ML Training Pipeline
Code-Trainer V6
A 6-phase pipeline to build and deploy a multimodal code generation model capable of generating source code from VS Code screenshot images.
32,658
Training Samples
HuggingFace dataset
32,727
Screenshot Captures
VS Code + Monaco
8
Languages
programming languages
6
Pipeline Phases
2 complete
Pipeline Progress 2/6 phases
01 Data Collection 02 Preprocessing 03 Vision Model 04 Qwen Fine-tuning 05 GGUF Deployment 06 Inference Agent
Base Model
Qwen2.5-Coder-14B-Instruct
Hardware
RTX 5060 Ti 16GB (Blackwell)
HF Dataset
cmndcntrlcyber/code-trainer-v6-dataset
Stack
Python · Core language
PyTorch · ML framework
Transformers · Model loading & training
PEFT · LoRA adapters
TRL · SFT training
Playwright · Screenshot capture
W&B · Experiment tracking
uv · Dependency management