EiffL commited on
Commit
423e97e
·
verified ·
1 Parent(s): 684a222

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +79 -3
README.md CHANGED
@@ -3,7 +3,83 @@ tags:
3
  - model_hub_mixin
4
  - pytorch_model_hub_mixin
5
  ---
 
6
 
7
- This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
8
- - Library: [More Information Needed]
9
- - Docs: [More Information Needed]
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  - model_hub_mixin
4
  - pytorch_model_hub_mixin
5
  ---
6
+ # AION-Base: Astronomical Omnimodal Network
7
 
8
+ [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
9
+ [![GitHub](https://img.shields.io/badge/GitHub-PolymathicAI%2FAION-blue)](https://github.com/PolymathicAI/AIO
10
+ N)
11
+ [![Paper](https://img.shields.io/badge/Paper-arXiv-red)](https://arxiv.org/abs/XXXX.XXXXX)
12
+
13
+ **AION-Base** is a 300M parameter large omnimodal model specifically designed for astronomical surveys. It
14
+ integrates 39 distinct astronomical data types and enables adaptation to a wide range of astronomical tasks
15
+ through multimodal masked modeling.
16
+
17
+ ## Model Details
18
+
19
+ - **Architecture**: Encoder-Decoder Transformer (12 blocks each, 768 dim, 12 heads)
20
+ - **Parameters**: 300M
21
+ - **Training**: Multimodal Masked Modeling (4M) on astronomical survey data
22
+ - **Modalities**: 39 data types including imaging, spectra, catalogs, and photometry
23
+
24
+ ## Quick Start
25
+
26
+ ```python
27
+ from aion import AION
28
+
29
+ # Load the pretrained model
30
+ model = AION.from_pretrained('polymathic-ai/aion-base')
31
+
32
+ # Your astronomical analysis begins here!
33
+ ```
34
+
35
+ ## Supported Data Types
36
+
37
+ AION-Base processes data from major astronomical surveys:
38
+
39
+ - Imaging: Legacy Survey, HSC Wide
40
+ - Spectra: SDSS, DESI
41
+ - Catalog: Legacy Survey entries
42
+ - Gaia: BP/RP spectra, parallax, coordinates, photometry
43
+ - Photometry: Legacy Survey (g,r,i,z + WISE), HSC (g,r,i,z,y)
44
+ - Shape: Ellipticity and morphological parameters
45
+
46
+ ## Installation
47
+
48
+ ```bash
49
+ pip install aion
50
+ ```
51
+
52
+ For development:
53
+ ```bash
54
+ pip install aion[torch,dev]
55
+ ```
56
+
57
+ ## Resources
58
+
59
+ - GitHub: https://github.com/PolymathicAI/AION
60
+ - Documentation: https://polymathic-ai.github.io/AION/
61
+ - Tutorial: https://colab.research.google.com/github/PolymathicAI/AION/blob/main/notebooks/Tutorial.ipynb
62
+
63
+ ## Model Variants
64
+
65
+ | Variant | Parameters | Encoder/Decoder | Model Dim |
66
+ |---------|------------|-----------------|-----------|
67
+ | Base | 300M | 12/12 blocks | 768 |
68
+ | Large | 800M | 24/24 blocks | 1024 |
69
+ | XLarge | 3B | 24/24 blocks | 2048 |
70
+
71
+ ## License
72
+
73
+ MIT License - see https://github.com/PolymathicAI/AION/blob/main/LICENSE for details.
74
+
75
+ ## Citation
76
+
77
+ @article{aion2024,
78
+ title={AION: Astronomical Omnimodal Network},
79
+ author={...},
80
+ journal={arXiv preprint arXiv:XXXX.XXXXX},
81
+ year={2025}
82
+ }
83
+
84
+ ---
85
+ Built with ❤ for the astronomical community by https://polymathic-ai.org/