| --- | |
| # Agent Q3 [Evo] HFRepo | |
| **Archived from:** `madDegen/agent-q3-core` | |
| **Archived to canonical repo:** `madDegen/Agent-Q3` | |
| ## About | |
| This README was originally the root README of the `madDegen/agent-q3-core` HuggingFace model repo, | |
| labeled **Agent Q3 [Evo] HFRepo**. | |
| Agent Q3 [Evo] is the self-improving evolution variant: | |
| - Unsloth LoRA fine-tuning (SFTTrainer on Llama-3.2-3B-Instruct-bnb-4bit) | |
| - arXiv ingestion pipeline → ChromaDB (384-dim nomic-embed-text) | |
| - DPO/RLHF feedback collector | |
| - LangGraph loop: ingest → train → benchmark → feedback (continues if score < 0.80) | |
| - 4 domain benchmarks: prediction_markets / solidity / langgraph / lora | |
| All code lives in the canonical repo at `madDegen/Agent-Q3` under the `evo/` directory. | |
| ## Links | |
| - Canonical Code Repo: https://huggingface.co/madDegen/Agent-Q3 | |
| - Canonical Dataset/Bucket: https://huggingface.co/datasets/madDegen/agent-q3 | |
| - Canonical Space: https://huggingface.co/spaces/madDegen/agent-q3-space | |
| - GitHub: https://github.com/MADdegen/Agent-Q3 | |