chore: suppress Qwen config warnings in greedy generation

swap: Qwen2.5-0.5B-Instruct (faster, reliable #### format) + fix dp2 detection bug

chore: gitignore __pycache__

fix: context-aware Step token detection for TinyLlama (handle multiple token IDs)

Restore top_p_presence_next field in TimestepArtifacts to match research code

Fix: remove top_p_presence_next arg not in TimestepArtifacts

HF Space: Reasoning Trajectory Demo