Update README.md

README.md

@@ -1,5 +1,10 @@
 ---
 license: apache-2.0
+language:
+- zh
+metrics:
+- accuracy
+pipeline_tag: audio-classification
 ---
 # hubert-base-ch-speech-emotion-recognition
 This model uses [TencentGameMate/chinese-hubert-base]([TencentGameMate/chinese-hubert-base · Hugging Face](https://huggingface.co/TencentGameMate/chinese-hubert-base)) as the pre-training model for training on the CASIA dataset.
@@ -140,4 +145,4 @@ predict(path, processor, model)
 
 *Loss curve of training set and verification set*
 
-<div> <img src="https://huggingface.co/xmj2002/hubert-base-ch-speech-emotion-recognition/resolve/main/loss.png" width = 80%/> </div>
+<div> <img src="https://huggingface.co/xmj2002/hubert-base-ch-speech-emotion-recognition/resolve/main/loss.png" width = 80%/> </div>