Automatic Speech Recognition
Transformers
ONNX
Safetensors
ExecuTorch
English
asr4all
feature-extraction
speech
ctc
streaming
on-device
edge
custom_code
Eval Results (legacy)
Eval Results
Instructions to use futo-org/asr4all-l with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use futo-org/asr4all-l with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="futo-org/asr4all-l", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("futo-org/asr4all-l", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Commit History
-
asr4all-l: streaming trunk + PCEC + wake VAD; ONNX per-component graphs
Verified signatureReplaces the previous large release. - ONNX now ships one graph per component with shared weights; the unified asr4all.onnx is removed (loading it pulled in every component). - ExecuTorch ships xnnpack_int8 and xnnpack_fp32; the fp16 and OpenVINO-delegate bundles are retired (OpenVINO runs the ONNX graphs). - Card carries measured RTFx across nine platform configurations and the wake VAD figures.b822e45
-
asr4all-l: initial artifacts — PCEC in-bundle (transformers + ExecuTorch), GPTQ int8, 32 s windows
Verified signature6ba9314