beverlyhillscop/cotton-disease-qwen25vl-3b

Cotton Disease Detection - Qwen2.5-VL Models

Projekt-Übersicht

Vision-Language Model für Cotton Disease Detection auf Drohnen (Jetson Orin 64GB). Binäre Klassifikation: healthy vs diseased

Ergebnisse

Empfehlung: 3B FP16 für Jetson Deployment

Verzeichnisstruktur

/workspace/cotton-disease-detection/

• —data/raw/Cotton Disease/train/ - Training Daten
• —data/raw/Cotton Disease/val/ - Validation/Test Daten
• —scripts/trainqwen25vllora_v2.py - 7B Training Script
• —scripts/trainqwen25vl3b_simple.py - 3B Training Script (EMPFOHLEN)
• —scripts/test3bfp16.py - 3B FP16 Test Script
• —outputs/qwen25vlmergedv2/ - 7B Merged Model (~16GB)
• —outputs/qwen25vl3bmerged/ - 3B Merged Model (~7GB) PRODUCTION
• —outputs/qwen25vlq4k_m.gguf - 7B GGUF INT4 (4.46 GB)

Training Commands

7B Model:

python3 /workspace/cotton-disease-detection/scripts/trainqwen25vllora_v2.py

• —Base: Qwen/Qwen2.5-VL-7B-Instruct
• —Time: ~55 min (A100)
• —Batch: 4

3B Model (Empfohlen):

python3 /workspace/cotton-disease-detection/scripts/trainqwen25vl3b_simple.py

• —Base: Qwen/Qwen2.5-VL-3B-Instruct
• —Time: ~157 min (A100)
• —Batch: 8

Testing

python3 /workspace/cotton-disease-detection/scripts/test3bfp16.py

Jetson Deployment

Kopieren auf Jetson Orin 64GB:

• —3B FP16: outputs/qwen25vl3bmerged/ (~7GB)
• —Alternative: outputs/qwen25vlq4k_m.gguf (4.46GB für llama.cpp)

Erwartete Jetson Performance:

• —3B FP16 PyTorch: ~400-500ms
• —3B FP16 TensorRT: ~100-150ms (geschätzt)

LoRA Config

r=16, alpha=32, dropout=0.05 targets: qproj, kproj, vproj, oproj, gateproj, upproj, down_proj

Changelog

2025-01-29: 3B Model (98% Accuracy, 264ms, 7.13GB) 2025-01-29: 7B Model v2 (98% Accuracy)