zeekay commited on
Commit
d69845c
·
verified ·
1 Parent(s): 4afbf96

Update model card: add zen/zenlm tags, fix branding

Browse files
Files changed (1) hide show
  1. README.md +29 -40
README.md CHANGED
@@ -1,48 +1,38 @@
1
  ---
2
- library_name: diffusers
3
- pipeline_tag: text-to-video
4
  tags:
5
- - diffusers
6
- - text-to-video
7
- - video-generation
8
- - world-model
9
  - zen
10
  - zenlm
11
  - hanzo
12
- license: apache-2.0
 
 
 
 
13
  ---
14
 
15
  # Zen World
16
 
17
- **Zen LM by Hanzo AI** — Large-scale world simulation and video generation model. Generates temporally consistent video from text descriptions.
 
 
18
 
19
- ## Specs
20
 
21
- | Property | Value |
22
- |----------|-------|
23
- | Parameters | 14B |
24
- | Architecture | Zen MoDE (Mixture of Distilled Experts) |
25
- | Task | World Simulation / Text-to-Video |
26
- | Resolution | 720p |
27
 
28
- ## Usage
29
 
30
  ```python
31
  from diffusers import AutoPipelineForText2Video
32
  import torch
33
 
34
- pipe = AutoPipelineForText2Video.from_pretrained(
35
- "zenlm/zen-world",
36
- torch_dtype=torch.bfloat16,
37
- )
38
  pipe = pipe.to("cuda")
39
 
40
- video_frames = pipe(
41
- "A bustling city square with people walking and pigeons flying",
42
- num_frames=81,
43
- height=480,
44
- width=832,
45
- ).frames[0]
46
  ```
47
 
48
  ## API Access
@@ -50,24 +40,23 @@ video_frames = pipe(
50
  ```python
51
  from openai import OpenAI
52
 
53
- client = OpenAI(
54
- base_url='https://api.hanzo.ai/v1',
55
- api_key='your-api-key',
56
- )
57
-
58
- response = client.chat.completions.create(
59
- model='zen-world',
60
- messages=[{"role": "user", "content": "Generate a world simulation of a forest in autumn"}],
61
  )
62
- print(response.choices[0].message.content)
63
  ```
64
 
65
- Get your API key at [console.hanzo.ai](https://console.hanzo.ai) — $5 free credit on signup.
 
 
 
 
 
 
66
 
67
  ## License
68
 
69
  Apache 2.0
70
-
71
- ---
72
-
73
- *Zen LM is developed by [Hanzo AI](https://hanzo.ai) — Frontier AI infrastructure.*
 
1
  ---
2
+ language: en
3
+ license: apache-2.0
4
  tags:
 
 
 
 
5
  - zen
6
  - zenlm
7
  - hanzo
8
+ - text-to-video
9
+ - world-model
10
+ - video-generation
11
+ pipeline_tag: text-to-video
12
+ library_name: diffusers
13
  ---
14
 
15
  # Zen World
16
 
17
+ World simulation model for interactive environment generation.
18
+
19
+ ## Overview
20
 
21
+ Built on **Zen MoDE (Mixture of Distilled Experts)** architecture with 13B parameters.
22
 
23
+ Developed by [Hanzo AI](https://hanzo.ai) and the [Zoo Labs Foundation](https://zoo.ngo).
 
 
 
 
 
24
 
25
+ ## Quick Start
26
 
27
  ```python
28
  from diffusers import AutoPipelineForText2Video
29
  import torch
30
 
31
+ model_id = "zenlm/zen-world"
32
+ pipe = AutoPipelineForText2Video.from_pretrained(model_id, torch_dtype=torch.bfloat16)
 
 
33
  pipe = pipe.to("cuda")
34
 
35
+ video_frames = pipe("A drone flying over a tropical coastline at golden hour").frames[0]
 
 
 
 
 
36
  ```
37
 
38
  ## API Access
 
40
  ```python
41
  from openai import OpenAI
42
 
43
+ client = OpenAI(base_url="https://api.hanzo.ai/v1", api_key="your-api-key")
44
+ response = client.images.generate(
45
+ model="zen-world",
46
+ prompt="A drone flying over a tropical coastline at golden hour",
47
+ size="1280x720",
 
 
 
48
  )
49
+ print(response.data[0].url)
50
  ```
51
 
52
+ ## Model Details
53
+
54
+ | Attribute | Value |
55
+ |-----------|-------|
56
+ | Parameters | 13B |
57
+ | Architecture | Zen MoDE |
58
+ | License | Apache 2.0 |
59
 
60
  ## License
61
 
62
  Apache 2.0