| moe_coarse_1e19 [swapped_residency]: validation loss at iteration 2 on test set | lm loss value: 4.414523E+00 | lm loss PPL: 8.264242E+01 | | |
| g1_tmoe_coarse_1e19 [swapped_unmasked]: validation loss at iteration 2 on test set | lm loss value: 3.793768E+00 | lm loss PPL: 4.442347E+01 | | |
| temporal_fine_g3_1e19 [swapped_unmasked]: validation loss at iteration 2 on test set | lm loss value: 3.768807E+00 | lm loss PPL: 4.332835E+01 | | |
| ALL SWAP EVALS DONE | |