RobAgrees
/

quantized-dia-1.6B-int8

model_hub_mixin

pytorch_model_hub_mixin

Model card Files Files and versions

RobAgrees commited on Apr 29, 2025

Commit

8eb0be6

·

verified ·

1 Parent(s): e9325c8

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -27,7 +27,7 @@ widget:
   It uses dynamic quantization for lighter deployment and faster inference.
   Original model: **float16**, ~6.4GB
-  Quantized model: **int8 dynamic**, ~6.4GB
   ## ⚡️ Quickstart

   It uses dynamic quantization for lighter deployment and faster inference.
   Original model: **float16**, ~6.4GB
+  Quantized model: **int8 dynamic**, ~6.4GB, ~20% faster inference
   ## ⚡️ Quickstart