Ginnipahwa05/Meta-Pytorch

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Sync latest validator fixes from dashboard repo

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Keep task scores safely away from validator boundaries

Sync Hugging Face repo with dashboard repo

Restore grader factory for inference startup

test

Apply strict safe-score clamping throughout graders

Sync Hugging Face repo with dashboard repo

Sync Hugging Face repo with dashboard repo

Add explicit grader metadata to tasks

Align task grader metadata with score policy

Update inference scoring output

Update grader logic

Clamp task scores to 0.001-0.999

Align submission score ranges across code and spec

Clamp exposed task grades to safe range

Keep task scores safely away from boundaries

Widen grader score clamp margin

Run three graded tasks and clamp score bounds

Clamp grader outputs to open interval

Require evaluator LLM proxy env vars

Use evaluator API_KEY for LLM proxy

Make server app runnable for validator

Add multi-mode deployment metadata

Add pyproject for OpenEnv validation

Add missing seed script for Space

Fix scripts package import for Space

Add Hugging Face Spaces README config

Add OpenEnv hackathon submission app

initial commit