GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
KhiredNetworks
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
KhiredNetworks
Model Tree
openai/whisper-large-v3-turbo
Input Modalities
Output Modalities
Supported Functionality
language=ur,
task=transcribe, greedy.133 held-out SimpleRishta calls, zero caller overlap with training, corpus C0 WER (normalized, PII-masked), identical decode settings for both models:
| Model | WER on SR calls |
|---|---|
openai/whisper-large-v3 (off-the-shelf) | 15.6% |
| this adapter | 8.9% |
Reference transcription: an independent high-accuracy Whisper-family model
(eval-only, never trained on). Both models scored identically against it.
Head-to-head on the same calls: this adapter wins 121 / loses 9 / ties 3.
Zero degenerate generations across all stability checks (repetition-rate and
length-ratio guards, clean greedy decoding). Aggregate protocol and numbers:
eval/aggregate-results.json.
Independent reproduction (live run): the 8.9% figure is re-derived end-to-end on a clean GPU in this public Kaggle notebook — model loaded, 133 calls decoded, fresh WER printed against the certified value (observed drift 0.0000): r12-sr-eval-wer-proof.
python
import torch, soundfile as sffrom peft import PeftModelfrom transformers import WhisperForConditionalGeneration, WhisperProcessor base = "openai/whisper-large-v3-turbo"processor = WhisperProcessor.from_pretrained(base)model = WhisperForConditionalGeneration.from_pretrained(base)model = PeftModel.from_pretrained(model, "KhiredNetworks/PakUrdu-Conversational-ASR").merge_and_unload() speech, sr = sf.read("call.wav", dtype="float32")inputs = processor(speech, sampling_rate=sr, return_tensors="pt")ids = model.generate(inputs.input_features, language="ur", task="transcribe", do_sample=False)print(processor.batch_decode(ids, skip_special_tokens=True)[0])A runnable version ships as inference_example.py.
Urdu script output; evaluation text normalized under a corpus C0 protocol
(Unicode normalization, digit/punctuation folding, brand canonicalization,
PII masking) before scoring. Decode: greedy (do_sample=False),
language=ur, task=transcribe, 16 kHz mono input. FP16 on GPU / FP32 on
CPU both supported; results certified on a single T4.
language=ur fixed in the intended decode config).LICENSE). Base model:
openai/whisper-large-v3-turbo — Whisper license and attribution chain
apply downstream.CITATION.cff (Khired Networks).import torch, soundfile as sffrom peft import PeftModelfrom transformers import WhisperForConditionalGeneration, WhisperProcessor base = "openai/whisper-large-v3-turbo"processor = WhisperProcessor.from_pretrained(base)model = WhisperForConditionalGeneration.from_pretrained(base)model = PeftModel.from_pretrained(model, "KhiredNetworks/PakUrdu-Conversational-ASR").merge_and_unload() speech, sr = sf.read("call.wav", dtype="float32")inputs = processor(speech, sampling_rate=sr, return_tensors="pt")ids = model.generate(inputs.input_features, language="ur", task="transcribe", do_sample=False)print(processor.batch_decode(ids, skip_special_tokens=True)[0])