https://huggingface.co/openai/whisper-small.en with ONNX weights to be compatible with Transformers.js.
If you haven't already, you can install the Transformers.js JavaScript library from NPM using:
npm i @huggingface/transformers
Example: Transcribe audio from a URL.
import { pipeline } from '@huggingface/transformers';
const transcriber = await pipeline('automatic-speech-recognition', 'Xenova/whisper-small.en');
const url = 'https://huggingface.co/datasets/Xenova/transformers.js-docs/resolve/main/jfk.wav';
const output = await transcriber(url);
Note: Having a separate repo for ONNX weights is intended to be a temporary solution until WebML gains more traction. If you would like to make your models web-ready, we recommend converting to ONNX using 🤗 Optimum and structuring your repo like this one (with ONNX weights located in a subfolder named onnx).
onnx/decoder_model_merged.onnx 615 MBonnx/decoder_model.onnx 615 MBonnx/decoder_with_past_model.onnx 558 MBonnx/encoder_model.onnx 353 MBonnx/decoder_model_uint8.onnx 315 MBonnx/decoder_model_int8.onnx 315 MBonnx/decoder_model_merged_fp16.onnx 309 MBonnx/decoder_model_fp16.onnx 308 MBonnx/decoder_with_past_model_uint8.onnx 301 MBonnx/decoder_with_past_model_int8.onnx 301 MBonnx/decoder_with_past_model_fp16.onnx 279 MBonnx/decoder_model_merged_q4.onnx 233 MBonnx/decoder_model_q4.onnx 233 MBonnx/decoder_model_merged_bnb4.onnx 226 MBonnx/decoder_model_bnb4.onnx 226 MBonnx/decoder_with_past_model_q4.onnx 224 MBonnx/decoder_with_past_model_bnb4.onnx 218 MBonnx/encoder_model_fp16.onnx 177 MBonnx/decoder_model_merged_uint8.onnx 157 MBonnx/decoder_model_merged_int8.onnx 157 MBonnx/decoder_model_merged_quantized.onnx 157 MBonnx/decoder_model_quantized.onnx 156 MBonnx/decoder_model_merged_q4f16.onnx 146 MBonnx/decoder_model_q4f16.onnx 145 MBonnx/decoder_with_past_model_quantized.onnx 142 MBonnx/decoder_with_past_model_q4f16.onnx 137 MBonnx/encoder_model_quantized.onnx 92 MBonnx/encoder_model_uint8.onnx 92 MBonnx/encoder_model_q4.onnx 66 MBonnx/encoder_model_bnb4.onnx 61 MBonnx/encoder_model_q4f16.onnx 54 MBopenai/whisper-small.en