GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
theshivam7
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
theshivam7
Model Tree
openai/whisper-medium
Input Modalities
Output Modalities
Supported Functionality
Requires 16 kHz mono audio (the pipeline below resamples for you).
python
from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="theshivam7/whisper-medium-indian-english")result = pipe("audio.wav")print(result["text"])For audio longer than 30 seconds, enable chunking:
python
result = pipe("long_audio.wav", chunk_length_s=30, stride_length_s=5)A full fine-tune (all 769M parameters) of Whisper Medium via transformers Seq2SeqTrainer,
adapted to Indian English academic/lecture speech. Standard Whisper encoder-decoder architecture,
no structural changes. Input: 16 kHz mono audio; output: English text.
Evaluated on Indian-accented English academic lecture speech (TIE_shorts). Not evaluated on other English varieties or domains.
⚠️ Speaker overlap (disclosed). The training data's official split shares speakers with the test split (100% of test speakers, and 100% of test clips, are from speakers also seen in training). There is no clip-level leakage, but part of the measured gain over the pretrained baseline likely reflects speaker adaptation rather than purely accent/content adaptation. See the benchmark repo for full methodology.
⚠️ No net gain, and this is the largest of a 3-size family. The +0.20pp WER change over the pretrained baseline (below) is not statistically significant (95% CI [−0.46, +1.03] pp, pHolm=0.774). Across the tiny/small/medium capacity study, gains shrink as model size grows and flip to null at this size; see the benchmark repo for the full capacity-study writeup. Do not assume this checkpoint outperforms the base model.
⚠️ Speaker-disjoint re-evaluation shows a regression, not just a null. Under a strict speaker-disjoint re-split (train speakers never appear in test) run with 3 seeds, one seed (42) shows a statistically significant WER regression (+1.75pp, 95% CI [+0.13, +4.17], pHolm=0.048) against this same 14.42% pretrained baseline; the other two seeds move the same direction (+0.38pp and +0.79pp) without reaching significance. A size-matched, speaker-overlapping control at the same ~567-clip training budget lands flat across all three seeds (-0.09, -0.02, -0.02 pp, all pHolm=1.000), so the regression traces to speaker-disjointness itself, not the smaller training set.
The implication is that the +0.20pp null above is not neutral. Once speaker overlap is removed, this fine-tune is worse than not fine-tuning at all, so the official-split number was propped up by adaptation to test speakers. Full tables:
results/tie/analysis/finetune_disjoint_control.md. Note that the per-clip transcripts for those six control runs were not retained, so that table is the scoring output as recorded rather than something recomputable from the repo.
Transcript field (English, transcribe task).load_best_model_at_end)mask_time_prob=0.05, mask_feature_prob=0.05)Checkpoint selection used validation WER (same normalization as the final metric). Training early-stopped at epoch 3 (patience 2), final training loss ≈ 0.39.
Corpus WER on TIE_shorts test split, transcript_clean normalization, both models decoded
through the identical HF transformers pipeline (engine-controlled comparison):
| Model | Corpus WER |
|---|---|
| openai/whisper-medium (pretrained baseline) | 14.42% |
| theshivam7/whisper-medium-indian-english (this model) | 14.61% |
Δ = +0.20 pp (95% CI [−0.46, +1.03], pHolm=0.774, not significant, see limitations above).
Part of Indian-ASR-Bench. See the repo for full benchmark methodology, error analysis, and reproduction instructions.