microsoft/mttl

Issues

• Train experts on a custom dataset

  #153 opened 2 months ago by vietanh125
  3

• Add Instructions for Running Polytropon in README

  #157 opened 2 months ago by kadarsh22
  0

• The question of reproducing the effect of mhr on the ni dataset

  #155 opened 3 months ago by TJL233
  2

• experts based on Phi 3 model with PEFT library

  #124 opened 5 months ago by AliEdalat
  18

• Issue When Trying to Replicate the Result

  #142 opened 4 months ago by thuuyen98
  5

• HardPromptConfig takes tokenizer which is a complex arg

  #89 opened 5 months ago by sordonia
  0

• The cli_dataset_create.py file was not found

  #119 opened 5 months ago by GuanghuiSong
  6

• On the details of Arrow Routing

  #107 opened 5 months ago by Weifan1226
  11

• Rename CoalescedLoRAExpertContainer to LoRAExpertContainer, e.g. remove LoraExpertContainer

  #96 opened 5 months ago by sordonia
  0

• Can we remove "modify_layers" argument? "Modify_modules" can be sufficient?

  #97 opened 7 months ago by oleksost
  1

• Packed batch

  #98 opened 7 months ago by sordonia
  5

• May I get the trained LoRA lib to reproduce the paper:Modular LLM?

  #72 opened 8 months ago by Weifan1226
  3

• FLAN

  #42 opened 9 months ago by sordonia
  0

• Create scripts for MHR

  #37 opened 9 months ago by sordonia
  0

• Issues when finetuning pretrained lora checkpoint for downstream tasks.

  #48 opened 9 months ago by prateeky2806
  1

• Documentation

  #53 opened 9 months ago by harsha-moparthy
  2

• Inquiry about Multi-GPU Training Support

  #43 opened 2 years ago by rongaoli
  0

• Action required: self-attest your goal for this repository

  #26 opened 2 years ago by microsoft-github-policy-service
  3