Training completed!

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: vinai/phobert-base-v2
+base_model: vinai/phobert-base
 tags:
 - generated_from_trainer
 metrics:
@@ -15,11 +15,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 # PhoBERT-cls-OCR
 
-This model is a fine-tuned version of [vinai/phobert-base-v2](https://huggingface.co/vinai/phobert-base-v2) on the None dataset.
+This model is a fine-tuned version of [vinai/phobert-base](https://huggingface.co/vinai/phobert-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6475
-- Accuracy: 0.8416
-- F1: 0.8422
+- Loss: 0.4288
+- Accuracy: 0.8812
+- F1: 0.8812
 
 ## Model description
 
@@ -44,22 +44,19 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 10
+- num_epochs: 7
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
-| 0.5849        | 1.0   | 25   | 0.4385          | 0.8218   | 0.8225 |
-| 0.3382        | 2.0   | 50   | 0.3785          | 0.8614   | 0.8599 |
-| 0.1897        | 3.0   | 75   | 0.4339          | 0.8515   | 0.8511 |
-| 0.1101        | 4.0   | 100  | 0.4626          | 0.8614   | 0.8599 |
-| 0.0798        | 5.0   | 125  | 0.5006          | 0.8713   | 0.8703 |
-| 0.0547        | 6.0   | 150  | 0.5670          | 0.8614   | 0.8614 |
-| 0.034         | 7.0   | 175  | 0.5568          | 0.8812   | 0.8806 |
-| 0.0121        | 8.0   | 200  | 0.6414          | 0.8317   | 0.8326 |
-| 0.0207        | 9.0   | 225  | 0.6713          | 0.8416   | 0.8427 |
-| 0.0203        | 10.0  | 250  | 0.6475          | 0.8416   | 0.8422 |
+| 0.5728        | 1.0   | 25   | 0.4112          | 0.8614   | 0.8590 |
+| 0.2892        | 2.0   | 50   | 0.3444          | 0.8515   | 0.8511 |
+| 0.1954        | 3.0   | 75   | 0.3638          | 0.8812   | 0.8816 |
+| 0.1387        | 4.0   | 100  | 0.3591          | 0.8812   | 0.8806 |
+| 0.1029        | 5.0   | 125  | 0.3809          | 0.8911   | 0.8908 |
+| 0.053         | 6.0   | 150  | 0.4145          | 0.8911   | 0.8903 |
+| 0.0527        | 7.0   | 175  | 0.4288          | 0.8812   | 0.8812 |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "vinai/phobert-base-v2",
+  "_name_or_path": "vinai/phobert-base",
   "architectures": [
     "RobertaForSequenceClassification"
   ],
@@ -7,6 +7,7 @@
   "bos_token_id": 0,
   "classifier_dropout": null,
   "eos_token_id": 2,
+  "gradient_checkpointing": false,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
   "hidden_size": 768,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a8511396f9ea5184235dd616f8f03bb00371169732c46fca287c6c7ce39b5ce1
+oid sha256:f4d680c566094d14f0c7d2318c1469000748ad8e0e4770240d47159617aa60fb
 size 540068209

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:73494ef25fd4750f7db3c8f601fa2adece4cbce7b471d0e3d9a1dfecdb6bbf52
+oid sha256:80326e334c5466457ee484bccbf56053310faea7cd1288387e41491a6158429b
 size 4027