Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
verilog-qwen2.5-coder-7b-v20-adapter-routing-pipeline – AI Model by Pablo-Flores-Mollinedo | AlphaNeural AI
You can deploy this model and start earning money today!
Pablo-Flores-Mollinedo
/
verilog-qwen2.5-coder-7b-v20-adapter-routing-pipeline
like
0
verilog
rtl
adapter-routing
lora
qwen2.5-coder
Qwen/Qwen2.5-Coder-7B-Instruct
adapter
apache-2.0
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
Verilog Adapter Routing Pipeline v20
Offline adapter-routing analysis over existing VerilogEval v2 direct builds.
Results
v9 default: 67/156 = 42.95%
simple category router: 71/156 = 45.51%
safer category router: 69/156 = 44.23%
oracle union across adapters: 86/156 = 55.13%
best existing pipeline v9 k=5 + v7b repair: 76/156 = 48.72%
Interpretation
Routing works: a simple prompt-category router improves over v9 direct by +4 tasks. The oracle union indicates substantial complementarity between adapters.
Caveat
Benchmark-targeted research pipeline, not a clean zero-shot leaderboard claim.
