tunib-ai/parallelformers

Issues

• Cross-node inference

  #48 opened 2 years ago by BDHU
  4

• Use this library for CNN networks like Unet

  #55 opened 7 months ago by cporrasn
  0

• Support for LLaMA

  #50 opened 2 years ago by IzzetYoung
  3

• Title: RuntimeError: Timed out initializing process group in store based barrier

  #54 opened a year ago by hugocool
  0

• Still in development?

  #53 opened a year ago by codeananda
  0

• RuntimeError: CUDA error: peer access is not supported between these two devices

  #44 opened 2 years ago by Dorcoh4
  1

• freeze_support()

  #22 opened 3 years ago by psinha30
  2

• freeze_support()

  #52 opened a year ago by vinnitu
  0

• Support for Falcon-7B and Falcon-40B models

  #51 opened a year ago by mahdyshabeeb
  0

• Bus error in parallelformers 1.2.7 for OPT model

  #37 opened 2 years ago by sindhuvahinis
  1

• Error using google/UL2 model

  #29 opened 2 years ago by dnhkng
  6

• [Feature Request] Add Bloom to the Auto Policy

  #36 opened 2 years ago by airsplay
  2

• OSError: [Errno 9] Bad file descriptor

  #42 opened 2 years ago by aws-stdun
  1

• Speed up results serialization

  #46 opened 2 years ago by mkardas
  0

• Add Vision Encoder Decoder model to parallelformers

  #45 opened 2 years ago by gagan3012
  0

• Bug with T511b inference

  #43 opened 2 years ago by ZeyiLiao
  0

• A bug with `n_fused`

  #41 opened 2 years ago by JiayiFeng
  4

• torch no_grad

  #40 opened 2 years ago by zelcookie
  1

• INT8 support

  #39 opened 2 years ago by volkerha
  0

• Support Codegen 12B

  #38 opened 2 years ago by Tiiiger
  0

• Can you please add Question Answering models like LayoutLMv2ForQuestionAnswering

  #35 opened 2 years ago by sujit420
  0

• Can you please add support for gpt_neox

  #34 opened 2 years ago by tahercoolguy
  2

• Support for GPT2-XL

  #33 opened 2 years ago by snoop2head
  3

• GPT2 parallelism does not work on the Tesla K80

  #27 opened 2 years ago by 0x7o
  1

• Issue running parallelformers test script in a VM

  #23 opened 3 years ago by Mehrad0711
  1

• How do I use this for zero shot classification tasks

  #12 opened 2 years ago by subhamkhemka
  1

• RuntimeError: Cannot re-initialize CUDA in forked subprocess

  #32 opened 2 years ago by cabal-daniel
  12

• Support for OPT

  #30 opened 2 years ago by mrzjy
  1

• RuntimeError: "LayerNormKernelImpl" not implemented for 'Half'

  #28 opened 2 years ago by samarthsarin
  2

• EncoderDecoder support

  #26 opened 3 years ago by d-miketa
  0

• Recommended way for cleaning up?

  #24 opened 3 years ago by creatorrr
  0

• AttributeError: Can't get attribute 'MegatronPolicy' on <module '__main__' (built-in)>

  #20 opened 3 years ago by Oaklight
  6

• GPU행업 이슈

  #19 opened 3 years ago by jason9693
  4

• GPT models hang on large token generation. Lower performance?

  #15 opened 3 years ago by mallorbc
  1

• 다중 Model 로드 방법

  #18 opened 3 years ago by Don9wanKim
  6

• AssertionError: Model should be on CPU before parallelization. It is more memory-efficient.

  #16 opened 3 years ago by juliensalinas
  29

• KoGPT3와 연동시 품질 이슈

  #17 opened 3 years ago by BangDaeng
  13

• Support for GPT-J

  #4 opened 3 years ago by andreamad8
  11

• Add guides about the number of GPUs to the documentation

  #10 opened 3 years ago by hyunwoongko
  0

• Integration Note with Huggingface Transformers & Microsoft DeepSpeed

  #11 opened 3 years ago by hyunwoongko
  1

• How can I parallelize the MegatronBertModel?

  #14 opened 3 years ago by kajyuuen
  1

• docker support

  #3 opened 3 years ago by hyunwoongko
  2

• Bug about `AlbertModel`

  #5 opened 3 years ago by hyunwoongko
  0