1. 13 2月, 2019 1 次提交
    • S
      update · e6f0e318
      Shiyuan Shang-Guan 提交于
      
      Former-commit-id: 9c0e29ba1a7107f42ef5365844daea86dbb65488
      e6f0e318
  2. 12 2月, 2019 2 次提交
  3. 10 2月, 2019 1 次提交
    • S
      add CHECK · 4a099902
      Shiyuan Shang-Guan 提交于
      
      Former-commit-id: 9904ab56a9a2b08630c741b565c1fe5ae5c97928
      4a099902
  4. 30 1月, 2019 1 次提交
  5. 29 1月, 2019 1 次提交
    • S
      Dev jxf bert top k (#1633) · 69c9ee85
      scxfjiang 提交于
      * top_k
      
      * dev top_k op
      
      * refine
      
      * fix bug
      
      * refactor top_k op, cooperate with gather op to get values now
      
      * customized TOPK_KERNEL_ENTRY in auto factory
      
      * batch gather op
      
      * refine
      
      * Backup: batch_gather op, pass compile
      
      * fix bugs, pass the test
      
      * fix no new line at the end of file
      
      * const
      
      * refine by review
      
      * fix bugs
      
      * rename: instance_dim -> instance_size
      
      * remove a blank line
      
      * refine coding style by Juncheng's suggestions, Bravo
      
      * refine top_k
      
      * more refine
      
      * compatible with new model parallel
      
      * refine
      
      * rename
      
      * cpu only in top_k
      
      
      Former-commit-id: 67ade80ed5d7e6ef3f3e1ca4edc541686e15d8f2
      69c9ee85
  6. 28 1月, 2019 1 次提交
  7. 25 1月, 2019 2 次提交
  8. 24 1月, 2019 1 次提交
    • S
      update · 1316afd5
      Shiyuan Shang-Guan 提交于
      
      Former-commit-id: 4281e1080da619cc19a7c5cb1f625de67827cefc
      1316afd5
  9. 23 1月, 2019 6 次提交
    • L
      Dev op graph piece size (#1637) · 156a89cc
      Li Xinqi 提交于
      * fix a bug in OpGraph::InferNoParallelBlobDesc
      
      * fix a bug in OpGraph::InferNoParallelBlobDesc
      
      
      Former-commit-id: c0b1071fc6fbe72f1207d02fdb794dc1076eb59a
      156a89cc
    • S
      Merge branch 'dev_bert' of https://github.com/Oneflow-Inc/oneflow into dev_multi-socket · d8e0912d
      Shiyuan Shang-Guan 提交于
      
      Former-commit-id: 82527249de96a849a3f85b567ff5173980783ba4
      d8e0912d
    • L
      Dev global op graph (#1636) · 2733025d
      Li Xinqi 提交于
      * Global<OpGraph> is only available duraing compilation
      
      * small record_piece_size for InferNoParallelBlobDesc
      
      
      Former-commit-id: 5eb1012703f8f9389ac8e2f16131bfd36411b0db
      2733025d
    • L
      Dev logical blob dim0 (#1635) · d408be08
      Li Xinqi 提交于
      * mem_shared_hint_id
      
      * sharable memory block
      
      * rm useless code
      
      * remove useless code
      
      * bugfix: no redundant edges
      
      * rename: MemBlockGroup => MemBlock
      
      * put constrcutor of SharableMemBlockNode into header file
      
      * bugfix
      
      * rename field: MemBlock.block_id => MemBlock.mem_block_id
      
      * replace piece_size with logical_blob_dim0
      
      * BlobParallelConf
      
      * BlobParallelDesc
      
      * infer out blob model_split_axis
      
      * int64_t => int32_t
      
      * InferOutBlobParallelDesc
      
      * gather out blob model split (#1624)
      
      * InferBlobParallelDesc
      
      * let variable op support kModelParallel
      
      * rename lbi2blob_desc_ => lbi2no_parallel_blob_desc_
      
      * Global<OpGraph>
      
      * SplitLogicalInputBlobDesc
      
      * ConcatOutputBlobDescs
      
      * rename: BlobDataParallel => DataBlobParallel; BlobModelParallel => ModelBlobParallel; BlobGridParallel => GridBlobParallel
      
      * OpGraph::CheckBlobDescs(...)
      
      * exact division is unnecessary
      
      * fix bugs
      
      * rename InferOutBlob* => InferOutputBlob
      
      * exact division in variable_op is unnecessary
      
      * bug fix
      
      * fix bugs
      
      * fix bugs
      
      * IsInputBlobAllowedModelSplit
      
      * use Global<OpGraph> to InferModelSize
      
      * add OpGraph::GetDataBalancedSplitter and OpGraph::GetModelBalancedSplitter
      
      * fix IdentityOp::IsInputBlobAllowedModelSplit
      
      * no implementation for pure virtual function Operator::IsInputBlobAllowedModelSplit
      
      * refine BlobParallelDesc: replace CopyParallelConf with operator=
      
      * refine ParallelDesc: remove unused functions
      
      * more checks on ParallelDesc
      
      * remove unused function Operator::MaxModelSplitNum
      
      * bugfix: SoleOp() => op_vec().at(0)
      
      
      Former-commit-id: be1f820b2927f7f79f55b7891f6575cdeb4b2053
      d408be08
    • S
      make comm_net_conf optional · d6090ce6
      Shiyuan Shang-Guan 提交于
      
      Former-commit-id: 28615613e22379ae2135af09fb2aab1f23b1ec6c
      d6090ce6
    • L
      Dev logical blob dim0 (#1625) · d91685b1
      Li Xinqi 提交于
      * mem_shared_hint_id
      
      * sharable memory block
      
      * rm useless code
      
      * remove useless code
      
      * bugfix: no redundant edges
      
      * rename: MemBlockGroup => MemBlock
      
      * put constrcutor of SharableMemBlockNode into header file
      
      * bugfix
      
      * rename field: MemBlock.block_id => MemBlock.mem_block_id
      
      * replace piece_size with logical_blob_dim0
      
      * BlobParallelConf
      
      * BlobParallelDesc
      
      * infer out blob model_split_axis
      
      * int64_t => int32_t
      
      * InferOutBlobParallelDesc
      
      * gather out blob model split (#1624)
      
      * InferBlobParallelDesc
      
      * let variable op support kModelParallel
      
      * rename lbi2blob_desc_ => lbi2no_parallel_blob_desc_
      
      * Global<OpGraph>
      
      * SplitLogicalInputBlobDesc
      
      * ConcatOutputBlobDescs
      
      * rename: BlobDataParallel => DataBlobParallel; BlobModelParallel => ModelBlobParallel; BlobGridParallel => GridBlobParallel
      
      * OpGraph::CheckBlobDescs(...)
      
      * exact division is unnecessary
      
      * fix bugs
      
      * rename InferOutBlob* => InferOutputBlob
      
      * exact division in variable_op is unnecessary
      
      * bug fix
      
      * fix bugs
      
      * fix bugs
      
      * IsInputBlobAllowedModelSplit
      
      * use Global<OpGraph> to InferModelSize
      
      * add OpGraph::GetDataBalancedSplitter and OpGraph::GetModelBalancedSplitter
      
      * fix IdentityOp::IsInputBlobAllowedModelSplit
      
      * no implementation for pure virtual function Operator::IsInputBlobAllowedModelSplit
      
      * refine BlobParallelDesc: replace CopyParallelConf with operator=
      
      * refine ParallelDesc: remove unused functions
      
      * more checks on ParallelDesc
      
      
      Former-commit-id: 2b78c6e1f37e514e39f1dc807ccce455190b00a7
      d91685b1
  10. 22 1月, 2019 2 次提交
  11. 21 1月, 2019 2 次提交
  12. 20 1月, 2019 2 次提交
  13. 19 1月, 2019 1 次提交
  14. 18 1月, 2019 4 次提交
  15. 17 1月, 2019 5 次提交
  16. 10 1月, 2019 2 次提交
    • S
      refine CHECK in AllReduce (#1618) · 52a6c519
      scxfjiang 提交于
      * refine CHECK in AllReduce
      
      * move ReduceConcatOpCtx definition to .cpp file
      
      
      Former-commit-id: 5a50f692cb92c5a6a7074be2063cbc1ec325c1ca
      52a6c519
    • L
      带策略的寄存器着色 (#1613) · d72a21e2
      Li Xinqi 提交于
      * mem_shared_hint_id
      
      * sharable memory block
      
      * rm useless code
      
      * remove useless code
      
      * bugfix: no redundant edges
      
      * rename: MemBlockGroup => MemBlock
      
      * put constrcutor of SharableMemBlockNode into header file
      
      * bugfix
      
      * rename field: MemBlock.block_id => MemBlock.mem_block_id
      
      
      Former-commit-id: 6a8fc14c2ba6bbe148a84458fa6119af16cbe672
      d72a21e2
  17. 09 1月, 2019 3 次提交
  18. 04 1月, 2019 1 次提交
  19. 02 1月, 2019 2 次提交
    • J
      Dev random shuffle (#1607) · bb0dfaa3
      Juncheng 提交于
      * random shuffle
      
      * fix
      
      * refine
      
      * refine
      
      * single thread
      
      * refine
      
      
      Former-commit-id: 0dbb1f3d7265f9c55a11b07695efd092cd81a83c
      bb0dfaa3
    • S
      Fix jxf reduce concat bug (#1606) · c5310e4c
      scxfjiang 提交于
      * refine logic to infer reduce_concat_op's elem_cnt of out blob, still have bugs...
      
      * add RoundUp in reduce_concat
      
      * CHECK_LE -> CHECK_EQ
      
      * add CHECK
      
      
      Former-commit-id: 962817e2a322ba6452c9966bae87fb5da9d4a86a
      c5310e4c