Automatic Speech Recognition
Transformers
Safetensors
jais_asr
text-generation
jais
ifm
arabic
arabic-dialects
speech-to-text
code-switching
diacritization
language-identification
open-weights
custom_code
Instructions to use IFM/Jais-ASR with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use IFM/Jais-ASR with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="IFM/Jais-ASR", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("IFM/Jais-ASR", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download vllm_plugin/pyproject.toml from IFM/Jais-ASR: direct link, hf CLI and curl.
- Browser
- Download file 739 Bytes
-
https://huggingface.co/IFM/Jais-ASR/resolve/main/vllm_plugin/pyproject.toml
- Command line
-
hf download hf://IFM/Jais-ASR/vllm_plugin/pyproject.toml
-
curl -L -o pyproject.toml https://huggingface.co/IFM/Jais-ASR/resolve/main/vllm_plugin/pyproject.toml
739 Bytes
| [build-system] | |
| requires = ["setuptools>=64"] | |
| build-backend = "setuptools.build_meta" | |
| [project] | |
| name = "jais-asr-vllm" | |
| version = "2.0.1" | |
| description = "vLLM integration for Jais-ASR: registers the model architecture (audio support) and the `jais_asr_tag` reasoning parser." | |
| requires-python = ">=3.10" | |
| dependencies = [] | |
| [project.entry-points."vllm.general_plugins"] | |
| # vLLM calls this in every engine process at startup, before the model is built. | |
| jais_asr = "jais_asr_vllm:register" | |
| [project.entry-points."vllm.logits_processors"] | |
| # Loaded by every serve; active only for requests that pass the vllm_xargs key. | |
| jais_asr_language = "jais_asr_vllm.logitsprocs:JaisASRLanguageLogitsProcessor" | |
| [tool.setuptools] | |
| packages = ["jais_asr_vllm"] | |