Add complete model with all weights (quantized linear + non-quantized other)

Update config.json with quantization information

Add processor_config.json from official DeepSeek OCR repo

Add modeling_deepseekv2.py from official DeepSeek OCR repo

Add modeling_deepseekocr.py from official DeepSeek OCR repo

Add deepencoder.py from official DeepSeek OCR repo

Add conversation.py from official DeepSeek OCR repo

Add configuration_deepseek_v2.py from official DeepSeek OCR repo

Add LICENSE from official DeepSeek OCR repo

Update README to clarify model.safetensors is the 4-bit packed model

Remove unpacked duplicate, keeping only 4-bit packed model

Remove unpacked duplicate, keeping only 4-bit packed model

Remove unpacked duplicate, keeping only 4-bit packed model

Upload GPTQ 4-bit packed model (75.0% size reduction, 4.0x compression)

initial commit