jointriple/brand_classification_2_20240626_model_1

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 4.9735
-- Accuracy: 0.7396
+- Loss: 5.0555
+- Accuracy: 0.7284
 
 ## Model description
 
@@ -38,24 +38,26 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 2e-05
-- train_batch_size: 128
-- eval_batch_size: 128
+- train_batch_size: 256
+- eval_batch_size: 256
 - seed: 42
 - gradient_accumulation_steps: 4
-- total_train_batch_size: 512
+- total_train_batch_size: 1024
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 5
+- num_epochs: 7
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 7.3089        | 1.0   | 592  | 6.6011          | 0.1957   |
-| 6.5028        | 2.0   | 1185 | 5.8580          | 0.5069   |
-| 5.9384        | 3.0   | 1778 | 5.3658          | 0.6636   |
-| 5.5484        | 4.0   | 2371 | 5.0733          | 0.7220   |
-| 5.2914        | 4.99  | 2960 | 4.9735          | 0.7396   |
+| 7.0753        | 1.0   | 296  | 6.9521          | 0.0573   |
+| 6.4689        | 2.0   | 593  | 6.3053          | 0.2966   |
+| 6.0338        | 3.0   | 889  | 5.8345          | 0.5124   |
+| 5.6788        | 4.0   | 1186 | 5.4963          | 0.6125   |
+| 5.4413        | 5.0   | 1482 | 5.2485          | 0.6947   |
+| 5.271         | 6.0   | 1779 | 5.1036          | 0.7218   |
+| 5.2718        | 6.99  | 2072 | 5.0555          | 0.7284   |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2cd84ec27ed1b68fa4dac847eab3e1dc438a51b0c24f106a3807cb99853778cd
+oid sha256:d98abd5f9e5ee134b10775868be7bdd923fe4d5cab6c90a4512d6a8f90f41e0b
 size 444369032

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9021cc0d929ee4828a226a2f43d69af6d3f6aafd4a19569edb4fc01d91b6b1c2
-size 4920
+oid sha256:05c5a36bdb092e119bcffd9c5856a3800ec61f39b9204079b5281a2b6a55892e
+size 4856