Paddle

Paddle copied to clipboard

SlogDeterminant kernel support complex 易用性提升

1

### PR Category User Experience ### PR Types Improvements ### Description SlogDeterminant kernel 支持complex64/complex128 修改方法与 https://github.com/PaddlePaddle/Paddle/pull/68390 相同，因为 slogdeterminant_grad_kernel 中的计算本身支持复数，所以不需要修改。 另外根据 slogdeterminant_grad_kernel 的计算推导结果重新修复了 determinant_grad ，并测试了 out_grad 不是 1+0j 而是任意复数的情况。

NKNaN

[OneDNN][PIR] Fix picodet performance drop in bf16

1

### PR Category Performance Optimization ### PR Types Performance ### Description Scale Op with bias skip bfloat16, since scale with bias add quant/dequant will make result different and lead MultiClassNMSKernel...

zhanglirong1999

【Hackathon 7th No.19】NO.19为 Paddle 新增 load_state_dict_from_url API v1

2

### PR Category User Experience ### PR Types Improvements ### Description 【Hackathon 7th No.19】NO.19为 Paddle 新增 load_state_dict_from_url API v1

zty-king

add pd_op.min converter for PIR TRT and Add ReduceCommonOpPattern

1

### PR Category Inference ### PR Types New features ### Description pcard-71500 add pd_op.min converter for PIR TRT and Add ReduceCommonOpPattern base class to simplify code

zhink

[Docathon][Add API Legend No.42] Add legend for as_complex API

1

### PR Category User Experience ### PR Types Docs ### Description update legend for as_complex api in https://github.com/PaddlePaddle/docs/pull/6913

Leoforever123

DataLoader随机出现报错

3

Paddle3.0安装完后运行会报“Illegal instruction”，但机器显示支持avx指令集，看到issue区也很多同样问题没解决方案，所以降级装了paddle 2.6 当前环境： Ubuntu 22.04 CUDA 12.4 Driver Version: 550.54.14，Nvidia A800 80G nvidia-cublas-cu12 12.6.1.4 nvidia-cudnn-cu12 9.4.0.58 paddlepaddle-gpu 2.6.1.post120 paddlesci 0.0.0 python 3.12.5 问题： DataLoader会随机出现报错，同样的代码有时能正常跑完，有时报这个错  可能相关代码： ``` dataList =...

rayLemond

PaddlePaddle安装PaddlePaddle，使用import paddle导入出现Illegal instruction

11

### 问题描述 Please describe your issue 

sheiyaoligun

paddle是否支持容错

4

### 请提出你的问题 Please ask your question 在training-operator v1.8中，可以看到paddle是支持弹性调度的，我想问一下paddle是否支持容错，类似于pytorch框架elastic中的RDZV，如果支持的话，可以提供一下相关的文档吗？

zhaizhch

fix FLAGS_enable_pir_api when enable_new_ir

1

### PR Category Inference ### PR Types Bug fixes ### Description pcard-71500 Fixed the issue where the PIR Pass log is still printed when the high-priority FLAGS_enable_pir_api environment variable takes...

zhink

[0-size Tensor No.6、7、284、285][0-size Tensor Job2 No.47、49、88、90]Add 0-size Tensor support for amin

2

### PR Category Execute Infrastructure ### PR Types Improvements ### Description [0-size Tensor No.6、7、284、285] paddle.amax paddle.amin paddle.Tensor.amax paddle.Tensor.amin [0-size Tensor Job2 No.47、49、88、90] 任务二中在同一个单测文件 paddle.max paddle.min paddle.Tensor.max paddle.Tensor.min 修改 paddle/phi/kernels/funcs/reduce_function.h 中检查...

co63oc