fix(train): accept wandb_v1_ key format in W&B gate check

fix(grpo): allow FATHOM_USE_VLLM=0 to bypass vLLM rollout and avoid IS-ratio collapse under QLoRA

fix(train): tolerate missing/invalid WANDB_API_KEY, surface key length in smoke

fix(reward): align prompt with SFT, soft format, real recursion signal (A.3 + A.4-bis)

fix(grpo): align sys_msg with SFT - emit <answer>...</answer>

fix(grpo): tail-truncate context inside _to_prompt to satisfy vllm input-length gate

fix(grpo): TRL 1.2 stock path + LoRA double-wrap fix + smoke memory caps

feat: phase 1 complete — smoke green on HF Jobs, training scripts, plot generator, Colab notebook, submission preflight

clean repo without secrets or data