Skip to content

falcon h1 force float32 when dtype is torch.float16 #3026

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

Merged
danielhanchen merged 1 commit into from
Jul 22, 2025
Merged

falcon h1 force float32 when dtype is torch.float16 #3026

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: 5 additions & 0 deletions unsloth/models/loader.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -584,6 +584,11 @@ def from_pretrained(

if transformers_version < Version("4.53.0"):
raise RuntimeError("Unsloth: Gemma 3N only works on transformers >= 4.53.0" + LATEST)
elif "falcon-h1" in lowered_model_name:
os.environ["UNSLOTH_FORCE_CUSTOM_DTYPE"] = \
"float16;torch.float32;torch.float16;"\
"if name.endswith(('q_proj', 'k_proj', 'v_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj', 'head')): module.to(torch.float16);"
os.environ["TRITON_F32_DEFAULT"] = "ieee"
Comment thread
danielhanchen marked this conversation as resolved.
else:
for check_model_name in DISABLE_COMPILE_MODEL_NAMES:
if check_model_name in lowered_model_name:
Expand Down