AnAnonymousHuggingface/

PTMan-a-pre-trained-model-for-Manchu-speech-recognition

-----inference.py: Inference and computing WER/CER.

-----300M-pretrained: Model pretrained on Tungusic language speech.

-----300M-pretrained-finetuned: Model pretrained on Tungusic language speech and fine-tuned on Manchu parallel corpora.

-----w2v_processed_manchu_test_dataset: The test set of the dataset used in the paper. 

-----vocab.json: Dictionary file.