ufal/factgenie

Issues

• Use factgenie structured output schema for LLM-evals in prompts

  #176 opened 9 days ago
  0

• Support no input data

  #162 opened 24 days ago
  0

• Overlapping annotations in browse mode

  #159 opened 25 days ago
  0

• Add tests to be run after adding a new class

  #157 opened a month ago
  0

• Document vLLM usage in wiki

  #156 opened a month ago
  0

• Hashes to identify input, outputs and output_annotations data entries

  #155 opened a month ago
  0

• Add cli command factgenie download data inputs, data outputs and campaigns

  #150 opened a month ago
  1

• Allow users to set OPENAI_API_KEY in the app config

  #147 opened 2 months ago
  3

• Add support for other commercial models

  #145 opened 2 months ago
  1

• Add available resources to factgenie list --available-for-download

  #144 opened a month ago
  1

• Support ESA for MT: Error-Span Annotation for Machine Translation

  #140 opened 25 days ago
  9

• Show resource size

  #139 opened 2 months ago
  0

• Imprint relevant parts of config into generated outputs

  #130 opened a month ago
  0

• Sliders do not correctly show the selected value

  #128 opened 2 months ago
  0

• Agregate statistics for flags, options and inputs

  #126 opened 2 months ago
  1

• Annotator checks

  #122 opened 2 months ago
  0

• Re-think the placement of user data (inputs, outputs, configs)

  #120 opened a month ago
  1

• Prolific URL arguments are not compatible with our URL arguments

  #119 opened 2 months ago
  0

• Examples are not being freed after `idle_time`

  #118 opened 2 months ago
  1

• Factgenie does not start because examples are not on disc

  #116 opened 2 months ago
  1

• Send the final message only after annotations are received

  #114 opened 2 months ago
  0

• Support adding Huggingface datasets through the web interface

  #113 opened 3 months ago
  0

• Flags and selects in LLM eval

  #112 opened 3 months ago
  0

• Prevent overlapping annotations from LLM evaluators

  #111 opened a month ago
  1

• Correctly calculate number of annotated outputs for incomplete splits

  #110 opened 2 months ago
  0

• Selectively remove collected annotations

  #109 opened 2 months ago
  0

• Pre-fill the prompt

  #108 opened a month ago
  1

• Document how to add notes / reasons for individual annotations

  #107 opened 3 months ago
  1

• Clear campaign results

  #106 opened 2 months ago
  0

• Edit campaign config after it is created

  #105 opened 2 months ago
  0

• Differentiate ourselves from other annotation frameworks

  #104 opened 3 months ago
  1

• Add free-form text inputs

  #102 opened 2 months ago
  0

• Add token-level inter-annotator agreement

  #99 opened 3 months ago
  0

• Checking user inputs

  #94 opened 3 months ago
  0

• Robust parsing of int and float model arguments

  #93 opened 3 months ago
  1

• Independent scrolling

  #90 opened 2 months ago
  0

• Use a permalink by default

  #89 opened 2 months ago
  0

• Advanced data management

  #88 opened 3 months ago
  1

• Visualize the outputs from flags, selects and sliders

  #87 opened 2 months ago
  0

• Index model outputs using ids

  #86 opened 3 months ago
  0

• Update wiki

  #84 opened 3 months ago
  0

• Add shuffling modes

  #82 opened 3 months ago
  1

• Prompting and output generation inside Factgenie

  #79 opened 3 months ago
  0

• Sliders, Select boxes, Text input field for annotating the whole output

  #77 opened 3 months ago
  0

• ESA for MT: Support Error Span Annotation for machine translation

  #76 opened 3 months ago
  0

• Support vLLM as backend

  #75 opened a month ago
  1

• Remove unused branches

  #74 opened a month ago
  1

• Make example datasets optional

  #73 opened 3 months ago
  1

• Improve the data selector

  #72 opened 3 months ago
  0

• Newlines are poorly rendered in annotation page

  #69 opened 4 months ago
  0