File size: 8,123 Bytes
ac3b3f0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
"""Ingest examples into Weaviate."""
import os
from pathlib import Path

import weaviate

WEAVIATE_URL = os.environ["WEAVIATE_URL"]
client = weaviate.Client(
    url=WEAVIATE_URL,
    additional_headers={"X-OpenAI-Api-Key": os.environ["OPENAI_API_KEY"]},
)

client.schema.delete_class("Rephrase")
client.schema.delete_class("QA")
client.schema.get()
schema = {
    "classes": [
        {
            "class": "Rephrase",
            "description": "Rephrase Examples",
            "vectorizer": "text2vec-openai",
            "moduleConfig": {
                "text2vec-openai": {
                    "model": "ada",
                    "modelVersion": "002",
                    "type": "text",
                }
            },
            "properties": [
                {
                    "dataType": ["text"],
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": False,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "content",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "question",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "answer",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "chat_history",
                },
            ],
        },
    ]
}

client.schema.create(schema)

documents = [
    {
        "question": "how do i load those?",
        "chat_history": "Human: What types of memory exist?\nAssistant: \n\nThere are a few different types of memory: Buffer, Summary, and Conversational Memory.",
        "answer": "How do I load Buffer, Summary, and Conversational Memory",
    },
    {
        "question": "how do i install this package?",
        "chat_history": "",
        "answer": "How do I install langchain?",
    },
    {
        "question": "how do I set serpapi_api_key?",
        "chat_history": "Human: can you write me a code snippet for that?\nAssistant: \n\nYes, you can create an Agent with a custom LLMChain in LangChain. Here is a [link](https://langchain.readthedocs.io/en/latest/modules/agents/examples/custom_agent.html) to the documentation that provides a code snippet for creating a custom Agent.",
        "answer": "How do I set the serpapi_api_key?",
    },
    {
        "question": "What are some methods for data augmented generation?",
        "chat_history": "Human: List all methods of an Agent class please\nAssistant: \n\nTo answer your question, you can find a list of all the methods of the Agent class in the [API reference documentation](https://langchain.readthedocs.io/en/latest/modules/agents/reference.html).",
        "answer": "What are some methods for data augmented generation?",
    },
    {
        "question": "can you write me a code snippet for that?",
        "chat_history": "Human: how do I create an agent with custom LLMChain?\nAssistant: \n\nTo create an Agent with a custom LLMChain in LangChain, you can use the [Custom Agent example](https://langchain.readthedocs.io/en/latest/modules/agents/examples/custom_agent.html). This example shows how to create a custom LLMChain and use an existing Agent class to parse the output. For more information on Agents and Tools, check out the [Key Concepts](https://langchain.readthedocs.io/en/latest/modules/agents/key_concepts.html) documentation.",
        "answer": "Can you provide a code snippet for creating an Agent with a custom LLMChain?",
    },
]
from langchain.prompts.example_selector.semantic_similarity import \
    sorted_values

for d in documents:
    d["content"] = " ".join(sorted_values(d))
with client.batch as batch:
    for text in documents:
        batch.add_data_object(
            text,
            "Rephrase",
        )

client.schema.get()
schema = {
    "classes": [
        {
            "class": "QA",
            "description": "Rephrase Examples",
            "vectorizer": "text2vec-openai",
            "moduleConfig": {
                "text2vec-openai": {
                    "model": "ada",
                    "modelVersion": "002",
                    "type": "text",
                }
            },
            "properties": [
                {
                    "dataType": ["text"],
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": False,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "content",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "question",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "answer",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "summaries",
                },
                {
                    "dataType": ["text"],
                    "description": "The link",
                    "moduleConfig": {
                        "text2vec-openai": {
                            "skip": True,
                            "vectorizePropertyName": False,
                        }
                    },
                    "name": "sources",
                },
            ],
        },
    ]
}

client.schema.create(schema)

documents = [
    {
        "question": "how do i install langchain?",
        "answer": "```pip install langchain```",
        "summaries": ">Example:\nContent:\n---------\nYou can pip install langchain package by running 'pip install langchain'\n----------\nSource: foo.html",
        "sources": "foo.html",
    },
    {
        "question": "how do i import an openai LLM?",
        "answer": "```from langchain.llm import OpenAI```",
        "summaries": ">Example:\nContent:\n---------\nyou can import the open ai wrapper (OpenAI) from the langchain.llm module\n----------\nSource: bar.html",
        "sources": "bar.html",
    },
]
from langchain.prompts.example_selector.semantic_similarity import \
    sorted_values

for d in documents:
    d["content"] = " ".join(sorted_values(d))
with client.batch as batch:
    for text in documents:
        batch.add_data_object(
            text,
            "QA",
        )