You can deploy this model and start earning money today!

chakibmed/
whispbook-qwen3-4b-speaker-lora-speaker_lora_v1_easy_evidence_only_cleaned

Name: whispbook-qwen3-4b-speaker-lora-speaker_lora_v1_easy_evidence_only_cleaned
Availability: InStock

Views

No views yet

whispbook-qwen3-4b-speaker-lora-speaker_lora_v1_easy_evidence_only_cleaned

LoRA adapter for speaker attribution on top of Qwen/Qwen3-4B-Instruct-2507.

Training base: mlx-community/Qwen3-4B-Instruct-2507-4bit.

python

import mlx_lm

model, tokenizer = mlx_lm.load("mlx-community/Qwen3-4B-Instruct-2507-4bit", adapter_path="path/to/adapter")

Run ID: speaker_lora_v1_easy_evidence_only_cleaned
Experiment: speaker_lora_v1_easy_evidence_only
Dataset slice: direct-attribution dialogue only, local randomized IDs, no narrator/unknown rows.
Evaluation split: outputs/qwen_json_v1_easy/test.jsonl