Update README.md

README.md

@@ -16,6 +16,33 @@ datasets:
 language:
   - en
 pipeline_tag: text-generation
+metrics:
+  - accuracy
+  - f1
+  - precision
+  - recall
+model-index:
+  - name: qlora-cyber-security-classifier
+    results:
+      - task:
+          type: text-classification
+          name: Cybersecurity threat classification
+        dataset:
+          name: SQL Injection + Phishing (held-out test split)
+          type: custom
+        metrics:
+          - type: accuracy
+            value: 0.99
+            name: Accuracy
+          - type: f1
+            value: 0.99
+            name: Weighted F1
+          - type: precision
+            value: 0.99
+            name: Weighted Precision
+          - type: recall
+            value: 0.99
+            name: Weighted Recall
 ---
 
 ## Model Information
@@ -176,10 +203,31 @@ classification accuracy** — see Evaluation below.
 
 ## Evaluation
 
-Precision/recall/F1 on the held-out test split have not yet been recorded in
-this model card. `[Add your classification_report output here once you have
-it — run the evaluation cell against test_records and paste the per-class
-precision/recall/F1 table.]`
+Evaluated on the held-out test split (1,532 examples) using exact-match
+comparison between the model's generated `Classification:` label and ground
+truth.
+
+| Class | Precision | Recall | F1-score | Support |
+| :--- | :---: | :---: | :---: | :---: |
+| benign | 1.00 | 1.00 | 1.00 | 585 |
+| legitimate | 0.99 | 0.96 | 0.97 | 203 |
+| phishing | 0.96 | 0.99 | 0.97 | 182 |
+| sql injection | 1.00 | 1.00 | 1.00 | 562 |
+| **accuracy** | | | **0.99** | 1532 |
+| macro avg | 0.99 | 0.99 | 0.99 | 1532 |
+| weighted avg | 0.99 | 0.99 | 0.99 | 1532 |
+
+**Overall test accuracy: 99%.** The SQL injection task (benign / sql
+injection) is essentially perfect on this test split. The phishing task
+(legitimate / phishing) is slightly softer, with legitimate URLs occasionally
+misclassified as phishing (96% recall) and phishing URLs very reliably
+caught (99% recall) — i.e., the model is a little more likely to over-flag a
+legitimate URL than to miss an actual phishing one.
+
+Note this reflects performance on a **held-out split of the same cleaned
+dataset** used for training — it does not measure generalization to
+attack patterns or URL structures outside that distribution (see
+[Limitations](#limitations)).
 
 ## Limitations
 
@@ -191,9 +239,10 @@ precision/recall/F1 table.]`
 - **Templated explanations:** the `Reason:` text is class-templated rather
   than generated per-example, so explanations are somewhat generic rather
   than deeply input-specific.
-- **No adversarial evaluation:** obfuscated SQL payloads (encoding, comment
-  tricks, case-mixing) and homoglyph/lookalike phishing domains were not
-  specifically tested.
+- **No adversarial evaluation:** the 99% accuracy above is on a clean
+  held-out split from the same source datasets. Obfuscated SQL payloads
+  (encoding, comment tricks, case-mixing) and homoglyph/lookalike phishing
+  domains were not specifically tested, and performance on those is unknown.
 - **Long training time relative to budget:** the run took ~9.35 hours rather
   than the intended ~1 hour, most likely due to 7B-parameter 4-bit inference
   overhead plus gradient checkpointing on a single T4 — worth profiling