kyy/llm_trainer
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

improve KTO impl., replace datasets

Browse Source
This commit is contained in:
hiyouga committed 2024-05-18 03:44:56 +08:00
1 parent 33a354548e
commit c450ee87a3
65 files changed
+46415 -2035053
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

examples/extras/llama_pro/llama3_freeze_sft.yaml
+1 -1
View File
@@ -10,7 +10,7 @@ freeze_trainable_modules: all
use_llama_pro: true
### dataset
dataset: identity,alpaca_gpt4_en
dataset: identity,alpaca_en_demo
template: llama3
cutoff_len: 1024
max_samples: 1000