Skip to content

fix: guard model.half() with dtype check in all rerankers #1574

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

Open
nedeadinside wants to merge 3 commits into
base: master
Choose a base branch
from
+11 −6
Open

fix: guard model.half() with dtype check in all rerankers #1574

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
99d954b
fix: guard model.half() with dtype check in all rerankers
nedeadinside Apr 16, 2026
fc17449
Merge pull request #1 from nedeadinside/fix/reranker-fp16-half-guard
nedeadinside Apr 16, 2026
afe28ad
refactor: replace self.use_fp16 mutation with local fp16 flag
nedeadinside Apr 16, 2026
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
3 changes: 2 additions & 1 deletion FlagEmbedding/inference/reranker/decoder_only/base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -297,7 +297,8 @@ def compute_score_single_gpu(
device = self.target_devices[0]

if device == "cpu": self.use_fp16 = False
if self.use_fp16: self.model.half()
if self.use_fp16 and next(self.model.parameters()).dtype != torch.float16:
self.model.half()

self.model.to(device)
self.model.eval()
Expand Down
3 changes: 2 additions & 1 deletion FlagEmbedding/inference/reranker/decoder_only/layerwise.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -179,7 +179,8 @@ def compute_score_single_gpu(
device = self.target_devices[0]

if device == "cpu": self.use_fp16 = False
if self.use_fp16: self.model.half()
if self.use_fp16 and next(self.model.parameters()).dtype != torch.float16:
self.model.half()

self.model.to(device)
self.model.eval()
Expand Down
3 changes: 2 additions & 1 deletion FlagEmbedding/inference/reranker/decoder_only/lightweight.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -258,7 +258,8 @@ def compute_score_single_gpu(
device = self.target_devices[0]

if device == "cpu": self.use_fp16 = False
if self.use_fp16: self.model.half()
if self.use_fp16 and next(self.model.parameters()).dtype != torch.float16:
self.model.half()

self.model.to(device)
self.model.eval()
Expand Down
5 changes: 3 additions & 2 deletions FlagEmbedding/inference/reranker/encoder_only/base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -110,8 +110,9 @@ def compute_score_single_gpu(
if device is None:
device = self.target_devices[0]

if device == "cpu": self.use_fp16 = False
if self.use_fp16: self.model.half()
use_fp16 = self.use_fp16 and device != "cpu"
if use_fp16 and next(self.model.parameters()).dtype != torch.float16:
self.model.half()

self.model.to(device)
self.model.eval()
Expand Down
3 changes: 2 additions & 1 deletion research/Matroyshka_reranker/inference/rank_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -203,7 +203,8 @@ def compute_score_single_gpu(
device = self.target_devices[0]

if device == "cpu": self.use_fp16 = False
if self.use_fp16: self.model.half()
if self.use_fp16 and next(self.model.parameters()).dtype != torch.float16:
self.model.half()

self.model.to(device)
self.model.eval()
Expand Down