Skip to content
View Lawson-Darrow's full-sized avatar
1 follower · 0 following

Achievements

Achievement: Pair Extraordinairex2Achievement: QuickdrawAchievement: YOLOAchievement: Pull Shark

Achievements

Achievement: Pair Extraordinairex2Achievement: QuickdrawAchievement: YOLOAchievement: Pull Shark

Highlights

  • Pro

Block or report Lawson-Darrow

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Lawson-Darrow/README.md

Hey, I'm Lawson 👋

  • Mostly computer vision and applied LLM systems (RAG, fine-tuning, model eval)
  • Just wrapped a B.S. in CS at Florida Tech
  • Currently building rag-conflict-eval: catching RAG answers that fake a consensus when their sources actually disagree
  • Supporting the open source libraries and tools I build on
  • More at ledarrow.dev

Pinned Loading

  1. rag-conflict-eval rag-conflict-eval Public

    Catch RAG answers that mishandle conflicting evidence: false consensus, stale sources. A conflict-aware eval + diagnostic for RAG.

    Python 1

  2. Biomedical-RAG-Agent Biomedical-RAG-Agent Public

    Grounded biomedical RAG agent with a frontier-vs-open evaluation harness.

    Python 2

  3. Text-to-SQL-Finetune Text-to-SQL-Finetune Public

    Fine-tuning small open models (Qwen2.5-Coder) for text-to-SQL; execution-accuracy eval across a size ladder vs a frontier baseline.

    Python 1

  4. Brain-Tumor-MRI-Classification Brain-Tumor-MRI-Classification Public

    Brain tumor MRI classification: ViT-B/16 vs CNNs vs classical, trained to convergence with a perceptual-hash data-leakage audit and dual official/deduplicated eval (94.9% / 87.8% acc, 0.99 AUC).

    Jupyter Notebook

  5. Skin-Cancer-Model Skin-Cancer-Model Public

    A Leakage-Controlled Benchmark for Skin-Lesion Classification

    Python 1

  6. Ship-Detection-Project Ship-Detection-Project Public

    Ship detection & localization in satellite imagery: YOLOv8 + Faster-RCNN on ShipRSImageNet with COCO mAP/IoU and area-stratified AP (0.80 mAP@0.5, 3-seed).

    Jupyter Notebook 1