Bump up version to v0.3.2 (#2968)

This version is for more model support. Add support for Gemma models (#2964) and OLMo models (#2832).

vllm/__init__.py

@@ -8,7 +8,7 @@
 from vllm.outputs import CompletionOutput, RequestOutput
 from vllm.sampling_params import SamplingParams
 
-__version__ = "0.3.1"
+__version__ = "0.3.2"
 
 __all__ = [
     "LLM",