GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
theelvace
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
theelvace
Model Tree
openai/whisper-small
Input Modalities
Output Modalities
Supported Functionality
62.45% WER on the FLEURS Igbo test set (969 samples), down from a 68.95% baseline — verified on the full test set, not estimated from training metrics.
ig_ng config) — 2,839 training examplespython
from transformers import WhisperProcessor, WhisperForConditionalGenerationfrom peft import PeftModelimport torch processor = WhisperProcessor.from_pretrained("openai/whisper-small")base_model = WhisperForConditionalGeneration.from_pretrained("openai/whisper-small")model = PeftModel.from_pretrained(base_model, "theelvace/whisper-small-igbo")model = model.merge_and_unload() # Igbo language token proxy — see note aboveforced_decoder_ids = [[1, 50325], [2, 50359], [3, 50363]] inputs = processor.feature_extractor(audio_array, sampling_rate=16000, return_tensors="pt")generated_ids = model.generate(inputs.input_features, forced_decoder_ids=forced_decoder_ids)transcription = processor.tokenizer.decode(generated_ids[0], skip_special_tokens=True)This repo also hosts ONNX exports (encoder, cross-attention initializer, KV-cache decoder) used to run this model fully on-device on Android — no cloud inference. See the Olu Igbo GitHub repo for the full mobile app, export scripts, and benchmarks on a Snapdragon 678 device.
MIT