End of training

README.md

@@ -9,13 +9,13 @@ model-index:
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/rayranger/huggingface/runs/9t8e6qv5)
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/rayranger/huggingface/runs/q0t4izjn)
 # khmer-trocr-base-printed
 
 This model was trained from scratch on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0783
-- Cer: 0.5191
+- Loss: 0.0635
+- Cer: 0.5175
 
 ## Model description
 
@@ -47,14 +47,14 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss | Cer    |
 |:-------------:|:------:|:----:|:---------------:|:------:|
-| 0.2278        | 0.4876 | 1000 | 0.2244          | 0.6099 |
-| 0.1517        | 0.9751 | 2000 | 0.1667          | 0.5711 |
-| 0.1147        | 1.4627 | 3000 | 0.1412          | 0.5578 |
-| 0.0927        | 1.9503 | 4000 | 0.1159          | 0.5552 |
-| 0.0669        | 2.4378 | 5000 | 0.1028          | 0.5458 |
-| 0.0551        | 2.9254 | 6000 | 0.0901          | 0.5270 |
-| 0.0304        | 3.4130 | 7000 | 0.0833          | 0.5170 |
-| 0.0182        | 3.9005 | 8000 | 0.0783          | 0.5191 |
+| 0.0662        | 0.4876 | 1000 | 0.1256          | 0.5637 |
+| 0.0435        | 0.9751 | 2000 | 0.1191          | 0.5692 |
+| 0.0412        | 1.4627 | 3000 | 0.0978          | 0.5385 |
+| 0.0275        | 1.9503 | 4000 | 0.0896          | 0.5530 |
+| 0.0212        | 2.4378 | 5000 | 0.0825          | 0.5518 |
+| 0.0155        | 2.9254 | 6000 | 0.0732          | 0.5297 |
+| 0.0084        | 3.4130 | 7000 | 0.0676          | 0.5107 |
+| 0.005         | 3.9005 | 8000 | 0.0635          | 0.5175 |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/kaggle/input/khmerocr-trocr-base/KhmerOCR_TrOCR",
+  "_name_or_path": "/kaggle/input/khmerocr-trocr-fine-tune-v1/KhmerOCR_TrOCR",
   "architectures": [
     "VisionEncoderDecoderModel"
   ],

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:480b91120705e08bbee9571fe7f7180c7342f31264384a18b789f35fdcbc7b88
+oid sha256:a83faf51a7f8ba96f4a9a47b864002fbeab6383f4db8c2ff2601c0ae77de5f30
 size 608335608

tokenizer_config.json

@@ -49,11 +49,18 @@
   "errors": "replace",
   "mask_token": "<mask>",
   "max_len": 150,
+  "max_length": 145,
   "model_max_length": 150,
+  "pad_to_multiple_of": null,
   "pad_token": "<pad>",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
   "processor_class": "TrOCRProcessor",
   "sep_token": "</s>",
+  "stride": 0,
   "tokenizer_class": "RobertaTokenizer",
   "trim_offsets": true,
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
   "unk_token": "<unk>"
 }