PaConvert

PaConvert copied to clipboard

【Hackathon 5th No.45】convert torch.cuda.comm.broadcast, torch.cuda.device -part

4

### PR Docs https://github.com/PaddlePaddle/docs/pull/6217 ### PR APIs - [x] convert torch.cuda.comm.broadcast - [x] convert torch.cuda.device

longranger2

【Hackathon 5th No.45】为torch.cuda.seed, torch.cuda.seed_all 添加test -part

5

### PR Docs https://github.com/PaddlePaddle/docs/pull/6217 ### PR APIs - [x] 添加test for torch.cuda.seed - [x] 添加test for torch.cuda.seed_all - https://github.com/PaddlePaddle/Paddle/issues/57262

longranger2

转换规则No65 torch.nn.functional.threshold

12

### PR Docs 根据表格信息，对应文档已存在，未找到对应PR ### PR APIs ```bash torch.nn.functional.threshold ``` ### API Discriptions https://github.com/PaddlePaddle/PaConvert/issues/112 case 65

Liyulingyue

转换规则 No. 10

10

### PR Docs * https://github.com/PaddlePaddle/PaConvert/issues/112 10 * https://github.com/PaddlePaddle/docs/pull/6007 ### PR APIs ```bash paddle.signal.stft ```

enkilee

转换规则 torch.hub.load_state_dict_from_url

3

### PR Docs https://github.com/PaddlePaddle/docs/pull/6090 ### PR APIs ```bash torch.hub.load_state_dict_from_url ```

LokeZhou

单测case覆盖2

1

### PR Docs 单测case覆盖 ### PR APIs ```bash torch.autograd.grad torch.bernoulli torch.chain_matmul torch.cumulative_trapezoid torch.distributions.Bernoulli torch.distributions.Categorical torch.distributions.Geometric torch.distributions.Multinomial torch.dsplit torch.hsplit torch.linalg.matrix_rank torch.linalg.solve_triangular torch.load torch.max torch.max_pool1d torch.nn.init.xavier_uniform_ torch.nn.utils.vector_to_parameters torch.nonzero ```

Xuxuanang

单测case覆盖

1

### PR Docs 单测case覆盖 ### PR APIs torch.cuda.amp.autocast torch.diff torch.distributed.rpc.shutdown torch.distributions.AffineTransform torch.distributions.ExpTransform torch.distributions.PowerTransform torch.histogramdd torch.hub.download_url_to_file torch.inference_mode torch.median torch.nanmedian torch.nn.CrossEntropyLoss torch.nn.KLDivLoss torch.nn.LayerNorm torch.nn.UpsamplingBilinear2d torch.nn.UpsamplingNearest2d torch.nn.functional.grid_sample torch.nn.functional.kl_div torch.nn.functional.pad torch.nn.functional.softmin torch.nn.init.xavier_normal_ ```bash ```

Xuxuanang

test

1

### PR Docs 测试pr ### PR APIs ```bash ```

xuxinyi389

got the wrong result when using paconvet

3

- Pytorch ```python import torch.backends.cudnn as cudnn cudnn.benchmark = True ``` - Paddle ```python import paddle False = True ```

GreatV

got the wrong result when using paconvet

3

- pytorch ```python import torch.nn as nn from functools import partial class test(nn.Module): def __init__(self, in_channels, out_channels, norm_func=nn.LayerNorm): super(test, self).__init__() self.norm = norm_func(in_channels) self.linear = nn.Linear(in_channels, out_channels) def forward(self, x):...

GreatV