Merge pull request #572 from robertknight/broadcast-matmul-zero-point

Broadcast zero point vector when converting batched matmul to non-batched

src/ops/matmul.rs

@@ -222,6 +222,18 @@ where
         // nb. We assume `a` is likely already contiguous, so this will be cheap.
         let a_contig = a.to_contiguous_in(pool).auto_return(pool);
         let a_matrix = a_contig.reshaped([num_a_matrices * a_rows, a_cols].as_slice());
+
+        // Broadcast zero point to match new row count.
+        let a_quant: Option<Vec<LhsT>> = a_quant.map(|a_quant| {
+            a_quant
+                .zero_point
+                .iter()
+                .copied()
+                .cycle()
+                .take(a_matrix.size(0))
+                .collect()
+        });
+
         let mut output = matmul_impl(
             pool,
             a_matrix.view(),
@@ -230,7 +242,9 @@ where
             strategy,
             bias,
             alpha,
-            a_quant,
+            a_quant.as_ref().map(|zero_point| QuantParams {
+                zero_point: zero_point.as_slice(),
+            }),
             b_quant,
         )?;
         output.reshape(out_shape);
@@ -1065,7 +1079,15 @@ mod tests {
                 b_zero_point: Some(Tensor::from([3, 4])),
                 expected_err: None,
             },
-            // A input which is a row vector
+            // LHS batch input with vector zero point
+            Case {
+                a: Tensor::zeros(&[3, 2, 2]),
+                b: Tensor::from([[5, 6], [7, 8]]),
+                a_zero_point: Some(Tensor::from([1, 2])),
+                b_zero_point: Some(Tensor::from([3, 4])),
+                expected_err: None,
+            },
+            // An input which is a row vector
             Case {
                 a: Tensor::from([[1, 2, 3, 4]]),
                 b: Tensor::from([[5, 6], [7, 8], [9, 10], [11, 12]]),