Fastest execution; struggles heavily with accents and background noise.
For multilingual audio where you want the output translated into English, simply append the translation flag: ./main -m models/ggml-medium.bin -f output.wav -tr Use code with caution. Optimizing Performance
Alternatively, if you have cloned the repository, use the included shell script: sh ./models/download-ggml-model.sh medium Use code with caution. 2. Run the Model ggml-medium.bin
The ggml-medium.bin file is a specific, pre-trained model checkpoint of OpenAI’s Whisper "Medium" model. It has been converted and quantized into the (now largely succeeded by and integrated into GGUF ecosystem developments, though still widely referred to by its original binary name in Whisper ecosystems).
./build/bin/whisper-cli -m models/ggml-medium.bin -f samples/my_audio_file.wav -osrt Use code with caution. System Requirements & Optimization if you have cloned the repository
You can directly download the pre-converted ggml-medium.bin (or ggml-medium.en.bin for English-only) directly from the Hugging Face Whisper.cpp Collection. 2. Basic Transcription
Are you integrating this into a (like Python, Node.js, or a video editor)? ggml-medium.bin
It is important to note that as of late 2023, the ggml-medium.bin file format is widely considered .
To smoothly run ggml-medium.bin inside a project like whisper.cpp , your hardware should meet these baselines: : At least 8 GB of system memory.