Update README.md

README.md

@@ -1,35 +1,55 @@
 ---
 license: apache-2.0
+language:
+- vi
 tags:
 - automatic-speech-recognition
-- geninhu/fpt-vi
-- generated_from_trainer
+- robust-speech-event
+- common-voice
 model-index:
 - name: xls-asr-vi-40h-1B
-  results: []
+  results:
+  - task: 
+      name: Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: Common Voice 7.0
+      type: mozilla-foundation/common_voice_7_0
+
+      args: vi
+    metrics:
+       - name: Test WER
+         type: wer
+         value: 34.210
+       - name: Test CER
+         type: cer
+         value: 19.938
 ---
 
-<!-- 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. -->
 
 # xls-asr-vi-40h-1B
 
-This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the GENINHU/FPT-VI - NA dataset.
-It achieves the following results on the evaluation set:
-- Loss: 4.1691
-- Wer: 0.4133
+This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the 40 hours of Vietnamese ASR data, including common_voice 7.0 and private dataset.
 
-## Model description
+### Benchmark WER result:
+| | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 7.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) |
+|---|---|---|
+|without LM| 25.93 | 34.21 |
+|with 4-grams LM|  | |
 
-More information needed
+### Benchmark CER result:
+| | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 7.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) |
+|---|---|---|
+|without LM| 9.243 | 19.938 |
+|with 4-grams LM|  | |
 
-## Intended uses & limitations
+## Evaluation
 
-More information needed
+Please use eval.py file to run evaluation
 
-## Training and evaluation data
-
-More information needed
+```python
+python eval_custom.py --model_id geninhu/xls-asr-vi-40h-1B --dataset mozilla-foundation/common_voice_7_0 --config vi --split test --log_outputs
+```
 
 ## Training procedure