List view

• Performant Mixtral 8x7B, 8x22B

  No due date

  •1/3 issues closed
  33% complete2 open1 closed

• Good quality Llama 3.1 405B in torch_xla_models

  Optimize on 1 and 2 pods of TPU v6e

  Overdue by 3 month(s)

  •

  Due by March 1, 2025

  •7/7 issues closed
  100% complete0 open7 closed

• Good quality Llama 3.1 8B and 70B in torch_xla_models

  The goal of this milestone is to ensure we can replace the hard-to-understand Llama reference implementation in https://github.com/pytorch-tpu/transformers/tree/flash_attention. The branch of the Huggingface fork is not ideal for engineering and for showing to interested users.

  Overdue by 4 month(s)

  •

  Due by January 30, 2025

  •6/7 issues closed
  85% complete1 open6 closed