Restrict scale's values of dequantizeLinear and quantizeLinear to be positive and nonzero (#906)

* Restrict scale of dequantizeLinear and quantizeLinear to be positive
* Update description for non-positive scale values
* remove the step of checking scale contain non-positive values
* Update prose to say values must be nonzero
* update descriptions for scale's values being negative
* refine descriptions for scale's values being negative
* Just normalize wording across two locations

---------

Co-authored-by: Dwayne Robinson <dwayner@microsoft.com>

Author: BruceDai

index.bs

@@ -4116,7 +4116,7 @@ partial dictionary MLOpSupportLimits {
 <div dfn-for="MLGraphBuilder/dequantizeLinear(input, scale, zeroPoint, options)" dfn-type=argument>
     **Arguments:**
         - <dfn>input</dfn>: an {{MLOperand}}. The input tensor.
-        - <dfn>scale</dfn>: an {{MLOperand}}. The scale tensor to multiply each input value by after adjusting by the zero point. It must be [=blockwise broadcastable=] with the input.
+        - <dfn>scale</dfn>: an {{MLOperand}}. The scale tensor to multiply each input value by after adjusting by the zero point. It must be [=blockwise broadcastable=] with the input. Values must be positive and nonzero, or else the behavior is [=implementation-defined=] (e.g. correct results, incorrect results, or compilation failure).
         - <dfn>zeroPoint</dfn>: an {{MLOperand}}. The zero point tensor to subtract from each input value. It has the same [=MLOperand/shape=] as the scale.
         - <dfn>options</dfn>: an {{MLOperatorOptions}}. Specifies the optional parameters of the operation.
 
@@ -4291,7 +4291,7 @@ partial dictionary MLOpSupportLimits {
 <div dfn-for="MLGraphBuilder/quantizeLinear(input, scale, zeroPoint, options)" dfn-type=argument>
     **Arguments:**
         - <dfn>input</dfn>: an {{MLOperand}}. The input tensor.
-        - <dfn>scale</dfn>: an {{MLOperand}}. The scale tensor to divide each input value by before adjusting by the zero point. It must be [=blockwise broadcastable=] with the input.
+        - <dfn>scale</dfn>: an {{MLOperand}}. The scale tensor to divide each input value by before adjusting by the zero point. It must be [=blockwise broadcastable=] with the input. Values must be positive and nonzero, or else behaviors are implementation dependent (e.g. correct results, incorrect results, or compilation failure).
         - <dfn>zeroPoint</dfn>: an {{MLOperand}}. The zero point tensor to add to each rescaled input value. It has the same [=MLOperand/shape=] as the scale.
         - <dfn>options</dfn>: an {{MLOperatorOptions}}. Specifies the optional parameters of the operation.