[torch.compile] allow candidate compile sizes (#10984)

Signed-off-by: youkaichao <youkaichao@gmail.com>

tests/engine/test_arg_utils.py

@@ -50,12 +50,12 @@ def test_compilation_config():
     args = parser.parse_args(["-O=3"])
     assert args.compilation_config.level == 3
 
-    # set to json
+    # set to string form of a dict
-    args = parser.parse_args(["--compilation-config", '{"level": 3}'])
+    args = parser.parse_args(["--compilation-config", "{'level': 3}"])
     assert args.compilation_config.level == 3
 
-    # set to json
+    # set to string form of a dict
-    args = parser.parse_args(['--compilation-config={"level": 3}'])
+    args = parser.parse_args(["--compilation-config={'level': 3}"])
     assert args.compilation_config.level == 3
 
 

vllm/config.py

@@ -1,3 +1,4 @@
+import ast
 import copy
 import enum
 import hashlib
@@ -2191,14 +2192,10 @@ class CompilationConfig(BaseModel):
         - use_inductor: whether to use inductor compilation.
             - False: inductor compilation is not used. graph runs in eager.
             - True: inductor compilation is used. one graph for symbolic shape
-                is compiled. In addition, compile for different sizes specified
+                is compiled. In addition, compile for cudagraph sizes that are
-                in inductor_compile_sizes, using configurations
+                in candidate_compile_sizes, using configurations
                 in inductor_compile_config.
-        - inductor_compile_sizes: sizes to compile for inductor.
+        - candidate_compile_sizes: sizes to compile for inductor.
-        - inductor_specialize_for_cudagraph_no_more_than: an optional integer
-            to specialize inductor for cudagraph sizes no more than the
-            specified size. It is useful when we want to specialize inductor
-            with a subset of cudagraph sizes.
         - inductor_compile_config: additional configurations for inductor.
             - None: use default configurations.
         - inductor_passes: additional passes for inductor. It is a dictionary
@@ -2227,8 +2224,7 @@ class CompilationConfig(BaseModel):
     ])
 
     use_inductor: bool = True
-    inductor_specialize_for_cudagraph_no_more_than: Optional[int] = None
+    candidate_compile_sizes: Optional[List[int]] = Field(default=None)
-    inductor_compile_sizes: Optional[List[int]] = Field(default=None)
     inductor_compile_config: Dict = Field(default_factory=dict)
     inductor_passes: Dict[str, str] = Field(default_factory=dict)
 
@@ -2294,7 +2290,9 @@ class CompilationConfig(BaseModel):
         """Parse the CLI value for the compilation config."""
         if cli_value in ["0", "1", "2", "3"]:
             return cls(level=int(cli_value))
-        return CompilationConfig.model_validate_json(cli_value)
+        # do not use `eval`, it is dangerous and can execute arbitrary code
+        dict_value = ast.literal_eval(cli_value)
+        return CompilationConfig.model_validate(dict_value)
 
     def model_post_init(self, __context: Any) -> None:
 
@@ -2355,18 +2353,20 @@ class CompilationConfig(BaseModel):
             logger.info(("cudagraph sizes specified by model runner"
                          " %s is overridden by config %s"),
                         sizes_to_specialize, self.cudagraph_capture_sizes)
-        if self.inductor_specialize_for_cudagraph_no_more_than is not None:
+
-            assert self.inductor_compile_sizes is None, (
+        if self.candidate_compile_sizes is None:
-                "inductor_compile_sizes should be None when "
+            self.candidate_compile_sizes = []
-                "inductor_specialize_for_cudagraph_no_more_than is not None")
+        self.compile_sizes = [
-            self.compile_sizes = [
+            x for x in self.candidate_compile_sizes if x in self.capture_sizes
-                x for x in self.capture_sizes
+        ]
-                if x <= self.inductor_specialize_for_cudagraph_no_more_than
+        ignored_sizes = [
-            ]
+            x for x in self.candidate_compile_sizes
-        else:
+            if x not in self.capture_sizes
-            if self.inductor_compile_sizes is None:
+        ]
-                self.inductor_compile_sizes = []
+        if ignored_sizes:
-            self.compile_sizes = self.inductor_compile_sizes
+            logger.warning(("candidate_compile_sizes %s are ignored "
+                            "because they are not cudagraph capture sizes."),
+                           ignored_sizes)
 
         # sort to make sure cudagraph capture sizes are in descending order
         self.capture_sizes.sort(reverse=True)

vllm/engine/arg_utils.py

@@ -209,12 +209,9 @@ class EngineArgs:
         # support `EngineArgs(compilation_config={...})`
         # without having to manually construct a
         # CompilationConfig object
-        if isinstance(self.compilation_config, (int)):
+        if isinstance(self.compilation_config, (int, dict)):
             self.compilation_config = CompilationConfig.from_cli(
                 str(self.compilation_config))
-        elif isinstance(self.compilation_config, (dict)):
-            self.compilation_config = CompilationConfig.from_cli(
-                json.dumps(self.compilation_config))
 
         # Setup plugins
         from vllm.plugins import load_general_plugins

vllm/entrypoints/llm.py

@@ -1,5 +1,4 @@
 import itertools
-import json
 import warnings
 from contextlib import contextmanager
 from typing import (Any, ClassVar, Dict, List, Optional, Sequence, Tuple, Type,
@@ -186,12 +185,9 @@ class LLM:
             kwargs["disable_log_stats"] = True
 
         if compilation_config is not None:
-            if isinstance(compilation_config, (int)):
+            if isinstance(compilation_config, (int, dict)):
                 compilation_config_instance = CompilationConfig.from_cli(
                     str(compilation_config))
-            elif isinstance(compilation_config, (dict)):
-                compilation_config_instance = CompilationConfig.from_cli(
-                    json.dumps(compilation_config))
             else:
                 compilation_config_instance = compilation_config
         else: