README.md

metadata

library_name: transformers
language:
  - en
license: apache-2.0
base_model: openai/whisper-tiny
tags:
  - wft
  - whisper
  - automatic-speech-recognition
  - audio
  - speech
  - generated_from_trainer
datasets:
  - hf-internal-testing/librispeech_asr_dummy
metrics:
  - wer
model-index:
  - name: wft-test-model
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: hf-internal-testing/librispeech_asr_dummy
          type: hf-internal-testing/librispeech_asr_dummy
        metrics:
          - type: wer
            value: 4.724409448818897
            name: Wer

wft-test-model

This model is a fine-tuned version of openai/whisper-tiny on the hf-internal-testing/librispeech_asr_dummy dataset. It achieves the following results on the evaluation set:

• Loss: 0.1218
• Wer: 4.7244
• Cer: 100.7102
• Decode Time: 0.5595
• Wer Time: 0.0060
• Cer Time: 0.0027

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: 0.0005
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer	Decode Time	Wer Time	Cer Time
2.4038	0.1	10	1.9871	299.6063	116.5483	0.5542	0.0112	0.0038
1.2202	1.01	20	1.1610	172.0472	91.6903	0.5178	0.0055	0.0028
0.8765	1.11	30	0.8079	31.8898	56.1080	0.4854	0.0051	0.0024
0.4415	2.02	40	0.6279	25.9843	82.5284	0.5218	0.0060	0.0028
0.4307	2.12	50	0.4509	16.9291	98.2955	0.5377	0.0056	0.0030
0.2363	3.03	60	0.2952	9.4488	102.9119	0.5310	0.0048	0.0027
0.2245	3.13	70	0.2046	7.4803	96.3778	0.5429	0.0056	0.0029
0.1053	4.04	80	0.1700	5.5118	96.4489	0.5306	0.0067	0.0028
0.0731	4.14	90	0.1320	4.7244	97.1591	0.5509	0.0050	0.0026
0.0782	5.05	100	0.1218	4.7244	100.7102	0.5595	0.0060	0.0027

Framework versions

• PEFT 0.13.2
• Transformers 4.45.2
• Pytorch 2.5.0
• Datasets 3.0.2
• Tokenizers 0.20.1