mingkaid/rl-prompt

Issues

• truncation_side=left as parameter in GPT2LMHeadModel.from_pretrained()

  #46 opened 3 months ago by Ancientpixels
  0

• Repeating tokens in optimized prompt

  #45 opened 7 months ago by AMJasser
  0

• seems have a bug in evaluate function

  #42 opened a year ago by A11en0
  1

• question: Is the model used in the policy network always distilgpt?

  #36 opened a year ago by pascalhuszar
  3

• Can we use this as a layer between user and RAG model?

  #43 opened a year ago by deepakdev1507
  1

• Questions on the Gradients of LLM

  #41 opened a year ago by Schwartz-Zha
  0

• Baselines results

  #40 opened a year ago by Davido111200
  0

• BrokenPipeError: [Errno 32] Broken pipe

  #37 opened a year ago by Xinhui-Zhu
  2

• Reproducibility and randomness

  #38 opened a year ago by YasamanJafari
  1

• Why does this method need so much steps?

  #39 opened a year ago by A11en0
  2

• About the RL training

  #34 opened a year ago by FayeXXX
  3

• About the prepended special character \u0120.

  #35 opened a year ago by guozix
  3

• Network is unreachable

  #32 opened a year ago by rabi-fei
  1

• Train using vertexai

  #33 opened a year ago by yguezpa
  1

• A question about ppl score

  #29 opened 2 years ago by FayeXXX
  7

• question

  #31 opened a year ago by 18712234451
  2

• CUDA Out of Memory training on shakespeare dataset

  #30 opened 2 years ago by DwyaneLQY
  2

• Clarification on the RL problem

  #28 opened 2 years ago by hv68
  1

• classification with gpt & training time

  #27 opened 2 years ago by MatthewCYM
  2

• classifcation with gpt

  #26 opened 2 years ago by MatthewCYM
  2

• RL-prompt MLP loss

  #25 opened 2 years ago by hv68
  1

• Scope of this project

  #24 opened 2 years ago by YujingYang666777
  2

• An error about Hydra when running examples/few-shot-classification

  #23 opened 2 years ago by plusnli
  7

• A question about prompt initialization

  #22 opened 2 years ago by beeevita
  3

• A question about how to judge the performance of the prompts after running "run_fsc.py" file?

  #12 opened 2 years ago by jasonyin718
  3

• ImportError

  #21 opened 2 years ago by beeevita
  2

• output data of your experiment

  #18 opened 2 years ago by li-jing-wen
  4

• question about using greedy search during inference

  #17 opened 2 years ago by lihenglin
  1

• assert len(prompt_strs) == len(source_strs) fails during inference

  #15 opened 2 years ago by mahdiabdollahpour
  4

• Transferring Prompts across LMs

  #13 opened 2 years ago by 52ie
  3

• some Doubts about a symbol

  #10 opened 2 years ago by oujieww
  2

• Could RLPROMPT be applied to zero-shot settings?

  #9 opened 2 years ago by shadowkiller33
  5

• RuntimeError

  #8 opened 2 years ago by Ericmututu
  2

• Lack of evaluation in `modular-interface` branch

  #6 opened 2 years ago by Dinxin
  4

• Could not find the exact place where the training set is input?

  #5 opened 2 years ago by Dinxin
  4

• The data in 'data/prompt-gpt2-vocab' folder seems meaningless?

  #4 opened 2 years ago by Dinxin
  1

• generate_prompts.py may be only usable for `input_specific=True` option?

  #3 opened 2 years ago by Dinxin
  1