[MISC] Dump model runner inputs when crashing (vllm-project#8305)

.github/ISSUE_TEMPLATE/400-bug report.yml

@@ -30,6 +30,15 @@ body:
       </details>
   validations:
     required: true
+- type: textarea
+  attributes:
+    label: Model Input Dumps
+    description: |
+      If you are facing crashing due to illegal memory access or other issues with model execution, vLLM may dump the problematic input of the model. In this case, you will see the message `Error in model execution (input dumped to /tmp/err_xxx.pkl)`. If you see this message, please zip the file (because GitHub doesn't support .pkl file format) and upload it here. This will help us to reproduce the issue and facilitate the debugging process.
+    placeholder: |
+      Upload the dumped input file.
+  validations:
+    required: false
 - type: textarea
   attributes:
     label: 🐛 Describe the bug

tests/basic_correctness/test_basic_correctness.py

@@ -3,12 +3,16 @@
 Run `pytest tests/basic_correctness/test_basic_correctness.py`.
 """
 import os
+import pickle
+import re
 import weakref
+from unittest.mock import patch
 
 import pytest
 
 from vllm import LLM
 from vllm.utils import is_hip
+from vllm.worker.model_runner import ModelInputForGPUWithSamplingMetadata
 
 from ..models.utils import check_outputs_equal
 
@@ -67,3 +71,29 @@ def test_models(
         name_0="hf",
         name_1="vllm",
     )
+
+
+def test_model_with_failure(vllm_runner) -> None:
+    try:
+        with patch("vllm.model_executor.models.opt.OPTForCausalLM.forward",
+                   side_effect=ValueError()):
+            with pytest.raises(ValueError) as exc_info:
+                vllm_runner("facebook/opt-125m",
+                            dtype="half",
+                            enforce_eager=False,
+                            gpu_memory_utilization=0.7)
+            matches = re.search(r"input dumped to (.+).pkl",
+                                str(exc_info.value))
+            assert matches is not None
+            filename = f"{matches.group(1)}.pkl"
+
+        with open(filename, "rb") as filep:
+            inputs = pickle.load(filep)
+
+        if any(key not in inputs for key in ("arg_1", "arg_2", "arg_3")):
+            raise AssertionError("Missing keys in dumped inputs. Dumped keys: "
+                                 f"{list(inputs.keys())}")
+        assert isinstance(inputs["arg_1"],
+                          ModelInputForGPUWithSamplingMetadata)
+    finally:
+        os.remove(filename)

vllm/worker/model_runner.py

@@ -53,7 +53,7 @@
     _add_attn_metadata_broadcastable_dict,
     _add_sampling_metadata_broadcastable_dict,
     _init_attn_metadata_from_tensor_dict,
-    _init_sampling_metadata_from_tensor_dict)
+    _init_sampling_metadata_from_tensor_dict, dump_input_when_exception)
 
 if TYPE_CHECKING:
     from vllm.attention.backends.abstract import AttentionBackend
@@ -1489,6 +1489,7 @@ def prepare_model_input(
                                    virtual_engine=virtual_engine)
 
     @torch.inference_mode()
+    @dump_input_when_exception(exclude_args=[0], exclude_kwargs=["self"])
     def execute_model(
         self,
         model_input: ModelInputForGPUWithSamplingMetadata,

vllm/worker/model_runner_base.py

@@ -1,5 +1,8 @@
 import dataclasses
+import pickle
 from abc import ABC, abstractmethod
+from datetime import datetime
+from functools import wraps
 from typing import (TYPE_CHECKING, Any, Dict, Generic, List, Optional, Type,
                     TypeVar)
 
@@ -98,6 +101,37 @@ def _init_frozen_model_input_from_tensor_dict(
     return tensor_dict
 
 
+def dump_input_when_exception(exclude_args: Optional[List[int]] = None,
+                              exclude_kwargs: Optional[List[str]] = None):
+
+    def _inner(func):
+
+        @wraps(func)
+        def _wrapper(*args, **kwargs):
+            try:
+                return func(*args, **kwargs)
+            except Exception as err:
+                timestamp = datetime.now().strftime("%Y%m%d-%H%M%S")
+                filename = f"/tmp/err_{func.__name__}_input_{timestamp}.pkl"
+                with open(filename, "wb") as filep:
+                    dumped_inputs = {
+                        k: v
+                        for k, v in kwargs.items()
+                        if k not in (exclude_kwargs or [])
+                    }
+                    for i, arg in enumerate(args):
+                        if i not in (exclude_args or []):
+                            dumped_inputs[f"arg_{i}"] = arg
+                    pickle.dump(dumped_inputs, filep)
+                raise type(err)(
+                    f"Error in model execution (input dumped to {filename}): "
+                    f"{str(err)}") from err
+
+        return _wrapper
+
+    return _inner
+
+
 class BroadcastableModelInput(ABC):
 
     @abstractmethod