Modelpublic

thunderboltc/whisper_sanlish_augmented_test1

sourceHugging Faceupdated 1mo agoView on Hugging Face

0likes22downloads

Model Card

geminiwhisperipa

Fine-tuned openai/whisper-small for Santali IPA/Sanlish transliteration (speech-to-text).

Last updated: 2026-08-31T22:08:11.029452Z — after epoch 10.18

Dataset split (80/10/10)

—train: 3094
—validation: 193
—test: 193

Latest evaluation metrics (epoch 10.18)

—test_loss: 0.7281994819641113
—test_wer: 33.021077283372364
—test_cer: 7.840385001002607

Training hyperparameters

—perdevicetrainbatchsize: 8
—gradientaccumulationsteps: 2
—effectivebatchsize: 16
—learning_rate: 2e-05
—warmup_steps: 150
—numtrainepochs: 15
—gradient_checkpointing: True
—fp16: True
—eval_strategy: IntervalStrategy.EPOCH
—save_strategy: SaveStrategy.EPOCH
—savetotallimit: 2
—loadbestmodelatend: True
—metricforbest_model: wer
—predictwithgenerate: True
—generationmaxlength: 225