Aller au contenu
ModelPrésentéExterne
Télécharger

Whisper Small Fa ONNX

This is an ONNX version of aictsharif/whisper-small-fa. It was automatically converted and uploaded using this Hugging Face Space.

Sourcehuggingface.co/onnx-community/whisper-small-fa-ONNX/tree/f8427a5ea7755dd334e57475c3e58e0672cb7a88
TypeText modelModèle de baseaictsharif/whisper-small-faVersion

Source du modèle

Type
Text model
Taille du fichier
245 MB
Modèle de base
aictsharif/whisper-small-fa
Version
vdecoder_model_bnb4
Mis à jour
22 sept. 2026
Téléchargements
16
Source
Hugging Face

Description de la source

whisper-small-fa (ONNX)

This is an ONNX version of aictsharif/whisper-small-fa. It was automatically converted and uploaded using this Hugging Face Space.

Usage with Transformers.js

See the pipeline documentation for automatic-speech-recognition: https://huggingface.co/docs/transformers.js/api/pipelines#module_pipelines.AutomaticSpeechRecognitionPipeline

Whisper small Fa - Common Voice

This model is a fine-tuned version of openai/whisper-small on the Common Voice 17.0 dataset.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-05
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • training_steps: 4000
  • mixed_precision_training: Native AMP

Test results

  • Best test WER (Word Error Rate): 0.419
  • Best test CER (Character Error Rate): 0.196

Usage

import torch
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline

device = "cuda:0" if torch.cuda.is_available() else "cpu"
torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32

model_id = "aictsharif/whisper-small-fa"
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(
    "automatic-speech-recognition",
    model=model,
    tokenizer=processor.tokenizer,
    feature_extractor=processor.feature_extractor,
    torch_dtype=torch_dtype,
    device=device,
)

result = pipe('sample.mp3')
print(result["text"])

Framework versions

  • Transformers 4.51.3
  • Pytorch 2.5.1+cu124
  • Datasets 3.5.0
  • Tokenizers 0.21.1

Sources

1 source
Hugging FacePrincipalVersion f8427a5ea7755dd334e57475c3e58e0672cb7a88
huggingface.co/onnx-community/whisper-small-fa-ONNX/tree/f8427a5ea7755dd334e57475c3e58e0672cb7a88

Vérifié 22 sept.

Artefacts du modèle

2 artefacts

onnx/decoder_model_bnb4.onnx

onnx · 245 MB · SHA-256 787922de9111…7750 · Hugging Face

Télécharger

onnx/decoder_model_q4.onnx

onnx · Q4 · 245 MB · SHA-256 787922de9111…7750 · Hugging Face

Télécharger

Extraits de sources

2 extraits
Readme22 sept.huggingface.co/onnx-community/whisper-small-fa-ONNX

--- library_name: transformers.js language: - fa license: apache-2.0 base_model: - aictsharif/whisper-small-fa tags: - generated_from_trainer datasets: - mozilla-foundation/common_voice_17_0 model-index: - name: Whisper small Fa - Common Voice results: [] pipeline_tag: automatic-speech-recognition --- # whisper-small-fa (ONNX) This is an ONNX version of [aictsharif/whisper-small-fa](https://huggingface.co/aictsharif/whisper-small-fa). It was automatically converted and uploaded using [this Hugging Face Space](https://huggingface.co/spaces/onnx-community/convert-to-onnx). ## Usage with Transformers.js See the pipeline documentation for `automatic-speech-recognition`: https://huggingface.co/docs/transformers.js/api/pipelines#module_pipelines.AutomaticSpeechRecognitionPipeline --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # Whisper small Fa - Common Voice This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 17.0 dataset. ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 1e-05 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 500 - training_steps: 4000 - mixed_precision_training: Native AMP ### Test results - Best test WER (Word Error Rate): 0.419 - Best test CER (Character Error Rate): 0.196 ### Usage ```python import torch from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline device = "cuda:0" if torch.cuda.is_available() else "cpu" torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32 model_id = "aictsharif/whisper-small-fa" 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( "automatic-speech-recognition", model=model, tokenizer=processor.tokenizer, feature_extractor=processor.feature_extractor, torch_dtype=torch_dtype, device=device, ) re...

Note22 sept.huggingface.co/onnx-community/whisper-small-fa-ONNX

Source context: 16 downloads · 0 likes · Pipeline automatic-speech-recognition · Library transformers.js · Repo onnx-community/whisper-small-fa-ONNX