HowieHwong/TrustLLM

Issues

• Can it be extended to suit my own fine-tuned huggingface model?

  #44 opened 2 months ago by mxtsecure
  2

• AttributeError : HARM_CATEGORY_DEROGATORY

  #43 opened 4 months ago by andytan0051
  5

• Add a unified path for dataset

  #19 opened a year ago by HowieHwong
  0

• unhashable type: 'list'

  #24 opened 2 months ago by jinz2014
  8

• KeyError when running `run_truthfulness`

  #42 opened 6 months ago by gauravjain14
  1

• Unable to reproduce the results of advGLUE

  #22 opened 6 months ago by AboveParadise
  4

• a key change from "mission_open-ended" to "mission_open_ended"

  #30 opened 6 months ago by jinz2014
  1

• Misleading default config documentation

  #41 opened 6 months ago by ic-768
  4

• It seems that the entire library is unavailable and full of errors.

  #33 opened 9 months ago by mxdlzg
  5

• Gold "answer" ID is not supposed to be 0 in the codah data set

  #39 opened 7 months ago by jinz2014
  3

• Reponse from the model is Ah, a chatbot! I'm happy to help you with any questions or tasks you may have. How can I assist you today?

  #26 opened 10 months ago by jinz2014
  6

• suggestion: add PKU-Alignment/BeaverTails and R-Judge dataset into safety section

  #38 opened 8 months ago by yananchen1989
  0

• The parameter "device" of the function "LLMGeneration" can not work !

  #37 opened 8 months ago by xingyu8wang
  0

• Strange and nonsensical results in generation

  #35 opened 8 months ago by eyb12
  4

• question about the ground truth for the Adversarial Factuality

  #34 opened 9 months ago by jinz2014
  1

• KeyError: 'res'

  #15 opened a year ago by rggithub7
  1

• Data Format

  #32 opened 10 months ago by redwyd
  2

• question about the mission_open_ended subset

  #31 opened 10 months ago by jinz2014
  1

• replacement of a GPT evaluator with an open-source evaluator

  #29 opened 10 months ago by jinz2014
  0

• "culture" dimension not available in awareness.json

  #27 opened 10 months ago by jinz2014
  1

• scripts to reproduce the results in the paper

  #25 opened a year ago by jinz2014
  1

• Wrong computation of metrics for implicit ethics

  #23 opened a year ago by Nikis14
  1

• 使用示例代码出错 Invalid device: cuda:0

  #21 opened a year ago by oyy2000
  2

• generation.py中调用函数prompt2conversation()参数缺少一个model_path

  #20 opened a year ago by liano3
  1

• AssertionError: Torch not compiled with CUDA enabled

  #14 opened a year ago by rggithub7
  1

• Empty generation results

  #11 opened a year ago by pearlmary
  11

• Typo in code utils/gpt_auto_eval.py

  #10 opened a year ago by chichidd
  0

• Add a `conda` install option for `trustllm`

  #7 opened a year ago by sugatoray
  2

• Add a workflow to publish the docs with github actions

  #5 opened a year ago by sugatoray
  0

• Update `.gitignore` file: use a standard file for python repositories

  #3 opened a year ago by sugatoray
  0

• Fix docs: incorrect python code in example

  #1 opened a year ago by sugatoray
  1