Fix: Python 3.14 compatibility with PyTorch 2.9

bitsandbytes/backends/default/ops.py

@@ -1,4 +1,5 @@
 from collections.abc import Sequence
+from functools import wraps
 from math import prod, sqrt
 from typing import Optional
 
@@ -8,6 +9,32 @@
 from ..utils import CODE
 
 
+def _try_torch_compile(func=None, **compile_kwargs):
+    """
+    Wrapper around torch.compile that falls back to the original function if compilation fails.
+    """
+
+    def decorator(fn):
+        try:
+            compiled_fn = torch.compile(fn, **compile_kwargs)
+
+            @wraps(fn)
+            def wrapper(*args, **kwargs):
+                try:
+                    return compiled_fn(*args, **kwargs)
+                except Exception:
+                    return fn(*args, **kwargs)
+
+            return wrapper
+        except Exception:
+            return fn
+
+    if func is None:
+        return decorator
+    else:
+        return decorator(func)
+
+
 @register_kernel("bitsandbytes::int8_mm_dequant", "default")
 def _(
     A: torch.Tensor,
@@ -332,7 +359,7 @@ def _(
 }
 
 
-@torch.compile
+@_try_torch_compile
 def _optimizer_precondition_32bit(
     g: torch.Tensor,
     p: torch.Tensor,
@@ -393,7 +420,7 @@ def _optimizer_precondition_32bit(
     unorm_vec.add_(total_norm)
 
 
-@torch.compile
+@_try_torch_compile
 def _optimizer_update_32bit(
     g: torch.Tensor,
     p: torch.Tensor,

tests/test_linear4bit.py

@@ -2,6 +2,7 @@
 import os
 import pickle
 import platform
+import sys
 from tempfile import TemporaryDirectory
 
 import pytest
@@ -320,6 +321,9 @@ def test_params4bit_real_serialization(device, quant_type, blocksize, compress_s
 @pytest.mark.parametrize("fullgraph", TRUE_FALSE, ids=id_formatter("fullgraph"))
 @pytest.mark.parametrize("mode", ["default", "reduce-overhead"], ids=id_formatter("mode"))
 @pytest.mark.skipif(torch.__version__ < (2, 4), reason="Not supported in torch < 2.4")
+@pytest.mark.skipif(
+    torch.__version__ < (2, 10) and sys.version_info >= (3, 14), reason="Not supported in Python 3.14 until torch 2.10"
+)
 def test_linear4bit_torch_compile(device, quant_type, compute_dtype, compress_statistics, bias, fullgraph, mode):
     if device == "hpu" and not is_supported_on_hpu(quant_type):
         pytest.skip("This configuration is not supported on HPU.")

tests/test_linear8bitlt.py

@@ -3,6 +3,7 @@
 import os
 import pickle
 import platform
+import sys
 from tempfile import TemporaryDirectory
 
 import pytest
@@ -234,6 +235,9 @@ def test_linear8bit_serialization(linear8bit):
 @pytest.mark.parametrize("fullgraph", TRUE_FALSE, ids=id_formatter("fullgraph"))
 @pytest.mark.parametrize("mode", ["default", "reduce-overhead"], ids=id_formatter("mode"))
 @pytest.mark.skipif(torch.__version__ < (2, 4), reason="Not supported in torch < 2.4")
+@pytest.mark.skipif(
+    torch.__version__ < (2, 10) and sys.version_info >= (3, 14), reason="Not supported in Python 3.14 until torch 2.10"
+)
 @pytest.mark.skipif(ROCM_WARP_SIZE_64, reason="this test is not supported on ROCm yet")
 def test_linear8bitlt_torch_compile(device, threshold, bias, fullgraph, mode):
     if device == "cuda" and platform.system() == "Windows":