Spaces:
Paused
Paused
Commit History
-
Sync Space with HCM-21-Private main (72f5f30): reward recalibration, seeded RNG streams, MCP integration, poster artifacts
Verified signature670ccf0
-
Optimize Docker build: add .dockerignore, layer caching, drop editable install
- Add .dockerignore to exclude .git/, __pycache__/, tests/, notebooks/ from build context - Separate dependency install from source copy for better layer caching on rebuilds - Replace editable install (-e) with standard install (not needed in production) Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>20f0740 -
Fix HF Space runtime error: replace failing base image with python:3.12-slim
The openenv-base image's init script tries to curl huggingface.co during startup, which fails DNS resolution inside the HF Spaces runtime (exit code 6), causing HTTP 503 and breaking WebSocket connections from Colab. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>f38d204 -
Fix GRPO all-zero rewards: parse JSON actions from completions
Remove environment_factory from GRPOTrainer — reward_func now parses JSON action objects from model text completions via regex and executes them in fresh HCM21Env instances. This bypasses TRL's tool-calling infrastructure which fails with a 0.6B model + Unsloth patches. Changes: - SYSTEM_PROMPT: instruct model to output JSON action lines - reward_func: parse actions, run in env, return scores - GRPOTrainer: remove environment_factory arg - Evaluation: single-shot generation matching training format - Fix BatchEncoding .shape bug in eval cell - Add CLAUDE.md with GRPO training architecture docs Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>1ac70ff -
Fix GRPOConfig: use fp16 for T4 GPU, remove deprecated logging_dir
T4 (Turing) doesn't support bf16 — switch to fp16. Replace deprecated logging_dir param with TENSORBOARD_LOGGING_DIR env var. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>8fdc935 -
Fix GRPOConfig: remove chat_template_kwargs unsupported by Unsloth
Unsloth's patched GRPOConfig strips chat_template_kwargs from its init signature, causing training to fail. Disable Qwen3 thinking mode via tokenizer.chat_template instead. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>2a6a311 -
Add demo script, API endpoints, and training notebook baselines
- Add demo.py: standalone heuristic agent for full 6-quarter episodes (CLI + JSON API, no LLM key required) - Add /demo, /hyperparameters, /scenarios API endpoints to HF Space - Update Dockerfile to include demo.py and agent/ directory - Enhance training notebook with baseline comparisons and reward curves: - Random baseline (5 seeds, ~0.15-0.25 floor) - Heuristic baseline via /demo API (~0.50-0.65 target) - Reward + loss curve plotting after GRPO training - Before/after evaluation on same seeds - Results summary with bar charts and JSON export Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>80d9920 -
Add usage guide with demo walkthrough and API reference
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>01d7974 -
Fix training notebook to use WebSocket client for stateful sessions
The OpenEnv REST endpoints (POST /reset, /step) are stateless. Updated the Colab training notebook to use GenericEnvClient (WebSocket) which maintains a stateful environment session per connection. Verified: full quarter cycle works end-to-end against production (https://paretooptimal-hcm21.hf.space). Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>7d19ba7 -
Deploy to HF Spaces + add TRL/Unsloth training notebook
- Add OpenEnv-compliant root-level wrappers (__init__.py, client.py, models.py, server/) - Update Dockerfile to use ghcr.io/meta-pytorch/openenv-base with health checks - Add HF Spaces frontmatter to README.md (sdk: docker, app_port: 8000) - Update pyproject.toml with [project.scripts] server entry point - Update openenv.yaml to point to server.app:app - Add notebooks/hcm21_trl_training.ipynb: GRPO training with Unsloth on Colab T4 - Passes `openenv validate` — deployed and running at https://huggingface.co/spaces/ParetoOptimal/hcm21 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>c7d603c -
Add TAM/growth analysis, hackathon context, and demo talk track
- README: Add OpenEnv Hackathon Statement 2 submission context at top - README: Add $34B+ TAM breakdown (HR Analytics, HCM Software, AI in HR) - README: Add growth inflection thesis, whitespace analysis, GTM paths - README: Add detailed Statement 2 compliance mapping table - DEMO.md: 1-minute demo talk track with timed sections and live commands Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>e5c66f5 -
Update README.md
Unverified signature1b85a9e -
Fix punctuation in README description
Unverified signatureCorrected punctuation in the description for clarity.b449973 -
Initial release: HCM:21 HR Productivity Measurement Environment
OpenEnv-compliant simulation for evaluating LLM agents on long-horizon HR decision-making using Fitz-enz frameworks (HCVA, HCROI, QIPS, Five Indexes of Change). 6-quarter episodes with 150-200+ steps, stochastic events, and cross-quarter dependencies. Includes: - Full environment (company sim, employee lifecycle, metrics, scoring) - HCM:21 phase system (Scanning/Planning/Producing/Controlling) - 4 task scenarios (high turnover, budget cuts, rapid growth, balanced) - Sample Claude-based agent with phase-aware prompting - 43 passing unit tests - Docker + OpenEnv manifest for competition submission Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>8697ae6