embracellm commited on
Commit
2265b48
1 Parent(s): 8febc1d

End of training

Browse files
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: openrail++
3
+ library_name: diffusers
4
+ tags:
5
+ - text-to-image
6
+ - text-to-image
7
+ - diffusers-training
8
+ - diffusers
9
+ - dora
10
+ - template:sd-lora
11
+ - stable-diffusion-xl
12
+ - stable-diffusion-xl-diffusers
13
+ base_model: stabilityai/stable-diffusion-xl-base-1.0
14
+ instance_prompt: a photo of sushi
15
+ widget: []
16
+ ---
17
+
18
+ <!-- This model card has been generated automatically according to the information the training script had access to. You
19
+ should probably proofread and complete it, then remove this comment. -->
20
+
21
+
22
+ # SDXL LoRA DreamBooth - embracellm/sushi_LoRA_2
23
+
24
+ <Gallery />
25
+
26
+ ## Model description
27
+
28
+ These are embracellm/sushi_LoRA_2 LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0.
29
+
30
+ The weights were trained using [DreamBooth](https://dreambooth.github.io/).
31
+
32
+ LoRA for the text encoder was enabled: False.
33
+
34
+ Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.
35
+
36
+ ## Trigger words
37
+
38
+ You should use a photo of sushi to trigger the image generation.
39
+
40
+ ## Download model
41
+
42
+ Weights for this model are available in Safetensors format.
43
+
44
+ [Download](embracellm/sushi_LoRA_2/tree/main) them in the Files & versions tab.
45
+
46
+
47
+
48
+ ## Intended uses & limitations
49
+
50
+ #### How to use
51
+
52
+ ```python
53
+ # TODO: add an example code snippet for running this diffusion pipeline
54
+ ```
55
+
56
+ #### Limitations and bias
57
+
58
+ [TODO: provide examples of latent issues and potential remediations]
59
+
60
+ ## Training details
61
+
62
+ [TODO: describe the data used to train the model]
logs/dreambooth-lora-sd-xl/1714203341.0314255/events.out.tfevents.1714203341.235b5debc8b2.11877.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa90faa7a82daed3f56c467d9f11e568a67fa78dea916426397f0f0c0223028c
3
+ size 3307
logs/dreambooth-lora-sd-xl/1714203341.038226/hparams.yml ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ adam_beta1: 0.9
2
+ adam_beta2: 0.999
3
+ adam_epsilon: 1.0e-08
4
+ adam_weight_decay: 0.0001
5
+ adam_weight_decay_text_encoder: 0.001
6
+ allow_tf32: false
7
+ cache_dir: null
8
+ caption_column: prompt
9
+ center_crop: false
10
+ checkpointing_steps: 717
11
+ checkpoints_total_limit: null
12
+ class_data_dir: null
13
+ class_prompt: null
14
+ dataloader_num_workers: 0
15
+ dataset_config_name: null
16
+ dataset_name: null
17
+ do_edm_style_training: false
18
+ enable_xformers_memory_efficient_attention: false
19
+ gradient_accumulation_steps: 3
20
+ gradient_checkpointing: true
21
+ hub_model_id: null
22
+ hub_token: null
23
+ image_column: image
24
+ instance_data_dir: sushi_try
25
+ instance_prompt: a photo of sushi
26
+ learning_rate: 0.0001
27
+ local_rank: -1
28
+ logging_dir: logs
29
+ lr_num_cycles: 1
30
+ lr_power: 1.0
31
+ lr_scheduler: constant
32
+ lr_warmup_steps: 0
33
+ max_grad_norm: 1.0
34
+ max_train_steps: 500
35
+ mixed_precision: fp16
36
+ num_class_images: 100
37
+ num_train_epochs: 250
38
+ num_validation_images: 4
39
+ optimizer: AdamW
40
+ output_dir: sushi_LoRA_2
41
+ output_kohya_format: false
42
+ pretrained_model_name_or_path: stabilityai/stable-diffusion-xl-base-1.0
43
+ pretrained_vae_model_name_or_path: madebyollin/sdxl-vae-fp16-fix
44
+ prior_generation_precision: null
45
+ prior_loss_weight: 1.0
46
+ prodigy_beta3: null
47
+ prodigy_decouple: true
48
+ prodigy_safeguard_warmup: true
49
+ prodigy_use_bias_correction: true
50
+ push_to_hub: false
51
+ random_flip: false
52
+ rank: 4
53
+ repeats: 1
54
+ report_to: tensorboard
55
+ resolution: 1024
56
+ resume_from_checkpoint: null
57
+ revision: null
58
+ sample_batch_size: 4
59
+ scale_lr: false
60
+ seed: 0
61
+ snr_gamma: 5.0
62
+ text_encoder_lr: 5.0e-06
63
+ train_batch_size: 1
64
+ train_text_encoder: false
65
+ use_8bit_adam: true
66
+ use_dora: false
67
+ validation_epochs: 50
68
+ validation_prompt: null
69
+ variant: null
70
+ with_prior_preservation: false
logs/dreambooth-lora-sd-xl/1714203463.2332785/events.out.tfevents.1714203463.235b5debc8b2.12630.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0cb7417d76d1111fc234237cfdfaaa84c1b202353b84c2bed096fd54723599b9
3
+ size 3307
logs/dreambooth-lora-sd-xl/1714203463.2397501/hparams.yml ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ adam_beta1: 0.9
2
+ adam_beta2: 0.999
3
+ adam_epsilon: 1.0e-08
4
+ adam_weight_decay: 0.0001
5
+ adam_weight_decay_text_encoder: 0.001
6
+ allow_tf32: false
7
+ cache_dir: null
8
+ caption_column: prompt
9
+ center_crop: false
10
+ checkpointing_steps: 717
11
+ checkpoints_total_limit: null
12
+ class_data_dir: null
13
+ class_prompt: null
14
+ dataloader_num_workers: 0
15
+ dataset_config_name: null
16
+ dataset_name: null
17
+ do_edm_style_training: false
18
+ enable_xformers_memory_efficient_attention: false
19
+ gradient_accumulation_steps: 3
20
+ gradient_checkpointing: true
21
+ hub_model_id: null
22
+ hub_token: null
23
+ image_column: image
24
+ instance_data_dir: sushi_try
25
+ instance_prompt: a photo of sushi
26
+ learning_rate: 0.0001
27
+ local_rank: -1
28
+ logging_dir: logs
29
+ lr_num_cycles: 1
30
+ lr_power: 1.0
31
+ lr_scheduler: constant
32
+ lr_warmup_steps: 0
33
+ max_grad_norm: 1.0
34
+ max_train_steps: 500
35
+ mixed_precision: fp16
36
+ num_class_images: 100
37
+ num_train_epochs: 250
38
+ num_validation_images: 4
39
+ optimizer: AdamW
40
+ output_dir: sushi_LoRA_2
41
+ output_kohya_format: false
42
+ pretrained_model_name_or_path: stabilityai/stable-diffusion-xl-base-1.0
43
+ pretrained_vae_model_name_or_path: madebyollin/sdxl-vae-fp16-fix
44
+ prior_generation_precision: null
45
+ prior_loss_weight: 1.0
46
+ prodigy_beta3: null
47
+ prodigy_decouple: true
48
+ prodigy_safeguard_warmup: true
49
+ prodigy_use_bias_correction: true
50
+ push_to_hub: false
51
+ random_flip: false
52
+ rank: 4
53
+ repeats: 1
54
+ report_to: tensorboard
55
+ resolution: 1024
56
+ resume_from_checkpoint: null
57
+ revision: null
58
+ sample_batch_size: 4
59
+ scale_lr: false
60
+ seed: 0
61
+ snr_gamma: 5.0
62
+ text_encoder_lr: 5.0e-06
63
+ train_batch_size: 1
64
+ train_text_encoder: false
65
+ use_8bit_adam: true
66
+ use_dora: false
67
+ validation_epochs: 50
68
+ validation_prompt: null
69
+ variant: null
70
+ with_prior_preservation: false
logs/dreambooth-lora-sd-xl/1714203920.794048/events.out.tfevents.1714203920.235b5debc8b2.14637.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dcd2ca8935c4e2a75bdf7f99bd5d0b4d3e62daf3bacb13b6b3d0b9134766f35b
3
+ size 3307
logs/dreambooth-lora-sd-xl/1714203920.8008738/hparams.yml ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ adam_beta1: 0.9
2
+ adam_beta2: 0.999
3
+ adam_epsilon: 1.0e-08
4
+ adam_weight_decay: 0.0001
5
+ adam_weight_decay_text_encoder: 0.001
6
+ allow_tf32: false
7
+ cache_dir: null
8
+ caption_column: prompt
9
+ center_crop: false
10
+ checkpointing_steps: 717
11
+ checkpoints_total_limit: null
12
+ class_data_dir: null
13
+ class_prompt: null
14
+ dataloader_num_workers: 0
15
+ dataset_config_name: null
16
+ dataset_name: null
17
+ do_edm_style_training: false
18
+ enable_xformers_memory_efficient_attention: false
19
+ gradient_accumulation_steps: 3
20
+ gradient_checkpointing: true
21
+ hub_model_id: null
22
+ hub_token: null
23
+ image_column: image
24
+ instance_data_dir: sushi_try
25
+ instance_prompt: a photo of sushi
26
+ learning_rate: 0.0001
27
+ local_rank: -1
28
+ logging_dir: logs
29
+ lr_num_cycles: 1
30
+ lr_power: 1.0
31
+ lr_scheduler: constant
32
+ lr_warmup_steps: 0
33
+ max_grad_norm: 1.0
34
+ max_train_steps: 500
35
+ mixed_precision: fp16
36
+ num_class_images: 100
37
+ num_train_epochs: 250
38
+ num_validation_images: 4
39
+ optimizer: AdamW
40
+ output_dir: sushi_LoRA_2
41
+ output_kohya_format: false
42
+ pretrained_model_name_or_path: stabilityai/stable-diffusion-xl-base-1.0
43
+ pretrained_vae_model_name_or_path: madebyollin/sdxl-vae-fp16-fix
44
+ prior_generation_precision: null
45
+ prior_loss_weight: 1.0
46
+ prodigy_beta3: null
47
+ prodigy_decouple: true
48
+ prodigy_safeguard_warmup: true
49
+ prodigy_use_bias_correction: true
50
+ push_to_hub: false
51
+ random_flip: false
52
+ rank: 4
53
+ repeats: 1
54
+ report_to: tensorboard
55
+ resolution: 1024
56
+ resume_from_checkpoint: null
57
+ revision: null
58
+ sample_batch_size: 4
59
+ scale_lr: false
60
+ seed: 0
61
+ snr_gamma: 5.0
62
+ text_encoder_lr: 5.0e-06
63
+ train_batch_size: 1
64
+ train_text_encoder: false
65
+ use_8bit_adam: true
66
+ use_dora: false
67
+ validation_epochs: 50
68
+ validation_prompt: null
69
+ variant: null
70
+ with_prior_preservation: false
logs/dreambooth-lora-sd-xl/events.out.tfevents.1714203341.235b5debc8b2.11877.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:071c9a1e799b0244c50f619a4ccdeea8b4c88c1726961582c6d6221274511637
3
+ size 88
logs/dreambooth-lora-sd-xl/events.out.tfevents.1714203463.235b5debc8b2.12630.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8044d25bdcac0b59a9a6c4d222f06a9579b740ea649f7840d09a759edc74afff
3
+ size 88
logs/dreambooth-lora-sd-xl/events.out.tfevents.1714203920.235b5debc8b2.14637.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4c7a34461b8c26fcce9b56cb2af2b648f35e319b37541c2fdd491dc10567984
3
+ size 83570
pytorch_lora_weights.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40e94ad6bd4fcada1439c5bb69518f3df408ccf65d6b777dd0da13a012c7786a
3
+ size 23390424