包大人

Do you mean document classification? Or short text like twitter? If it is twitter, you can just see it as one sentence. What`s more, document classification is quite different from...

after `self.h_pool = tf.concat(3, pooled_outputs)` the shape is [batch_size, 1, 1, num_filters * len(filter_sizes)]

有进展吗？

@Ablustrund 中旬已过哈哈

@Ablustrund 还有后续不，已经快24年了。

@Ablustrund 加油加油...