renaming and clarifying max_length parameter

predict.py

@@ -20,10 +20,10 @@ def predict(
         self,
         prompt: str = Input(description=f"Prompt to send to LLaMA."),
         n: int = Input(description="Number of output sequences to generate", default=1, ge=1, le=5),
-        max_length: int = Input(
-            description="Maximum number of tokens to generate. A word is generally 2-3 tokens",
+        total_tokens: int = Input(
+            description="Maximum number of tokens for input + generation. A word is generally 2-3 tokens",
             ge=1,
-            default=50
+            default=2000
         ),
         temperature: float = Input(
             description="Adjusts randomness of outputs, greater than 1 is random and 0 is deterministic, 0.75 is a good starting value.",
@@ -50,7 +50,7 @@ def predict(
         outputs = self.model.generate(
             input,
             num_return_sequences=n,
-            max_length=max_length,
+            max_length=total_tokens,
             do_sample=True,
             temperature=temperature,
             top_p=top_p,