Computes a safe mean of the losses. Args: losses: `Tensor` whose elements contain individual loss measurements. num_present: The number of measurable elements in `losses`. Returns: A scalar representing the mean of `losses`. If `num_present` is zero, then zero is returned.
(losses, num_present)
| 73 | |
| 74 | |
| 75 | def _safe_mean(losses, num_present): |
| 76 | """Computes a safe mean of the losses. |
| 77 | |
| 78 | Args: |
| 79 | losses: `Tensor` whose elements contain individual loss measurements. |
| 80 | num_present: The number of measurable elements in `losses`. |
| 81 | |
| 82 | Returns: |
| 83 | A scalar representing the mean of `losses`. If `num_present` is zero, |
| 84 | then zero is returned. |
| 85 | """ |
| 86 | total_loss = math_ops.reduce_sum(losses) |
| 87 | return math_ops.div_no_nan(total_loss, num_present, name="value") |
| 88 | |
| 89 | |
| 90 | def _num_present(losses, weights, per_batch=False): |
no test coverage detected