Oneflow-Inc/oneflow issues and pull requests

#10666 - Segmentation fault (core dumped) in `flow.eye+ diag`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10662 - Aborted (core dumped) in `flow.cuda.synchronize`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10657 - Floating point exception (core dumped) in `flow.argsort`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10655 - Segmentation fault (core dumped) in `oneflow.logical_or`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10654 - Floating point exception (core dumped) in `oneflow.view`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10653 - Floating point exception (core dumped) in `oneflow.reshape`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10652 - Aborted (core dumped) in `oneflow.index_add`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10651 - Segmentation fault (core dumped) in `oneflow._oneflow_internal.autograd.Function.FunctionCtx.mark_non_differentiable`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10650 - Segmentation fault (core dumped) in `flow.scatter_nd`

Issue - State: open - Opened by Daisy2ang 2 months ago
Labels: bug, community

#10647 - adaptation_of_mlu_device_based_on_refactor_xccl_primitive

Pull Request - State: closed - Opened by Flowingsun007 3 months ago - 5 comments
Labels: enhancement, feature, system

#10646 - oneflow.quantile perform differently on cpu and cuda

Issue - State: open - Opened by Redempt1onzzZZ 3 months ago
Labels: bug, community

#10645 - Revert "add interpolate_like for cpu"

Pull Request - State: open - Opened by ShawnXuan 3 months ago

#10643 - Dev fused nms

Pull Request - State: open - Opened by ShawnXuan 3 months ago
Labels: op, api

#10642 - Add NPU support for NMS with scores input

Pull Request - State: closed - Opened by ShawnXuan 3 months ago - 4 comments
Labels: feature, op

#10641 - Support int64 output dtype when use NPU.

Pull Request - State: closed - Opened by MagicalMogu 3 months ago - 7 comments
Labels: feature, op, api

#10639 - Aborted (core dumped) in `flow.nn.BCELoss` when logits contain negative values

Issue - State: open - Opened by deepliao 4 months ago
Labels: bug, community

#10638 - # Aborted (core dumped) in `flow.nn.Conv2d`

Issue - State: open - Opened by deepliao 4 months ago
Labels: bug, community

#10637 - Aborted (core dumped) in flow.unbind

Issue - State: open - Opened by deepliao 4 months ago
Labels: bug, community

#10636 - Segmentation fault (core dumped) in flow.DoubleTensor

Issue - State: open - Opened by deepliao 4 months ago
Labels: bug, community

#10635 - Segmentation fault (core dumped) in oneflow.Tensor.add(self, other)

Issue - State: open - Opened by deepliao 4 months ago
Labels: bug, community

#10634 - Segmentation fault (core dumped) in flow.IntTensor

Issue - State: open - Opened by deepliao 4 months ago
Labels: bug, community

#10633 - Dev nll loss

Pull Request - State: open - Opened by ShawnXuan 4 months ago

#10632 - fix npu resource

Pull Request - State: open - Opened by ShawnXuan 4 months ago

#10632 - fix npu resource

Pull Request - State: closed - Opened by ShawnXuan 4 months ago

#10631 - update to ubuntu-22.04

Pull Request - State: closed - Opened by Flowingsun007 5 months ago - 18 comments
Labels: purge, automerge, ci, need-clean-ccache

#10631 - update to ubuntu-22.04

Pull Request - State: closed - Opened by Flowingsun007 5 months ago - 18 comments
Labels: purge, automerge, ci, need-clean-ccache

#10630 - Import error

Issue - State: open - Opened by eayicati 5 months ago
Labels: bug, community

#10629 - [Question]: Cudnn 9.X support?

Issue - State: open - Opened by dimentary 5 months ago

#10628 - Aborted (core dumped) in flow.optim.LAMB

Issue - State: open - Opened by deepliao 6 months ago
Labels: bug, community

#10627 - Aborted (core dumped) in flow.logspace

Issue - State: open - Opened by deepliao 6 months ago

#10626 - Aborted (core dumped) in flow.linalg.vector_norm

Issue - State: open - Opened by deepliao 6 months ago

#10625 - Aborted (core dumped) in flow.Tensor

Issue - State: open - Opened by deepliao 6 months ago

#10624 - Aborted (core dumped) in flow.topk

Issue - State: open - Opened by deepliao 6 months ago

#10623 - support BUILD_NPU

Pull Request - State: closed - Opened by Flowingsun007 6 months ago - 2 comments
Labels: bug, eager, system

#10623 - support BUILD_NPU

Pull Request - State: closed - Opened by Flowingsun007 6 months ago - 2 comments
Labels: bug, eager, system

#10622 - cache@v2->v4

Pull Request - State: closed - Opened by ShawnXuan 6 months ago - 2 comments
Labels: enhancement, ci

#10621 - Modified README.md

Pull Request - State: closed - Opened by davidliao-commits 6 months ago - 2 comments

#10620 - Modify dropout mask tensor shape and bits format for NPU

Pull Request - State: open - Opened by MagicalMogu 6 months ago - 1 comment

#10620 - Modify dropout mask tensor shape and bits format for NPU

Pull Request - State: open - Opened by MagicalMogu 6 months ago

#10619 - upload-artifact@v3 -> v4

Pull Request - State: closed - Opened by ShawnXuan 6 months ago - 2 comments
Labels: enhancement, ci

#10619 - upload-artifact@v3 -> v4

Pull Request - State: closed - Opened by ShawnXuan 6 months ago - 2 comments
Labels: enhancement, ci

#10618 - adaptation of multi devices based on refactor_xccl_primitive

Pull Request - State: closed - Opened by Flowingsun007 7 months ago - 16 comments
Labels: enhancement, system, need-clean-ccache

#10618 - adaptation of multi devices based on refactor_xccl_primitive

Pull Request - State: closed - Opened by Flowingsun007 7 months ago - 16 comments
Labels: enhancement, system, need-clean-ccache

#10617 - NewCclCommMgr on runtime

Pull Request - State: closed - Opened by ShawnXuan 7 months ago - 5 comments
Labels: bug, eager, system

#10617 - NewCclCommMgr on runtime

Pull Request - State: open - Opened by ShawnXuan 7 months ago

#10616 - [Question]: How to deploy DeepSeek-R1 model using OneFlow framework?

Issue - State: open - Opened by 2catycm 7 months ago

#10615 - oneflow import failed when built with libudev-dev

Issue - State: open - Opened by xuchunmei000 8 months ago
Labels: bug, community

#10614 - support fuse layer norm grad for npu

Pull Request - State: closed - Opened by ShawnXuan 8 months ago - 7 comments
Labels: enhancement, op

#10614 - support fuse layer norm grad for npu

Pull Request - State: closed - Opened by ShawnXuan 8 months ago - 7 comments
Labels: enhancement, op

#10613 - Dev refactor xccl primitive

Pull Request - State: closed - Opened by Flowingsun007 8 months ago - 5 comments
Labels: enhancement, automerge, ci, eager, system

#10613 - Dev refactor xccl primitive

Pull Request - State: closed - Opened by Flowingsun007 8 months ago - 5 comments
Labels: enhancement, automerge, ci, eager, system

#10612 - add fuse_layer_norm_grad functor and op

Pull Request - State: closed - Opened by crazy-JiangDongHua 8 months ago - 5 comments

#10612 - add fuse_layer_norm_grad functor and op

Pull Request - State: open - Opened by crazy-JiangDongHua 8 months ago

#10611 - Only release custom OneFlow once a week

Pull Request - State: closed - Opened by jackalcooper 8 months ago
Labels: enhancement, ci

#10611 - Only release custom OneFlow once a week

Pull Request - State: closed - Opened by jackalcooper 8 months ago
Labels: enhancement, ci

#10610 - Got Exception "libcudnn_cnn_train-fb9074d0.so.8.9.0: undefined symbol: _ZN5cudnn3cnn5infer22queryClusterPropertiesERPhS3_, version libcudnn_cnn_infer.so.8"

Issue - State: open - Opened by wangmouren609 8 months ago

#10609 - fix bug of group norm and layer norm for npu

Pull Request - State: closed - Opened by crazy-JiangDongHua 9 months ago - 8 comments
Labels: enhancement, automerge, op

#10609 - fix bug of group norm and layer norm for npu

Pull Request - State: closed - Opened by crazy-JiangDongHua 9 months ago - 8 comments
Labels: enhancement, automerge, op

#10608 - update error tol

Pull Request - State: closed - Opened by ShawnXuan 9 months ago - 2 comments
Labels: bug, ci

#10608 - update error tol

Pull Request - State: closed - Opened by ShawnXuan 9 months ago - 2 comments
Labels: bug, ci

#10607 - For ci demo

Pull Request - State: closed - Opened by ShawnXuan 9 months ago
Labels: bug, ci

#10607 - For ci demo

Pull Request - State: closed - Opened by ShawnXuan 9 months ago
Labels: bug, ci

#10606 - Use cu118 in canary

Pull Request - State: closed - Opened by jackalcooper 9 months ago
Labels: enhancement, build

#10606 - Use cu118 in canary

Pull Request - State: closed - Opened by jackalcooper 9 months ago
Labels: enhancement, build

#10605 - fix autocast to support global tensor

Pull Request - State: open - Opened by fpzh2011 9 months ago - 6 comments
Labels: enhancement, op, global

#10605 - fix autocast to support global tensor

Pull Request - State: open - Opened by fpzh2011 9 months ago

#10604 - fixed version nvidia-cudnn-cu11 test docker

Pull Request - State: closed - Opened by jackalcooper 9 months ago - 2 comments
Labels: bug, ci

#10604 - fixed version nvidia-cudnn-cu11 test docker

Pull Request - State: closed - Opened by jackalcooper 9 months ago - 2 comments
Labels: bug, ci

#10603 - Llm train fp16

Pull Request - State: closed - Opened by ShawnXuan 9 months ago - 1 comment
Labels: feature, ci, op, interface

#10603 - Llm train fp16

Pull Request - State: closed - Opened by ShawnXuan 9 months ago - 1 comment
Labels: feature, ci, op, interface