Spaces:
Runtime error
Runtime error
ms-analytics
commited on
Commit
•
b6c5ec2
1
Parent(s):
e98358f
Update description.md
Browse files- description.md +30 -1
description.md
CHANGED
@@ -1 +1,30 @@
|
|
1 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
OFA (Open-domain Feature-based Architecture) is an open-source, unified sequence-to-sequence learning framework developed by Open-domain Feature-based Architecture System (OFA-Sys).
|
2 |
+
|
3 |
+
It was released in ICML 2022[1].
|
4 |
+
|
5 |
+
OFA is designed to unify modalities and tasks such as
|
6 |
+
image captioning,
|
7 |
+
o VQA, visual grounding,
|
8 |
+
o text-to-image generation,
|
9 |
+
o text classification,
|
10 |
+
o text generation,
|
11 |
+
o image classification.
|
12 |
+
|
13 |
+
It provides step-by-step instructions for pretraining and finetuning with checkpoints available from both official repository and Hugging Face. [2]
|
14 |
+
|
15 |
+
OFA also provides online demos for interactive experience with its pretrained and finetuned models. [3]
|
16 |
+
|
17 |
+
Furthermore, it offers Colab notebooks to understand the procedures better. OFA is even more powerful with the release of MuE (Multi-modal Embeddings) which accelerates OFA with little performance degradation. [1]
|
18 |
+
|
19 |
+
It also provides OFA-OCR for Chinese text recognition [1] and MMSpeech, an ASR pre-training method based on OFA. [1]
|
20 |
+
|
21 |
+
Moreover, a Chinese version of OFA is released. [1]
|
22 |
+
|
23 |
+
|
24 |
+
References:
|
25 |
+
|
26 |
+
[1] https://github.com/OFA-Sys/OFA
|
27 |
+
|
28 |
+
[2] https://github.com/OFA-Sys
|
29 |
+
|
30 |
+
[3] https://github.com/OFA-Sys/OFA/blob/main/colab.md
|