1. 05 9月, 2023 1 次提交
    • G
      [Auto Parallel]: Support std::vector<phi::Tensor> input and output for DistTensor. (#56602) · d2fedeac
      Ghost Screaming 提交于
      * [WIP] Support std::vector<phi::Tensor> input and output for DistTensor.
      Concat forward and backward are verified.
      
      * Polish code for new dist tensor implementation.
      
      * Fix bug of DistTensor upgrade. Add support functions for std::vector<Tensor> -> std::vector<Tensor>.
      
      * Add support for DistTensor type of std::vector<phi::Tensor> as input or output of operators.
      Following testcases are passed.
      1. concat: std::vector<phi::Tensor> -> phi::Tensor
      2. unbind: phi::Tensor -> std::vector<phi::Tensor>
      3. broadcast_tensors: std::vector<phi::Tensor> -> std::vector<phi::Tensor>
      
      * Polish code. Remove useless comments.
      
      * Add update_loss_scaling in skip_op_lists.
      
      * Polish code.
      d2fedeac
  2. 22 8月, 2023 1 次提交
    • C
      [AutoParallel] Polish dist tensor design (#56368) · 8495377a
      Chen Weihang 提交于
      * polish dist teensor design
      
      * adjust constructor
      
      * polish details
      
      * polish details design
      
      * fix compile error
      
      * refactor init tensor impl
      
      * fix reshard test
      
      * polish details
      
      * add unittest for coverage
      8495377a
  3. 16 8月, 2023 1 次提交
    • C
      [AutoParallel] Dygraph basic impl for semi auto parallel (#55698) · 7039bef3
      Chen Weihang 提交于
      * add phi forward api gen impl
      
      * add phi backward gen code
      
      * polish api code gen impl
      
      * polish code gen impl
      
      * remove auto_paralel namespace
      
      * add dygraph forward impl
      
      * add for_auto_parallel cond
      
      * fix code gen errors
      
      * add dygraph backward impl
      
      * resolve conflict with develop
      
      * refactor dist api gen impl
      
      * revert origin api gen impl
      
      * replace template for override func
      
      * fix dnnl marco error
      
      * revert third_party change
      
      * add with distributed marco
      
      * Update grad_tensor_holder.cc details
      
      * merge dist tensor constructor
      
      * change test tensor to replicate
      
      * fx typo
      
      * resolve conflict with develop
      
      * fix out dim error
      7039bef3