jbilcke-hf HF staff commited on
Commit
062f791
1 Parent(s): 6cd7de7

saving checkpoint-500

Browse files
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ ---
3
+ base_model: stabilityai/stable-diffusion-xl-base-1.0
4
+ instance_prompt: starfield-style
5
+ tags:
6
+ - stable-diffusion-xl
7
+ - stable-diffusion-xl-diffusers
8
+ - text-to-image
9
+ - diffusers
10
+ - lora
11
+ inference: false
12
+ datasets:
13
+ - jbilcke-hf/starfield
14
+ ---
15
+
16
+ # LoRA DreamBooth - jbilcke-hf/sdxl-starfield
17
+ ## MODEL IS CURRENTLY TRAINING ...
18
+ Last checkpoint saved: checkpoint-500
19
+ These are LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0 trained on @fffiloni's SD-XL trainer.
20
+ The weights were trained on the concept prompt:
21
+ ```
22
+ starfield-style
23
+ ```
24
+ Use this keyword to trigger your custom model in your prompts.
25
+ LoRA for the text encoder was enabled: False.
26
+ Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.
27
+ ## Usage
28
+ Make sure to upgrade diffusers to >= 0.19.0:
29
+ ```
30
+ pip install diffusers --upgrade
31
+ ```
32
+ In addition make sure to install transformers, safetensors, accelerate as well as the invisible watermark:
33
+ ```
34
+ pip install invisible_watermark transformers accelerate safetensors
35
+ ```
36
+ To just use the base model, you can run:
37
+ ```python
38
+ import torch
39
+ from diffusers import DiffusionPipeline, AutoencoderKL
40
+ device = "cuda" if torch.cuda.is_available() else "cpu"
41
+ vae = AutoencoderKL.from_pretrained('madebyollin/sdxl-vae-fp16-fix', torch_dtype=torch.float16)
42
+ pipe = DiffusionPipeline.from_pretrained(
43
+ "stabilityai/stable-diffusion-xl-base-1.0",
44
+ vae=vae, torch_dtype=torch.float16, variant="fp16",
45
+ use_safetensors=True
46
+ )
47
+ pipe.to(device)
48
+ # This is where you load your trained weights
49
+ specific_safetensors = "pytorch_lora_weights.safetensors"
50
+ lora_scale = 0.9
51
+ pipe.load_lora_weights(
52
+ 'jbilcke-hf/sdxl-starfield',
53
+ weight_name = specific_safetensors,
54
+ # use_auth_token = True
55
+ )
56
+ prompt = "A majestic starfield-style jumping from a big stone at night"
57
+ image = pipe(
58
+ prompt=prompt,
59
+ num_inference_steps=50,
60
+ cross_attention_kwargs={"scale": lora_scale}
61
+ ).images[0]
62
+ ```
checkpoint-500/optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a818d36da5790ae08e25e8268f2e7359f120158d9f68dfdc5bd05ac1974018e8
3
+ size 14990599
checkpoint-500/pytorch_lora_weights.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d122363dc32e1b79843c9686d2424c61e7de6758ced7709398f10aca46653752
3
+ size 23401064
checkpoint-500/random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35af3a6036d18c82a86944172d52de6ee8448bcccdc5f5ee8fc6e92660361208
3
+ size 14599
checkpoint-500/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3f196a54202bb4ba1220e8c59f42f9cda0702d68ea83147d814c2fb2f36b8f2
3
+ size 557
checkpoint-500/scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62783ff6f0ae2c2ccc177b097e430c03940c5e70c38a191df273c5cf7ab1227c
3
+ size 563