-
### Describe the bug
In the following G2P sample, the text "To be or not to be, that is the question" is used.
https://huggingface.co/speechbrain/soundchoice-g2p
In "wordemb/util.py", "expand_to_…
-
Hi!
I had issues uploading the solved jupyter notebook of the last homework. I tried to upload it to my previously forked branch, in which only the early files were there (and I cannot find the file …
-
-
您好,请问:在运行 main_for_evaluation() 时,程序并未报错而是一直卡在运行状态无变化。
经调试发现程序运行到:
with tf.variable_scope("model"):
mtest = network.GRU(is_training=False, word_embeddings=wordembedding, settings=test_s…
-
I just downloaded OpenNMT & attempted to train using the demo data set and ended up with an illegal memory access error.
Command that I ran -
`CUDA_LAUNCH_BLOCKING=1 th train.lua -data data/demo-tr…
-
-
- runtimeHome - explain what is it used for
- sub - remove this property if it is not used
- tokenization - rename to "preprocessing"
- typeToks - remove, assume "tagger" value is the only one whic…
-
-
The "run.bat" script has the options `text`, `read_vocab` and `train_file`.
- Which is the input format expected for these files (e.g. the a dump `https://dumps.wikimedia.org/enwiki/latest/enwiki-late…
-
请教下,假设有2张卡,下面代码中各层如何设置0和1:
def auto_configure_device_map(num_gpus):
# transformer.layers 占用 28 层
num_trans_layers = 28
# bugfix: 在linux中调用torch.embedding传入的weight,input不在同一device…