cmedina/agenx-lora-training
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: use 8-bit AdamW optimizer to save VRAM

Browse Source
This commit is contained in:
Christian Medina committed 2026-07-02 17:53:05 -04:00
1 parent 92999b7783
commit 691e5d3611
2 files changed
+5 -3
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

training/configs/ornith-35b-lora.yaml
+1
View File
@@ -42,6 +42,7 @@ train_params:
save_steps: 100
save_total_limit: 3
output_dir: ../../output/ornith-35b-lora
optim: adamw_bnb_8bit # 8-bit optimizer to save VRAM
# Precision
mixed_precision: bf16