HazyResearch/metal

Issues

• Not support Networkx 3.2.1

  #214 opened a year ago by xinyi-zhao
  0

• Singular error

  #213 opened 3 years ago by prionnet
  0

• Problems with networkx when using LabelFunction

  #212 opened 5 years ago by stenpiren
  1

• Bug in Metal.multitask.mt_end_model, line 114.

  #177 opened 5 years ago by kexiao17
  2

• I think there are some errors in the mt_end_model.py

  #200 opened 6 years ago by WenxiongLiao
  1

• when applying metal, how to estimate noisy-label probability for multilabel problem of a single task

  #198 opened 5 years ago by yyht
  2

• Make special (but common) case of k=2 more efficient

  #133 opened 5 years ago by bhancock8
  1

• Set `padding_idx=0` by default in `nn.Embeddings`

  #107 opened 5 years ago by jason-fries
  1

• Add gradient clipping in Classifier._train_model()

  #136 opened 5 years ago by bhancock8
  1

• Supporting "default" class

  #208 opened 5 years ago by astarostap
  1

• Pass around RandomState instead of overwriting seeds

  #140 opened 5 years ago by bhancock8
  1

• Speed up unit tests

  #138 opened 5 years ago by bhancock8
  1

• Add AWS runner + kwargs in Tuner object

  #147 opened 5 years ago by jdunnmon
  1

• Recommendation: Use logging rather than prints

  #170 opened 5 years ago by lukehsiao
  1

• Change `'Could not find kwarg "{k}" in destination dict.'` to `warning.warn` command.

  #119 opened 5 years ago by vincentschen
  0

• End model seems to always do better when assuming a 50/50 class balance

  #207 opened 5 years ago by astarostap
  1

• Question on LF influence on Snorkel predictions based on coverage

  #209 opened 5 years ago by astarostap
  1

• Change LF stat "empirical accuracy" to "precision"

  #210 opened 5 years ago by astarostap
  1

• How does an end model improve upon Snorkel?

  #206 opened 5 years ago by astarostap
  1

• Potential bug for the predict_proba function in label_model.py

  #202 opened 5 years ago by georgezgf
  0

• How to use weak supervised learning with MMTL

  #203 opened 5 years ago by yzj19870824
  1

• Generative model snorkel vs Generative model metal

  #179 opened 5 years ago by Wronskia
  3

• When is v0.5 being released ? Would it be significant refactor ?

  #172 opened 5 years ago by shishircc
  2

• Using BERT with MeTaL

  #178 opened 5 years ago by Peter-Devine
  6

• links to basic snorkel tutorials are broken.

  #204 opened 5 years ago by jennxf
  1

• Change model saving/loading in `Classifier`

  #106 opened 6 years ago by jason-fries
  1

• Keep track of changes as CHANGELOG

  #199 opened 5 years ago by HiromuHota
  2

• issue keras or Tensorflow in the future?

  #196 opened 5 years ago by WenxiongLiao
  1

• MMTL training accuracies swinging between two datasets

  #191 opened 6 years ago by Peter-Devine
  3

• multiple

  #197 opened 6 years ago by yzj19870824
  0

• [QUESTION] Mixing textual and numerical inputs with multi-branch models

  #184 opened 6 years ago by f-lng
  2

• Checkpoint cleanup for simultaneous experiments

  #128 opened 6 years ago by jdunnmon
  1

• KeyError: 'METALHOME' error from MMTL_Basics tutorial

  #182 opened 6 years ago by Peter-Devine
  1

• Annotate multiple entity/relation in text dataset [example / guideline]

  #167 opened 6 years ago by msank00
  0

• Make log_valid_metrics, checkpoint_metric, and validation metric compatible

  #166 opened 6 years ago by jdunnmon
  1

• Startup problem - Basics.ipynb ( Need to know what I am missing)

  #152 opened 6 years ago by rekhasinghal
  1

• tqdm error in RandomSearchTuner

  #143 opened 6 years ago by mlu355
  1

• metal/util/pred_to_prob Squeezing 1d tensor with 1 element to 0d tensor

  #141 opened 6 years ago by dliangsta
  1

• Validate on arbitrary metrics and directions

  #124 opened 6 years ago by bhancock8
  1

• Postpone averaging loss as long as possible

  #123 opened 6 years ago by bhancock8
  3

• grid search creates overlapping logs in tensorboard

  #120 opened 6 years ago by vincentschen
  2

• Allow finer fidelity logging

  #129 opened 6 years ago by bhancock8
  4

• Upgrade to pytorch 1.0

  #132 opened 6 years ago by bhancock8
  1

• Replace .cuda() calls with .to(device) calls

  #135 opened 6 years ago by bhancock8
  0

• `LSTMModule` in main or contrib?

  #114 opened 6 years ago by ajratner
  6

• Report Tuner search results in pandas dataframe for easy sorting and grouping

  #131 opened 6 years ago by bhancock8
  0

• Checking on Basic Test Case

  #127 opened 6 years ago by CatalinVoss
  3

• Unusually Long Training Time

  #113 opened 6 years ago by danich1
  7

• Current 0.1 pip version fails on import with syntax error

  #111 opened 6 years ago by CatalinVoss
  1

• Easy support for searching over all EndModel sub-module hyperparams

  #104 opened 6 years ago by ajratner
  0