[quant][fix] Add bias once in conv_fused

[jerryzh168]

Stack from ghstack:

• [quant][fix] Add bias once in conv_fused #48593 [quant][fix] Add bias once in conv_fused

Summary:
Previously _conv_forward will add self.bias to the result, so bias is added twice in qat ConvBn module
this PR added a bias argument to _conv_forward and _conv_forward is called with zero bias
in ConvBn module

fixes: #48514

Test Plan:

Reviewers:

Subscribers:

Tasks:

Tags:

Differential Revision: D25222215

[dr-ci]

💊 CI failures summary and remediations

As of commit 17e2dbb (more details on the Dr. CI page):

---

• 2/3 failures possibly* introduced in this PR
  • 1/2 non-CircleCI failure(s)
• 1/3 broken upstream at merge base af520d9 on Nov 30 from 7:48am to 3:47pm PDT (15 commits; 5bb2a87 - d74f2d2)

---

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

pytorch_linux_xenial_cuda10_2_cudnn7_py3_gcc7_test2 (1/1)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

Dec 01 00:16:36 ERROR [0.020s]: test_qat_prepare_device_affinity (quantization.test_quantize_fx.TestQuantizeFx)

Dec 01 00:16:36   File "/opt/conda/lib/python3.6/site-packages/torch/nn/intrinsic/qat/modules/conv_fused.py", line 372, in forward 
Dec 01 00:16:36     return F.relu(ConvBn2d._forward(self, input)) 
Dec 01 00:16:36   File "/opt/conda/lib/python3.6/site-packages/torch/nn/intrinsic/qat/modules/conv_fused.py", line 98, in _forward 
Dec 01 00:16:36     conv = self._conv_forward(input, scaled_weight, self.zero_bias) 
Dec 01 00:16:36   File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 390, in _conv_forward 
Dec 01 00:16:36     self.padding, self.dilation, self.groups) 
Dec 01 00:16:36 RuntimeError: Input type (torch.cuda.FloatTensor) and bias type (torch.FloatTensor) should be the same 
Dec 01 00:16:36  
Dec 01 00:16:36  
Dec 01 00:16:36 ====================================================================== 
Dec 01 00:16:36 ERROR [0.020s]: test_qat_prepare_device_affinity (quantization.test_quantize_fx.TestQuantizeFx) 
Dec 01 00:16:36 ---------------------------------------------------------------------- 
Dec 01 00:16:36 Traceback (most recent call last): 
Dec 01 00:16:36   File "/opt/conda/lib/python3.6/site-packages/torch/testing/_internal/common_quantized.py", line 122, in test_fn 
Dec 01 00:16:36     qfunction(*args, **kwargs) 
Dec 01 00:16:36   File "/var/lib/jenkins/workspace/test/quantization/test_quantize_fx.py", line 467, in test_qat_prepare_device_affinity 
Dec 01 00:16:36     model(input) 
Dec 01 00:16:36   File "/opt/conda/lib/python3.6/site-packages/torch/fx/graph_module.py", line 216, in wrapped_call 
Dec 01 00:16:36     return cls_call(self, *args, **kwargs) 
Dec 01 00:16:36   File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 744, in _call_impl 
Dec 01 00:16:36     result = self.forward(*input, **kwargs) 

---

🚧 1 fixed upstream failure:

These were probably caused by upstream breakages that were already fixed.

Please rebase on the viable/strict branch (expand for instructions)

If your commit is newer than viable/strict, you can try basing on an older, stable commit:

git fetch https://github.com/pytorch/pytorch viable/strict
git rebase --onto FETCH_HEAD $(git merge-base origin/master HEAD)

If your commit is older than viable/strict:

git fetch https://github.com/pytorch/pytorch viable/strict
git rebase FETCH_HEAD

Check out the recency history of this "viable master" tracking branch.

• pytorch_bazel_build on Nov 30 from 7:48am to 3:47pm PDT (15 commits; 5bb2a87 - d74f2d2)
  • 🔁 rerun

---

ci.pytorch.org: 1 failed

• Failed: pr/pytorch-linux-bionic-rocm3.9-py3.6

---

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group.

See how this bot performed.

This comment has been revised 24 times.

[facebook-github-bot]

This pull request has been merged in d2e4298.

[facebook-github-bot]

This pull request has been reverted by c81f2d9.