YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Qwen3 Chat Summarization
PEFT LoRA adapter for Vietnamese chat summarization.
Training & Model Details
- Base Model:
Qwen/Qwen3-1.7B - Adapter: LoRA rank 32, alpha 128
- Target Modules: {'q_proj', 'v_proj', 'k_proj', 'o_proj'}
- Max Sequence Length: 1024
- Max New Tokens: 70
- Vietnamese System Instruction with An/Bình/Chi/Dũng One-shot.
Evaluation on Held-out Test Split (300 samples)
| Metric | Value |
|---|---|
| Mean BLEU | 0.184186 |
| Mean ROUGE-1 F1 | 0.525737 |
| Mean ROUGE-2 F1 | 0.252915 |
| Mean ROUGE-L F1 | 0.454015 |
| Mean METEOR | 0.448288 |
| Latency mean (sec) | 3.389575 |
| Latency p50 (sec) | 3.301413 |
| Latency p95 (sec) | 4.525797 |
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support