Update README.md
Browse files
README.md
CHANGED
|
@@ -46,11 +46,15 @@ source venv/bin/activate
|
|
| 46 |
pip install -e .
|
| 47 |
```
|
| 48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
Run zero-shot classification with:
|
| 50 |
```shell
|
| 51 |
python inference.py --run-classification \
|
| 52 |
--model-name "Llama3-MS-CLIP-Base" \
|
| 53 |
-
--images "path/to/
|
| 54 |
--class-names "class 1" "class 2" "class 3"
|
| 55 |
```
|
| 56 |
|
|
@@ -58,7 +62,7 @@ Run text-to-image retrieval with:
|
|
| 58 |
```shell
|
| 59 |
python inference.py --run_retrieval\
|
| 60 |
--model-name "Llama3-MS-CLIP-Base" \
|
| 61 |
-
--images "path/to/
|
| 62 |
--query "Your query text" \
|
| 63 |
--top-k 5 # Number of retrieved images
|
| 64 |
```
|
|
|
|
| 46 |
pip install -e .
|
| 47 |
```
|
| 48 |
|
| 49 |
+
### Infernece
|
| 50 |
+
|
| 51 |
+
We provide an easy-to-use inference script that automatically downloads the weights from Hugging Face. You just need to provide a path to folder with Sentinel-2 L2A files (all 12 bands) and `class_names` or a `query`.
|
| 52 |
+
|
| 53 |
Run zero-shot classification with:
|
| 54 |
```shell
|
| 55 |
python inference.py --run-classification \
|
| 56 |
--model-name "Llama3-MS-CLIP-Base" \
|
| 57 |
+
--images "path/to/sentinel2_l2a_files/" \
|
| 58 |
--class-names "class 1" "class 2" "class 3"
|
| 59 |
```
|
| 60 |
|
|
|
|
| 62 |
```shell
|
| 63 |
python inference.py --run_retrieval\
|
| 64 |
--model-name "Llama3-MS-CLIP-Base" \
|
| 65 |
+
--images "path/to/sentinel2_l2a_files/" \
|
| 66 |
--query "Your query text" \
|
| 67 |
--top-k 5 # Number of retrieved images
|
| 68 |
```
|