summaryrefslogtreecommitdiff
path: root/test
AgeCommit message (Expand)AuthorFilesLines
2019-04-24fix rocm test (#19663)HEADmasterElias Ellison1-12/+24
2019-04-23Enable assignment for QTensor in pytorch frontend (#19530)Jerry Zhang1-3/+3
2019-04-23Dont introduce aliasing in CSE or Constant Pooling (#19576)eellison3-38/+61
2019-04-23builtin ivalues sort (#19572)Elias Ellison1-0/+37
2019-04-23fix lint (#19632)Jerry Zhang1-0/+1
2019-04-23Add base support to torch.logspace, default base=10 (#19542)Phúc Lê2-0/+9
2019-04-23add torch.tensor requires grad (#19445)Elias Ellison3-5/+67
2019-04-23Fix lack of state init for adagrad and add share_memory flag (#17679)Kaiyu Shi1-0/+89
2019-04-23dispatch max_pools with no indices, expose max_pools to torch namespace (#19449)Wanchao Liang1-2/+2
2019-04-23Adds `fakeQuantizePerTensorAffineOp` to pytorch (#19387)Jerry Zhang1-0/+85
2019-04-23Fix the Division by Zero Bug of CosineAnnealingLR (#19180)Chandler Zuo1-2/+3
2019-04-23Add docs and test guaranteeing indices from torch.nonzero ordered C-style (#1...jhultman1-29/+53
2019-04-23Remove unnecessary printing from testsTongzhou Wang2-9/+18
2019-04-23Fix lr_scheduler's last_epoch value at the time of initialization (BC BREAKIN...Bado Lee1-1/+6
2019-04-23add torch.cuda.synchronize(device=None) (#19573)SsnL1-0/+15
2019-04-22Revert D15039713: [pytorch][PR] add torch.tensor requires gradMichael Suo2-64/+3
2019-04-22add torch.tensor requires grad (#19445)Elias Ellison2-3/+64
2019-04-22Don't create FusionGroups for known-CPU producer values (#19342)James Reed2-8/+13
2019-04-22IRParser: optionally create name->value map of the parsed IR. (#19551)Mikhail Zolotukhin1-1/+10
2019-04-22Add manual_seed in script (#19510)Elias Ellison1-0/+11
2019-04-22Rename potri to cholesky_inverse (#19498)vishwakftw2-13/+15
2019-04-21Generate only one Type class per backend (#19295)Roy Li3-20/+9
2019-04-21Make complex its own backend (#19275)Roy Li1-7/+25
2019-04-21Add ScalarType argument to Type::options() (#19270)Roy Li2-7/+7
2019-04-20Revert D14689639: [pytorch] Allow passing lists as trace inputs.Michael Suo1-42/+0
2019-04-19Make Observer class as template Quant class for QuantConfig (#19418)Nishant Pandit1-135/+124
2019-04-19Improve embedding_bag add kernel (#19329)James Reed1-8/+11
2019-04-19Make finding unused model parameters optional (#19515)Pieter Noordhuis1-0/+63
2019-04-19Add minimalistic implementation of subgraph matcher. (#19322)Mikhail Zolotukhin2-0/+415
2019-04-19First step at container aliasing (#18710)eellison2-0/+214
2019-04-19Allow passing lists as trace inputs.Eric Faust1-0/+42
2019-04-19Make empty_affine_quantized private (#19446)Jerry Zhang1-1/+1
2019-04-19Fix ord() when dealing with utf8 chars (#19423)David Riazati1-0/+7
2019-04-19Fix copied optimizer (#19308)barrh1-0/+5
2019-04-19Remove no-fork workaround for running tests with ROCm (#19436)Junjie Bai2-29/+2
2019-04-19Delete defunct test/ffi directory. (#19168)Edward Yang6-55/+0
2019-04-18Allow passing dicts as trace inputs. (#18092)Eric Faust2-13/+95
2019-04-18skip test_trace_c10_ops if _caffe2 is not built (#19099)Lu Fang1-0/+5
2019-04-18Strip doc_string from exported ONNX models (#18882)Lara Haidar-Ahmad1-0/+24
2019-04-18Step 4: add support for unique with dim=None (#18651)Xiang Gao1-6/+6
2019-04-18allow bools to be used as attributes (#19440)Michael Suo1-1/+2
2019-04-18Fix test build (#19444)David Riazati1-2/+3
2019-04-18pow scalar exponent / base autodiff, fusion (#19324)Thomas Viehmann1-0/+21
2019-04-18Allow optionals arguments from C++ (#19311)David Riazati1-0/+16
2019-04-18Recursively find tensors in DDP module output (#19360)Pieter Noordhuis1-0/+90
2019-04-18Fix pickling torch.float32 (#18045)Ailing Zhang1-0/+7
2019-04-18Respect order of Parameters in rnn.py (#18198)David Riazati1-0/+24
2019-04-18Add device-specific cuFFT plan caches (#19300)Tongzhou Wang1-4/+46
2019-04-18Add constant folding to ONNX graph during export (Resubmission) (#18698)Spandan Tiwari3-10/+182
2019-04-17Allow DDP to wrap multi-GPU modules (#19271)Shen Li1-11/+190