Upload model with metrics - 2024-12-12 01:56
Browse files- README.md +24 -7
- checkpoint-35/model.safetensors +1 -1
- checkpoint-35/trainer_state.json +19 -19
- checkpoint-35/training_args.bin +3 -0
- checkpoint-64/model.safetensors +1 -1
- checkpoint-64/trainer_state.json +19 -19
- checkpoint-64/training_args.bin +3 -0
- metrics.json +19 -20
- model.safetensors +1 -1
- trainer_state.json +23 -23
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -30,6 +30,30 @@ This model is trained for document classification using vision transformers (DiT
|
|
| 30 |
* Learning Rate Schedule: cosine_with_restarts
|
| 31 |
* Warmup Ratio: 0.1
|
| 32 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
## Usage
|
| 34 |
|
| 35 |
```python
|
|
@@ -48,10 +72,3 @@ inputs = processor(image, return_tensors="pt")
|
|
| 48 |
outputs = model(**inputs)
|
| 49 |
predicted_label = outputs.logits.argmax(-1).item()
|
| 50 |
```
|
| 51 |
-
|
| 52 |
-
## Model Performance
|
| 53 |
-
|
| 54 |
-
The model was evaluated on a held-out test set. The metrics above show the model's performance on both training and evaluation data. Key metrics to note:
|
| 55 |
-
- Accuracy: Overall classification accuracy
|
| 56 |
-
- F1 Score: Weighted average F1-score across all classes
|
| 57 |
-
- Precision and Recall: Weighted averages across all classes
|
|
|
|
| 30 |
* Learning Rate Schedule: cosine_with_restarts
|
| 31 |
* Warmup Ratio: 0.1
|
| 32 |
|
| 33 |
+
## Training and Evaluation Metrics
|
| 34 |
+
### Training Metrics
|
| 35 |
+
* Loss: 5.2299
|
| 36 |
+
* Grad Norm: 5.5809
|
| 37 |
+
* Learning Rate: 0.0010
|
| 38 |
+
* Epoch: 0.5426
|
| 39 |
+
* Step: 35.0000
|
| 40 |
+
|
| 41 |
+
### Evaluation Metrics
|
| 42 |
+
* Loss: 5.0066
|
| 43 |
+
* Accuracy: 0.1351
|
| 44 |
+
* Weighted F1: 0.0662
|
| 45 |
+
* Micro F1: 0.1351
|
| 46 |
+
* Macro F1: 0.0380
|
| 47 |
+
* Weighted Recall: 0.1351
|
| 48 |
+
* Micro Recall: 0.1351
|
| 49 |
+
* Macro Recall: 0.0646
|
| 50 |
+
* Weighted Precision: 0.0548
|
| 51 |
+
* Micro Precision: 0.1351
|
| 52 |
+
* Macro Precision: 0.0333
|
| 53 |
+
* Runtime: 8.2574
|
| 54 |
+
* Samples Per Second: 441.1820
|
| 55 |
+
* Steps Per Second: 3.5120
|
| 56 |
+
|
| 57 |
## Usage
|
| 58 |
|
| 59 |
```python
|
|
|
|
| 72 |
outputs = model(**inputs)
|
| 73 |
predicted_label = outputs.logits.argmax(-1).item()
|
| 74 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-35/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1213526036
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e72a88f97c761c0a531cb086303001d50691997a2b12f53e5ba97ac052c09560
|
| 3 |
size 1213526036
|
checkpoint-35/trainer_state.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "ds3-img-classification/checkpoint-35",
|
| 4 |
"epoch": 0.5426356589147286,
|
| 5 |
"eval_steps": 35,
|
|
@@ -10,34 +10,34 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.015503875968992248,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 0.0002857142857142857,
|
| 15 |
-
"loss":
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.5426356589147286,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 0.0010275543423681622,
|
| 22 |
-
"loss": 5.
|
| 23 |
"step": 35
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.5426356589147286,
|
| 27 |
-
"eval_accuracy": 0.
|
| 28 |
-
"eval_loss": 5.
|
| 29 |
-
"eval_macro_f1": 0.
|
| 30 |
-
"eval_macro_precision": 0.
|
| 31 |
-
"eval_macro_recall": 0.
|
| 32 |
-
"eval_micro_f1": 0.
|
| 33 |
-
"eval_micro_precision": 0.
|
| 34 |
-
"eval_micro_recall": 0.
|
| 35 |
-
"eval_runtime": 8.
|
| 36 |
-
"eval_samples_per_second":
|
| 37 |
-
"eval_steps_per_second": 3.
|
| 38 |
-
"eval_weighted_f1": 0.
|
| 39 |
-
"eval_weighted_precision": 0.
|
| 40 |
-
"eval_weighted_recall": 0.
|
| 41 |
"step": 35
|
| 42 |
}
|
| 43 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.1350535273126544,
|
| 3 |
"best_model_checkpoint": "ds3-img-classification/checkpoint-35",
|
| 4 |
"epoch": 0.5426356589147286,
|
| 5 |
"eval_steps": 35,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.015503875968992248,
|
| 13 |
+
"grad_norm": 16.251585006713867,
|
| 14 |
"learning_rate": 0.0002857142857142857,
|
| 15 |
+
"loss": 6.0665,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.5426356589147286,
|
| 20 |
+
"grad_norm": 5.580934524536133,
|
| 21 |
"learning_rate": 0.0010275543423681622,
|
| 22 |
+
"loss": 5.2299,
|
| 23 |
"step": 35
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.5426356589147286,
|
| 27 |
+
"eval_accuracy": 0.1350535273126544,
|
| 28 |
+
"eval_loss": 5.006561279296875,
|
| 29 |
+
"eval_macro_f1": 0.038001266460012215,
|
| 30 |
+
"eval_macro_precision": 0.033336047367919855,
|
| 31 |
+
"eval_macro_recall": 0.06460349403258135,
|
| 32 |
+
"eval_micro_f1": 0.1350535273126544,
|
| 33 |
+
"eval_micro_precision": 0.1350535273126544,
|
| 34 |
+
"eval_micro_recall": 0.1350535273126544,
|
| 35 |
+
"eval_runtime": 8.2574,
|
| 36 |
+
"eval_samples_per_second": 441.182,
|
| 37 |
+
"eval_steps_per_second": 3.512,
|
| 38 |
+
"eval_weighted_f1": 0.0662231556079451,
|
| 39 |
+
"eval_weighted_precision": 0.054814565244902975,
|
| 40 |
+
"eval_weighted_recall": 0.1350535273126544,
|
| 41 |
"step": 35
|
| 42 |
}
|
| 43 |
],
|
checkpoint-35/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6cc8ffd94025c9e00fadf213825a92f261358a56e78b613f233178b10436ef6
|
| 3 |
+
size 5368
|
checkpoint-64/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1213526036
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df2901e8559ae0de2bd216d039cc92f33736cca45db6be8c08de0bbfb4fdb663
|
| 3 |
size 1213526036
|
checkpoint-64/trainer_state.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "ds3-img-classification/checkpoint-35",
|
| 4 |
"epoch": 0.9922480620155039,
|
| 5 |
"eval_steps": 35,
|
|
@@ -10,34 +10,34 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.015503875968992248,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 0.0002857142857142857,
|
| 15 |
-
"loss":
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.5426356589147286,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 0.0010275543423681622,
|
| 22 |
-
"loss": 5.
|
| 23 |
"step": 35
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.5426356589147286,
|
| 27 |
-
"eval_accuracy": 0.
|
| 28 |
-
"eval_loss": 5.
|
| 29 |
-
"eval_macro_f1": 0.
|
| 30 |
-
"eval_macro_precision": 0.
|
| 31 |
-
"eval_macro_recall": 0.
|
| 32 |
-
"eval_micro_f1": 0.
|
| 33 |
-
"eval_micro_precision": 0.
|
| 34 |
-
"eval_micro_recall": 0.
|
| 35 |
-
"eval_runtime": 8.
|
| 36 |
-
"eval_samples_per_second":
|
| 37 |
-
"eval_steps_per_second": 3.
|
| 38 |
-
"eval_weighted_f1": 0.
|
| 39 |
-
"eval_weighted_precision": 0.
|
| 40 |
-
"eval_weighted_recall": 0.
|
| 41 |
"step": 35
|
| 42 |
}
|
| 43 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.1350535273126544,
|
| 3 |
"best_model_checkpoint": "ds3-img-classification/checkpoint-35",
|
| 4 |
"epoch": 0.9922480620155039,
|
| 5 |
"eval_steps": 35,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.015503875968992248,
|
| 13 |
+
"grad_norm": 16.251585006713867,
|
| 14 |
"learning_rate": 0.0002857142857142857,
|
| 15 |
+
"loss": 6.0665,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.5426356589147286,
|
| 20 |
+
"grad_norm": 5.580934524536133,
|
| 21 |
"learning_rate": 0.0010275543423681622,
|
| 22 |
+
"loss": 5.2299,
|
| 23 |
"step": 35
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.5426356589147286,
|
| 27 |
+
"eval_accuracy": 0.1350535273126544,
|
| 28 |
+
"eval_loss": 5.006561279296875,
|
| 29 |
+
"eval_macro_f1": 0.038001266460012215,
|
| 30 |
+
"eval_macro_precision": 0.033336047367919855,
|
| 31 |
+
"eval_macro_recall": 0.06460349403258135,
|
| 32 |
+
"eval_micro_f1": 0.1350535273126544,
|
| 33 |
+
"eval_micro_precision": 0.1350535273126544,
|
| 34 |
+
"eval_micro_recall": 0.1350535273126544,
|
| 35 |
+
"eval_runtime": 8.2574,
|
| 36 |
+
"eval_samples_per_second": 441.182,
|
| 37 |
+
"eval_steps_per_second": 3.512,
|
| 38 |
+
"eval_weighted_f1": 0.0662231556079451,
|
| 39 |
+
"eval_weighted_precision": 0.054814565244902975,
|
| 40 |
+
"eval_weighted_recall": 0.1350535273126544,
|
| 41 |
"step": 35
|
| 42 |
}
|
| 43 |
],
|
checkpoint-64/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6cc8ffd94025c9e00fadf213825a92f261358a56e78b613f233178b10436ef6
|
| 3 |
+
size 5368
|
metrics.json
CHANGED
|
@@ -1,26 +1,25 @@
|
|
| 1 |
{
|
| 2 |
"train": {
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
-
"
|
| 8 |
-
"epoch": 0.9922480620155039
|
| 9 |
},
|
| 10 |
"eval": {
|
| 11 |
-
"eval_loss": 5.
|
| 12 |
-
"eval_accuracy": 0.
|
| 13 |
-
"eval_weighted_f1": 0.
|
| 14 |
-
"eval_micro_f1": 0.
|
| 15 |
-
"eval_macro_f1": 0.
|
| 16 |
-
"eval_weighted_recall": 0.
|
| 17 |
-
"eval_micro_recall": 0.
|
| 18 |
-
"eval_macro_recall": 0.
|
| 19 |
-
"eval_weighted_precision": 0.
|
| 20 |
-
"eval_micro_precision": 0.
|
| 21 |
-
"eval_macro_precision": 0.
|
| 22 |
-
"eval_runtime": 8.
|
| 23 |
-
"eval_samples_per_second":
|
| 24 |
-
"eval_steps_per_second": 3.
|
| 25 |
}
|
| 26 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"train": {
|
| 3 |
+
"loss": 5.2299,
|
| 4 |
+
"grad_norm": 5.580934524536133,
|
| 5 |
+
"learning_rate": 0.0010275543423681622,
|
| 6 |
+
"epoch": 0.5426356589147286,
|
| 7 |
+
"step": 35.0
|
|
|
|
| 8 |
},
|
| 9 |
"eval": {
|
| 10 |
+
"eval_loss": 5.006561279296875,
|
| 11 |
+
"eval_accuracy": 0.1350535273126544,
|
| 12 |
+
"eval_weighted_f1": 0.0662231556079451,
|
| 13 |
+
"eval_micro_f1": 0.1350535273126544,
|
| 14 |
+
"eval_macro_f1": 0.038001266460012215,
|
| 15 |
+
"eval_weighted_recall": 0.1350535273126544,
|
| 16 |
+
"eval_micro_recall": 0.1350535273126544,
|
| 17 |
+
"eval_macro_recall": 0.06460349403258135,
|
| 18 |
+
"eval_weighted_precision": 0.054814565244902975,
|
| 19 |
+
"eval_micro_precision": 0.1350535273126544,
|
| 20 |
+
"eval_macro_precision": 0.033336047367919855,
|
| 21 |
+
"eval_runtime": 8.2574,
|
| 22 |
+
"eval_samples_per_second": 441.182,
|
| 23 |
+
"eval_steps_per_second": 3.512
|
| 24 |
}
|
| 25 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1213526036
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e72a88f97c761c0a531cb086303001d50691997a2b12f53e5ba97ac052c09560
|
| 3 |
size 1213526036
|
trainer_state.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "ds3-img-classification/checkpoint-35",
|
| 4 |
"epoch": 0.9922480620155039,
|
| 5 |
"eval_steps": 35,
|
|
@@ -10,44 +10,44 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.015503875968992248,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 0.0002857142857142857,
|
| 15 |
-
"loss":
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.5426356589147286,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 0.0010275543423681622,
|
| 22 |
-
"loss": 5.
|
| 23 |
"step": 35
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.5426356589147286,
|
| 27 |
-
"eval_accuracy": 0.
|
| 28 |
-
"eval_loss": 5.
|
| 29 |
-
"eval_macro_f1": 0.
|
| 30 |
-
"eval_macro_precision": 0.
|
| 31 |
-
"eval_macro_recall": 0.
|
| 32 |
-
"eval_micro_f1": 0.
|
| 33 |
-
"eval_micro_precision": 0.
|
| 34 |
-
"eval_micro_recall": 0.
|
| 35 |
-
"eval_runtime": 8.
|
| 36 |
-
"eval_samples_per_second":
|
| 37 |
-
"eval_steps_per_second": 3.
|
| 38 |
-
"eval_weighted_f1": 0.
|
| 39 |
-
"eval_weighted_precision": 0.
|
| 40 |
-
"eval_weighted_recall": 0.
|
| 41 |
"step": 35
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 0.9922480620155039,
|
| 45 |
"step": 64,
|
| 46 |
"total_flos": 8.978215898519175e+18,
|
| 47 |
-
"train_loss": 4.
|
| 48 |
-
"train_runtime": 163.
|
| 49 |
-
"train_samples_per_second": 200.
|
| 50 |
-
"train_steps_per_second": 0.
|
| 51 |
}
|
| 52 |
],
|
| 53 |
"logging_steps": 35,
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.1350535273126544,
|
| 3 |
"best_model_checkpoint": "ds3-img-classification/checkpoint-35",
|
| 4 |
"epoch": 0.9922480620155039,
|
| 5 |
"eval_steps": 35,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.015503875968992248,
|
| 13 |
+
"grad_norm": 16.251585006713867,
|
| 14 |
"learning_rate": 0.0002857142857142857,
|
| 15 |
+
"loss": 6.0665,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.5426356589147286,
|
| 20 |
+
"grad_norm": 5.580934524536133,
|
| 21 |
"learning_rate": 0.0010275543423681622,
|
| 22 |
+
"loss": 5.2299,
|
| 23 |
"step": 35
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.5426356589147286,
|
| 27 |
+
"eval_accuracy": 0.1350535273126544,
|
| 28 |
+
"eval_loss": 5.006561279296875,
|
| 29 |
+
"eval_macro_f1": 0.038001266460012215,
|
| 30 |
+
"eval_macro_precision": 0.033336047367919855,
|
| 31 |
+
"eval_macro_recall": 0.06460349403258135,
|
| 32 |
+
"eval_micro_f1": 0.1350535273126544,
|
| 33 |
+
"eval_micro_precision": 0.1350535273126544,
|
| 34 |
+
"eval_micro_recall": 0.1350535273126544,
|
| 35 |
+
"eval_runtime": 8.2574,
|
| 36 |
+
"eval_samples_per_second": 441.182,
|
| 37 |
+
"eval_steps_per_second": 3.512,
|
| 38 |
+
"eval_weighted_f1": 0.0662231556079451,
|
| 39 |
+
"eval_weighted_precision": 0.054814565244902975,
|
| 40 |
+
"eval_weighted_recall": 0.1350535273126544,
|
| 41 |
"step": 35
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 0.9922480620155039,
|
| 45 |
"step": 64,
|
| 46 |
"total_flos": 8.978215898519175e+18,
|
| 47 |
+
"train_loss": 4.950647212564945,
|
| 48 |
+
"train_runtime": 163.8995,
|
| 49 |
+
"train_samples_per_second": 200.037,
|
| 50 |
+
"train_steps_per_second": 0.39
|
| 51 |
}
|
| 52 |
],
|
| 53 |
"logging_steps": 35,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6cc8ffd94025c9e00fadf213825a92f261358a56e78b613f233178b10436ef6
|
| 3 |
size 5368
|