unbreak float8 static quant tutorial

Summary:

The tutorial was broken, fixing it.

Test Plan:

```
python tutorials/calibration_flow/static_quant.py
```

Reviewers:

Subscribers:

Tasks:

Tags:

ghstack-source-id: aebfc44af855a8c39ddcf1eb06c6e13e4c12d19a
ghstack-comment-id: 2655560001
Pull Request resolved: #1709

torchao/dtypes/floatx/float8_layout.py

@@ -253,6 +253,7 @@ def _linear_fp8_act_fp8_weight_impl(
 ):
     """Implements matmul between FP8 input and FP8 weight with compute using _scaled_mm"""
     scaled_mm_config = weight_tensor._layout.mm_config
+    assert scaled_mm_config is not None
     out_shape = get_out_shape(input_tensor.shape, weight_tensor.shape)
 
     # Weight tensor preprocessing

tutorials/calibration_flow/static_quant.py

@@ -163,12 +163,13 @@ def __init__(
                 weight, weight_scale, weight_zero_point, block_size, self.target_dtype
             )
         elif self.target_dtype == torch.float8_e4m3fn:
+            mm_config = Float8MMConfig(use_fast_accum=True)
             self.qweight = to_affine_quantized_floatx_static(
                 weight,
                 weight_scale,
                 block_size,
                 target_dtype,
-                Float8Layout(mm_config=None),
+                Float8Layout(mm_config=mm_config),
             )
         else:
             raise ValueError(f"Unsupported target dtype {self.target_dtype}")