wenyu332
wenyu332
> This is a good job. Thank you. > And > I remember the f1-score > 88 some days ago, but the f1-score < 88 I ran last night. >...
> > > This is a good job. Thank you. > > > And > > > I remember the f1-score > 88 some days ago, but the f1-score <...
> Micro-averaged result (excluding Other): > P = 71/1953 = 3.64% R = 71/2263 = 3.14% F1 = 3.37% > > MACRO-averaged result (excluding Other): > P = 6.65% R...
> > > Micro-averaged result (excluding Other): > > > P = 71/1953 = 3.64% R = 71/2263 = 3.14% F1 = 3.37% > > > MACRO-averaged result (excluding Other):...
@LeoWood 你好，我尝试将eHealth-base模型转换为pytorch版本的，发现缺少saved_weights.pdparams，不知道您这边解决了吗，谢谢。
模型我已经上传到百度云盘了，下载链接我放到了redeme中，你可以看一下
不好意思，多卡的我也不太会，如果我后面学会了我会更新上来。。。
这个你可以在google上面找找有没有大神有转换工具，我觉得应该有开源的，但我没有尝试过。。。
找到了 可以这样转换 import numpy as np import tensorflow as tf import torch v_list=[] for k,v in torch.load('pytorch_model.bin',map_location='cpu').items(): v_list.append(np.array(v)) def change(ckpt_path, new_ckpt_path): index = 0 with tf.Session() as sess: for var_name,...
恩，crf和hmm中出现这种情况了，我会修改一下提交上来的，谢谢