google/paxml

Issues

• How to do gradient accumulation?

  #104 opened 4 months ago by gramesh-amd
  0

• Question about Integration Timeline for PR #99

  #102 opened 5 months ago by Santiago-Castellano
  8

• LoRA support in transformers

  #96 opened 7 months ago by tanmayshishodia
  1

• Error running tutorial notebook

  #95 opened 7 months ago by ajikmr
  0

• lingvo issue while installing paxml in vscode

  #89 opened 8 months ago by juneedpk
  0

• DEADLINE_EXCEEDED on 1024 GPUs.

  #77 opened 9 months ago by mhugues
  0

• Jax + tpu and AQT int8 train model loss is abnormal

  #71 opened 10 months ago by Lisennlp
  0

• Use bfloat16 for eval

  #66 opened a year ago by tbaker2
  1

• [Question] Very low MFU(30%~35%) when train bf16 Llama2 and GPT model with single SXM4 A100 machine.

  #65 opened a year ago by MoFHeka
  0

• [Feature Request] Need ZeRo-1/2 to cooperate with PP+TP+DP. Which may more faster than FSDP sometimes.

  #64 opened a year ago by MoFHeka
  0

• Remaining tutorials

  #58 opened a year ago by rahulbatra85
  0

• How to continue training from a checkpoint?

  #37 opened a year ago by lkm2835
  0

• Int8 checkpoint

  #35 opened 2 years ago by wx-x
  0

• ARM64 Build

  #30 opened 2 years ago by joker-eph
  4

• Installing paxml from source failed due to dependency problem

  #25 opened 2 years ago by yhtang
  5

• Pipeline Parallelism: USE_REPEATED_LAYERS bug

  #5 opened 2 years ago by abhinavgoel95
  3

• Pipeline Parallelism: F external/org_tensorflow/tensorflow/compiler/xla/array.h:446] Check failed: n < sizes_size Fatal Python error: Aborted

  #4 opened 2 years ago by abhinavgoel95
  3

• Error running Common Crawl example

  #11 opened 2 years ago by RobertLiJN
  2

• ERROR: error loading package 'paxml'

  #1 opened 2 years ago by sharathts
  7

• Unexpected Overheads with Activation Checkpointing with Pipeline Parallelism

  #17 opened 2 years ago by abhinavgoel95
  1

• What does USE_REPEATED_LAYER do?

  #3 opened 2 years ago by abhinavgoel95
  1