Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

fix: torch_dtype mistral default to fp32 #1050

Merged
merged 1 commit into from
Jan 9, 2024
Merged

fix: torch_dtype mistral default to fp32 #1050

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 4 additions & 1 deletion src/axolotl/utils/models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -589,7 +589,10 @@ def load_model(

# LlamaRMSNorm layers are in fp32 after kbit_training or full finetune, so we need to
# convert them back to fp16/bf16 for flash-attn compatibility.
if needs_fa2_dtype or (cfg.flash_attention and cfg.is_llama_derived_model):
if needs_fa2_dtype or (
cfg.flash_attention
and (cfg.is_llama_derived_model or cfg.is_mistral_derived_model)
):
LOG.info("converting modules to %s for flash attention", cfg.torch_dtype)
for name, module in model.named_modules():
if "norm" in name:
Expand Down