update text

app.py

@@ -11,6 +11,7 @@ import os
 import backoff
 from functools import lru_cache
 from huggingface_hub import list_models, ModelFilter, login
+import copy
 
 os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"
 
@@ -127,18 +128,18 @@ def predict_subset(model_id, token):
 
 
 with gr.Blocks() as demo:
-    with gr.Tab("Random image gallery"):
+    with gr.Tab("Random Image Gallery"):
         gr.Markdown(
-            """## Random image gallery
-        This is a random image gallery. 
-        You can refresh the images by clicking the refresh button."""
+        """## Random image gallery
+        This tab allows you to explore images in your ARCH collection. You can refresh the images by clicking the refresh button.
+        **Please note** not all images will be displayed as some images may not available via the original URLS anymore."""
         )
         button = gr.Button("Refresh")
         gallery = gr.Gallery().style(grid=9, height="1400")
         button.click(return_random_sample, [], [gallery])
-    with gr.Tab("image search"):
+    with gr.Tab("Image Search"):
         gr.Markdown(
-            """## Image search 
+        """## Image search 
         You can search for images by entering a search term and clicking the search button.
         You can also change the number of images to be returned.
         This model uses the [clip-ViT-B-16](https://huggingface.co/sentence-transformers/clip-ViT-B-16) model to embed your images and search term"""
@@ -148,27 +149,37 @@ with gr.Blocks() as demo:
         button = gr.Button("search")
         gallery = gr.Gallery().style(grid=3)
         button.click(get_nearest_k_examples, [text, k], [gallery])
-        gr.Markdown("""### More info
-        ![https://raw.githubusercontent.com/UKPLab/sentence-transformers/master/docs/img/ImageSearch.png](https://raw.githubusercontent.com/UKPLab/sentence-transformers/master/docs/img/ImageSearch.png)""")
-    # with gr.Tab("Export for label studio"):
-    #     button = gr.Button("Export")
-    #     dataset2 = copy.deepcopy(dataset)
-    #     # dataset2 = dataset2.remove_columns('image')
-    #     # dataset2 = dataset2.rename_column("url", "image")
-    #     csv = dataset2.to_csv("label_studio.csv")
-    #     csv_file = gr.File("label_studio.csv")
-    #     button.click(dataset.save_to_disk, [], [csv_file])
-    with gr.Tab("predict"):
+        # gr.Markdown(
+        #     """### More info
+        # ![https://raw.githubusercontent.com/UKPLab/sentence-transformers/master/docs/img/ImageSearch.png](https://raw.githubusercontent.com/UKPLab/sentence-transformers/master/docs/img/ImageSearch.png)"""
+        # )
+ 
+    with gr.Tab("Image Classification Model Tester"):
         gr.Markdown(
-            """## Image classification model tester
-        You can use this to test out [image classification models](https://huggingface.co/models?pipeline_tag=image-classification) on the Hugging Face Hub. """
+        """## Image classification model tester
+        You can use this to test out [image classification models](https://huggingface.co/models?pipeline_tag=image-classification) on the Hugging Face Hub:
+        - To use this tab you will need to have a Hugging Face account and a valid token. 
+        - You can get a token from your [Hugging Face account page](https://huggingface.co/settings/token). 
+        - Input this token into the token box and then input a valid image classification model id from the Hub. For example `microsoft/resnet-50`
+        
+        This tab uses Hugging Face's [Inference API](https://huggingface.co/docs/api-inference/index) to make predictions. It will randomly select 10 images from your dataset and make predictions on them using your chosen model.
+        **Please note** the predictions will take some time since the model needs to be loaded for inference first. If you make a second batch of prediction using the same model the predictions should be quicker."""
         )
         token = gr.Textbox(label="token", type="password")
-        model_id = gr.Textbox(label="model_id")
+        model_id = gr.Textbox(label="model_id", value="microsoft/resnet-50")
         button = gr.Button("predict")
         gr.Markdown("## Results")
         plot = gr.BarPlot(x="labels", y="freqs", width=600, height=400, vertical=False)
         gallery = gr.Gallery()
         button.click(predict_subset, [model_id, token], [gallery, plot])
-
+    with gr.Tab("Export to Label Studio format"):
+        gr.Markdown("""
+        ## Export to Label Studio format
+        This will export the current dataset to a csv file in Label Studio format. You can then import this into Label Studio to label your images.""")
+        dataset2 = copy.deepcopy(dataset)
+        dataset2 = dataset2.remove_columns('image')
+        dataset2 = dataset2.rename_column("url", "image")
+        csv = dataset2.to_csv("label_studio.csv")
+        csv_file = gr.File("label_studio.csv")
+        button.click(dataset.save_to_disk, [], [csv_file])
 demo.launch(enable_queue=True)