Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / TreB1eN/InsightFace_Pytorch issues and pull requests

#50 - Fine tuning on the model

Issue - State: open - Opened by DecentMakeover over 5 years ago - 2 comments

#49 - Question about training

Issue - State: closed - Opened by Po-Shin over 5 years ago - 1 comment

#49 - Question about training

Issue - State: closed - Opened by Po-Shin over 5 years ago - 1 comment

#47 - Save generated .pth value to Database

Issue - State: open - Opened by chungyau97 over 5 years ago

#47 - Save generated .pth value to Database

Issue - State: open - Opened by chungyau97 over 5 years ago

#46 - margin weight?

Issue - State: open - Opened by nvlong21 over 5 years ago - 4 comments

#46 - margin weight?

Issue - State: open - Opened by nvlong21 over 5 years ago - 4 comments

#45 - Run on CPU?

Issue - State: open - Opened by DecentMakeover almost 6 years ago - 5 comments

#45 - Run on CPU?

Issue - State: open - Opened by DecentMakeover almost 6 years ago - 5 comments

#44 - What version of python ?

Issue - State: open - Opened by DecentMakeover almost 6 years ago - 1 comment

#44 - What version of python ?

Issue - State: open - Opened by DecentMakeover almost 6 years ago - 1 comment

#43 - Maybe a bug in mtcnn.py

Issue - State: open - Opened by y-tai almost 6 years ago - 1 comment

#43 - Maybe a bug in mtcnn.py

Issue - State: open - Opened by y-tai almost 6 years ago - 1 comment

#40 - about test_transform params for model_ir_se50.pth

Issue - State: open - Opened by windspire almost 6 years ago - 1 comment

#40 - about test_transform params for model_ir_se50.pth

Issue - State: open - Opened by windspire almost 6 years ago - 1 comment

#39 - pre-trained head .pth?

Issue - State: open - Opened by thnkim almost 6 years ago

#39 - pre-trained head .pth?

Issue - State: open - Opened by thnkim almost 6 years ago

#38 - code seems do not support resume training from a saved weight?

Issue - State: open - Opened by qq184861643 almost 6 years ago - 9 comments

#38 - code seems do not support resume training from a saved weight?

Issue - State: open - Opened by qq184861643 almost 6 years ago - 9 comments

#37 - maybe few differences with mxnet

Issue - State: open - Opened by ruiming46zrm almost 6 years ago - 7 comments

#37 - maybe few differences with mxnet

Issue - State: open - Opened by ruiming46zrm almost 6 years ago - 7 comments

#36 - Do you know where is the source of vgg2_fp?

Issue - State: open - Opened by twmht almost 6 years ago - 1 comment

#36 - Do you know where is the source of vgg2_fp?

Issue - State: open - Opened by twmht almost 6 years ago - 1 comment

#35 - Disordered RGB Channels after Running "python prepare_data.py" (fixed).

Issue - State: closed - Opened by ZhaoJ9014 almost 6 years ago - 1 comment

#35 - Disordered RGB Channels after Running "python prepare_data.py" (fixed).

Issue - State: closed - Opened by ZhaoJ9014 almost 6 years ago - 1 comment

#34 - how to load pre-trained model and continue training?

Issue - State: closed - Opened by Royzon almost 6 years ago - 3 comments

#34 - how to load pre-trained model and continue training?

Issue - State: closed - Opened by Royzon almost 6 years ago - 3 comments

#33 - Pre-trained Model performs poorly on MegaFace challenge-1

Issue - State: closed - Opened by qq184861643 almost 6 years ago

#33 - Pre-trained Model performs poorly on MegaFace challenge-1

Issue - State: closed - Opened by qq184861643 almost 6 years ago

#32 - how to train on multi-GPUs

Issue - State: closed - Opened by Royzon almost 6 years ago - 2 comments

#32 - how to train on multi-GPUs

Issue - State: closed - Opened by Royzon almost 6 years ago - 2 comments

#31 - which is model folder ?

Issue - State: closed - Opened by Yaoxingtian almost 6 years ago - 1 comment

#31 - which is model folder ?

Issue - State: closed - Opened by Yaoxingtian almost 6 years ago - 1 comment

#30 - What dataset was used for training IR-SE50 model?

Issue - State: closed - Opened by Archelunch almost 6 years ago - 4 comments

#30 - What dataset was used for training IR-SE50 model?

Issue - State: closed - Opened by Archelunch almost 6 years ago - 4 comments

#29 - Did you reproduce megaface ?

Issue - State: open - Opened by twmht almost 6 years ago - 27 comments

#28 - is bgr format image necessary in training and validating? thankyou

Issue - State: closed - Opened by ruiming46zrm almost 6 years ago - 1 comment

#28 - is bgr format image necessary in training and validating? thankyou

Issue - State: closed - Opened by ruiming46zrm almost 6 years ago - 1 comment

#27 - Do huge batch size necessary?

Issue - State: open - Opened by Jasonbaby almost 6 years ago - 8 comments

#26 - No such file or directory: 'work_space/save/model_final.pth'

Issue - State: closed - Opened by while-true-work almost 6 years ago - 6 comments

#25 - I get the acc 56% by 5 epoch which make me saaaaaaaaaaad

Issue - State: closed - Opened by ruiming46zrm almost 6 years ago - 7 comments

#24 - No such file or directory: 'data/facebank/facebank.pth'

Issue - State: open - Opened by YueXiNPU almost 6 years ago - 13 comments

#23 - Why is the verification using l2 distance?

Issue - State: open - Opened by timost1234 almost 6 years ago - 1 comment

#22 - why do you divide CE loss by batch_size?

Issue - State: open - Opened by thnkim almost 6 years ago - 1 comment

#21 - why trans.Normalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5])

Issue - State: open - Opened by tengshaofeng almost 6 years ago - 3 comments

#20 - No such file or directory: 'data/facebank/names.npy

Issue - State: closed - Opened by zhanglaplace almost 6 years ago - 3 comments

#19 - Have you ever transferred your pytorch model into caffemodel?

Issue - State: closed - Opened by hustzeyu almost 6 years ago - 1 comment

#18 - Why pre-trained model has lower performance than that from MXNET?

Issue - State: open - Opened by tengerye almost 6 years ago - 3 comments

#17 - 内存问题

Issue - State: closed - Opened by shizhenjun about 6 years ago - 1 comment

#16 - Download link about mentioned 'emore' dataset in README.md

Issue - State: open - Opened by ShiinaMitsuki about 6 years ago - 6 comments

#15 - Train on Own Dataset

Issue - State: open - Opened by leacornelio about 6 years ago - 3 comments

#14 - Maybe a bug in pytorch 1.0 in computing the cos_theta in model.py

Issue - State: closed - Opened by tiandunx about 6 years ago - 3 comments

#13 - Some errors about image color

Issue - State: closed - Opened by lvvvmd about 6 years ago - 8 comments

#12 - training paras for mobilefacenet

Issue - State: closed - Opened by cvtower about 6 years ago - 24 comments

#11 - NameError: name 'module' is not defined

Issue - State: closed - Opened by qianchen94 about 6 years ago - 2 comments

#10 - NameError: name 'module' is not defined

Issue - State: closed - Opened by qianchen94 about 6 years ago

#9 - NameError: name 'module' is not defined

Issue - State: closed - Opened by qianchen94 about 6 years ago

#8 - Do you need image alignment during training or evalution?

Issue - State: closed - Opened by x7177214 about 6 years ago - 7 comments

#7 - the reference to mobileFaceNet

Issue - State: closed - Opened by twmht about 6 years ago

#6 - Maybe a bug

Issue - State: closed - Opened by JingyuanHu about 6 years ago - 14 comments

#5 - GPU memory required to train r100

Issue - State: closed - Opened by JingyuanHu about 6 years ago - 5 comments

#4 - Environment setting problem(POSIXPATH error)

Issue - State: closed - Opened by ming1144 about 6 years ago - 4 comments

#3 - MobileFaceNet is not right

Issue - State: closed - Opened by tensorboy about 6 years ago - 1 comment

#2 - Learning rate

Issue - State: closed - Opened by tensorboy over 6 years ago - 5 comments

#1 - Training

Issue - State: closed - Opened by tensorboy over 6 years ago - 5 comments