ahatamiz commited on
Commit
ede60d3
1 Parent(s): 29ac3ef

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +49 -5
README.md CHANGED
@@ -1,5 +1,49 @@
1
- ---
2
- license: other
3
- license_name: nvclv1
4
- license_link: LICENSE
5
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: nvclv1
4
+ license_link: LICENSE
5
+ datasets:
6
+ - ILSVRC/imagenet-1k
7
+ pipeline_tag: image-classification
8
+ ---
9
+
10
+
11
+ [**MambaVision: A Hybrid Mamba-Transformer Vision Backbone**](https://arxiv.org/abs/2407.08083).
12
+
13
+ ### Model Overview
14
+
15
+ We introduce a novel mixer block by creating a symmetric path without SSM to enhance the modeling of global context. MambaVision has a hierarchical architecture that employs both self-attention and mixer blocks.
16
+
17
+
18
+ ### Model Performance
19
+
20
+ MambaVision demonstrates a strong performance by achieving a new SOTA Pareto-front in
21
+ terms of Top-1 accuracy and throughput.
22
+
23
+ <p align="center">
24
+ <img src="https://github.com/NVlabs/MambaVision/assets/26806394/79dcf841-3966-4b77-883d-76cd5e1d4320" width=42% height=42%
25
+ class="center">
26
+ </p>
27
+
28
+
29
+ ### Model Usage
30
+
31
+ You must first login into HuggingFace to pull the model:
32
+
33
+ ```Bash
34
+ huggingface-cli login
35
+ ```
36
+
37
+ The model can be simply used according to:
38
+
39
+ ```Python
40
+ access_token = "<YOUR ACCESS TOKEN"
41
+ model = AutoModel.from_pretrained("nvidia/MambaVision-T-2K", trust_remote_code=True)
42
+ ```
43
+
44
+
45
+ ### License:
46
+
47
+ [NVIDIA Source Code License-NC](https://huggingface.co/nvidia/MambaVision-T2-1K/blob/main/LICENSE)
48
+
49
+