summaryrefslogtreecommitdiff
path: root/caffe2/proto
AgeCommit message (Expand)AuthorFilesLines
2019-02-05Add XLA / TPU device type, backend type and type id (#16763)Alex Şuhan1-1/+2
2019-02-01Introduce backend extensions (overriding operators on custom backends)Roy Li1-1/+2
2019-01-10Move files to/from c10/core and c10/util (#15316)Sebastian Messmer2-2/+2
2018-12-28keep extra_info of each op in ProfDagStats (#15244)Dong Li1-0/+4
2018-12-21Metadata for input/output formats in model file proto. (#15252)Pritam Damania1-1/+3
2018-12-03Restore device when import jit script module (#14454)Lu Fang1-1/+3
2018-11-30Use a zip archive as our container format (#14521)Zachary DeVito1-5/+0
2018-11-29Switch import/export to python printing (#14400)Zachary DeVito1-53/+37
2018-11-20Add tensor table in ModelDef and use it for jit script serialization and dese...Lu Fang1-2/+7
2018-11-19Store the optimize flag in module (#14166)Lu Fang1-0/+4
2018-11-14Use the torch.proto to store script module (#13736)Lu Fang2-1/+9
2018-11-04add explicit fpga context (#13318)Hector Yuen1-2/+3
2018-10-26reduce Device to 32bits (#12767)Roy Li1-1/+1
2018-10-24Back out "Migrate DeviceOption.numa_node_id to DeviceOption.device_id"Junjie Bai2-7/+5
2018-10-24Move Device and DeviceType to c10Edward Yang1-1/+1
2018-10-19Migrate DeviceOption.numa_node_id to DeviceOption.device_idJunjie Bai2-5/+7
2018-10-17remove ATen/Error.h and ATen/core/Error.h (#12792)Yangqing Jia1-1/+1
2018-10-16New serialization format (#12384)Lu Fang2-545/+164
2018-10-10Migrate HIP to use DeviceOption.device_id and delete DeviceOption.hip_gpu_idJunjie Bai2-9/+3
2018-10-10Remove ExtractDeviceOption from StaticContext (#12304)Jerry Zhang1-0/+7
2018-10-09Rename cuda_gpu_id to device_id in DeviceOption (#12456)Junjie Bai2-3/+3
2018-10-02Revert D10098106: [pytorch][PR] [WIP] New version of PT1 model formatDmytro Dzhulgakov2-121/+529
2018-10-01Back out "Revert D10123245: Back out "codemod cuda_gpu_id to device_id"" (#12...Junjie Bai2-3/+45
2018-10-01New version of PT1 model format (#12149)Lu Fang2-529/+121
2018-10-01Revert D10123245: Back out "codemod cuda_gpu_id to device_id"Rick Ratmansky1-1/+1
2018-10-01Back out "codemod cuda_gpu_id to device_id"Yang Liu1-1/+1
2018-10-01Back out "[pytorch][PR] Revert "Move CreateContext to global registry (#11688...Jerry Zhang1-1/+18
2018-09-27codemod cuda_gpu_id to device_id (#12022)Junjie Bai1-1/+1
2018-09-25Revert "Move CreateContext to global registry (#11688)" (#12049)Edward Yang1-31/+1
2018-09-24Move CreateContext to global registry (#11688)Jerry Zhang1-1/+31
2018-09-20codemod: caffe::float16 -> at::Half (#11785)Roy Li2-7/+7
2018-09-19Add the missing header (#11864)Lu Fang1-1/+2
2018-09-19OperatorDef <==> NodeProto Conversion (#11621)Lu Fang2-2/+9
2018-09-11New Serialization ProtoLu Fang1-0/+550
2018-09-07Fix more warnings (#11320)Tongzhou Wang1-6/+0
2018-09-05caffe2::DeviceType -> at::DeviceType (#11254)Jerry Zhang2-10/+94
2018-08-28Replace direct include of caffe2.pb.h with an intermediary header caffe2_pb.h...Edward Yang1-0/+2
2018-08-27Change the default value of DeviceOption.numa_node_id from -1 to 0 (#10877)Junjie Bai1-1/+1
2018-08-16Additional changes to make GPU builds work (#10507)Orion Reblitz-Richardson1-2/+4
2018-08-10Remove caffe1 specific proto (#10380)Yangqing Jia1-0/+15
2018-08-03Introduce at::DeviceType, which subsumes at::Device::Type and (partially) caf...Edward Yang1-1/+2
2018-07-02Adds serialization path (#9035)Yavuz Yetim1-0/+1
2018-06-26Update from Facebook (#8887)Orion Reblitz-Richardson1-2/+31
2018-06-13fix old comment to point to the right file (#8416)Yangqing Jia1-1/+1
2018-06-04[Caffe2] Enabling AMD GPU Backend for Caffe2 (#7955)bddppq1-1/+4
2018-06-01Update from facebook 1ee4edd286a3 (#8040)Bram Wasti1-0/+1
2018-05-23Revert "[Caffe2] Enabling AMD GPU Backend for Caffe2" (#7802)bddppq1-4/+1
2018-05-23[Caffe2] Enabling AMD GPU Backend for Caffe2 (#7566)Peter Yeh1-1/+4
2018-04-22[feature request] [Caffe2] Enable MKLDNN support for inference (#6699)Jinghui1-1/+2
2018-04-20[caffe2][opencl] Add OpenCL context (#6777)Bram Wasti1-1/+2