active_genie 0.0.20 → 0.0.22
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/README.md +1 -1
- data/VERSION +1 -1
- metadata +14 -111
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 3fdcde2460a0cd018a5c140a4ad9018b5a48149789921a75ac2fc71601408d44
|
|
4
|
+
data.tar.gz: 0df8c2d70f45f1786da6622c71feff04743b5ec0479d5cd2c4d79e30d885306d
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: a3f22c2e6af49dba9301403652def1cd85cd0c5ac7e64195de28c0cdae79afadbacae79771c741a20cf6b0e854e2e2088571c0802074dfd49b6e37eef1e92ef7
|
|
7
|
+
data.tar.gz: 3837b2687720a22633b3b06f012c81b431cb18ff097ea9d2abc98364db0c65a73f8998117e3522e821933c95f2e897703162485d76bb4ccd1d96eb0ebe4adc2c
|
data/README.md
CHANGED
|
@@ -4,7 +4,7 @@
|
|
|
4
4
|
[](https://badge.fury.io/rb/active_genie)
|
|
5
5
|
[](https://github.com/roriz/active_genie/actions/workflows/benchmark.yml)
|
|
6
6
|
|
|
7
|
-
ActiveGenie is a Ruby gem that provides valuable solutions powered by Generative AI (GenAI) models. Just like Lodash or ActiveStorage, ActiveGenie brings a set of Modules reach real value fast and
|
|
7
|
+
ActiveGenie is a Ruby gem that provides valuable solutions powered by Generative AI (GenAI) models. Just like Lodash or ActiveStorage, ActiveGenie brings a set of Modules to reach real value fast and reliably.
|
|
8
8
|
ActiveGenie is backed by a custom benchmarking system that ensures consistent quality and performance across different models and providers in every release.
|
|
9
9
|
|
|
10
10
|
## Installation
|
data/VERSION
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
0.0.
|
|
1
|
+
0.0.22
|
metadata
CHANGED
|
@@ -1,121 +1,26 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: active_genie
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.0.
|
|
4
|
+
version: 0.0.22
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Radamés Roriz
|
|
8
|
-
autorequire:
|
|
9
8
|
bindir: bin
|
|
10
9
|
cert_chain: []
|
|
11
|
-
date:
|
|
10
|
+
date: 1980-01-02 00:00:00.000000000 Z
|
|
12
11
|
dependencies: []
|
|
13
12
|
description: "# ActiveGenie \U0001F9DE♂️\n> The lodash for GenAI, stop reinventing
|
|
14
|
-
the wheel\n\
|
|
15
|
-
|
|
16
|
-
|
|
17
|
-
|
|
18
|
-
|
|
19
|
-
|
|
20
|
-
|
|
21
|
-
\
|
|
22
|
-
|
|
23
|
-
|
|
24
|
-
|
|
25
|
-
analysis, handling informal language and complex expressions.\n\n```ruby\ntext =
|
|
26
|
-
\"Nike Air Max 90 - Size 42 - $199.99\"\nschema = {\n brand: { \n type: 'string',\n
|
|
27
|
-
\ enum: [\"Nike\", \"Adidas\", \"Puma\"]\n },\n price: { \n type: 'number',\n
|
|
28
|
-
\ minimum: 0\n },\n size: {\n type: 'number',\n minimum: 35,\n maximum:
|
|
29
|
-
46\n }\n}\n\nresult = ActiveGenie::DataExtractor.call(\n text,\n schema,\n config:
|
|
30
|
-
{ provider: :openai, model: 'gpt-4o-mini' } # optional\n)\n# => { \n# brand:
|
|
31
|
-
\"Nike\", \n# brand_explanation: \"Brand name found at start of text\",\n#
|
|
32
|
-
\ price: 199.99,\n# price_explanation: \"Price found in USD format at end\",\n#
|
|
33
|
-
\ size: 42,\n# size_explanation: \"Size explicitly stated in the middle\"\n#
|
|
34
|
-
\ }\n```\n\n*Recommended model*: `gpt-4o-mini`\n\nFeatures:\n- Structured data
|
|
35
|
-
extraction with type validation\n- Schema-based extraction with custom constraints\n-
|
|
36
|
-
Informal text analysis (litotes, hedging)\n- Detailed explanations for extracted
|
|
37
|
-
values\n\nSee the [Data Extractor README](lib/active_genie/data_extractor/README.md)
|
|
38
|
-
for informal text processing, advanced schemas, and detailed interface documentation.\n\n###
|
|
39
|
-
Scoring\nText evaluation system that provides detailed scoring and feedback using
|
|
40
|
-
multiple expert reviewers. Get balanced scoring through AI-powered expert reviewers
|
|
41
|
-
that automatically adapt to your content.\n\n```ruby\ntext = \"The code implements
|
|
42
|
-
a binary search algorithm with O(log n) complexity\"\ncriteria = \"Evaluate technical
|
|
43
|
-
accuracy and clarity\"\n\nresult = ActiveGenie::Scoring.basic(\n text,\n criteria,\n
|
|
44
|
-
\ config: { provider: :anthropic, model: 'claude-3-5-haiku-20241022' } # optional\n)\n#
|
|
45
|
-
=> {\n# algorithm_expert_score: 95,\n# algorithm_expert_reasoning: \"Accurately
|
|
46
|
-
describes binary search and its complexity\",\n# technical_writer_score: 90,\n#
|
|
47
|
-
\ technical_writer_reasoning: \"Clear and concise explanation of the algorithm\",\n#
|
|
48
|
-
\ final_score: 92.5\n# }\n```\n\n*Recommended model*: `claude-3-5-haiku-20241022`\n\nFeatures:\n-
|
|
49
|
-
Multi-reviewer evaluation with automatic expert selection\n- Detailed feedback with
|
|
50
|
-
scoring reasoning\n- Customizable reviewer weights\n- Flexible evaluation criteria\n\nSee
|
|
51
|
-
the [Scoring README](lib/active_genie/scoring/README.md) for advanced usage, custom
|
|
52
|
-
reviewers, and detailed interface documentation.\n\n### Battle\nAI-powered battle
|
|
53
|
-
evaluation system that determines winners between two players based on specified
|
|
54
|
-
criteria.\n\n```ruby\nrequire 'active_genie'\n\nplayer_1 = \"Implementation uses
|
|
55
|
-
dependency injection for better testability\"\nplayer_2 = \"Code has high test coverage
|
|
56
|
-
but tightly coupled components\"\ncriteria = \"Evaluate code quality and maintainability\"\n\nresult
|
|
57
|
-
= ActiveGenie::Battle.call(\n player_1,\n player_2,\n criteria,\n config: {
|
|
58
|
-
provider: :google, model: 'gemini-2.0-flash-lite' } # optional\n)\n# => {\n# winner_player:
|
|
59
|
-
\"Implementation uses dependency injection for better testability\",\n# reasoning:
|
|
60
|
-
\"Player 1 implementation demonstrates better maintainability through dependency
|
|
61
|
-
injection, \n# which allows for easier testing and component replacement.
|
|
62
|
-
While Player 2 has good test coverage, \n# the tight coupling makes
|
|
63
|
-
the code harder to maintain and modify.\",\n# what_could_be_changed_to_avoid_draw:
|
|
64
|
-
\"Focus on specific architectural patterns and design principles\"\n# }\n```\n\n*Recommended
|
|
65
|
-
model*: `claude-3-5-haiku`\n\nFeatures:\n- Multi-reviewer evaluation with automatic
|
|
66
|
-
expert selection\n- Detailed feedback with scoring reasoning\n- Customizable reviewer
|
|
67
|
-
weights\n- Flexible evaluation criteria\n\nSee the [Battle README](lib/active_genie/battle/README.md)
|
|
68
|
-
for advanced usage, custom reviewers, and detailed interface documentation.\n\n###
|
|
69
|
-
Ranking\nThe Ranking module provides competitive ranking through multi-stage evaluation:\n\n```ruby\nrequire
|
|
70
|
-
'active_genie'\n\nplayers = ['REST API', 'GraphQL API', 'SOAP API', 'gRPC API',
|
|
71
|
-
'Websocket API']\ncriteria = \"Best one to be used into a high changing environment\"\n\nresult
|
|
72
|
-
= ActiveGenie::Ranking.call(\n players,\n criteria,\n config: { provider: :google,
|
|
73
|
-
model: 'gemini-2.0-flash-lite' } # optional\n)\n# => {\n# winner_player: \"gRPC
|
|
74
|
-
API\",\n# reasoning: \"gRPC API is the best one to be used into a high changing
|
|
75
|
-
environment\",\n# }\n```\n\n*Recommended model*: `gemini-2.0-flash-lite`\n\n-
|
|
76
|
-
**Multi-phase ranking system** combining expert scoring and ELO algorithms\n- **Automatic
|
|
77
|
-
elimination** of inconsistent performers using statistical analysis\n- **Dynamic
|
|
78
|
-
ranking adjustments** based on simulated pairwise battles, from bottom to top\n\nSee
|
|
79
|
-
the [Ranking README](lib/active_genie/ranking/README.md) for implementation details,
|
|
80
|
-
configuration, and advanced ranking strategies.\n\n### Text Summarizer (Future)\n###
|
|
81
|
-
Categorizer (Future)\n### Language detector (Future)\n### Translator (Future)\n###
|
|
82
|
-
Sentiment analyzer (Future)\n\n## Benchmarking \U0001F9EA\n\nActiveGenie includes
|
|
83
|
-
a comprehensive benchmarking system to ensure consistent, high-quality outputs across
|
|
84
|
-
different LLM models and providers.\n\n```ruby\n# Run all benchmarks\nbundle exec
|
|
85
|
-
rake active_genie:benchmark\n\n# Run benchmarks for a specific module\nbundle exec
|
|
86
|
-
rake active_genie:benchmark[data_extractor]\n```\n\n### Latest Results\n\n| Model
|
|
87
|
-
| Overall Precision |\n|-------|-------------------|\n| claude-3-5-haiku-20241022
|
|
88
|
-
| 92.25% |\n| gemini-2.0-flash-lite | 84.25% |\n| gpt-4o-mini | 62.75% |\n| deepseek-chat
|
|
89
|
-
| 57.25% |\n\nSee the [Benchmark README](benchmark/README.md) for detailed results,
|
|
90
|
-
methodology, and how to contribute to our test suite.\n\n## Configuration\n\n| Config
|
|
91
|
-
| Description | Default |\n|--------|-------------|---------|\n| `provider` | LLM
|
|
92
|
-
provider (openai, anthropic, etc) | `nil` |\n| `model` | Model to use | `nil` |\n|
|
|
93
|
-
`api_key` | Provider API key | `nil` |\n| `timeout` | Request timeout in seconds
|
|
94
|
-
| `5` |\n| `max_retries` | Maximum retry attempts | `3` |\n\n> **Note:** Each module
|
|
95
|
-
can append its own set of configuration, see the individual module documentation
|
|
96
|
-
for details.\n\n## How to create a new provider\n\nActiveGenie supports adding custom
|
|
97
|
-
providers to integrate with different LLM services. To create a new provider:\n\n1.
|
|
98
|
-
Create a configuration class for your provider in `lib/active_genie/configuration/providers/`:\n\n```ruby\n#
|
|
99
|
-
Example: lib/active_genie/configuration/providers/internal_company_api_config.rb\nmodule
|
|
100
|
-
ActiveGenie\n module Configuration::Providers\n class InternalCompanyApiConfig
|
|
101
|
-
< BaseConfig\n NAME = :internal_company_api\n \n # API key accessor
|
|
102
|
-
with environment variable fallback\n def api_key\n @api_key || ENV['INTERNAL_COMPANY_API_KEY']\n
|
|
103
|
-
\ end\n \n # Base API URL\n def api_url\n @api_url ||
|
|
104
|
-
'https://api.internal-company.com/v1'\n end\n \n # Client instantiation\n
|
|
105
|
-
\ def client\n @client ||= ::ActiveGenie::Clients::InternalCompanyApiClient.new(self)\n
|
|
106
|
-
\ end\n \n # Model tier definitions\n def lower_tier_model\n
|
|
107
|
-
\ @lower_tier_model || 'internal-basic'\n end\n \n def middle_tier_model\n
|
|
108
|
-
\ @middle_tier_model || 'internal-standard'\n end\n \n def
|
|
109
|
-
upper_tier_model\n @upper_tier_model || 'internal-premium'\n end\n end\n
|
|
110
|
-
\ end\nend\n```\n\n2. Register your provider in your configuration:\n\n```ruby\n#
|
|
111
|
-
In config/initializers/active_genie.rb\nActiveGenie.configure do |config|\n # Register
|
|
112
|
-
your custom provider\n config.providers.register(InternalCompanyApi::Configuration)\n
|
|
113
|
-
\ \n # Configure your provider\n config.internal_company_api.api_key = ENV['INTERNAL_COMPANY_API_KEY']\nend\n```\n\n##
|
|
114
|
-
Contributing\n\n1. Fork the repository\n2. Create your feature branch (`git checkout
|
|
115
|
-
-b feature/amazing-feature`)\n3. Commit your changes (`git commit -m 'Add amazing
|
|
116
|
-
feature'`)\n4. Push to the branch (`git push origin feature/amazing-feature`)\n5.
|
|
117
|
-
Open a Pull Request\n\n## License\n\nThis project is licensed under the Apache License
|
|
118
|
-
2.0 License - see the [LICENSE](LICENSE) file for details.\n"
|
|
13
|
+
the wheel\n\nActiveGenie is a Ruby gem that provides valuable solutions powered
|
|
14
|
+
by Generative AI (GenAI) models. Just like Lodash or ActiveStorage, ActiveGenie
|
|
15
|
+
brings a set of Modules to reach real value fast and reliably.\nActiveGenie is backed
|
|
16
|
+
by a custom benchmarking system that ensures consistent quality and performance
|
|
17
|
+
across different models and providers in every release.\n\n## Features\n\n* **Data
|
|
18
|
+
Extractor:** (100% precision) Extract structured data from text using AI-powered
|
|
19
|
+
analysis\n* **Scoring:** (69% precision) Text evaluation system with detailed scoring
|
|
20
|
+
and feedback\n* **Battle:** (100% precision) AI-powered battle evaluation system
|
|
21
|
+
between two players\n* **Ranking:** (100% precision) Multi-phase ranking system
|
|
22
|
+
for multiple options\n\nSee the [GitHub repository](https://github.com/Roriz/active_genie)
|
|
23
|
+
for complete documentation and examples.\n"
|
|
119
24
|
email:
|
|
120
25
|
- radames@roriz.dev
|
|
121
26
|
executables: []
|
|
@@ -175,7 +80,6 @@ metadata:
|
|
|
175
80
|
changelog_uri: https://github.com/Roriz/active_genie/blob/master/CHANGELOG.md
|
|
176
81
|
bug_tracker_uri: https://github.com/Roriz/active_genie/issues
|
|
177
82
|
rubygems_mfa_required: 'true'
|
|
178
|
-
post_install_message:
|
|
179
83
|
rdoc_options: []
|
|
180
84
|
require_paths:
|
|
181
85
|
- lib
|
|
@@ -190,8 +94,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
|
190
94
|
- !ruby/object:Gem::Version
|
|
191
95
|
version: '0'
|
|
192
96
|
requirements: []
|
|
193
|
-
rubygems_version: 3.
|
|
194
|
-
signing_key:
|
|
97
|
+
rubygems_version: 3.6.7
|
|
195
98
|
specification_version: 4
|
|
196
99
|
summary: Transform your Ruby application with powerful, production-ready GenAI features
|
|
197
100
|
test_files: []
|