add image

docs/README.md

@@ -1,10 +1,7 @@
-# Clay model documentation
+# Clay documentation
 
 <!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->
 
-This site cover the documenation on how the Clay model is done, and how
-to use it.
-
 ## Overview
 
 Clay is a [foundational model](#) of Earth using Earth Observation data.
@@ -23,88 +20,43 @@ This fine-tuning supports various tasks, including classification
 as creating RGB imagery from SAR data. Moreover, users can further
 enhance model performance by incorporating higher-resolution data.
 
-Currently in `v0`, it uses ten bands from Sentinel-2, two from
-Sentinel-1, and a Digital Elevation Models (DEM).
-
-The model code lives on
-[Github](https://github.com/Clay-foundation/model), the model weights on
-[Huggin Face](#), and we maintain a set of embeddgins on [Source
-Cooperative](#).
-
-Documentation for this model is made using [nbdev](), making every page
-a python notebook anyone can use, practical code examples for each
-functionality, and use case.
+This documentation uses [nbdev](#), which combines documentation, code
+samples and an SDK. This means that every page is also a python notebook
+anyone can use, with practical code examples for each functionality, and
+use case. Moreover, you can install `pip install clay` and use the same
+functions.
 
 Clay is open source, open data and open for business.
 
-Clay is a fiscally sponsored project of [Radiant Earth](), a USA
-registered 501(c)3 non-profit.
-
-## How Clay is developed
+## Where is what
 
-Development of Clay model code happens in the open on [this
-repo](https://github.com/Clay-foundation/model). It is licensed as
-Apache.
-
-We have high-level goals to define each version: \* `v0`: Released
-Dec’23. Development version aiming to have a functional architecture,
-small sample of trainning data, and notional embeddgins. See [v0 Release
-notes](#). \* `v1`: Released planned for MArch ’24
+- The Clay model **code** lives on
+  [Github](https://github.com/Clay-foundation/model). License:
+  [Apache](https://github.com/Clay-foundation/model/LICENSE).
+- The Clay model **weights** live on [Huggin Face](#). License:
+  [OpenRAIL-M](https://github.com/Clay-foundation/model/blob/main/LICENSE-MODEL.md).
+- The Clay **documentation** lives on here. License: [CC-BY](#).
+- The Clay **SDK** lives on [PyPi](#). License: [Apache](#).
+- We maintain a set of **embeddings** on [Source Cooperative](#).
+  License: [ODC-BY](#).
 
 ## How to use Clay
 
-The model can be used in two main ways: 1. Directly, generating semantic
-**embeddigns**. E.g. Similarity search. 3. **Fine-tunning** output of
-the model: With additional data as labels, and optionally different
-input data (e.g. higher-res images).: 1. Classification: E.g. flood
-detection, crop types, damage assessment, deforestation monitor… 2.
-Regression tasks: E.g. Carbon stock above ground, crop yields, … 3.
-**Generative** tasks: E.g. Estimate RGB imagery wihtout clouds or from
-Synthetic Aperture Radar (SAR) data. to leverage clay’s pre-processing
-for your mode.
-
-### Embeddings
-
-Clay can be directly used to generate semantic embeddings. These
-embeddings are powerful in understanding and interpreting complex
-spatial and temporal patterns in EO data.
-
-A prime example is “Similarity Search”: Utilize embeddings to find
-similar patterns or features within large datasets. This is particularly
-useful in environmental monitoring and land use classification.
-
-See [Embeddings]()
+The model can be used in two main ways:
 
-### Fine-tunning
+1.  Directly, use it to make inference. See [Model](Model.html)
+    1.  Check and run Benchmarks on the model. See
+        [Benchmarks](Benchmarks.html)
+2.  Generating semantic **embeddings**. E.g. for Similarity search. See
+    [Embeddings](Embeddings.html).
+3.  **Fine-tunning** the model for other tasks, or for other input data.
+    E.g. flood detection, crop yields, … See
+    [Fine-tunning](Fine-tunning.html).
+4.  **Generative** tasks: E.g. Estimate RGB imagery wihtout clouds or
+    from Synthetic Aperture Radar (SAR) data. See
+    [Generative](Generative.html).
 
-As a foundational model, Clay is highly adaptable through fine-tuning,
-allowing it to tackle specific tasks by using additional labeled data
-and, optionally, different input data like higher-resolution images.
+–
 
-#### Classification tasks
-
-Apply Clay to classify various geographical features and phenomena, such
-as flood detection, crop type identification, damage assessment, and
-deforestation monitoring.
-
-Fine-tune Clay with labeled data relevant to your classification task.
-This involves retraining the model’s final layers to recognize and
-categorize specific features or conditions in EO data.
-
-#### Regresion tasks
-
-Estimate quantitative information like carbon stock above ground, crop
-yields, and other environmental metrics.
-
-Regression tasks require fine-tuning with data that provide a measurable
-output. Clay learns to predict these continuous values from EO data,
-enabling precise environmental assessments.
-
-### Generative AI
-
-Clay can generate new data representations, such as estimating RGB
-imagery without clouds or converting SAR data into RGB imagery.
-
-This involves training Clay to recognize patterns in SAR data and
-translate them into corresponding RGB formats, a process that can be
-critical in cloud-penetrated regions or for night-time imagery analysis.
+Clay is a fiscally sponsored project of [Radiant Earth](), a USA
+registered 501(c)3 non-profit.