Updated for exact values

openshift-psap · Feb 22, 2024 · 6ebfa88 · 6ebfa88
1 parent 7e4a066
commit 6ebfa88
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 1 deletion.
diff --git a/language/gpt-j/SUT.py b/language/gpt-j/SUT.py
@@ -124,7 +124,7 @@ def __init__(self,
 
         if not batch_size:
             if device == "cpu": # Also applies to API server mode
-                batch_size = 13368
+                batch_size = 31192
             else:
                 batch_size = 32  # Reduce to 8 if using 4 GPUs, 16 for 8.
         self.batch_size = batch_size

diff --git a/language/gpt-j/user.conf b/language/gpt-j/user.conf
@@ -3,3 +3,4 @@
 # Model maybe '*' as wildcard. In that case the value applies to all models.
 # All times are in milli seconds
 *.Server.target_qps = 120
+*.Offline.min_query_count = 93576