Commit History

8 days ago

  1. Clean up ephemeral _ephemeral_combo_exp4_priming_results.tar.gz

    Verified signature
    302f2ce
  2. Upload ephemeral combo Exp4 priming report for local pull

    Verified signature
    a825563
  3. Clean up ephemeral combo Exp4 priming staging bundle

    Verified signature
    ae4d8e9
  4. Upload ephemeral combo Exp4 priming staging bundle for A100-80GB

    Verified signature
    b96265a
  5. Clean up ephemeral combo Exp4 priming staging bundle

    Verified signature
    0513005
  6. Upload ephemeral combo Exp4 priming staging bundle for A100-80GB

    Verified signature
    2e9f802
  7. Clean up ephemeral _ephemeral_exp4_results_bundle.tar.gz

    Verified signature
    4c9fbc5
  8. Upload ephemeral experimental-four Arm 3 checkpoint bundle for local pull

    Verified signature
    edee3aa
  9. Clean up ephemeral experimental-four staging bundle

    Verified signature
    bb3e3fc
  10. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    83c7279
  11. Clean up ephemeral combo Exp4 priming staging bundle

    Verified signature
    c9f0afc
  12. Upload ephemeral combo Exp4 priming staging bundle for A100-80GB

    Verified signature
    0d352fa
  13. Clean up ephemeral _ephemeral_exp4_results_bundle.tar.gz

    Verified signature
    66e5d6a
  14. Upload ephemeral experimental-four output bundle for local pull

    Verified signature
    a74a75d
  15. Clean up ephemeral experimental-four staging bundle

    Verified signature
    aa7ccf2
  16. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    51edd45
  17. Clean up ephemeral experimental-four staging bundle

    Verified signature
    f826e36
  18. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    2caa16a
  19. Clean up ephemeral _ephemeral_exp4_results_bundle.tar.gz

    Verified signature
    c3311da
  20. Upload ephemeral experimental-four output bundle for local pull

    Verified signature
    a0bdad5
  21. Clean up ephemeral experimental-four staging bundle

    Verified signature
    a33ef73
  22. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    08dff1e
  23. Clean up ephemeral experimental-four staging bundle

    Verified signature
    1561ba4
  24. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    d43661c
  25. Clean up ephemeral experimental-four staging bundle

    Verified signature
    cb3d76b
  26. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    ed5d743
  27. Clean up ephemeral experimental-four staging bundle

    Verified signature
    4dafc1e
  28. Upload ephemeral experimental-four staging bundle for A100-80GB

    Verified signature
    afd14dd

9 days ago

  1. Publish remediated strict Train-40-isolated 240x checkpoint (max_abs_diff=0.0 noise invariance) & A100-80GB zero-helper-leakage <think> evaluation results (93.39% MBPP / 78.57% MBPP+ / 79.88%-83.54% HE+)

    Verified signature
    cc334a1
  2. Clean up ephemeral A100 evaluation staging bundle

    Verified signature
    18c13cb
  3. Upload ephemeral Clean 240x <think> evaluation staging bundle for A100-80GB

    Verified signature
    2494f49
  4. Add clean 240x <think> Chain-of-Thought evaluation results (HumanEval+ 83.54%, MBPP 95.24%, MBPP+ 79.89%, BCB-Hard 30.41%)

    Verified signature
    866d3ea
  5. Upload submissions/ProbioticFarmer--Qwen3.8-27B-BarronPolyStep-RealCoder-v3--main--bigcodebench-hard-instruct--custom-0-1-sanitized_calibrated.jsonl from Clean 240x <think> A100-80GB evaluation

    Verified signature
    8c7d6ff
  6. Upload submissions/mbppplus_378_polystep_think_samples.jsonl from Clean 240x <think> A100-80GB evaluation

    Verified signature
    144bc04
  7. Upload submissions/humanevalplus_164_polystep_think_samples.jsonl from Clean 240x <think> A100-80GB evaluation

    Verified signature
    4dfd512
  8. Upload clean_240x_think_eval_results.json from Clean 240x <think> A100-80GB evaluation

    Verified signature
    57b3bcd
  9. Upload Clean 240x <think> Chain-of-Thought A100-80GB evaluation results

    Verified signature
    c60da0d
  10. Upload lightweight Clean 240x <think> CoT evaluation bundle for A100-80GB

    Verified signature
    9295334
  11. Upload lightweight Clean 240x <think> CoT evaluation bundle for A100-80GB

    Verified signature
    9c3c3d1
  12. Update to clean 240x Barron-PolyStep RealCoder-v3 (20,400 steps / 3.30M vertex evals, zero leakage)

    Verified signature
    41fa122

10 days ago

  1. Update model card: 280x RealCoder-v5 (24,000 steps + 40x DeepSeek Distillation), HE+ 79.27%, MBPP+ 77.78%, BCB-Hard 29.73%

    Verified signature
    54a49c8
  2. Upload submissions/ProbioticFarmer--Qwen3.8-27B-BarronPolyStep-RealCoder-v3--main--bigcodebench-hard-complete--custom-0-1-sanitized_calibrated_eval_results.json from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    65332d6
  3. Upload submissions/ProbioticFarmer--Qwen3.8-27B-BarronPolyStep-RealCoder-v3--main--bigcodebench-hard-complete--custom-0-1-sanitized_calibrated.jsonl from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    4aa3b8b
  4. Upload submissions/ProbioticFarmer_Qwen3.8-27B-BarronPolyStep-RealCoder-v3_ProbioticFarmer.json from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    3517ad8
  5. Upload submissions/EVALPLUS_LEADERBOARD_SUBMISSION.md from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    458e9cd
  6. Upload submissions/mbppplus_378_polystep_samples.jsonl from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    b68ff8d
  7. Upload submissions/humanevalplus_164_polystep_samples.jsonl from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    fa623b4
  8. Upload LEADERBOARD_ALL3_REPORT.md from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    aece980
  9. Upload leaderboard_all3_hf_results.json from 3-Step Leaderboard A100-80GB pipeline

    Verified signature
    83e14d1
  10. Upload 3-Step Public Leaderboard A100-80GB results, BCB-Hard 148, MBPP+ 378, and Merge Verification

    Verified signature
    d98ac25