docs: added colab link and fize model card links

README.md

@@ -4,9 +4,11 @@ license: apache-2.0
 
 # Model Card for granite-geospatial-flood-detection-uki
 
+[<b><i>>>Try it on Colab<<</i></b>](https://colab.research.google.com/github/ibm-granite/geospatial/blob/main/uki-flooddetection/notebooks/1_getting_started.ipynb)
+
 This is a fine-tuned geospatial foundation model for detecting flood and surface water in the United Kingdom and Ireland using multispectral and synthetic aperture radar (SAR) satellite imagery. 
 
-The model predicts areas of water from Sentinel-2 and Sentinel-1 SAR imagery and was trained on flood events from the United Kingdom and Ireland, recorded by the [Copernicus Emergency Management service (CEMS)](https://emergency.copernicus.eu/). Please see the [model description](#model-description) and [training Data](#training-data) below for more details. The model was fine-tuned from the [granite-geospatial-uki model]((https://huggingface.co/ibm-granite/granite-geospatial-uki), which has also been released. 
+The model predicts areas of water from Sentinel-2 and Sentinel-1 SAR imagery and was trained on flood events from the United Kingdom and Ireland, recorded by the [Copernicus Emergency Management service (CEMS)](https://emergency.copernicus.eu/). Please see the [model description](#model-description) and [training Data](#training-data) below for more details. The model was fine-tuned from the [granite-geospatial-uki model](https://huggingface.co/ibm-granite/granite-geospatial-uki), which has also been released. 
 
 You can use granite-geospatial-flood-detection-uki to run inference and detect flood events. Our experiments have shown that while granite-geospatial-flood-detection-uki works best in the UK and Ireland, the base model granite-geospatial-uki can also be successfully fine-tuned for flood detection in other locations. Please see the [granite-geospatial-uki model card](https://huggingface.co/ibm-granite/granite-geospatial-uki) for details.
 
@@ -20,7 +22,7 @@ You can use granite-geospatial-flood-detection-uki to run inference and detect f
 
 1. [Inference notebook](https://github.com/ibm-granite/geospatial/blob/main/uki-flooddetection/notebooks/1_getting_started.ipynb), for running the granite-geospatial-flood-detection-uki model.
 
-2. [Fine-tuning](https://github.com/ibm-granite/geospatial/blob/main/uki-flooddetection/notebooks/2_fine_tuning.ipynb) notebooks for fine tuning on other locations using the base model [granite-geospatial-uki](./granite-geospatial-uki-model-card.md)
+2. [Fine-tuning](https://github.com/ibm-granite/geospatial/blob/main/uki-flooddetection/notebooks/2_fine_tuning.ipynb) notebooks for fine tuning on other locations using the base model [granite-geospatial-uki](https://huggingface.co/ibm-granite/granite-geospatial-uki)
 
 
 ## Model Description
@@ -46,7 +48,7 @@ The UK and Ireland data consisted of 69, 512x512 pixel images, from 6 separate f
 
 - The segmented flood outlines and permanent water bodies were taken from [Copernicus EMS](https://emergency.copernicus.eu/) and all water (flood and permanent water) was considered the same for the purpose of training. 
 - Sentinel-2 L2 surface reflectance values and the cloud mask were obtained from [Sentinel Hub](https://www.sentinel-hub.com/).
-- Sentinel-1 SAR VV and VH backscatter ($\sigma_0$) was accessed from [Sentinel Hub](https://www.sentinel-hub.com/) and normalized using $10log(\sigma_0)$, where pixels with $10log(\sigma_0) > 10$ are set to 10, and $10log(\sigma_0) < -35$ are set to -35.
+- Sentinel-1 SAR VV and VH backscatter \\(\sigma_0\\) was accessed from [Sentinel Hub](https://www.sentinel-hub.com/) and normalized using \\(10log(\sigma_0)\\), where pixels with \\(10log(\sigma_0) > 10\\) are set to \\(10\\) and \\(10log(\sigma_0) < -35\\) are set to \\(-35\\).
 - To get the cloud mask we looked at the [scene classification values](https://custom-scripts.sentinel-hub.com/custom-scripts/sentinel-2/scene-classification/). We assigned a cloud mask value of 1 to scene classification values 8, 9 and 10. All other scene classification values were assigned a cloud mask value of 0.