Add files using upload-large-folder tool
Browse files- .gitattributes +2 -0
- dit/config.json +289 -0
- dit/model.flashpack +3 -0
- model_index.json +17 -0
- sampler/scheduler_config.json +12 -0
- vae/config.json +57 -0
- vae/model.flashpack +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
vae/model.flashpack filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
dit/model.flashpack filter=lfs diff=lfs merge=lfs -text
|
dit/config.json
ADDED
|
@@ -0,0 +1,289 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "NaDiT",
|
| 3 |
+
"_diffusers_version": "0.35.2",
|
| 4 |
+
"ada": "single",
|
| 5 |
+
"block_type": [
|
| 6 |
+
"mmdit_sr",
|
| 7 |
+
"mmdit_sr",
|
| 8 |
+
"mmdit_sr",
|
| 9 |
+
"mmdit_sr",
|
| 10 |
+
"mmdit_sr",
|
| 11 |
+
"mmdit_sr",
|
| 12 |
+
"mmdit_sr",
|
| 13 |
+
"mmdit_sr",
|
| 14 |
+
"mmdit_sr",
|
| 15 |
+
"mmdit_sr",
|
| 16 |
+
"mmdit_sr",
|
| 17 |
+
"mmdit_sr",
|
| 18 |
+
"mmdit_sr",
|
| 19 |
+
"mmdit_sr",
|
| 20 |
+
"mmdit_sr",
|
| 21 |
+
"mmdit_sr",
|
| 22 |
+
"mmdit_sr",
|
| 23 |
+
"mmdit_sr",
|
| 24 |
+
"mmdit_sr",
|
| 25 |
+
"mmdit_sr",
|
| 26 |
+
"mmdit_sr",
|
| 27 |
+
"mmdit_sr",
|
| 28 |
+
"mmdit_sr",
|
| 29 |
+
"mmdit_sr",
|
| 30 |
+
"mmdit_sr",
|
| 31 |
+
"mmdit_sr",
|
| 32 |
+
"mmdit_sr",
|
| 33 |
+
"mmdit_sr",
|
| 34 |
+
"mmdit_sr",
|
| 35 |
+
"mmdit_sr",
|
| 36 |
+
"mmdit_sr",
|
| 37 |
+
"mmdit_sr",
|
| 38 |
+
"mmdit_sr",
|
| 39 |
+
"mmdit_sr",
|
| 40 |
+
"mmdit_sr",
|
| 41 |
+
"mmdit_sr"
|
| 42 |
+
],
|
| 43 |
+
"dtype": "bfloat16",
|
| 44 |
+
"emb_dim": 18432,
|
| 45 |
+
"expand_ratio": 4,
|
| 46 |
+
"head_dim": 128,
|
| 47 |
+
"heads": 24,
|
| 48 |
+
"mlp_type": "normal",
|
| 49 |
+
"norm": "fusedrms",
|
| 50 |
+
"norm_eps": 1e-05,
|
| 51 |
+
"num_layers": 36,
|
| 52 |
+
"patch_size": [
|
| 53 |
+
1,
|
| 54 |
+
2,
|
| 55 |
+
2
|
| 56 |
+
],
|
| 57 |
+
"qk_bias": false,
|
| 58 |
+
"qk_norm": "fusedrms",
|
| 59 |
+
"qk_rope": true,
|
| 60 |
+
"shared_mlp": false,
|
| 61 |
+
"shared_qkv": false,
|
| 62 |
+
"temporal_shifted": false,
|
| 63 |
+
"temporal_window_size": null,
|
| 64 |
+
"txt_dim": 3072,
|
| 65 |
+
"txt_in_dim": 5120,
|
| 66 |
+
"vid_dim": 3072,
|
| 67 |
+
"vid_in_channels": 33,
|
| 68 |
+
"vid_out_channels": 16,
|
| 69 |
+
"window": [
|
| 70 |
+
[
|
| 71 |
+
4,
|
| 72 |
+
3,
|
| 73 |
+
3
|
| 74 |
+
],
|
| 75 |
+
[
|
| 76 |
+
4,
|
| 77 |
+
3,
|
| 78 |
+
3
|
| 79 |
+
],
|
| 80 |
+
[
|
| 81 |
+
4,
|
| 82 |
+
3,
|
| 83 |
+
3
|
| 84 |
+
],
|
| 85 |
+
[
|
| 86 |
+
4,
|
| 87 |
+
3,
|
| 88 |
+
3
|
| 89 |
+
],
|
| 90 |
+
[
|
| 91 |
+
4,
|
| 92 |
+
3,
|
| 93 |
+
3
|
| 94 |
+
],
|
| 95 |
+
[
|
| 96 |
+
4,
|
| 97 |
+
3,
|
| 98 |
+
3
|
| 99 |
+
],
|
| 100 |
+
[
|
| 101 |
+
4,
|
| 102 |
+
3,
|
| 103 |
+
3
|
| 104 |
+
],
|
| 105 |
+
[
|
| 106 |
+
4,
|
| 107 |
+
3,
|
| 108 |
+
3
|
| 109 |
+
],
|
| 110 |
+
[
|
| 111 |
+
4,
|
| 112 |
+
3,
|
| 113 |
+
3
|
| 114 |
+
],
|
| 115 |
+
[
|
| 116 |
+
4,
|
| 117 |
+
3,
|
| 118 |
+
3
|
| 119 |
+
],
|
| 120 |
+
[
|
| 121 |
+
4,
|
| 122 |
+
3,
|
| 123 |
+
3
|
| 124 |
+
],
|
| 125 |
+
[
|
| 126 |
+
4,
|
| 127 |
+
3,
|
| 128 |
+
3
|
| 129 |
+
],
|
| 130 |
+
[
|
| 131 |
+
4,
|
| 132 |
+
3,
|
| 133 |
+
3
|
| 134 |
+
],
|
| 135 |
+
[
|
| 136 |
+
4,
|
| 137 |
+
3,
|
| 138 |
+
3
|
| 139 |
+
],
|
| 140 |
+
[
|
| 141 |
+
4,
|
| 142 |
+
3,
|
| 143 |
+
3
|
| 144 |
+
],
|
| 145 |
+
[
|
| 146 |
+
4,
|
| 147 |
+
3,
|
| 148 |
+
3
|
| 149 |
+
],
|
| 150 |
+
[
|
| 151 |
+
4,
|
| 152 |
+
3,
|
| 153 |
+
3
|
| 154 |
+
],
|
| 155 |
+
[
|
| 156 |
+
4,
|
| 157 |
+
3,
|
| 158 |
+
3
|
| 159 |
+
],
|
| 160 |
+
[
|
| 161 |
+
4,
|
| 162 |
+
3,
|
| 163 |
+
3
|
| 164 |
+
],
|
| 165 |
+
[
|
| 166 |
+
4,
|
| 167 |
+
3,
|
| 168 |
+
3
|
| 169 |
+
],
|
| 170 |
+
[
|
| 171 |
+
4,
|
| 172 |
+
3,
|
| 173 |
+
3
|
| 174 |
+
],
|
| 175 |
+
[
|
| 176 |
+
4,
|
| 177 |
+
3,
|
| 178 |
+
3
|
| 179 |
+
],
|
| 180 |
+
[
|
| 181 |
+
4,
|
| 182 |
+
3,
|
| 183 |
+
3
|
| 184 |
+
],
|
| 185 |
+
[
|
| 186 |
+
4,
|
| 187 |
+
3,
|
| 188 |
+
3
|
| 189 |
+
],
|
| 190 |
+
[
|
| 191 |
+
4,
|
| 192 |
+
3,
|
| 193 |
+
3
|
| 194 |
+
],
|
| 195 |
+
[
|
| 196 |
+
4,
|
| 197 |
+
3,
|
| 198 |
+
3
|
| 199 |
+
],
|
| 200 |
+
[
|
| 201 |
+
4,
|
| 202 |
+
3,
|
| 203 |
+
3
|
| 204 |
+
],
|
| 205 |
+
[
|
| 206 |
+
4,
|
| 207 |
+
3,
|
| 208 |
+
3
|
| 209 |
+
],
|
| 210 |
+
[
|
| 211 |
+
4,
|
| 212 |
+
3,
|
| 213 |
+
3
|
| 214 |
+
],
|
| 215 |
+
[
|
| 216 |
+
4,
|
| 217 |
+
3,
|
| 218 |
+
3
|
| 219 |
+
],
|
| 220 |
+
[
|
| 221 |
+
4,
|
| 222 |
+
3,
|
| 223 |
+
3
|
| 224 |
+
],
|
| 225 |
+
[
|
| 226 |
+
4,
|
| 227 |
+
3,
|
| 228 |
+
3
|
| 229 |
+
],
|
| 230 |
+
[
|
| 231 |
+
4,
|
| 232 |
+
3,
|
| 233 |
+
3
|
| 234 |
+
],
|
| 235 |
+
[
|
| 236 |
+
4,
|
| 237 |
+
3,
|
| 238 |
+
3
|
| 239 |
+
],
|
| 240 |
+
[
|
| 241 |
+
4,
|
| 242 |
+
3,
|
| 243 |
+
3
|
| 244 |
+
],
|
| 245 |
+
[
|
| 246 |
+
4,
|
| 247 |
+
3,
|
| 248 |
+
3
|
| 249 |
+
]
|
| 250 |
+
],
|
| 251 |
+
"window_method": [
|
| 252 |
+
"720pwin_by_size_bysize",
|
| 253 |
+
"720pswin_by_size_bysize",
|
| 254 |
+
"720pwin_by_size_bysize",
|
| 255 |
+
"720pswin_by_size_bysize",
|
| 256 |
+
"720pwin_by_size_bysize",
|
| 257 |
+
"720pswin_by_size_bysize",
|
| 258 |
+
"720pwin_by_size_bysize",
|
| 259 |
+
"720pswin_by_size_bysize",
|
| 260 |
+
"720pwin_by_size_bysize",
|
| 261 |
+
"720pswin_by_size_bysize",
|
| 262 |
+
"720pwin_by_size_bysize",
|
| 263 |
+
"720pswin_by_size_bysize",
|
| 264 |
+
"720pwin_by_size_bysize",
|
| 265 |
+
"720pswin_by_size_bysize",
|
| 266 |
+
"720pwin_by_size_bysize",
|
| 267 |
+
"720pswin_by_size_bysize",
|
| 268 |
+
"720pwin_by_size_bysize",
|
| 269 |
+
"720pswin_by_size_bysize",
|
| 270 |
+
"720pwin_by_size_bysize",
|
| 271 |
+
"720pswin_by_size_bysize",
|
| 272 |
+
"720pwin_by_size_bysize",
|
| 273 |
+
"720pswin_by_size_bysize",
|
| 274 |
+
"720pwin_by_size_bysize",
|
| 275 |
+
"720pswin_by_size_bysize",
|
| 276 |
+
"720pwin_by_size_bysize",
|
| 277 |
+
"720pswin_by_size_bysize",
|
| 278 |
+
"720pwin_by_size_bysize",
|
| 279 |
+
"720pswin_by_size_bysize",
|
| 280 |
+
"720pwin_by_size_bysize",
|
| 281 |
+
"720pswin_by_size_bysize",
|
| 282 |
+
"720pwin_by_size_bysize",
|
| 283 |
+
"720pswin_by_size_bysize",
|
| 284 |
+
"720pwin_by_size_bysize",
|
| 285 |
+
"720pswin_by_size_bysize",
|
| 286 |
+
"720pwin_by_size_bysize",
|
| 287 |
+
"720pswin_by_size_bysize"
|
| 288 |
+
]
|
| 289 |
+
}
|
dit/model.flashpack
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0f714ee2e397e7a21528486470ad452e098aade629c23d0a559912da1104cfb8
|
| 3 |
+
size 16479325158
|
model_index.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "SeedVRPipeline",
|
| 3 |
+
"_diffusers_version": "0.35.2",
|
| 4 |
+
"dit": [
|
| 5 |
+
"seedvr.models.dit.nadit",
|
| 6 |
+
"NaDiT"
|
| 7 |
+
],
|
| 8 |
+
"sampler": [
|
| 9 |
+
"seedvr.common.diffusion.samplers.euler",
|
| 10 |
+
"EulerSampler"
|
| 11 |
+
],
|
| 12 |
+
"transform_timesteps": true,
|
| 13 |
+
"vae": [
|
| 14 |
+
"seedvr.models.video_vae_v3.modules.attn_video_vae",
|
| 15 |
+
"VideoAutoencoderKLWrapper"
|
| 16 |
+
]
|
| 17 |
+
}
|
sampler/scheduler_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "EulerSampler",
|
| 3 |
+
"_diffusers_version": "0.35.2",
|
| 4 |
+
"device": "cuda",
|
| 5 |
+
"prediction_type": "v_lerp",
|
| 6 |
+
"return_endpoint": true,
|
| 7 |
+
"schedule_t": 1000.0,
|
| 8 |
+
"schedule_type": "lerp",
|
| 9 |
+
"timesteps_shift": 1.0,
|
| 10 |
+
"timesteps_steps": 1,
|
| 11 |
+
"timesteps_type": "uniform_trailing"
|
| 12 |
+
}
|
vae/config.json
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "VideoAutoencoderKLWrapper",
|
| 3 |
+
"_diffusers_version": "0.35.2",
|
| 4 |
+
"act_fn": "silu",
|
| 5 |
+
"attention": true,
|
| 6 |
+
"block_out_channels": [
|
| 7 |
+
128,
|
| 8 |
+
256,
|
| 9 |
+
512,
|
| 10 |
+
512
|
| 11 |
+
],
|
| 12 |
+
"down_block_types": [
|
| 13 |
+
"DownEncoderBlock3D",
|
| 14 |
+
"DownEncoderBlock3D",
|
| 15 |
+
"DownEncoderBlock3D",
|
| 16 |
+
"DownEncoderBlock3D"
|
| 17 |
+
],
|
| 18 |
+
"extra_cond_dim": null,
|
| 19 |
+
"force_upcast": true,
|
| 20 |
+
"freeze_encoder": false,
|
| 21 |
+
"gradient_checkpoint": false,
|
| 22 |
+
"grouping": false,
|
| 23 |
+
"in_channels": 3,
|
| 24 |
+
"inflation_mode": "pad",
|
| 25 |
+
"latent_channels": 16,
|
| 26 |
+
"latents_mean": null,
|
| 27 |
+
"latents_std": null,
|
| 28 |
+
"layers_per_block": 2,
|
| 29 |
+
"memory_limit": {
|
| 30 |
+
"conv_max_mem": 0.5,
|
| 31 |
+
"norm_max_mem": 0.5
|
| 32 |
+
},
|
| 33 |
+
"mid_block_add_attention": true,
|
| 34 |
+
"norm_num_groups": 32,
|
| 35 |
+
"out_channels": 3,
|
| 36 |
+
"sample_size": 32,
|
| 37 |
+
"scaling_factor": 0.9152,
|
| 38 |
+
"shift_factor": null,
|
| 39 |
+
"slicing": {
|
| 40 |
+
"memory_device": "same",
|
| 41 |
+
"split_size": 4
|
| 42 |
+
},
|
| 43 |
+
"slicing_sample_min_size": 4,
|
| 44 |
+
"slicing_up_num": 0,
|
| 45 |
+
"spatial_downsample_factor": 8,
|
| 46 |
+
"temporal_downsample_factor": 4,
|
| 47 |
+
"temporal_scale_num": 2,
|
| 48 |
+
"time_receptive_field": "full",
|
| 49 |
+
"up_block_types": [
|
| 50 |
+
"UpDecoderBlock3D",
|
| 51 |
+
"UpDecoderBlock3D",
|
| 52 |
+
"UpDecoderBlock3D",
|
| 53 |
+
"UpDecoderBlock3D"
|
| 54 |
+
],
|
| 55 |
+
"use_post_quant_conv": false,
|
| 56 |
+
"use_quant_conv": false
|
| 57 |
+
}
|
vae/model.flashpack
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78beb228cb7b11e15eddbab9e30028b7c60484acd8ccd05d196e8ebe7bc7ed85
|
| 3 |
+
size 1002618763
|