Added filenames on line 1 and removed RESET_DBS, fixed logging bug

This commit is contained in:
henri123lemoine
2023-06-27 07:21:45 -04:00
parent abb506511e
commit 560b132e67
5 changed files with 13 additions and 9 deletions
+1 -1
View File
@@ -1,4 +1,4 @@
# dataset/pinecone_handler.py
# dataset/pinecone_db_handler.py
import json
import pinecone
+3 -4
View File
@@ -1,3 +1,5 @@
# dataset/settings.py
from pathlib import Path
### FILE PATHS ###
@@ -16,7 +18,4 @@ EMBEDDINGS_RATE_LIMIT = 3500
PINECONE_INDEX_NAME = "stampy-chat-embeddings-test"
PINECONE_VALUES_DIMS = EMBEDDINGS_DIMS
PINECONE_METRIC = "cosine"
PINECONE_METADATA_ENTRIES = ["entry_id", "source", "title", "authors", "text"]
### MISC ###
RESET_DBS = False
PINECONE_METADATA_ENTRIES = ["entry_id", "source", "title", "authors", "text"]
+2
View File
@@ -1,3 +1,5 @@
# dataset/sql_db_handler.py
import sqlite3
from typing import List, Dict, Any
+2
View File
@@ -1,3 +1,5 @@
# dataset/text_splitter.py
import re
from typing import List
import tiktoken
+5 -4
View File
@@ -10,17 +10,18 @@ else:
import openai
openai.api_key = os.environ.get('OPENAI_API_KEY')
import logging
logging.basicConfig(level=logging.INFO)
from dataset.update_dataset import ARDUpdater
from dataset.settings import RESET_DBS
def update_sql_and_pinecone_dbs():
updater = ARDUpdater(
min_tokens_per_block=200,
max_tokens_per_block=300,
)
if RESET_DBS:
updater.reset_dbs()
updater.update(['gwern_blog'])
updater.update()
def main():