Commits · 5c4519e8edd55b12e14e3cf3573f5a51b5784735 · Grießhaber Daniel / evoprompt

Jul 25, 2024
- Change repo structure and improve logging · 5c4519e8
  Max Kimmich authored 8 months ago
  
  5c4519e8
Jun 04, 2024
- Allow to generate unique paraphrases · 37d991ad
  Max Kimmich authored 9 months ago
  
  37d991ad
May 28, 2024
- Allow to use pre-trained llama model from HF hub · 5b1c997f
  Max Kimmich authored 9 months ago
  
  5b1c997f
May 27, 2024
- add missing parameter in QuestionAnswering.evaluate · f09ffba8
  Grießhaber Daniel authored 10 months ago
  
  f09ffba8
- add option to disable early stopping · eb8df827
  Grießhaber Daniel authored 10 months ago
  
  eb8df827
- add manual prompt evolution endpoints · f73cc390
  Grießhaber Daniel authored 10 months ago
  
  f73cc390
May 24, 2024
- updated frontend version · 57f9c9ac
  Grießhaber Daniel authored 10 months ago
  
  57f9c9ac
- add endpoints for frontend integration of run capability · c2ad8f7d
  Grießhaber Daniel authored 10 months ago
  
  c2ad8f7d
May 22, 2024
- save snapshot during evolution process · 7fcaaca7
  Grießhaber Daniel authored 10 months ago
  
  7fcaaca7
May 21, 2024
- move to thread based separation of backend and optimizer · 3cef8395
  Grießhaber Daniel authored 10 months ago
  
  3cef8395
- fix reverse proxy serving during debug · 8226548c
  Grießhaber Daniel authored 10 months ago
  
  8226548c
- Merge branch 'master' into integrate-frontend · e7ccb8aa
  Grießhaber Daniel authored 10 months ago
  
  e7ccb8aa
May 17, 2024
- Fix frontend serving · 2c0102c4
  Max Kimmich authored 10 months ago
  
  2c0102c4
- Merge branch 'backend' into integrate-frontend · c8362914
  Max Kimmich authored 10 months ago
  
  c8362914
- Merge branch 'backend' · d819c1e6
  Max Kimmich authored 10 months ago
  
  d819c1e6
- Remove PyTorch dependency · a4495a8f
  Max Kimmich authored 10 months ago
  
  a4495a8f
- Add optimization backend for HTTP API · 9976307b
  Max Kimmich authored 10 months ago
  
  9976307b
May 13, 2024
- add frontend as submodule · 942aa86d
  Grießhaber Daniel authored 10 months ago
  
  942aa86d
Apr 29, 2024
- Merge branch 'early-stopping' into backend · 79a99169
  Grießhaber Daniel authored 10 months ago
  
  79a99169
- add support for llama3 · 4f82a862
  Grießhaber Daniel authored 10 months ago
  
  4f82a862
- Add HTTP API endpoints · aa579769
  Max Kimmich authored 10 months ago
  
  aa579769
Apr 25, 2024
- added parent based early stopping monitor · b77a62b7
  Grießhaber Daniel authored 11 months ago
  
  b77a62b7
- implement moment-based early stopping · 28441f6d
  Grießhaber Daniel authored 11 months ago
  
  28441f6d
- move cli parameters to local modules · d47234e3
  Grießhaber Daniel authored 11 months ago
  
  d47234e3
- refactor evaluation to have a central point to implement early stopping · 82f55d87
  Grießhaber Daniel authored 11 months ago
  
  82f55d87
Apr 16, 2024
- Refactor more · cbd240bc
  Max Kimmich authored 11 months ago
  
  cbd240bc
Apr 10, 2024
- Re-add model usage for evolution and evaluation · 7886a510
  Max Kimmich authored 11 months ago
  
  7886a510
Apr 08, 2024
- Refactor code for better modularization · 3f29409c
  Max Kimmich authored 11 months ago
  
  3f29409c
Apr 04, 2024
- Improve code readability · dee5c448
  Max Kimmich authored 11 months ago
  
  dee5c448
Mar 19, 2024
- Create unique run names · 300ba08e
  Max Kimmich authored 1 year ago
  
  300ba08e
- Add more replacements for context for QA · 0aff2b5c
  Max Kimmich authored 1 year ago
  
  0aff2b5c
- Adding console handler for logging · 8ee4b027
  Max Kimmich authored 1 year ago
  
  8ee4b027
- Add system message for OpenAI model · 356db033
  Max Kimmich authored 1 year ago
  
  356db033
Mar 18, 2024
- Fix llama2 prompt format · ec6c2f39
  Max Kimmich authored 1 year ago
  
  ec6c2f39
- add model usage counter for individual prompts and models · 14c65c4d
  Grießhaber Daniel authored 1 year ago
  
  14c65c4d
Mar 14, 2024
- make sure cli options are logged after run directory is initialized · 6f2fb95b
  Grießhaber Daniel authored 1 year ago
  
  6f2fb95b
Mar 13, 2024
- make experiment name generation compatible with llama model that can't follow... · c69122db
  Grießhaber Daniel authored 1 year ago
  
  make experiment name generation compatible with llama model that can't follow simple instructions :^)
  c69122db
- let the llm create the run name · 5218f36a
  Grießhaber Daniel authored 1 year ago
  
  5218f36a
- make --chat and --use-grammar mutually exclusive for evaluation task · 2f3cbc02
  Grießhaber Daniel authored 1 year ago
  
  2f3cbc02
- add --use-grammar flag to diable grammars in tasks · 4c276ebc
  Grießhaber Daniel authored 1 year ago
  
  4c276ebc