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

PyTorch Hub amp.autocast() inference #2641

Merged
merged 1 commit into from
Mar 28, 2021
Merged

PyTorch Hub amp.autocast() inference #2641

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
PyTorch Hub amp.autocast() inference 
I think this should help speed up CUDA inference, as currently models may be running in FP32 inference mode on CUDA devices unnecesarily.
  • Loading branch information
@glenn-jocher
glenn-jocher committed Mar 28, 2021
commit ce34cba88efe399bef964abe65dc83751cd8283b
17 changes: 9 additions & 8 deletions models/common.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
import torch
import torch.nn as nn
from PIL import Image
from torch.cuda import amp

from utils.datasets import letterbox
from utils.general import non_max_suppression, make_divisible, scale_coords, xyxy2xywh
Expand Down Expand Up @@ -219,17 +220,17 @@ def forward(self, imgs, size=640, augment=False, profile=False):
x = torch.from_numpy(x).to(p.device).type_as(p) / 255. # uint8 to fp16/32
t.append(time_synchronized())

# Inference
with torch.no_grad():
with torch.no_grad(), amp.autocast(enabled=p.device.type != 'cpu'):
# Inference
y = self.model(x, augment, profile)[0] # forward
t.append(time_synchronized())
t.append(time_synchronized())

# Post-process
y = non_max_suppression(y, conf_thres=self.conf, iou_thres=self.iou, classes=self.classes) # NMS
for i in range(n):
scale_coords(shape1, y[i][:, :4], shape0[i])
t.append(time_synchronized())
# Post-process
y = non_max_suppression(y, conf_thres=self.conf, iou_thres=self.iou, classes=self.classes) # NMS
for i in range(n):
scale_coords(shape1, y[i][:, :4], shape0[i])

t.append(time_synchronized())
return Detections(imgs, y, files, t, self.names, x.shape)


Expand Down