allow offload quant (#10)

* allow offload quant

* rm cuda

* refine and pass test

comfy/model_patcher.py

@@ -57,7 +57,7 @@ def to_mmap(t: torch.Tensor, filename: Optional[str] = None) -> torch.Tensor:
     """
     # Create temporary file
     if filename is None:
-        temp_file = tempfile.mktemp(suffix='.pt', prefix='comfy_mmap_')
+        temp_file = tempfile.mkstemp(suffix='.pt', prefix='comfy_mmap_')[1]
     else:
         temp_file = filename
     
@@ -65,12 +65,10 @@ def to_mmap(t: torch.Tensor, filename: Optional[str] = None) -> torch.Tensor:
     cpu_tensor = t.cpu()
     torch.save(cpu_tensor, temp_file)
     
-    # If we created a CPU copy from CUDA, delete it to free memory
+    # If we created a CPU copy from other device, delete it to free memory
-    if t.is_cuda:
+    if not t.device.type == 'cpu':
         del cpu_tensor
         gc.collect()
-        if torch.cuda.is_available():
-            torch.cuda.empty_cache()
     
     # Load with mmap - this doesn't load all data into RAM
     mmap_tensor = torch.load(temp_file, map_location='cpu', mmap=True, weights_only=False)
@@ -110,15 +108,9 @@ def model_to_mmap(model: torch.nn.Module):
     logging.debug(f"Converting model {model.__class__.__name__} to mmap, current free cpu memory: {free_cpu_mem/(1024*1024*1024)} GB")
     
     def convert_fn(t):
-        """Convert function for _apply()
-        
-        - For Parameters: modify .data and return the Parameter object
-        - For buffers (plain Tensors): return new MemoryMappedTensor
-        """
         if isinstance(t, QuantizedTensor):
-            logging.debug(f"QuantizedTensor detected, skipping mmap conversion, tensor meta info: size {t.size()}, dtype {t.dtype}, device {t.device}, is_contiguous {t.is_contiguous()}")
+            logging.debug(f"QuantizedTensor detected, tensor meta info: size {t.size()}, dtype {t.dtype}, device {t.device}, is_contiguous {t.is_contiguous()}")
-            return t
+        if isinstance(t, torch.nn.Parameter):
-        elif isinstance(t, torch.nn.Parameter):
             new_tensor = to_mmap(t.detach())
             return torch.nn.Parameter(new_tensor, requires_grad=t.requires_grad)
         elif isinstance(t, torch.Tensor):

comfy/quant_ops.py

@@ -130,7 +130,19 @@ class QuantizedTensor(torch.Tensor):
             layout_type: Layout class (subclass of QuantizedLayout)
             layout_params: Dict with layout-specific parameters
         """
-        return torch.Tensor._make_wrapper_subclass(cls, qdata.shape, device=qdata.device, dtype=qdata.dtype, requires_grad=False)
+        # Use as_subclass so the QuantizedTensor instance shares the same
+        # storage and metadata as the underlying qdata tensor. This ensures
+        # torch.save/torch.load and the torch serialization storage scanning
+        # see a valid underlying storage (fixes data_ptr errors).
+        if not isinstance(qdata, torch.Tensor):
+            raise TypeError("qdata must be a torch.Tensor")
+        obj = qdata.as_subclass(cls)
+        # Ensure grad flag is consistent for quantized tensors
+        try:
+            obj.requires_grad_(False)
+        except Exception:
+            pass
+        return obj
 
     def __init__(self, qdata, layout_type, layout_params):
         self._qdata = qdata
@@ -575,3 +587,34 @@ def fp8_func(func, args, kwargs):
         ar[0] = plain_input
         return QuantizedTensor(func(*ar, **kwargs), "TensorCoreFP8Layout", input_tensor._layout_params)
     return func(*args, **kwargs)
+
+def _rebuild_quantized_tensor(qdata, layout_type, layout_params):
+    """Rebuild QuantizedTensor during unpickling when qdata is already a tensor."""
+    return QuantizedTensor(qdata, layout_type, layout_params)
+
+
+def _rebuild_quantized_tensor_from_base(qdata_reduce, layout_type, layout_params):
+    """Rebuild QuantizedTensor during unpickling given the base tensor's reduce tuple.
+
+    qdata_reduce is the tuple returned by qdata.__reduce_ex__(protocol) on the original
+    inner tensor. We call the provided rebuild function with its args to recreate the
+    inner tensor, then wrap it in QuantizedTensor.
+    """
+    rebuild_fn, rebuild_args = qdata_reduce
+    qdata = rebuild_fn(*rebuild_args)
+    return QuantizedTensor(qdata, layout_type, layout_params)
+
+
+# Register custom globals with torch.serialization so torch.load(..., weights_only=True)
+# accepts these during unpickling. Wrapped in try/except for older PyTorch versions.
+try:
+    import torch as _torch_serial
+    if hasattr(_torch_serial, "serialization") and hasattr(_torch_serial.serialization, "add_safe_globals"):
+        _torch_serial.serialization.add_safe_globals([
+            QuantizedTensor,
+            _rebuild_quantized_tensor,
+            _rebuild_quantized_tensor_from_base,
+        ])
+except Exception:
+    # If add_safe_globals doesn't exist or registration fails, we silently continue.
+    pass

tests-unit/comfy_quant/test_quant_registry.py

@@ -47,6 +47,29 @@ class TestQuantizedTensor(unittest.TestCase):
         self.assertEqual(dequantized.dtype, torch.float32)
         self.assertTrue(torch.allclose(dequantized, torch.ones(10, 20) * 3.0, rtol=0.1))
 
+    def test_save_load(self):
+        """Test creating a QuantizedTensor with TensorCoreFP8Layout"""
+        fp8_data = torch.randn(256, 128, dtype=torch.float32).to(torch.float8_e4m3fn)
+        scale = torch.tensor(2.0)
+        layout_params = {'scale': scale, 'orig_dtype': torch.bfloat16}
+
+        qt = QuantizedTensor(fp8_data, "TensorCoreFP8Layout", layout_params)
+
+        self.assertIsInstance(qt, QuantizedTensor)
+        self.assertEqual(qt.shape, (256, 128))
+        self.assertEqual(qt.dtype, torch.float8_e4m3fn)
+        self.assertEqual(qt._layout_params['scale'], scale)
+        self.assertEqual(qt._layout_params['orig_dtype'], torch.bfloat16)
+        self.assertEqual(qt._layout_type, "TensorCoreFP8Layout")
+
+        torch.save(qt, "test.pt")
+        loaded_qt = torch.load("test.pt", weights_only=False)
+        # loaded_qt = torch.load("test.pt", map_location='cpu', mmap=True, weights_only=False)
+
+        self.assertEqual(loaded_qt._layout_type, "TensorCoreFP8Layout")
+        self.assertEqual(loaded_qt._layout_params['scale'], scale)
+        self.assertEqual(loaded_qt._layout_params['orig_dtype'], torch.bfloat16)
+
     def test_from_float(self):
         """Test creating QuantizedTensor from float tensor"""
         float_tensor = torch.randn(64, 32, dtype=torch.float32)

tests/inference/test_model_mmap.py

@@ -5,6 +5,11 @@ import psutil
 import os
 import gc
 import tempfile
+import sys
+
+# Ensure the project root is on the Python path (so `import comfy` works when running tests from this folder)
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..')))
+
 from comfy.model_patcher import model_to_mmap, to_mmap