Skip to content

Commit 98d43df

Browse files
mendonkmendonk
authored
[RAG-40] - Docs - langchain migration (#145)
* init * nav * edits
1 parent c905676 commit 98d43df

File tree

2 files changed

+218
-0
lines changed

2 files changed

+218
-0
lines changed

‎docs/modules/ROOT/nav.adoc

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -4,6 +4,7 @@
44
* xref:ROOT:quickstart.adoc[]
55
* xref:ROOT:what-is-rag.adoc[]
66
* xref:ROOT:dev-environment.adoc[]
7+
* xref:ROOT:migration.adoc[]
78
89
.Introduction to RAG
910
* xref:intro-to-rag:index.adoc[]

‎docs/modules/ROOT/pages/migration.adoc

Lines changed: 217 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,217 @@
1+
= Migration
2+
3+
Migrating an existing LangChain application to RAGStack is easy - just change your `requirements.txt` or `pyproject.toml` file to use `ragstack-ai`.
4+
5+
RAGStack contains the below packages as of version 0.3.1. When RAGStack is installed, these packages are replaced with the stable, tested `ragstack-ai` versions listed below. For the latest list, see xref:ROOT:changelog.adoc[].
6+
[%autowidth]
7+
[cols="2*",options="header"]
8+
|===
9+
| Library | Version
10+
11+
| astrapy
12+
| >=0.6.2,<0.7.0
13+
14+
| cassio
15+
| >=0.1.3,<0.2.0
16+
17+
| langchain
18+
| ==0.0.349
19+
20+
| llama-index
21+
| ==0.9.14
22+
23+
| unstructured
24+
| >=0.10,<0.11
25+
26+
|===
27+
28+
== Example migration
29+
30+
Here is a simple LangChain application that loads a dataset from HuggingFace and embeds the document objects in AstraDB.
31+
32+
.migration.py
33+
[%collapsible%open]
34+
====
35+
[source,python]
36+
----
37+
import os
38+
from datasets import load_dataset
39+
from dotenv import load_dotenv
40+
from langchain.document_loaders import PyPDFDirectoryLoader
41+
from langchain.vectorstores.astradb import AstraDB
42+
from langchain.embeddings import OpenAIEmbeddings
43+
from langchain.schema import Document
44+
45+
load_dotenv()
46+
47+
ASTRA_DB_APPLICATION_TOKEN = os.environ.get("ASTRA_DB_APPLICATION_TOKEN")
48+
ASTRA_DB_API_ENDPOINT = os.environ.get("ASTRA_DB_API_ENDPOINT")
49+
OPEN_AI_API_KEY = os.environ.get("OPENAI_API_KEY")
50+
ASTRA_DB_COLLECTION = os.environ.get("ASTRA_DB_COLLECTION")
51+
52+
embedding = OpenAIEmbeddings()
53+
vstore = AstraDB(
54+
embedding=embedding,
55+
collection_name="test3",
56+
token=os.environ["ASTRA_DB_APPLICATION_TOKEN"],
57+
api_endpoint=os.environ["ASTRA_DB_API_ENDPOINT"],
58+
)
59+
print(vstore.astra_db.collection(ASTRA_DB_COLLECTION).find())
60+
61+
philo_dataset = load_dataset("datastax/philosopher-quotes")["train"]
62+
print("An example entry:")
63+
print(philo_dataset[16])
64+
65+
docs = []
66+
for entry in philo_dataset:
67+
metadata = {"author": entry["author"]}
68+
if entry["tags"]:
69+
for tag in entry["tags"].split(";"):
70+
metadata[tag] = "y"
71+
doc = Document(page_content=entry["quote"], metadata=metadata)
72+
docs.append(doc)
73+
74+
inserted_ids = vstore.add_documents(docs)
75+
print(f"\nInserted {len(inserted_ids)} documents.")
76+
77+
print(vstore.astra_db.collection(ASTRA_DB_COLLECTION).find())
78+
79+
vstore.clear()
80+
----
81+
====
82+
83+
. This application requires installation of the following packages:
84+
+
85+
[source,python]
86+
----
87+
pip install langchain datasets openai astrapy tiktoken python-dotenv
88+
----
89+
+
90+
. You decide you want to use RAGStack's pinned, tested version of LangChain (`langchain-0.0.349`) instead of the latest version of LangChain (`langchain-0.0.350`).
91+
+
92+
. Install the `ragstack-ai` package with the `--upgrade-strategy="only-if-needed"` option.
93+
This ensures pip will not upgrade any packages that are already installed, unless required by the `ragstack-ai` package.
94+
+
95+
[source,python]
96+
----
97+
pip install ragstack-ai --upgrade-strategy="only-if-needed"
98+
----
99+
+
100+
[NOTE]
101+
====
102+
If you're having trouble with your migration, try uninstalling your current LangChain package and reinstalling the `ragstack-ai` package.
103+
[source,python]
104+
----
105+
pip uninstall langchain
106+
Successfully uninstalled langchain-0.0.350
107+
pip install ragstack-ai --upgrade-strategy="only-if-needed"
108+
----
109+
====
110+
+
111+
. Once the `ragstack-ai` package is installed, run `pip list` to see your current list of packages.
112+
Notice that the installed version of langchain is `0.0.349`.
113+
+
114+
.Pip list
115+
[%collapsible%open]
116+
====
117+
[source,console]
118+
----
119+
Package Version
120+
------------------- ------------
121+
aiohttp 3.9.1
122+
aiosignal 1.3.1
123+
annotated-types 0.6.0
124+
anyio 4.1.0
125+
astrapy 0.6.2
126+
attrs 23.1.0
127+
backoff 2.2.1
128+
beautifulsoup4 4.12.2
129+
cassandra-driver 3.28.0
130+
cassio 0.1.3
131+
certifi 2023.11.17
132+
chardet 5.2.0
133+
charset-normalizer 3.3.2
134+
click 8.1.7
135+
dataclasses-json 0.6.3
136+
datasets 2.15.0
137+
Deprecated 1.2.14
138+
dill 0.3.7
139+
distro 1.8.0
140+
emoji 2.9.0
141+
filelock 3.13.1
142+
filetype 1.2.0
143+
frozenlist 1.4.0
144+
fsspec 2023.10.0
145+
geomet 0.2.1.post1
146+
greenlet 3.0.2
147+
h11 0.14.0
148+
h2 4.1.0
149+
hpack 4.0.0
150+
httpcore 1.0.2
151+
httpx 0.25.2
152+
huggingface-hub 0.19.4
153+
hyperframe 6.0.1
154+
idna 3.6
155+
joblib 1.3.2
156+
jsonpatch 1.33
157+
jsonpointer 2.4
158+
langchain 0.0.349
159+
langchain-community 0.0.1
160+
langchain-core 0.0.13
161+
langdetect 1.0.9
162+
langsmith 0.0.69
163+
llama-index 0.9.14
164+
lxml 4.9.3
165+
marshmallow 3.20.1
166+
multidict 6.0.4
167+
multiprocess 0.70.15
168+
mypy-extensions 1.0.0
169+
nest-asyncio 1.5.8
170+
nltk 3.8.1
171+
numpy 1.26.2
172+
openai 1.3.8
173+
packaging 23.2
174+
pandas 2.1.4
175+
pip 23.2.1
176+
pyarrow 14.0.1
177+
pyarrow-hotfix 0.6
178+
pydantic 2.5.2
179+
pydantic_core 2.14.5
180+
python-dateutil 2.8.2
181+
python-dotenv 1.0.0
182+
python-iso639 2023.12.11
183+
python-magic 0.4.27
184+
pytz 2023.3.post1
185+
PyYAML 6.0.1
186+
ragstack-ai 0.3.1
187+
rapidfuzz 3.5.2
188+
regex 2023.10.3
189+
requests 2.31.0
190+
setuptools 65.5.0
191+
six 1.16.0
192+
sniffio 1.3.0
193+
soupsieve 2.5
194+
SQLAlchemy 2.0.23
195+
tabulate 0.9.0
196+
tenacity 8.2.3
197+
tiktoken 0.5.2
198+
tqdm 4.66.1
199+
typing_extensions 4.9.0
200+
typing-inspect 0.9.0
201+
tzdata 2023.3
202+
unstructured 0.10.30
203+
urllib3 2.1.0
204+
wrapt 1.16.0
205+
xxhash 3.4.1
206+
yarl 1.9.4
207+
----
208+
====
209+
+
210+
. Run your application...
211+
+
212+
[source,python]
213+
----
214+
python3 migration.py
215+
----
216+
217+
...and you should see the same output as before, with no changes to your code required!

0 commit comments

Comments
 (0)