summaryrefslogtreecommitdiff
path: root/caffe2/python/examples
AgeCommit message (Expand)AuthorFilesLines
2019-02-13Make timeout in resnet50_trainer configurable (#17058)Junjie Bai1-5/+12
2019-02-09Change the default image size from 227 to 224 in resnet50 trainer (#16924)Junjie Bai1-1/+1
2019-01-22fix typo in resnet50_trainer.pyrohithkrn1-1/+1
2019-01-15Enhance cpu support on gloo based multi-nodes mode. (#11330)Shane Li1-17/+40
2018-12-07Add __init__.py so files get picked up on install (#14898)Orion Reblitz-Richardson1-0/+0
2018-11-29Unify cuda and hip device types in Caffe2 python front end (#14221)rohithkrn2-2/+2
2018-09-12Add resnext model to OSS (#11468)Xiaodong Wang1-17/+74
2018-04-17Update from Facebook (#6692)Orion Reblitz-Richardson1-13/+1
2018-03-27Remove Apache headers from source.Orion Reblitz-Richardson3-45/+0
2018-03-09Fix one more naming issue in resnet50_trainer.py for PR 2205Kutta Srinivasan1-2/+2
2018-03-09Clean up naming of FP16-related code, add commentsKutta Srinivasan1-1/+1
2018-01-24Add support for multi-device batch normalization through an option to data_pa...Lukasz Wesolowski1-0/+1
2018-01-18Checking for positive epoch size before running epochAnirban Roychowdhury1-0/+6
2017-11-29Allow shifting of activations / ops to other GPUs in data parallel modelAapo Kyrola1-2/+16
2017-11-03Fix LoadModel() in resnet50_trainerAapo Kyrola1-0/+10
2017-11-01fix fp16 issues with resnet trainerAapo Kyrola1-7/+15
2017-10-30Added FP16 compute support to FC OpAapo Kyrola1-7/+5
2017-10-24Updated resnet50_trainer and resnet for more FP16 supportAapo Kyrola1-10/+24
2017-10-10Fix ImageInput op in resnet50_trainer.pyLuke Yeager1-2/+4
2017-10-10Add missing is_test argument in ImageInput opsJunjie Bai1-1/+2
2017-09-28Re-license to ApacheYangqing Jia3-0/+45
2017-09-15for CPU-data parallel, allow sharing modelAapo Kyrola1-1/+2
2017-09-15Allow run of example resnet50_trainer without training dataPieter Noordhuis1-21/+53
2017-09-14Resnet50 should param init net before creating test netAapo Kyrola1-4/+3
2017-09-08Create Gloo common world through MPI rendezvousPieter Noordhuis1-5/+23
2017-09-08Add transport/interface arguments to CreateCommonWorld operatorPieter Noordhuis1-0/+11
2017-08-17Add fp16 and tensorcore support to resnet50_trainerLuke Yeager1-10/+44
2017-08-15Important typo in resnet50_trainerLuke Yeager1-1/+1
2017-08-03Add prefix argument to FileStoreHandlerPieter Noordhuis1-0/+1
2017-07-27warn about using test/val model with init_params=True + fixed some casesAapo Kyrola1-1/+1
2017-07-21Populate learning rate blob name into data_parallel_model and fix resnet50_tr...Ahmed Taei1-2/+5
2017-06-23Fixed typoHyungsuk Kang1-1/+1
2017-06-20CPU data parallel modelAapo Kyrola1-6/+13
2017-06-19Deprecate CNNModelHelper in lmdb_create_exampleXiaoti Hu1-4/+3
2017-06-07resnet50: fetch right lrAapo Kyrola1-1/+1
2017-06-07Fixes range/xrange for Python 3Thomas Dudziak2-2/+2
2017-06-0616117716 [Caffe2 OSS] make char-rnn exapmle use build_sgdXiangyu Wang1-13/+9
2017-06-02add_weight_decay + restore weight decay to resnet50_trainerAapo Kyrola1-0/+1
2017-05-30add optimizer support to data_parallel_model; Use MomentumSGDUpdateAapo Kyrola1-34/+10
2017-05-19Refactor data_parallel_model initial sync and checkpointingAapo Kyrola1-15/+3
2017-05-16rnn with brewYiming Wu1-10/+16
2017-05-15new resnet building with brewYiming Wu1-21/+25
2017-05-05Add model saving and loading to resnet50_trainer.pyHeng Wang1-3/+94
2017-05-02Inference code for seq2seq modelYury Zemlyanskiy2-1135/+0
2017-04-25resnet train print loss and accuracyYiming Wu1-0/+5
2017-04-20Fix char_rnn LSTM importJay Mahadeokar1-3/+3
2017-04-19fix a few typosinspire991-2/+2
2017-04-18RNNCell, LSTMCell, LSTMWithAttentionCellYury Zemlyanskiy2-7/+7
2017-04-12Add throughput information to resnet50_trainerPieter Noordhuis1-4/+8
2017-04-12Update resnet50_trainer examplePieter Noordhuis1-47/+54