update model card README.md

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: allenai/scibert_scivocab_uncased
+base_model: allenai/longformer-large-4096
 tags:
 - generated_from_trainer
 metrics:
@@ -14,10 +14,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # patentClassfication3
 
-This model is a fine-tuned version of [allenai/scibert_scivocab_uncased](https://huggingface.co/allenai/scibert_scivocab_uncased) on the None dataset.
+This model is a fine-tuned version of [allenai/longformer-large-4096](https://huggingface.co/allenai/longformer-large-4096) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5828
-- Accuracy: 0.6901
+- Loss: 0.5709
+- Accuracy: 0.7283
 
 ## Model description
 
@@ -36,24 +36,25 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 2.51444e-05
+- learning_rate: 1.4489735300181872e-05
 - train_batch_size: 8
 - eval_batch_size: 8
-- seed: 61
-- gradient_accumulation_steps: 8
-- total_train_batch_size: 64
+- seed: 3
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
 - lr_scheduler_warmup_steps: 240
-- num_epochs: 2
+- num_epochs: 5
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Accuracy |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.6511        | 1.0   | 554  | 0.6841          | 0.6125   |
-| 0.5721        | 2.0   | 1108 | 0.5828          | 0.6901   |
+| Training Loss | Epoch | Step  | Validation Loss | Accuracy |
+|:-------------:|:-----:|:-----:|:---------------:|:--------:|
+| 0.6069        | 1.0   | 4438  | 0.5878          | 0.6927   |
+| 0.559         | 2.0   | 8876  | 0.5991          | 0.7026   |
+| 0.5133        | 3.0   | 13314 | 0.5709          | 0.7283   |
+| 0.4348        | 4.0   | 17752 | 0.6091          | 0.7228   |
+| 0.376         | 5.0   | 22190 | 0.7537          | 0.7117   |
 
 
 ### Framework versions