-
在32G 的卡上跑,clip 默认设置的200。 batch-size=1都会溢出。也是服了。
然后发现preprocess里面的clip 其实没有clip掉。最长的句子竟然会出现1146个char 溢出应该也正常吧。。。
修改了代码,贴上来,有遇到同样问题的人可以参考一下。
```
import argparse
import numpy as np
import os
imp…
-
**Prerequisites**
> Please fill in by replacing `[ ]` with `[x]`.
* [x ] Are you running the latest `bert-as-service`?
* [x] Did you follow [the installation](https://github.com/hanxiao/bert-as…
-
我使用BiLSTMCRF做ner任务,embedding为cn-Bert-base-wwm, 模型开始训练时,Trainer设置save_path 为"./save_model" ,训练完成后生成文件:best_BiLSTMCRF_f_2021-07-06-12-14-12-018221; 随后我想加载现有训练结果继续训练,遂编写callback 为CheckPointCallback(save_…
-
提问时请尽可能提供如下信息:
### 基本信息
- 你使用的**操作系统**: Ubuntu 16.04
- 你使用的**Python**版本: 3.6.9
- 你使用的**Tensorflow**版本: 1.15.0
- 你使用的**Keras**版本: 2.4.3
- 你使用的**bert4keras**版本: 0.7.4
- 你使用纯**keras**还是**tf.kera…
-
请教一下,正在尝试训练 https://github.com/fishaudio/Bert-VITS2 v1.0.1 版本的。我只训练了中文环境的,所以只使用了中文Bert。 目前训练到14万步,基本效果出来了,但是训练过程突然僵死了。已经过了10多个小时,日志也不打,也没有新的checkpoint存储,显卡一直占用,进程也没有退出。就是没有任何反应,不知道在干嘛。请问这种情况如何处理呢?大概率…
-
您在掩码中文词时的实现如下:
```
for index in index_set:
covered_indexes.add(index)
masked_token = None
# 80% of the time, replace with [MASK]
if rng.random() < 0.8:
mas…
-
## Description
1. Run GluonNLP [full suite of tests](https://github.com/dmlc/gluon-nlp/tree/master/tests) with `pytest` on `mxnet-cu102==2.0.0b20201022` will introduce threading error (see Error Me…
-
I followed your instructions using my data.
Since the batch_size was too big for my data i changed it to 6.
Then i got this error during evaluation:
`
08/23/2019 17:50:14 - INFO - root - …
-
```bash
python s2s_pipeline.py --local_mac_optimal_settings
```
It seems this is done running setup and ready for me to start speaking? My mic is set to MacBook Pro Microphone. I say somethin…
-
Hi,
Thank you for providing `mixed_large_24_model.bin`.
I was trying to convert this model using `scripts/convert_bert_from_uer_to_google.py`.
Everything goes fine without any alerts.
However, wh…