langchainrb 0.3.5 → 0.3.7
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.env.example +11 -0
- data/CHANGELOG.md +9 -0
- data/Gemfile.lock +15 -3
- data/README.md +28 -10
- data/lib/agent/chain_of_thought_agent/chain_of_thought_agent.rb +2 -2
- data/lib/dependency_helper.rb +8 -0
- data/lib/langchain.rb +25 -1
- data/lib/llm/base.rb +15 -0
- data/lib/llm/cohere.rb +5 -1
- data/lib/llm/hugging_face.rb +3 -2
- data/lib/llm/openai.rb +22 -1
- data/lib/loader.rb +26 -0
- data/lib/loaders/base.rb +17 -0
- data/lib/loaders/pdf.rb +34 -0
- data/lib/loaders/text.rb +22 -0
- data/lib/tool/base.rb +1 -2
- data/lib/vectorsearch/base.rb +39 -7
- data/lib/vectorsearch/milvus.rb +0 -4
- data/lib/vectorsearch/pinecone.rb +1 -1
- data/lib/vectorsearch/qdrant.rb +1 -1
- data/lib/vectorsearch/weaviate.rb +1 -1
- data/lib/version.rb +1 -1
- metadata +23 -5
- data/lib/logging.rb +0 -13
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 93a3fcc195fbdf55ec52402c1db2f11c929069c03afa90477259e6bf2f542957
|
4
|
+
data.tar.gz: 737e456d831e40e8c388a1986f2483f9dff3934c8b4e05a9456529e017075637
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 23619f8e256a9856eb113afce8eef94f759beb84644a87d9b67cac4fac9d5aedfc06978b3baedad1d988308b3cdb72d32e19a0f5e21d8b8431e8c9ff04eda548
|
7
|
+
data.tar.gz: 6cd0fcc55553a5472e2ac6c69a6e49dbd4b52fe8bcc899c04710ae40397c39947a3418649737c0d21a99086f85736d52f5968b0cc8af761f186eedf790ba85db
|
data/.env.example
ADDED
data/CHANGELOG.md
CHANGED
@@ -1,5 +1,14 @@
|
|
1
1
|
## [Unreleased]
|
2
2
|
|
3
|
+
## [0.3.7] - 2023-05-19
|
4
|
+
- Loaders
|
5
|
+
- Introduce `Loaders::Text` to parse .txt files
|
6
|
+
- Introduec `Loaders::PDF` to parse .pdf files
|
7
|
+
|
8
|
+
## [0.3.6] - 2023-05-17
|
9
|
+
- LLMs
|
10
|
+
- Bump `hugging-face` gem version
|
11
|
+
|
3
12
|
## [0.3.5] - 2023-05-16
|
4
13
|
- Bug fixes
|
5
14
|
|
data/Gemfile.lock
CHANGED
@@ -1,11 +1,12 @@
|
|
1
1
|
PATH
|
2
2
|
remote: .
|
3
3
|
specs:
|
4
|
-
langchainrb (0.3.
|
4
|
+
langchainrb (0.3.7)
|
5
5
|
|
6
6
|
GEM
|
7
7
|
remote: https://rubygems.org/
|
8
8
|
specs:
|
9
|
+
Ascii85 (1.0.3)
|
9
10
|
actionpack (7.0.4.3)
|
10
11
|
actionview (= 7.0.4.3)
|
11
12
|
activesupport (= 7.0.4.3)
|
@@ -26,6 +27,7 @@ GEM
|
|
26
27
|
tzinfo (~> 2.0)
|
27
28
|
addressable (2.8.4)
|
28
29
|
public_suffix (>= 2.0.2, < 6.0)
|
30
|
+
afm (0.2.2)
|
29
31
|
ast (2.4.2)
|
30
32
|
builder (3.2.4)
|
31
33
|
byebug (11.1.3)
|
@@ -114,10 +116,11 @@ GEM
|
|
114
116
|
graphql-client (0.18.0)
|
115
117
|
activesupport (>= 3.0)
|
116
118
|
graphql
|
119
|
+
hashery (2.1.2)
|
117
120
|
httparty (0.21.0)
|
118
121
|
mini_mime (>= 1.0.0)
|
119
122
|
multi_xml (>= 0.5.2)
|
120
|
-
hugging-face (0.3.
|
123
|
+
hugging-face (0.3.3)
|
121
124
|
faraday (~> 1.0)
|
122
125
|
i18n (1.13.0)
|
123
126
|
concurrent-ruby (~> 1.0)
|
@@ -144,6 +147,12 @@ GEM
|
|
144
147
|
parallel (1.23.0)
|
145
148
|
parser (3.2.2.1)
|
146
149
|
ast (~> 2.4.1)
|
150
|
+
pdf-reader (1.4.1)
|
151
|
+
Ascii85 (~> 1.0.0)
|
152
|
+
afm (~> 0.2.1)
|
153
|
+
hashery (~> 2.0)
|
154
|
+
ruby-rc4
|
155
|
+
ttfunk
|
147
156
|
pinecone (0.1.71)
|
148
157
|
dry-struct (~> 1.6.0)
|
149
158
|
dry-validation (~> 1.10.0)
|
@@ -211,6 +220,7 @@ GEM
|
|
211
220
|
faraday (>= 1)
|
212
221
|
faraday-multipart (>= 1)
|
213
222
|
ruby-progressbar (1.13.0)
|
223
|
+
ruby-rc4 (0.1.5)
|
214
224
|
ruby2_keywords (0.0.5)
|
215
225
|
standard (1.28.2)
|
216
226
|
language_server-protocol (~> 3.17.0.2)
|
@@ -228,6 +238,7 @@ GEM
|
|
228
238
|
thor (1.2.1)
|
229
239
|
treetop (1.6.12)
|
230
240
|
polyglot (~> 0.3)
|
241
|
+
ttfunk (1.7.0)
|
231
242
|
tzinfo (2.0.6)
|
232
243
|
concurrent-ruby (~> 1.0)
|
233
244
|
unicode-display_width (2.4.2)
|
@@ -249,9 +260,10 @@ DEPENDENCIES
|
|
249
260
|
dotenv-rails (~> 2.7.6)
|
250
261
|
eqn (~> 1.6.5)
|
251
262
|
google_search_results (~> 2.0.0)
|
252
|
-
hugging-face (~> 0.3.
|
263
|
+
hugging-face (~> 0.3.3)
|
253
264
|
langchainrb!
|
254
265
|
milvus (~> 0.9.0)
|
266
|
+
pdf-reader (~> 1.4)
|
255
267
|
pinecone (~> 0.1.6)
|
256
268
|
pry-byebug (~> 3.10.0)
|
257
269
|
qdrant-ruby (~> 0.9.0)
|
data/README.md
CHANGED
@@ -62,7 +62,7 @@ client.create_default_schema
|
|
62
62
|
```
|
63
63
|
|
64
64
|
```ruby
|
65
|
-
# Store
|
65
|
+
# Store plain texts in your vector search database
|
66
66
|
client.add_texts(
|
67
67
|
texts: [
|
68
68
|
"Begin by preheating your oven to 375°F (190°C). Prepare four boneless, skinless chicken breasts by cutting a pocket into the side of each breast, being careful not to cut all the way through. Season the chicken with salt and pepper to taste. In a large skillet, melt 2 tablespoons of unsalted butter over medium heat. Add 1 small diced onion and 2 minced garlic cloves, and cook until softened, about 3-4 minutes. Add 8 ounces of fresh spinach and cook until wilted, about 3 minutes. Remove the skillet from heat and let the mixture cool slightly.",
|
@@ -70,7 +70,13 @@ client.add_texts(
|
|
70
70
|
]
|
71
71
|
)
|
72
72
|
```
|
73
|
+
```ruby
|
74
|
+
# Store the contents of your files in your vector search database
|
75
|
+
my_pdf = Langchain.root.join("path/to/my.pdf")
|
76
|
+
my_text = Langchain.root.join("path/to/my.txt")
|
73
77
|
|
78
|
+
client.add_data(paths: [my_pdf, my_text])
|
79
|
+
```
|
74
80
|
```ruby
|
75
81
|
# Retrieve similar documents based on the query string passed in
|
76
82
|
client.similarity_search(
|
@@ -233,11 +239,21 @@ agent.run(question: "How many full soccer fields would be needed to cover the di
|
|
233
239
|
|
234
240
|
#### Available Tools 🛠️
|
235
241
|
|
236
|
-
| Name
|
237
|
-
|
|
238
|
-
| "calculator" | Useful for getting the result of a math expression | |
|
239
|
-
| "search"
|
240
|
-
| "wikipedia"
|
242
|
+
| Name | Description | ENV Requirements | Gem Requirements |
|
243
|
+
| ------------ | :------------------------------------------------: | :-----------------------------------------------------------: | :---------------------------------------: |
|
244
|
+
| "calculator" | Useful for getting the result of a math expression | | `gem "eqn", "~> 1.6.5"` |
|
245
|
+
| "search" | A wrapper around Google Search | `ENV["SERPAPI_API_KEY"]` (https://serpapi.com/manage-api-key) | `gem "google_search_results", "~> 2.0.0"` | |
|
246
|
+
| "wikipedia" | Calls Wikipedia API to retrieve the summary | | `gem "wikipedia-client", "~> 1.17.0"` |
|
247
|
+
|
248
|
+
|
249
|
+
#### Loaders 🚚
|
250
|
+
|
251
|
+
Need to read data from various sources? Load it up.
|
252
|
+
|
253
|
+
| Name | Class | Gem Requirements |
|
254
|
+
| ---- | ------------- | :--------------------------: |
|
255
|
+
| pdf | Loaders::PDF | `gem "pdf-reader", "~> 1.4"` |
|
256
|
+
| text | Loaders::Text | |
|
241
257
|
|
242
258
|
|
243
259
|
## Logging
|
@@ -251,9 +267,10 @@ Langchain.logger.level = :info
|
|
251
267
|
|
252
268
|
## Development
|
253
269
|
|
254
|
-
|
255
|
-
|
256
|
-
|
270
|
+
1. `git clone https://github.com/andreibondarev/langchainrb.git`
|
271
|
+
2. `cp .env.example .env`, then fill out the environment variables in `.env`
|
272
|
+
3. `rspec spec/` to ensure that the tests pass
|
273
|
+
4. `bin/console` to load the gem in a REPL session. Feel free to add your own instances of LLMs, Tools, Agents, etc. and experiment with them.
|
257
274
|
|
258
275
|
## Core Contributors
|
259
276
|
[<img style="border-radius:50%" alt="Andrei Bondarev" src="https://avatars.githubusercontent.com/u/541665?v=4" width="80" height="80" class="avatar">](https://github.com/andreibondarev)
|
@@ -261,8 +278,9 @@ To install this gem onto your local machine, run `bundle exec rake install`. To
|
|
261
278
|
## Honorary Contributors
|
262
279
|
[<img style="border-radius:50%" alt="Andrei Bondarev" src="https://avatars.githubusercontent.com/u/541665?v=4" width="80" height="80" class="avatar">](https://github.com/andreibondarev)
|
263
280
|
[<img style="border-radius:50%" alt="Rafael Figueiredo" src="https://avatars.githubusercontent.com/u/35845775?v=4" width="80" height="80" class="avatar">](https://github.com/rafaelqfigueiredo)
|
281
|
+
[<img style="border-radius:50%" alt="Ricky Chilcott" src="https://avatars.githubusercontent.com/u/445759?v=4" width="80" height="80" class="avatar">](https://github.com/rickychilcott)
|
264
282
|
|
265
|
-
(Criteria
|
283
|
+
(Criteria for becoming an Honorary Contributor or Core Contributor is pending...)
|
266
284
|
|
267
285
|
## Contributing
|
268
286
|
|
@@ -43,7 +43,7 @@ module Agent
|
|
43
43
|
|
44
44
|
loop do
|
45
45
|
Langchain.logger.info("Agent: Passing the prompt to the #{llm} LLM")
|
46
|
-
response = llm_client.
|
46
|
+
response = llm_client.complete(
|
47
47
|
prompt: prompt,
|
48
48
|
stop_sequences: ["Observation:"],
|
49
49
|
max_tokens: 500
|
@@ -100,7 +100,7 @@ module Agent
|
|
100
100
|
# @return [PromptTemplate] PromptTemplate instance
|
101
101
|
def prompt_template
|
102
102
|
@template ||= Prompt.load_from_path(
|
103
|
-
file_path:
|
103
|
+
file_path: Langchain.root.join("agent/chain_of_thought_agent/chain_of_thought_agent_prompt.json")
|
104
104
|
)
|
105
105
|
end
|
106
106
|
end
|
data/lib/dependency_helper.rb
CHANGED
@@ -1,5 +1,13 @@
|
|
1
1
|
# frozen_string_literal: true
|
2
2
|
|
3
|
+
# This method requires and loads the given gem, and then checks to see if the version of the gem meets the requirements listed in `langchain.gemspec`
|
4
|
+
# This solution was built to avoid auto-loading every single gem in the Gemfile when the developer will mostly likely be only using a few of them.
|
5
|
+
#
|
6
|
+
# @param gem_name [String] The name of the gem to load
|
7
|
+
# @return [Boolean] Whether or not the gem was loaded successfully
|
8
|
+
# @raise [LoadError] If the gem is not installed
|
9
|
+
# @raise [LoadError] If the gem is installed, but the version does not meet the requirements
|
10
|
+
#
|
3
11
|
def depends_on(gem_name)
|
4
12
|
gem(gem_name) # require the gem
|
5
13
|
|
data/lib/langchain.rb
CHANGED
@@ -1,8 +1,21 @@
|
|
1
1
|
# frozen_string_literal: true
|
2
2
|
|
3
|
+
require "logger"
|
4
|
+
|
3
5
|
require_relative "./version"
|
4
6
|
require_relative "./dependency_helper"
|
5
|
-
|
7
|
+
module Langchain
|
8
|
+
class << self
|
9
|
+
attr_accessor :default_loaders
|
10
|
+
attr_accessor :logger
|
11
|
+
|
12
|
+
attr_reader :root
|
13
|
+
end
|
14
|
+
|
15
|
+
@logger ||= ::Logger.new($stdout, level: :warn, formatter: ->(severity, datetime, progname, msg) { "[LangChain.rb] #{msg}\n" })
|
16
|
+
|
17
|
+
@root = Pathname.new(__dir__)
|
18
|
+
end
|
6
19
|
|
7
20
|
module Agent
|
8
21
|
autoload :Base, "agent/base"
|
@@ -38,3 +51,14 @@ module Tool
|
|
38
51
|
autoload :SerpApi, "tool/serp_api"
|
39
52
|
autoload :Wikipedia, "tool/wikipedia"
|
40
53
|
end
|
54
|
+
|
55
|
+
module Loaders
|
56
|
+
autoload :Base, "loaders/base"
|
57
|
+
autoload :PDF, "loaders/pdf"
|
58
|
+
autoload :Text, "loaders/text"
|
59
|
+
end
|
60
|
+
|
61
|
+
autoload :Loader, "loader"
|
62
|
+
|
63
|
+
# Load the default Loaders
|
64
|
+
Langchain.default_loaders ||= [::Loaders::Text, ::Loaders::PDF]
|
data/lib/llm/base.rb
CHANGED
@@ -16,6 +16,21 @@ module LLM
|
|
16
16
|
self.class.const_get(:DEFAULTS).dig(:dimension)
|
17
17
|
end
|
18
18
|
|
19
|
+
# Method supported by an LLM that generates a response for a given chat-style prompt
|
20
|
+
def chat(...)
|
21
|
+
raise NotImplementedError, "#{self.class.name} does not support chat"
|
22
|
+
end
|
23
|
+
|
24
|
+
# Method supported by an LLM that completes a given prompt
|
25
|
+
def complete(...)
|
26
|
+
raise NotImplementedError, "#{self.class.name} does not support completion"
|
27
|
+
end
|
28
|
+
|
29
|
+
# Method supported by an LLM that generates an embedding for a given text or array of texts
|
30
|
+
def embed(...)
|
31
|
+
raise NotImplementedError, "#{self.class.name} does not support generating embeddings"
|
32
|
+
end
|
33
|
+
|
19
34
|
# Ensure that the LLM value passed in is supported
|
20
35
|
# @param llm [Symbol] The LLM to use
|
21
36
|
def self.validate_llm!(llm:)
|
data/lib/llm/cohere.rb
CHANGED
@@ -47,7 +47,11 @@ module LLM
|
|
47
47
|
response.dig("generations").first.dig("text")
|
48
48
|
end
|
49
49
|
|
50
|
-
|
50
|
+
# Cohere does not have a dedicated chat endpoint, so instead we call `complete()`
|
51
|
+
def chat(...)
|
52
|
+
complete(...)
|
53
|
+
end
|
54
|
+
|
51
55
|
alias_method :generate_embedding, :embed
|
52
56
|
end
|
53
57
|
end
|
data/lib/llm/hugging_face.rb
CHANGED
@@ -25,8 +25,9 @@ module LLM
|
|
25
25
|
# @param text [String] The text to embed
|
26
26
|
# @return [Array] The embedding
|
27
27
|
def embed(text:)
|
28
|
-
|
29
|
-
input: text
|
28
|
+
client.embedding(
|
29
|
+
input: text,
|
30
|
+
model: DEFAULTS[:embeddings_model_name]
|
30
31
|
)
|
31
32
|
end
|
32
33
|
end
|
data/lib/llm/openai.rb
CHANGED
@@ -5,6 +5,7 @@ module LLM
|
|
5
5
|
DEFAULTS = {
|
6
6
|
temperature: 0.0,
|
7
7
|
completion_model_name: "text-davinci-003",
|
8
|
+
chat_completion_model_name: "gpt-3.5-turbo",
|
8
9
|
embeddings_model_name: "text-embedding-ada-002",
|
9
10
|
dimension: 1536
|
10
11
|
}.freeze
|
@@ -50,7 +51,27 @@ module LLM
|
|
50
51
|
response.dig("choices", 0, "text")
|
51
52
|
end
|
52
53
|
|
53
|
-
|
54
|
+
# Generate a chat completion for a given prompt
|
55
|
+
# @param prompt [String] The prompt to generate a chat completion for
|
56
|
+
# @return [String] The chat completion
|
57
|
+
def chat(prompt:, **params)
|
58
|
+
default_params = {
|
59
|
+
model: DEFAULTS[:chat_completion_model_name],
|
60
|
+
temperature: DEFAULTS[:temperature],
|
61
|
+
# TODO: Figure out how to introduce persisted conversations
|
62
|
+
messages: [{role: "user", content: prompt}]
|
63
|
+
}
|
64
|
+
|
65
|
+
if params[:stop_sequences]
|
66
|
+
default_params[:stop] = params.delete(:stop_sequences)
|
67
|
+
end
|
68
|
+
|
69
|
+
default_params.merge!(params)
|
70
|
+
|
71
|
+
response = client.chat(parameters: default_params)
|
72
|
+
response.dig("choices", 0, "message", "content")
|
73
|
+
end
|
74
|
+
|
54
75
|
alias_method :generate_embedding, :embed
|
55
76
|
end
|
56
77
|
end
|
data/lib/loader.rb
ADDED
@@ -0,0 +1,26 @@
|
|
1
|
+
module Loader
|
2
|
+
def self.with(*loaders)
|
3
|
+
LoaderSet.new(loaders)
|
4
|
+
end
|
5
|
+
|
6
|
+
class LoaderSet
|
7
|
+
def initialize(loaders)
|
8
|
+
@loaders = Array(loaders)
|
9
|
+
end
|
10
|
+
|
11
|
+
def load(*paths)
|
12
|
+
Array(paths)
|
13
|
+
.flatten
|
14
|
+
.map { |path| first_loadable_loader(path)&.load }
|
15
|
+
.compact
|
16
|
+
end
|
17
|
+
|
18
|
+
def first_loadable_loader(path)
|
19
|
+
@loaders
|
20
|
+
.each do |loader_klass|
|
21
|
+
loader_instance = loader_klass.new(path)
|
22
|
+
return(loader_instance) if loader_instance.loadable?
|
23
|
+
end
|
24
|
+
end
|
25
|
+
end
|
26
|
+
end
|
data/lib/loaders/base.rb
ADDED
data/lib/loaders/pdf.rb
ADDED
@@ -0,0 +1,34 @@
|
|
1
|
+
module Loaders
|
2
|
+
class PDF < Base
|
3
|
+
#
|
4
|
+
# This Loader parses PDF files into text.
|
5
|
+
# If you'd like to use it directly you can do so like this:
|
6
|
+
# Loaders::PDF.new("path/to/my.pdf").load
|
7
|
+
#
|
8
|
+
# This parser is also invoked when you're adding data to a Vectorsearch DB:
|
9
|
+
# qdrant = Vectorsearch::Qdrant.new(...)
|
10
|
+
# path = Langchain.root.join("path/to/my.pdf")
|
11
|
+
# qdrant.add_data(path: path)
|
12
|
+
#
|
13
|
+
|
14
|
+
def initialize(path)
|
15
|
+
depends_on "pdf-reader"
|
16
|
+
require "pdf-reader"
|
17
|
+
|
18
|
+
@path = path
|
19
|
+
end
|
20
|
+
|
21
|
+
# Check that the file is a PDF file
|
22
|
+
def loadable?
|
23
|
+
@path.to_s.end_with?(".pdf")
|
24
|
+
end
|
25
|
+
|
26
|
+
def load
|
27
|
+
::PDF::Reader
|
28
|
+
.new(@path)
|
29
|
+
.pages
|
30
|
+
.map(&:text)
|
31
|
+
.join("\n\n")
|
32
|
+
end
|
33
|
+
end
|
34
|
+
end
|
data/lib/loaders/text.rb
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
module Loaders
|
2
|
+
class Text < Base
|
3
|
+
#
|
4
|
+
# This Loader parses .txt files.
|
5
|
+
# If you'd like to use it directly you can do so like this:
|
6
|
+
# Loaders::Text.new("path/to/my.txt").load
|
7
|
+
#
|
8
|
+
# This parser is also invoked when you're adding data to a Vectorsearch DB:
|
9
|
+
# qdrant = Vectorsearch::Qdrant.new(...)
|
10
|
+
# path = Langchain.root.join("path/to/my.txt")
|
11
|
+
# qdrant.add_data(path: path)
|
12
|
+
#
|
13
|
+
|
14
|
+
def loadable?
|
15
|
+
true
|
16
|
+
end
|
17
|
+
|
18
|
+
def load
|
19
|
+
@path.read
|
20
|
+
end
|
21
|
+
end
|
22
|
+
end
|
data/lib/tool/base.rb
CHANGED
data/lib/vectorsearch/base.rb
CHANGED
@@ -19,24 +19,38 @@ module Vectorsearch
|
|
19
19
|
@llm_api_key = llm_api_key
|
20
20
|
|
21
21
|
@llm_client = LLM.const_get(LLM::Base::LLMS.fetch(llm)).new(api_key: llm_api_key)
|
22
|
+
|
23
|
+
@loaders = Langchain.default_loaders
|
22
24
|
end
|
23
25
|
|
26
|
+
# Method supported by Vectorsearch DB to create a default schema
|
24
27
|
def create_default_schema
|
25
|
-
raise NotImplementedError
|
28
|
+
raise NotImplementedError, "#{self.class.name} does not support creating a default schema"
|
29
|
+
end
|
30
|
+
|
31
|
+
# Method supported by Vectorsearch DB to add a list of texts to the index
|
32
|
+
def add_texts(...)
|
33
|
+
raise NotImplementedError, "#{self.class.name} does not support adding texts"
|
26
34
|
end
|
27
35
|
|
28
|
-
|
29
|
-
|
36
|
+
# Method supported by Vectorsearch DB to search for similar texts in the index
|
37
|
+
def similarity_search(...)
|
38
|
+
raise NotImplementedError, "#{self.class.name} does not support similarity search"
|
30
39
|
end
|
31
40
|
|
32
|
-
#
|
33
|
-
|
34
|
-
|
41
|
+
# Method supported by Vectorsearch DB to search for similar texts in the index by the passed in vector.
|
42
|
+
# You must generate your own vector using the same LLM that generated the embeddings stored in the Vectorsearch DB.
|
43
|
+
def similarity_search_by_vector(...)
|
44
|
+
raise NotImplementedError, "#{self.class.name} does not support similarity search by vector"
|
45
|
+
end
|
46
|
+
|
47
|
+
# Method supported by Vectorsearch DB to answer a question given a context (data) pulled from your Vectorsearch DB.
|
48
|
+
def ask(...)
|
49
|
+
raise NotImplementedError, "#{self.class.name} does not support asking questions"
|
35
50
|
end
|
36
51
|
|
37
52
|
def_delegators :llm_client,
|
38
53
|
:generate_embedding,
|
39
|
-
:generate_completion,
|
40
54
|
:default_dimension
|
41
55
|
|
42
56
|
def generate_prompt(question:, context:)
|
@@ -56,5 +70,23 @@ module Vectorsearch
|
|
56
70
|
|
57
71
|
prompt_template.format(question: question)
|
58
72
|
end
|
73
|
+
|
74
|
+
def add_data(path: nil, paths: nil)
|
75
|
+
raise ArgumentError, "Either path or paths must be provided" if path.nil? && paths.nil?
|
76
|
+
raise ArgumentError, "Either path or paths must be provided, not both" if !path.nil? && !paths.nil?
|
77
|
+
|
78
|
+
texts =
|
79
|
+
Loader
|
80
|
+
.with(*loaders)
|
81
|
+
.load(path || paths)
|
82
|
+
|
83
|
+
add_texts(texts: texts)
|
84
|
+
end
|
85
|
+
|
86
|
+
attr_reader :loaders
|
87
|
+
|
88
|
+
def add_loader(*loaders)
|
89
|
+
loaders.each { |loader| @loaders << loader }
|
90
|
+
end
|
59
91
|
end
|
60
92
|
end
|
data/lib/vectorsearch/milvus.rb
CHANGED
data/lib/vectorsearch/qdrant.rb
CHANGED
data/lib/version.rb
CHANGED
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: langchainrb
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.3.
|
4
|
+
version: 0.3.7
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Andrei Bondarev
|
8
8
|
autorequire:
|
9
9
|
bindir: exe
|
10
10
|
cert_chain: []
|
11
|
-
date: 2023-05-
|
11
|
+
date: 2023-05-19 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: dotenv-rails
|
@@ -86,14 +86,14 @@ dependencies:
|
|
86
86
|
requirements:
|
87
87
|
- - "~>"
|
88
88
|
- !ruby/object:Gem::Version
|
89
|
-
version: 0.3.
|
89
|
+
version: 0.3.3
|
90
90
|
type: :development
|
91
91
|
prerelease: false
|
92
92
|
version_requirements: !ruby/object:Gem::Requirement
|
93
93
|
requirements:
|
94
94
|
- - "~>"
|
95
95
|
- !ruby/object:Gem::Version
|
96
|
-
version: 0.3.
|
96
|
+
version: 0.3.3
|
97
97
|
- !ruby/object:Gem::Dependency
|
98
98
|
name: milvus
|
99
99
|
requirement: !ruby/object:Gem::Requirement
|
@@ -108,6 +108,20 @@ dependencies:
|
|
108
108
|
- - "~>"
|
109
109
|
- !ruby/object:Gem::Version
|
110
110
|
version: 0.9.0
|
111
|
+
- !ruby/object:Gem::Dependency
|
112
|
+
name: pdf-reader
|
113
|
+
requirement: !ruby/object:Gem::Requirement
|
114
|
+
requirements:
|
115
|
+
- - "~>"
|
116
|
+
- !ruby/object:Gem::Version
|
117
|
+
version: '1.4'
|
118
|
+
type: :development
|
119
|
+
prerelease: false
|
120
|
+
version_requirements: !ruby/object:Gem::Requirement
|
121
|
+
requirements:
|
122
|
+
- - "~>"
|
123
|
+
- !ruby/object:Gem::Version
|
124
|
+
version: '1.4'
|
111
125
|
- !ruby/object:Gem::Dependency
|
112
126
|
name: pinecone
|
113
127
|
requirement: !ruby/object:Gem::Requirement
|
@@ -185,6 +199,7 @@ executables: []
|
|
185
199
|
extensions: []
|
186
200
|
extra_rdoc_files: []
|
187
201
|
files:
|
202
|
+
- ".env.example"
|
188
203
|
- ".rspec"
|
189
204
|
- CHANGELOG.md
|
190
205
|
- Gemfile
|
@@ -207,7 +222,10 @@ files:
|
|
207
222
|
- lib/llm/cohere.rb
|
208
223
|
- lib/llm/hugging_face.rb
|
209
224
|
- lib/llm/openai.rb
|
210
|
-
- lib/
|
225
|
+
- lib/loader.rb
|
226
|
+
- lib/loaders/base.rb
|
227
|
+
- lib/loaders/pdf.rb
|
228
|
+
- lib/loaders/text.rb
|
211
229
|
- lib/prompt/base.rb
|
212
230
|
- lib/prompt/few_shot_prompt_template.rb
|
213
231
|
- lib/prompt/loading.rb
|
data/lib/logging.rb
DELETED
@@ -1,13 +0,0 @@
|
|
1
|
-
# frozen_string_literal: true
|
2
|
-
|
3
|
-
require "logger"
|
4
|
-
|
5
|
-
module Langchain
|
6
|
-
def self.logger
|
7
|
-
@@logger ||= Logger.new($stdout, level: :warn, formatter: ->(severity, datetime, progname, msg) { "[LangChain.rb] #{msg}\n" })
|
8
|
-
end
|
9
|
-
|
10
|
-
def self.logger=(instance)
|
11
|
-
@@logger = instance
|
12
|
-
end
|
13
|
-
end
|