Warmup IPEX models at init

Author: ofirzaf

optimum/intel/ipex/modeling_base.py

@@ -15,6 +15,7 @@
 
 import logging
 import os
+from functools import wraps
 from pathlib import Path
 from tempfile import TemporaryDirectory
 from typing import Optional, Tuple, Union
@@ -43,7 +44,7 @@
 from optimum.modeling_base import OptimizedModel
 from optimum.utils import NormalizedConfigManager
 
-from ..generation.modeling import jit_trace
+from ..generation.modeling import jit_trace, prepare_jit_inputs
 from ..utils.import_utils import is_torch_version
 from ..utils.modeling_utils import MULTI_QUERY_ATTN_MODELS, patch_decoder_attention_mask
 
@@ -62,6 +63,7 @@ def __init__(
         model,
         config: PretrainedConfig = None,
         model_save_dir: Optional[Union[str, Path, TemporaryDirectory]] = None,
+        initial_warmup: bool = True,
         **kwargs,
     ):
         OptimizedModel.__init__(self, model=model, config=config)
@@ -79,6 +81,8 @@ def __init__(
         AutoConfig.register(self.base_model_prefix, AutoConfig)
         if hasattr(self.auto_model_class, "register"):
             self.auto_model_class.register(AutoConfig, self.__class__)
+        if initial_warmup:
+            self._init_warmup()
 
     @classmethod
     def _from_transformers(
@@ -222,6 +226,14 @@ def _call_model(self, *args, **kwargs):
             out = self.model(*args, **kwargs)
         return out
 
+    def _init_warmup(self):
+        # warmup, the first 2 forwards of an IPEX model include some preprocessing steps and
+        # the results of the compute are unpredictable
+        use_cache = getattr(self, "use_cache", getattr(self.config, "use_cache", False))
+        dummy_inputs = prepare_jit_inputs(self, self.export_feature, use_cache)
+        for _ in range(2):
+            self(**dummy_inputs)
+
 
 class IPEXModelForSequenceClassification(IPEXModel):
     auto_model_class = AutoModelForSequenceClassification
@@ -280,8 +292,9 @@ class IPEXModelForQuestionAnswering(IPEXModel):
     auto_model_class = AutoModelForQuestionAnswering
     export_feature = "question-answering"
 
+    @wraps(IPEXModel.forward)
     def forward(self, *args, **kwargs):
-        outputs = self._call_model(*args, **kwargs)
+        outputs = super().forward(*args, **kwargs)
         start_logits = outputs["start_logits"] if isinstance(outputs, dict) else outputs[0]
         end_logits = outputs["end_logits"] if isinstance(outputs, dict) else outputs[1]
         return ModelOutput(start_logits=start_logits, end_logits=end_logits)
@@ -299,7 +312,8 @@ def __init__(
         use_cache: bool = True,
         **kwargs,
     ):
-        super().__init__(model, config, model_save_dir=model_save_dir)
+        # Perform the initial warmup at the end of __init__
+        super().__init__(model, config, model_save_dir=model_save_dir, initial_warmup=False)
 
         self.normalized_config = NormalizedConfigManager.get_normalized_config_class(config.model_type)(config)
         self.model_dtype = kwargs.get("model_dtype", self.dtype)
@@ -315,6 +329,7 @@ def __init__(
         config.is_decoder = True
         config.is_encoder_decoder = False
         self.generation_config = GenerationConfig.from_model_config(config)
+        self._init_warmup()
 
     def _prepare_past_key_values(self, input_ids):
         model_type = self.config.model_type.replace("_", "-")