Spaces:
Running
Running
Clement Vachet
commited on
Commit
·
25e9ef7
1
Parent(s):
5eda9a8
doc: add menu and deployment section
Browse files
README.md
CHANGED
|
@@ -14,18 +14,32 @@ short_description: Object detection via Gradio
|
|
| 14 |
|
| 15 |
Aim: AI-driven object detection (on COCO image dataset)
|
| 16 |
|
| 17 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
-
###
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
> python detect_torch.py
|
| 21 |
|
| 22 |
-
### 2. Use of transformers library
|
| 23 |
> python detect_transformers.py
|
| 24 |
|
| 25 |
-
### 3. Use of HuggingFace pipeline library
|
| 26 |
> python detect_pipeline.py
|
| 27 |
|
| 28 |
-
##
|
| 29 |
Use of Gradio library for web interface
|
| 30 |
|
| 31 |
Command line:
|
|
@@ -33,11 +47,11 @@ Command line:
|
|
| 33 |
|
| 34 |
<b>Note:</b> The Gradio app should now be accessible at http://localhost:7860
|
| 35 |
|
| 36 |
-
##
|
| 37 |
|
| 38 |
<b>Note:</b> Use of existing Gradio server (running locally, in a Docker container, or in the cloud as a HuggingFace space or AWS)
|
| 39 |
|
| 40 |
-
### 1. Creation of docker container
|
| 41 |
|
| 42 |
Command lines:
|
| 43 |
> sudo docker build -t gradio-app .
|
|
@@ -46,6 +60,21 @@ Command lines:
|
|
| 46 |
|
| 47 |
The Gradio app should now be accessible at http://localhost:7860
|
| 48 |
|
| 49 |
-
### 2. Direct inference via API
|
| 50 |
Command line:
|
| 51 |
> python inference_API.py
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
Aim: AI-driven object detection (on COCO image dataset)
|
| 16 |
|
| 17 |
+
Machine learning models:
|
| 18 |
+
- facebook/detr-resnet-50,
|
| 19 |
+
- facebook/detr-resnet-101,
|
| 20 |
+
- hustvl/yolos-tiny,
|
| 21 |
+
- hustvl/yolos-small
|
| 22 |
|
| 23 |
+
### <b>Table of contents:</b>
|
| 24 |
+
- [Execution via command line](#1-execution-via-command-line)
|
| 25 |
+
- [Execution via User Interface ](#2-execution-via-user-interface)
|
| 26 |
+
- [Execution via Gradio client API](#3-execution-via-gradio-client-api)
|
| 27 |
+
- [Deployment on Hugging Face](#4-deployment-on-hugging-face)
|
| 28 |
+
- [Deployment on Docker Hub](#5-deployment-on-docker-hub)
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
## 1. Execution via command line
|
| 32 |
+
|
| 33 |
+
### 1.1. Use of torch library
|
| 34 |
> python detect_torch.py
|
| 35 |
|
| 36 |
+
### 1.2. Use of transformers library
|
| 37 |
> python detect_transformers.py
|
| 38 |
|
| 39 |
+
### 1.3. Use of HuggingFace pipeline library
|
| 40 |
> python detect_pipeline.py
|
| 41 |
|
| 42 |
+
## 2. Execution via User Interface
|
| 43 |
Use of Gradio library for web interface
|
| 44 |
|
| 45 |
Command line:
|
|
|
|
| 47 |
|
| 48 |
<b>Note:</b> The Gradio app should now be accessible at http://localhost:7860
|
| 49 |
|
| 50 |
+
## 3. Execution via Gradio client API
|
| 51 |
|
| 52 |
<b>Note:</b> Use of existing Gradio server (running locally, in a Docker container, or in the cloud as a HuggingFace space or AWS)
|
| 53 |
|
| 54 |
+
### 3.1. Creation of docker container
|
| 55 |
|
| 56 |
Command lines:
|
| 57 |
> sudo docker build -t gradio-app .
|
|
|
|
| 60 |
|
| 61 |
The Gradio app should now be accessible at http://localhost:7860
|
| 62 |
|
| 63 |
+
### 3.2. Direct inference via API
|
| 64 |
Command line:
|
| 65 |
> python inference_API.py
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
## 4. Deployment on Hugging Face
|
| 69 |
+
|
| 70 |
+
This web application is available on Hugging Face, via a Gradio space
|
| 71 |
+
|
| 72 |
+
URL: https://huggingface.co/spaces/cvachet/object_detection_gradio
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
## 5. Deployment on Docker Hub
|
| 76 |
+
|
| 77 |
+
This web application is available as a container on Docker Hub
|
| 78 |
+
|
| 79 |
+
URL: https://hub.docker.com/r/cvachet/object-detection-gradio
|
| 80 |
+
|