=== Upload Instructions ===

1. Navigate to the upload directory:
   cd flowamp_upload_small

2. Initialize git repository:
   git init
   git add .
   git commit -m "Initial FlowAMP upload (small version)"

3. Add Hugging Face remote:
   git remote add origin https://huggingface.co/esunAI/FlowAMP

4. Push to Hugging Face:
   git push -u origin main

=== Files Included ===

Core Model:
- final_flow_model.py: Main FlowAMP model architecture
- final_sequence_encoder.py: ESM-2 sequence encoding
- final_sequence_decoder.py: Sequence decoding and generation
- compressor_with_embeddings.py: Embedding compression/decompression
- cfg_dataset.py: CFG dataset and dataloader

Training:
- amp_flow_training_single_gpu_full_data.py: Single GPU training
- amp_flow_training_multi_gpu.py: Multi-GPU training
- launch_*.sh: Training launch scripts

Models:
- normalization_stats.pt: Preprocessing statistics
- MODEL_FILES_INFO.md: Information about missing large model files

Apex Integration:
- apex/AMP_DL_model_twohead.py: Apex model architecture
- apex/predict.py: MIC prediction script

Documentation:
- README.md: Comprehensive project documentation
- model_card.md: Hugging Face model card
- usage_example.py: Usage demonstration
- requirements.txt: Python dependencies

=== Note ===
This is a smaller version without large model files due to Hugging Face size limits.
See MODEL_FILES_INFO.md for details on obtaining model weights.