pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Move api.settings to common.settings (#11036)

Browse Source 
### What problem does this PR solve?

As title

### Type of change

- [x] Refactoring

---------

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
This commit is contained in:
Jin Hai
2025-11-06 09:36:38 +08:00
committed by GitHub
parent 87c9a054d3
commit f98b24c9bf
68 changed files with 675 additions and 718 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
rag/flow/tokenizer/tokenizer.py
View File

@ -27,7 +27,7 @@ from common.connection_utils import timeout
from rag.flow.base import ProcessBase, ProcessParamBase
from rag.flow.tokenizer.schema import TokenizerFromUpstream
from rag.nlp import rag_tokenizer
from rag.settings import EMBEDDING_BATCH_SIZE
from common import settings
from rag.svr.task_executor import embed_limiter
from common.token_utils import truncate
@ -82,16 +82,16 @@ class Tokenizer(ProcessBase):
return embedding_model.encode([truncate(c, embedding_model.max_length - 10) for c in txts])
cnts_ = np.array([])
for i in range(0, len(texts), EMBEDDING_BATCH_SIZE):
for i in range(0, len(texts), settings.EMBEDDING_BATCH_SIZE):
async with embed_limiter:
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(texts[i : i + EMBEDDING_BATCH_SIZE]))
vts, c = await trio.to_thread.run_sync(lambda: batch_encode(texts[i : i + settings.EMBEDDING_BATCH_SIZE]))
if len(cnts_) == 0:
cnts_ = vts
else:
cnts_ = np.concatenate((cnts_, vts), axis=0)
token_count += c
if i % 33 == 32:
self.callback(i * 1.0 / len(texts) / parts / EMBEDDING_BATCH_SIZE + 0.5 * (parts - 1))
self.callback(i * 1.0 / len(texts) / parts / settings.EMBEDDING_BATCH_SIZE + 0.5 * (parts - 1))
cnts = cnts_
title_w = float(self._param.filename_embd_weight)