Alexander Dokuchaev

### 🐛 Describe the bug ```python import torch import torch.nn.functional as F weight = torch.ones([16, 1, 32]) bias = torch.ones([16]) stride, padding, dilation, groups = (1, 16, 1, 16) input_1...

[PT] CUDA 12.1

2

### Changes Removed --index-url to run test with cuda12.1

Use tabulate instead of texttable

2

### Changes Using tabulate instead of texttable package to create tables. ### Reason for changes Avoid installation two packages to same task. Tabulate is optional dependencies of pandas package, that...

[PT] Extractor for fused convolution

1

### Changes - Get layer attributes by args, kwargs for `conv`, `batch_norm`, `group_norm` and `linear` functions. - Extract submodules for FBC for custom `conv` and `batch norm` modules. - Add...

### Changes ### Reason for changes ### Related tickets ### Tests

[WIP] Use const graph for nncf.quantize

1

### Changes ### Reason for changes ### Related tickets ### Tests

cast tensor only if detect overflow

1

### Changes ### Reason for changes ### Related tickets ### Tests

Improve Tensor performance

1

### Changes Add custom custom implementation of functools.singledispatch with custom arguments 'wrapper_type' to select wrapper function. Remove `np.errstate` from divide function because it degrades performance, to disable warning add `disable_error_handling`...

Tensor dispatcher

1

### Changes Added dispatcher for tensor to simplify register function, that base on signature of functions. Removed _binary_op_nowarn, _binary_reverse_op_nowarn