Update requirements.txt

requirements.txt

@@ -1,133 +1,85 @@
 # ============================================
-# ULTIMATE REQUIREMENTS FOR MAXIMUM RESOURCE USAGE
-# Target: 32GB RAM + 24GB GPU VRAM (L4)
+# WORKING REQUIREMENTS FOR MAXIMUM RESOURCE USAGE
+# Fixed for HuggingFace Spaces deployment
+# Target: 32GB RAM + 24GB GPU VRAM
 # ============================================
 
 # Core Dependencies
-streamlit==1.48.0
-opencv-python-headless==4.8.1.78
-numpy==1.24.4
-Pillow==10.0.1
-
-# Video Processing - ENHANCED for 16K support
-imageio==2.25.0
-imageio-ffmpeg==0.4.8
-ffmpeg-python==0.2.0
-moviepy==1.0.3
-av>=0.5.2
-vidgear[core]>=0.3.0  # For 16K video streaming
-
-# KRITISK: PyTorch with CUDA 11.8 for L4 GPU
---extra-index-url https://download.pytorch.org/whl/cu118
-torch==2.1.0+cu118
-torchvision==0.16.0+cu118
-torchaudio==2.1.0+cu118  # For video with audio
-
-# DEEP LEARNING for Real-Time Training
-transformers>=4.35.0  # HuggingFace transformers
-accelerate>=0.24.0  # GPU memory optimization
-datasets>=2.14.0  # For loading training data
-tokenizers>=0.14.0  # Fast tokenizers
-safetensors>=0.4.0  # Efficient tensor serialization
-bitsandbytes>=0.41.0  # 8-bit optimization
-peft>=0.6.0  # Parameter efficient fine-tuning
-
-# Computer Vision Models
-timm>=0.9.0  # PyTorch Image Models
-albumentations>=1.3.0  # Advanced augmentations
-segmentation-models-pytorch>=0.3.0  # Segmentation models
-ultralytics>=8.0.0  # YOLO models
-detectron2 @ git+https://github.com/facebookresearch/detectron2.git  # Facebook's detection
-
-# ONNX Runtime GPU - MAXIMUM PERFORMANCE
+streamlit==1.32.2
+opencv-python-headless==4.9.0.80
+numpy==1.24.3
+Pillow==10.2.0
+
+# System Monitoring - ESSENTIAL
+psutil==5.9.8
+GPUtil==1.4.0
+nvidia-ml-py==12.535.161
+py3nvml==0.2.7
+
+# PyTorch with CUDA (HuggingFace provides CUDA 11.8)
+torch==2.1.2
+torchvision==0.16.2
+torchaudio==2.1.2
+
+# Transformers & HuggingFace
+transformers==4.37.2
+accelerate==0.26.1
+datasets==2.16.1
+tokenizers==0.15.1
+safetensors==0.4.2
+huggingface-hub==0.20.3
+
+# Background Removal
+rembg==2.0.53
 onnxruntime-gpu==1.16.3
-onnx>=1.15.0
-onnxconverter-common>=1.13.0
-
-# Background Removal - GPU OPTIMIZED
-rembg[gpu]==2.0.50
-pymatting==1.1.8
-backgroundremover>=0.2.0  # Additional BG removal
-transparent-background>=1.2.0  # Another BG removal model
-
-# Image Processing & Enhancement
-scipy>=1.7.0
-scikit-image>=0.19.0
-matplotlib>=3.5.0
-kornia>=0.7.0  # GPU-accelerated image processing
-wand>=0.6.0  # ImageMagick binding for advanced processing
-colour-science>=0.4.0  # Color science algorithms
-
-# AI Training & Optimization
-wandb>=0.15.0  # Weights & Biases for tracking
-tensorboard>=2.14.0  # Training visualization
-optuna>=3.3.0  # Hyperparameter optimization
-ray[tune]>=2.7.0  # Distributed training
-pytorch-lightning>=2.0.0  # Training framework
 
-# System Monitoring - ENHANCED
-psutil>=5.9.0
-nvidia-ml-py3>=7.352.0
-pynvml>=11.5.0
-gpustat>=1.1.0  # GPU monitoring
-py3nvml>=0.2.0  # NVIDIA management
-prometheus-client>=0.18.0  # Metrics export
-
-# Memory Management
-memory-profiler>=0.60.0
-pympler>=1.0.0  # Memory tracking
-objgraph>=3.5.0  # Object tracking
-guppy3>=3.1.0  # Memory profiling
-
-# Parallel Processing
-joblib>=1.3.0
-dask[complete]>=2023.5.0  # Distributed computing
-multiprocess>=0.70.0
-concurrent-futures-extensions>=0.1.0
-
-# Caching & Storage
-diskcache>=5.6.0
-lmdb>=1.4.0  # Lightning memory-mapped database
-h5py>=3.9.0  # HDF5 for large datasets
-zarr>=2.16.0  # Compressed N-dimensional arrays
+# Computer Vision Models
+timm==0.9.12
+albumentations==1.3.1
+ultralytics==8.1.18
+supervision==0.17.1
+
+# Image Processing
+scipy==1.12.0
+scikit-image==0.22.0
+matplotlib==3.8.2
+imageio==2.33.1
+imageio-ffmpeg==0.4.9
+
+# Video Processing
+opencv-contrib-python==4.9.0.80
+moviepy==1.0.3
+av==11.0.0
 
-# Networking & API
-requests>=2.31.0
-aiohttp>=3.8.0
-httpx>=0.24.0
-websockets>=11.0
+# ML Libraries
+xgboost==2.0.3
+lightgbm==4.3.0
+scikit-learn==1.4.0
 
 # Data Processing
-pandas>=2.0.0
-polars>=0.19.0  # Faster pandas alternative
-vaex>=4.17.0  # Out-of-core DataFrames
-dask-ml>=2023.3.0  # Scalable machine learning
+pandas==2.2.0
+pyarrow==15.0.0
 
-# Additional ML Libraries
-xgboost>=2.0.0  # Gradient boosting
-lightgbm>=4.0.0  # Fast gradient boosting
-catboost>=1.2.0  # Categorical boosting
-prophet>=1.1.0  # Time series forecasting
+# Memory Profiling
+memory-profiler==0.61.0
+pympler==1.0.1
+objgraph==3.6.1
 
-# Audio Processing (for video with audio)
-librosa>=0.10.0
-soundfile>=0.12.0
-pydub>=0.25.0
-pedalboard>=0.8.0  # Spotify's audio effects
+# Parallel Processing
+joblib==1.3.2
+threadpoolctl==3.2.0
 
 # Utilities
-tqdm>=4.64.0
-rich>=13.5.0  # Rich terminal output
-typer>=0.9.0  # CLI building
-pydantic>=2.4.0  # Data validation
-omegaconf>=2.3.0  # Configuration management
-
-# Git packages for advanced models
-git+https://github.com/pq-yang/MatAnyone.git
-git+https://github.com/facebookresearch/segment-anything.git
-git+https://github.com/IDEA-Research/GroundingDINO.git
-
-# Optional: More models for RAM usage
-# git+https://github.com/microsoft/unilm.git
-# git+https://github.com/openai/CLIP.git
-# git+https://github.com/CompVis/stable-diffusion.git
+tqdm==4.66.1
+rich==13.7.0
+python-dotenv==1.0.1
+
+# Additional Vision Models (if space allows)
+segment-anything==1.0
+groundingdino-py==0.4.0
+
+# Optional Heavy Libraries (comment out if build fails)
+# detectron2 - requires manual installation
+# tensorflow==2.15.0  # Another 2GB
+# jax==0.4.23  # Another GPU framework
+# kornia==0.7.1  # GPU accelerated vision