Hugging Face's logo

FelixK7
/
whisper-large-lv

Automatic Speech Recognition
Transformers
Safetensors
Latvian
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Community
whisper-large-lv / README.md
FelixK7's picture
FelixK7
End of training
75826f9 verified
preview code
|
raw
history blame contribute delete
1.88 kB
metadata
library_name: transformers
language:
  - lv
license: apache-2.0
base_model: AiLab-IMCS-UL/whisper-large-v3-lv-late-cv19
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: Whisper large LV - Felikss Kleins
    results: []

Whisper large LV - Felikss Kleins

This model is a fine-tuned version of AiLab-IMCS-UL/whisper-large-v3-lv-late-cv19 on the Recorded Voice dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4380
  • Wer: 25.9259

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 1e-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 64
  • optimizer: Use adamw_hf 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: 5000

Training results

Training Loss Epoch Step Validation Loss Wer
0.036 48.7805 1000 0.4135 19.7531
0.0196 97.5610 2000 0.4086 22.2222
0.0126 146.3415 3000 0.4710 23.4568
0.0117 195.1220 4000 0.4380 25.9259

Framework versions

  • Transformers 4.46.1
  • Pytorch 2.3.1
  • Datasets 3.1.0
  • Tokenizers 0.20.2