Commit
•
4e5adba
1
Parent(s):
d95f05a
🍻 cheers
Browse files- README.md +10 -6
- all_results.json +16 -0
- config.json +1 -1
- eval_results.json +12 -0
- train_results.json +8 -0
- trainer_state.json +997 -0
README.md
CHANGED
|
@@ -1,7 +1,11 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: facebook/dinov2-large
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -15,13 +19,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# dino-large-2023_12_08-with_custom_head-imgsize1036
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- F1 Micro: 0.
|
| 22 |
-
- F1 Macro: 0.
|
| 23 |
-
- Roc Auc: 0.
|
| 24 |
-
- Accuracy: 0.
|
| 25 |
- Learning Rate: 0.0001
|
| 26 |
|
| 27 |
## Model description
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- eng
|
| 4 |
license: apache-2.0
|
| 5 |
base_model: facebook/dinov2-large
|
| 6 |
tags:
|
| 7 |
+
- multilabel-image-classification
|
| 8 |
+
- multilabel
|
| 9 |
- generated_from_trainer
|
| 10 |
metrics:
|
| 11 |
- accuracy
|
|
|
|
| 19 |
|
| 20 |
# dino-large-2023_12_08-with_custom_head-imgsize1036
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [facebook/dinov2-large](https://huggingface.co/facebook/dinov2-large) on the multilabel_complete_dataset dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.1017
|
| 25 |
+
- F1 Micro: 0.8547
|
| 26 |
+
- F1 Macro: 0.8214
|
| 27 |
+
- Roc Auc: 0.9081
|
| 28 |
+
- Accuracy: 0.5649
|
| 29 |
- Learning Rate: 0.0001
|
| 30 |
|
| 31 |
## Model description
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_accuracy": 0.5649419218585006,
|
| 4 |
+
"eval_f1_macro": 0.8213721230972145,
|
| 5 |
+
"eval_f1_micro": 0.8546501276040122,
|
| 6 |
+
"eval_loss": 0.10166258364915848,
|
| 7 |
+
"eval_roc_auc": 0.9081373500899811,
|
| 8 |
+
"eval_runtime": 735.6363,
|
| 9 |
+
"eval_samples_per_second": 3.862,
|
| 10 |
+
"eval_steps_per_second": 0.242,
|
| 11 |
+
"learning_rate": 0.0001,
|
| 12 |
+
"train_loss": 0.15577314120620045,
|
| 13 |
+
"train_runtime": 143618.003,
|
| 14 |
+
"train_samples_per_second": 2.983,
|
| 15 |
+
"train_steps_per_second": 0.187
|
| 16 |
+
}
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "facebook/dinov2-
|
| 3 |
"apply_layernorm": true,
|
| 4 |
"architectures": [
|
| 5 |
"NewheadDinov2ForImageClassification"
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "facebook/dinov2-large2023_12_08",
|
| 3 |
"apply_layernorm": true,
|
| 4 |
"architectures": [
|
| 5 |
"NewheadDinov2ForImageClassification"
|
eval_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_accuracy": 0.5649419218585006,
|
| 4 |
+
"eval_f1_macro": 0.8213721230972145,
|
| 5 |
+
"eval_f1_micro": 0.8546501276040122,
|
| 6 |
+
"eval_loss": 0.10166258364915848,
|
| 7 |
+
"eval_roc_auc": 0.9081373500899811,
|
| 8 |
+
"eval_runtime": 735.6363,
|
| 9 |
+
"eval_samples_per_second": 3.862,
|
| 10 |
+
"eval_steps_per_second": 0.242,
|
| 11 |
+
"learning_rate": 0.0001
|
| 12 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"learning_rate": 0.0001,
|
| 4 |
+
"train_loss": 0.15577314120620045,
|
| 5 |
+
"train_runtime": 143618.003,
|
| 6 |
+
"train_samples_per_second": 2.983,
|
| 7 |
+
"train_steps_per_second": 0.187
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,997 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.10334235429763794,
|
| 3 |
+
"best_model_checkpoint": "/home1/datahome/mcontini/multilabelTest/huggingface_multilabel/models/dino-large-2023_12_08-with_custom_head-imgsize1036/checkpoint-26264",
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 26800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.93,
|
| 13 |
+
"learning_rate": 0.01,
|
| 14 |
+
"loss": 0.2502,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.0,
|
| 19 |
+
"eval_accuracy": 0.4030010718113612,
|
| 20 |
+
"eval_f1_macro": 0.5121125903883165,
|
| 21 |
+
"eval_f1_micro": 0.6426868905742146,
|
| 22 |
+
"eval_loss": 0.19902026653289795,
|
| 23 |
+
"eval_roc_auc": 0.7509953312991865,
|
| 24 |
+
"eval_runtime": 684.3444,
|
| 25 |
+
"eval_samples_per_second": 4.09,
|
| 26 |
+
"eval_steps_per_second": 0.256,
|
| 27 |
+
"learning_rate": 0.01,
|
| 28 |
+
"step": 536
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 1.87,
|
| 32 |
+
"learning_rate": 0.01,
|
| 33 |
+
"loss": 0.2164,
|
| 34 |
+
"step": 1000
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.0,
|
| 38 |
+
"eval_accuracy": 0.4047874240800286,
|
| 39 |
+
"eval_f1_macro": 0.6295963337884725,
|
| 40 |
+
"eval_f1_micro": 0.6769131832797427,
|
| 41 |
+
"eval_loss": 1.1450586318969727,
|
| 42 |
+
"eval_roc_auc": 0.7895166431869363,
|
| 43 |
+
"eval_runtime": 697.1722,
|
| 44 |
+
"eval_samples_per_second": 4.015,
|
| 45 |
+
"eval_steps_per_second": 0.251,
|
| 46 |
+
"learning_rate": 0.01,
|
| 47 |
+
"step": 1072
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.8,
|
| 51 |
+
"learning_rate": 0.01,
|
| 52 |
+
"loss": 0.2149,
|
| 53 |
+
"step": 1500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 3.0,
|
| 57 |
+
"eval_accuracy": 0.3994283672740264,
|
| 58 |
+
"eval_f1_macro": 0.5251308791060187,
|
| 59 |
+
"eval_f1_micro": 0.651589690873045,
|
| 60 |
+
"eval_loss": 0.2338775098323822,
|
| 61 |
+
"eval_roc_auc": 0.7524284144947219,
|
| 62 |
+
"eval_runtime": 691.846,
|
| 63 |
+
"eval_samples_per_second": 4.046,
|
| 64 |
+
"eval_steps_per_second": 0.253,
|
| 65 |
+
"learning_rate": 0.01,
|
| 66 |
+
"step": 1608
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 3.73,
|
| 70 |
+
"learning_rate": 0.01,
|
| 71 |
+
"loss": 0.2159,
|
| 72 |
+
"step": 2000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 4.0,
|
| 76 |
+
"eval_accuracy": 0.4115755627009646,
|
| 77 |
+
"eval_f1_macro": 0.6028196862535703,
|
| 78 |
+
"eval_f1_micro": 0.7376345424002986,
|
| 79 |
+
"eval_loss": 0.16987967491149902,
|
| 80 |
+
"eval_roc_auc": 0.8296525341756016,
|
| 81 |
+
"eval_runtime": 692.4855,
|
| 82 |
+
"eval_samples_per_second": 4.042,
|
| 83 |
+
"eval_steps_per_second": 0.253,
|
| 84 |
+
"learning_rate": 0.01,
|
| 85 |
+
"step": 2144
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 4.66,
|
| 89 |
+
"learning_rate": 0.01,
|
| 90 |
+
"loss": 0.2171,
|
| 91 |
+
"step": 2500
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 5.0,
|
| 95 |
+
"eval_accuracy": 0.43944265809217575,
|
| 96 |
+
"eval_f1_macro": 0.63036934398975,
|
| 97 |
+
"eval_f1_micro": 0.7363523166655744,
|
| 98 |
+
"eval_loss": 0.16497960686683655,
|
| 99 |
+
"eval_roc_auc": 0.8166227130928913,
|
| 100 |
+
"eval_runtime": 679.5828,
|
| 101 |
+
"eval_samples_per_second": 4.119,
|
| 102 |
+
"eval_steps_per_second": 0.258,
|
| 103 |
+
"learning_rate": 0.01,
|
| 104 |
+
"step": 2680
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 5.6,
|
| 108 |
+
"learning_rate": 0.01,
|
| 109 |
+
"loss": 0.2166,
|
| 110 |
+
"step": 3000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 6.0,
|
| 114 |
+
"eval_accuracy": 0.39228295819935693,
|
| 115 |
+
"eval_f1_macro": 0.5122569229120599,
|
| 116 |
+
"eval_f1_micro": 0.6852550663871418,
|
| 117 |
+
"eval_loss": 0.1748417615890503,
|
| 118 |
+
"eval_roc_auc": 0.7773210612716831,
|
| 119 |
+
"eval_runtime": 677.7081,
|
| 120 |
+
"eval_samples_per_second": 4.13,
|
| 121 |
+
"eval_steps_per_second": 0.258,
|
| 122 |
+
"learning_rate": 0.01,
|
| 123 |
+
"step": 3216
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 6.53,
|
| 127 |
+
"learning_rate": 0.01,
|
| 128 |
+
"loss": 0.2081,
|
| 129 |
+
"step": 3500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 7.0,
|
| 133 |
+
"eval_accuracy": 0.4347981421936406,
|
| 134 |
+
"eval_f1_macro": 0.6129188202062517,
|
| 135 |
+
"eval_f1_micro": 0.7454787740338855,
|
| 136 |
+
"eval_loss": 0.1636398434638977,
|
| 137 |
+
"eval_roc_auc": 0.8290983004904813,
|
| 138 |
+
"eval_runtime": 695.4612,
|
| 139 |
+
"eval_samples_per_second": 4.025,
|
| 140 |
+
"eval_steps_per_second": 0.252,
|
| 141 |
+
"learning_rate": 0.01,
|
| 142 |
+
"step": 3752
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 7.46,
|
| 146 |
+
"learning_rate": 0.01,
|
| 147 |
+
"loss": 0.2111,
|
| 148 |
+
"step": 4000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 8.0,
|
| 152 |
+
"eval_accuracy": 0.42765273311897106,
|
| 153 |
+
"eval_f1_macro": 0.6444819502221721,
|
| 154 |
+
"eval_f1_micro": 0.7542994588093808,
|
| 155 |
+
"eval_loss": 0.16511820256710052,
|
| 156 |
+
"eval_roc_auc": 0.847560960415324,
|
| 157 |
+
"eval_runtime": 698.6301,
|
| 158 |
+
"eval_samples_per_second": 4.006,
|
| 159 |
+
"eval_steps_per_second": 0.25,
|
| 160 |
+
"learning_rate": 0.01,
|
| 161 |
+
"step": 4288
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 8.4,
|
| 165 |
+
"learning_rate": 0.01,
|
| 166 |
+
"loss": 0.209,
|
| 167 |
+
"step": 4500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 9.0,
|
| 171 |
+
"eval_accuracy": 0.43586995355484104,
|
| 172 |
+
"eval_f1_macro": 0.6522075091299645,
|
| 173 |
+
"eval_f1_micro": 0.7062128001070019,
|
| 174 |
+
"eval_loss": 0.1750405728816986,
|
| 175 |
+
"eval_roc_auc": 0.7966230241843142,
|
| 176 |
+
"eval_runtime": 689.5145,
|
| 177 |
+
"eval_samples_per_second": 4.059,
|
| 178 |
+
"eval_steps_per_second": 0.254,
|
| 179 |
+
"learning_rate": 0.01,
|
| 180 |
+
"step": 4824
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 9.33,
|
| 184 |
+
"learning_rate": 0.01,
|
| 185 |
+
"loss": 0.2107,
|
| 186 |
+
"step": 5000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 10.0,
|
| 190 |
+
"eval_accuracy": 0.3829939264022865,
|
| 191 |
+
"eval_f1_macro": 0.5923770104096581,
|
| 192 |
+
"eval_f1_micro": 0.7244064088539991,
|
| 193 |
+
"eval_loss": 0.17505380511283875,
|
| 194 |
+
"eval_roc_auc": 0.8145635827712373,
|
| 195 |
+
"eval_runtime": 706.021,
|
| 196 |
+
"eval_samples_per_second": 3.964,
|
| 197 |
+
"eval_steps_per_second": 0.248,
|
| 198 |
+
"learning_rate": 0.01,
|
| 199 |
+
"step": 5360
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 10.26,
|
| 203 |
+
"learning_rate": 0.01,
|
| 204 |
+
"loss": 0.2162,
|
| 205 |
+
"step": 5500
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 11.0,
|
| 209 |
+
"eval_accuracy": 0.42515183994283673,
|
| 210 |
+
"eval_f1_macro": 0.678019114323704,
|
| 211 |
+
"eval_f1_micro": 0.7505816381315993,
|
| 212 |
+
"eval_loss": 0.22285762429237366,
|
| 213 |
+
"eval_roc_auc": 0.8475316447819295,
|
| 214 |
+
"eval_runtime": 709.1839,
|
| 215 |
+
"eval_samples_per_second": 3.947,
|
| 216 |
+
"eval_steps_per_second": 0.247,
|
| 217 |
+
"learning_rate": 0.01,
|
| 218 |
+
"step": 5896
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 11.19,
|
| 222 |
+
"learning_rate": 0.01,
|
| 223 |
+
"loss": 0.2153,
|
| 224 |
+
"step": 6000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 12.0,
|
| 228 |
+
"eval_accuracy": 0.4105037513397642,
|
| 229 |
+
"eval_f1_macro": 0.6543045937545326,
|
| 230 |
+
"eval_f1_micro": 0.7501448603546181,
|
| 231 |
+
"eval_loss": 0.1740036904811859,
|
| 232 |
+
"eval_roc_auc": 0.8550110327013883,
|
| 233 |
+
"eval_runtime": 710.4504,
|
| 234 |
+
"eval_samples_per_second": 3.94,
|
| 235 |
+
"eval_steps_per_second": 0.246,
|
| 236 |
+
"learning_rate": 0.01,
|
| 237 |
+
"step": 6432
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 12.13,
|
| 241 |
+
"learning_rate": 0.01,
|
| 242 |
+
"loss": 0.2197,
|
| 243 |
+
"step": 6500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 13.0,
|
| 247 |
+
"eval_accuracy": 0.41872097177563417,
|
| 248 |
+
"eval_f1_macro": 0.6605308887240204,
|
| 249 |
+
"eval_f1_micro": 0.748710897215538,
|
| 250 |
+
"eval_loss": 0.1744571030139923,
|
| 251 |
+
"eval_roc_auc": 0.857218077620948,
|
| 252 |
+
"eval_runtime": 687.247,
|
| 253 |
+
"eval_samples_per_second": 4.073,
|
| 254 |
+
"eval_steps_per_second": 0.255,
|
| 255 |
+
"learning_rate": 0.01,
|
| 256 |
+
"step": 6968
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 13.06,
|
| 260 |
+
"learning_rate": 0.001,
|
| 261 |
+
"loss": 0.2128,
|
| 262 |
+
"step": 7000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 13.99,
|
| 266 |
+
"learning_rate": 0.001,
|
| 267 |
+
"loss": 0.18,
|
| 268 |
+
"step": 7500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 14.0,
|
| 272 |
+
"eval_accuracy": 0.5058949624866024,
|
| 273 |
+
"eval_f1_macro": 0.7455097936778116,
|
| 274 |
+
"eval_f1_micro": 0.8036385688295937,
|
| 275 |
+
"eval_loss": 0.13480359315872192,
|
| 276 |
+
"eval_roc_auc": 0.8730946172404269,
|
| 277 |
+
"eval_runtime": 700.8684,
|
| 278 |
+
"eval_samples_per_second": 3.994,
|
| 279 |
+
"eval_steps_per_second": 0.25,
|
| 280 |
+
"learning_rate": 0.001,
|
| 281 |
+
"step": 7504
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 14.93,
|
| 285 |
+
"learning_rate": 0.001,
|
| 286 |
+
"loss": 0.164,
|
| 287 |
+
"step": 8000
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 15.0,
|
| 291 |
+
"eval_accuracy": 0.5173276170060735,
|
| 292 |
+
"eval_f1_macro": 0.7782933504010985,
|
| 293 |
+
"eval_f1_micro": 0.8159655831739961,
|
| 294 |
+
"eval_loss": 0.13084079325199127,
|
| 295 |
+
"eval_roc_auc": 0.8844182661393802,
|
| 296 |
+
"eval_runtime": 699.1842,
|
| 297 |
+
"eval_samples_per_second": 4.003,
|
| 298 |
+
"eval_steps_per_second": 0.25,
|
| 299 |
+
"learning_rate": 0.001,
|
| 300 |
+
"step": 8040
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 15.86,
|
| 304 |
+
"learning_rate": 0.001,
|
| 305 |
+
"loss": 0.162,
|
| 306 |
+
"step": 8500
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 16.0,
|
| 310 |
+
"eval_accuracy": 0.5201857806359415,
|
| 311 |
+
"eval_f1_macro": 0.7530230371766228,
|
| 312 |
+
"eval_f1_micro": 0.8187878039748179,
|
| 313 |
+
"eval_loss": 0.1304706484079361,
|
| 314 |
+
"eval_roc_auc": 0.8764409254464951,
|
| 315 |
+
"eval_runtime": 704.2301,
|
| 316 |
+
"eval_samples_per_second": 3.975,
|
| 317 |
+
"eval_steps_per_second": 0.248,
|
| 318 |
+
"learning_rate": 0.001,
|
| 319 |
+
"step": 8576
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 16.79,
|
| 323 |
+
"learning_rate": 0.001,
|
| 324 |
+
"loss": 0.1548,
|
| 325 |
+
"step": 9000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 17.0,
|
| 329 |
+
"eval_accuracy": 0.5248302965344765,
|
| 330 |
+
"eval_f1_macro": 0.788713547954132,
|
| 331 |
+
"eval_f1_micro": 0.829077551987677,
|
| 332 |
+
"eval_loss": 0.12419470399618149,
|
| 333 |
+
"eval_roc_auc": 0.8945177280625883,
|
| 334 |
+
"eval_runtime": 725.0968,
|
| 335 |
+
"eval_samples_per_second": 3.86,
|
| 336 |
+
"eval_steps_per_second": 0.241,
|
| 337 |
+
"learning_rate": 0.001,
|
| 338 |
+
"step": 9112
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 17.72,
|
| 342 |
+
"learning_rate": 0.001,
|
| 343 |
+
"loss": 0.1532,
|
| 344 |
+
"step": 9500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 18.0,
|
| 348 |
+
"eval_accuracy": 0.5226866738120758,
|
| 349 |
+
"eval_f1_macro": 0.782300591476402,
|
| 350 |
+
"eval_f1_micro": 0.8292102759276879,
|
| 351 |
+
"eval_loss": 0.12466239929199219,
|
| 352 |
+
"eval_roc_auc": 0.8934405856125267,
|
| 353 |
+
"eval_runtime": 737.9752,
|
| 354 |
+
"eval_samples_per_second": 3.793,
|
| 355 |
+
"eval_steps_per_second": 0.237,
|
| 356 |
+
"learning_rate": 0.001,
|
| 357 |
+
"step": 9648
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 18.66,
|
| 361 |
+
"learning_rate": 0.001,
|
| 362 |
+
"loss": 0.152,
|
| 363 |
+
"step": 10000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 19.0,
|
| 367 |
+
"eval_accuracy": 0.5280457306180779,
|
| 368 |
+
"eval_f1_macro": 0.7687831529578533,
|
| 369 |
+
"eval_f1_micro": 0.8237944471505114,
|
| 370 |
+
"eval_loss": 0.1271921843290329,
|
| 371 |
+
"eval_roc_auc": 0.8832304052589448,
|
| 372 |
+
"eval_runtime": 742.4217,
|
| 373 |
+
"eval_samples_per_second": 3.77,
|
| 374 |
+
"eval_steps_per_second": 0.236,
|
| 375 |
+
"learning_rate": 0.001,
|
| 376 |
+
"step": 10184
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 19.59,
|
| 380 |
+
"learning_rate": 0.001,
|
| 381 |
+
"loss": 0.1479,
|
| 382 |
+
"step": 10500
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 20.0,
|
| 386 |
+
"eval_accuracy": 0.5287602715255448,
|
| 387 |
+
"eval_f1_macro": 0.7783155153237968,
|
| 388 |
+
"eval_f1_micro": 0.8279668813247469,
|
| 389 |
+
"eval_loss": 0.12385135143995285,
|
| 390 |
+
"eval_roc_auc": 0.8833930136365329,
|
| 391 |
+
"eval_runtime": 714.6841,
|
| 392 |
+
"eval_samples_per_second": 3.916,
|
| 393 |
+
"eval_steps_per_second": 0.245,
|
| 394 |
+
"learning_rate": 0.001,
|
| 395 |
+
"step": 10720
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 20.52,
|
| 399 |
+
"learning_rate": 0.001,
|
| 400 |
+
"loss": 0.1483,
|
| 401 |
+
"step": 11000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 21.0,
|
| 405 |
+
"eval_accuracy": 0.534119328331547,
|
| 406 |
+
"eval_f1_macro": 0.791367242205616,
|
| 407 |
+
"eval_f1_micro": 0.8361280949209999,
|
| 408 |
+
"eval_loss": 0.1375582218170166,
|
| 409 |
+
"eval_roc_auc": 0.8918987306282016,
|
| 410 |
+
"eval_runtime": 704.0932,
|
| 411 |
+
"eval_samples_per_second": 3.975,
|
| 412 |
+
"eval_steps_per_second": 0.249,
|
| 413 |
+
"learning_rate": 0.001,
|
| 414 |
+
"step": 11256
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 21.46,
|
| 418 |
+
"learning_rate": 0.001,
|
| 419 |
+
"loss": 0.1448,
|
| 420 |
+
"step": 11500
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 22.0,
|
| 424 |
+
"eval_accuracy": 0.5380493033226152,
|
| 425 |
+
"eval_f1_macro": 0.7773902467356547,
|
| 426 |
+
"eval_f1_micro": 0.8291756052712227,
|
| 427 |
+
"eval_loss": 0.12673501670360565,
|
| 428 |
+
"eval_roc_auc": 0.8842480463699955,
|
| 429 |
+
"eval_runtime": 694.3206,
|
| 430 |
+
"eval_samples_per_second": 4.031,
|
| 431 |
+
"eval_steps_per_second": 0.252,
|
| 432 |
+
"learning_rate": 0.001,
|
| 433 |
+
"step": 11792
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 22.39,
|
| 437 |
+
"learning_rate": 0.001,
|
| 438 |
+
"loss": 0.1456,
|
| 439 |
+
"step": 12000
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 23.0,
|
| 443 |
+
"eval_accuracy": 0.5448374419435512,
|
| 444 |
+
"eval_f1_macro": 0.7913781838771046,
|
| 445 |
+
"eval_f1_micro": 0.8334247577253612,
|
| 446 |
+
"eval_loss": 0.12168043851852417,
|
| 447 |
+
"eval_roc_auc": 0.8883093845683582,
|
| 448 |
+
"eval_runtime": 716.6426,
|
| 449 |
+
"eval_samples_per_second": 3.906,
|
| 450 |
+
"eval_steps_per_second": 0.244,
|
| 451 |
+
"learning_rate": 0.001,
|
| 452 |
+
"step": 12328
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 23.32,
|
| 456 |
+
"learning_rate": 0.001,
|
| 457 |
+
"loss": 0.1441,
|
| 458 |
+
"step": 12500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 24.0,
|
| 462 |
+
"eval_accuracy": 0.5380493033226152,
|
| 463 |
+
"eval_f1_macro": 0.7852190635254378,
|
| 464 |
+
"eval_f1_micro": 0.8282665343094677,
|
| 465 |
+
"eval_loss": 0.11930090188980103,
|
| 466 |
+
"eval_roc_auc": 0.8801416627925638,
|
| 467 |
+
"eval_runtime": 736.3569,
|
| 468 |
+
"eval_samples_per_second": 3.801,
|
| 469 |
+
"eval_steps_per_second": 0.238,
|
| 470 |
+
"learning_rate": 0.001,
|
| 471 |
+
"step": 12864
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 24.25,
|
| 475 |
+
"learning_rate": 0.001,
|
| 476 |
+
"loss": 0.1406,
|
| 477 |
+
"step": 13000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 25.0,
|
| 481 |
+
"eval_accuracy": 0.534119328331547,
|
| 482 |
+
"eval_f1_macro": 0.8019722412424143,
|
| 483 |
+
"eval_f1_micro": 0.8392282958199356,
|
| 484 |
+
"eval_loss": 0.11846613883972168,
|
| 485 |
+
"eval_roc_auc": 0.89877394155691,
|
| 486 |
+
"eval_runtime": 732.3109,
|
| 487 |
+
"eval_samples_per_second": 3.822,
|
| 488 |
+
"eval_steps_per_second": 0.239,
|
| 489 |
+
"learning_rate": 0.001,
|
| 490 |
+
"step": 13400
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 25.19,
|
| 494 |
+
"learning_rate": 0.001,
|
| 495 |
+
"loss": 0.1416,
|
| 496 |
+
"step": 13500
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 26.0,
|
| 500 |
+
"eval_accuracy": 0.5441229010360843,
|
| 501 |
+
"eval_f1_macro": 0.7850806686320094,
|
| 502 |
+
"eval_f1_micro": 0.8350723045945452,
|
| 503 |
+
"eval_loss": 0.1295102834701538,
|
| 504 |
+
"eval_roc_auc": 0.888897831813434,
|
| 505 |
+
"eval_runtime": 736.4837,
|
| 506 |
+
"eval_samples_per_second": 3.8,
|
| 507 |
+
"eval_steps_per_second": 0.238,
|
| 508 |
+
"learning_rate": 0.001,
|
| 509 |
+
"step": 13936
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 26.12,
|
| 513 |
+
"learning_rate": 0.001,
|
| 514 |
+
"loss": 0.1417,
|
| 515 |
+
"step": 14000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 27.0,
|
| 519 |
+
"eval_accuracy": 0.5305466237942122,
|
| 520 |
+
"eval_f1_macro": 0.7699298453614014,
|
| 521 |
+
"eval_f1_micro": 0.8287484510532839,
|
| 522 |
+
"eval_loss": 0.1389976292848587,
|
| 523 |
+
"eval_roc_auc": 0.8808089652631531,
|
| 524 |
+
"eval_runtime": 714.8246,
|
| 525 |
+
"eval_samples_per_second": 3.916,
|
| 526 |
+
"eval_steps_per_second": 0.245,
|
| 527 |
+
"learning_rate": 0.001,
|
| 528 |
+
"step": 14472
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 27.05,
|
| 532 |
+
"learning_rate": 0.001,
|
| 533 |
+
"loss": 0.1452,
|
| 534 |
+
"step": 14500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 27.99,
|
| 538 |
+
"learning_rate": 0.001,
|
| 539 |
+
"loss": 0.142,
|
| 540 |
+
"step": 15000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 28.0,
|
| 544 |
+
"eval_accuracy": 0.5441229010360843,
|
| 545 |
+
"eval_f1_macro": 0.7857164640115839,
|
| 546 |
+
"eval_f1_micro": 0.8328267477203647,
|
| 547 |
+
"eval_loss": 0.12560133635997772,
|
| 548 |
+
"eval_roc_auc": 0.888767490354217,
|
| 549 |
+
"eval_runtime": 711.9043,
|
| 550 |
+
"eval_samples_per_second": 3.932,
|
| 551 |
+
"eval_steps_per_second": 0.246,
|
| 552 |
+
"learning_rate": 0.001,
|
| 553 |
+
"step": 15008
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 28.92,
|
| 557 |
+
"learning_rate": 0.001,
|
| 558 |
+
"loss": 0.14,
|
| 559 |
+
"step": 15500
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 29.0,
|
| 563 |
+
"eval_accuracy": 0.5359056806002144,
|
| 564 |
+
"eval_f1_macro": 0.7758987493304471,
|
| 565 |
+
"eval_f1_micro": 0.8291174469664172,
|
| 566 |
+
"eval_loss": 0.12682993710041046,
|
| 567 |
+
"eval_roc_auc": 0.8815429141058391,
|
| 568 |
+
"eval_runtime": 725.5421,
|
| 569 |
+
"eval_samples_per_second": 3.858,
|
| 570 |
+
"eval_steps_per_second": 0.241,
|
| 571 |
+
"learning_rate": 0.001,
|
| 572 |
+
"step": 15544
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 29.85,
|
| 576 |
+
"learning_rate": 0.001,
|
| 577 |
+
"loss": 0.1415,
|
| 578 |
+
"step": 16000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 30.0,
|
| 582 |
+
"eval_accuracy": 0.5419792783136834,
|
| 583 |
+
"eval_f1_macro": 0.7674795938272058,
|
| 584 |
+
"eval_f1_micro": 0.8239833027638985,
|
| 585 |
+
"eval_loss": 0.13736343383789062,
|
| 586 |
+
"eval_roc_auc": 0.8722214417609943,
|
| 587 |
+
"eval_runtime": 716.5365,
|
| 588 |
+
"eval_samples_per_second": 3.906,
|
| 589 |
+
"eval_steps_per_second": 0.244,
|
| 590 |
+
"learning_rate": 0.001,
|
| 591 |
+
"step": 16080
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 30.78,
|
| 595 |
+
"learning_rate": 0.001,
|
| 596 |
+
"loss": 0.1414,
|
| 597 |
+
"step": 16500
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 31.0,
|
| 601 |
+
"eval_accuracy": 0.5405501964987496,
|
| 602 |
+
"eval_f1_macro": 0.7795434508186554,
|
| 603 |
+
"eval_f1_micro": 0.8309928553830993,
|
| 604 |
+
"eval_loss": 0.12812598049640656,
|
| 605 |
+
"eval_roc_auc": 0.8838076169713958,
|
| 606 |
+
"eval_runtime": 695.9054,
|
| 607 |
+
"eval_samples_per_second": 4.022,
|
| 608 |
+
"eval_steps_per_second": 0.251,
|
| 609 |
+
"learning_rate": 0.001,
|
| 610 |
+
"step": 16616
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 31.72,
|
| 614 |
+
"learning_rate": 0.0001,
|
| 615 |
+
"loss": 0.1349,
|
| 616 |
+
"step": 17000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 32.0,
|
| 620 |
+
"eval_accuracy": 0.5512683101107538,
|
| 621 |
+
"eval_f1_macro": 0.7927055189210745,
|
| 622 |
+
"eval_f1_micro": 0.8388680866736234,
|
| 623 |
+
"eval_loss": 0.11437654495239258,
|
| 624 |
+
"eval_roc_auc": 0.8891921900071421,
|
| 625 |
+
"eval_runtime": 701.3503,
|
| 626 |
+
"eval_samples_per_second": 3.991,
|
| 627 |
+
"eval_steps_per_second": 0.25,
|
| 628 |
+
"learning_rate": 0.0001,
|
| 629 |
+
"step": 17152
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 32.65,
|
| 633 |
+
"learning_rate": 0.0001,
|
| 634 |
+
"loss": 0.1294,
|
| 635 |
+
"step": 17500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 33.0,
|
| 639 |
+
"eval_accuracy": 0.5534119328331547,
|
| 640 |
+
"eval_f1_macro": 0.7990834574660737,
|
| 641 |
+
"eval_f1_micro": 0.8414029503580829,
|
| 642 |
+
"eval_loss": 0.10969647765159607,
|
| 643 |
+
"eval_roc_auc": 0.8914615216569188,
|
| 644 |
+
"eval_runtime": 698.1361,
|
| 645 |
+
"eval_samples_per_second": 4.009,
|
| 646 |
+
"eval_steps_per_second": 0.251,
|
| 647 |
+
"learning_rate": 0.0001,
|
| 648 |
+
"step": 17688
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 33.58,
|
| 652 |
+
"learning_rate": 0.0001,
|
| 653 |
+
"loss": 0.1281,
|
| 654 |
+
"step": 18000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 34.0,
|
| 658 |
+
"eval_accuracy": 0.5519828510182208,
|
| 659 |
+
"eval_f1_macro": 0.7981650632699254,
|
| 660 |
+
"eval_f1_micro": 0.8424912902634313,
|
| 661 |
+
"eval_loss": 0.11602061241865158,
|
| 662 |
+
"eval_roc_auc": 0.8925180364347453,
|
| 663 |
+
"eval_runtime": 712.6765,
|
| 664 |
+
"eval_samples_per_second": 3.927,
|
| 665 |
+
"eval_steps_per_second": 0.246,
|
| 666 |
+
"learning_rate": 0.0001,
|
| 667 |
+
"step": 18224
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 34.51,
|
| 671 |
+
"learning_rate": 0.0001,
|
| 672 |
+
"loss": 0.1274,
|
| 673 |
+
"step": 18500
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 35.0,
|
| 677 |
+
"eval_accuracy": 0.5576991782779565,
|
| 678 |
+
"eval_f1_macro": 0.7999145597601702,
|
| 679 |
+
"eval_f1_micro": 0.8441375939390237,
|
| 680 |
+
"eval_loss": 0.12442068755626678,
|
| 681 |
+
"eval_roc_auc": 0.893549294756116,
|
| 682 |
+
"eval_runtime": 707.8053,
|
| 683 |
+
"eval_samples_per_second": 3.954,
|
| 684 |
+
"eval_steps_per_second": 0.247,
|
| 685 |
+
"learning_rate": 0.0001,
|
| 686 |
+
"step": 18760
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 35.45,
|
| 690 |
+
"learning_rate": 0.0001,
|
| 691 |
+
"loss": 0.1243,
|
| 692 |
+
"step": 19000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 36.0,
|
| 696 |
+
"eval_accuracy": 0.555912826009289,
|
| 697 |
+
"eval_f1_macro": 0.7991433380498747,
|
| 698 |
+
"eval_f1_micro": 0.8434315360385399,
|
| 699 |
+
"eval_loss": 0.10998840630054474,
|
| 700 |
+
"eval_roc_auc": 0.889847418195485,
|
| 701 |
+
"eval_runtime": 712.761,
|
| 702 |
+
"eval_samples_per_second": 3.927,
|
| 703 |
+
"eval_steps_per_second": 0.246,
|
| 704 |
+
"learning_rate": 0.0001,
|
| 705 |
+
"step": 19296
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 36.38,
|
| 709 |
+
"learning_rate": 0.0001,
|
| 710 |
+
"loss": 0.1231,
|
| 711 |
+
"step": 19500
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 37.0,
|
| 715 |
+
"eval_accuracy": 0.564130046445159,
|
| 716 |
+
"eval_f1_macro": 0.8086403056854424,
|
| 717 |
+
"eval_f1_micro": 0.8485288775880857,
|
| 718 |
+
"eval_loss": 0.10725793987512589,
|
| 719 |
+
"eval_roc_auc": 0.8988640009924486,
|
| 720 |
+
"eval_runtime": 714.0011,
|
| 721 |
+
"eval_samples_per_second": 3.92,
|
| 722 |
+
"eval_steps_per_second": 0.245,
|
| 723 |
+
"learning_rate": 0.0001,
|
| 724 |
+
"step": 19832
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 37.31,
|
| 728 |
+
"learning_rate": 0.0001,
|
| 729 |
+
"loss": 0.1245,
|
| 730 |
+
"step": 20000
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 38.0,
|
| 734 |
+
"eval_accuracy": 0.5602000714540908,
|
| 735 |
+
"eval_f1_macro": 0.8053806942991502,
|
| 736 |
+
"eval_f1_micro": 0.8455705566857777,
|
| 737 |
+
"eval_loss": 0.1091923713684082,
|
| 738 |
+
"eval_roc_auc": 0.8915942447938312,
|
| 739 |
+
"eval_runtime": 713.3543,
|
| 740 |
+
"eval_samples_per_second": 3.924,
|
| 741 |
+
"eval_steps_per_second": 0.245,
|
| 742 |
+
"learning_rate": 0.0001,
|
| 743 |
+
"step": 20368
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 38.25,
|
| 747 |
+
"learning_rate": 0.0001,
|
| 748 |
+
"loss": 0.1197,
|
| 749 |
+
"step": 20500
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 39.0,
|
| 753 |
+
"eval_accuracy": 0.5623436941764915,
|
| 754 |
+
"eval_f1_macro": 0.8111886307990975,
|
| 755 |
+
"eval_f1_micro": 0.8482584066530071,
|
| 756 |
+
"eval_loss": 0.10690516978502274,
|
| 757 |
+
"eval_roc_auc": 0.9001525872337657,
|
| 758 |
+
"eval_runtime": 742.3723,
|
| 759 |
+
"eval_samples_per_second": 3.77,
|
| 760 |
+
"eval_steps_per_second": 0.236,
|
| 761 |
+
"learning_rate": 0.0001,
|
| 762 |
+
"step": 20904
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 39.18,
|
| 766 |
+
"learning_rate": 0.0001,
|
| 767 |
+
"loss": 0.1242,
|
| 768 |
+
"step": 21000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 40.0,
|
| 772 |
+
"eval_accuracy": 0.5637727759914255,
|
| 773 |
+
"eval_f1_macro": 0.8080882298039329,
|
| 774 |
+
"eval_f1_micro": 0.8468215158924205,
|
| 775 |
+
"eval_loss": 0.10654206573963165,
|
| 776 |
+
"eval_roc_auc": 0.8949146236875923,
|
| 777 |
+
"eval_runtime": 748.0644,
|
| 778 |
+
"eval_samples_per_second": 3.742,
|
| 779 |
+
"eval_steps_per_second": 0.234,
|
| 780 |
+
"learning_rate": 0.0001,
|
| 781 |
+
"step": 21440
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 40.11,
|
| 785 |
+
"learning_rate": 0.0001,
|
| 786 |
+
"loss": 0.1167,
|
| 787 |
+
"step": 21500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 41.0,
|
| 791 |
+
"eval_accuracy": 0.5591282600928903,
|
| 792 |
+
"eval_f1_macro": 0.8043452814541342,
|
| 793 |
+
"eval_f1_micro": 0.8462199312714777,
|
| 794 |
+
"eval_loss": 0.10834041237831116,
|
| 795 |
+
"eval_roc_auc": 0.8933701178832314,
|
| 796 |
+
"eval_runtime": 743.3565,
|
| 797 |
+
"eval_samples_per_second": 3.765,
|
| 798 |
+
"eval_steps_per_second": 0.235,
|
| 799 |
+
"learning_rate": 0.0001,
|
| 800 |
+
"step": 21976
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 41.04,
|
| 804 |
+
"learning_rate": 0.0001,
|
| 805 |
+
"loss": 0.1201,
|
| 806 |
+
"step": 22000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 41.98,
|
| 810 |
+
"learning_rate": 0.0001,
|
| 811 |
+
"loss": 0.1179,
|
| 812 |
+
"step": 22500
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 42.0,
|
| 816 |
+
"eval_accuracy": 0.5691318327974276,
|
| 817 |
+
"eval_f1_macro": 0.8089823958516517,
|
| 818 |
+
"eval_f1_micro": 0.8505242623750304,
|
| 819 |
+
"eval_loss": 0.10719550400972366,
|
| 820 |
+
"eval_roc_auc": 0.8978175825326844,
|
| 821 |
+
"eval_runtime": 698.4214,
|
| 822 |
+
"eval_samples_per_second": 4.008,
|
| 823 |
+
"eval_steps_per_second": 0.251,
|
| 824 |
+
"learning_rate": 0.0001,
|
| 825 |
+
"step": 22512
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 42.91,
|
| 829 |
+
"learning_rate": 0.0001,
|
| 830 |
+
"loss": 0.1186,
|
| 831 |
+
"step": 23000
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 43.0,
|
| 835 |
+
"eval_accuracy": 0.5612718828152912,
|
| 836 |
+
"eval_f1_macro": 0.8071782588939219,
|
| 837 |
+
"eval_f1_micro": 0.8482877132686878,
|
| 838 |
+
"eval_loss": 0.11790523678064346,
|
| 839 |
+
"eval_roc_auc": 0.8944810701797377,
|
| 840 |
+
"eval_runtime": 695.4681,
|
| 841 |
+
"eval_samples_per_second": 4.025,
|
| 842 |
+
"eval_steps_per_second": 0.252,
|
| 843 |
+
"learning_rate": 0.0001,
|
| 844 |
+
"step": 23048
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 43.84,
|
| 848 |
+
"learning_rate": 0.0001,
|
| 849 |
+
"loss": 0.1174,
|
| 850 |
+
"step": 23500
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 44.0,
|
| 854 |
+
"eval_accuracy": 0.5655591282600929,
|
| 855 |
+
"eval_f1_macro": 0.808033951861079,
|
| 856 |
+
"eval_f1_micro": 0.8477235124701268,
|
| 857 |
+
"eval_loss": 0.10680884122848511,
|
| 858 |
+
"eval_roc_auc": 0.8946411623084602,
|
| 859 |
+
"eval_runtime": 698.3846,
|
| 860 |
+
"eval_samples_per_second": 4.008,
|
| 861 |
+
"eval_steps_per_second": 0.251,
|
| 862 |
+
"learning_rate": 0.0001,
|
| 863 |
+
"step": 23584
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 44.78,
|
| 867 |
+
"learning_rate": 0.0001,
|
| 868 |
+
"loss": 0.1153,
|
| 869 |
+
"step": 24000
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 45.0,
|
| 873 |
+
"eval_accuracy": 0.571632725973562,
|
| 874 |
+
"eval_f1_macro": 0.8193083754252073,
|
| 875 |
+
"eval_f1_micro": 0.85338323172572,
|
| 876 |
+
"eval_loss": 0.10468064993619919,
|
| 877 |
+
"eval_roc_auc": 0.9025354227051225,
|
| 878 |
+
"eval_runtime": 732.8049,
|
| 879 |
+
"eval_samples_per_second": 3.82,
|
| 880 |
+
"eval_steps_per_second": 0.239,
|
| 881 |
+
"learning_rate": 0.0001,
|
| 882 |
+
"step": 24120
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 45.71,
|
| 886 |
+
"learning_rate": 0.0001,
|
| 887 |
+
"loss": 0.1167,
|
| 888 |
+
"step": 24500
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 46.0,
|
| 892 |
+
"eval_accuracy": 0.5666309396212933,
|
| 893 |
+
"eval_f1_macro": 0.822920967846307,
|
| 894 |
+
"eval_f1_micro": 0.8535311572700296,
|
| 895 |
+
"eval_loss": 0.10616234689950943,
|
| 896 |
+
"eval_roc_auc": 0.9080353429554627,
|
| 897 |
+
"eval_runtime": 730.6388,
|
| 898 |
+
"eval_samples_per_second": 3.831,
|
| 899 |
+
"eval_steps_per_second": 0.24,
|
| 900 |
+
"learning_rate": 0.0001,
|
| 901 |
+
"step": 24656
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 46.64,
|
| 905 |
+
"learning_rate": 0.0001,
|
| 906 |
+
"loss": 0.1162,
|
| 907 |
+
"step": 25000
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 47.0,
|
| 911 |
+
"eval_accuracy": 0.5694891032511611,
|
| 912 |
+
"eval_f1_macro": 0.8179688260529804,
|
| 913 |
+
"eval_f1_micro": 0.8521570528356762,
|
| 914 |
+
"eval_loss": 0.10597173124551773,
|
| 915 |
+
"eval_roc_auc": 0.9006400086530192,
|
| 916 |
+
"eval_runtime": 707.3305,
|
| 917 |
+
"eval_samples_per_second": 3.957,
|
| 918 |
+
"eval_steps_per_second": 0.247,
|
| 919 |
+
"learning_rate": 0.0001,
|
| 920 |
+
"step": 25192
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 47.57,
|
| 924 |
+
"learning_rate": 0.0001,
|
| 925 |
+
"loss": 0.1145,
|
| 926 |
+
"step": 25500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 48.0,
|
| 930 |
+
"eval_accuracy": 0.5744908896034298,
|
| 931 |
+
"eval_f1_macro": 0.8153624729345684,
|
| 932 |
+
"eval_f1_micro": 0.8529072240111792,
|
| 933 |
+
"eval_loss": 0.10410725325345993,
|
| 934 |
+
"eval_roc_auc": 0.9002018263609444,
|
| 935 |
+
"eval_runtime": 693.6803,
|
| 936 |
+
"eval_samples_per_second": 4.035,
|
| 937 |
+
"eval_steps_per_second": 0.252,
|
| 938 |
+
"learning_rate": 0.0001,
|
| 939 |
+
"step": 25728
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 48.51,
|
| 943 |
+
"learning_rate": 0.0001,
|
| 944 |
+
"loss": 0.1143,
|
| 945 |
+
"step": 26000
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 49.0,
|
| 949 |
+
"eval_accuracy": 0.5652018578063595,
|
| 950 |
+
"eval_f1_macro": 0.8181584192698232,
|
| 951 |
+
"eval_f1_micro": 0.8542054951001022,
|
| 952 |
+
"eval_loss": 0.10334235429763794,
|
| 953 |
+
"eval_roc_auc": 0.9042645729162101,
|
| 954 |
+
"eval_runtime": 700.9047,
|
| 955 |
+
"eval_samples_per_second": 3.993,
|
| 956 |
+
"eval_steps_per_second": 0.25,
|
| 957 |
+
"learning_rate": 0.0001,
|
| 958 |
+
"step": 26264
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 49.44,
|
| 962 |
+
"learning_rate": 0.0001,
|
| 963 |
+
"loss": 0.1129,
|
| 964 |
+
"step": 26500
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 50.0,
|
| 968 |
+
"eval_accuracy": 0.5719899964272954,
|
| 969 |
+
"eval_f1_macro": 0.8102121985255263,
|
| 970 |
+
"eval_f1_micro": 0.850807689945335,
|
| 971 |
+
"eval_loss": 0.10544609278440475,
|
| 972 |
+
"eval_roc_auc": 0.8956334123918851,
|
| 973 |
+
"eval_runtime": 741.7594,
|
| 974 |
+
"eval_samples_per_second": 3.773,
|
| 975 |
+
"eval_steps_per_second": 0.236,
|
| 976 |
+
"learning_rate": 0.0001,
|
| 977 |
+
"step": 26800
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 50.0,
|
| 981 |
+
"learning_rate": 0.0001,
|
| 982 |
+
"step": 26800,
|
| 983 |
+
"total_flos": 1.1878568451307192e+20,
|
| 984 |
+
"train_loss": 0.15577314120620045,
|
| 985 |
+
"train_runtime": 143618.003,
|
| 986 |
+
"train_samples_per_second": 2.983,
|
| 987 |
+
"train_steps_per_second": 0.187
|
| 988 |
+
}
|
| 989 |
+
],
|
| 990 |
+
"logging_steps": 500,
|
| 991 |
+
"max_steps": 26800,
|
| 992 |
+
"num_train_epochs": 50,
|
| 993 |
+
"save_steps": 500,
|
| 994 |
+
"total_flos": 1.1878568451307192e+20,
|
| 995 |
+
"trial_name": null,
|
| 996 |
+
"trial_params": null
|
| 997 |
+
}
|