Function validate

fairseq_cli/train.py:194–246 · view source on GitHub ↗

Evaluate the model on the validation set(s) and return the losses.

(args, trainer, task, epoch_itr, subsets)

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

mainFunction · 0.70

trainFunction · 0.70

AverageMeterClass · 0.90

get_batch_iteratorMethod · 0.80

datasetMethod · 0.80

get_modelMethod · 0.80

get_meterMethod · 0.80

get_valid_statsFunction · 0.70

next_epoch_itrMethod · 0.45

max_positionsMethod · 0.45

resetMethod · 0.45

valid_stepMethod · 0.45

updateMethod · 0.45

printMethod · 0.45

no test coverage detected