[eager][utils] fix GRAD_SLOT_META_TEMPLATE fmt error

[gouzil]

PR Category

User Experience

PR Types

Bug fixes

Description

修复 GLOG_v>=6 下格式化输出的错误

在 log_v 6 下 const char* GRAD_SLOT_META_TEMPLATE = " {SlotSize: [%d]: %s} "; 需要传大小和内容，而修复前只传了内容

I0409 13:34:50.035048 3900504768 eager_op_function.cc:28054] Running Eager Final State API: arange
I0409 13:34:50.035063 3900504768 eager_op_function.cc:28056] args count: 2
I0409 13:34:50.035090 3900504768 dygraph_functions.cc:56816] Running AD API: arange
I0409 13:34:50.035135 3900504768 dygraph_functions.cc:56840]  No Type Promotion for arange_ad_func api.
I0409 13:34:50.035171 3900504768 dygraph_functions.cc:56862] Running C++ API: arange
Assertion failed: (0 && "tinyformat: Too many conversion specifiers in format string"), function formatImpl, file tinyformat.h, line 816.
[1]    12681 abort      python test/dygraph_to_static/test_no_gradient.py

修复后

I0409 14:24:41.228745 3900504768 eager_op_function.cc:28054] Running Eager Final State API: arange
I0409 14:24:41.228746 3900504768 eager_op_function.cc:28056] args count: 2
I0409 14:24:41.228751 3900504768 dygraph_functions.cc:56816] Running AD API: arange
I0409 14:24:41.228753 3900504768 dygraph_functions.cc:56840]  No Type Promotion for arange_ad_func api.
I0409 14:24:41.228755 3900504768 dygraph_functions.cc:56862] Running C++ API: arange
I0409 14:24:41.228794 3900504768 dygraph_functions.cc:56879] { Input: [
( start , [{Name: None, Initialized: 1, Ptr: 0x12f4864d8,TensorInfo: [ Type: DenseTensor, Dtype: int32, Place: Place(cpu), Shape: 1, DistAttr: Unknown ], ADInfo:[ Grad: [ {Name: None, Initialized: 0, Ptr: 0x0,TensorInfo: [ Unknown ], ADInfo:[ None ]} ],  GradNode: [ BackwardOutMeta: [  {SlotSize: [1]: SlotID: 0, StopGradients: 0, , Edges[ { NULL Edge } ]}  ], BackwardInMeta: [  {SlotSize: [1]: SlotID: 0, StopGradients: 0, , Edges[ { NULL Edge } ]}  ] ], StopGradient: [ 1 ] ]}]),
( end , [{Name: None, Initialized: 1, Ptr: 0x12f4fcf68,TensorInfo: [ Type: DenseTensor, Dtype: int32, Place: Place(cpu), Shape: 1, DistAttr: Unknown ], ADInfo:[ Grad: [ {Name: None, Initialized: 0, Ptr: 0x0,TensorInfo: [ Unknown ], ADInfo:[ None ]} ],  GradNode: [ BackwardOutMeta: [  {SlotSize: [1]: SlotID: 0, StopGradients: 0, , Edges[ { NULL Edge } ]}  ], BackwardInMeta: [  {SlotSize: [1]: SlotID: 0, StopGradients: 0, , Edges[ { NULL Edge } ]}  ] ], StopGradient: [ 1 ] ]}]),
( step , [{Name: None, Initialized: 1, Ptr: 0x12f4dca28,TensorInfo: [ Type: DenseTensor, Dtype: int32, Place: Place(cpu), Shape: 1, DistAttr: Unknown ], ADInfo:[ Grad: [ {Name: None, Initialized: 0, Ptr: 0x0,TensorInfo: [ Unknown ], ADInfo:[ None ]} ],  GradNode: [ BackwardOutMeta: [  {SlotSize: [1]: SlotID: 0, StopGradients: 0, , Edges[ { NULL Edge } ]}  ], BackwardInMeta: [  {SlotSize: [1]: SlotID: 0, StopGradients: 0, , Edges[ { NULL Edge } ]}  ] ], StopGradient: [ 1 ] ]}]), ]}

[paddle-bot]

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

[SigureMo]

LGTMeow