mikeee commited on
Commit
a088911
1 Parent(s): 63eef71

Upload hlm-dataset-README.md

Browse files
Files changed (1) hide show
  1. hlm-dataset-README.md +71 -0
hlm-dataset-README.md ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ # For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/datasetcard.md?plain=1
3
+ # Doc / guide: https://huggingface.co/docs/hub/datasets-cards
4
+ {}
5
+ ---
6
+
7
+ # Dataset Card for `hlm-paraphrase-multilingual-mpnet-base-v2`
8
+
9
+ ### Dataset Summary
10
+
11
+ Chromadb vectorstore for 红楼梦, created with
12
+ ```
13
+ import os
14
+ from langchain.document_loaders import TextLoader
15
+ from langchain.embeddings import SentenceTransformerEmbeddings
16
+ from langchain.text_splitter import RecursiveCharacterTextSplitter
17
+ from langchain.vectorstores import Chroma
18
+
19
+ model_name = 'paraphrase-multilingual-mpnet-base-v2'
20
+ embedding = SentenceTransformerEmbeddings(model_name=model_name)
21
+
22
+ url = 'https://raw.githubusercontent.com/ffreemt/multilingual-dokugpt/master/docs/hlm.txt'
23
+ os.system(f'wget -c {url}')
24
+ doc = TextLoader('hlm.txt').load()
25
+
26
+ text_splitter = RecursiveCharacterTextSplitter(
27
+ separators=["\n\n", "\n", ".", "!", "?", ",", " ", ""],
28
+ chunk_size=620,
29
+ chunk_overlap=60,
30
+ length_function=len
31
+ )
32
+
33
+ doc_chunks = text_splitter.split_documents(doc)
34
+ client_settings = Settings(chroma_db_impl="duckdb+parquet", anonymized_telemetry=False, persist_directory='db')
35
+
36
+ # takes 8-20 minutes on CPU
37
+ vectorstore = Chroma.from_documents(
38
+ documents=doc_chunks,
39
+ embedding=embedding,
40
+ persist_directory='db',
41
+ client_settings=client_settings,
42
+ )
43
+ vectorstore.persist()
44
+ ```
45
+
46
+ ### How to use
47
+
48
+ Download the files to a local directory, e.g., `db`
49
+
50
+
51
+ ```python
52
+ from langchain.embeddings import SentenceTransformerEmbeddings
53
+ from langchain.vectorstores import Chroma
54
+ from chromadb.config import Settings
55
+
56
+ model_name = 'paraphrase-multilingual-mpnet-base-v2'
57
+ embedding = SentenceTransformerEmbeddings(model_name=model_name)
58
+
59
+ path_to_db_parent = '...'
60
+ client_settings = Settings(chroma_db_impl="duckdb+parquet", anonymized_telemetry=False, persist_directory=f'{path_to_db_parent}/db')
61
+
62
+ db = Chroma(
63
+ # persist_directory='docs',
64
+ embedding_function=embedding,
65
+ client_settings=client_settings,
66
+ )
67
+
68
+ res = db.search("红楼梦主线", search_type="similarity", k=2)
69
+ print(res)
70
+ # [Document(page_content='通灵宝玉正面图式\u3000通灵宝玉反面图式\n\n\n\n玉宝灵通\u3000\u3000\u3000\u3000\u3000三二一\n\n仙莫\u3000\u3000\u3000\u3000\u3000\u3000知疗除\n\n寿失\u3000\u3000\u3000\u3000\u3000\u3000祸冤邪\n\n恒莫\u3000\u3000\u3000\u3000\u3000\u3000福疾崇\n\n昌忘\n\n\n\n宝钗看毕,【甲戌双行。。。
71
+ ```