Improve model card
#1
by
nielsr
HF staff
- opened
README.md
CHANGED
|
@@ -2,10 +2,18 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
language:
|
| 4 |
- en
|
|
|
|
|
|
|
|
|
|
| 5 |
---
|
|
|
|
| 6 |
## Open-MAGVIT2: Democratizing Autoregressive Visual Generation
|
| 7 |
|
| 8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
Until now, VQGAN, the initial tokenizer is still acting an indispensible role in mainstream tasks, especially autoregressive visual generation. Limited by the bottleneck of the size of codebook and the utilization of code, the capability of AR generation with VQGAN is underestimated.
|
| 11 |
|
|
@@ -18,3 +26,7 @@ ImageNet 128 × 128:
|
|
| 18 |
|
| 19 |
ImageNet 256 × 256:
|
| 20 |
- Model [ImageNet_256_Base.ckpt](https://huggingface.co/TencentARC/Open-MAGVIT2/blob/main/imagenet_256_B.ckpt)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
language:
|
| 4 |
- en
|
| 5 |
+
pipeline_tag: image-feature-extraction
|
| 6 |
+
tags:
|
| 7 |
+
- image-to-image
|
| 8 |
---
|
| 9 |
+
|
| 10 |
## Open-MAGVIT2: Democratizing Autoregressive Visual Generation
|
| 11 |
|
| 12 |
+
Code: https://github.com/TencentARC/Open-MAGVIT2
|
| 13 |
+
|
| 14 |
+
Paper: https://huggingface.co/papers/2409.04410
|
| 15 |
+
|
| 16 |
+
## Introduction
|
| 17 |
|
| 18 |
Until now, VQGAN, the initial tokenizer is still acting an indispensible role in mainstream tasks, especially autoregressive visual generation. Limited by the bottleneck of the size of codebook and the utilization of code, the capability of AR generation with VQGAN is underestimated.
|
| 19 |
|
|
|
|
| 26 |
|
| 27 |
ImageNet 256 × 256:
|
| 28 |
- Model [ImageNet_256_Base.ckpt](https://huggingface.co/TencentARC/Open-MAGVIT2/blob/main/imagenet_256_B.ckpt)
|
| 29 |
+
|
| 30 |
+
## Usage
|
| 31 |
+
|
| 32 |
+
Refer to the Github repository which includes [scripts](https://github.com/TencentARC/Open-MAGVIT2/tree/main/scripts) for training, evaluation and inference.
|