Add custom API title and description with shapiron branding and live base URL 64de1cc shapiron commited on Mar 12
Working GRPO training script (dry run passed on H100) f82bba7 shapiron Claude Sonnet 4.6 commited on Mar 8
Fix GRPO tensor mismatch: pre-format prompts as strings, left-pad, shrink completion length a7b51dc shapiron Claude Sonnet 4.6 commited on Mar 8
Fix: remove fast_inference flag (requires vLLM not installed on H100) 022dcfb shapiron Claude Sonnet 4.6 commited on Mar 8
Add Auditron multi-agent procurement oversight environment b2a4e1f shapiron Claude Sonnet 4.6 commited on Mar 8