logo
Browse Source

Count chunk

Signed-off-by: Jael Gu <mengjia.gu@zilliz.com>
main
Jael Gu 2 years ago
parent
commit
723998f09f
  1. 2
      osschat_insert.py

2
osschat_insert.py

@ -98,8 +98,8 @@ def osschat_insert_pipe(config):
p = ( p = (
pipe.input('doc', 'project_name') pipe.input('doc', 'project_name')
.map('doc', 'text', data_loader) .map('doc', 'text', data_loader)
.map('doc', 'token_count', ops.token_counter(config.token_model))
.flat_map('text', 'chunk', text_split_op) .flat_map('text', 'chunk', text_split_op)
.map('chunk', 'token_count', ops.token_counter(config.token_model))
.map('chunk', 'embedding', sentence_embedding_op, config=sentence_embedding_config) .map('chunk', 'embedding', sentence_embedding_op, config=sentence_embedding_config)
) )
if config.embedding_normalize: if config.embedding_normalize:

Loading…
Cancel
Save