Update README.md

README.md

@@ -1,5 +1,4 @@
 ---
-license: gemma
 base_model: google/gemma-3-270m
 pipeline_tag: text-classification
 language:
@@ -12,6 +11,9 @@ tags:
 metrics:
 - accuracy
 - f1
+license: other
+license_name: link-attribution
+license_link: https://dejan.ai/link-attribution/
 ---
 
 # Gemma 3 270M: 7-level sentiment classifier
@@ -38,6 +40,8 @@ On a held-out test set of 8,749 segments (from documents never seen in training)
 | Exact or one level off | **95.2%** |
 | Macro-F1 | 0.833 |
 
+![image](https://cdn-uploads.huggingface.co/production/uploads/64732e7f7be71eb8b1b572a8/HyGInCobx5fLKgxO6t0ra.png)
+
 | Label | F1 |
 |---|---|
 | very positive | 0.88 |
@@ -48,8 +52,14 @@ On a held-out test set of 8,749 segments (from documents never seen in training)
 | negative | 0.79 |
 | very negative | 0.89 |
 
+
+![image](https://cdn-uploads.huggingface.co/production/uploads/64732e7f7be71eb8b1b572a8/Tnc55aKYHEz_FOKJXj3U9.png)
+
 Most errors are one level off. The hardest boundary is "somewhat negative" versus "negative".
 
+
+![image](https://cdn-uploads.huggingface.co/production/uploads/64732e7f7be71eb8b1b572a8/Urc74fNkhEu6GLCJCTCnM.png)
+
 **Important:** the test set is synthetic and was labelled by the same process as the training data (see below). These scores measure agreement with that labelling. The model has not yet been evaluated on human-labelled, real-world text.
 
 ## Usage
@@ -93,4 +103,4 @@ Splits are by document (80/10/10), so no document has segments in both training
 
 ## License
 
-This model is a derivative of Gemma and is provided under and subject to the [Gemma Terms of Use](https://ai.google.dev/gemma/terms). Use must also follow the [Gemma Prohibited Use Policy](https://ai.google.dev/gemma/prohibited_use_policy).
+This model is a derivative of Gemma and is provided under and subject to the [Gemma Terms of Use](https://ai.google.dev/gemma/terms). Use must also follow the [Gemma Prohibited Use Policy](https://ai.google.dev/gemma/prohibited_use_policy).