File size: 1,771 Bytes
0c4969a
 
 
 
 
 
 
 
 
14774a0
9e69087
14774a0
 
 
 
 
 
 
 
 
 
3699b97
14774a0
 
3699b97
 
 
 
 
14774a0
 
9e69087
 
 
14774a0
 
 
9e69087
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14774a0
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
64
65
---
library_name: diffusers
base_model: runwayml/stable-diffusion-v1-5
tags:
- text-to-image
license: creativeml-openrail-m
inference: false
---

## yujiepan/dreamshaper-8-lcm-openvino

This model applies `latent-consistency/lcm-lora-sdv1-5` to base model `Lykon/dreamshaper-8`, and is converted to OpenVINO format.

#### Usage

```python
from optimum.intel.openvino.modeling_diffusion import OVStableDiffusionPipeline
pipeline = OVStableDiffusionPipeline.from_pretrained(
    'yujiepan/dreamshaper-8-lcm-openvino',
    device='CPU',
)
prompt = 'cute dog typing at a laptop, 4k, details'
images = pipeline(prompt=prompt, num_inference_steps=8, guidance_scale=1.0).images
```

![output image](./assets/cute-dog-typing-at-a-laptop-4k-details.png)




#### TODO

- The fp16 base model is converted to openvino in fp32, which is unnecessary.

#### Scripts

The model is generated by the following codes:

```python
import torch
from diffusers import AutoPipelineForText2Image, LCMScheduler
from optimum.intel.openvino.modeling_diffusion import OVStableDiffusionPipeline

base_model_id = "Lykon/dreamshaper-8"
adapter_id = "latent-consistency/lcm-lora-sdv1-5"
save_torch_folder = './dreamshaper-8-lcm'
save_ov_folder = './dreamshaper-8-lcm-openvino'

torch_pipeline = AutoPipelineForText2Image.from_pretrained(
    base_model_id, torch_dtype=torch.float16, variant="fp16")
torch_pipeline.scheduler = LCMScheduler.from_config(
    torch_pipeline.scheduler.config)
# load and fuse lcm lora
torch_pipeline.load_lora_weights(adapter_id)
torch_pipeline.fuse_lora()
torch_pipeline.save_pretrained(save_torch_folder)

ov_pipeline = OVStableDiffusionPipeline.from_pretrained(
    save_torch_folder,
    device='CPU',
    export=True,
)
ov_pipeline.save_pretrained(save_ov_folder)
```