fffiloni commited on
Commit
84d244a
1 Parent(s): c2d3b3c

better model card

Browse files
Files changed (1) hide show
  1. train_dreambooth_lora_sdxl.py +7 -3
train_dreambooth_lora_sdxl.py CHANGED
@@ -90,7 +90,11 @@ inference: false
90
 
91
  These are LoRA adaption weights for {base_model}.
92
 
93
- The weights were trained on the concept prompt: `{prompt}` using [DreamBooth](https://dreambooth.github.io/).
 
 
 
 
94
 
95
  LoRA for the text encoder was enabled: {train_text_encoder}.
96
 
@@ -114,7 +118,7 @@ To just use the base model, you can run:
114
  import torch
115
  from diffusers import DiffusionPipeline, AutoencoderKL
116
 
117
- vae = AutoencoderKL.from_pretrained({vae_path}, torch_dtype=torch.float16)
118
 
119
  pipe = DiffusionPipeline.from_pretrained(
120
  "stabilityai/stable-diffusion-xl-base-1.0",
@@ -123,7 +127,7 @@ pipe = DiffusionPipeline.from_pretrained(
123
  )
124
 
125
  # This is where you load your trained weights
126
- pipe.load_lora_weights({repo_id})
127
 
128
  pipe.to("cuda")
129
 
 
90
 
91
  These are LoRA adaption weights for {base_model}.
92
 
93
+ The weights were trained on the concept prompt:
94
+
95
+ `{prompt}`
96
+
97
+ Use this keyword to trigger your custom model in your prompts :)
98
 
99
  LoRA for the text encoder was enabled: {train_text_encoder}.
100
 
 
118
  import torch
119
  from diffusers import DiffusionPipeline, AutoencoderKL
120
 
121
+ vae = AutoencoderKL.from_pretrained('{vae_path}', torch_dtype=torch.float16)
122
 
123
  pipe = DiffusionPipeline.from_pretrained(
124
  "stabilityai/stable-diffusion-xl-base-1.0",
 
127
  )
128
 
129
  # This is where you load your trained weights
130
+ pipe.load_lora_weights('{repo_id}')
131
 
132
  pipe.to("cuda")
133