Prime:一个简单的模块在具有SoTA性能的主要NMT数据集上始终优于自我关注和Transformer模型 （211个子文件） 
make.bat 805B docutils.conf 25B libbleu.cpp 3KB module.cpp 791B theme_overrides.css 192B convert_model.lua 3KB convert_dictionary.lua 897B Makefile 607B README.md 15KB README.md 11KB README.md 8KB README.md 4KB README.md 4KB README.md 3KB README.md 3KB README.md 3KB README.md 3KB README.md 2KB README.md 1KB CONTRIBUTING.md 1KB README.md 897B README.md 625B CODE_OF_CONDUCT.md 241B PATENTS 2KB transformer_bm.py 48KB transformer.py 38KB lightconv.py 35KB multihead_attention.py 29KB fconv.py 26KB sequence_generator.py 26KB options.py 24KB lstm.py 24KB combine_transformer.py 23KB test_binaries.py 23KB fconv_self_att.py 23KB test_noising.py 19KB trainer.py 19KB semisupervised_translation.py 19KB multilingual_translation.py 15KB masked_lm.py 15KB indexed_dataset.py 14KB fp16_optimizer.py 14KB fairseq_model.py 14KB transformer_lm.py 13KB block_pair_dataset.py 13KB masked_lm_dataset.py 12KB noising.py 12KB train.py 11KB checkpoint_utils.py 11KB lightweight_convolution.py 11KB dynamic_convolution.py 11KB utils.py 10KB test_sequence_generator.py 10KB dictionary.py 10KB preprocess.py 10KB fairseq_task.py 10KB lightconv_lm.py 10KB downsampled_multihead_attention.py 10KB language_modeling.py 9KB progress_bar.py 9KB translation_moe.py 9KB multilingual_transformer.py 9KB language_pair_dataset.py 9KB adafactor.py 9KB search.py 9KB iterators.py 9KB translation.py 8KB layer_mha.py 8KB eval_lm.py 8KB generate.py 7KB monolingual_dataset.py 7KB utils.py 7KB adaptive_softmax.py 7KB transformer_sentence_encoder.py 7KB data_utils.py 6KB distributed_utils.py 6KB cross_lingual_lm.py 6KB legacy_distributed_data_parallel.py 6KB token_block_dataset.py 6KB masked_lm_loss.py 6KB backtranslation_dataset.py 6KB transformer_from_pretrained_xlm.py 6KB adam.py 6KB score_moe.py 6KB interactive.py 6KB lamb.py 6KB adamod.py 5KB character_token_embedder.py 5KB average_checkpoints.py 5KB cosine_lr_scheduler_bound.py 5KB average_checkpoints.py 5KB multi_corpus_sampled_dataset.py 5KB cosine_lr_scheduler_cut.py 5KB rm_pt.py 5KB cosine_lr_scheduler.py 5KB fconv_lm.py 5KB masked_lm.py 5KB __init__.py 5KB test_train.py 5KB layer_norm.py 4KB共 211 条
- 1
- 2
- 3
资源评论
