v4 notebook: fix dtype Half/BFloat16 mismatch (explicit bf16), fix tied embeddings path, fix max_length warning

v4 notebook: fix TypeError crash, suppress warnings, update paths to CWD, add V3 task-aware system prompts

Fix total_mem → total_memory in V4 notebook (PyTorch API)

Add V4 Instruct-Only GRPO notebook implementing ADR-002

ADR-002: V4 Instruct-Only GRPO — revises dual-model plan based on model repo audit

Add comprehensive investigation report — performance audit, unexplored alternatives, literature-backed recommendations

Add session checkpoint: v3 launch decision with full context

apply v3 task-aware thinking controls and delete deprecated notebook

Add v3 thinking control patch - task-aware system prompts + think efficiency reward

Initial commit: Tucano2-Commerce GRPO v3 training pipeline

Upload grpo_vertex_v3.ipynb

Rename notebooks/grpo_vertex_v3.ipynb to notebooks/DEPRECATED_grpo_vertex_v3.ipynb

Delete grpo_vertex_v3.md

tools: add md-to-ipynb converter script

feat: add v3 notebook (.ipynb) — ready for Vertex AI Workbench

feat: add GRPO v3 implementation with entropy collapse fixes

Create grpo_vertex_v2_ipynb.md

docs: add ADR-001 next steps with detailed execution plans

docs: add project documentation

initial commit