Skip to content

Update for Torch-TensorRT #14

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

Open
shunk031 wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Update for Torch-TensorRT #14

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
65c9a75
update model
shunk031 Mar 19, 2025
f89646a
add torch-tensorrt
shunk031 Mar 19, 2025
ab758d6
add tests for exporting the MVANet model
shunk031 Mar 19, 2025
892d19e
update for linter
shunk031 Mar 19, 2025
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
58 changes: 34 additions & 24 deletions mvanet/model/MVANet.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ def __init__(
device=torch.device("cuda" if torch.cuda.is_available() else "cpu"),
)

def __call__(self, b, h, w):
def forward(self, b, h, w):
mask = torch.zeros([b, h, w], dtype=torch.bool, device=self.dim_t.device)
assert mask is not None
not_mask = ~mask
Expand Down Expand Up @@ -127,8 +127,7 @@ def __init__(self, d_model, num_heads, pool_ratios=[1, 4, 8]):
self.dropout2 = nn.Dropout(0.1)
self.activation = get_activation_fn("relu")
self.pool_ratios = pool_ratios
self.p_poses = []
self.g_pos = None

self.positional_encoding = PositionEmbeddingSine(
num_pos_feats=d_model // 2, normalize=True
)
Expand All @@ -140,30 +139,31 @@ def forward(self, l, g):
"""
b, c, h, w = l.size()
# 4,c,h,w -> 1,c,2h,2w
breakpoint()
concated_locs = rearrange(l, "(hg wg b) c h w -> b c (hg h) (wg w)", hg=2, wg=2)

pools = []
p_poses, pools = [], []
for pool_ratio in self.pool_ratios:
# b,c,h,w
tgt_hw = (round(h / pool_ratio), round(w / pool_ratio))
pool = F.adaptive_avg_pool2d(concated_locs, tgt_hw)
pools.append(rearrange(pool, "b c h w -> (h w) b c"))
if self.g_pos is None:
pos_emb = self.positional_encoding(
pool.shape[0], pool.shape[2], pool.shape[3]
)
pos_emb = rearrange(pos_emb, "b c h w -> (h w) b c")
self.p_poses.append(pos_emb)

pos_emb = self.positional_encoding(
pool.shape[0], pool.shape[2], pool.shape[3]
)
pos_emb = rearrange(pos_emb, "b c h w -> (h w) b c")
p_poses.append(pos_emb)
pools = torch.cat(pools, 0)
if self.g_pos is None:
self.p_poses = torch.cat(self.p_poses, dim=0)
pos_emb = self.positional_encoding(g.shape[0], g.shape[2], g.shape[3])
self.g_pos = rearrange(pos_emb, "b c h w -> (h w) b c")

p_poses = torch.cat(p_poses, dim=0)
pos_emb = self.positional_encoding(g.shape[0], g.shape[2], g.shape[3])
g_pos = rearrange(pos_emb, "b c h w -> (h w) b c")

# attention between glb (q) & multisensory concated-locs (k,v)
g_hw_b_c = rearrange(g, "b c h w -> (h w) b c")
g_hw_b_c = g_hw_b_c + self.dropout1(
self.attention[0](g_hw_b_c + self.g_pos, pools + self.p_poses, pools)[0]
self.attention[0](g_hw_b_c + g_pos, pools + p_poses, pools)[0]
)
g_hw_b_c = self.norm1(g_hw_b_c)
g_hw_b_c = g_hw_b_c + self.dropout2(
Expand Down Expand Up @@ -219,22 +219,32 @@ def __init__(self, d_model, num_heads, pool_ratios=[1, 4, 8]):
self.dropout2 = nn.Dropout(0.1)
self.activation = get_activation_fn("relu")
self.pool_ratios = pool_ratios
self.p_poses = []
self.g_pos = None

self.positional_encoding = PositionEmbeddingSine(
num_pos_feats=d_model // 2, normalize=True
)

# def calc_concated_locs(self, l: torch.Tensor) -> torch.Tensor:
# # shape: (hg, wg, b, c, h, w)
# l = l.view(2, 2, -1, l.size(1), l.size(2), l.size(3))
# # shape: (b, c, hg, h, wg, w)
# l = l.permute(2, 3, 0, 4, 1, 5).contiguous()
# # shape: (b, c, hg*h, wg*w)
# l = l.view(-1, l.size(1), l.size(2) * l.size(3), l.size(4) * l.size(5))

# return l

def forward(self, l, g):
"""
l: 4,c,h,w
g: 1,c,h,w
"""
b, c, h, w = l.size()

# 4,c,h,w -> 1,c,2h,2w
concated_locs = rearrange(l, "(hg wg b) c h w -> b c (hg h) (wg w)", hg=2, wg=2)
self.p_poses = []
pools = []

pools, p_poses = [], []
for pool_ratio in self.pool_ratios:
# b,c,h,w
tgt_hw = (round(h / pool_ratio), round(w / pool_ratio))
Expand All @@ -245,17 +255,17 @@ def forward(self, l, g):
pool.shape[0], pool.shape[2], pool.shape[3]
)
pos_emb = rearrange(pos_emb, "b c h w -> (h w) b c")
self.p_poses.append(pos_emb)
p_poses.append(pos_emb)
pools = torch.cat(pools, 0)
# if self.g_pos is None:
self.p_poses = torch.cat(self.p_poses, dim=0)

p_poses = torch.cat(p_poses, dim=0)
pos_emb = self.positional_encoding(g.shape[0], g.shape[2], g.shape[3])
self.g_pos = rearrange(pos_emb, "b c h w -> (h w) b c")
g_pos = rearrange(pos_emb, "b c h w -> (h w) b c")

# attention between glb (q) & multisensory concated-locs (k,v)
g_hw_b_c = rearrange(g, "b c h w -> (h w) b c")
g_hw_b_c = g_hw_b_c + self.dropout1(
self.attention[0](g_hw_b_c + self.g_pos, pools + self.p_poses, pools)[0]
self.attention[0](g_hw_b_c + g_pos, pools + p_poses, pools)[0]
)
g_hw_b_c = self.norm1(g_hw_b_c)
g_hw_b_c = g_hw_b_c + self.dropout2(
Expand Down
Loading