ord-retro-qwen25-15b LoRA adapter
QLoRA adapter fine-tuned on ORD-derived single-step reaction records.
Task:
target molecule SMILES -> one single-step JSON synthesis proposal.
Training setup:
- Base model: Qwen/Qwen2.5-1.5B-Instruct
- Max sequence length: 2048
- Max steps: 2500
- LoRA rank: 32
- BF16: True
Limitations:
- Academic/diploma demonstration only.
- Not authoritative chemistry guidance.
- Product-only retrosynthesis is under-constrained.
- Generated routes must be manually checked.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support