Address review comments

apache · Sep 8, 2022 · 1d89d43 · 1d89d43
1 parent 56eb957
commit 1d89d43
Show file tree

Hide file tree

Showing 2 changed files with 39 additions and 44 deletions.
diff --git a/python/tvm/topi/hexagon/qnn/qadd_qsub_qmul.py b/python/tvm/topi/hexagon/qnn/qadd_qsub_qmul.py
@@ -80,44 +80,44 @@ def get_int_scale(
         corr = zero_point_M << rsh
 
         return scale_fixed_point, rsh, corr
+
+    a_scale_f = scale_A * C_recip
+    b_scale_f = scale_B * C_recip
+    scale_fixed_point_a, rsh_a = get_fixed_point_value(a_scale_f, "int16")
+    scale_fixed_point_b, rsh_b = get_fixed_point_value(b_scale_f, "int16")
+
+    # Here we have two exp_scale_factors rsh_a and rsh_b.
+    # To avoid complexity, we want to use a common exp_scale_factor and
+    # we want to use the lowest of the two.
+
+    # Since, either of scale_fixed_point_a or scale_fixed_point_b has already been multiplied
+    # by 2^max(rsh_a, rsh_b) in topi.hexagon.utils.get_fixed_point_value,
+    # we want to undo that by right shifting that scale_fixed_point value
+    # by the difference of rsh_a and rsh_b.
+
+    # This results into having a common exp_scale_factor for both scale_fixed_point_a
+    # and scale_fixed_point_b.
+
+    # We also set rsh here which is used to adjust the zero_point_M and compute the corr value,
+    # computation of which comes from the original equation of the op's compute.
+
+    if rsh_a > rsh_b:
+        scale_fixed_point_a = scale_fixed_point_a >> (rsh_a - rsh_b)
+        rsh = rsh_b
+    else:
+        scale_fixed_point_b = scale_fixed_point_b >> (rsh_b - rsh_a)
+        rsh = rsh_a
+
+    if op == "qadd":
+        corr = (zero_point_M << rsh) - (
+            zero_point_A * scale_fixed_point_a + zero_point_B * scale_fixed_point_b
+        )
     else:
-        a_scale_f = scale_A * C_recip
-        b_scale_f = scale_B * C_recip
-        scale_fixed_point_a, rsh_a = get_fixed_point_value(a_scale_f, "int16")
-        scale_fixed_point_b, rsh_b = get_fixed_point_value(b_scale_f, "int16")
-
-        # Here we have two exp_scale_factors rsh_a and rsh_b.
-        # To avoid complexity, we want to use a common exp_scale_factor and
-        # we want to use the lowest of the two.
-
-        # Since, either of scale_fixed_point_a or scale_fixed_point_b has already been multiplied
-        # by 2^max(rsh_a, rsh_b) in topi.hexagon.utils.get_fixed_point_value,
-        # we want to undo that by right shifting that scale_fixed_point value
-        # by the difference of rsh_a and rsh_b.
-
-        # This results into having a common exp_scale_factor for both scale_fixed_point_a
-        # and scale_fixed_point_b.
-
-        # We also set rsh here which is used to adjust the zero_point_M and compute the corr value,
-        # computation of which comes from the original equation of the op's compute.
-
-        if rsh_a > rsh_b:
-            scale_fixed_point_a = scale_fixed_point_a >> (rsh_a - rsh_b)
-            rsh = rsh_b
-        else:
-            scale_fixed_point_b = scale_fixed_point_b >> (rsh_b - rsh_a)
-            rsh = rsh_a
-
-        if op == "qadd":
-            corr = (zero_point_M << rsh) - (
-                zero_point_A * scale_fixed_point_a + zero_point_B * scale_fixed_point_b
-            )
-        else:
-            corr = (zero_point_M << rsh) - (
-                zero_point_A * scale_fixed_point_a - zero_point_B * scale_fixed_point_b
-            )
-
-        return scale_fixed_point_a, scale_fixed_point_b, rsh, corr
+        corr = (zero_point_M << rsh) - (
+            zero_point_A * scale_fixed_point_a - zero_point_B * scale_fixed_point_b
+        )
+
+    return scale_fixed_point_a, scale_fixed_point_b, rsh, corr
 
 
 def qadd_broadcast_compute(

diff --git a/tests/python/contrib/test_hexagon/infrastructure.py b/tests/python/contrib/test_hexagon/infrastructure.py
@@ -347,12 +347,7 @@ def quantize_np(arr_np: numpy.ndarray, dtype: str):
     elif fmin > 0:
         fmin = 0.0
 
-    if fmax == fmin:
-        scale = fmax
-        zero_point = 0
-    else:
-        scale = (fmax - fmin) / (qmax - qmin)
-        zero_point = numpy.rint((fmax * qmin - fmin * qmax) / (fmax - fmin)).astype("int32")
-
+    scale = (fmax - fmin) / (qmax - qmin)
+    zero_point = numpy.rint((fmax * qmin - fmin * qmax) / (fmax - fmin)).astype("int32")
     quant_np = (arr_np / scale + zero_point).astype(dtype)
     return quant_np, scale, zero_point