You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
zhanghangit a2e70f02e6 update code and readme 2 years ago
..
.ipynb_checkpoints update code and readme 2 years ago
__pycache__ update code and readme 2 years ago
benchmark update code and readme 2 years ago
clib update code and readme 2 years ago
config update code and readme 2 years ago
criterions update code and readme 2 years ago
data update code and readme 2 years ago
dataclass update code and readme 2 years ago
distributed update code and readme 2 years ago
logging update code and readme 2 years ago
model_parallel update code and readme 2 years ago
models update code and readme 2 years ago
modules update code and readme 2 years ago
optim update code and readme 2 years ago
scoring update code and readme 2 years ago
tasks update code and readme 2 years ago
__init__.py update code and readme 2 years ago
binarizer.py update code and readme 2 years ago
checkpoint_utils.py update code and readme 2 years ago
file_io.py update code and readme 2 years ago
file_utils.py update code and readme 2 years ago
hub_utils.py update code and readme 2 years ago
incremental_decoding_utils.py update code and readme 2 years ago
iterative_refinement_generator.py update code and readme 2 years ago
libbase.cpython-38-x86_64-linux-gnu.so update code and readme 2 years ago
libbleu.cpython-36m-x86_64-linux-gnu.so update code and readme 2 years ago
libbleu.cpython-38-x86_64-linux-gnu.so update code and readme 2 years ago
libnat.cpython-38-x86_64-linux-gnu.so update code and readme 2 years ago
libnat_cuda.cpython-38-x86_64-linux-gnu.so update code and readme 2 years ago
nan_detector.py update code and readme 2 years ago
ngram_repeat_block.py update code and readme 2 years ago
ngram_repeat_block_cuda.cpython-38-x86_64-linux-gnu.so update code and readme 2 years ago
options.py update code and readme 2 years ago
pdb.py update code and readme 2 years ago
quantization_utils.py update code and readme 2 years ago
registry.py update code and readme 2 years ago
search.py update code and readme 2 years ago
sequence_generator.py update code and readme 2 years ago
sequence_scorer.py update code and readme 2 years ago
token_generation_constraints.py update code and readme 2 years ago
tokenizer.py update code and readme 2 years ago
trainer.py update code and readme 2 years ago
utils.py update code and readme 2 years ago
version.py update code and readme 2 years ago
version.txt update code and readme 2 years ago

鹏程-通言模型 通言模型是在M2M-100模型结构上进行改进的多语种机器翻译模型,通过参数复用和增量式训练,将模型参数从1.2B提升至13.2B,在一带一路多个小语种的翻译上大幅提升。

Text Python C++ Cuda other