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

Convert Sinusoid position encoding to torch format #59

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Convert Sinusoid position encoding to torch format #59

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
180246f
Update requirements.txt with pillow
daniegr May 23, 2023
d63407c
Convert get_sinusoid_encoding_table to torch format
daniegr May 23, 2023
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
25 changes: 9 additions & 16 deletions models/multimodal_preprocessors.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,21 +25,14 @@

def get_sinusoid_encoding_table(n_position, d_hid):
"""Sinusoid position encoding table"""

sinusoid_table = torch.FloatTensor(
[pos_i for pos_i in range(n_position)]).unsqueeze(1) * torch.FloatTensor(
[1 / 10000 ** (2 * (hid_j // 2) / d_hid) for hid_j in range(d_hid)]).unsqueeze(0)
sinusoid_table[:, 0::2] = torch.sin(sinusoid_table[:, 0::2]) # dim 2i
sinusoid_table[:, 1::2] = torch.cos(sinusoid_table[:, 1::2]) # dim 2i+1

# TODO: make it with torch instead of numpy
def get_position_angle_vec(position):
return [
position / np.power(10000, 2 * (hid_j // 2) / d_hid)
for hid_j in range(d_hid)
]

sinusoid_table = np.array(
[get_position_angle_vec(pos_i) for pos_i in range(n_position)]
)
sinusoid_table[:, 0::2] = np.sin(sinusoid_table[:, 0::2]) # dim 2i
sinusoid_table[:, 1::2] = np.cos(sinusoid_table[:, 1::2]) # dim 2i+1

return torch.FloatTensor(sinusoid_table).unsqueeze(0)
return sinusoid_table.unsqueeze(0)


def interpolate_pos_encoding_2d(target_spatial_size, pos_embed):
Expand Down Expand Up @@ -197,7 +190,7 @@ def __init__(
self,
rgbt_stem: PatchEmbedGeneric,
depth_stem: Optional[PatchEmbedGeneric],
img_size: Tuple = (3, 224, 224),
img_size: List = [3, 224, 224],
num_cls_tokens: int = 1,
pos_embed_fn: Optional[Callable] = None,
use_type_embed: bool = False,
Expand Down Expand Up @@ -609,7 +602,7 @@ def __init__(
kernel_size: int,
imu_stem: PatchEmbedGeneric,
embed_dim: int,
img_size: Tuple = (6, 2000),
img_size: List = [6, 2000],
num_cls_tokens: int = 1,
pos_embed_fn: Optional[Callable] = None,
init_param_style: str = "openclip",
Expand Down
1 change: 1 addition & 0 deletions requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,3 +15,4 @@ matplotlib
types-regex
mayavi
cartopy
pillow