roubaofeipi commited on
Commit
5d83c7c
1 Parent(s): 36a67df

Upload 12 files

Browse files
configs/inference/controlnet_c_3b_canny.yaml ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3.6B
3
+ dtype: bfloat16
4
+
5
+ # ControlNet specific
6
+ controlnet_blocks: [0, 4, 8, 12, 51, 55, 59, 63]
7
+ controlnet_filter: CannyFilter
8
+ controlnet_filter_params:
9
+ resize: 224
10
+
11
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
12
+ previewer_checkpoint_path: models/previewer.safetensors
13
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
14
+ controlnet_checkpoint_path: models/canny.safetensors
configs/inference/controlnet_c_3b_identity.yaml ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3.6B
3
+ dtype: bfloat16
4
+
5
+ # ControlNet specific
6
+ controlnet_bottleneck_mode: 'simple'
7
+ controlnet_blocks: [0, 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]
8
+ controlnet_filter: IdentityFilter
9
+ controlnet_filter_params:
10
+ max_faces: 4
11
+ p_drop: 0.00
12
+ p_full: 0.0
13
+
14
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
15
+ previewer_checkpoint_path: models/previewer.safetensors
16
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
17
+ controlnet_checkpoint_path:
configs/inference/controlnet_c_3b_inpainting.yaml ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3.6B
3
+ dtype: bfloat16
4
+
5
+ # ControlNet specific
6
+ controlnet_blocks: [0, 4, 8, 12, 51, 55, 59, 63]
7
+ controlnet_filter: InpaintFilter
8
+ controlnet_filter_params:
9
+ thresold: [0.04, 0.4]
10
+ p_outpaint: 0.4
11
+
12
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
13
+ previewer_checkpoint_path: models/previewer.safetensors
14
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
15
+ controlnet_checkpoint_path: models/inpainting.safetensors
configs/inference/controlnet_c_3b_sr.yaml ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3.6B
3
+ dtype: bfloat16
4
+
5
+ # ControlNet specific
6
+ controlnet_bottleneck_mode: 'large'
7
+ controlnet_blocks: [0, 4, 8, 12, 51, 55, 59, 63]
8
+ controlnet_filter: SREffnetFilter
9
+ controlnet_filter_params:
10
+ scale_factor: 0.5
11
+
12
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
13
+ previewer_checkpoint_path: models/previewer.safetensors
14
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
15
+ controlnet_checkpoint_path: models/super_resolution.safetensors
configs/inference/lora_c_3b.yaml ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3.6B
3
+ dtype: bfloat16
4
+
5
+ # LoRA specific
6
+ module_filters: ['.attn']
7
+ rank: 4
8
+ train_tokens:
9
+ # - ['^snail', null] # token starts with "snail" -> "snail" & "snails", don't need to be reinitialized
10
+ - ['[fernando]', '^dog</w>'] # custom token [snail], initialize as avg of snail & snails
11
+
12
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
13
+ previewer_checkpoint_path: models/previewer.safetensors
14
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
15
+ lora_checkpoint_path: models/lora_fernando_10k.safetensors
configs/inference/stage_b_1b.yaml ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 700M
3
+ dtype: bfloat16
4
+
5
+ # For demonstration purposes in reconstruct_images.ipynb
6
+ webdataset_path: path to your dataset
7
+ batch_size: 1
8
+ image_size: 2048
9
+ grad_accum_steps: 1
10
+
11
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
12
+ stage_a_checkpoint_path: models/stage_a.safetensors
13
+ generator_checkpoint_path: models/stage_b_lite_bf16.safetensors
configs/inference/stage_b_3b.yaml ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3B
3
+ dtype: bfloat16
4
+
5
+ # For demonstration purposes in reconstruct_images.ipynb
6
+ webdataset_path: path to your dataset
7
+ batch_size: 4
8
+ image_size: 1024
9
+ grad_accum_steps: 1
10
+
11
+ effnet_checkpoint_path: path to effnet of stablecascade / effnet_encoder.safetensors
12
+ stage_a_checkpoint_path: path to effnet of stablecascade stage a decoder/stage_a.safetensors
13
+ generator_checkpoint_path: path to effnet of stablecascade stage b decoer heavy version bf16/stage_b_lite_bf16.safetensors
configs/inference/stage_c_1b.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 1B
3
+ dtype: bfloat16
4
+
5
+ effnet_checkpoint_path: path to effnet of stablecascade / effnet_encoder.safetensors
6
+ previewer_checkpoint_path: path to previewer of stablecascade/previewer.safetensors
7
+ generator_checkpoint_path: path to generator of stablecascade stage c lite version bf16 /stage_c_lite_bf16.safetensors
configs/inference/stage_c_3b.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ model_version: 3.6B
3
+ dtype: bfloat16
4
+
5
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
6
+ previewer_checkpoint_path: models/previewer.safetensors
7
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
configs/training/cfg_control_lr.yaml ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ experiment_id: Ultrapixel_controlnet
3
+
4
+ checkpoint_path: checkpoint output path
5
+ output_path: visual results output path
6
+ model_version: 3.6B
7
+ dtype: float32
8
+ # # WandB
9
+ # wandb_project: StableCascade
10
+ # wandb_entity: wandb_username
11
+ #module_filters: ['.depthwise', '.mapper', '.attn', '.channelwise' ]
12
+ #rank: 32
13
+ # TRAINING PARAMS
14
+ lr: 1.0e-4
15
+ batch_size: 12
16
+ #image_size: [1536, 2048, 2560, 3072, 4096]
17
+ image_size: [1024, 2048, 2560, 3072, 3584, 3840, 4096, 4608]
18
+ #image_size: [ 1024, 1536, 2048, 2560, 3072, 3584, 3840, 4096, 4608]
19
+ #image_size: [ 1024, 1280]
20
+ multi_aspect_ratio: [1/1, 1/2, 1/3, 2/3, 3/4, 1/5, 2/5, 3/5, 4/5, 1/6, 5/6, 9/16]
21
+ grad_accum_steps: 2
22
+ updates: 40000
23
+ backup_every: 5000
24
+ save_every: 256
25
+ warmup_updates: 1
26
+ use_fsdp: True
27
+
28
+ # ControlNet specific
29
+ controlnet_blocks: [0, 4, 8, 12, 51, 55, 59, 63]
30
+ controlnet_filter: CannyFilter
31
+ controlnet_filter_params:
32
+ resize: 224
33
+ # offset_noise: 0.1
34
+
35
+ # GDF
36
+ adaptive_loss_weight: True
37
+
38
+ ema_start_iters: 10
39
+ ema_iters: 50
40
+ ema_beta: 0.9
41
+
42
+ webdataset_path: path to your training dataset
43
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
44
+ previewer_checkpoint_path: models/previewer.safetensors
45
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
46
+ controlnet_checkpoint_path: models/canny.safetensors
47
+
48
+
configs/training/lora_personalization.yaml ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ experiment_id: roubao_cat_personalized
3
+
4
+ checkpoint_path: checkpoint output path
5
+ output_path: visual results output path
6
+ model_version: 3.6B
7
+ dtype: float32
8
+
9
+ module_filters: [ '.attn']
10
+ rank: 4
11
+ train_tokens:
12
+ # - ['^snail', null] # token starts with "snail" -> "snail" & "snails", don't need to be reinitialized
13
+ - ['[roubaobao]', '^cat</w>'] # custom token [snail], initialize as avg of snail & snails
14
+ # TRAINING PARAMS
15
+ lr: 1.0e-4
16
+ batch_size: 4
17
+
18
+ image_size: [1024, 2048, 2560, 3072, 3584, 3840, 4096, 4608]
19
+ multi_aspect_ratio: [1/1, 1/2, 1/3, 2/3, 3/4, 1/5, 2/5, 3/5, 4/5, 1/6, 5/6, 9/16]
20
+ grad_accum_steps: 2
21
+ updates: 40000
22
+ backup_every: 5000
23
+ save_every: 512
24
+ warmup_updates: 1
25
+ use_ddp: True
26
+
27
+ # GDF
28
+ adaptive_loss_weight: True
29
+
30
+
31
+ tmp_prompt: a photo of a cat [roubaobao]
32
+ webdataset_path: path to your personalized training dataset
33
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
34
+ previewer_checkpoint_path: models/previewer.safetensors
35
+ generator_checkpoint_path: models/stage_c_bf16.safetensors
36
+ ultrapixel_path: models/ultrapixel_t2i.safetensors
37
+
38
+
configs/training/t2i.yaml ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # GLOBAL STUFF
2
+ experiment_id: ultrapixel_t2i
3
+ #strc_fixlrt_norm3_lite_1024_hrft_newdata
4
+ checkpoint_path: checkpoint output path #output model directory
5
+ output_path: visual results output path #experiment output directory
6
+ model_version: 3.6B # finetune large stage c model of stablecascade
7
+ dtype: float32
8
+
9
+
10
+ # TRAINING PARAMS
11
+ lr: 1.0e-4
12
+ batch_size: 4 # gpu_number * num_per_gpu * grad_accum_steps
13
+ image_size: [1024, 2048, 2560, 3072, 3584, 3840, 4096, 4608] # possible image resolution
14
+ multi_aspect_ratio: [1/1, 1/2, 1/3, 2/3, 3/4, 1/5, 2/5, 3/5, 4/5, 1/6, 5/6, 9/16]
15
+ grad_accum_steps: 2
16
+ updates: 40000
17
+ backup_every: 5000
18
+ save_every: 256
19
+ warmup_updates: 1
20
+ use_ddp: True
21
+
22
+ # GDF
23
+ adaptive_loss_weight: True
24
+
25
+
26
+ webdataset_path: path to your personalized training dataset
27
+ effnet_checkpoint_path: models/effnet_encoder.safetensors
28
+ previewer_checkpoint_path: models/previewer.safetensors
29
+ generator_checkpoint_path: models/stage_c_bf16.safetensors