[Bug Report] ttnn.clip interface does not follow Pytorch

[kevinwuTT]

Describe the bug
ttnn.clip has recently changed such that the arguments min and max are keywords. However, the equivalent arguments for torch.clip and also torch.clamp and Hardtanh are positional instead.

https://pytorch.org/docs/stable/generated/torch.clip.html

torch.clip(input, min=None, max=None, *, out=None) → Tensor

https://pytorch.org/docs/main/generated/torch.clamp.html
https://pytorch.org/docs/stable/generated/torch.nn.Hardtanh.html#torch.nn.Hardtanh

This changed was observed since tags/v0.53.0-rc21 but could have been earlier.

To Reproduce

import ttnn
import torch

tensor = torch.rand((4, 4), dtype=torch.bfloat16)

torch_clip = torch.clip(tensor, 0.5, 0.5)

with ttnn.manage_device(device_id=0) as device:
    ttnn_from_torch = ttnn.from_torch(tensor, layout = ttnn.TILE_LAYOUT, dtype = ttnn.bfloat16, device = device)
    ttnn_clip = ttnn.clip(ttnn_from_torch, -0.5, 0.5)
    ttnn_to_torch = ttnn.to_torch(ttnn_clip)

Error message:

TypeError: __call__(): incompatible function arguments. The following argument types are supported:
    1. (self: ttnn._ttnn.operations.unary.clip_t, input_tensor: ttnn._ttnn.tensor.Tensor, *, min: Optional[float] = None, max: Optional[float] = None, memory_config: Optional[ttnn._ttnn.tensor.MemoryConfig] = None) -> ttnn._ttnn.tensor.Tensor

Expected behavior
If the min and max args are positional like torch.clip then ttnn.clip would not show any error.

Please complete the following environment information:

• OS: Ubuntu 20.04.6 LTS
• tags/v0.53.0-rc21

[VirdhatchaniKN]

Hi @kevinwuTT
#14127 Should Fix this issue. Can you check this?

[kevinwuTT]

@VirdhatchaniKN
I can confirm it works with this branch. Thanks!

[VirdhatchaniKN]

Will close issue after PR gets merged : #14127