Preparing the catalog and filters…Katalog ve filtreler hazırlanıyor…
Starter code
module trainability_profile;
param p0: Angle = 0.1;
param p1: Angle = 0.2;
param p2: Angle = 0.3;
param p3: Angle = 0.4;
param p4: Angle = 0.5;
param p5: Angle = 0.6;
fn main() {
let q = qreg[4];
Ry(q[0], p0);
Ry(q[1], p1);
Ry(q[2], p2);
Ry(q[3], p3);
CNOT(q[0], q[1]);
CNOT(q[1], q[2]);
CNOT(q[2], q[3]);
Ry(q[0], p4);
Ry(q[3], p5);
let result = expect Z(q[0]);
return result;
}Expected output
NM-QML-301 appears as a warning
Execution is not blocked
The profile remains deterministic
Low variance is marked below 1e-3
Debugging checklist
Parse the starter and find NM-QML-301
Reduce one threshold dimension
Open Analyze trainability
Compare shallow and deeper profiles
Name one mitigation
Extension challenge
Replace global depth with layer-wise growth and record how the warning and variance change.