search

xtreme1-io / xtreme1

Xtreme1 is an all-in-one data labeling and annotation platform for multimodal data training and supports 3D LiDAR point cloud, image, and LLM.
https://www.basic.ai
Apache License 2.0
853 stars 141 forks source link

Upload dataset #217

Open aj148 opened 6 months ago

aj148 commented 6 months ago

When uploading according to this doc, I had some trouble

According to the form file upload format (Content-Type: multipart/form-data), use the PUT method to request the upload address to directly upload the compressed package to the object storage service;

Can you give an idea of how this requests looks like?

also for step 4, how do you know what the dataset id should be?

jaggerwang commented 6 months ago

You can using a s3 compatible client sdk (recommend minio) to construct this request, as we use minio as the object storage service. As the dataset id, you can create a new dataset, or query an existing one to get the id.

aj148 commented 6 months ago

do we have a bucket that exits already within xtreme1? also this looks different than what the sdk has for import

jaggerwang commented 6 months ago

Sorry, it's different to upload with minio sdk. We have already and an example for uploading through presigend url in the api doc, https://apifox.com/apidoc/shared-2f8468c3-a269-4e3d-9df7-c589bda69450/api-153037674 . image

VeeranjaneyuluToka commented 5 months ago

Hi, Is there away to annotate without uploading the dataset? I mean i wanted to have local copy (in my host machine) and annotate without uploading at all.

jaggerwang commented 5 months ago

It is currently not supported, and the dataset management is also part of the Xtreme1 function. What you expect is a pure tool. However, because Xtreme1 is privately deployed in your local machine, so there is no data security problem.