File size: 2,338 Bytes
b6af722
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import copy

from cosmos_predict1.diffusion.networks.general_dit import GeneralDIT
from cosmos_predict1.diffusion.networks.general_dit_multiview import MultiviewGeneralDIT
from cosmos_predict1.utils.lazy_config import LazyCall as L
from cosmos_predict1.utils.lazy_config import LazyDict

FADITV2Config: LazyDict = L(GeneralDIT)(
    max_img_h=240,
    max_img_w=240,
    max_frames=128,
    in_channels=16,
    out_channels=16,
    patch_spatial=2,
    patch_temporal=1,
    model_channels=4096,
    block_config="FA-CA-MLP",
    num_blocks=28,
    num_heads=32,
    concat_padding_mask=True,
    pos_emb_cls="rope3d",
    pos_emb_learnable=False,
    pos_emb_interpolation="crop",
    block_x_format="THWBD",
    affline_emb_norm=True,
    use_adaln_lora=True,
    adaln_lora_dim=256,
)


FADITV2_14B_Config = copy.deepcopy(FADITV2Config)
FADITV2_14B_Config.model_channels = 5120
FADITV2_14B_Config.num_heads = 40
FADITV2_14B_Config.num_blocks = 36


FADITV2_Multiview_Config: LazyDict = L(MultiviewGeneralDIT)(
    max_img_h=240,
    max_img_w=240,
    max_frames=128,
    in_channels=16,
    out_channels=16,
    patch_spatial=2,
    patch_temporal=1,
    model_channels=4096,
    block_config="FA-CA-MLP",
    num_blocks=28,
    num_heads=32,
    concat_padding_mask=True,
    pos_emb_cls="rope3d",
    pos_emb_learnable=False,
    pos_emb_interpolation="crop",
    block_x_format="THWBD",
    affline_emb_norm=True,
    use_adaln_lora=True,
    adaln_lora_dim=256,
    n_views=6,
    view_condition_dim=6,
    add_repeat_frame_embedding=True,
    rope_h_extrapolation_ratio=1.0,
    rope_w_extrapolation_ratio=1.0,
    rope_t_extrapolation_ratio=1.0,
)