Update README.md

README.md

@@ -9,16 +9,22 @@ pinned: false
 
 <img src="https://github.com/user-attachments/assets/82709377-2a46-460e-b2d6-45e6b0be98d7" width=355px height=60px />
 
-With <em>release v0.1.2</em> comes integration for the Hugging Face ecosystem.
-<code><a target="_blank" href="https://github.com/aurelienmorgan/retrain-pipelines">retrain-pipelines</a></code> runs which take advantage of this do benefit from automated&nbsp;:
-  - dataset and model versioning with dynamically-generated READMEs with full fine-grained lineage
-  - execution source-code uploading
-  - execution pipeline-card uploading
-  - an inspector for model performance history across versions
-  - and all other good things that make the <code><a target="_blank" href="https://github.com/aurelienmorgan/retrain-pipelines">retrain-pipelines</a></code>' DNA, like highly-customizable pipeline templates to get started fast&nbsp;!
+<b>Wondering if the new version of your model is better than the last&nbsp;? Did continuous training improve, or decrease perf&nbsp;? Does sorting through model cards feel like a never-ending chore&nbsp;?</b>
 
-Browse model versions &nbsp;:
+With <code><a target="_blank" href="https://github.com/aurelienmorgan/retrain-pipelines">retrain-pipelines</a></code>, you can overlook your model versions, hassle-free &nbsp;:
 
 | with the inspector | on the Hub |
 |---|---|
-| <a target="_blank" href="https://github.com/user-attachments/assets/9cbbb624-9cef-4e36-b982-e0897fa81d93"><img alt="inspector.webm" width="250" height="210" src="https://github.com/user-attachments/assets/f9d1ce14-a4c3-48a0-91fc-ab141fbbd750" /></a> | <a target="_blank" href="https://github.com/user-attachments/assets/d3fb9a55-69c8-40a4-b458-52b0217f559e"><img alt="hub.webm" width="250" src="https://github.com/user-attachments/assets/41627b60-8de9-4846-a807-821c139cb70a" /></a> |
+| <a target="_blank" href="https://github.com/user-attachments/assets/6863f30d-bfaa-4a9a-946f-fb8b90e8bba0"><img alt="inspector.webm" width="250" src="https://github.com/user-attachments/assets/8ed27954-e875-4a71-81c9-8435648de139" /></a> | <a target="_blank" href="https://github.com/user-attachments/assets/d3fb9a55-69c8-40a4-b458-52b0217f559e"><img alt="hub.webm" width="250" src="https://github.com/user-attachments/assets/77549e82-ead0-4625-945b-3d73598bde2d" /></a> |
+
+<hr />
+
+<code><a target="_blank" href="https://github.com/aurelienmorgan/retrain-pipelines">retrain-pipelines</a></code> focuses on continuous learning (retraining through dataset updates/refinements), enabling the creation of a new version of a model and the comparison of its performance against that of its best predecessor.
+
+<em>Release v0.1.2</em> brings integration for the Hugging Face ecosystem.
+<code><a target="_blank" href="https://github.com/aurelienmorgan/retrain-pipelines">retrain-pipelines</a></code> runs which take advantage of this do benefit from automated&nbsp;:
+  - dataset and model versioning with dynamically-generated READMEs and full fine-grained lineage
+  - execution source-code upload
+  - execution pipeline-card upload
+  - an inspector for model performance history across versions
+  - and all other good things that make the <code><a target="_blank" href="https://github.com/aurelienmorgan/retrain-pipelines">retrain-pipelines</a></code>' DNA, like highly-customizable pipeline templates to get started fast&nbsp;!