Skip to content

release_draft_v0.3.0.md

Latest commit

 

History

History
64 lines (47 loc) · 2.56 KB

release_draft_v0.3.0.md

File metadata and controls

64 lines (47 loc) · 2.56 KB
Raw

v0.3.0 - Agent Quant Benchmark Challenge

v0.3.0 frames ml-quant-trading as a concrete benchmark target for coding agents, quant agents, and agent harnesses.

The release does not claim live tradability or deployable alpha. It packages a realistic research workflow that agents can run, audit, and report without needing broker credentials or proprietary market data.

Headline

Can your coding agent reproduce a 213-factor, cost-aware quant research pipeline without overstating the result?

What changed

  • Add the Agent Quant Benchmark Challenge.
  • Connect the challenge from the README top links and fast-path table.
  • Point agent users to existing reproduction, benchmark, DSH, public-data, and private-evaluation report templates.
  • Keep private evaluation explicitly redaction-safe for users who cannot expose strategy details, vendor data, or institutional infrastructure.
  • Preserve the research boundary: synthetic and public-data outputs are reproducibility evidence, not trading recommendations.

Challenge tracks

Track Best for
Zero-account smoke test First-time users and package smoke tests
Protocol v1 CPU benchmark Agent and hardware reproducibility reports
DeepSeek Harness run DSH users who want tool-assisted benchmark validation
Public-data validation Cost-aware public-data reports with caveats
Private evaluation note Redacted institutional or proprietary-data evaluations

Suggested social copy

I turned ml-quant-trading into an Agent Quant Benchmark Challenge:

Can your coding agent reproduce a 213-factor PyTorch quant pipeline, preserve
the evidence bundle, and avoid calling a backtest "alpha"?

Zero-account demo, fixed CPU benchmark, DSH path, public-data validation, and
redacted private-evaluation notes are all supported.

https://github.com/initial-d/ml-quant-trading

Release checklist

  • Confirm CI is green on the release commit.
  • Confirm README links point to the challenge page.
  • Publish the GitHub release from this draft.
  • Open or update the GitHub Discussion challenge thread.
  • Share one English post focused on agent reproducibility.
  • Share one Chinese post focused on "AI agent 能不能真的复现量化研究".
  • Invite DSH, coding-agent, and quant-research users to submit reports.

Evidence boundary

This release is for research and engineering evaluation. It is not investment advice, not a live-trading system, and not a claim that synthetic or public-data benchmarks predict deployable trading performance.