Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add support for AQTLayout, PlainAQTLayout and TensorCoreTiledAQTLayout #278

Merged
merged 3 commits into from
May 29, 2024
Merged

Add support for AQTLayout, PlainAQTLayout and TensorCoreTiledAQTLayout #278

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f08f6de
Add support for `AQTStorage` and `PlainAQTStorage`
jerryzh168 May 25, 2024
ec3f79f
Merge branch 'main' into layout
jerryzh168 May 29, 2024
4758982
Merge branch 'main' into layout
jerryzh168 May 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions test/quantization/test_quant_api.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -110,8 +110,8 @@ def __init__(self, m=64, n=32, k=64):
self.linear1 = torch.nn.Linear(m, n, bias=False).to(torch.float)
self.linear2 = torch.nn.Linear(n, k, bias=False).to(torch.float)

def example_inputs(self, batch_size=1):
return (torch.randn(batch_size, self.linear1.in_features).to(torch.float),)
def example_inputs(self, batch_size=1, dtype=torch.float, device="cpu"):
return (torch.randn(batch_size, self.linear1.in_features, dtype=dtype, device=device),)

def forward(self, x):
x = self.linear1(x)
Expand Down Expand Up @@ -450,7 +450,7 @@ def test_quantized_tensor_subclass_int4(self):
# use 1024 so that we don't need padding
m = ToyLinearModel(1024, 1024, 1024).eval().to(torch.bfloat16).to("cuda")
m_copy = copy.deepcopy(m)
example_inputs = tuple(map(lambda x: x.to(torch.bfloat16).to("cuda"), m.example_inputs()))
example_inputs = m.example_inputs(dtype=torch.bfloat16, device="cuda")

groupsize = 32
m = quantize(m, get_apply_int4wo_quant(groupsize=groupsize))
Expand Down Expand Up @@ -496,7 +496,7 @@ def test_quantized_tensor_subclass_int8_dyn_quant(self):
m = ToyLinearModel(1024, 1024, 1024).eval().to(torch.bfloat16).to("cuda")
m_copy = copy.deepcopy(m)
# setting batch_size to 20 to be compatible with the kernel
example_inputs = tuple(map(lambda x: x.to(torch.bfloat16).to("cuda"), m.example_inputs(batch_size=20)))
example_inputs = m.example_inputs(batch_size=20, dtype=torch.bfloat16, device="cuda")
m = quantize(m, get_apply_int8dyn_quant())

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't see any tests explicitly testing TensorCoreTiledAQTLayout

assert isinstance(m.linear1.weight, LinearActQuantizedTensor)
Expand Down
Loading
Loading