- 25 8月, 2023 12 次提交
-
-
由 JZ-LIANG 提交于
* add embedding backward rule * update backward api * revert api * matmul inferbackward * update unitest
-
由 liuruyan 提交于
* fix windows_open_blas error. * fix cmakelists.
-
由 LiYuRio 提交于
* support create shard dist tesnor * support non-zero shard to replicated * change reshard signature
-
由 hong 提交于
* support new ir load combine * update * polish code * remove print * update * update * update * polish code * fix bug * polish code * fix compile bug * fix bug * revert code * remove useless code * polish code
-
由 Yuanle Liu 提交于
* auto mixed precision inference support white list * update * update * update * move down identity_op_clean_pass * fix code style
-
由 xiaoxiaohehe001 提交于
* add_bias_and_simplify_mmha
-
由 ronnywang 提交于
-
由 Frank Lin 提交于
-
由 hong 提交于
* update * fix comile error
-
由 Yichen Zhang 提交于
* fix small bug in reshape spmd rule * small fix in unit test
-
由 Leo Chen 提交于
* bind python interface for pass manager * add ut * revert unused change
-
由 niuliling123 提交于
-
- 24 8月, 2023 8 次提交
-
-
由 niuliling123 提交于
-
由 Leo Chen 提交于
* shard_tensor support static graph * add comments * add dy2static ut * use property in c++ side
-
由 xiaoguoguo626807 提交于
* modify sum with divide net bug mutablesttribute * delete prin
-
由 kangguangli 提交于
* check_unregistered_ops * fix
-
由 liuruyan 提交于
* add symbolicDimProduct & symbolicDimMgr without method shape_constraint related * split ddim in phi, add a target ddim, used by pd_type * add pd_type.cc to ir_shape CMakeLists
-
由 Aurelius84 提交于
[NewIR]Add NOT_FOR_INFER to prune Inference Library Size and Split VJP CodeGen into pd_op_vjp.cc (#56352) * [NewIR]Prune Inference Library Size and Remove IR Dialect * remove options * add NOT_FOR_INFER * fix pd_vjp.cc * polish deps * fix code style * fix unittest * fix cmake * fix inference CI
-
由 YuanRisheng 提交于
-
由 csy0225 提交于
-
- 23 8月, 2023 6 次提交
-
-
由 zhangbo9674 提交于
* fix inplace with set_parameter * fix bug * fix bug * fix bug * fix bug * fix bug * refine code * refine code
-
由 wanghuancoder 提交于
* support ir fill constant
-
由 Leo Chen 提交于
* Integrate quantize/dequantize linear and add config for explicit quantization * Fix the build error * Add macro for TRT version < 8.0 * Remove qdq UT from windows * Fix UT failure * Check TRT version in qdq UT * Test tensorrt_explicit_enabled API * Disable QDQ UT if TRT version < 8.5 * Add quantization postfix into public APIs * Apply code formatter * Fix the UT failure for explicit quantization * Apply code formatter on modified files * Correct the year in copyright
-
由 Charles-hit 提交于
-
由 risemeup1 提交于
* fix coverage compile error * test
-
由 Travis-Lee 提交于
-
- 22 8月, 2023 5 次提交
-
-
由 bukejiyu 提交于
* modify tensorrt ci timeout * activation ci bug fix * Update CMakeLists.txt * Update CMakeLists.txt * comment out int8 mode test_trt_dynamic_shape_groupnorm * fix test_sync_batch_norm_op Segmentation fault * fix timeout * Update CMakeLists.txt
-
由 Yuanle Liu 提交于
-
由 lijin23 提交于
* add inverse kernel for xpu * add more kernels * add index_put kernel for xpu * add index_put kernel for xpu * remove unused headers * refine test * wait to avoid memory bugs for xpu * refine inverse
-
由 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
-
由 FormlessUnit 提交于
* Modify kernels to support quantized_matmul --------- Co-authored-by: Nsuperxf <1208713646@qq.com>
-
- 21 8月, 2023 9 次提交
-
-
由 Aurelius84 提交于
* [NewIR]Split PaddleDialect and KernelDialect and Implement Layered Compailation * remove ir_core DEPS * fix header files * fix conflict * ignore test_assert UT * fix code style * fix UT
-
由 cyber-pioneer 提交于
* change prim forward in New IR * polish code * polish code * test case
-
由 feifei-111 提交于
-
由 zhanglirong1999 提交于
* add cpu quantize pass unit test
-
由 Fisher 提交于
graph_compiler_util.h/cc: 整合GraphCompiler与ParallelCompiler共同持有的数据结构CompilationStage、CompilationStatus、CompilationContext、CompilationResult Parallel Compiler: 整合数据结构至CompilationContext 支持分阶段编译,通过指定CompilationContext::Stage 添加编译状态信息,包括状态CompilationResult::Status和信息CompilationResult::message 一个Task对应一个fusion_group,每一阶段编译完成后,将编译结果放入CompilationResult数组的对应下标中,删去原Task中内部的局部变量,省去了MergeResult Graph Compiler: 丰富CompilationResult,提供编译各阶段的中间结果 整合数据结构至CompilationContext 添加编译状态信息,包括状态CompilationResult::Status和信息CompilationResult::message 启用单测 其他:相关单测、前端接口、paddle2cinn适配CompilationContext
-
由 Lu Qi 提交于
-
由 Ryan 提交于
* add complex numel * change test && add doc
-
由 LiYuRio 提交于
-
由 xiaoguoguo626807 提交于
* support ir api form prim * convert vector of int to intarray * support ir api for prim * Add more gen api * concat python api to concat_grad * fix gen conflict * support vjp prim mode in new ir * remove useless code * add vjp autogen v1.0 * add test for prim * resolve type conflict * modify utils * remove useless code * add split op and modify some bug of vectorType * fix conflict * add concat python test --------- Co-authored-by: Ncyber-pioneer <chenzhuo@tju.edu.cn> Co-authored-by: NCharles-hit <wanghao107@baidu.com> Co-authored-by: N0x45f <wangzhen45@baidu.com> Co-authored-by: Nchenzhiyang <1792266893@qq.com> Co-authored-by: NChen Zhiyang <chenzhiyang99@126.com>
-