zkdeng
/

convnextv2-tiny-22k-384-finetuned-spiderTraining5-100

Generated from Trainer

Model card Files Files and versions Community

zkdeng commited on Aug 19

Commit

8b8c690

•

1 Parent(s): cabcb87

Model save

Files changed (1) hide show

README.md +12 -11

README.md CHANGED Viewed

@@ -20,11 +20,11 @@ should probably proofread and complete it, then remove this comment. -->
 This model is a fine-tuned version of [facebook/convnextv2-tiny-22k-384](https://huggingface.co/facebook/convnextv2-tiny-22k-384) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4642
-- Accuracy: 0.9
-- Precision: 0.9035
-- Recall: 0.9103
-- F1: 0.9024
 ## Model description
@@ -43,10 +43,11 @@ More information needed
 ### Training hyperparameters
 The following hyperparameters were used during training:
-- learning_rate: 5e-05
 - train_batch_size: 16
 - eval_batch_size: 16
 - seed: 42
 - gradient_accumulation_steps: 4
 - total_train_batch_size: 64
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
@@ -58,11 +59,11 @@ The following hyperparameters were used during training:
 | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
-| No log        | 0.96  | 6    | 1.1802          | 0.68     | 0.6789    | 0.6869 | 0.6751 |
-| 1.3233        | 1.92  | 12   | 0.7867          | 0.84     | 0.8603    | 0.8570 | 0.8388 |
-| 1.3233        | 2.88  | 18   | 0.5761          | 0.9      | 0.9035    | 0.9103 | 0.9024 |
-| 0.6617        | 4.0   | 25   | 0.4823          | 0.9      | 0.9035    | 0.9103 | 0.9024 |
-| 0.4323        | 4.8   | 30   | 0.4642          | 0.9      | 0.9035    | 0.9103 | 0.9024 |
 ### Framework versions

 This model is a fine-tuned version of [facebook/convnextv2-tiny-22k-384](https://huggingface.co/facebook/convnextv2-tiny-22k-384) on an unknown dataset.
 It achieves the following results on the evaluation set:
+- Loss: 0.7208
+- Accuracy: 0.88
+- Precision: 0.8822
+- Recall: 0.8936
+- F1: 0.8805
 ## Model description
 ### Training hyperparameters
 The following hyperparameters were used during training:
+- learning_rate: 3e-05
 - train_batch_size: 16
 - eval_batch_size: 16
 - seed: 42
+- distributed_type: multi-GPU
 - gradient_accumulation_steps: 4
 - total_train_batch_size: 64
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
+| No log        | 0.96  | 6    | 1.3602          | 0.52     | 0.5198    | 0.5390 | 0.5169 |
+| 1.4377        | 1.92  | 12   | 1.0444          | 0.78     | 0.7903    | 0.7869 | 0.7804 |
+| 1.4377        | 2.88  | 18   | 0.8560          | 0.86     | 0.8636    | 0.8736 | 0.8598 |
+| 0.9338        | 4.0   | 25   | 0.7443          | 0.86     | 0.8636    | 0.8736 | 0.8598 |
+| 0.7051        | 4.8   | 30   | 0.7208          | 0.88     | 0.8822    | 0.8936 | 0.8805 |
 ### Framework versions