diff --git a/astrai/config/train_config.py b/astrai/config/train_config.py
index dcf9ff6..82d514d 100644
--- a/astrai/config/train_config.py
+++ b/astrai/config/train_config.py
@@ -45,6 +45,12 @@ class TrainConfig(BaseConfig):
         default_factory=list,
         metadata={"help": "Module types to enable activation checkpointing for."},
     )
+    compile_mode: Optional[str] = field(
+        default=None,
+        metadata={
+            "help": "torch.compile mode: 'default', 'reduce-overhead', 'max-autotune', or None to disable."
+        },
+    )
 
     # checkpoint setting
     start_epoch: int = field(default=0, metadata={"help": "Start epoch for training."})
diff --git a/astrai/trainer/train_context.py b/astrai/trainer/train_context.py
index 3771ee5..8b7c51a 100644
--- a/astrai/trainer/train_context.py
+++ b/astrai/trainer/train_context.py
@@ -1,3 +1,4 @@
+import logging
 import threading
 from dataclasses import dataclass, field
 from pathlib import Path
@@ -19,6 +20,8 @@ from astrai.tokenize import AutoTokenizer
 from astrai.trainer.rollout import RolloutGenerator, RolloutRunner
 from astrai.trainer.strategy import BaseStrategy, StrategyFactory, create_ref_model
 
+logger = logging.getLogger(__name__)
+
 
 @dataclass
 class TrainContext:
@@ -124,6 +127,9 @@ class TrainContextBuilder:
                 )
             if preloaded_state_dict is not None:
                 m.load_state_dict(preloaded_state_dict, strict=False)
+            if cfg.compile_mode is not None:
+                logger.info("torch.compile enabled (mode=%s)", cfg.compile_mode)
+                m = torch.compile(m, mode=cfg.compile_mode)
             return m
 
         context = TrainContext(
diff --git a/scripts/tools/train.py b/scripts/tools/train.py
index abbb758..7d9a2d6 100644
--- a/scripts/tools/train.py
+++ b/scripts/tools/train.py
@@ -208,6 +208,13 @@ _START_METHODS = ["spawn", "fork", "forkserver"]
     default=False,
     help="Enable activation checkpointing.",
 )
+@click.option(
+    "--compile",
+    "compile_mode",
+    type=click.Choice(["default", "reduce-overhead", "max-autotune"]),
+    default=None,
+    help="torch.compile mode. Omit to disable.",
+)
 @click.option(
     "--ckpt_interval", type=int, default=5000, help="Steps between checkpoints."
 )
@@ -495,6 +502,7 @@ def train(
     )
 
     grad_ckpt_modules = [DecoderBlock] if gradient_checkpointing else []
+    compile_mode = kwargs.pop("compile_mode", None)
 
     collate_fn = None
     if train_type == "dpo":
@@ -532,6 +540,7 @@ def train(
         val_step=val_step,
         metrics=metrics,
         gradient_checkpointing_modules=grad_ckpt_modules,
+        compile_mode=compile_mode,
         executor_kwargs=executor_kwargs,
         extra_kwargs=strategy_kwargs,
         neftune_alpha=neftune_alpha,