Add more information to quantized linear module and added some logs

[jerryzh168]

Summary:
Fixes #771

Test Plan:
python test/dtypes/test_affine_quantized_tensor.py -k test_print_quantized_module

Example output:

Linear(in_features=128, out_features=256, weight=AffineQuantizedTensor(shape=torch.Size([256, 128]), block_size=(1, 128), device=cuda:0, layout_type=PlainLayoutType(), layout_tensor_dtype=torch.int8, quant_min=None, quant_max=None))
.Linear(in_features=128, out_features=256, weight=LinearActivationQuantizedTensor(activation=<function _int8_asymm_per_token_quant at 0x7feb1d146820>, weight=AffineQuantizedTensor(shape=torch.Size([256, 128]), block_size=(1, 32), device=cuda:0, layout_type=PlainLayoutType(), layout_tensor_dtype=torch.int8, quant_min=-8, quant_max=7)))
.Linear(in_features=128, out_features=256, weight=LinearActivationQuantizedTensor(activation=<function _int8_symm_per_token_reduced_range_quant at 0x7feb1d146af0>, weight=AffineQuantizedTensor(shape=torch.Size([256, 128]), block_size=(1, 128), device=cuda:0, layout_type=PlainLayoutType(), layout_tensor_dtype=torch.int8, quant_min=None, quant_max=None)))
.Linear(in_features=128, out_features=256, weight=AffineQuantizedTensor(shape=torch.Size([256, 128]), block_size=(1, 32), device=cuda:0, layout_type=TensorCoreTiledLayoutType(inner_k_tiles=8), layout_tensor_dtype=torch.int32, quant_min=0, quant_max=15))
.Linear(in_features=128, out_features=256, weight=LinearActivationQuantizedTensor(activation=<function _int8_symm_per_token_reduced_range_quant at 0x7feb1d146af0>, weight=AffineQuantizedTensor(shape=torch.Size([256, 128]), block_size=(1, 128), device=cuda:0, layout_type=SemiSparseLayoutType(), layout_tensor_dtype=torch.int8, quant_min=None, quant_max=None)))

Reviewers:

Subscribers:

Tasks:

Tags:

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/782

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit f7e0bbd with merge base 6080986 ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[vkuzo]

optional: is there anything we can do to preserve any preexisting custom repr?

[jerryzh168]

actually there will be a recursive call here, since we are overriding extra_repr()

[vkuzo]

looks great, thanks for working on this!