Commit History

Autor SHA1 Mensaxe Data
  Justin Lee 4768a41a20 merge utils, added configs and sample notebook to run optimizer hai 3 meses
  Justin Lee 3174e5bd2a code handoff hai 3 meses
  Justin Lee 423231e139 updated mmlu and harness hai 3 meses
  Justin Lee 52c5a76165 made changes to utils hai 4 meses
  Justin Lee 7a014b3e00 update readme hai 4 meses
  Justin Lee d4638ba575 updated gitignore hai 4 meses
  Justin Lee d214437e3e Stop tracking files in eval_results/meta-llama__Llama-3.3-70B-Instruct hai 4 meses
  Justin Lee e1d64ca2f4 update gitignore, added mmlu 0shot and ran a bunch of test hai 4 meses
  Justin Lee 479b1fbbd7 updated mmlu meta-eval for prompt migration hai 4 meses
  Justin Lee caeddccb8d update utils hai 4 meses
  Justin Lee f8a6c7d79f running mmlu pro with meta eval - fixed error hai 4 meses
  Chester Hu 07b191b17e Merge pull request #2 from pia-papanna/tools-refactory-chester hai 11 meses
  Justin Lee dc406b4769 setup meta-eval for benchmark, ray error hai 4 meses
  Justin Lee 21e04c29bf update mmlu pro hai 4 meses
  Justin Lee e19b9e9e34 added fix split, gitignore and download mmlu script hai 4 meses
  Justin Lee 8d3a0479e5 updated env file hai 4 meses
  Justin Lee 9ffb292272 added inspect and modified harness hai 4 meses
  Justin Lee eea96618cf batching and parallelization, ran on baseline and lite hai 4 meses
  Justin Lee 4fd5f29414 revert to previous changes hai 4 meses
  Justin Lee a6f448f362 <Replace this line with a title. Use 1 line only, 67 chars or less> hai 4 meses
  Justin Lee becbe77ff3 attempt to fix json output format in eval hai 4 meses
  Justin Lee 03f2b8eddd change gpu parallel size docs hai 4 meses
  Justin Lee 0bec41f86a updated readme hai 4 meses
  Justin Lee 2776a35314 harness runcode hai 4 meses
  Justin Lee 314b6a874a added updated llama-mmlu-pro and added human-eva hai 4 meses
  Justin Lee 5730a84b8a beef up readme hai 4 meses
  Justin Lee 62b53676fb update harness notebook hai 4 meses
  Justin Lee 1e4c6d22dd update harness notebook hai 4 meses
  Justin Lee e52e1d1ab4 updated prompt migration to use benchmark and also mipro, added meta implementation hai 4 meses
  Justin Lee 4d75fe97b5 update dir hai 4 meses