |
+
"eval_loss": 0.5534961223602295,
|
| 447 |
+
"eval_pr_auc": 0.9766965382800533,
|
| 448 |
+
"eval_precision": 0.9620853080568721,
|
| 449 |
+
"eval_recall": 1.0,
|
| 450 |
+
"eval_roc_auc": 0.6854679802955665,
|
| 451 |
+
"eval_runtime": 4.4444,
|
| 452 |
+
"eval_samples_per_second": 47.926,
|
| 453 |
+
"eval_steps_per_second": 0.9,
|
| 454 |
+
"step": 72
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 18.857142857142858,
|
| 458 |
+
"grad_norm": 0.7703476548194885,
|
| 459 |
+
"learning_rate": 0.00048400557839041057,
|
| 460 |
+
"loss": 0.5121,
|
| 461 |
+
"step": 75
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 19.0,
|
| 465 |
+
"eval_accuracy": 0.9624413145539906,
|
| 466 |
+
"eval_best_threshold": 0.06578465551137924,
|
| 467 |
+
"eval_f1": 0.9806763285024155,
|
| 468 |
+
"eval_loss": 0.5378168225288391,
|
| 469 |
+
"eval_pr_auc": 0.9891977624351356,
|
| 470 |
+
"eval_precision": 0.9620853080568721,
|
| 471 |
+
"eval_recall": 1.0,
|
| 472 |
+
"eval_roc_auc": 0.8330049261083744,
|
| 473 |
+
"eval_runtime": 4.5334,
|
| 474 |
+
"eval_samples_per_second": 46.984,
|
| 475 |
+
"eval_steps_per_second": 0.882,
|
| 476 |
+
"step": 76
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 19.571428571428573,
|
| 480 |
+
"grad_norm": 1.074415922164917,
|
| 481 |
+
"learning_rate": 0.00048258015411605095,
|
| 482 |
+
"loss": 0.4896,
|
| 483 |
+
"step": 78
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 20.0,
|
| 487 |
+
"eval_accuracy": 0.9624413145539906,
|
| 488 |
+
"eval_best_threshold": 0.29690900444984436,
|
| 489 |
+
"eval_f1": 0.9806763285024155,
|
| 490 |
+
"eval_loss": 0.57256019115448,
|
| 491 |
+
"eval_pr_auc": 0.9841532218125894,
|
| 492 |
+
"eval_precision": 0.9620853080568721,
|
| 493 |
+
"eval_recall": 1.0,
|
| 494 |
+
"eval_roc_auc": 0.758128078817734,
|
| 495 |
+
"eval_runtime": 4.6069,
|
| 496 |
+
"eval_samples_per_second": 46.235,
|
| 497 |
+
"eval_steps_per_second": 0.868,
|
| 498 |
+
"step": 80
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 20.285714285714285,
|
| 502 |
+
"grad_norm": 2.890902042388916,
|
| 503 |
+
"learning_rate": 0.00048109617903167303,
|
| 504 |
+
"loss": 0.403,
|
| 505 |
+
"step": 81
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 21.0,
|
| 509 |
+
"grad_norm": 0.32196927070617676,
|
| 510 |
+
"learning_rate": 0.0004795540267200685,
|
| 511 |
+
"loss": 0.4986,
|
| 512 |
+
"step": 84
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 21.0,
|
| 516 |
+
"eval_accuracy": 0.9624413145539906,
|
| 517 |
+
"eval_best_threshold": 0.24242974817752838,
|
| 518 |
+
"eval_f1": 0.9806763285024155,
|
| 519 |
+
"eval_loss": 0.5482825040817261,
|
| 520 |
+
"eval_pr_auc": 0.9836054537100154,
|
| 521 |
+
"eval_precision": 0.9620853080568721,
|
| 522 |
+
"eval_recall": 1.0,
|
| 523 |
+
"eval_roc_auc": 0.7773399014778325,
|
| 524 |
+
"eval_runtime": 4.3525,
|
| 525 |
+
"eval_samples_per_second": 48.938,
|
| 526 |
+
"eval_steps_per_second": 0.919,
|
| 527 |
+
"step": 84
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 21.857142857142858,
|
| 531 |
+
"grad_norm": 0.31015127897262573,
|
| 532 |
+
"learning_rate": 0.00047795408540983475,
|
| 533 |
+
"loss": 0.532,
|
| 534 |
+
"step": 87
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 22.0,
|
| 538 |
+
"eval_accuracy": 0.9624413145539906,
|
| 539 |
+
"eval_best_threshold": 0.07094483077526093,
|
| 540 |
+
"eval_f1": 0.9806763285024155,
|
| 541 |
+
"eval_loss": 0.5339506268501282,
|
| 542 |
+
"eval_pr_auc": 0.9872064689219997,
|
| 543 |
+
"eval_precision": 0.9620853080568721,
|
| 544 |
+
"eval_recall": 1.0,
|
| 545 |
+
"eval_roc_auc": 0.8167487684729065,
|
| 546 |
+
"eval_runtime": 4.5492,
|
| 547 |
+
"eval_samples_per_second": 46.822,
|
| 548 |
+
"eval_steps_per_second": 0.879,
|
| 549 |
+
"step": 88
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 22.571428571428573,
|
| 553 |
+
"grad_norm": 0.4279550313949585,
|
| 554 |
+
"learning_rate": 0.0004762967578776406,
|
| 555 |
+
"loss": 0.4404,
|
| 556 |
+
"step": 90
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 23.0,
|
| 560 |
+
"eval_accuracy": 0.9671361502347418,
|
| 561 |
+
"eval_best_threshold": 0.4337865710258484,
|
| 562 |
+
"eval_f1": 0.9830508474576272,
|
| 563 |
+
"eval_loss": 0.5571203231811523,
|
| 564 |
+
"eval_pr_auc": 0.9819982800942599,
|
| 565 |
+
"eval_precision": 0.9666666666666667,
|
| 566 |
+
"eval_recall": 1.0,
|
| 567 |
+
"eval_roc_auc": 0.7495073891625617,
|
| 568 |
+
"eval_runtime": 4.8091,
|
| 569 |
+
"eval_samples_per_second": 44.291,
|
| 570 |
+
"eval_steps_per_second": 0.832,
|
| 571 |
+
"step": 92
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 23.285714285714285,
|
| 575 |
+
"grad_norm": 1.652251958847046,
|
| 576 |
+
"learning_rate": 0.00047458246134682926,
|
| 577 |
+
"loss": 0.4665,
|
| 578 |
+
"step": 93
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 24.0,
|
| 582 |
+
"grad_norm": 0.5141251087188721,
|
| 583 |
+
"learning_rate": 0.0004728116273823847,
|
| 584 |
+
"loss": 0.473,
|
| 585 |
+
"step": 96
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 24.0,
|
| 589 |
+
"eval_accuracy": 0.9671361502347418,
|
| 590 |
+
"eval_best_threshold": 0.32589587569236755,
|
| 591 |
+
"eval_f1": 0.9830508474576272,
|
| 592 |
+
"eval_loss": 0.5704872012138367,
|
| 593 |
+
"eval_pr_auc": 0.9865814075621451,
|
| 594 |
+
"eval_precision": 0.9666666666666667,
|
| 595 |
+
"eval_recall": 1.0,
|
| 596 |
+
"eval_roc_auc": 0.7926108374384236,
|
| 597 |
+
"eval_runtime": 4.3422,
|
| 598 |
+
"eval_samples_per_second": 49.054,
|
| 599 |
+
"eval_steps_per_second": 0.921,
|
| 600 |
+
"step": 96
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 24.857142857142858,
|
| 604 |
+
"grad_norm": 2.2477593421936035,
|
| 605 |
+
"learning_rate": 0.0004709847017822876,
|
| 606 |
+
"loss": 0.578,
|
| 607 |
+
"step": 99
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 25.0,
|
| 611 |
+
"eval_accuracy": 0.9671361502347418,
|
| 612 |
+
"eval_best_threshold": 0.32844865322113037,
|
| 613 |
+
"eval_f1": 0.9830508474576272,
|
| 614 |
+
"eval_loss": 0.5548251271247864,
|
| 615 |
+
"eval_pr_auc": 0.9830996192936517,
|
| 616 |
+
"eval_precision": 0.9666666666666667,
|
| 617 |
+
"eval_recall": 1.0,
|
| 618 |
+
"eval_roc_auc": 0.7620689655172413,
|
| 619 |
+
"eval_runtime": 4.2924,
|
| 620 |
+
"eval_samples_per_second": 49.623,
|
| 621 |
+
"eval_steps_per_second": 0.932,
|
| 622 |
+
"step": 100
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 25.571428571428573,
|
| 626 |
+
"grad_norm": 3.829688310623169,
|
| 627 |
+
"learning_rate": 0.00046910214446528766,
|
| 628 |
+
"loss": 0.4749,
|
| 629 |
+
"step": 102
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 26.0,
|
| 633 |
+
"eval_accuracy": 0.9624413145539906,
|
| 634 |
+
"eval_best_threshold": 0.0979498103260994,
|
| 635 |
+
"eval_f1": 0.9806763285024155,
|
| 636 |
+
"eval_loss": 0.4922328591346741,
|
| 637 |
+
"eval_pr_auc": 0.9846144860590845,
|
| 638 |
+
"eval_precision": 0.9620853080568721,
|
| 639 |
+
"eval_recall": 1.0,
|
| 640 |
+
"eval_roc_auc": 0.7921182266009853,
|
| 641 |
+
"eval_runtime": 4.1671,
|
| 642 |
+
"eval_samples_per_second": 51.115,
|
| 643 |
+
"eval_steps_per_second": 0.96,
|
| 644 |
+
"step": 104
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 26.285714285714285,
|
| 648 |
+
"grad_norm": 0.6822347044944763,
|
| 649 |
+
"learning_rate": 0.00046716442935512215,
|
| 650 |
+
"loss": 0.4634,
|
| 651 |
+
"step": 105
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 27.0,
|
| 655 |
+
"grad_norm": 0.3724188506603241,
|
| 656 |
+
"learning_rate": 0.00046517204426120757,
|
| 657 |
+
"loss": 0.3946,
|
| 658 |
+
"step": 108
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 27.0,
|
| 662 |
+
"eval_accuracy": 0.9671361502347418,
|
| 663 |
+
"eval_best_threshold": 0.14669741690158844,
|
| 664 |
+
"eval_f1": 0.9830508474576272,
|
| 665 |
+
"eval_loss": 0.5319708585739136,
|
| 666 |
+
"eval_pr_auc": 0.9846531299061106,
|
| 667 |
+
"eval_precision": 0.9666666666666667,
|
| 668 |
+
"eval_recall": 1.0,
|
| 669 |
+
"eval_roc_auc": 0.7795566502463055,
|
| 670 |
+
"eval_runtime": 4.0384,
|
| 671 |
+
"eval_samples_per_second": 52.743,
|
| 672 |
+
"eval_steps_per_second": 0.99,
|
| 673 |
+
"step": 108
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 27.857142857142858,
|
| 677 |
+
"grad_norm": 3.3339312076568604,
|
| 678 |
+
"learning_rate": 0.0004631254907558365,
|
| 679 |
+
"loss": 0.6008,
|
| 680 |
+
"step": 111
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 28.0,
|
| 684 |
+
"eval_accuracy": 0.9624413145539906,
|
| 685 |
+
"eval_best_threshold": 0.4017758369445801,
|
| 686 |
+
"eval_f1": 0.9804878048780488,
|
| 687 |
+
"eval_loss": 0.4973604679107666,
|
| 688 |
+
"eval_pr_auc": 0.9854492936449624,
|
| 689 |
+
"eval_precision": 0.9710144927536232,
|
| 690 |
+
"eval_recall": 0.9901477832512315,
|
| 691 |
+
"eval_roc_auc": 0.7903940886699506,
|
| 692 |
+
"eval_runtime": 4.4645,
|
| 693 |
+
"eval_samples_per_second": 47.71,
|
| 694 |
+
"eval_steps_per_second": 0.896,
|
| 695 |
+
"step": 112
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 28.571428571428573,
|
| 699 |
+
"grad_norm": 1.0172795057296753,
|
| 700 |
+
"learning_rate": 0.00046102528404790965,
|
| 701 |
+
"loss": 0.52,
|
| 702 |
+
"step": 114
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 29.0,
|
| 706 |
+
"eval_accuracy": 0.9577464788732394,
|
| 707 |
+
"eval_best_threshold": 0.02784844860434532,
|
| 708 |
+
"eval_f1": 0.9783132530120482,
|
| 709 |
+
"eval_loss": 0.5491310358047485,
|
| 710 |
+
"eval_pr_auc": 0.986877369995463,
|
| 711 |
+
"eval_precision": 0.9575471698113207,
|
| 712 |
+
"eval_recall": 1.0,
|
| 713 |
+
"eval_roc_auc": 0.8157635467980296,
|
| 714 |
+
"eval_runtime": 4.223,
|
| 715 |
+
"eval_samples_per_second": 50.438,
|
| 716 |
+
"eval_steps_per_second": 0.947,
|
| 717 |
+
"step": 116
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 29.285714285714285,
|
| 721 |
+
"grad_norm": 1.3928241729736328,
|
| 722 |
+
"learning_rate": 0.0004588719528532341,
|
| 723 |
+
"loss": 0.4332,
|
| 724 |
+
"step": 117
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 30.0,
|
| 728 |
+
"grad_norm": 0.13795076310634613,
|
| 729 |
+
"learning_rate": 0.00045666603926142283,
|
| 730 |
+
"loss": 0.3921,
|
| 731 |
+
"step": 120
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 30.0,
|
| 735 |
+
"eval_accuracy": 0.9624413145539906,
|
| 736 |
+
"eval_best_threshold": 0.3900836706161499,
|
| 737 |
+
"eval_f1": 0.9805825242718447,
|
| 738 |
+
"eval_loss": 0.5215495228767395,
|
| 739 |
+
"eval_pr_auc": 0.9885726401668996,
|
| 740 |
+
"eval_precision": 0.9665071770334929,
|
| 741 |
+
"eval_recall": 0.9950738916256158,
|
| 742 |
+
"eval_roc_auc": 0.8231527093596059,
|
| 743 |
+
"eval_runtime": 3.9438,
|
| 744 |
+
"eval_samples_per_second": 54.008,
|
| 745 |
+
"eval_steps_per_second": 1.014,
|
| 746 |
+
"step": 120
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 30.857142857142858,
|
| 750 |
+
"grad_norm": 1.1264947652816772,
|
| 751 |
+
"learning_rate": 0.0004544080985994258,
|
| 752 |
+
"loss": 0.4964,
|
| 753 |
+
"step": 123
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 31.0,
|
| 757 |
+
"eval_accuracy": 0.9624413145539906,
|
| 758 |
+
"eval_best_threshold": 0.06064055114984512,
|
| 759 |
+
"eval_f1": 0.9806763285024155,
|
| 760 |
+
"eval_loss": 0.5333446264266968,
|
| 761 |
+
"eval_pr_auc": 0.9872200029984016,
|
| 762 |
+
"eval_precision": 0.9620853080568721,
|
| 763 |
+
"eval_recall": 1.0,
|
| 764 |
+
"eval_roc_auc": 0.8251231527093597,
|
| 765 |
+
"eval_runtime": 4.334,
|
| 766 |
+
"eval_samples_per_second": 49.147,
|
| 767 |
+
"eval_steps_per_second": 0.923,
|
| 768 |
+
"step": 124
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 31.571428571428573,
|
| 772 |
+
"grad_norm": 1.147207260131836,
|
| 773 |
+
"learning_rate": 0.0004520986992917297,
|
| 774 |
+
"loss": 0.4592,
|
| 775 |
+
"step": 126
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 32.0,
|
| 779 |
+
"eval_accuracy": 0.9671361502347418,
|
| 780 |
+
"eval_best_threshold": 0.2888699173927307,
|
| 781 |
+
"eval_f1": 0.9829683698296837,
|
| 782 |
+
"eval_loss": 0.47561803460121155,
|
| 783 |
+
"eval_pr_auc": 0.98778851386531,
|
| 784 |
+
"eval_precision": 0.9711538461538461,
|
| 785 |
+
"eval_recall": 0.9950738916256158,
|
| 786 |
+
"eval_roc_auc": 0.8290640394088671,
|
| 787 |
+
"eval_runtime": 4.5973,
|
| 788 |
+
"eval_samples_per_second": 46.331,
|
| 789 |
+
"eval_steps_per_second": 0.87,
|
| 790 |
+
"step": 128
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 32.285714285714285,
|
| 794 |
+
"grad_norm": 0.5480005145072937,
|
| 795 |
+
"learning_rate": 0.00044973842271726027,
|
| 796 |
+
"loss": 0.4054,
|
| 797 |
+
"step": 129
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 33.0,
|
| 801 |
+
"grad_norm": 1.6971485614776611,
|
| 802 |
+
"learning_rate": 0.00044732786306302303,
|
| 803 |
+
"loss": 0.476,
|
| 804 |
+
"step": 132
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 33.0,
|
| 808 |
+
"eval_accuracy": 0.9671361502347418,
|
| 809 |
+
"eval_best_threshold": 0.2422952502965927,
|
| 810 |
+
"eval_f1": 0.9830508474576272,
|
| 811 |
+
"eval_loss": 0.47018811106681824,
|
| 812 |
+
"eval_pr_auc": 0.989969192055292,
|
| 813 |
+
"eval_precision": 0.9666666666666667,
|
| 814 |
+
"eval_recall": 1.0,
|
| 815 |
+
"eval_roc_auc": 0.8450738916256157,
|
| 816 |
+
"eval_runtime": 4.2741,
|
| 817 |
+
"eval_samples_per_second": 49.835,
|
| 818 |
+
"eval_steps_per_second": 0.936,
|
| 819 |
+
"step": 132
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 33.857142857142854,
|
| 823 |
+
"grad_norm": 1.4494688510894775,
|
| 824 |
+
"learning_rate": 0.0004448676271745197,
|
| 825 |
+
"loss": 0.5734,
|
| 826 |
+
"step": 135
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 34.0,
|
| 830 |
+
"eval_accuracy": 0.9671361502347418,
|
| 831 |
+
"eval_best_threshold": 0.22772686183452606,
|
| 832 |
+
"eval_f1": 0.9830508474576272,
|
| 833 |
+
"eval_loss": 0.47546666860580444,
|
| 834 |
+
"eval_pr_auc": 0.9894810176977142,
|
| 835 |
+
"eval_precision": 0.9666666666666667,
|
| 836 |
+
"eval_recall": 1.0,
|
| 837 |
+
"eval_roc_auc": 0.8379310344827586,
|
| 838 |
+
"eval_runtime": 4.3512,
|
| 839 |
+
"eval_samples_per_second": 48.952,
|
| 840 |
+
"eval_steps_per_second": 0.919,
|
| 841 |
+
"step": 136
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 34.57142857142857,
|
| 845 |
+
"grad_norm": 0.8203570246696472,
|
| 846 |
+
"learning_rate": 0.00044235833440297857,
|
| 847 |
+
"loss": 0.3077,
|
| 848 |
+
"step": 138
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 35.0,
|
| 852 |
+
"eval_accuracy": 0.9624413145539906,
|
| 853 |
+
"eval_best_threshold": 0.05028397589921951,
|
| 854 |
+
"eval_f1": 0.9806763285024155,
|
| 855 |
+
"eval_loss": 0.46481361985206604,
|
| 856 |
+
"eval_pr_auc": 0.9884215676732423,
|
| 857 |
+
"eval_precision": 0.9620853080568721,
|
| 858 |
+
"eval_recall": 1.0,
|
| 859 |
+
"eval_roc_auc": 0.8428571428571429,
|
| 860 |
+
"eval_runtime": 4.0747,
|
| 861 |
+
"eval_samples_per_second": 52.274,
|
| 862 |
+
"eval_steps_per_second": 0.982,
|
| 863 |
+
"step": 140
|
| 864 |
+
}
|
| 865 |
+
],
|
| 866 |
+
"logging_steps": 3,
|
| 867 |
+
"max_steps": 600,
|
| 868 |
+
"num_input_tokens_seen": 0,
|
| 869 |
+
"num_train_epochs": 200,
|
| 870 |
+
"save_steps": 500,
|
| 871 |
+
"stateful_callbacks": {
|
| 872 |
+
"EarlyStoppingCallback": {
|
| 873 |
+
"args": {
|
| 874 |
+
"early_stopping_patience": 20,
|
| 875 |
+
"early_stopping_threshold": 0.0
|
| 876 |
+
},
|
| 877 |
+
"attributes": {
|
| 878 |
+
"early_stopping_patience_counter": 20
|
| 879 |
+
}
|
| 880 |
+
},
|
| 881 |
+
"TrainerControl": {
|
| 882 |
+
"args": {
|
| 883 |
+
"should_epoch_stop": false,
|
| 884 |
+
"should_evaluate": false,
|
| 885 |
+
"should_log": false,
|
| 886 |
+
"should_save": true,
|
| 887 |
+
"should_training_stop": true
|
| 888 |
+
},
|
| 889 |
+
"attributes": {}
|
| 890 |
+
}
|
| 891 |
+
},
|
| 892 |
+
"total_flos": 1.0621405284091822e+19,
|
| 893 |
+
"train_batch_size": 128,
|
| 894 |
+
"trial_name": null,
|
| 895 |
+
"trial_params": null
|
| 896 |
+
}
|
checkpoint-140/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f603d30a49833e95de39ef9d1ed2771533f8d9b708df00b15be6b4ad228368bd
|
| 3 |
+
size 5368
|
checkpoint-60/config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-large-patch16-224-in21k",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 1024,
|
| 11 |
+
"image_size": 256,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 4096,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"model_type": "vit",
|
| 16 |
+
"num_attention_heads": 16,
|
| 17 |
+
"num_channels": 3,
|
| 18 |
+
"num_hidden_layers": 24,
|
| 19 |
+
"patch_size": 16,
|
| 20 |
+
"problem_type": "single_label_classification",
|
| 21 |
+
"qkv_bias": true,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.49.0"
|
| 24 |
+
}
|
checkpoint-60/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7603c376108527effe45eab5fca348a5c0335c2c3085408192150b8fbdd97a3
|
| 3 |
+
size 1213507024
|
checkpoint-60/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a91bc74d81e2604b56d12ab5149b6c8668a2b2d367b45e22218bc28b40009614
|
| 3 |
+
size 2427249066
|
checkpoint-60/preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": false,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTFeatureExtractor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"resample": 2,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-60/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88a8c6076f3c05cdf1125c1235b4a613ac1ac913542e9b26d9987d0d57db6272
|
| 3 |
+
size 14244
|
checkpoint-60/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7565f867f9c8d3f5c8aa2ea6bd211664218909ee6750a4ca50a8bb62aa0732f
|
| 3 |
+
size 988
|
checkpoint-60/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f881e0232dd1e145e78a55a6abf77ba5bec2e9fb9008686b6cf1fbde772fafb
|
| 3 |
+
size 1064
|
checkpoint-60/trainer_state.json
ADDED
|
@@ -0,0 +1,414 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9830508474576272,
|
| 3 |
+
"best_model_checkpoint": "/tmp/logs/GenetikaPlus/clf_soma_model_v1.0.1/checkpoint-60",
|
| 4 |
+
"epoch": 15.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 60,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2857142857142857,
|
| 13 |
+
"grad_norm": 3.202357053756714,
|
| 14 |
+
"learning_rate": 8.333333333333333e-05,
|
| 15 |
+
"loss": 0.6786,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.8571428571428571,
|
| 20 |
+
"grad_norm": 4.824620246887207,
|
| 21 |
+
"learning_rate": 0.00025,
|
| 22 |
+
"loss": 0.7317,
|
| 23 |
+
"step": 3
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_accuracy": 0.9577464788732394,
|
| 28 |
+
"eval_best_threshold": 0.3037576377391815,
|
| 29 |
+
"eval_f1": 0.9783132530120482,
|
| 30 |
+
"eval_loss": 0.6482536792755127,
|
| 31 |
+
"eval_pr_auc": 0.9798696623714835,
|
| 32 |
+
"eval_precision": 0.9575471698113207,
|
| 33 |
+
"eval_recall": 1.0,
|
| 34 |
+
"eval_roc_auc": 0.7300492610837439,
|
| 35 |
+
"eval_runtime": 3.6503,
|
| 36 |
+
"eval_samples_per_second": 58.351,
|
| 37 |
+
"eval_steps_per_second": 1.096,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.5714285714285714,
|
| 42 |
+
"grad_norm": 8.443906784057617,
|
| 43 |
+
"learning_rate": 0.0005,
|
| 44 |
+
"loss": 0.9287,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.0,
|
| 49 |
+
"eval_accuracy": 0.9577464788732394,
|
| 50 |
+
"eval_best_threshold": 0.09376094490289688,
|
| 51 |
+
"eval_f1": 0.9783132530120482,
|
| 52 |
+
"eval_loss": 1.2041555643081665,
|
| 53 |
+
"eval_pr_auc": 0.9823049604214897,
|
| 54 |
+
"eval_precision": 0.9575471698113207,
|
| 55 |
+
"eval_recall": 1.0,
|
| 56 |
+
"eval_roc_auc": 0.7192118226600985,
|
| 57 |
+
"eval_runtime": 3.8529,
|
| 58 |
+
"eval_samples_per_second": 55.283,
|
| 59 |
+
"eval_steps_per_second": 1.038,
|
| 60 |
+
"step": 8
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 2.2857142857142856,
|
| 64 |
+
"grad_norm": 7.837808609008789,
|
| 65 |
+
"learning_rate": 0.0004999685319184688,
|
| 66 |
+
"loss": 1.0591,
|
| 67 |
+
"step": 9
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 3.0,
|
| 71 |
+
"grad_norm": 0.33617040514945984,
|
| 72 |
+
"learning_rate": 0.0004998741355957963,
|
| 73 |
+
"loss": 0.6429,
|
| 74 |
+
"step": 12
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 3.0,
|
| 78 |
+
"eval_accuracy": 0.9530516431924883,
|
| 79 |
+
"eval_best_threshold": 0.11066323518753052,
|
| 80 |
+
"eval_f1": 0.9759615384615384,
|
| 81 |
+
"eval_loss": 0.6775969862937927,
|
| 82 |
+
"eval_pr_auc": 0.9570451344092105,
|
| 83 |
+
"eval_precision": 0.9530516431924883,
|
| 84 |
+
"eval_recall": 1.0,
|
| 85 |
+
"eval_roc_auc": 0.4647783251231527,
|
| 86 |
+
"eval_runtime": 4.3875,
|
| 87 |
+
"eval_samples_per_second": 48.547,
|
| 88 |
+
"eval_steps_per_second": 0.912,
|
| 89 |
+
"step": 12
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 3.857142857142857,
|
| 93 |
+
"grad_norm": 2.0180256366729736,
|
| 94 |
+
"learning_rate": 0.000499716834795752,
|
| 95 |
+
"loss": 0.7296,
|
| 96 |
+
"step": 15
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 4.0,
|
| 100 |
+
"eval_accuracy": 0.9577464788732394,
|
| 101 |
+
"eval_best_threshold": 0.488298624753952,
|
| 102 |
+
"eval_f1": 0.9783132530120482,
|
| 103 |
+
"eval_loss": 0.6953739523887634,
|
| 104 |
+
"eval_pr_auc": 0.9387681415619215,
|
| 105 |
+
"eval_precision": 0.9575471698113207,
|
| 106 |
+
"eval_recall": 1.0,
|
| 107 |
+
"eval_roc_auc": 0.3980295566502463,
|
| 108 |
+
"eval_runtime": 3.9434,
|
| 109 |
+
"eval_samples_per_second": 54.014,
|
| 110 |
+
"eval_steps_per_second": 1.014,
|
| 111 |
+
"step": 16
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 4.571428571428571,
|
| 115 |
+
"grad_norm": 0.6065838932991028,
|
| 116 |
+
"learning_rate": 0.0004994966691179711,
|
| 117 |
+
"loss": 0.5691,
|
| 118 |
+
"step": 18
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 5.0,
|
| 122 |
+
"eval_accuracy": 0.9577464788732394,
|
| 123 |
+
"eval_best_threshold": 0.4559290409088135,
|
| 124 |
+
"eval_f1": 0.9783132530120482,
|
| 125 |
+
"eval_loss": 0.6963727474212646,
|
| 126 |
+
"eval_pr_auc": 0.9437271902417405,
|
| 127 |
+
"eval_precision": 0.9575471698113207,
|
| 128 |
+
"eval_recall": 1.0,
|
| 129 |
+
"eval_roc_auc": 0.42167487684729066,
|
| 130 |
+
"eval_runtime": 4.451,
|
| 131 |
+
"eval_samples_per_second": 47.854,
|
| 132 |
+
"eval_steps_per_second": 0.899,
|
| 133 |
+
"step": 20
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 5.285714285714286,
|
| 137 |
+
"grad_norm": 0.7243921160697937,
|
| 138 |
+
"learning_rate": 0.0004992136939879857,
|
| 139 |
+
"loss": 0.5782,
|
| 140 |
+
"step": 21
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 6.0,
|
| 144 |
+
"grad_norm": 0.2275872528553009,
|
| 145 |
+
"learning_rate": 0.0004988679806432712,
|
| 146 |
+
"loss": 0.5628,
|
| 147 |
+
"step": 24
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 6.0,
|
| 151 |
+
"eval_accuracy": 0.9624413145539906,
|
| 152 |
+
"eval_best_threshold": 0.40434572100639343,
|
| 153 |
+
"eval_f1": 0.9806763285024155,
|
| 154 |
+
"eval_loss": 0.6847261786460876,
|
| 155 |
+
"eval_pr_auc": 0.9548760246480454,
|
| 156 |
+
"eval_precision": 0.9620853080568721,
|
| 157 |
+
"eval_recall": 1.0,
|
| 158 |
+
"eval_roc_auc": 0.5586206896551724,
|
| 159 |
+
"eval_runtime": 4.6725,
|
| 160 |
+
"eval_samples_per_second": 45.586,
|
| 161 |
+
"eval_steps_per_second": 0.856,
|
| 162 |
+
"step": 24
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 6.857142857142857,
|
| 166 |
+
"grad_norm": 1.0703548192977905,
|
| 167 |
+
"learning_rate": 0.0004984596161153135,
|
| 168 |
+
"loss": 0.6532,
|
| 169 |
+
"step": 27
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 7.0,
|
| 173 |
+
"eval_accuracy": 0.9624413145539906,
|
| 174 |
+
"eval_best_threshold": 0.3936326503753662,
|
| 175 |
+
"eval_f1": 0.9805825242718447,
|
| 176 |
+
"eval_loss": 0.6566980481147766,
|
| 177 |
+
"eval_pr_auc": 0.9620782247005666,
|
| 178 |
+
"eval_precision": 0.9665071770334929,
|
| 179 |
+
"eval_recall": 0.9950738916256158,
|
| 180 |
+
"eval_roc_auc": 0.5571428571428572,
|
| 181 |
+
"eval_runtime": 4.3448,
|
| 182 |
+
"eval_samples_per_second": 49.024,
|
| 183 |
+
"eval_steps_per_second": 0.921,
|
| 184 |
+
"step": 28
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 7.571428571428571,
|
| 188 |
+
"grad_norm": 2.041855573654175,
|
| 189 |
+
"learning_rate": 0.0004979887032076989,
|
| 190 |
+
"loss": 0.5593,
|
| 191 |
+
"step": 30
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 8.0,
|
| 195 |
+
"eval_accuracy": 0.9624413145539906,
|
| 196 |
+
"eval_best_threshold": 0.2755194306373596,
|
| 197 |
+
"eval_f1": 0.9806763285024155,
|
| 198 |
+
"eval_loss": 0.6331985592842102,
|
| 199 |
+
"eval_pr_auc": 0.9751121096020542,
|
| 200 |
+
"eval_precision": 0.9620853080568721,
|
| 201 |
+
"eval_recall": 1.0,
|
| 202 |
+
"eval_roc_auc": 0.6586206896551724,
|
| 203 |
+
"eval_runtime": 4.7804,
|
| 204 |
+
"eval_samples_per_second": 44.557,
|
| 205 |
+
"eval_steps_per_second": 0.837,
|
| 206 |
+
"step": 32
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 8.285714285714286,
|
| 210 |
+
"grad_norm": 0.8734021186828613,
|
| 211 |
+
"learning_rate": 0.0004974553604702333,
|
| 212 |
+
"loss": 0.5002,
|
| 213 |
+
"step": 33
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 9.0,
|
| 217 |
+
"grad_norm": Infinity,
|
| 218 |
+
"learning_rate": 0.0004970651810649666,
|
| 219 |
+
"loss": 0.5346,
|
| 220 |
+
"step": 36
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 9.0,
|
| 224 |
+
"eval_accuracy": 0.9577464788732394,
|
| 225 |
+
"eval_best_threshold": 0.14823205769062042,
|
| 226 |
+
"eval_f1": 0.9783132530120482,
|
| 227 |
+
"eval_loss": 0.6035900712013245,
|
| 228 |
+
"eval_pr_auc": 0.9819011692535515,
|
| 229 |
+
"eval_precision": 0.9575471698113207,
|
| 230 |
+
"eval_recall": 1.0,
|
| 231 |
+
"eval_roc_auc": 0.7339901477832512,
|
| 232 |
+
"eval_runtime": 4.0039,
|
| 233 |
+
"eval_samples_per_second": 53.198,
|
| 234 |
+
"eval_steps_per_second": 0.999,
|
| 235 |
+
"step": 36
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 9.857142857142858,
|
| 239 |
+
"grad_norm": 1.1991403102874756,
|
| 240 |
+
"learning_rate": 0.0004964280947263676,
|
| 241 |
+
"loss": 0.5955,
|
| 242 |
+
"step": 39
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 10.0,
|
| 246 |
+
"eval_accuracy": 0.9624413145539906,
|
| 247 |
+
"eval_best_threshold": 0.5386950373649597,
|
| 248 |
+
"eval_f1": 0.9805825242718447,
|
| 249 |
+
"eval_loss": 0.6214358806610107,
|
| 250 |
+
"eval_pr_auc": 0.9624581599102329,
|
| 251 |
+
"eval_precision": 0.9665071770334929,
|
| 252 |
+
"eval_recall": 0.9950738916256158,
|
| 253 |
+
"eval_roc_auc": 0.6002463054187192,
|
| 254 |
+
"eval_runtime": 4.5609,
|
| 255 |
+
"eval_samples_per_second": 46.701,
|
| 256 |
+
"eval_steps_per_second": 0.877,
|
| 257 |
+
"step": 40
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 10.571428571428571,
|
| 261 |
+
"grad_norm": 2.6938207149505615,
|
| 262 |
+
"learning_rate": 0.0004957289714327572,
|
| 263 |
+
"loss": 0.4496,
|
| 264 |
+
"step": 42
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 11.0,
|
| 268 |
+
"eval_accuracy": 0.9530516431924883,
|
| 269 |
+
"eval_best_threshold": 0.04157886654138565,
|
| 270 |
+
"eval_f1": 0.9759615384615384,
|
| 271 |
+
"eval_loss": 0.8543573617935181,
|
| 272 |
+
"eval_pr_auc": 0.9879501355451374,
|
| 273 |
+
"eval_precision": 0.9530516431924883,
|
| 274 |
+
"eval_recall": 1.0,
|
| 275 |
+
"eval_roc_auc": 0.7901477832512316,
|
| 276 |
+
"eval_runtime": 4.6335,
|
| 277 |
+
"eval_samples_per_second": 45.969,
|
| 278 |
+
"eval_steps_per_second": 0.863,
|
| 279 |
+
"step": 44
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 11.285714285714286,
|
| 283 |
+
"grad_norm": 3.554097890853882,
|
| 284 |
+
"learning_rate": 0.0004949679871846857,
|
| 285 |
+
"loss": 0.6118,
|
| 286 |
+
"step": 45
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 12.0,
|
| 290 |
+
"grad_norm": 2.869584083557129,
|
| 291 |
+
"learning_rate": 0.0004941453335558681,
|
| 292 |
+
"loss": 0.6877,
|
| 293 |
+
"step": 48
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 12.0,
|
| 297 |
+
"eval_accuracy": 0.9530516431924883,
|
| 298 |
+
"eval_best_threshold": 0.17342814803123474,
|
| 299 |
+
"eval_f1": 0.9759615384615384,
|
| 300 |
+
"eval_loss": 0.5579107403755188,
|
| 301 |
+
"eval_pr_auc": 0.9873731974164195,
|
| 302 |
+
"eval_precision": 0.9530516431924883,
|
| 303 |
+
"eval_recall": 1.0,
|
| 304 |
+
"eval_roc_auc": 0.7891625615763547,
|
| 305 |
+
"eval_runtime": 4.1544,
|
| 306 |
+
"eval_samples_per_second": 51.271,
|
| 307 |
+
"eval_steps_per_second": 0.963,
|
| 308 |
+
"step": 48
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 12.857142857142858,
|
| 312 |
+
"grad_norm": 1.1457290649414062,
|
| 313 |
+
"learning_rate": 0.0004932612176449559,
|
| 314 |
+
"loss": 0.5374,
|
| 315 |
+
"step": 51
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 13.0,
|
| 319 |
+
"eval_accuracy": 0.9624413145539906,
|
| 320 |
+
"eval_best_threshold": 0.5808722376823425,
|
| 321 |
+
"eval_f1": 0.9806763285024155,
|
| 322 |
+
"eval_loss": 0.5983880758285522,
|
| 323 |
+
"eval_pr_auc": 0.9837750859728142,
|
| 324 |
+
"eval_precision": 0.9620853080568721,
|
| 325 |
+
"eval_recall": 1.0,
|
| 326 |
+
"eval_roc_auc": 0.7578817733990147,
|
| 327 |
+
"eval_runtime": 4.6879,
|
| 328 |
+
"eval_samples_per_second": 45.436,
|
| 329 |
+
"eval_steps_per_second": 0.853,
|
| 330 |
+
"step": 52
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 13.571428571428571,
|
| 334 |
+
"grad_norm": 2.2151918411254883,
|
| 335 |
+
"learning_rate": 0.000492315862023402,
|
| 336 |
+
"loss": 0.5329,
|
| 337 |
+
"step": 54
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 14.0,
|
| 341 |
+
"eval_accuracy": 0.9624413145539906,
|
| 342 |
+
"eval_best_threshold": 0.1792430430650711,
|
| 343 |
+
"eval_f1": 0.9806763285024155,
|
| 344 |
+
"eval_loss": 0.5941116213798523,
|
| 345 |
+
"eval_pr_auc": 0.987113937851267,
|
| 346 |
+
"eval_precision": 0.9620853080568721,
|
| 347 |
+
"eval_recall": 1.0,
|
| 348 |
+
"eval_roc_auc": 0.7970443349753695,
|
| 349 |
+
"eval_runtime": 4.3199,
|
| 350 |
+
"eval_samples_per_second": 49.306,
|
| 351 |
+
"eval_steps_per_second": 0.926,
|
| 352 |
+
"step": 56
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 14.285714285714286,
|
| 356 |
+
"grad_norm": 2.5402119159698486,
|
| 357 |
+
"learning_rate": 0.0004913095046794281,
|
| 358 |
+
"loss": 0.4644,
|
| 359 |
+
"step": 57
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 15.0,
|
| 363 |
+
"grad_norm": 2.0666849613189697,
|
| 364 |
+
"learning_rate": 0.0004902423989581143,
|
| 365 |
+
"loss": 0.5655,
|
| 366 |
+
"step": 60
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 15.0,
|
| 370 |
+
"eval_accuracy": 0.9671361502347418,
|
| 371 |
+
"eval_best_threshold": 0.09999631345272064,
|
| 372 |
+
"eval_f1": 0.9830508474576272,
|
| 373 |
+
"eval_loss": 0.6284329891204834,
|
| 374 |
+
"eval_pr_auc": 0.9873903178253213,
|
| 375 |
+
"eval_precision": 0.9666666666666667,
|
| 376 |
+
"eval_recall": 1.0,
|
| 377 |
+
"eval_roc_auc": 0.8014778325123153,
|
| 378 |
+
"eval_runtime": 3.9163,
|
| 379 |
+
"eval_samples_per_second": 54.388,
|
| 380 |
+
"eval_steps_per_second": 1.021,
|
| 381 |
+
"step": 60
|
| 382 |
+
}
|
| 383 |
+
],
|
| 384 |
+
"logging_steps": 3,
|
| 385 |
+
"max_steps": 600,
|
| 386 |
+
"num_input_tokens_seen": 0,
|
| 387 |
+
"num_train_epochs": 200,
|
| 388 |
+
"save_steps": 500,
|
| 389 |
+
"stateful_callbacks": {
|
| 390 |
+
"EarlyStoppingCallback": {
|
| 391 |
+
"args": {
|
| 392 |
+
"early_stopping_patience": 20,
|
| 393 |
+
"early_stopping_threshold": 0.0
|
| 394 |
+
},
|
| 395 |
+
"attributes": {
|
| 396 |
+
"early_stopping_patience_counter": 0
|
| 397 |
+
}
|
| 398 |
+
},
|
| 399 |
+
"TrainerControl": {
|
| 400 |
+
"args": {
|
| 401 |
+
"should_epoch_stop": false,
|
| 402 |
+
"should_evaluate": false,
|
| 403 |
+
"should_log": false,
|
| 404 |
+
"should_save": true,
|
| 405 |
+
"should_training_stop": false
|
| 406 |
+
},
|
| 407 |
+
"attributes": {}
|
| 408 |
+
}
|
| 409 |
+
},
|
| 410 |
+
"total_flos": 4.5520308360393523e+18,
|
| 411 |
+
"train_batch_size": 128,
|
| 412 |
+
"trial_name": null,
|
| 413 |
+
"trial_params": null
|
| 414 |
+
}
|
checkpoint-60/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f603d30a49833e95de39ef9d1ed2771533f8d9b708df00b15be6b4ad228368bd
|
| 3 |
+
size 5368
|
config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/vit-large-patch16-224-in21k",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ViTForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"encoder_stride": 16,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.0,
|
| 10 |
+
"hidden_size": 1024,
|
| 11 |
+
"image_size": 256,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 4096,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"model_type": "vit",
|
| 16 |
+
"num_attention_heads": 16,
|
| 17 |
+
"num_channels": 3,
|
| 18 |
+
"num_hidden_layers": 24,
|
| 19 |
+
"patch_size": 16,
|
| 20 |
+
"problem_type": "single_label_classification",
|
| 21 |
+
"qkv_bias": true,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.49.0"
|
| 24 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7de864ac2885746ef4be671e778f1df03aee0141c6847dbba7da6de34864c462
|
| 3 |
+
size 1213507024
|
pr_curve.png
ADDED
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": false,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTFeatureExtractor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"resample": 2,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 224,
|
| 21 |
+
"width": 224
|
| 22 |
+
}
|
| 23 |
+
}
|
roc_curve.png
ADDED
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f603d30a49833e95de39ef9d1ed2771533f8d9b708df00b15be6b4ad228368bd
|
| 3 |
+
size 5368
|