Merge pull request #125 from russellb/max-tokens-default

llmblock: Set a more reasonable default for num_tokens

src/instructlab/sdg/llmblock.py

@@ -78,7 +78,7 @@ def __init__(
         self.defaults = {
             "model": self.ctx.model_id,
             "temperature": 0,
-            "max_tokens": 12000,
+            "max_tokens": 4096,
         }
 
         # Whether the LLM server supports a list of input prompts