Onocom commited on
Commit
47a29bf
1 Parent(s): 77a304b

Upload 2 files

Browse files
Ono-CrossFusion/Ono-CrossFusion - V6 - fix.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb70f7a0e8d11d86fd6915560e7b4c6cc4e9827f8cb218320963b74951d8e227
3
+ size 2132710878
Ono-CrossFusion/Ono-CrossFusion - V6 - fix.yaml ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ model:
2
+ base_learning_rate: 1.0e-04
3
+ target: ldm.models.diffusion.ddpm.LatentDiffusion
4
+ params:
5
+ parameterization: "v"
6
+ linear_start: 0.00085
7
+ linear_end: 0.0120
8
+ num_timesteps_cond: 1
9
+ log_every_t: 200
10
+ timesteps: 1000
11
+ first_stage_key: "jpg"
12
+ cond_stage_key: "txt"
13
+ image_size: 64
14
+ channels: 4
15
+ cond_stage_trainable: false # Note: different from the one we trained before
16
+ conditioning_key: crossattn
17
+ monitor: val/loss_simple_ema
18
+ scale_factor: 0.18215
19
+ use_ema: False
20
+
21
+ scheduler_config: # 10000 warmup steps
22
+ target: ldm.lr_scheduler.LambdaLinearScheduler
23
+ params:
24
+ warm_up_steps: [ 10000 ]
25
+ cycle_lengths: [ 10000000000000 ] # incredibly large number to prevent corner cases
26
+ f_start: [ 1.e-6 ]
27
+ f_max: [ 1. ]
28
+ f_min: [ 1. ]
29
+
30
+ unet_config:
31
+ target: ldm.modules.diffusionmodules.openaimodel.UNetModel
32
+ params:
33
+ use_fp16: True
34
+ image_size: 32 # unused
35
+ in_channels: 4
36
+ out_channels: 4
37
+ model_channels: 320
38
+ attention_resolutions: [ 4, 2, 1 ]
39
+ num_res_blocks: 2
40
+ channel_mult: [ 1, 2, 4, 4 ]
41
+ num_heads: 8
42
+ use_spatial_transformer: True
43
+ transformer_depth: 1
44
+ context_dim: 768
45
+ use_checkpoint: True
46
+ legacy: False
47
+
48
+ first_stage_config:
49
+ target: ldm.models.autoencoder.AutoencoderKL
50
+ params:
51
+ embed_dim: 4
52
+ monitor: val/rec_loss
53
+ ddconfig:
54
+ double_z: true
55
+ z_channels: 4
56
+ resolution: 256
57
+ in_channels: 3
58
+ out_ch: 3
59
+ ch: 128
60
+ ch_mult:
61
+ - 1
62
+ - 2
63
+ - 4
64
+ - 4
65
+ num_res_blocks: 2
66
+ attn_resolutions: []
67
+ dropout: 0.0
68
+ lossconfig:
69
+ target: torch.nn.Identity
70
+
71
+ cond_stage_config:
72
+ target: ldm.modules.encoders.modules.FrozenCLIPEmbedder