Fix 627 #642
Fix 627 #642
Conversation
|
|
||
| assert feature_frame_num == label_frame_num | ||
|
|
||
| if self._split_sentence_threshold == -1 or self._split_perturb == -1 or self._split_sub_sentence_len == -1 or self._split_sentence_threshold >= feature_frame_num: |
There was a problem hiding this comment.
Please keep one line under the 80-character limit, which is also suitable for comments
|
LGTM |
| @@ -61,10 +61,21 @@ class SampleInfoBucket(object): | |||
| label_bin_paths (list|tuple): Files containing the binary label data. | |||
| label_desc_paths (list|tuple): Files containing the description of | |||
| samples' label data. | |||
| split_perturb(int): split long sentence' perturb sub-sentence length value. | |||
| split_sentence_threshold(int): sentence length large than | |||
There was a problem hiding this comment.
sentence --> Sentence
large --> larger
operator --> operation
sentence length large than split_sentence_threshold trigger split operator. -->
Sentence whose length larger than the value will trigger split operation.
| @@ -61,10 +61,21 @@ class SampleInfoBucket(object): | |||
| label_bin_paths (list|tuple): Files containing the binary label data. | |||
| label_desc_paths (list|tuple): Files containing the description of | |||
| samples' label data. | |||
| split_perturb(int): split long sentence' perturb sub-sentence length value. | |||
There was a problem hiding this comment.
split_perturb(int) --> split_perturb (int)
There was a problem hiding this comment.
split long sentence' perturb sub-sentence length value. Please refine this comment.
There was a problem hiding this comment.
Random perturb sub-sentence length when split long sentence
There was a problem hiding this comment.
How about Maximum perturbation value for length of sub-sentence when splitting long sentence.?
| remain_frame_num = feature_frame_num | ||
| while True: | ||
| if remain_frame_num > self._split_sentence_threshold: | ||
| cur_frame_len = self._split_sub_sentence_len + random.randint( |
There was a problem hiding this comment.
Please use self._rng instead using random.randint directly.
There was a problem hiding this comment.
self._rng.randint(0, self._split_perturb)
| @@ -244,11 +291,20 @@ def read_bytes(fpath, start, size): | |||
| sample_info.feature_start, | |||
| sample_info.feature_size) | |||
|
|
|||
| assert sample_info.feature_frame_num * sample_info.feature_dim * 4 == len( | |||
| @@ -273,7 +329,8 @@ def read_bytes(fpath, start, size): | |||
| time.sleep(0.001) | |||
|
|
|||
| # drop long sentence | |||
| if self._drop_frame_len >= sample_data[0].shape[0]: | |||
| if self._drop_frame_len == -1 or self._drop_frame_len >= sample_data[ | |||
split long sentence, drop sentence close by -1