nshepperd/gpt-2

Issues

• Training on TPU

  #89 opened 3 years ago by Dhanachandra
  1

• GPT-2 doesn't have Encode.py or Train.py.

  #88 opened 3 years ago by thegoldenboy542
  0

• how to use trained data to generate text?

  #87 opened 3 years ago by bag7dad
  1

• ModuleNotFoundError: No module named 'Sampler'

  #79 opened 4 years ago by haiitsstrawberry
  3

• avg stays in 2.6-2.9 range

  #86 opened 3 years ago by freedmann2
  0

• ModuleNotFoundError: No module named 'encoder'

  #73 opened 3 years ago by shreesha345
  7

• How can I use 1558m data or 775 data instead of using 177m data to train my own model?

  #57 opened 4 years ago by drizzt00s
  2

• Encoder doesn't work

  #83 opened 4 years ago by ruskyvisky
  1

• I can't train my dataset

  #80 opened 4 years ago by shreesha345
  3

• this parameter cannot be zero when i using my file history.npz

  #82 opened 4 years ago by WiNE-iNEFF
  0

• How to generate interactive conditional samples after retraining on custom dataset?

  #51 opened 5 years ago by nikilp
  4

• Need some help getting Tensor Rematerialization to work

  #81 opened 4 years ago by WhereAmO
  7

• Dockerfile no longer accurate?

  #78 opened 4 years ago by teledemic
  2

• GPT2 Fine Tuning does not support Ampere (RTX 3000s) Cards

  #77 opened 4 years ago by imakemoneymoves
  0

• File "encode.py", line 23

  #76 opened 4 years ago by Jucamhc
  1

• Cannot download the models

  #75 opened 4 years ago by giorgiogiudice
  4

• OOM on 345M with GPU

  #69 opened 4 years ago by babaraza
  4

• gpt2 translation task

  #65 opened 4 years ago by jzl0166
  1

• Where to enter model name as parameter

  #74 opened 4 years ago by kalyons1
  1

• "ModelNotFoundError": No model named "encoder"

  #72 opened 4 years ago by shreesha345
  0

• Error on running Encode.py: "ModuleNotFoundError: No module named 'regex'"

  #54 opened 4 years ago by samysa-kr
  2

• Early Stopping

  #68 opened 4 years ago by bala1802
  0

• how do I train gpt-2 using multiple encoding files?

  #64 opened 4 years ago by ca3games
  1

• refine-tuning by GPU generating repeated words

  #60 opened 4 years ago by drizzt00s
  1

• How to train in multiple gpu

  #66 opened 4 years ago by teja835
  1

• Error when calculating the validation loss - indices[0,1200] = 1200 is not in [0, 1024)

  #63 opened 4 years ago by yijunzhouzoey
  1

• Process gets killed when training

  #61 opened 4 years ago by 50417
  1

• Encoding large single text files is not working

  #59 opened 4 years ago by 0TT0mation
  1

• Why the label of training is like this

  #58 opened 4 years ago by SchenbergZY
  0

• Image GPT Training

  #56 opened 4 years ago by cellininicholas
  0

• Training on distributed machine is slow. Using 8 Nvidia V100.

  #28 opened 5 years ago by dimeldo
  8

• How to ACTUALLY train 345M on Multiple GPU using train-horovod.py?

  #53 opened 5 years ago by shamiul94
  0

• What is the minimum size of GPU I need to set batch_size more than 1 to train 345M model using train.py?

  #52 opened 5 years ago by shamiul94
  0

• Encoding on GPU

  #43 opened 5 years ago by Zer0-dev115
  2

• Train.py issues windows 10

  #35 opened 5 years ago by veldermon
  1

• Sample Length

  #50 opened 5 years ago by cachilders
  0

• How to get the model embeddings?

  #47 opened 5 years ago by Leggerla
  0

• Is possible to use this train script for another language dataset? In order to train it from start in a new language.

  #49 opened 5 years ago by nikkon3
  2

• Manipulating how a sample should start

  #48 opened 5 years ago by fabpapi
  0

• NameError: name 'How' is not defined

  #45 opened 5 years ago by mikkokotila
  0

• dataset

  #44 opened 5 years ago by AlgoRhythm-Technologies
  1

• Encode of a new dataset, confused about <|endoftext|> encoding

  #33 opened 5 years ago by AliceZhang2016
  4

• Finetuning on the Full Model - OOM 1558M

  #37 opened 5 years ago by vince-lynch
  10

• Batch size in training GPT_2

  #41 opened 5 years ago by ngocpham97
  1

• Apologies, but HELP

  #32 opened 5 years ago by OdincoGaming
  4

• About Perplexity

  #36 opened 5 years ago by curly0613
  0

• Unicode characters each considered as a token

  #34 opened 5 years ago by Masum06
  2

• Consued about vocab and encoder

  #31 opened 5 years ago by weiguowilliam
  0

• Sampling structure looks weird. Maybe becuase I'm structuring my data wrongly?

  #29 opened 5 years ago by dimeldo
  0

• Train loss

  #26 opened 5 years ago by alecalma
  0