bhavin-gajjar/laravel-coder-lora-train

Laravel Coder LoRA — end-to-end training data Instruction-tuning datasets for Laravel Bob (Qwen / CodeLlama / DeepSeek modelfiles).Built from official Laravel docs v10.x–v13.x plus version-detection examples. Each config has a single schema — do not mix raw (Alpaca) with chat / lora_* (messages). Configs Config Path Rows Schema raw raw/laravel_training.jsonl 1253 instruction, input, output, topic, version chat chat/laravel_training_chat.jsonl 1253… See the full description on the dataset page: https://huggingface.co/datasets/bhavin-gajjar/laravel-coder-lora-train.