Search results

Optimization: Replace shift().fillna(0) with shift(fill_value=0) for cleaner and faster Series manipulation

Status: Open.

#14 In zphang/minimal-llama;

· SaFE-APIOpt opened on Apr 18, 2025

How to correctly load and merge finetuned LLaMA models in different formats?

Status: Open.

#11 In zphang/minimal-llama;

· chenmiaomiao opened on May 12, 2023

this training process did not consider decoder_attention_mask?

Status: Open.

#10 In zphang/minimal-llama;

· zlh1992 opened on Mar 24, 2023

PEFT + PP bug?

Status: Open.

#9 In zphang/minimal-llama;

· hexafluoride opened on Mar 17, 2023

hi

Status: Open.

#8 In zphang/minimal-llama;

· nusZhaoYikai opened on Mar 16, 2023

Bitsandbytes not compatible with Windows!

Status: Open.

#7 In zphang/minimal-llama;

· JustAHippo opened on Mar 15, 2023

Fine-tuning with Naive Pipeline Parallel: NaN after optimizer step

Status: Open.

#4 In zphang/minimal-llama;

· fs4r opened on Mar 12, 2023

one step ,no data.json file

Status: Open.

#3 In zphang/minimal-llama;

· jiangyong888 opened on Mar 11, 2023

RuntimeError: module must have its parameters and buffers on device cuda:0 (device_ids[0]) but found one of them on device: cuda:1 0%|

Status: Open.

#2 In zphang/minimal-llama;

· gadiluna opened on Mar 10, 2023

AttributeError: module transformers.models.llama has no attribute LLaMATokenizer

Status: Open.

#1 In zphang/minimal-llama;

· corranmac opened on Mar 9, 2023