mirror of
https://github.com/facebookresearch/fairseq.git
synced 2024-08-16 20:10:40 +03:00
- no more FP16Trainer, we just have an FP16Optimizer wrapper - most of the distributed code is moved to a new wrapper class called DistributedFairseqModel, which behaves like DistributedDataParallel and a FairseqModel at the same time - Trainer now requires an extra dummy_batch argument at initialization, which we do fwd/bwd on when there's an uneven number of batches per worker. We hide the gradients from these dummy batches by multiplying the loss by 0 - Trainer.train_step now takes a list of samples, which will allow cleaner --update-freq |
||
---|---|---|
.. | ||
_static | ||
command_line_tools.rst | ||
conf.py | ||
criterions.rst | ||
data.rst | ||
docutils.conf | ||
getting_started.rst | ||
index.rst | ||
lr_scheduler.rst | ||
make.bat | ||
Makefile | ||
models.rst | ||
modules.rst | ||
optim.rst | ||
overview.rst | ||
requirements.txt | ||
tasks.rst | ||
tutorial_classifying_names.rst | ||
tutorial_simple_lstm.rst |