whisper-web
whisper-web copied to clipboard
xenova
ML-powered speech recognition directly in your browser
I want to cry so bad, so far I haven't found out how onnx c++ runs word level timestamps。Someone please show me the way.
Hello, I’m using whisper-web (experimental-webgpu branch) with local models, (env.allowLocalModels = true and env.localModelPath = "./models"), and facing challenges in setting distinct dtype values for encoder_model and decoder_model_merged with a...
It's more of a question rather than an issue with the repo, but I tried to build a solution that basically mirrors what this does transcription capability wise, but when...
Added lightweight sentiment analysis for the transcribed text snippets and breakdown of sentiments used in the recording. Model used: `Cohee/distilbert-base-uncased-go-emotions-onnx`
I test on the demo site. I record in Marathi language. In the settings, I select "Multilingual" and set the source language to Marathi and select task to perform as...
Every time I restart the browser, I need to complete the download of the model file again, which takes a long time. How can I specify a directory for the...
