manage default (greedy) gen_kwargs in vllm (#1341)

* manage default (greedy) gen_kwargs in vllm better

* mirror HF `do_sample`

* just need to set temp=0 for greedy

lm_eval/models/vllm_causallms.py

@@ -170,14 +170,8 @@ def _model_generate(
         stop: Optional[List[str]] = None,
         **kwargs,
     ):
-        if "do_sample" in kwargs.keys():
-            kwargs.pop("do_sample")
         if generate:
-            # hf defaults
-            kwargs["skip_special_tokens"] = kwargs.get("skip_special_tokens", False)
-            kwargs["spaces_between_special_tokens"] = kwargs.get(
-                "spaces_between_special_tokens", False
-            )
+            kwargs = self.modify_gen_kwargs(kwargs)
             sampling_params = SamplingParams(max_tokens=max_tokens, stop=stop, **kwargs)
         else:
             sampling_params = SamplingParams(
@@ -438,3 +432,16 @@ def _parse_logprobs(tokens: List, outputs, ctxlen: int) -> Tuple[float, bool]:
                     break
 
         return continuation_logprobs, is_greedy
+
+    @staticmethod
+    def modify_gen_kwargs(kwargs: dict) -> dict:
+        # sampling_params
+        do_sample = kwargs.pop("do_sample", False)
+        if do_sample is not True:
+            kwargs["temperature"] = 0.0
+        # hf defaults
+        kwargs["skip_special_tokens"] = kwargs.get("skip_special_tokens", False)
+        kwargs["spaces_between_special_tokens"] = kwargs.get(
+            "spaces_between_special_tokens", False
+        )
+        return kwargs