helenai commited on
Commit
b3dfd77
1 Parent(s): 7a241ba

commit files to HF hub

Browse files
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: openrail++
3
+ tags:
4
+ - stable-diffusion
5
+ - text-to-image
6
+ - openvino
7
+
8
+ ---
9
+
10
+ # OpenVINO Stable Diffusion
11
+
12
+ ## stabilityai/stable-diffusion-2-1
13
+
14
+ This repository contains the models from [stabilityai/stable-diffusion-2-1](https://huggingface.co/stabilityai/stable-diffusion-2-1) converted to
15
+ OpenVINO, for accelerated inference on CPU or Intel GPU with OpenVINO's integration into Optimum:
16
+ [optimum-intel](https://github.com/huggingface/optimum-intel#openvino). The model weights are stored with FP16
17
+ precision, which reduces the size of the model by half.
18
+
19
+ Please check out the [source model repository](https://huggingface.co/stabilityai/stable-diffusion-2-1) for more information about the model and its license.
20
+
21
+ To install the requirements for this demo, do `pip install optimum[openvino]`. This installs all the necessary dependencies,
22
+ including Transformers and OpenVINO. For more detailed steps, please see this [installation guide](https://github.com/helena-intel/optimum-intel/wiki/OpenVINO-Integration-Installation-Guide).
23
+
24
+ The simplest way to generate an image with stable diffusion takes only two lines of code, as shown below. The first line downloads the
25
+ model from the Hugging Face hub (if it has not been downloaded before) and loads it; the second line generates an image.
26
+
27
+ ```python
28
+ from optimum.intel.openvino import OVStableDiffusionPipeline
29
+
30
+ stable_diffusion = OVStableDiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-2-1")
31
+ images = stable_diffusion("a random image").images
32
+ ```
33
+
34
+ The following example code uses static shapes for even faster inference. Using larger image sizes will
35
+ require more memory and take longer to generate.
36
+
37
+ If you have an 11th generation or later Intel Core processor, you can use the integrated GPU for inference, and if you have an Intel
38
+ discrete GPU, you can use that. Add the line `stable_diffusion.to("GPU")` before `stable_diffusion.compile()` in the example below.
39
+ Model loading will take some time the first time, but will be faster after that, because the model will be cached. On GPU, for stable
40
+ diffusion only static shapes are supported at the moment.
41
+
42
+
43
+ ```python
44
+ from optimum.intel.openvino.modeling_diffusion import OVStableDiffusionPipeline
45
+
46
+ batch_size = 1
47
+ num_images_per_prompt = 1
48
+ height = 256
49
+ width = 256
50
+
51
+ # load the model and reshape to static shapes for faster inference
52
+ model_id = "stabilityai/stable-diffusion-2-1"
53
+ stable_diffusion = OVStableDiffusionPipeline.from_pretrained(model_id, compile=False)
54
+ stable_diffusion.reshape( batch_size=batch_size, height=height, width=width, num_images_per_prompt=num_images_per_prompt)
55
+ stable_diffusion.compile()
56
+
57
+ # generate image!
58
+ prompt = "a random image"
59
+ images = stable_diffusion(prompt, height=height, width=width, num_images_per_prompt=num_images_per_prompt).images
60
+ images[0].save("result.png")
61
+ ```
62
+
feature_extractor/preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 224,
4
+ "width": 224
5
+ },
6
+ "do_center_crop": true,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "feature_extractor_type": "CLIPFeatureExtractor",
12
+ "image_mean": [
13
+ 0.48145466,
14
+ 0.4578275,
15
+ 0.40821073
16
+ ],
17
+ "image_processor_type": "CLIPFeatureExtractor",
18
+ "image_std": [
19
+ 0.26862954,
20
+ 0.26130258,
21
+ 0.27577711
22
+ ],
23
+ "resample": 3,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "shortest_edge": 224
27
+ }
28
+ }
inference.py ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from optimum.intel.openvino.modeling_diffusion import OVStableDiffusionPipeline
2
+
3
+ batch_size = 1
4
+ num_images_per_prompt = 1
5
+ height = 256
6
+ width = 256
7
+
8
+ # load the model and reshape to static shapes for faster inference
9
+ model_id = "helenai/stabilityai-stable-diffusion-2-1-ov"
10
+ stable_diffusion = OVStableDiffusionPipeline.from_pretrained(model_id, compile=False)
11
+ stable_diffusion.reshape( batch_size=batch_size, height=height, width=width, num_images_per_prompt=num_images_per_prompt)
12
+ stable_diffusion.compile()
13
+
14
+ # generate image!
15
+ prompt = "a random image"
16
+ images = stable_diffusion(prompt, height=height, width=width, num_images_per_prompt=num_images_per_prompt).images
17
+ images[0].save("result.png")
18
+
model_index.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "OVStableDiffusionPipeline",
3
+ "_diffusers_version": "0.13.1",
4
+ "feature_extractor": [
5
+ "transformers",
6
+ "CLIPFeatureExtractor"
7
+ ],
8
+ "requires_safety_checker": false,
9
+ "safety_checker": [
10
+ null,
11
+ null
12
+ ],
13
+ "scheduler": [
14
+ "diffusers",
15
+ "DDIMScheduler"
16
+ ],
17
+ "text_encoder": [
18
+ "optimum",
19
+ "OVModelTextEncoder"
20
+ ],
21
+ "tokenizer": [
22
+ "transformers",
23
+ "CLIPTokenizer"
24
+ ],
25
+ "unet": [
26
+ "optimum",
27
+ "OVModelUnet"
28
+ ],
29
+ "vae_decoder": [
30
+ "optimum",
31
+ "OVModelVaeDecoder"
32
+ ]
33
+ }
scheduler/scheduler_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_class_name": "DDIMScheduler",
3
+ "_diffusers_version": "0.13.1",
4
+ "beta_end": 0.012,
5
+ "beta_schedule": "scaled_linear",
6
+ "beta_start": 0.00085,
7
+ "clip_sample": false,
8
+ "num_train_timesteps": 1000,
9
+ "prediction_type": "v_prediction",
10
+ "set_alpha_to_one": false,
11
+ "skip_prk_steps": true,
12
+ "steps_offset": 1,
13
+ "trained_betas": null
14
+ }
text_encoder/openvino_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3d5702e9aefe3f9426713201632b99005f17d77cf6f1bbfb54525a473206ffa
3
+ size 680776428
text_encoder/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|startoftext|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "!",
17
+ "unk_token": {
18
+ "content": "<|endoftext|>",
19
+ "lstrip": false,
20
+ "normalized": true,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": {
4
+ "__type": "AddedToken",
5
+ "content": "<|startoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false
10
+ },
11
+ "do_lower_case": true,
12
+ "eos_token": {
13
+ "__type": "AddedToken",
14
+ "content": "<|endoftext|>",
15
+ "lstrip": false,
16
+ "normalized": true,
17
+ "rstrip": false,
18
+ "single_word": false
19
+ },
20
+ "errors": "replace",
21
+ "model_max_length": 77,
22
+ "name_or_path": "models/stabilityai-stable-diffusion-2-1-ov/tokenizer",
23
+ "pad_token": "<|endoftext|>",
24
+ "special_tokens_map_file": "./special_tokens_map.json",
25
+ "tokenizer_class": "CLIPTokenizer",
26
+ "unk_token": {
27
+ "__type": "AddedToken",
28
+ "content": "<|endoftext|>",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false
33
+ }
34
+ }
tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
unet/openvino_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d753b9848294357fe4d0b0d818ac220a3a763b291434cd6c0fc9100c77774f7b
3
+ size 1731822168
unet/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff
 
vae_decoder/openvino_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bea41feef073b8e247dceeb0ec95d6e523a3c5acc1310882e22b6245509390d9
3
+ size 98980700
vae_decoder/openvino_model.xml ADDED
The diff for this file is too large to render. See raw diff