Issues

• Any plans to upgrade with flax.nnx?

  young-geng/EasyLM#118 · Filter by author YinpeiDai opened on Aug 15, 2025
  1

• Too small initializer variance

  young-geng/EasyLM#115 · Filter by author LeoXinhaoLee opened on Aug 19, 2024

• Dataset for the paper: The False Promise of Imitating Proprietary LLMs (https://arxiv.org/pdf/2305.15717)

  young-geng/EasyLM#112 · Filter by author manandey opened on May 14, 2024

• [Bug] Error in Evaluation

  young-geng/EasyLM#111 · Filter by author LeoXinhaoLee opened on Mar 18, 2024

• Llama 7b Pretraining Dtype

  young-geng/EasyLM#110 · Filter by author LeoXinhaoLee opened on Mar 15, 2024

• ERROR: Accessing retired flag 'jax_enable_async_collective_offload'

  young-geng/EasyLM#109 · Filter by author LeoXinhaoLee opened on Mar 15, 2024
  2

• How does accumulate_gradient_steps work?

  young-geng/EasyLM#108 · Filter by author VictorSanh opened on Mar 8, 2024

• Mistral

  young-geng/EasyLM#104 · Filter by author peregilk opened on Nov 23, 2023
  5

• Serving errors: deprecated dependencies and structure error

  young-geng/EasyLM#103 · Filter by author sjw8793 opened on Nov 7, 2023
  2

• why 'LLaMATokenizer' object has no attribute 'sp_model'?

  young-geng/EasyLM#101 · Filter by author zepen opened on Oct 24, 2023
  3

• Conflicting dependencies for jax[cuda11-pip]==0.4.14

  young-geng/EasyLM#100 · Filter by author gbacon opened on Oct 23, 2023

• Does EasyLM work for LLama-2?

  young-geng/EasyLM#99 · Filter by author rafael-wampfler opened on Oct 19, 2023

• OOM trying to pretrain llama 7b on v4-256

  young-geng/EasyLM#98 · Filter by author redbrain opened on Oct 8, 2023
  7

• Missing config for Open LLaMA 3B

  young-geng/EasyLM#97 · Filter by author jcole75 opened on Oct 4, 2023
  1

• What is the full batch size if mesh_dim is set to 1,1,-1, on TPUv3-8?

  young-geng/EasyLM#94 · Filter by author TPFRL opened on Sep 23, 2023
  3

• how to train llama2-7b in a100 80G gpu?

  young-geng/EasyLM#87 · Filter by author brewswang opened on Aug 4, 2023
  3

• Questions about inference mode

  young-geng/EasyLM#86 · Filter by author Tengfei09 opened on Aug 2, 2023
  2

• TPU Installation broken because of change in Orbax

  young-geng/EasyLM#84 · Filter by author akhilkedia opened on Aug 1, 2023
  1

• Feature request: Use Orbax for checkpointing.

  young-geng/EasyLM#80 · Filter by author OhadRubin opened on Jul 25, 2023
  2

• LlaMa Pretraining in A100 80G

  young-geng/EasyLM#79 · Filter by author mohammadaminabbasi opened on Jul 25, 2023
  1

• Use EasyLM to pre-train llama-7B using Nvidia GPU

  young-geng/EasyLM#78 · Filter by author zhpacer opened on Jul 24, 2023
  2

• LLaMA 2 support for pre-training

  young-geng/EasyLM#77 · Filter by author philschmid opened on Jul 19, 2023
  6

• example fine-tune alpaca

  young-geng/EasyLM#70 · Filter by author ehartford opened on Jul 3, 2023
  1

• [Doc error]: Outdated doc for LLAMA

  young-geng/EasyLM#59 · Filter by author llCurious opened on Jun 12, 2023
  1

• For 30B LLama model, can server be supported by configuring mesh_dims on tpu v3-8 (128g)? I tried 8,1 and 4,1 but they don't seem to work.

  young-geng/EasyLM#35 · Filter by author joytianya opened on Apr 19, 2023
  2