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

Do evaluation with gloo backend, and only on process 0 #6

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Do evaluation with gloo backend, and only on process 0 #6

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
88 changes: 49 additions & 39 deletions ibl/evaluators.py
100644 → 100755
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,8 +2,6 @@
import time
from collections import OrderedDict
import numpy as np
from sklearn.preprocessing import normalize
from sklearn.metrics import pairwise_distances

import torch
import torch.nn.functional as F
Expand Down Expand Up @@ -73,34 +71,50 @@ def extract_features(model, data_loader, dataset, print_freq=10,
if (pca is not None):
del pca

cpu_group = dist.new_group(None, backend="gloo")
if (sync_gather):
# all gather features in parallel
# cost more GPU memory but less time
features = torch.cat(features).cuda(gpu)
all_features = [torch.empty_like(features) for _ in range(world_size)]
dist.all_gather(all_features, features)
del features
all_features = torch.cat(all_features).cpu()[:len(dataset)]
features_dict = OrderedDict()
for fname, output in zip(dataset, all_features):
features_dict[fname[0]] = output
del all_features
# cost more memory but less time
features = torch.cat(features)
if rank == 0:
# gather all features on rank 0
all_features = [torch.empty_like(features) for _ in range(world_size)]
dist.gather(features, all_features, group=cpu_group)
del features
all_features = torch.cat(all_features)[:len(dataset)]
features_dict = OrderedDict()
for fname, output in zip(dataset, all_features):
features_dict[fname[0]] = output
del all_features
else:
# if not rank 0 then just send, dont receive
dist.gather(features, group=cpu_group)
else:
# broadcast features in sequence
# cost more time but less GPU memory
bc_features = torch.cat(features).cuda(gpu)
features_dict = OrderedDict()
for k in range(world_size):
bc_features.data.copy_(torch.cat(features))
if (rank==0):
# cost more time but less memory (maybe?)
features = torch.cat(features)

if rank == 0:
features_dict = OrderedDict()
for k in range(world_size):
print("gathering features from rank no.{}".format(k))
dist.broadcast(bc_features, k)
l = bc_features.cpu().size(0)
for fname, output in zip(dataset[k*l:(k+1)*l], bc_features.cpu()):
features_dict[fname[0]] = output
del bc_features, features

return features_dict
if k != 0:
dist.recv(features, src=k, group=cpu_group)

l = features.size(0)
for fname, output in zip(dataset[k*l:(k+1)*l], features):
features_dict[fname[0]] = output.clone()
else:
# if not rank 0 then send only
dist.send(features, dst=0, group=cpu_group)
del features

if rank == 0:
return features_dict
else:
# return empty dict as we're not duplicating data
return {}

def pairwise_distance(features, query=None, gallery=None, metric=None):
if query is None and gallery is None:
Expand All @@ -113,8 +127,7 @@ def pairwise_distance(features, query=None, gallery=None, metric=None):
dist_m = dist_m.expand(n, n) - 2 * torch.mm(x, x.t())
return dist_m, None, None

if (dist.get_rank()==0):
print ("===> Start calculating pairwise distances")
print ("===> Start calculating pairwise distances")
x = torch.cat([features[f].unsqueeze(0) for f, _, _, _ in query], 0)
y = torch.cat([features[f].unsqueeze(0) for f, _, _, _ in gallery], 0)

Expand Down Expand Up @@ -144,8 +157,7 @@ def evaluate_all(distmat, gt, gallery, recall_topk=[1, 5, 10], nms=False):
del distmat
db_ids = [db[1] for db in gallery]

if (dist.get_rank()==0):
print("===> Start calculating recalls")
print("===> Start calculating recalls")
correct_at_n = np.zeros(len(recall_topk))

for qIx, pred in enumerate(sort_idx):
Expand All @@ -160,13 +172,11 @@ def evaluate_all(distmat, gt, gallery, recall_topk=[1, 5, 10], nms=False):
recalls = correct_at_n / len(gt)
del sort_idx

if (dist.get_rank()==0):
print('Recall Scores:')
for i, k in enumerate(recall_topk):
print(' top-{:<4}{:12.1%}'.format(k, recalls[i]))
print('Recall Scores:')
for i, k in enumerate(recall_topk):
print(' top-{:<4}{:12.1%}'.format(k, recalls[i]))
return recalls


class Evaluator(object):
def __init__(self, model):
super(Evaluator, self).__init__()
Expand All @@ -186,16 +196,16 @@ def evaluate(self, query_loader, dataset, query, gallery, ground_truth, gallery_
features = extract_features(self.model, query_loader, dataset,
vlad=vlad, pca=pca, gpu=gpu, sync_gather=sync_gather)

distmat, _, _ = pairwise_distance(features, query, gallery)
recalls = evaluate_all(distmat, ground_truth, gallery, nms=nms)
if (not rerank):
return recalls

if (self.rank==0):
distmat, _, _ = pairwise_distance(features, query, gallery)
recalls = evaluate_all(distmat, ground_truth, gallery, nms=nms)
if (not rerank):
return recalls

print('Applying re-ranking ...')
distmat_gg, _, _ = pairwise_distance(features, gallery, gallery)
distmat_qq, _, _ = pairwise_distance(features, query, query)
distmat = re_ranking(distmat.numpy(), distmat_qq.numpy(), distmat_gg.numpy(),
k1=rr_topk, k2=1, lambda_value=lambda_value)

return evaluate_all(distmat, ground_truth, gallery, nms=nms)
return evaluate_all(distmat, ground_truth, gallery, nms=nms)