Downloads · 30 days
0
forge-optimizer-qwen3.6-35b-a3b
squaredcuber/forge-optimizer-qwen3.6-35b-a3b
forge-optimizer-qwen3.6-35b-a3b is a machine learning model from squaredcuber. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
A LoRA fine-tune of Qwen3.6-35B-A3B (MoE) that turns unoptimized backend code into efficient, correct code — the skill measured by forger-bench, an efficiency-aware benchmark for AI-generated InsForge SDK code.
Access
Public
Updated Jun 7, 2026
Repo size
7.5 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors7.4 GB · 100%
At a glance
- License
- apache-2.0
- Access
- Public
- Created
- Jun 6, 2026
- Updated
- Jun 7, 2026
- SHA
- 1820044f
Base models
Model card
From the Hugging Face model README
forge-optimizer
A LoRA fine-tune of Qwen3.6-35B-A3B (MoE) that turns unoptimized backend code into efficient, correct code — the skill measured by forger-bench, an efficiency-aware benchmark for AI-generated InsForge SDK code.
Trained with Unsloth (bf16 LoRA) + an agentic GRPO loop adopting CUDA-Agent (arXiv 2602.24286): the model writes a solution, the forger-bench grader runs+verifies+ measures real server metrics, and a discrete milestone reward (-1 incorrect/scaleBug, 1 wasteful, 2 beats-naive, 3 near-optimal) drives RL.
Contamination control
Never trained on a sealed test task; held-out concepts measure optimization skill vs template memorization.
- License
- apache-2.0
- Created
- Jun 6, 2026
- Updated
- Jun 7, 2026