trungkienbkhn

For more information, I executed some benchmarks for Faster whisper with FlashAttention in [here](https://github.com/SYSTRAN/faster-whisper/issues/598#issuecomment-2158053951).

FYI, FW does not support change subtitle length. Sometimes model creates segments for 30s because it doesn't detect audio well and can't split the segments, maybe your audio has a...

@Arche151 , no. You need to convert it through Ctranslate2. Example: ``` ct2-transformers-converter --model sanchit-gandhi/distil-whisper-large-v3-de-kd --output_dir distil-whisper-large-v3-de-kd-ct2 --copy_files tokenizer.json preprocessor_config.json --quantization float16 ``` Then, when initializing the Whisper model, you...

Since your model was converted from Distil model, so you should add option `condition_on_previous_text=False` when transcribing. For more info, see this comment: https://github.com/SYSTRAN/faster-whisper/pull/557#issuecomment-1837394755

@Arche151 , could you try again with compute_type="default" (or remove this command when initializing whisper model) ?

@tuocheng0824 I just downloaded this model and everything is still working properly, please verify your Internet connection and try again. ``` config.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 2.39k/2.39k [00:00

@prem1303 , hello. FW 0.10.0 is broken tag, could you update FW and Ctranslate2 to latest version (FW 1.0.1 + Ctranslate2 4.2.0 + CUDA 12) ?

@prem1303 Could you show your code and attach example audio ? I will try testing them to evaluate throughput. I think you could try using smaller models or distil models...

@George0828Zhang , hello. Tks for your idea. I created a new [PR](https://github.com/SYSTRAN/faster-whisper/pull/807) to implement this.

@George0828Zhang , I think that if you want to handle tokenizer and preprocessor with other initialization data, you could edit tokenizer.json and preprocessor_config.json files in your custom FW model instead...