From e9c6b1e75270bccefa35d7310ba226b6ee554e87 Mon Sep 17 00:00:00 2001 From: sangbumchoi Date: Tue, 20 Aug 2024 15:06:11 +0000 Subject: [PATCH] add docs --- docs/source/en/model_doc/vitpose.md | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/docs/source/en/model_doc/vitpose.md b/docs/source/en/model_doc/vitpose.md index 12aeee91d4d70f..0e8855ae7a119c 100644 --- a/docs/source/en/model_doc/vitpose.md +++ b/docs/source/en/model_doc/vitpose.md @@ -26,7 +26,10 @@ The original code can be found [here](https://github.com/ViTAE-Transformer/ViTPo ## Usage Tips -The current model utilizes a 2-step inference pipeline. The first step involves placing a bounding box around the region corresponding to the person. After that, the second step uses ViTPose to predict the keypoints. +- To enable MoE(Mixture or Expert) function in backbone, user have to give appropriate input indices into the backbone model. + However, it is not used in default parameters. +- The current model utilizes a 2-step inference pipeline. The first step involves placing a bounding box around the region corresponding to the person. + After that, the second step uses ViTPose to predict the keypoints. ```py >>> import torch