Fix: Add chat_template to tokenizer_config.json

Update README: Describe as Merged (base + adapter fused) format

Fix: Upload correct full-precision fp16 model (was MLX quantized)

Fix: Remove MLX quantization config for transformers compatibility

Fix: Add quant_method to quantization_config

Improve model card with professional structure (inspired by Chess Tutor)

Add YAML metadata to model card

Upload DJ-AI-Radio - Tanglish DJ Radio model

initial commit