GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
bijaykumarsingh
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
bijaykumarsingh
Model Tree
Qwen/Qwen2-Audio-7B-Instruct
Input Modalities
Output Modalities
Supported Functionality
Qwen/Qwen2-Audio-7B-Instructas)Evaluated on a strictly speaker-disjoint held-out test split (2,266 utterances, 31 unique speakers, 3.18 hours, 22,596 total words) with standardized Unicode NFC and punctuation stripping:
| Metric | Score | 95% Bootstrap CI |
|---|---|---|
| Word Error Rate (WER) | 27.97% | [26.83%, 29.09%] |
| Character Error Rate (CER) | 10.92% | — |
| Match Error Rate (MER) | 27.36% | — |
| Word Information Lost (WIL) | 43.86% | — |
| Training Wall-Clock Time | 11.70 hours | (NVIDIA A40 48GB) |
python
from transformers import AutoProcessor, AutoModelForCausalLMfrom peft import PeftModel base_model_id = "Qwen/Qwen2-Audio-7B-Instruct"adapter_id = "bijaykumarsingh/qwen2-audio-7b-assamese-lora" processor = AutoProcessor.from_pretrained(base_model_id)base_model = AutoModelForCausalLM.from_pretrained(base_model_id, device_map="auto")model = PeftModel.from_pretrained(base_model, adapter_id)model.eval()bibtex
@article{singh2026assamese_lalm, title={Architectural Inductive Bias Trumps Parameter Scale: Adapting Large Audio-Language Models to Low-Resource Assamese ASR}, author={Singh, Bijay Kumar}, journal={arXiv preprint}, year={2026}}from transformers import AutoProcessor, AutoModelForCausalLMfrom peft import PeftModel base_model_id = "Qwen/Qwen2-Audio-7B-Instruct"adapter_id = "bijaykumarsingh/qwen2-audio-7b-assamese-lora" processor = AutoProcessor.from_pretrained(base_model_id)base_model = AutoModelForCausalLM.from_pretrained(base_model_id, device_map="auto")model = PeftModel.from_pretrained(base_model, adapter_id)model.eval()@article{singh2026assamese_lalm, title={Architectural Inductive Bias Trumps Parameter Scale: Adapting Large Audio-Language Models to Low-Resource Assamese ASR}, author={Singh, Bijay Kumar}, journal={arXiv preprint}, year={2026}}