| #!/usr/bin/env python3 | |
| """ | |
| Quillan-Ronin v5.3.0 Training Script for Google Colab | |
| Optimized for GPU acceleration with visual progress tracking | |
| """ | |
| # Install dependencies first | |
| !pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121 | |
| !pip install matplotlib numpy scipy | |
| # Check GPU availability | |
| import torch | |
| print(f"🔥 CUDA available: {torch.cuda.is_available()}") | |
| if torch.cuda.is_available(): | |
| print(f"🎮 GPU: {torch.cuda.get_device_name(0)}") | |
| print(f"🧠 Memory: {torch.cuda.get_device_properties(0).total_memory / 1024**3:.1f} GB") | |
| # Mount Google Drive for data access | |
| from google.colab import drive | |
| drive.mount('/content/drive') | |
| # Upload your training data to Google Drive first | |
| # Then update these paths accordingly | |
| DATA_PATH = "/content/drive/MyDrive/QuillanData/" | |
| # Copy the training script and run | |
| !cp /content/drive/MyDrive/Quillan/train_full_multimodal.py . | |
| !python train_full_multimodal.py | |
| # After training, download the results | |
| from google.colab import files | |
| !files.download("best_multimodal_quillan.pt") | |
| !files.download("training_progress_step_*.png") | |