Hackernews posts about HumanEval

  1. Beating GPT-4 on HumanEval with a fine-tuned CodeLlama-34B (www.phind.com)
    813 points by rushingcreek about 1 year ago | 285 comments
  2. Refact Code LLM: 1.6B LLM for code that reaches 32% HumanEval (refact.ai)
    181 points by kateklink about 1 year ago | 100 comments
  3. 50% on HumanEval with just 1.3B model (twitter.com)
    76 points by sytelus over 1 year ago | 7 comments
  4. Replit announces open source 2.7B code LLM that beats OpenAI Codex at HumanEval (twitter.com)
    11 points by swyx over 1 year ago | 1 comments
  5. InstructCodeT5: 16B model beats every model in HumanEval (twitter.com)
    6 points by debdut over 1 year ago | discuss
  6. Telling Mixtral it is "ChatGPT developed by OpenAI" boosts humaneval score by 6% (twitter.com)
    4 points by reqo 11 months ago | discuss
  7. WizardCoder-34B-Python surpasses GPT-4 on HumanEval (twitter.com)
    3 points by Xeophon about 1 year ago | 1 comments
  8. Fine-tuned CodeLlama beats GPT-4 on HumanEval (huggingface.co)
    3 points by jumploops about 1 year ago | discuss
  9. BigCodeBench: The Next Generation of HumanEval (github.com)
    2 points by t1t1 5 months ago | 1 comments
  10. Code Generation on HumanEval Leaderboard (paperswithcode.com)
    2 points by bcjordan over 1 year ago | discuss
  11. WizardCoder 34B surpasses GPT4 on HumanEval (twitter.com)
    1 points by bytefactory about 1 year ago | 1 comments
  12. Textbooks is all you need: new LLM 51% HumanEval. Better one >10x more params (twitter.com)
    1 points by andromaton over 1 year ago | 1 comments
  13. Azure GPT 3.5 completion endpoint bumps HumanEval from <50% to 74% (twitter.com)
    1 points by convexstrictly over 1 year ago | 1 comments
  14. Show HN: Aflow achieves 96.2% on HumanEval at 4.55% of GPT-4's cost (twitter.com)
    1 points by metagpt about 1 month ago | discuss
  15. HumanEval is saturated: new coding LLM benchmark released (bigcode-bench.github.io)
    1 points by eitanturok 5 months ago | discuss
  16. Running HumanEval Safely with Riza (riza.io)
    1 points by stanleydrew 5 months ago | discuss
  17. Future of NLG evaluation: LLMs and high quality human eval? (ehudreiter.com)
    1 points by fzliu over 1 year ago | discuss
  18. Phind Model beats GPT-4 at coding, with GPT-3.5 speed and 16k context (www.phind.com)
    891 points by rushingcreek about 1 year ago | 347 comments
  19. Show HN: Use Code Llama as Drop-In Replacement for Copilot Chat (continue.dev)
    187 points by sestinj about 1 year ago | 52 comments
  20. Show HN: Fine-Tuning Index of Open-Source LLMs vs. OpenAI (predibase.com)
    4 points by abhaym 6 months ago | discuss
  21. RES-Q: A New Benchmark for Differentiating Frontier Model Capability
    3 points by becklabs 5 months ago | discuss
  22. The most widely used benchmarks for evaluating LLMs
    1 points by kavaivaleri 7 months ago | 1 comments
  23. Beat GPT-4o at Python by searching with 100 dumb LLaMAs (modal.com)
    4 points by thundergolfer 4 months ago | 1 comments
  24. GPT4 Learning from Reflection (github.com)
    3 points by agomez314 over 1 year ago | discuss
  25. Beat GPT-4o at Python with 100 dumb LLaMAs (modal.com)
    1 points by pierremenard 4 months ago | discuss
  26. A thorough comparison of GPT-3.5 vs. GPT-4 programming ability (github.com)
    1 points by jamesmurdza over 1 year ago | discuss
  27. HumaneAI pin maker selling itself for $1B (gizmodo.com)
    26 points by alangibson 6 months ago | 46 comments
  28. Humanimals (twitter.com)
    9 points by dr_dshiv 9 months ago | 1 comments
  29. Refactoring Humanely and "Accidental Pomodoro" (melatonin.dev)
    6 points by wlll about 1 month ago | discuss
  30. How to Kill Bugs Humanely (reducing-suffering.org)
    3 points by lukas099 over 1 year ago | 4 comments