Model reference · open weights

whisper-large

Available as managed deployment Audio metythorn · community Speech→text 1 variants 4k dl/mo

whisper-large is an open-weight audio or speech model from metythorn. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.

Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.

What it is

Released bymetythorn
TypeAudio & music
TaskSpeech→text
Parameters (lead)1.5B
Released2025-11-21
Popularity4k downloads / month
LicenceOpen weights

About

What whisper-large is

Fine-tuned variant of openai/whisper-large-v3 for Khmer automatic speech recognition. The model was trained with the utilities in whisper and is intended for transcription workloads that prioritize Khmer text normalization, including numerals, currency, and date expressions.

Read the full model card

Model Card

AttributeValue
Base modelopenai/whisper-large-v3
LanguageKhmer (km-KH)
TaskAutomatic Speech Recognition (speech-to-text)
Sample rate16 kHz audio, automatically resampled
Input lengthUp to 30 s clips (truncated during batching)
Finetuning dataasr_mixed_dataset.txt (internal manifests, normalized through dataset_builder.segment_text)
Epochs10
Batch size2 (gradient accumulation 1)
OptimizerAdamW (managed by Seq2SeqTrainer)
Learning rate1e-6 with cosine scheduler & 1k warmup steps
NormalizationKhmer-specific regex and rule-based normalization (khmerspeech, khmercut)
DatasetTraining with Mixed Khmer & English audio with 199K samples (225 hours), train all khmer public dataset + humaned label dataset
Training TimeTraining with Mixed precision with RTX-5090 VRAM 32GB for 10 days

Limitations: performance has been validated only on internal validation/test splits. Long-form audio, accents outside the training distribution, or noisy backgrounds may degrade accuracy.

Inference Examples

import torch
import torchaudio
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline

AUDIO_PATH = "audio_path.wav"

device = "cuda:0" if torch.cuda.is_available() else "cpu"
torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
model_id = "metythorn/whisper-large-v3"
model = AutoModelForSpeechSeq2Seq.from_pretrained(
    model_id,
    torch_dtype=torch_dtype,
    low_cpu_mem_usage=True,
    use_safetensors=True,
)
model.to(device)
processor = AutoProcessor.from_pretrained(model_id)

pipe = pipeline(
    task="automatic-speech-recognition",
    model=model,
    tokenizer=processor.tokenizer,
    feature_extractor=processor.feature_extractor,
    torch_dtype=torch_dtype,
    device=device,
)

speech_waveform, sr = torchaudio.load(AUDIO_PATH)

# Whisper expects 16kHz mono
if sr != 16000:
    speech_waveform = torchaudio.functional.resample(
        speech_waveform,
        orig_freq=sr,
        new_freq=16000
    )
speech_waveform = speech_waveform.squeeze().numpy()
result = pipe(speech_waveform)

print("Transcription:", result["text"])

From the published model card. Full card on the HuggingFace links in the sidebar.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys metythorn-whisper-large for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (metythorn-whisper-large below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/audio/transcriptions \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -F model="metythorn-whisper-large" -F file=@audio.mp3

Create an account — your API key is available in the console. 3M free tokens every 30 days with every new account.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms