Commit
·
061b46a
1
Parent(s):
e79b407
Add evaluation results and output files to model card
Browse files- README.md +36 -0
- checkpoint-84/config.json +23 -0
- checkpoint-84/model.safetensors +3 -0
- checkpoint-84/optimizer.pt +3 -0
- checkpoint-84/preprocessor_config.json +23 -0
- checkpoint-84/rng_state.pth +3 -0
- checkpoint-84/scheduler.pt +3 -0
- checkpoint-84/trainer_state.json +840 -0
- checkpoint-84/training_args.bin +3 -0
- config.json +23 -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: 0.50
|
| 9 |
+
- roc_auc: 0.37
|
| 10 |
+
- best threshold according to F1: 0.03
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Binary Classification Model
|
| 14 |
+
|
| 15 |
+
## Evaluation Results
|
| 16 |
+
|
| 17 |
+
**Average Precision:** 0.50
|
| 18 |
+
**ROC AUC:** 0.37
|
| 19 |
+
**best threshold according to F1: 0.03
|
| 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-84/config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "GenetikaPlus/binary_classification_model_v3.1.5_Junction",
|
| 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": 256,
|
| 11 |
+
"image_size": 32,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 512,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"model_type": "vit",
|
| 16 |
+
"num_attention_heads": 8,
|
| 17 |
+
"num_channels": 3,
|
| 18 |
+
"num_hidden_layers": 4,
|
| 19 |
+
"patch_size": 4,
|
| 20 |
+
"qkv_bias": true,
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.48.0"
|
| 23 |
+
}
|
checkpoint-84/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eeeac60794353121bf077b31d3539872cd5f5fd95e5f488b53128f4dd2804954
|
| 3 |
+
size 8563512
|
checkpoint-84/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:463825b38620a46cd1ae415bb27541784040a351411f83c6f7792840c246b3da
|
| 3 |
+
size 17171514
|
checkpoint-84/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": 32,
|
| 21 |
+
"width": 32
|
| 22 |
+
}
|
| 23 |
+
}
|
checkpoint-84/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:965d5995676d59a44b47d9ced162417cf72cc56567ea7643d7e0802dcfc82c43
|
| 3 |
+
size 14244
|
checkpoint-84/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3370cf7c7ef59896f4e4a55bcb471682405d2132b95a209aac680ae21fcb19b1
|
| 3 |
+
size 1000
|
checkpoint-84/trainer_state.json
ADDED
|
@@ -0,0 +1,840 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7708737864077669,
|
| 3 |
+
"best_model_checkpoint": "/tmp/logs/binary_classification_model_v3.1.5_Junction_batch1/checkpoint-84",
|
| 4 |
+
"epoch": 14.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 84,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.16666666666666666,
|
| 13 |
+
"grad_norm": 736285.125,
|
| 14 |
+
"learning_rate": 0.00016666666666666666,
|
| 15 |
+
"loss": 0.7654,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.3333333333333333,
|
| 20 |
+
"grad_norm": 1333572.375,
|
| 21 |
+
"learning_rate": 0.0003333333333333333,
|
| 22 |
+
"loss": 0.8052,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.5,
|
| 27 |
+
"grad_norm": 1624097.125,
|
| 28 |
+
"learning_rate": 0.0005,
|
| 29 |
+
"loss": 2.2863,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.6666666666666666,
|
| 34 |
+
"grad_norm": 671227.0625,
|
| 35 |
+
"learning_rate": 0.0006666666666666666,
|
| 36 |
+
"loss": 0.9079,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.8333333333333334,
|
| 41 |
+
"grad_norm": 545356.625,
|
| 42 |
+
"learning_rate": 0.0008333333333333333,
|
| 43 |
+
"loss": 0.958,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.0,
|
| 48 |
+
"grad_norm": 413819.90625,
|
| 49 |
+
"learning_rate": 0.001,
|
| 50 |
+
"loss": 0.856,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.0,
|
| 55 |
+
"eval_accuracy": 0.5584881068752037,
|
| 56 |
+
"eval_best_threshold": 0.5805500745773315,
|
| 57 |
+
"eval_f1": 0.7138331573389651,
|
| 58 |
+
"eval_loss": 0.6886702179908752,
|
| 59 |
+
"eval_pr_auc": 0.6352806545912864,
|
| 60 |
+
"eval_precision": 0.556470200856108,
|
| 61 |
+
"eval_recall": 0.9952885747938751,
|
| 62 |
+
"eval_roc_auc": 0.6032108397144622,
|
| 63 |
+
"eval_runtime": 2.3315,
|
| 64 |
+
"eval_samples_per_second": 1316.324,
|
| 65 |
+
"eval_steps_per_second": 2.573,
|
| 66 |
+
"step": 6
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.1666666666666667,
|
| 70 |
+
"grad_norm": 38779.44921875,
|
| 71 |
+
"learning_rate": 0.0011666666666666668,
|
| 72 |
+
"loss": 0.6853,
|
| 73 |
+
"step": 7
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 1.3333333333333333,
|
| 77 |
+
"grad_norm": 471674.8125,
|
| 78 |
+
"learning_rate": 0.0013333333333333333,
|
| 79 |
+
"loss": 0.8944,
|
| 80 |
+
"step": 8
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.5,
|
| 84 |
+
"grad_norm": 534391.0625,
|
| 85 |
+
"learning_rate": 0.0015,
|
| 86 |
+
"loss": 0.9852,
|
| 87 |
+
"step": 9
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.6666666666666665,
|
| 91 |
+
"grad_norm": 454181.25,
|
| 92 |
+
"learning_rate": 0.0016666666666666666,
|
| 93 |
+
"loss": 0.8852,
|
| 94 |
+
"step": 10
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.8333333333333335,
|
| 98 |
+
"grad_norm": 191708.96875,
|
| 99 |
+
"learning_rate": 0.0018333333333333333,
|
| 100 |
+
"loss": 0.7181,
|
| 101 |
+
"step": 11
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.0,
|
| 105 |
+
"grad_norm": 138568.640625,
|
| 106 |
+
"learning_rate": 0.002,
|
| 107 |
+
"loss": 0.7003,
|
| 108 |
+
"step": 12
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.0,
|
| 112 |
+
"eval_accuracy": 0.5897686542847833,
|
| 113 |
+
"eval_best_threshold": 0.7087586522102356,
|
| 114 |
+
"eval_f1": 0.7176496972415339,
|
| 115 |
+
"eval_loss": 0.7489995956420898,
|
| 116 |
+
"eval_pr_auc": 0.6418343639245951,
|
| 117 |
+
"eval_precision": 0.5795001810938066,
|
| 118 |
+
"eval_recall": 0.9422850412249706,
|
| 119 |
+
"eval_roc_auc": 0.6285457040032509,
|
| 120 |
+
"eval_runtime": 2.4193,
|
| 121 |
+
"eval_samples_per_second": 1268.569,
|
| 122 |
+
"eval_steps_per_second": 2.48,
|
| 123 |
+
"step": 12
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.1666666666666665,
|
| 127 |
+
"grad_norm": 241500.28125,
|
| 128 |
+
"learning_rate": 0.002166666666666667,
|
| 129 |
+
"loss": 0.7381,
|
| 130 |
+
"step": 13
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.3333333333333335,
|
| 134 |
+
"grad_norm": 210473.890625,
|
| 135 |
+
"learning_rate": 0.0023333333333333335,
|
| 136 |
+
"loss": 0.7209,
|
| 137 |
+
"step": 14
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.5,
|
| 141 |
+
"grad_norm": 159988.078125,
|
| 142 |
+
"learning_rate": 0.0025,
|
| 143 |
+
"loss": 0.7087,
|
| 144 |
+
"step": 15
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.6666666666666665,
|
| 148 |
+
"grad_norm": 81561.6875,
|
| 149 |
+
"learning_rate": 0.0026666666666666666,
|
| 150 |
+
"loss": 0.685,
|
| 151 |
+
"step": 16
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.8333333333333335,
|
| 155 |
+
"grad_norm": 116087.46875,
|
| 156 |
+
"learning_rate": 0.002833333333333333,
|
| 157 |
+
"loss": 0.6949,
|
| 158 |
+
"step": 17
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 3.0,
|
| 162 |
+
"grad_norm": 89483.1953125,
|
| 163 |
+
"learning_rate": 0.003,
|
| 164 |
+
"loss": 0.6868,
|
| 165 |
+
"step": 18
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 3.0,
|
| 169 |
+
"eval_accuracy": 0.6865428478331704,
|
| 170 |
+
"eval_best_threshold": 0.5599992275238037,
|
| 171 |
+
"eval_f1": 0.7575604838709677,
|
| 172 |
+
"eval_loss": 0.6744044423103333,
|
| 173 |
+
"eval_pr_auc": 0.6810929123667648,
|
| 174 |
+
"eval_precision": 0.6621145374449339,
|
| 175 |
+
"eval_recall": 0.8851590106007067,
|
| 176 |
+
"eval_roc_auc": 0.6815071405927426,
|
| 177 |
+
"eval_runtime": 2.5065,
|
| 178 |
+
"eval_samples_per_second": 1224.396,
|
| 179 |
+
"eval_steps_per_second": 2.394,
|
| 180 |
+
"step": 18
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 3.1666666666666665,
|
| 184 |
+
"grad_norm": 6684.7294921875,
|
| 185 |
+
"learning_rate": 0.0031666666666666666,
|
| 186 |
+
"loss": 0.6702,
|
| 187 |
+
"step": 19
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 3.3333333333333335,
|
| 191 |
+
"grad_norm": 209005.765625,
|
| 192 |
+
"learning_rate": 0.003333333333333333,
|
| 193 |
+
"loss": 0.7009,
|
| 194 |
+
"step": 20
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 3.5,
|
| 198 |
+
"grad_norm": 323864.125,
|
| 199 |
+
"learning_rate": 0.0034999999999999996,
|
| 200 |
+
"loss": 0.7666,
|
| 201 |
+
"step": 21
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 3.6666666666666665,
|
| 205 |
+
"grad_norm": 307526.84375,
|
| 206 |
+
"learning_rate": 0.0036666666666666666,
|
| 207 |
+
"loss": 0.7608,
|
| 208 |
+
"step": 22
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 3.8333333333333335,
|
| 212 |
+
"grad_norm": 156517.046875,
|
| 213 |
+
"learning_rate": 0.0038333333333333336,
|
| 214 |
+
"loss": 0.6678,
|
| 215 |
+
"step": 23
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 4.0,
|
| 219 |
+
"grad_norm": 174099.984375,
|
| 220 |
+
"learning_rate": 0.004,
|
| 221 |
+
"loss": 0.6636,
|
| 222 |
+
"step": 24
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 4.0,
|
| 226 |
+
"eval_accuracy": 0.6881720430107527,
|
| 227 |
+
"eval_best_threshold": 0.30409353971481323,
|
| 228 |
+
"eval_f1": 0.7567979669631512,
|
| 229 |
+
"eval_loss": 0.7221760749816895,
|
| 230 |
+
"eval_pr_auc": 0.6879511756542679,
|
| 231 |
+
"eval_precision": 0.6656236030397854,
|
| 232 |
+
"eval_recall": 0.8769140164899882,
|
| 233 |
+
"eval_roc_auc": 0.6760302376589268,
|
| 234 |
+
"eval_runtime": 2.5115,
|
| 235 |
+
"eval_samples_per_second": 1221.98,
|
| 236 |
+
"eval_steps_per_second": 2.389,
|
| 237 |
+
"step": 24
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 4.166666666666667,
|
| 241 |
+
"grad_norm": 282579.03125,
|
| 242 |
+
"learning_rate": 0.004166666666666667,
|
| 243 |
+
"loss": 0.7214,
|
| 244 |
+
"step": 25
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 4.333333333333333,
|
| 248 |
+
"grad_norm": 263304.78125,
|
| 249 |
+
"learning_rate": 0.004333333333333334,
|
| 250 |
+
"loss": 0.7127,
|
| 251 |
+
"step": 26
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 4.5,
|
| 255 |
+
"grad_norm": 71898.953125,
|
| 256 |
+
"learning_rate": 0.0045000000000000005,
|
| 257 |
+
"loss": 0.6439,
|
| 258 |
+
"step": 27
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 4.666666666666667,
|
| 262 |
+
"grad_norm": 267385.34375,
|
| 263 |
+
"learning_rate": 0.004666666666666667,
|
| 264 |
+
"loss": 0.7345,
|
| 265 |
+
"step": 28
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 4.833333333333333,
|
| 269 |
+
"grad_norm": 274028.96875,
|
| 270 |
+
"learning_rate": 0.004833333333333334,
|
| 271 |
+
"loss": 0.7395,
|
| 272 |
+
"step": 29
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 5.0,
|
| 276 |
+
"grad_norm": 176626.65625,
|
| 277 |
+
"learning_rate": 0.005,
|
| 278 |
+
"loss": 0.6697,
|
| 279 |
+
"step": 30
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 5.0,
|
| 283 |
+
"eval_accuracy": 0.6924079504724666,
|
| 284 |
+
"eval_best_threshold": 0.5539383292198181,
|
| 285 |
+
"eval_f1": 0.7580727831881087,
|
| 286 |
+
"eval_loss": 0.6356029510498047,
|
| 287 |
+
"eval_pr_auc": 0.6910960136731504,
|
| 288 |
+
"eval_precision": 0.6710526315789473,
|
| 289 |
+
"eval_recall": 0.8710247349823321,
|
| 290 |
+
"eval_roc_auc": 0.6834678718430487,
|
| 291 |
+
"eval_runtime": 2.4937,
|
| 292 |
+
"eval_samples_per_second": 1230.682,
|
| 293 |
+
"eval_steps_per_second": 2.406,
|
| 294 |
+
"step": 30
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 5.166666666666667,
|
| 298 |
+
"grad_norm": 14253.6220703125,
|
| 299 |
+
"learning_rate": 0.004999830770009406,
|
| 300 |
+
"loss": 0.6372,
|
| 301 |
+
"step": 31
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 5.333333333333333,
|
| 305 |
+
"grad_norm": 66843.328125,
|
| 306 |
+
"learning_rate": 0.004999323102948654,
|
| 307 |
+
"loss": 0.6313,
|
| 308 |
+
"step": 32
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 5.5,
|
| 312 |
+
"grad_norm": 15686.796875,
|
| 313 |
+
"learning_rate": 0.004998477067547739,
|
| 314 |
+
"loss": 0.6291,
|
| 315 |
+
"step": 33
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 5.666666666666667,
|
| 319 |
+
"grad_norm": 13507.8310546875,
|
| 320 |
+
"learning_rate": 0.004997292778346311,
|
| 321 |
+
"loss": 0.6279,
|
| 322 |
+
"step": 34
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 5.833333333333333,
|
| 326 |
+
"grad_norm": 133211.515625,
|
| 327 |
+
"learning_rate": 0.004995770395678171,
|
| 328 |
+
"loss": 0.6277,
|
| 329 |
+
"step": 35
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 6.0,
|
| 333 |
+
"grad_norm": 85637.015625,
|
| 334 |
+
"learning_rate": 0.004993910125649561,
|
| 335 |
+
"loss": 0.62,
|
| 336 |
+
"step": 36
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 6.0,
|
| 340 |
+
"eval_accuracy": 0.6810035842293907,
|
| 341 |
+
"eval_best_threshold": 0.5501854419708252,
|
| 342 |
+
"eval_f1": 0.7665156212735511,
|
| 343 |
+
"eval_loss": 0.638248860836029,
|
| 344 |
+
"eval_pr_auc": 0.6844694185868021,
|
| 345 |
+
"eval_precision": 0.6440881763527054,
|
| 346 |
+
"eval_recall": 0.9464075382803298,
|
| 347 |
+
"eval_roc_auc": 0.6936909944251571,
|
| 348 |
+
"eval_runtime": 2.495,
|
| 349 |
+
"eval_samples_per_second": 1230.046,
|
| 350 |
+
"eval_steps_per_second": 2.405,
|
| 351 |
+
"step": 36
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 6.166666666666667,
|
| 355 |
+
"grad_norm": 157243.15625,
|
| 356 |
+
"learning_rate": 0.004991712220111265,
|
| 357 |
+
"loss": 0.6502,
|
| 358 |
+
"step": 37
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 6.333333333333333,
|
| 362 |
+
"grad_norm": 164708.90625,
|
| 363 |
+
"learning_rate": 0.004989176976624511,
|
| 364 |
+
"loss": 0.645,
|
| 365 |
+
"step": 38
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 6.5,
|
| 369 |
+
"grad_norm": 26995.654296875,
|
| 370 |
+
"learning_rate": 0.0049863047384206835,
|
| 371 |
+
"loss": 0.6175,
|
| 372 |
+
"step": 39
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 6.666666666666667,
|
| 376 |
+
"grad_norm": 27075.607421875,
|
| 377 |
+
"learning_rate": 0.004983095894354857,
|
| 378 |
+
"loss": 0.5887,
|
| 379 |
+
"step": 40
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 6.833333333333333,
|
| 383 |
+
"grad_norm": 99286.6328125,
|
| 384 |
+
"learning_rate": 0.0049795508788531544,
|
| 385 |
+
"loss": 0.5984,
|
| 386 |
+
"step": 41
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 7.0,
|
| 390 |
+
"grad_norm": 42803.4375,
|
| 391 |
+
"learning_rate": 0.004975670171853926,
|
| 392 |
+
"loss": 0.6223,
|
| 393 |
+
"step": 42
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 7.0,
|
| 397 |
+
"eval_accuracy": 0.6842619745845552,
|
| 398 |
+
"eval_best_threshold": 0.36684346199035645,
|
| 399 |
+
"eval_f1": 0.7680153220014364,
|
| 400 |
+
"eval_loss": 0.6090383529663086,
|
| 401 |
+
"eval_pr_auc": 0.6967211615556601,
|
| 402 |
+
"eval_precision": 0.6470350947962888,
|
| 403 |
+
"eval_recall": 0.944640753828033,
|
| 404 |
+
"eval_roc_auc": 0.710033428438142,
|
| 405 |
+
"eval_runtime": 2.4866,
|
| 406 |
+
"eval_samples_per_second": 1234.232,
|
| 407 |
+
"eval_steps_per_second": 2.413,
|
| 408 |
+
"step": 42
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.166666666666667,
|
| 412 |
+
"grad_norm": 90881.265625,
|
| 413 |
+
"learning_rate": 0.004971454298742779,
|
| 414 |
+
"loss": 0.6043,
|
| 415 |
+
"step": 43
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.333333333333333,
|
| 419 |
+
"grad_norm": 9821.1396484375,
|
| 420 |
+
"learning_rate": 0.004966903830281449,
|
| 421 |
+
"loss": 0.6071,
|
| 422 |
+
"step": 44
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.5,
|
| 426 |
+
"grad_norm": 167938.828125,
|
| 427 |
+
"learning_rate": 0.00496201938253052,
|
| 428 |
+
"loss": 0.6508,
|
| 429 |
+
"step": 45
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.666666666666667,
|
| 433 |
+
"grad_norm": 45160.44140625,
|
| 434 |
+
"learning_rate": 0.0049568016167660335,
|
| 435 |
+
"loss": 0.6175,
|
| 436 |
+
"step": 46
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.833333333333333,
|
| 440 |
+
"grad_norm": 225251.8125,
|
| 441 |
+
"learning_rate": 0.004951251239389948,
|
| 442 |
+
"loss": 0.682,
|
| 443 |
+
"step": 47
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 8.0,
|
| 447 |
+
"grad_norm": 271435.625,
|
| 448 |
+
"learning_rate": 0.004945369001834514,
|
| 449 |
+
"loss": 0.7134,
|
| 450 |
+
"step": 48
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 8.0,
|
| 454 |
+
"eval_accuracy": 0.6871945259042033,
|
| 455 |
+
"eval_best_threshold": 0.7288259863853455,
|
| 456 |
+
"eval_f1": 0.7659678205753291,
|
| 457 |
+
"eval_loss": 0.7107804417610168,
|
| 458 |
+
"eval_pr_auc": 0.6962413104496704,
|
| 459 |
+
"eval_precision": 0.6534941763727121,
|
| 460 |
+
"eval_recall": 0.925206124852768,
|
| 461 |
+
"eval_roc_auc": 0.7097486294855836,
|
| 462 |
+
"eval_runtime": 2.4945,
|
| 463 |
+
"eval_samples_per_second": 1230.312,
|
| 464 |
+
"eval_steps_per_second": 2.405,
|
| 465 |
+
"step": 48
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 8.166666666666666,
|
| 469 |
+
"grad_norm": 281774.375,
|
| 470 |
+
"learning_rate": 0.004939155700460536,
|
| 471 |
+
"loss": 0.7262,
|
| 472 |
+
"step": 49
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 8.333333333333334,
|
| 476 |
+
"grad_norm": 161943.453125,
|
| 477 |
+
"learning_rate": 0.004932612176449559,
|
| 478 |
+
"loss": 0.6575,
|
| 479 |
+
"step": 50
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 8.5,
|
| 483 |
+
"grad_norm": 36589.8828125,
|
| 484 |
+
"learning_rate": 0.004925739315689991,
|
| 485 |
+
"loss": 0.6269,
|
| 486 |
+
"step": 51
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 8.666666666666666,
|
| 490 |
+
"grad_norm": 81931.7265625,
|
| 491 |
+
"learning_rate": 0.00491853804865716,
|
| 492 |
+
"loss": 0.6326,
|
| 493 |
+
"step": 52
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.833333333333334,
|
| 497 |
+
"grad_norm": 26346.548828125,
|
| 498 |
+
"learning_rate": 0.004911009350287348,
|
| 499 |
+
"loss": 0.613,
|
| 500 |
+
"step": 53
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 9.0,
|
| 504 |
+
"grad_norm": 44889.1796875,
|
| 505 |
+
"learning_rate": 0.004903154239845798,
|
| 506 |
+
"loss": 0.6092,
|
| 507 |
+
"step": 54
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 9.0,
|
| 511 |
+
"eval_accuracy": 0.6855653307266211,
|
| 512 |
+
"eval_best_threshold": 0.451673686504364,
|
| 513 |
+
"eval_f1": 0.7684185265178786,
|
| 514 |
+
"eval_loss": 0.5992416739463806,
|
| 515 |
+
"eval_pr_auc": 0.6957104663478834,
|
| 516 |
+
"eval_precision": 0.648440664236533,
|
| 517 |
+
"eval_recall": 0.9428739693757362,
|
| 518 |
+
"eval_roc_auc": 0.7098253061266568,
|
| 519 |
+
"eval_runtime": 2.5332,
|
| 520 |
+
"eval_samples_per_second": 1211.496,
|
| 521 |
+
"eval_steps_per_second": 2.369,
|
| 522 |
+
"step": 54
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 9.166666666666666,
|
| 526 |
+
"grad_norm": 35008.66796875,
|
| 527 |
+
"learning_rate": 0.004894973780788722,
|
| 528 |
+
"loss": 0.6076,
|
| 529 |
+
"step": 55
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 9.333333333333334,
|
| 533 |
+
"grad_norm": 60022.11328125,
|
| 534 |
+
"learning_rate": 0.00488646908061933,
|
| 535 |
+
"loss": 0.6102,
|
| 536 |
+
"step": 56
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.5,
|
| 540 |
+
"grad_norm": 43722.4765625,
|
| 541 |
+
"learning_rate": 0.0048776412907378846,
|
| 542 |
+
"loss": 0.6014,
|
| 543 |
+
"step": 57
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.666666666666666,
|
| 547 |
+
"grad_norm": 79412.609375,
|
| 548 |
+
"learning_rate": 0.004868491606285823,
|
| 549 |
+
"loss": 0.6359,
|
| 550 |
+
"step": 58
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.833333333333334,
|
| 554 |
+
"grad_norm": 47201.9453125,
|
| 555 |
+
"learning_rate": 0.004859021265983959,
|
| 556 |
+
"loss": 0.617,
|
| 557 |
+
"step": 59
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 10.0,
|
| 561 |
+
"grad_norm": 74637.984375,
|
| 562 |
+
"learning_rate": 0.0048492315519647715,
|
| 563 |
+
"loss": 0.6094,
|
| 564 |
+
"step": 60
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 10.0,
|
| 568 |
+
"eval_accuracy": 0.6946888237210818,
|
| 569 |
+
"eval_best_threshold": 0.4449542760848999,
|
| 570 |
+
"eval_f1": 0.7616382599847367,
|
| 571 |
+
"eval_loss": 0.6147498488426208,
|
| 572 |
+
"eval_pr_auc": 0.6914527562797227,
|
| 573 |
+
"eval_precision": 0.6703985669502911,
|
| 574 |
+
"eval_recall": 0.8816254416961131,
|
| 575 |
+
"eval_roc_auc": 0.7052023704895019,
|
| 576 |
+
"eval_runtime": 2.5228,
|
| 577 |
+
"eval_samples_per_second": 1216.496,
|
| 578 |
+
"eval_steps_per_second": 2.378,
|
| 579 |
+
"step": 60
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 10.166666666666666,
|
| 583 |
+
"grad_norm": 95656.8359375,
|
| 584 |
+
"learning_rate": 0.004839123789598829,
|
| 585 |
+
"loss": 0.6225,
|
| 586 |
+
"step": 61
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 10.333333333333334,
|
| 590 |
+
"grad_norm": 26628.71875,
|
| 591 |
+
"learning_rate": 0.004828699347315356,
|
| 592 |
+
"loss": 0.6025,
|
| 593 |
+
"step": 62
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 10.5,
|
| 597 |
+
"grad_norm": 97382.2421875,
|
| 598 |
+
"learning_rate": 0.004817959636416969,
|
| 599 |
+
"loss": 0.6135,
|
| 600 |
+
"step": 63
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 10.666666666666666,
|
| 604 |
+
"grad_norm": 110094.703125,
|
| 605 |
+
"learning_rate": 0.004806906110888606,
|
| 606 |
+
"loss": 0.6178,
|
| 607 |
+
"step": 64
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 10.833333333333334,
|
| 611 |
+
"grad_norm": 66554.7265625,
|
| 612 |
+
"learning_rate": 0.004795540267200685,
|
| 613 |
+
"loss": 0.607,
|
| 614 |
+
"step": 65
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 11.0,
|
| 618 |
+
"grad_norm": 106479.515625,
|
| 619 |
+
"learning_rate": 0.004783863644106502,
|
| 620 |
+
"loss": 0.6209,
|
| 621 |
+
"step": 66
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 11.0,
|
| 625 |
+
"eval_accuracy": 0.6852394916911045,
|
| 626 |
+
"eval_best_threshold": 0.3707140386104584,
|
| 627 |
+
"eval_f1": 0.7657613967022309,
|
| 628 |
+
"eval_loss": 0.6221640110015869,
|
| 629 |
+
"eval_pr_auc": 0.6986347246977063,
|
| 630 |
+
"eval_precision": 0.6508656224237428,
|
| 631 |
+
"eval_recall": 0.9299175500588928,
|
| 632 |
+
"eval_roc_auc": 0.7001586798387256,
|
| 633 |
+
"eval_runtime": 2.5822,
|
| 634 |
+
"eval_samples_per_second": 1188.505,
|
| 635 |
+
"eval_steps_per_second": 2.324,
|
| 636 |
+
"step": 66
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 11.166666666666666,
|
| 640 |
+
"grad_norm": 137435.59375,
|
| 641 |
+
"learning_rate": 0.004771877822433911,
|
| 642 |
+
"loss": 0.6311,
|
| 643 |
+
"step": 67
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 11.333333333333334,
|
| 647 |
+
"grad_norm": 83377.578125,
|
| 648 |
+
"learning_rate": 0.004759584424871301,
|
| 649 |
+
"loss": 0.6013,
|
| 650 |
+
"step": 68
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 11.5,
|
| 654 |
+
"grad_norm": 83074.90625,
|
| 655 |
+
"learning_rate": 0.004746985115747918,
|
| 656 |
+
"loss": 0.6002,
|
| 657 |
+
"step": 69
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 11.666666666666666,
|
| 661 |
+
"grad_norm": 158799.578125,
|
| 662 |
+
"learning_rate": 0.004734081600808531,
|
| 663 |
+
"loss": 0.6513,
|
| 664 |
+
"step": 70
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 11.833333333333334,
|
| 668 |
+
"grad_norm": 131248.9375,
|
| 669 |
+
"learning_rate": 0.00472087562698251,
|
| 670 |
+
"loss": 0.6265,
|
| 671 |
+
"step": 71
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 12.0,
|
| 675 |
+
"grad_norm": 47012.6796875,
|
| 676 |
+
"learning_rate": 0.0047073689821473174,
|
| 677 |
+
"loss": 0.5974,
|
| 678 |
+
"step": 72
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 12.0,
|
| 682 |
+
"eval_accuracy": 0.6871945259042033,
|
| 683 |
+
"eval_best_threshold": 0.3037373721599579,
|
| 684 |
+
"eval_f1": 0.7706641184902054,
|
| 685 |
+
"eval_loss": 0.6228871941566467,
|
| 686 |
+
"eval_pr_auc": 0.701660429140792,
|
| 687 |
+
"eval_precision": 0.6483118971061094,
|
| 688 |
+
"eval_recall": 0.9499411071849234,
|
| 689 |
+
"eval_roc_auc": 0.7101066685911,
|
| 690 |
+
"eval_runtime": 2.5154,
|
| 691 |
+
"eval_samples_per_second": 1220.097,
|
| 692 |
+
"eval_steps_per_second": 2.385,
|
| 693 |
+
"step": 72
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 12.166666666666666,
|
| 697 |
+
"grad_norm": 155522.734375,
|
| 698 |
+
"learning_rate": 0.004693563494886455,
|
| 699 |
+
"loss": 0.627,
|
| 700 |
+
"step": 73
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 12.333333333333334,
|
| 704 |
+
"grad_norm": 228481.4375,
|
| 705 |
+
"learning_rate": 0.0046794610342419055,
|
| 706 |
+
"loss": 0.6672,
|
| 707 |
+
"step": 74
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 12.5,
|
| 711 |
+
"grad_norm": 181537.359375,
|
| 712 |
+
"learning_rate": 0.004665063509461097,
|
| 713 |
+
"loss": 0.6483,
|
| 714 |
+
"step": 75
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 12.666666666666666,
|
| 718 |
+
"grad_norm": 106007.8359375,
|
| 719 |
+
"learning_rate": 0.004650372869738414,
|
| 720 |
+
"loss": 0.6159,
|
| 721 |
+
"step": 76
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 12.833333333333334,
|
| 725 |
+
"grad_norm": 29989.439453125,
|
| 726 |
+
"learning_rate": 0.0046353911039513146,
|
| 727 |
+
"loss": 0.608,
|
| 728 |
+
"step": 77
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 13.0,
|
| 732 |
+
"grad_norm": 78397.3125,
|
| 733 |
+
"learning_rate": 0.004620120240391065,
|
| 734 |
+
"loss": 0.607,
|
| 735 |
+
"step": 78
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 13.0,
|
| 739 |
+
"eval_accuracy": 0.6845878136200717,
|
| 740 |
+
"eval_best_threshold": 0.4396921992301941,
|
| 741 |
+
"eval_f1": 0.7661835748792271,
|
| 742 |
+
"eval_loss": 0.5964340567588806,
|
| 743 |
+
"eval_pr_auc": 0.7024174776610554,
|
| 744 |
+
"eval_precision": 0.6494676494676495,
|
| 745 |
+
"eval_recall": 0.934040047114252,
|
| 746 |
+
"eval_roc_auc": 0.7138369764402966,
|
| 747 |
+
"eval_runtime": 2.5505,
|
| 748 |
+
"eval_samples_per_second": 1203.299,
|
| 749 |
+
"eval_steps_per_second": 2.352,
|
| 750 |
+
"step": 78
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 13.166666666666666,
|
| 754 |
+
"grad_norm": 63548.2265625,
|
| 755 |
+
"learning_rate": 0.004604562346488143,
|
| 756 |
+
"loss": 0.6052,
|
| 757 |
+
"step": 79
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 13.333333333333334,
|
| 761 |
+
"grad_norm": 40460.96875,
|
| 762 |
+
"learning_rate": 0.004588719528532341,
|
| 763 |
+
"loss": 0.609,
|
| 764 |
+
"step": 80
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 13.5,
|
| 768 |
+
"grad_norm": 66499.796875,
|
| 769 |
+
"learning_rate": 0.004572593931387604,
|
| 770 |
+
"loss": 0.6036,
|
| 771 |
+
"step": 81
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 13.666666666666666,
|
| 775 |
+
"grad_norm": 50256.63671875,
|
| 776 |
+
"learning_rate": 0.004556187738201656,
|
| 777 |
+
"loss": 0.5862,
|
| 778 |
+
"step": 82
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 13.833333333333334,
|
| 782 |
+
"grad_norm": 67104.109375,
|
| 783 |
+
"learning_rate": 0.00453950317011043,
|
| 784 |
+
"loss": 0.6084,
|
| 785 |
+
"step": 83
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 14.0,
|
| 789 |
+
"grad_norm": 78011.8125,
|
| 790 |
+
"learning_rate": 0.0045225424859373685,
|
| 791 |
+
"loss": 0.6091,
|
| 792 |
+
"step": 84
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 14.0,
|
| 796 |
+
"eval_accuracy": 0.6924079504724666,
|
| 797 |
+
"eval_best_threshold": 0.4554637670516968,
|
| 798 |
+
"eval_f1": 0.7708737864077669,
|
| 799 |
+
"eval_loss": 0.5902236104011536,
|
| 800 |
+
"eval_pr_auc": 0.7002916428750828,
|
| 801 |
+
"eval_precision": 0.6556564822460776,
|
| 802 |
+
"eval_recall": 0.9352179034157833,
|
| 803 |
+
"eval_roc_auc": 0.7163127513468885,
|
| 804 |
+
"eval_runtime": 2.5694,
|
| 805 |
+
"eval_samples_per_second": 1194.43,
|
| 806 |
+
"eval_steps_per_second": 2.335,
|
| 807 |
+
"step": 84
|
| 808 |
+
}
|
| 809 |
+
],
|
| 810 |
+
"logging_steps": 1,
|
| 811 |
+
"max_steps": 300,
|
| 812 |
+
"num_input_tokens_seen": 0,
|
| 813 |
+
"num_train_epochs": 50,
|
| 814 |
+
"save_steps": 500,
|
| 815 |
+
"stateful_callbacks": {
|
| 816 |
+
"EarlyStoppingCallback": {
|
| 817 |
+
"args": {
|
| 818 |
+
"early_stopping_patience": 5,
|
| 819 |
+
"early_stopping_threshold": 0.0
|
| 820 |
+
},
|
| 821 |
+
"attributes": {
|
| 822 |
+
"early_stopping_patience_counter": 0
|
| 823 |
+
}
|
| 824 |
+
},
|
| 825 |
+
"TrainerControl": {
|
| 826 |
+
"args": {
|
| 827 |
+
"should_epoch_stop": false,
|
| 828 |
+
"should_evaluate": false,
|
| 829 |
+
"should_log": false,
|
| 830 |
+
"should_save": true,
|
| 831 |
+
"should_training_stop": false
|
| 832 |
+
},
|
| 833 |
+
"attributes": {}
|
| 834 |
+
}
|
| 835 |
+
},
|
| 836 |
+
"total_flos": 6773336776507392.0,
|
| 837 |
+
"train_batch_size": 1024,
|
| 838 |
+
"trial_name": null,
|
| 839 |
+
"trial_params": null
|
| 840 |
+
}
|
checkpoint-84/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a63829309e19cfae4389603a792e6515e9c8585d46a666258924cab2c306c769
|
| 3 |
+
size 5432
|
config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "GenetikaPlus/binary_classification_model_v3.1.5_Junction",
|
| 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": 256,
|
| 11 |
+
"image_size": 32,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 512,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"model_type": "vit",
|
| 16 |
+
"num_attention_heads": 8,
|
| 17 |
+
"num_channels": 3,
|
| 18 |
+
"num_hidden_layers": 4,
|
| 19 |
+
"patch_size": 4,
|
| 20 |
+
"qkv_bias": true,
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.48.0"
|
| 23 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:516e97c20ed900325d474136bb1fdf6f01e76f82cf984f06f99c325965d6a3b7
|
| 3 |
+
size 8563512
|
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": 32,
|
| 21 |
+
"width": 32
|
| 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:a63829309e19cfae4389603a792e6515e9c8585d46a666258924cab2c306c769
|
| 3 |
+
size 5432
|