Fix device mismatch in notebook forward pass and generation cells

Remove redundant detect_scenario from LossDebugger

Add Code CPT pipeline for injecting Python code capability

Update 02_model notebook: switch default to debug_10m and improve model summary display

Translate all notebook content from Korean to English

Clear notebook outputs and add real checkpoint diagnosis cell

Add missing mock_history cases for full diagnose_status coverage

Fix debugger reference data and scenario detection logic

Update upload notebook for pretrained LLaMA 2 tokenizer

Standardize notebooks for Colab/local dual environment support

Add configurable wandb log directory path

Clean up error output from model notebook

Use LLaMA 2 pretrained tokenizer and remove tokenizer_mode option

Fix BPE tokenizer ByteLevel decoder and update evaluation notebook

feat(notebook): add HuggingFace Hub upload notebook

feat(training): add LossDebugger 5-level diagnostic framework

refactor(notebook): replace manual TrainConfig args with preset classmethod

refactor(notebook): replace manual param counting with torchinfo.summary in 02_model

Initial commit: LLM-1B-Lab project setup