Instructions to use Pablo-Flores-Mollinedo/verilog-qwen2.5-coder-7b-v21-comb-vector-specialist-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Pablo-Flores-Mollinedo/verilog-qwen2.5-coder-7b-v21-comb-vector-specialist-lora with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-Coder-7B-Instruct") model = PeftModel.from_pretrained(base_model, "Pablo-Flores-Mollinedo/verilog-qwen2.5-coder-7b-v21-comb-vector-specialist-lora") - Notebooks
- Google Colab
- Kaggle
Verilog Qwen2.5-Coder 7B v21 comb_vector Specialist LoRA
Specialist adapter trained from v9 for comb/vector-style Verilog tasks with v9 retention.
Parent
- Base:
Qwen/Qwen2.5-Coder-7B-Instruct - Parent adapter:
Pablo-Flores-Mollinedo/verilog-qwen2.5-coder-7b-v9-auto-distilled-direct-lora
VerilogEval v2 direct result
{
"total": 156,
"compile": 136,
"passed": 69,
"compile_pct": 87.18,
"pass_pct": 44.23
}
Comparison:
- v9 direct: 67/156 = 42.95%
- v21 comb_vector: 69/156 = 44.23%
Caveat: benchmark-targeted research adapter, not a clean zero-shot leaderboard claim.
- Downloads last month
- 14
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support