burtenshaw
/

GemmaCoder3-12B

Image-Text-to-Text

Generated from Trainer

text-generation-inference

Model card Files Files and versions

burtenshaw HF Staff commited on Apr 1

Commit

b490f7d

·

verified ·

1 Parent(s): a27e42d

Update README.md

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -23,9 +23,9 @@ The model improves performance on the LiveCodeBench benchmark, but diminishes on
 | Benchmark   | GemmaCoder-12B |  Gemma3-12B-it |
 |-------------|----------------|----------------|
-| Winogrande  | 63.9%         | **74.3%**         |
-| MMLU        | 61.0%         | **74.5%**         |
-| HellaSwag   | 54.0%         | **84.2%**         |
 | LiveCodeBench | **32.9%**         | 21.9%         |
 ## Quick start

 | Benchmark   | GemmaCoder-12B |  Gemma3-12B-it |
 |-------------|----------------|----------------|
+| Winogrande  | 63.9%         | **54.2%**         |
+| MMLU        | 61.0%         | **69.5%**         |
+| HellaSwag   | 54.0%         | **63.5%**         |
 | LiveCodeBench | **32.9%**         | 21.9%         |
 ## Quick start