Several smaller fixes to make adapters work

Also, reorganized the pattern of __init__ inside providers so
configuration can stay lightweight

llama_toolchain/inference/meta_reference/__init__.py

@@ -5,4 +5,15 @@
 # the root directory of this source tree.
 
 from .config import MetaReferenceImplConfig  # noqa
-from .inference import get_provider_impl  # noqa
+
+
+async def get_provider_impl(config: MetaReferenceImplConfig, _deps):
+    from .inference import MetaReferenceInferenceImpl
+
+    assert isinstance(
+        config, MetaReferenceImplConfig
+    ), f"Unexpected config type: {type(config)}"
+
+    impl = MetaReferenceInferenceImpl(config)
+    await impl.initialize()
+    return impl