IonGrozea's picture
Update card with normalized WER/CER
1da7556 verified
|
Raw History Blame Contribute Delete
1.73 kB
metadata
language:
  - ro
license: apache-2.0
tags:
  - automatic-speech-recognition
  - romanian
  - whisper
  - fine-tuned
metrics:
  - wer
  - cer
model-index:
  - name: whisper-medium_ro-80mel
    results:
      - task:
          type: automatic-speech-recognition
        dataset:
          name: Romanian merged corpus (80-mel)
          type: custom
        metrics:
          - type: wer
            value: 0.0656
          - type: cer
            value: 0.0228

whisper-medium_ro-80mel — Romanian Fine-tune

Fine-tuned openai/whisper-medium on a merged Romanian corpus (~445h). Input: 80-mel spectrograms.

Performance (beam=5, full val)

Metric Value
WER 0.0656
CER 0.0228
Evaluated on 25,348 validation samples (52.75h).

Training

Parameter Value
Base openai/whisper-medium
GPU NVIDIA A100 80GB PCIe
Features 80-mel
Effective batch 128 (16×8)
Optimizer adamw_torch
LR / scheduler 4e-06 / cosine
Warmup 500
Max epochs 8
Precision BF16
Selection eval_loss (early stop 5)
Date 2026-09-21

Dataset

Dataset Samples Hours
Echo 163,050 374.33h
VoxPopuli 27,060 85.20h
RODigits 15,389 37.55h
RO-Smart (TTS) 10,572 17.28h
CommonVoice-25 19,850 22.15h
FLEURS 4,161 13.73h
USPDATRO 2,637 4.31h
TOTAL 242,719 554.55h

Notes

  • decoder_start_token_id=50258 (SOT); positional label masking; begin_suppress_tokens=[220].
  • Checkpoint selection on eval_loss; final WER via standalone beam=5 generate.