The useful starting point
Trying a technology on one sample can reveal something useful about that sample and setup. It does not establish performance across every workload. Record the question before the experiment so a pleasing result does not quietly redefine the task.
A small step to try
Keep the version, input, conditions and result together. Note what was not tested and whether the example was selected after several attempts. Separate measured outcomes from your interpretation. A concise experiment note can support a practical next question, while a general recommendation or capability claim requires evidence beyond one informal trial.