facebookarchive/fb.resnet.torch issues and pull requests

#100 - Added :cuda() calls to load_model in classify.lua and extract-features.lua

Pull Request - State: closed - Opened by Cubbee about 8 years ago - 3 comments
Labels: CLA Signed

#99 - Fixed model-not-found print statement.

Pull Request - State: closed - Opened by arunmallya about 8 years ago - 3 comments
Labels: CLA Signed

#98 - Use FloatTensor(3, 256, 256):fill(.5) as image in case of reading error

Pull Request - State: open - Opened by zhangtemplar about 8 years ago - 3 comments

#97 - What about concatenating the residual instead of adding?

Issue - State: open - Opened by bernardohenz about 8 years ago - 1 comment

#96 - Sampling batch instances by reverse class frequency in dataset, against imbalanced data

Issue - State: closed - Opened by erogol about 8 years ago - 2 comments

#95 - How to save a model created by nngraph?

Issue - State: closed - Opened by bearpaw about 8 years ago - 2 comments

#94 - Data value

Issue - State: closed - Opened by albicasty about 8 years ago - 2 comments

#93 - Update README.md

Pull Request - State: closed - Opened by tornadomeet over 8 years ago - 3 comments
Labels: CLA Signed

#92 - question on shareGradInput

Issue - State: closed - Opened by superzrx over 8 years ago

#91 - New dataset support

Pull Request - State: closed - Opened by ltrottier over 8 years ago - 1 comment
Labels: CLA Signed

#90 - ResNet training hangs at random iter

Issue - State: closed - Opened by ysoo1133 over 8 years ago - 7 comments

#89 - error rate

Issue - State: closed - Opened by christopher5106 over 8 years ago - 5 comments

#88 - Resuming from checkpoints

Issue - State: closed - Opened by masabdi over 8 years ago - 2 comments

#87 - optnet problem

Issue - State: open - Opened by superzrx over 8 years ago - 19 comments

#86 - Why the software history was not kept?

Issue - State: closed - Opened by gustavopinto over 8 years ago - 2 comments

#85 - Large Memory Consumption When fine-tuning with sharegradInput=true option

Issue - State: closed - Opened by zacurr over 8 years ago - 2 comments

#84 - Memory doubled after the first round

Issue - State: closed - Opened by jonbakerfish over 8 years ago - 3 comments

#83 - Using classify.lua for custom dataset

Issue - State: closed - Opened by dapurv5 over 8 years ago - 2 comments

#82 - No longer shareGradInput after the first epoch if nGPU > 1

Issue - State: closed - Opened by Cysu over 8 years ago - 3 comments

#81 - Request: Adding Half Precision/fp16 statement

Issue - State: open - Opened by mastrcheif over 8 years ago - 2 comments

#80 - Add missed 'jpeg' extension

Pull Request - State: closed - Opened by Bihaqo over 8 years ago - 4 comments

#79 - Predict probabilities for test images

Pull Request - State: closed - Opened by Bihaqo over 8 years ago - 2 comments
Labels: CLA Signed

#78 - Mean and std for a different dataset

Issue - State: closed - Opened by afshindn over 8 years ago - 4 comments

#77 - Different values in Target and Prediction tensor

Issue - State: closed - Opened by jessika0991 over 8 years ago - 4 comments

#76 - Flag to specify folder for checkpoints

Pull Request - State: closed - Opened by Bihaqo over 8 years ago - 2 comments

#75 - The averaged top1, top5 and loss scores are incorrectly computed by m…

Pull Request - State: closed - Opened by ltrottier over 8 years ago - 1 comment

#74 - Fitting of the Model Over multiple GPUs

Issue - State: closed - Opened by tm157 over 8 years ago - 1 comment

#73 - Multi GPU query

Issue - State: closed - Opened by akileshbadrinaaraayanan over 8 years ago - 4 comments

#72 - Normalization params for custom dataset?

Issue - State: closed - Opened by siddhu95 over 8 years ago - 1 comment

#71 - Multiple GPU support Problem

Issue - State: closed - Opened by tm157 over 8 years ago - 3 comments

#70 - replace 'table.getn' with '#'

Pull Request - State: closed - Opened by aychang95 over 8 years ago - 3 comments
Labels: CLA Signed

#69 - getting it to work with Nvidia Digits

Issue - State: closed - Opened by mastrcheif over 8 years ago - 1 comment

#68 - 'getn' deprecated and removed in 'pretrained/extract-features.lua' line 70

Issue - State: closed - Opened by aychang95 over 8 years ago - 1 comment

#67 - Added test and train logger and a opt.save option

Pull Request - State: closed - Opened by spk921 over 8 years ago - 1 comment

#66 - Use optnet and clearState

Pull Request - State: closed - Opened by szagoruyko over 8 years ago - 7 comments
Labels: CLA Signed

#65 - Two enqueue in loop

Issue - State: closed - Opened by lichengunc over 8 years ago - 5 comments

#64 - Add inception-resnet-v2 model

Pull Request - State: open - Opened by lim0606 over 8 years ago - 8 comments
Labels: CLA Signed

#63 - Can't getParameters after adding additional layer in pre-trained resnet 34 model

Issue - State: closed - Opened by spk921 over 8 years ago - 1 comment

#62 - Multi-GPU Hanging With More than 2 GPUs

Issue - State: closed - Opened by stephenrawls over 8 years ago - 5 comments

#61 - Multi-GPU slow down.

Issue - State: open - Opened by shuni1001 over 8 years ago - 21 comments

#60 - How can I run GPU trained network in CPU for testing ?

Issue - State: closed - Opened by erogol over 8 years ago - 2 comments

#59 - Bias variable.

Issue - State: closed - Opened by omkarparkhi over 8 years ago - 2 comments

#58 - Problem while accessing the activation outputs of intermediate layers.

Issue - State: closed - Opened by ajeetksingh over 8 years ago - 4 comments

#57 - What's the meaning of following statements?

Issue - State: closed - Opened by ajeetksingh over 8 years ago - 4 comments

#56 - Problem while initializing the network with same set of weights?

Issue - State: closed - Opened by ajeetksingh over 8 years ago - 6 comments

#55 - How can we stop shuffling in the dataset?

Issue - State: closed - Opened by ajeetksingh over 8 years ago - 1 comment

#54 - Converting pretrained model to nngraph fails

Issue - State: open - Opened by 1adrianb over 8 years ago - 1 comment

#53 - Licence?

Issue - State: closed - Opened by ghost over 8 years ago - 2 comments

#52 - ResNet training hangs at random iter

Issue - State: closed - Opened by Fchaubard over 8 years ago - 3 comments

#51 - Pretrained models download really slow.

Issue - State: closed - Opened by murthypramod over 8 years ago - 2 comments

#50 - Do the extracting in batch mode, useful for extracting many images

Pull Request - State: closed - Opened by iamaaditya over 8 years ago - 3 comments
Labels: CLA Signed

#49 - Inconsistent model description and pre-trained model

Issue - State: open - Opened by ljk628 over 8 years ago - 4 comments

#48 - Fix shareGradInput for pre-activation ResNet

Pull Request - State: closed - Opened by colesbury over 8 years ago - 4 comments
Labels: CLA Signed

#47 - Reproducing resnet-200 results

Issue - State: closed - Opened by dubur over 8 years ago - 4 comments

#46 - Where's the SoftMax?

Issue - State: closed - Opened by roryhr over 8 years ago - 1 comment

#45 - error while extracting features via pretrained resnet-200

Issue - State: closed - Opened by DelightRun over 8 years ago - 3 comments

#44 - threads dojob waiting forever

Issue - State: closed - Opened by ouceduxzk over 8 years ago - 12 comments

#43 - Issues reproducing results when training from scratch

Issue - State: open - Opened by yanii over 8 years ago - 10 comments

#42 - Fix typos and pre-activation types in preresnet

Pull Request - State: closed - Opened by Cysu over 8 years ago - 3 comments
Labels: CLA Signed

#41 - Model definition of preresnet

Issue - State: closed - Opened by Cysu over 8 years ago

#40 - Retrained models really big

Issue - State: closed - Opened by tunguz over 8 years ago - 9 comments

#39 - Indicate GPU_id when training in Multi-gpu mode

Issue - State: closed - Opened by hli2020 over 8 years ago - 5 comments

#38 - 8.5 GB models

Issue - State: closed - Opened by pedropgusmao over 8 years ago - 4 comments

#37 - Minor bugfix for type A shortcut

Pull Request - State: open - Opened by lim0606 over 8 years ago - 5 comments
Labels: CLA Signed

#36 - Sizings, questions, possible error in preresnet

Issue - State: closed - Opened by karpathy over 8 years ago - 5 comments

#35 - cudnn error

Issue - State: closed - Opened by zcyang over 8 years ago - 3 comments

#34 - resume error

Issue - State: closed - Opened by BrettLL over 8 years ago - 9 comments

#33 - The classifier result of ResNet-200

Issue - State: closed - Opened by wlw208dzy over 8 years ago - 6 comments

#32 - added model:evaluate() for feature extraction

Pull Request - State: closed - Opened by ffmpbgrnn over 8 years ago - 1 comment
Labels: CLA Signed

#31 - scale size

Issue - State: closed - Opened by ffmpbgrnn over 8 years ago - 2 comments

#30 - How to implement the random lighting noise in datasets/transforms.lua?

Issue - State: closed - Opened by bryanyzhu over 8 years ago - 2 comments

#29 - how to use augmentation in trainning step

Issue - State: closed - Opened by gjtjx over 8 years ago - 1 comment

#28 - model:evaluate() missing in classify.lua

Issue - State: closed - Opened by nbalakumaran over 8 years ago - 1 comment

#27 - features almost the same

Issue - State: closed - Opened by gjtjx over 8 years ago - 1 comment

#26 - finetune the model for extrating feature

Issue - State: closed - Opened by BrettLL over 8 years ago - 4 comments

#25 - Bounding box for ImageNet data

Issue - State: closed - Opened by nazneenrajani over 8 years ago - 2 comments

#24 - the generated model is in a big size

Issue - State: closed - Opened by BrettLL over 8 years ago - 5 comments

#23 - pre-trained models on CIFAR10

Issue - State: open - Opened by yangky11 over 8 years ago - 3 comments

#22 - Unable to install cudnn bindings on Ubuntu 14.04 LTS

Issue - State: closed - Opened by tunguz over 8 years ago - 1 comment

#21 - Converting models to Caffe.

Issue - State: open - Opened by omkarparkhi over 8 years ago - 11 comments

#20 - 1202 Layer Net

Issue - State: closed - Opened by yueatsprograms over 8 years ago - 3 comments

#19 - Finetuning does not work for binary classification

Issue - State: closed - Opened by arushk1 over 8 years ago - 1 comment

#18 - ResNet 152

Issue - State: closed - Opened by yueatsprograms over 8 years ago - 4 comments

#17 - Training from scratch does not work

Issue - State: closed - Opened by mrastegari over 8 years ago - 21 comments

#16 - Did this project support on cuDNN 3

Issue - State: closed - Opened by samuel1208 over 8 years ago - 2 comments

#15 - File write issue

Issue - State: closed - Opened by arushk1 over 8 years ago - 3 comments

#14 - bad argument #3 to 'narrow'

Issue - State: closed - Opened by arushk1 over 8 years ago - 1 comment

#13 - SpatialBatchNormalization not working

Issue - State: closed - Opened by arushk1 over 8 years ago - 1 comment

#12 - Suggestions on nThreads

Issue - State: closed - Opened by yueatsprograms over 8 years ago - 3 comments

#11 - shareGradInput Problem

Issue - State: closed - Opened by yueatsprograms over 8 years ago - 2 comments

#10 - Your Batch Size?

Issue - State: closed - Opened by yueatsprograms over 8 years ago - 3 comments

#9 - Add pre-trained model classification script

Pull Request - State: closed - Opened by maraoz over 8 years ago - 4 comments
Labels: CLA Signed

#8 - Mismatch in layers' input and output

Issue - State: closed - Opened by AkankshaJ over 8 years ago

#7 - What's good hyper parameter for finetuning on resnet

Issue - State: closed - Opened by fairylrt over 8 years ago - 1 comment

#6 - Add support for checkpointing:

Pull Request - State: closed - Opened by colesbury over 8 years ago

#5 - backend for BN and MaxPooling

Issue - State: closed - Opened by ljk628 over 8 years ago - 2 comments

#4 - Issue with Finetunning

Issue - State: closed - Opened by melgor over 8 years ago - 2 comments

#3 - Feature extraction error

Issue - State: closed - Opened by hiwonjoon over 8 years ago - 4 comments

#2 - multi-gpu work issue

Issue - State: closed - Opened by arjunrajanna almost 9 years ago - 4 comments

#1 - b

Issue - State: closed - Opened by soumith almost 9 years ago

GitHub / facebookarchive/fb.resnet.torch issues and pull requests