kingabzpro / automatic-speech-recognition Open weights updated 4 months ago

wav2vec2-large-xls-r-300m-Urdu

Name: wav2vec2-large-xls-r-300m-Urdu
Author: kingabzpro

A fine-tuned XLS-R 300M CTC model for Urdu automatic speech recognition. It transcribes 16 kHz mono audio and includes an optional 5-gram KenLM decoder.

Params

320 M

Context

—

Downloads 30d

1.1 M

Likes

Commercial use: allowed apache-2.0 Not gated SAFETENSORS 1 languages View on Hugging Face ↗

Download history

daily snapshots · 106 days

▲ 148 K in the last 30 days (15.7%)

1.9 M681 K

Jul 28Sep 1Oct 6Nov 10

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	f32	1.3 GB	1.9 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Architecture: Wav2Vec2ForCTC
Parameters: 320 M
Tensor type: F32
Vocabulary: 57
Layers / heads: 24 / 16
Licence: apache-2.0
First seen on the Hub: 2022-03-02
Training datasets: mozilla-foundation/common_voice_8_0
Common Voice 8 (reported): 16.7
Added to our catalog: 2026-07-28

Compare with any automatic-speech-recognition model

Popular comparisons

whisperkit-coreml 51.6 M wav2vec2-large-xlsr-53-japanese 16.2 M speaker-diarization-3.1 6.7 M faster-whisper-small 5.7 M speaker-diarization-community-1 5.1 M wav2vec2-large-xlsr-53-russian 3.8 M