jev-training-data/logs/train-progress.log
2026-10-15 08:59:00 +00:00

18 lines
929 B
Plaintext

[jev] device=cuda base=xlm-roberta-base
Some weights of XLMRobertaForSequenceClassification were not initialized from the model checkpoint at xlm-roberta-base and are newly initialized: ['classifier.dense.bias', 'classifier.dense.weight', 'classifier.out_proj.bias', 'classifier.out_proj.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
[jev] train=35848 calib=5291
[jev] ep0 step100/3360 loss=0.3059
[jev] ep0 step200/3360 loss=0.1194
[jev] ep0 step300/3360 loss=0.2336
[jev] ep0 step400/3360 loss=0.2320
[jev] ep0 step500/3360 loss=0.1973
[jev] ep0 step600/3360 loss=0.2068
[jev] ep0 step700/3360 loss=0.1626
[jev] ep0 step800/3360 loss=0.2940
[jev] ep0 step900/3360 loss=0.2423
[jev] ep0 step1000/3360 loss=0.1728
[jev] ep0 step1100/3360 loss=0.2031
[jev] epoch 0 calibration: {'auc': 0.9924, 'ece': 0.0056, 'n': 5291}
[jev] ep1 step1200/3360 loss=0.0928