pan-x-c EE-LLM issues - Githubissues

pan-x-c / EE-LLM

EE-LLM is a framework for large-scale training and inference of early-exit (EE) large language models (LLMs).

Other

48 stars 5 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

[ENHANCEMENT] Could u provide the EE-Tuning scripts for llama-2-7b model?

#19 Noblezhong opened 1 month ago
2
[QUESTION] Questions on the performance of EE-LLM on HELM benchmark

#18 Marmot-C opened 4 months ago
11
[ENHANCEMENT] Supporting Beam Search in EE-LLM

#17 Marmot-C closed 4 months ago
2
Update BibTeX for EE-LLM (ICML 2024).

#16 yanxi-chen closed 5 months ago
1
[QUESTION] How can I convert checkpoint tuned by EE-Tuning to Huggingface format?

#15 Mr-lonely0 opened 5 months ago
12
[ENHANCEMENT] Could you provide a script to preprocess data?

#14 Mr-lonely0 opened 5 months ago
2
[BUG] AssertionError: causal mask is only for self attention

#13 Sunt-ing opened 5 months ago
2
[BUG] ValueError: optimizer got an empty parameter list

#12 Mr-lonely0 opened 5 months ago
6
[BUG] Missing fields in `ee_inference_server.sh`

#11 Sunt-ing closed 5 months ago
26
[QUESTION] Run checkpoint on Huggingface Transformers

#10 Sunt-ing closed 5 months ago
4
[QUESTION] The default setting regarding adding an early exit layer.

#9 BenWu11 closed 7 months ago
1
[QUESTION]Using HELM tool for EE-model in Megatron format

#8 BenWu11 closed 7 months ago
1
[QUESTION] Any way to convert ckpt from Megatron-LM into huggingface format

#7 BenWu11 closed 7 months ago
4
[QUESTION] Recreating Figure 9 in the paper

#6 CPA872 opened 7 months ago
4
[BUG] Unable to convert provided checkpoint

#5 CPA872 closed 7 months ago
2
Fix the bug that the checkpoint conversion script cannot support Megatron-LM native checkpoint

#4 pan-x-c closed 7 months ago
0
EE on a Megatron checkpoint

#3 iniverno closed 7 months ago
2
Update README.md

#2 eltociear closed 9 months ago
0
Release EE-Tuning

#1 pan-x-c closed 9 months ago
0