- 01 11月, 2021 1 次提交
-
-
由 Zhanghuihong Guan 提交于
* initial commit, add code for async construct tensor from numpy array * inital commit to change Maybe to Optional * delete redundant code * replace Maybe with Optional * fix compile errors * format code * changes based on review * format code, fix based on review * format code * fix multiclient type * changes based on review * changes based on review * unify calling to IsMultiClirnt * refector multi_client related code * restore InMultiClient interface * double check for unnecessary changes * remove unnecessary changes * format code * Update oneflow/api/python/symbol/job_conf_symbol.cpp * Update oneflow/api/python/symbol/op_conf_symbol.cpp * Update oneflow/api/python/symbol/op_node_signature_symbol.cpp * Update oneflow/core/common/optional.h * Update oneflow/api/python/symbol/string_symbol.cpp * Update oneflow/api/python/symbol/scope_symbol.cpp * Update oneflow/api/python/symbol/placement_symbol.cpp * Update oneflow/api/python/symbol/op_conf_symbol.cpp Co-authored-by: NHoujiang Chen <chenhoujiangcug@gmail.com> Co-authored-by: NTwice <i@twice.moe>
-
- 18 10月, 2021 1 次提交
-
-
由 guo ran 提交于
-
- 15 10月, 2021 1 次提交
-
-
由 guo ran 提交于
* refactor slice boxing * refine * slice boxing node use compute stream * refine * refine * refine * refine
-
- 23 9月, 2021 2 次提交
-
-
由 leaves-zwx 提交于
* LazyBuildTensor * fix slice get sbp * LazyBuildTensor -> BuildTensor * fix adam sbp * fix hierarchy condition * fix hierarchy parallel conf * sparse_softmax_cross_entropy_ms functor * debug code * rm debug code * s to b Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
由 Juncheng 提交于
Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 11 9月, 2021 1 次提交
-
-
由 cheng cheng 提交于
* Fix bug of Multi-Client src tick output order * Add input/output ctrl edge to DstSubTick for order io and callback_notify * add test scripts * remove note * auto format by CI * add note of sleep * auto format by CI Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org>
-
- 08 9月, 2021 1 次提交
-
-
由 Juncheng 提交于
Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 06 9月, 2021 1 次提交
-
-
由 Juncheng 提交于
* Remove IDMgr::GetGpuPhyIdFromThrdId/IDMgr::GetDeviceTypeFromThrdId * CHECK(new_task_id_) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 31 8月, 2021 1 次提交
-
-
由 Twice 提交于
* for all: use hash combine * for all: add Hash(T...) * util: clang format * rename Hash(size_T*, T...) to AddHash * clang format * apply Co-authored-by: NJuncheng <liujuncheng1022@gmail.com> * clang format * clang format * fix Co-authored-by: NShenghang Tsai <jackalcooper@gmail.com> Co-authored-by: NJuncheng <liujuncheng1022@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 26 8月, 2021 1 次提交
-
-
由 Twice 提交于
* functional_api: fix build error in mac os * functional_api: fix interpreter_test * for all: replace device_type.pb.h with device_type.h * value_types: impl hash * revert tools change * CI: mac only * cmake: fix of_functional_obj * cmake: fix of_functional_obj * CI: revert Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 19 8月, 2021 2 次提交
-
-
由 Juncheng 提交于
Co-authored-by: Nleaves-zwx <kunta0932@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
由 Juncheng 提交于
Co-authored-by: Nleaves-zwx <kunta0932@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 17 8月, 2021 1 次提交
-
-
由 Juncheng 提交于
* Remove GlobalWorkStreamId/GlobalThrdId * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 15 8月, 2021 1 次提交
-
-
由 Tianyu Zhao 提交于
* Rename `ParallelDistribution` to `NdSbp` * Rename `ParallelDistribution` to `NdSbp` * Rename `ParallelDistribution` to `NdSbp` * auto format by CI * Rename `ParallelDistribution` to `NdSbp` Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 12 8月, 2021 1 次提交
-
-
由 Tianyu Zhao 提交于
* Rename `parallel_distribution` to `nd_sbp` * Rename filenames containing `parallel_distribution` * auto format by CI * Rename `parallel_distribution` to `nd_sbp` * auto format by CI * Rename `parallel_distribution` to `nd_sbp` * auto format by CI Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 07 8月, 2021 1 次提交
-
-
由 Yinggang Wang 提交于
* fix bugs in shareing EagerBlobObject::blob_desc_.shape and EagerBlobObject::blob_.shape * feat(EmptyOp): add flow.empty * docs(EmptyOp): add doctest and refine document * docs(EmptyOp): refine document * refactor(Tensor): Tensor constructor use empty_op * refactor(Tensor): remove useless code * feat(EmptyOp): support construct in given device and add consistent_empty op * feat(EmptyOp): support unpacked tuple shape * refine array functor code * docs(EmptyOp): update empty op document * refine code * docs(EmptyOp): add test and document for consistent empty op * update document * fix merge bugs * fix(*): fix infer distribution * test(EmptyOp): fix ConsistentEmptyOp CPU_ONLY test bug * fix(*): init shape when InitBlob * fix(*): Constant and Empty Op use broadcast sbp * fix(indexing): replace MakeTensor with functional::Empty * fix(*): fix compile bug * refine code * fix(nnGraph): make eager tensor * auto format by CI * fix(Stride): infer stride before initializing shape Co-authored-by: NXinqi Li <lixinqi0703106@163.com> Co-authored-by: NLi Xinqi <lixinqi2010@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org>
-
- 06 8月, 2021 1 次提交
-
-
由 Juncheng 提交于
-
- 02 8月, 2021 1 次提交
-
-
由 Luyang 提交于
* 0-dim tensor support * test case * add more test * refine * update * update default constructor * reconstuct * merge master * remove notes * remove useless codes * fix comments * fix comment * add test case * format * refine * refine * refine * refine * MirroredTensorMeta::MirroredTensorMeta() * support 0-dim slice * support 0-dim slice grad * refine * auto format by CI * refine * refine * auto format by CI * refine * fix slice bug * auto format by CI * fix resnet50 0-im loss uasge * fix 0-dim tensor usage in test cases * add skip test * auto format by CI * fix test_dataset * check blobdesc.shape init * auto format by CI * remove useless empty shape init * fix l1loss 0-dim error * auto format by CI * fix argmax op test * fix add_n op test * auto format by CI * fix bce loss op test * auto format by CI * fix squeeze op test * fix conv2d op test * fix xpu_shape for clip_grad_norm * auto format by CI * resolve confilct * fix multi-cpu slice_copier 0-dim bug * auto format by CI * add memory copy for 0-dim * auto format by CI * support copy0dim * refine * auto format by CI * remove unuse codes * fix check for kldivloss * gpu 0-dim copy * auto format by CI * fix clip_grad_norm doctest * fix reduce_ops doctest * fix argmax doctest * fix loss module doctests * fix math_ops doctests * fix norm modules doctest Co-authored-by: NXinqi Li <lixinqi0703106@163.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 16 7月, 2021 1 次提交
-
-
由 Li Xinqi 提交于
* refactor job_pass by maybe_system * remove useless files Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 14 7月, 2021 1 次提交
-
-
由 liufengwei0103 提交于
* Modified the OutputArgModifyFn interface * maybe error stack from CheckAndConstructOp to OutputArgModifier callback function * maybe error stack from CheckAndConstructOp to OutputArgModifier callback function * OutputArgModifier return maybe part_1 * maybe error stack from CheckAndConstructOp to OutputArgModifier callback function * add JUST for hander in ForEachOperator Co-authored-by:
aishangjj <702572275@qq.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 01 7月, 2021 2 次提交
-
-
由 Li Xinqi 提交于
* Device::compute_dep_object_ * sequantialize instructions in the same stream. * refactor AttrMap * refactor Tensor * Export ConsistentTensor::is_cuda * remove ConsistentTensor::blob_object * refactor TensorImpl * minor fix * fix compiler' complains * Implements EagerConsistentTensorImpl::New * minor fix * fix compiler complains * remove unused code * skip test_creating_consistent_tensor * backup code * Symbol::shared_from_symbol * remove redundant header file includes * fix bug in Symbol::shared_from_symbol * symbolize ParallelDesc and ParallelDistribution * symbolize Scope::GetParallelDesc() * IsScalarType * fix compiler complains * InputConsistentTensorMeta * refactor Scope with PlacementScope * fix bug in exporting Scope to python * backup code * refactor DType * fix compiler complains * backup code * DType is only allowed to be used in python code * backup code * dtype api bugfix * fix error on exiting Signed-off-by: Ndaquexian <daquexian566@gmail.com> * lazily get rank Signed-off-by: Ndaquexian <daquexian566@gmail.com> * Export const DType* into python * minor fix * fix bug * refine * refactor signature of OpExpr::InferLogicalShapeAndDtype * fix bug * backup_code * fix bug * refactor SbpXXX to cfg::SbpXXX * merge refactor_sbp_to_cfg_sbp * fix bug * Infer ConsistentTensorMeta * Implement EagerConsistentInterpret::ApplyImpl * 1) move XXXTensorMeta into the new file tensor_meta.h; 2) add new Class ConsistentTensorInferCache * add class ConsistentTensorInferResult * remove unused OpArgMutConsistentTensorMeta::parallel_distribution_ * fix stack-overflow bug in Tensor::mut_eager_mirrored_tensor_impl * ignore empty parallel distribution constaint * fix bug * add explicit of cfg * fix xla compile bug * auto format by CI * fix according comment * fix bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by:
clackhan <han_binbin@163.com> Co-authored-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: NShenghang Tsai <jackalcooper@gmail.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org>
-
由 daquexian 提交于
* add missing JUST Signed-off-by: Ndaquexian <daquexian566@gmail.com> * remove redundant header Signed-off-by: Ndaquexian <daquexian566@gmail.com> * add missing JUST in master Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix compile error on gcc5 Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 30 6月, 2021 1 次提交
-
-
由 Juncheng 提交于
Co-authored-by: Nguo ran <360112263@qq.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 21 6月, 2021 1 次提交
-
-
由 leaves-zwx 提交于
* MemZoneId Former-commit-id: 7550a129f15554c5a6e480b728079e431c00be25 * move mem zone id source code Former-commit-id: 3859fc2a0fcda2fb23e57e886a0e3f1c0833d111 * revert Former-commit-id: 5cf3ad7caebe787918d1ca1c0467415656d9b491 * refine GetProxyNode using MemZoneId Former-commit-id: fba035f20b44b1acce2900b86b5bd24654e0d982 * refactor MemZoneId121 Former-commit-id: 0868a6139f1cf20dc7474d0a88714e03721c8e8e * replace using IDMgr interface Former-commit-id: 98b5db9ed879cd1d8197efd174c6d680bec69560 * fix linkage * rm useless comment * replace IsGpuMemZone * format * rm deprecated mem zone api in IDMgr * fix merge conflict error * refine mem zone id to include node index * revert added header * direct init device_id * address review * address review Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 15 6月, 2021 1 次提交
-
-
由 liufengwei0103 提交于
* refactor SbpXXX to cfg::SbpXXX * modify ParallelDistributionHint4InputArgNameAndIndex to be const function * fix sbp to cfg::sbp in job_pass * fix bug ToProto, InitFromProto and pb passed to cfg * auto format by CI * fix gpt segment fault * fix xla * tmp commit * tmp commit * fix xla compile error * [fix bug] return tmp in model_io_v2 * auto format by CI Co-authored-by: Nlixinqi <lixinqi0703106@163.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: NpoohRui <yuruil@qq.com>
-
- 03 6月, 2021 2 次提交
-
-
由 Shenghang Tsai 提交于
* Remove redundant copyright header * ci check if license duplicated * refine * refine * refine * address review Co-authored-by: Nliujuncheng <liujuncheng1022@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
由 Hongsheng Wang 提交于
* Add scalar support of greater less module (#4841) * add scalar input support * format Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9b11938c * Refine optimizer (#4840) * refactor(Optim): refine optimizer codes * docs(SGD): add document for SGD * docs(SGD): fix code * test(Adam): fix test_optim_adam bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: cd6ffac6 * add docstring (#4846) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: da82bb8c * fix eager with unknow symbol id (#4752) * fix eager with unknow symbol id * minor fix * fix conflict * remove unnnecessary function * remove unnecessary header * remove unnecessary methods Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: aa9f6f76 * Dev fix linear module (#4836) * add broadcast matmul support * refine * add batch matmul support * remove redundant test case * linear module support high dimension input * format * fix linear Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f1ccf2a3 * Add rmsprop optimizer (#4834) * add rmsprop optimizer * fix rmsprop optimizer bug * fix rmsprop optimizer bug * add rmsprop optimizer docs * add rmsprop docs * fix comment * fix comment * fix comment * fix comment Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f476d48d * add adamw optimizer (#4824) * init adamw optimizer * fix adamw optimizer bug * fix comment * fix comment * code format * fix comment Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f15a8aea * add experimental apis (#4817) * add experimental apis * merge master fix conflict * revert flow._oneflow_internal.dtype to flow.dtype * refine * fix test optimizer * update module docs * fix unit tests * fix matmul module test * fix adamw and rmsprop tests * fix crossentropy loss grad Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c50ff3cc * remove MakeParallelDescByDevice, fix the missing setting of parallel_desc in InstructionMsg copy constructor (#4850) Signed-off-by: Ndaquexian <daquexian566@gmail.com> Former-commit-id: d5d7ef56 * add experimental (#4856) Former-commit-id: 505d4865 * A more efficient implementation of NLL Loss (#4854) * A more efficient implementation of NLL Loss * A more efficient implementation of NLL Loss Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: ad5b0129 * Add where module (#4845) * add broadcast_like module * add where module, still has bug * fix where module bug * fix where module bug * fix bug and add where module * fix where module commnet * code format * fix where module * code format Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: edf9a190 * dev_compare_cfg_file (#4860) * dev_compare_cfg_file * add def of org_content * minor fix Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 5e399d38 * stateful local opkernel: return a temp parallel ctx (#4857) * add temp parallel ctx for single card Signed-off-by: Ndaquexian <daquexian566@gmail.com> * add TODO comment Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f450ff89 * Optimize memory occupancy for interface 1.0 (#4844) * Do not save inputs in function nodes even if requires_grad is true. * Allocate raw memory with actual size. Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: e5dadaf5 * use less event records (#4861) * use less event records * more comments Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 218f61e3 * Refine cast grad func. (#4853) Former-commit-id: dfe7759e * copy eager blob object to/from numpy in c++, use busy loop to wait (#4839) * numpy: create np arr in python and copy in c++, use busy loop to wait Signed-off-by: Ndaquexian <daquexian566@gmail.com> * reformat Signed-off-by: Ndaquexian <daquexian566@gmail.com> * add CopyBetweenMirroredTensorAndNumpy Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 87c0c3d8 * align squeeze module with torch (#4855) * align squeeze module with torch * fix comment * fix argmax bug * fix bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: a57b7649 * Fix adam weight decay param (#4835) * fix adam weight decay * fix adam weight decay * fix comment * fix comment * fix commnet * fix commnet * fix commnet * fix bug * fix(Adam): fix Adam test bug * revert adam test threshold to 1e-3 * fix(Adam): fix adam test bug and adjust param to increase error Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Nwyg1997 <wyg19970408@gmail.com> Former-commit-id: 1d5f743d * Fix groupnorm (#4848) * fix GroupNorm and modify test case * add grad op for reshape_like op Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9bf14b8b * use composed attr map in contexts (#4838) * use composed attr map Signed-off-by: Ndaquexian <daquexian566@gmail.com> * move implementation to .cpp Signed-off-by: Ndaquexian <daquexian566@gmail.com> * OpExpr::New returns Maybe Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9d78fdba * flow.Size support negative index and add test (#4870) * feat(PySize): support negative index and add test * style(*): refine code * format code Former-commit-id: 3519d2e7 * fix export experimental docs bug (#4867) * fix export experimental docs bug * fix export experimental docs bug * fix export experimental docs bug Co-authored-by: NYao Chi <later@usopp.net> Former-commit-id: d78c2bc6 * reorder VirtualMachine fields (#4873) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 1ce9f262 * Align module params with torch (#4865) * align mean module * allow negative dim param * support tuple of negative dim param * refine * format Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 373cefce * Generate cfg header and source files in parallel and prevent rebuild from scratch when Python version changes (#4876) * refine * Update cfg.cmake * Update cfg.cmake Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 72204213 * fix interpreter determin output leaf and grad (#4872) * fix interpreter determin output leaf and grad * fix GradMode get * simplify * add test for no_grad Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: de05e7a5 * support crossentropy loss 3dim (#4875) * support crossentropy loss 3dim * merge conflict Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4e689d25 * supoort nllloss 3dim (#4874) * supoort nllloss 3dim * supoort nllloss 3dim * merge conflict Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 30a75cee * Device compute dep object (#4862) * Device::compute_dep_object_ * sequantialize instructions in the same stream. * adjust atexit sort Co-authored-by: NHoujiang Chen <chenhoujiangcug@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by:
clackhan <han_binbin@163.com> Former-commit-id: 55a223cf * remove cambricon quantization test (#4879) Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9cddc629 * Use dlopen to call ibverbs APIs (#4852) * check in naive struct * refine * refine * refine * refine * add functions * refine * refine * refine * fmt * refine * refine * refine * refine * refine * refine * add note * refine * refine * refine * refine * refine * refine * refine * refine * rm include * revert cmakelist changes * refine * address review * rename * address review * address review * remove glog dependency * fix * refine * refine * print lib path in stdout * address review * address review * fix * support ONEFLOW_LIBIBVERBS_PATH * add case * update init_cluster_env.py for ONEFLOW_LIBIBVERBS_PATH * fix comment * address review Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 75f11b82 * Add copy user op (#4842) * copy user op * add to module and tensor.to interface * remove unnecessary code * backward for tensor.to * remove capture of input * support cpu only tensor * module to (#4858) * remove backward kernel and op * friendly deal with when tensor.grad is None * minor fix * minor fix * revert * suport 1m1d only * skip test normalization * skip test normalization * skip conv * support construct device using string * minor fix * minor fix * use maybe * fix device id type for device infer ctx * skip batchnorm * skip some tensor test case Co-authored-by: NXiaoyu Xu <xiaoyulink@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 2d5fae50 * Fix reduce sum grad func. (#4882) * Fix reduce sum grad func. * Fix zeros op Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: a4022285 * align dim size funtion (#4880) * align dim size funtion * fix dim usage Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4592d467 * support expand and repeat op int datatype (#4883) * support expand and repeat op int datatype * code format Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 2939e8cb * return LocalTensor (TensorTuple) directly from op expr __call__ (#4864) * expose local tensor Signed-off-by: Ndaquexian <daquexian566@gmail.com> * mt19937 -> minstd_rand Signed-off-by: Ndaquexian <daquexian566@gmail.com> * revert unnecessary diff Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix comments Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 72b76a29 * Support custom parameters for optimizer (#4881) * feat(Optim): support custom parameters for optimizer * feat(Adam): adam support custom parameters * feat(Adamw): adamw support custom parameters * feat(RMSprop): rmsprop support custom parameters * style(Optim): refine adam and adamw Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Former-commit-id: a26f7080 * align transpose module with pytorch (#4877) * align transpose module with pytorch * fix comment * align tranpose module * support expand and repeat op int datatype * fix bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 2a5bc359 * Add ones like op (#4889) * Add ones like op. Conflicts: oneflow/api/foreign_lock_helper.h oneflow/api/python/autograd/autograd.cpp oneflow/api/python/framework/tensor.cpp oneflow/core/framework/op_interpreter/op_interpreter.cpp oneflow/core/framework/op_interpreter/op_interpreter_util.cpp oneflow/core/framework/tensor.cpp oneflow/core/framework/tensor_impl.cpp oneflow/core/framework/tensor_impl.h * Add ones_like unittest. * Use SwithCase * Fix typo * undef * Bugfix * Fix merge conflicits Co-authored-by: hjchen2 <hjchen2> Co-authored-by: NYinggang Wang <wyg19970408@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7b4c8c50 * Cpu support conv module (#4894) * support expand and repeat op int datatype * support conv cpu module * support conv cpu module * support conv cpu module * support conv cpu module Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 59fcd027 * Async cuda stream type (#4895) * add class AsyncCudaStreamType * fix bug * remove useless headfile Co-authored-by: Nlixinqi <lixinqi0703106@163.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 8784b21d * skip infer instr when physical operand != nullptr, remove unused code (#4868) * Disable infer instruction if instruction type has physical operand Signed-off-by: Ndaquexian <daquexian566@gmail.com> * remove more infer instructions Signed-off-by: Ndaquexian <daquexian566@gmail.com> * raise UNIMPLEMENTED() in infer Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix hanging on exit Signed-off-by: Ndaquexian <daquexian566@gmail.com> * reformat Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix typo Signed-off-by: Ndaquexian <daquexian566@gmail.com> * wrap results by Tensor() in .to() Signed-off-by: Ndaquexian <daquexian566@gmail.com> * set need_check_mem_case to false for copy op Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: NLi Xinqi <lixinqi2010@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 3137f51c * fix and check to with module on forwad and backward (#4897) * fix and check to on forwad and backward * add todo Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 585afc09 * add JUST (#4891) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 98b88e20 * Fix docs bug (#4892) * support expand and repeat op int datatype * fix modules docs bug * fix docs bug * fix docstring bug * fix docs bug * fix docs bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: fb164b58 * Add warning when no param update (#4896) * style(Optim): add warning when no param update * style(Optim): add TODO Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7e9902aa * Add gather embedding module (#4826) * add gather module * add gather module * add test case * add embedding module * fix comments * update embedding module and test case * refine * fix comment * fix comment * fix comment Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: NBBuf <1182563586@qq.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 986becca * Support create cpu only tensor (#4863) * support create cpu tensor * add empty op * remove skip tensor test case: * remove skip tensor test case * remove TODO Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: dab6ba23 * Add permute module (#4901) * support expand and repeat op int datatype * fix modules docs bug * fix docs bug * fix docstring bug * fix docs bug * fix docs bug * add permute module Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 698e8cd5 * remove useless code in expand module test (#4903) Former-commit-id: ac06d3c8 * Add conv docs (#4904) * remove useless code in expand module test * add conv2d docs Former-commit-id: 422ced27 * Fix eager test bug (#4678) * skip test_gpt_data_loader in eager mode * 1_node_fix_egaer_test_bug * remove useless head file * skip tensor and module * skip 2-D sbp in eager mode * fix error * fix bug and remove some skip under eager * fix error * del oneflow_api * rm test_tensor.py * skip test_summary in eager mode * skip test_stateful_local_kernel under cpu only mode * add class AsyncCudaStreamType * fix bug * import os * remove BlobObject::is_python_shutting_down_ * fix error * sikp 2d sbp * minor fix * refine comment * make of_format Co-authored-by: Nlixinqi <lixinqi0703106@163.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 425bd439 * cache cudnn handle in bn infer (#4906) Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 516b3e5e * Fix ones zeros like (#4907) * feat(xxxLikeOp): ones_like and zeros_like use user op * fix(Optim): fix learning rate device error bug * style(*): format codes * style(*): use int instead of np.int * test(Optim): add optimizer gpu test (#4908) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 6d7c6d8b * Bump nccl from 2.8.3 to v2.9.8 (#4899) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 3294cf2a * Add prelu module (#4902) * support expand and repeat op int datatype * add prelu module * add prelu module * add prelu module * fix comments * fix comment * fix comment * add backward test * fix comment Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 138af616 * add InitEagerSession for eager mode (#4589) * try to merge eager ofrecord to master branch * refine * temp fix * try to add seed but fails * try to add seed but failsclear * use global function to init mirror/conssitent flag * fix test * add modules * fix record modules * fix destruction order * fix mirror gen seed * skip record unit test * remove TODO Co-authored-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Nmosout <mosout@qq.com> Co-authored-by: NLdpe2G <liangdepeng@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: a139f49e * add hardtanh module (#4914) Former-commit-id: d3c91a97 * fix_matmul_module_test_ci_bug (#4905) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 5c4a0d74 * Cpu support of batchnorm layernorm module (#4890) * add rsqrt moduel * add batchnorm module cpu support * refine * update * fix param ini * add reduce series modules * add batchnorm,layernorm modules and test cases * refine * update .rst * refine according to comments * refine * update * fix layernorm bug * refine * remove additional license Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9262e1ae * Add flow.tensor (#4829) * add flow.tensor * small fix * fix dtype * Update oneflow/python/framework/tensor.py Co-authored-by: Ndaquexian <daquexian566@gmail.com> * deal with multi-dimension list or tuple * remove list * Update oneflow/python/framework/tensor.py Co-authored-by: Ndaquexian <daquexian566@gmail.com> * add unit test case * format Co-authored-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c43caa88 * Feat lr scheduler (#4921) * feat(LrScheduler): add ConsineScheduler * feat(LrScheduler): update cosine_scheduler and add test * feat(LrScheduler): refine codes * style(*): format codes * docs(LrScheduler): add document * docs(LrScheduler): refine documents Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f7881ce7 * Fix CommNetIf::RegisterMemory/UnRegisterMemory lock scope (#4918) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 1a7c26b1 * add leakyrelu module (#4912) * add leakyrelu module * code format * update docs * update docs Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 2b74432f * Support cast in tensor.to (#4917) * support cast in to * refactor to interface * refine doc * refine doc * refine kwargs * add test case support tensor * minor fix and test case * format Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9fac6f36 * add hard swish module (#4915) * add hard swish module * fix bug * fix bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 98f937f1 * Replace oneflow_worker with worker agent (#4900) * naive impl * refine * refine * add log * refine * refine * refine * refine * add todo * refine * refine * sync dynamic libs * refine * fix docker cmd * fix rank * refine * refine * add callbacks simple rpc * refine * refine * fix * refine * refine * refine * fix conn * support tradional mode * refine * refine * refine * rm * refine * refine * refine * refine todo * refine * refine * rm unused * rm todo * revert * refine * add log * refine * refine * fix order * refine * refine * refine * refine * refine * refine * refine * rm * rename * add comment * refine * rm * refine * refine * refine * refine * add todo * add info * refine * refine * refine * add back some legacy code * refine * refine * refine * refine * refine * rm oneflow_worker exe * rm log * fix bug * support --cmd * add check * refine * fix * fmt Former-commit-id: 37c63928 * add hardsigmoid module (#4919) * add hardsigmoid module * refine docs Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 937ccaf3 * Refactor tensor (#4916) * refactor Tensor * Export ConsistentTensor::is_cuda * minor fix * minor fix Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 087c9753 * add relu6 module (#4925) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c013fac1 * add cuda test for add method(module) (#4888) * copy user op * add to module and tensor.to interface * remove unnecessary code * backward for tensor.to * remove capture of input * support cpu only tensor * module to (#4858) * remove backward kernel and op * friendly deal with when tensor.grad is None * minor fix * minor fix * revert * suport 1m1d only * skip test normalization * skip test normalization * skip conv * support construct device using string * minor fix * minor fix * use maybe * fix device id type for device infer ctx * skip batchnorm * skip some tensor test case * startup of add backward * startup of add gpu test * refine * add cuda test for Linear module * refine after sum fixed * gpu backward * gpu backward crashed * retain grad * refine according to comments of WangYinggang * refine: construct specified device tensor * refine testcase * refine: specifiy device when construct in test case * refien testcase for linear * refine * refien to_device * refine import statement * refine import path * remove useless _to_device fun Co-authored-by: NpoohRui <yuruil@qq.com> Co-authored-by: NYurui Li <32978179+poohRui@users.noreply.github.com> Co-authored-by: NXiaoyu Xu <xiaoyulink@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 2e268dfc * add upsample module (#4923) * add upsample module * add upsample2d unittest * add upsample2d unittest * add docs * add UpsamplingNearest2d and UpsamplingBilinear2d module * code format and add docs * add more unit_test Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 8125b43f * add elu module (#4924) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4dfc375a * improve ofrecord unit test (#4920) * improve ofrecord unit test * remove no_grad * fix codes according to review * fix format Former-commit-id: 984b1f08 * align inplace param (#4933) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 8b0bdcdc * refactor SGDUpdate and MomentumUpdate UserOp (#4930) * refactor(ModelUpdate): SGDUpdate and MomentumUpdate use optional input for learning rate * fix(*): fix bugs * style(*): refine code Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 49d53fec * Dev support tensor slice (#4898) * add scalar input support * format * update tensor slice function * add slice slice_update module * add slice funtion in tensor * refine * fix tensor slice * fix bug * add tenser slice test case * add logical_slice_assign module * fix LogicalSliceAssign kernel to support eager local Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix tests Signed-off-by: Ndaquexian <daquexian566@gmail.com> * update export strategy * refine * fix docs * add more test case * fix comments * refine according to comments * add TODO item * format Co-authored-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4c958482 * add tensor.zeros_() and SoftSyncStream instr (#4927) * add tensor.zeros_() and soft sync stream instr Signed-off-by: Ndaquexian <daquexian566@gmail.com> * separate cpu and gpu version of SoftSyncStream * Remove SyncAutoMemset * fix compile error Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix wrong parallel_desc() Signed-off-by: Ndaquexian <daquexian566@gmail.com> * remove unused code Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 0e7f1c6a * Only upload log if distributed test fails (#4934) * only upload log if distributed test fails * refine * refine * reduce timeout * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 08d62afa * Add logsigmoid softplus module (#4929) * add logsigmodi and softplus module * code format * fix docs Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 0a636c6f * Add module backward test (#4926) * add scalar input support * format * update tensor slice function * add slice slice_update module * add slice funtion in tensor * refine * fix tensor slice * fix bug * add tenser slice test case * update softmax testcase * refine softmax backward * add logsoftmax backward test * add maskedfill backward test case * add sigmoid backward test * rewrite transpose bacckward op * add transpose backward test * format * refine * rm useless code * Fix transose unittest. * refine according to comments * update * refine * refine * format * fix backward testcase * fix perm param * fix bug * numpy method to cal sigmoid grad * format * refine * refine comments Co-authored-by: Nhjchen2 <chenhoujiangcug@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7911d713 * fix flow.save (#4941) Signed-off-by: Ndaquexian <daquexian566@gmail.com> Former-commit-id: dbd9d76e * add math.abs (module) Former-commit-id: 9670c24a707e01d5445196dc01c145bda792995d * add math.abs (module) * fix zero point in fake quantization pass (#4586) * fix zero point Signed-off-by: Ndaquexian <daquexian566@gmail.com> * round zero_point in fake quant kernel to align with onnx Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: bc9ec6c4 * also allow ONEFLOW_DEBUG (#4950) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f65019f2 * Add Device Descriptor (#4939) * Add Device Descriptor * format * refine * refine * check cuda version * check cuda version * fix * fix * fix WorkSize * handle more error Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 1cc67da5 * Refactor Optimizers for eager (#4938) * refactor(Adam): refactor Adam to use dynamic learning rate * refactor(Adamw): refactor Adamw Optimizer * refactor(Rmsprop): refactor Rmsprop to use dynamic learning rate Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 49060956 * Add instructions on making sys env permanent (#4949) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 64d3a1d9 * add gradient functions for dim_gather op (#4913) * start up * bugs left * add backward test case * fix bugs * refine testcase * refine * replace optrait with composedAttrs * refine Former-commit-id: e400bc0b * add gradient funs for unary and binary math op (#4961) * add gradient funs for unary and binary math op * add test exp and pow example * refine pow test case * refine * rename register macro Former-commit-id: a559f632 * Refactor consistent tensor (#4937) * refactor Tensor * Export ConsistentTensor::is_cuda * remove ConsistentTensor::blob_object * minor fix * minor fix * fix compiler complains * remove unused code * skip test_creating_consistent_tensor * del useless function Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by:
clackhan <han_binbin@163.com> Former-commit-id: 662cda36 * add backward test case Former-commit-id: 28353e85d5d1aac17aefbcab9f15d78e9aa3bb0e * add backward test case * support tensorrt7 qat (#4958) * support tensorrt7 * Ignore handling bias * support label quantization Former-commit-id: 53881e39 * modify math.abs test case Former-commit-id: 1aa043e7221e71f3d049bcc89b8bd673f132b4bc * modify math.abs test case * fix softmax testcase (#4948) * add scalar input support * format * fix softmax testcase * fix logsoftmax test Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 10ca2bc0 * add module backward (#4935) * add reshape module backward * add expand module backward * add expand backward * add expand module backward * add expand module test * add squeeze module backward * code format * add repeat module backward * code format * fix bug * fix comment * align expand module with torch * align repeat module with torch * fix comment * fix comment * fix comment * fix code format * fix confilict * fix bug * fix comment * add module backward * fix pow bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 5362a0d8 * rewrite unsqueeze backward (#4966) * rewrite unsqueeze backward * fix comments * fix comment Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: ec1c80ff * add module backward (#4942) * add exp module backward * add greater test * add less module test * add negative module backward * code format * add matmul backward * add broadcast_matmul_backward * code format * add batch_matmul backward * add argmax module test * delete unuseless code * fix comment * fix comment * fix comment * fix comment * fix commet * code format * fix bug * fix pow bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: d20a9953 * fix activation ci bug (#4980) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4e974d7b * Hashable attr map (#4951) * Device::compute_dep_object_ * sequantialize instructions in the same stream. * refactor AttrMap * remove redundant header file includes Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: ded8a780 * Fix Global<CommNet>::Delete() (#4981) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 8aade7e7 * delete unused headfile Former-commit-id: 2b3c27a907fbc75c537cabbbaaf5818efb2e2a29 * delete unused headfile * update test case format Former-commit-id: 9cda51fb978fe91245d8c40a4d3fc6a10f2dd4dc * update test case format * add of_softmax_use_fast_math (#4979) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: ba6c64e7 * NetIB device enumeration (#4974) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 3f1038d6 * Fix local tensor requires grad (#4992) * fix(Tensor): add requires_grad setter for ExportTensor * test(Tensor): refine tensor autograd test Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 09552182 * Support localtensor slice (#4985) * add scalar input support * format * register local tensor slice methods Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4f4ce1f2 * Symbol::shared_from_symbol (#4969) * Symbol::shared_from_symbol * fix bug in Symbol::shared_from_symbol Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: a30eb43f * CI formats code automatically (#4983) * check in * qq mail * wrong fmt * auto format by CI * use youarefly@qq.com * wrong fmt * auto format by CI * use ci-bot@oneflow.org * wrong fmt * auto format by CI Co-authored-by: Noneflow-ci-bot <373331853@qq.com> Co-authored-by: Noneflow-ci-bot <youarefly@qq.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9cfee9b1 * Eager consistent tensor (#4984) * Device::compute_dep_object_ * sequantialize instructions in the same stream. * refactor AttrMap * refactor Tensor * Export ConsistentTensor::is_cuda * remove ConsistentTensor::blob_object * refactor TensorImpl * minor fix * fix compiler' complains * Implements EagerConsistentTensorImpl::New * minor fix * fix compiler complains * remove unused code * skip test_creating_consistent_tensor * backup code * Symbol::shared_from_symbol * remove redundant header file includes * fix bug in Symbol::shared_from_symbol * symbolize ParallelDesc and ParallelDistribution Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by:
clackhan <han_binbin@163.com> Former-commit-id: 3356bcad * add resnet50 model test (#4957) * add resnet50 model test * udpate script * udpate script * relax tolerant * run resnet50 in 1n1d * fix format * test resnet50 fun parameters * add resnet50 with and without bn test * fix resnet50 without bn train overflow * change assertEqual to assertTrue Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 01278a66 * Add arg where module (#4998) * add argmax test * add argwhere module * add argwhere module * code format * update unit_test * fix commet * update docs Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 816983aa * fix hierarchical_sub_task_graph_builder condition (#4990) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 0cc74104 * Add __str__ and tolist for tensor (#4928) * add __str__ and tolist * initial tensor printer * reorginized tensor str * minor fix * user nparray2string * Add FunctionNode op_name (#4970) * feat(FunctionNode): add op_type_name * style(FunctionNode): rename op_name to op_type_name * add test case for numel * style(OpExpr): rename type_name to op_type_name (#4976) * add test for tensor str * minor fix * minor fix * support for local tensor * support for local tensor * format * fix typo Co-authored-by: NYinggang Wang <wyg19970408@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: cf2ec98e * Add squeeze module backward (#5007) * add argmax test * add squeeze module backward * fix conflict Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 10c40f28 * RPC backend local supports barrier of barrier_num > 1 (#4968) * check in changes * refine * fix * erase when barrier exits Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c3371486 * add activation module backward test (#4967) * add tanh module backward test * add gelu module backward * fix activation ci bug * add reshape module backward * add tensor reshape module and code format * add permute module backward * fix conflict * add argmax test * fix permute module bug * add prelu module cpu backward * fix prelu gpu backward bug * code format * restruct hardtanh module test * add hardtanh backward * add hardswish backward * add hardsigmoid module backward * add relu module backward * add relu6 module backward * add elu module backward * fix comments Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c1873f74 * Add sys_ptrace for build docker container (#5005) * add sys_ptrace build docker container * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 659f59e1 * Refine pythonpath in cmake (#5002) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7b80f84f * Refactor scope parallel desc (#4996) * Symbol::shared_from_symbol * fix bug in Symbol::shared_from_symbol * symbolize Scope::GetParallelDesc() * IsScalarType * fix compiler complains * fix bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by:
clackhan <han_binbin@163.com> Former-commit-id: 0be87a94 * add arange module backward (#4978) * add arange module backward * update * refine * fix comments * refine * fix docs Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 3878aea7 * CI skips resnet50 to prevent segfault (#5017) * CI skip resnet50 * fix Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 13b46125 * NetSocket device enumeration (#4997) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c38ccebb * Add device unmatch info (#4989) * add argmax test * add device unmatch error * delete unuse code * delete unuse code * refine code * fix var name error bug * add prelu exception get * add prelu exception get * add prelu exception get * add prelu exception get * add exception * code format * fix commet * add more error information * refine error info Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 8b0905b8 * BindFwBwObaPairs skip parallel_cast (#4986) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 3a8bee33 * rewrite matmul op backward (#4988) * rewrite matmul op backward * refine * update * fix comments * refine * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: f06bf705 * Add concat module backward (#5013) * add argmax test * add concat module backward * add concat backward impl * add concat module backward * fix concat module backward bug * fix concat module backward bug * fix concat module backward bug * add concat module backward * delete unuse code * fix comments * fix comments * fix comments * fix comments Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 21f1bddb * rewrite dropout backward (#5014) * rewrite dropout backward * refine * fix comments * auto format by CI Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: bd778bf3 * fix has_grad template (#4962) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 1559abd4 * Upload core files optionally (#5020) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7b5eb78c * Fix docs bug (#5019) * add argmax test * fix oneflow docstring bug Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 1e9c6ff4 * Doesn't allow CI to run PRs in parallel (#5016) * update commit * fix sha Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: d770a56d * try fix (#5029) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 5b7c7649 * add cosh module (#4943) * add cosh module * fix the calculation of cosh backward * add testcase of cosh Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7142ed99 * Fix log level (#5009) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 539d6093 * oop oneflow.Model with training, validation and checkpoint (#4972) * trainer structure * add test * add nnmodel api * nn Model draft * try run global_func in Model * fit to be refined * model run global_func train & eval * nn Model for function style execution draf test pass * refactor nn model * nn model with nessary component * format * rm nn prefix of Model * flow.Model multi-task numpy-input * (flow.Model)op_dataload support multi job * (flow.Model) auto job_func signature for numpy input * (flow.Model)support auto numpy input job * (flow.Model) nump input multi job train test pass * (flow.Model)fix classmethod * fix test * (oneflow.Model)training_step multi output, refine according to pep8 * (oneflow.Model)pep8 check pass by flake8 * pytorch-style module Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix typo, update parameter Signed-off-by: Ndaquexian <daquexian566@gmail.com> * Model refine * Model fix typo * oneflow.Model optimizer variable lazy get, numpy job signature to DataModule * oneflow.Model merge and format * oneflow.Model: comment empty func to be overried * Optimizer: lazy get var add check and tips * oneflow.Model: refactor * oneflow.Model: refactor 2 * add TODO, remove unused import, set consistent to True in parameter Signed-off-by: Ndaquexian <daquexian566@gmail.com> * oneflow.Model: ModelStage -> SubStep, TrainStage -> TrainStep * fix format * oneflow.Model: SubStep to SubModel * oneflow.Model: infer_oneflow_data_placeholder and _infer_job_signature * set placement of parameter Signed-off-by: Ndaquexian <daquexian566@gmail.com> * reformat Signed-off-by: Ndaquexian <daquexian566@gmail.com> * add __init__.py in oneflow.python.nn.modules Signed-off-by: Ndaquexian <daquexian566@gmail.com> * add todo for GetCurrentJobName() * fix typo * oneflow.Model: refine error message * fix format * OOPModel: import new Module * oneflow.Model: rm FunctionConfig in Model * oneflow.Model config_exe to config_execution * OOPModel: add and test naive validate * oneflow.Model: merge module * Optimizer: user mode to confirm that Optimizer.Variable() is called inside a job * merge master * oop model : predict demo * model inherit new module * refine * refine oneflow.mode * add test oop model * add test * no_grad on ones_like * fix has_grad template * format * fix data input * fix * check oop model * add model checkpoint * rm useless code Co-authored-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: NLi Xinqi <lixinqi2010@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 121b1423 * Add upsample module backward (#5025) * add argmax test * add upsample module backward * update upsample unittest * fix unittest bug * refine upsample backward * code format * fix comment * fix comment Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 86d4a672 * Prevent CMake from using highest version of python3 (#5034) * Use conda python if available * refine * refine * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 01e1a714 * rewrite slice backward (#5018) * rewrite slice backward * remove unuse .h Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c1880c01 * fix qat (#5038) * fix qat * format * refine * add comment * Update test.yml * Update test_quantize_op.py * Update test_quantize_op.py Co-authored-by: NShenghang Tsai <jackalcooper@gmail.com> Former-commit-id: 2dfb5b56 * Support convert tensorbuffer to list of numpy (#4940) * support convert tensorbuffer to list of numpy * improve speed * remove useless codes * get tensor_buffer shapes and dtypes by single function * add __eq__ and __hash__ to DType * add dynamic_out to tensor_buffer_to_list_of_tensors_v2 Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 577568af * Add JUST (#5041) Former-commit-id: 994b0df0 * Add broadcast like module backward (#5037) * add argmax test * add broadcastlike module backward, bug need fixed * fix broadcast_like backward bug * auto format by CI * refine code Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: e7bb21bb * Fix segfault caused by zlib in conda when share lib is enabled (#5045) * check in changes * address review * address review Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 9bec9735 * Fix norm grad func to support dynamic attrs. (#5043) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: cc1c554c * Fix segfault in new interface (#5042) * fix data race about composed attr map Signed-off-by: Ndaquexian <daquexian566@gmail.com> * move ResetPrior before ChooseOpKernel Signed-off-by: Ndaquexian <daquexian566@gmail.com> * delete vm before others Signed-off-by: Ndaquexian <daquexian566@gmail.com> * revert deletion order change, sync by atexit Signed-off-by: Ndaquexian <daquexian566@gmail.com> * add comments Signed-off-by: Ndaquexian <daquexian566@gmail.com> * rename Signed-off-by: Ndaquexian <daquexian566@gmail.com> * fix multi machine bug Signed-off-by: Ndaquexian <daquexian566@gmail.com> * auto format by CI Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Former-commit-id: 68a5c10b * Add where module backward (#5035) * add argmax test * add where module backward * fix where module unit_test bug * add zerolike and where op function * add backward code * add broadcast like backward * refine * fix where module backward bug * rebuild test * fix comment * fix comment * fix comment * fix comment * fix comments Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: dec89a96 * Query system status if CI failed (#5052) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 26bbcfcd * fix(vm): add virtual mechine backpressure (#5050) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: cd21df2f * change in_edages/out_edages to from SKIPLIST to LIST (#5047) * change in_edages/out_edages to from SKIPLIST to LIST * minor fix * refine Co-authored-by: NLi Xinqi <lixinqi2010@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: b9568262 * Stop heartbeat and add barrier before Global<CtrlServer>::Delete() (#5010) refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: e861dd3b * fix error on exiting (#5053) * fix error on exiting Signed-off-by: Ndaquexian <daquexian566@gmail.com> * lazily get rank Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 51335128 * Restore eager r50 case (#5055) Former-commit-id: 64665e75 * Add argmax softplus logsigmoid module backward (#5049) * add argmax test * add argmax module backward, bug need fixed * add leakyrelu module backward * delete argmax backward test * delete argmax backward test * add softplus module backward * code format * add softplus module backward * add logsigmoid module backward Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 06b549b3 * add doctest (#5046) * add doctest * refine * update modules doctest Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 7394ec5f * refactor DType (#5024) * refactor DType * fix compiler complains * DType is only allowed to be used in python code * dtype api bugfix * fix error on exiting Signed-off-by: Ndaquexian <daquexian566@gmail.com> * lazily get rank Signed-off-by: Ndaquexian <daquexian566@gmail.com> * Export const DType* into python Co-authored-by:
binbinHan <han_binbin@163.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Ndaquexian <daquexian566@gmail.com> Former-commit-id: c2b2eb25 * remove try_init_session in new interface (#5061) Signed-off-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 912bc97d * Rewrite batch broadcast matmul backward (#5012) * rewrite matmul op backward * refine * update * fix comments * refine * refine * rewrite batch broadcast matmul backward * refine * refine * refine * refine * Add JUST * restructure matmul series module backward * refine * fix comments * fix comments * refine Co-authored-by: Nhjchen2 <chenhoujiangcug@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: c4f277a1 * modify format Former-commit-id: 5bdb45adbb0881263ef5fe0afe13dd606a8e6ff6 * modify format * run make of_format Former-commit-id: e976b9b0f88bc2e91fddb3053dab5bd5001e8318 * run make of_format * lock cmake version in manylinux cmake (#5057) Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: fe662377 * Doctest support in CI (#4973) * check in changes * refine * fix * add test on obj * add relu example * run doctest in ci * dont delete python * address review * address review * address review * address review * address review Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 56f6348d * Update README for 0.4.0 (#4965) * refine * remove content * refine * require py36 * address review * refine * address review Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: cc316206 * Add step lr and lambda lr (#5063) * feat(StepLR): add StepLR * feat(LambdaLR): add LambdaLR * docs(LambdaLR): fix document * style(LambdaLR): add comment Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 4c005b7c * Release cuda 112 (#5060) * Nightly for cu112 * add arg * nightly Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 72cfa991 * add flow.asin, flow.Tensor.asin, flow.arcsin, flow.Tensor.arcsin, flow.asinh, flow.Tensor.asinh, flow.arcsinh, flow.Tensor.arcsinh (#4955) * add flow.asin and torch.arcsin * add torch.asin and torch.arcsin * add torch.sin and torch.arcsin * add torch.asin and torch.arcsin * add torch.asin and torch.arcsin * add torch.asin and torch.arcsin * add torch.asin and torch.arcsin * add torch.asin and torch.arcsin * update test_asin.py including forward and backward * Update test_math_ops.py remove asin testcase * update testcase including forward and backward * add torch.asinh and torch.Tensor.asinh * update testcase of asin and asinh * update testcase of asin and asinh * update testcase of asin and asinh * update testcase * update testcase * make format * update license * update testcase * check in * qq mail * wrong fmt * auto format by CI * use youarefly@qq.com * wrong fmt * auto format by CI * use ci-bot@oneflow.org * wrong fmt * auto format by CI * mv op testcase to test_tensor.py * auto format by CI * update docstring * update docstring * update doctest * update doctest * auto format by CI * update arcsinh * auto format by CI Co-authored-by: N陈岱渊 <chendy@zhejianglab.com> Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: Njackalcooper <jackalcooper@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Noneflow-ci-bot <373331853@qq.com> Co-authored-by: Noneflow-ci-bot <youarefly@qq.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Former-commit-id: 95748870 * Add modules doctest bbuf (#5058) * add argmax test * add nllloss doctest * add doctest * add crossentropyloss doctest * add expand module doctest * add squeeze module doctest * add repeat module doctest * add exp module doctest * add argmax module doctest * add matmul module doctest * add greater module doctest * add less module doctest * add negative module doctest * add linear module doctest * add tanh module doctest * add gelu module doctest * add reshape module doctest * add transpose module doctest * add where module doctest * add permute module doctest * add prelu module doctest * add hardtanh module doctest * add activation module doctest * add activation module doctest * add upsample module doctest * auto format by CI Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Former-commit-id: 58183f87 * remove origin testCase in test_math_ops Former-commit-id: 3a4bae704e2698225b15734fb440a6f8ddef5120 * remove origin testCase in test_math_ops * Add tensor detach python api (#5068) * add argmax test * add tensor detach python api * delete unuse code Former-commit-id: 9b30b7c9 * Delete preprocessor_internal.h.REMOVED.git-id * Delete nn_ops.py.REMOVED.git-id Co-authored-by: NLyon <flowingsun007@163.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: NYinggang Wang <wyg19970408@gmail.com> Co-authored-by: NYurui Li <32978179+poohRui@users.noreply.github.com> Co-authored-by: NXiaoyu Zhang <35585791+BBuf@users.noreply.github.com> Co-authored-by: NLiang Depeng <liangdepeng@gmail.com> Co-authored-by: Ndaquexian <daquexian566@gmail.com> Co-authored-by:
binbinHan <han_binbin@163.com> Co-authored-by: NHoujiang Chen <chenhoujiangcug@gmail.com> Co-authored-by: NLi Xinqi <lixinqi2010@gmail.com> Co-authored-by: NShijie <821898965@qq.com> Co-authored-by: NYao Chi <later@usopp.net> Co-authored-by: NShenghang Tsai <jackalcooper@gmail.com> Co-authored-by: NXiaoyu Xu <xiaoyulink@gmail.com> Co-authored-by: Nlixinqi <lixinqi0703106@163.com> Co-authored-by: NBBuf <1182563586@qq.com> Co-authored-by: NJuncheng <liujuncheng1022@gmail.com> Co-authored-by: Nmosout <mosout@qq.com> Co-authored-by: NpoohRui <yuruil@qq.com> Co-authored-by: Nguo ran <360112263@qq.com> Co-authored-by: Noneflow-ci-bot <373331853@qq.com> Co-authored-by: Noneflow-ci-bot <youarefly@qq.com> Co-authored-by: Noneflow-ci-bot <ci-bot@oneflow.org> Co-authored-by: NYongtaoShi <73167956+YongtaoShi@users.noreply.github.com> Co-authored-by:
yayeoCddy <Dy_Chen95@163.com> Co-authored-by: N陈岱渊 <chendy@zhejianglab.com>
-
- 02 6月, 2021 1 次提交
-
-
由 Juncheng 提交于
Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 26 5月, 2021 1 次提交
-
-
由 guo ran 提交于
Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 08 5月, 2021 1 次提交
-
-
由 Shenghang Tsai 提交于
* Serialize proto in binary rather than text * move del ops out from loop * refine * Skip GenCollectiveBoxingPlan if no CollectiveBoxingTaskNode * multi core to proto * copy pointers explicitly * make toproto const method * reorder * larger tol * Update test_layers_conv1d.py * fix deadlock * remove ForeignCallBack in Operator::ToOpAttribute Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Nliujuncheng <liujuncheng1022@gmail.com> Co-authored-by:
clackhan <han_binbin@163.com>
-
- 05 5月, 2021 1 次提交
-
-
由 cheng cheng 提交于
* Fw/Bw support double compute stream * NCCL comm create by stream id * 2D NCCL logical kernel support BW independent stream * StreamIndex: NcclComputeStream for each subgraph insert nccl logical. * refactor code * refine code for review * Add WITH_CUDA in DoJobPass(InsertNcclLogicalOpPass)
-
- 29 4月, 2021 1 次提交
-
-
由 cheng cheng 提交于
* Pipeline Parallelism: checkpointing insert identity buffer op * fix complier err * identity buffer op custom out regst num * fix bug and runnable * Chain merge divide fw/bw; MemChain ignore merge; copyhd regst num hack * Pipeline buffer pass * Pipeline runnable * rollback NOT merge mem chain hack * pipeline_stage_id_hint and rollback checkpointing buffer * Pipeline buffer only. test pass. * rollback repeat hack * Remove CopyHd Hack; Add buffer cross label loader and loss * refine code for review & fix for new dtype infer * add note Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 28 4月, 2021 2 次提交
-
-
由 guo ran 提交于
* b21 boxing add ctrl_edge * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com> Co-authored-by: Ncheng cheng <472491134@qq.com>
-
由 levi 提交于
* add memory detect info * small fix in opattrref optimize * use bitset * refactor using vector * refine * refine * rename * refine * address review * address review * refine * refine * address review * smaller BITSET_SIZE * refine * refine * refine * refine nameing * refine * refine * refine * update * delete swp file * small update * format fix * format modify * format modify * Update compiler.cpp fix for comment * Update reshape_user_op_util.cpp bug about reshape is fixed Co-authored-by: Njackalcooper <jackalcooper@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 23 4月, 2021 1 次提交
-
-
由 Shenghang Tsai 提交于
* use bitset * refactor using vector * refine * refine * rename * refine * address review * address review * refine * refine * address review * smaller BITSET_SIZE * refine * refine * refine * refine nameing * refine * refine * refine Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 19 4月, 2021 2 次提交
-
-
由 cheng cheng 提交于
* NCCL logical refine timeshape * Insert nccl ops after acc interface * Inser NCCL ops after acc implement; need refine or add new acc_tick_op * deadlock * speed up and run * add acc tick fix deadlocak ; and add nccl comm debug log * refine log: rm cc_debug_log and cclog * use reference for speed up * refine code for review * fix for review Co-authored-by: NJuncheng <liujuncheng1022@gmail.com> Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
由 cheng cheng 提交于
* Remove RtBlobDesc * refine code for RuntimeBlobShapeInferHelper::BlobDesc4BnInOp Co-authored-by: Noneflow-ci-bot <69100618+oneflow-ci-bot@users.noreply.github.com>
-
- 12 4月, 2021 1 次提交
-
-
由 Juncheng 提交于
-
- 07 4月, 2021 1 次提交
-
-
由 Juncheng 提交于
* Fix include cuda header * Fix
-
- 06 4月, 2021 1 次提交
-
-
由 guo ran 提交于
* Nccl support s1 to B and P to s1 * refine
-