diff options
| author | Dominic Symes <dominic.symes@arm.com> | 2023-02-20 14:38:41 +0000 |
|---|---|---|
| committer | Eric Kunze <eric.kunze@arm.com> | 2023-02-24 19:09:43 +0000 |
| commit | 8f7688d719bad4bd5c71a5ab6e0d05929a9fc045 (patch) | |
| tree | e9cb8fdad7d1d3d34b181aa52f29968538d356b0 | |
| parent | 580fdd133ad239e491d239fe4fdc16b442161b31 (diff) | |
| download | specification-8f7688d719bad4bd5c71a5ab6e0d05929a9fc045.tar.gz | |
Main inference conformance additions
Add a check on the error bias for dot
product calculations and slightly increase the
error variance bound for dot products.
Change-Id: I518704d1fb1b109695b0f90f7ca6733850e6a8c7
Signed-off-by: Dominic Symes <dominic.symes@arm.com>
| -rw-r--r-- | chapters/introduction.adoc | 17 |
1 files changed, 10 insertions, 7 deletions
diff --git a/chapters/introduction.adoc b/chapters/introduction.adoc index 5765f22..cae23d4 100644 --- a/chapters/introduction.adoc +++ b/chapters/introduction.adoc @@ -300,14 +300,17 @@ For each data test set (S, N) consisting of T tests the following must hold: * if `y_bnd[t] == 0` then ** `y_imp[t]` must be zero and set `y_err[t] = 0` * if `y_bnd[t] > 0` then set: -** `y_err[t] = abs(y_imp[t] - y_ref[t]) * (1<<(M+1)) / y_bnd[t]` calculated using fp64 arithmetic -* For each test t the following must be satified: +** `y_err[t] = (y_imp[t] - y_ref[t]) * (1<<(M+1)) / y_bnd[t]` calculated using fp64 arithmetic +* For each test t the following must be satisfied: ** `y_ref[t], y_bnd[t], y_imp[t]` must be finite -** `y_err[t] \<= N` -* Calculate the mean-square (variance from 0) of y_err using 64 bit IEEE floating point arithmetic: -** `y_var = (y_err[0] * y_err[0] + ... + y_err[N-1] * y_err[N-1]) / T` -* The mean-square must satisfy: -** `y_var \<= N/3` +** `abs(y_err[t]) \<= N` +* Calculate the sum of y_err using fp64 arithmetic: +** `y_err_sum = y_err[0] + .... + y_err[T-1]` +* Calculate the sum of y_err squared using fp64 arithmetic: +** `y_err_sumsq = y_err[0] * y_err[0] + ... + y_err[T-1] * y_err[T-1]` +* The error sum and sum squares must satisfy the following. The first equation bounds the bias and the second the error variance. +** `abs(y_err_sum) \<= 2*sqrt(N*T)` +** `y_err_sumsq \<= 0.4*N*T` === Tensor Definitions |