File size: 5,031 Bytes
33d23e1 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 | # SPDX-FileCopyrightText: © 2025 Tenstorrent USA, Inc.
# SPDX-License-Identifier: Apache-2.0
import pytest
import torch
import ttnn
from ...layers.conv2d import Conv2d
from ...parallel.config import vae_all_gather
from ...parallel.manager import CCLManager
from ...utils.check import assert_quality
@pytest.mark.parametrize("mesh_device", [(1, 4)], indirect=True)
@pytest.mark.parametrize(
"device_params",
[{"fabric_config": ttnn.FabricConfig.FABRIC_1D, "l1_small_size": 32768, "trace_region_size": 20000000}],
indirect=True,
)
@pytest.mark.parametrize(
(
"batch",
"height",
"width",
"in_channels",
"out_channels",
"kernel",
"padding",
"stride",
"shard_input",
"mesh_axis",
),
[
(1, 128, 128, 16, 512, 3, 1, 1, False, 1),
(1, 128, 128, 512, 512, 3, 1, 1, True, 1),
(1, 256, 256, 512, 512, 3, 1, 1, True, 1),
(1, 512, 512, 512, 512, 3, 1, 1, True, 1),
(1, 512, 512, 512, 256, 3, 1, 1, True, 1),
(1, 512, 512, 256, 256, 3, 1, 1, True, 1),
(1, 1024, 1024, 256, 256, 3, 1, 1, True, 1),
(1, 1024, 1024, 256, 128, 3, 1, 1, True, 1),
(1, 1024, 1024, 128, 128, 3, 1, 1, True, 1),
(1, 1024, 1024, 128, 3, 3, 1, 1, False, None),
],
)
def test_conv2d(
*,
mesh_device: ttnn.Device,
batch: int,
height: int,
width: int,
in_channels: int,
out_channels: int,
kernel: int,
padding: int,
stride: int,
shard_input: bool,
mesh_axis: int,
) -> None:
ccl_manager = CCLManager(mesh_device, topology=ttnn.Topology.Linear)
# models
torch_model = torch.nn.Conv2d(
in_channels=in_channels, out_channels=out_channels, kernel_size=kernel, padding=padding, stride=stride
)
torch.nn.init.normal_(torch_model.weight)
torch.nn.init.normal_(torch_model.bias)
torch_model.eval()
tt_model = Conv2d.from_torch(
torch_ref=torch_model, mesh_device=mesh_device, out_mesh_axis=mesh_axis, ccl_manager=ccl_manager
)
torch_input = torch.randn(batch, in_channels, height, width)
tt_input_tensor = ttnn.from_torch(
torch_input.permute(0, 2, 3, 1),
dtype=ttnn.bfloat16,
device=mesh_device,
mesh_mapper=ttnn.ShardTensorToMesh(mesh_device, dim=-1) if shard_input else None,
layout=ttnn.TILE_LAYOUT,
)
with torch.no_grad():
torch_output = torch_model(torch_input)
tt_out = tt_model(tt_input_tensor)
if mesh_axis is not None:
tt_out = vae_all_gather(ccl_manager, tt_out)
tt_final_out_torch = ttnn.to_torch(ttnn.get_device_tensors(tt_out)[0]).permute(0, 3, 1, 2)
assert_quality(torch_output, tt_final_out_torch, pcc=0.999_500)
@pytest.mark.parametrize("mesh_device", [(1, 4)], indirect=True)
@pytest.mark.parametrize(
"device_params",
[{"fabric_config": ttnn.FabricConfig.FABRIC_1D, "l1_small_size": 32768, "trace_region_size": 20000000}],
indirect=True,
)
@pytest.mark.parametrize(
(
"batch",
"height",
"width",
"in_channels",
"out_channels",
"kernel",
"padding",
"stride",
"mesh_axis",
),
[
(1, 128, 128, 16, 512, 3, 1, 1, 1),
(1, 128, 128, 512, 512, 3, 1, 1, 1),
(1, 256, 256, 512, 512, 3, 1, 1, 1),
(1, 512, 512, 512, 512, 3, 1, 1, 1),
(1, 512, 512, 512, 256, 3, 1, 1, 1),
(1, 512, 512, 256, 256, 3, 1, 1, 1),
(1, 1024, 1024, 256, 256, 3, 1, 1, 1),
(1, 1024, 1024, 256, 128, 3, 1, 1, 1),
(1, 1024, 1024, 128, 128, 3, 1, 1, 1),
],
)
def test_conv2d_input_parallel(
*,
mesh_device: ttnn.Device,
batch: int,
height: int,
width: int,
in_channels: int,
out_channels: int,
kernel: int,
padding: int,
stride: int,
mesh_axis: int,
) -> None:
ccl_manager = CCLManager(mesh_device, topology=ttnn.Topology.Linear)
# models
torch_model = torch.nn.Conv2d(
in_channels=in_channels, out_channels=out_channels, kernel_size=kernel, padding=padding, stride=stride
)
torch.nn.init.normal_(torch_model.weight)
torch.nn.init.normal_(torch_model.bias)
torch_model.eval()
tt_model = Conv2d.from_torch(
torch_ref=torch_model, mesh_device=mesh_device, in_mesh_axis=mesh_axis, ccl_manager=ccl_manager
)
torch_input = torch.randn(batch, in_channels, height, width)
tt_input_tensor = ttnn.from_torch(
torch_input.permute(0, 2, 3, 1),
dtype=ttnn.bfloat16,
device=mesh_device,
mesh_mapper=ttnn.ShardTensorToMesh(mesh_device, dim=-1),
layout=ttnn.TILE_LAYOUT,
)
with torch.no_grad():
torch_output = torch_model(torch_input)
tt_out = tt_model(tt_input_tensor)
tt_out = vae_all_gather(ccl_manager, tt_out)
tt_final_out_torch = ttnn.to_torch(ttnn.get_device_tensors(tt_out)[0]).permute(0, 3, 1, 2)
assert_quality(torch_output, tt_final_out_torch, pcc=0.999_500)
|