Extras: Codebase

cognesy · Sep 16, 2024 · 55731cb · 55731cb
1 parent c64dbc7
commit 55731cb
Show file tree

Hide file tree

Showing 163 changed files with 1,252 additions and 429 deletions.
diff --git a/config/embed.php b/config/embed.php
@@ -0,0 +1,97 @@
+<?php
+use Cognesy\Instructor\ApiClient\Enums\ClientType;
+use Cognesy\Instructor\Utils\Env;
+
+return [
+    'defaultConnection' => 'openai',
+    'connections' => [
+        'azure' => [
+            'clientType' => ClientType::Azure->value,
+            'apiUrl' => Env::get('AZURE_OPENAI_BASE_URI', 'openai.azure.com'),
+            'apiKey' => Env::get('AZURE_OPENAI_API_KEY', ''),
+            'endpoint' => Env::get('AZURE_OPENAI_EMBED_ENDPOINT', '/embeddings'),
+            'metadata' => [
+                'apiVersion' => Env::get('AZURE_OPENAI_API_VERSION', '2023-03-15-preview'),
+                'resourceName' => Env::get('AZURE_OPENAI_RESOURCE_NAME', 'instructor-dev'),
+                'deploymentId' => Env::get('AZURE_OPENAI_DEPLOYMENT_NAME', 'gpt-4o-mini'),
+            ],
+            'defaultModel' => Env::get('AZURE_OPENAI_EMBED_MODEL', 'text-embedding-3-small'),
+            'defaultDimensions' => Env::get('AZURE_OPENAI_EMBED_DIMENSIONS', 1536),
+            'maxInputs' => Env::get('AZURE_OPENAI_MAX_INPUTS', 16),
+            'connectTimeout' => Env::get('AZURE_OPENAI_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('AZURE_OPENAI_REQUEST_TIMEOUT', 30),
+        ],
+        'cohere' => [
+            'clientType' => ClientType::Cohere->value,
+            'apiUrl' => Env::get('COHERE_API_URL', 'https://api.cohere.ai/v1'),
+            'apiKey' => Env::get('COHERE_API_KEY', ''),
+            'endpoint' => Env::get('COHERE_EMBED_ENDPOINT', '/embed'),
+            'defaultModel' => Env::get('COHERE_EMBED_MODEL', 'embed-multilingual-v3.0'),
+            'defaultDimensions' => Env::get('COHERE_EMBED_DIMENSIONS', 1024),
+            'maxInputs' => Env::get('COHERE_MAX_INPUTS', 96),
+            'connectTimeout' => Env::get('COHERE_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('COHERE_REQUEST_TIMEOUT', 30),
+        ],
+        'gemini' => [
+            'clientType' => ClientType::Gemini->value,
+            'apiUrl' => Env::get('GEMINI_API_URL', 'https://generativelanguage.googleapis.com/v1beta'),
+            'apiKey' => Env::get('GEMINI_API_KEY', ''),
+            'endpoint' => Env::get('GEMINI_EMBED_ENDPOINT', '/{model}:batchEmbedContents'),
+            'defaultModel' => Env::get('GEMINI_EMBED_MODEL', 'models/text-embedding-004'),
+            'defaultDimensions' => Env::get('GEMINI_EMBED_DIMENSIONS', 768),
+            'maxInputs' => Env::get('GEMINI_MAX_INPUTS', 100), // max 2048 tokens
+            'connectTimeout' => Env::get('GEMINI_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('GEMINI_REQUEST_TIMEOUT', 30),
+        ],
+        'jina' => [
+            'clientType' => ClientType::Jina->value,
+            'apiUrl' => Env::get('JINA_API_URL', 'https://api.jina.ai/v1'),
+            'apiKey' => Env::get('JINA_API_KEY', ''),
+            'endpoint' => Env::get('JINA_EMBED_ENDPOINT', '/embeddings'),
+            'metadata' => [
+                'organization' => ''
+            ],
+            'defaultModel' => Env::get('JINA_EMBED_MODEL', 'jina-embeddings-v2-base-en'),
+            'defaultDimensions' => Env::get('JINA_EMBED_DIMENSIONS', 768),
+            'maxInputs' => Env::get('JINA_MAX_INPUTS', 500), // max 8192 tokens
+            'connectTimeout' => Env::get('JINA_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('JINA_REQUEST_TIMEOUT', 30),
+        ],
+        'mistral' => [
+            'clientType' => ClientType::Mistral->value,
+            'apiUrl' => Env::get('MISTRAL_API_URL', 'https://api.mistral.ai/v1'),
+            'apiKey' => Env::get('MISTRAL_API_KEY', ''),
+            'endpoint' => Env::get('MISTRAL_EMBED_ENDPOINT', '/embeddings'),
+            'defaultModel' => Env::get('MISTRAL_EMBED_MODEL', 'mistral-embed'),
+            'defaultDimensions' => Env::get('MISTRAL_EMBED_DIMENSIONS', 1024),
+            'maxInputs' => Env::get('MISTRAL_MAX_INPUTS', 512), // max 512 tokens
+            'connectTimeout' => Env::get('MISTRAL_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('MISTRAL_REQUEST_TIMEOUT', 30),
+        ],
+        'ollama' => [
+            'clientType' => ClientType::Ollama->value,
+            'apiUrl' => Env::get('OLLAMA_API_URL', 'http://localhost:11434/v1'),
+            'apiKey' => Env::get('OLLAMA_API_KEY', ''),
+            'endpoint' => Env::get('OLLAMA_EMBED_ENDPOINT', '/embeddings'),
+            'defaultModel' => Env::get('OLLAMA_EMBED_MODEL', 'nomic-embed-text'),
+            'defaultDimensions' => Env::get('OLLAMA_EMBED_DIMENSIONS', 1024),
+            'maxInputs' => Env::get('OLLAMA_MAX_INPUTS', 512),
+            'connectTimeout' => Env::get('OLLAMA_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('OLLAMA_REQUEST_TIMEOUT', 30),
+        ],
+        'openai' => [
+            'clientType' => ClientType::OpenAI->value,
+            'apiUrl' => Env::get('OPENAI_API_URL', 'https://api.openai.com/v1'),
+            'apiKey' => Env::get('OPENAI_API_KEY', ''),
+            'endpoint' => Env::get('OPENAI_EMBED_ENDPOINT', '/embeddings'),
+            'metadata' => [
+                'organization' => ''
+            ],
+            'defaultModel' => Env::get('OPENAI_EMBED_MODEL', 'text-embedding-3-small'),
+            'defaultDimensions' => Env::get('OPENAI_EMBED_DIMENSIONS', 1536),
+            'maxInputs' => Env::get('OPENAI_MAX_INPUTS', 2048), // max 8192 tokens
+            'connectTimeout' => Env::get('OPENAI_CONNECT_TIMEOUT', 3),
+            'requestTimeout' => Env::get('OPENAI_REQUEST_TIMEOUT', 30),
+        ],
+    ],
+];
diff --git a/docs/cookbook/examples/api_support/anthropic.mdx b/docs/cookbook/examples/api_support/anthropic.mdx
@@ -20,10 +20,8 @@ Mode compatibility:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Anthropic\AnthropicClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -40,7 +38,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('anthropic');
 
 $user = $instructor->respond(

diff --git a/docs/cookbook/examples/api_support/azure_openai.mdx b/docs/cookbook/examples/api_support/azure_openai.mdx
@@ -17,10 +17,8 @@ using Azure OpenAI service console.
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Azure\AzureClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -37,7 +35,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('azure');
 
 // Call with your model name and preferred execution mode

diff --git a/docs/cookbook/examples/api_support/cohere.mdx b/docs/cookbook/examples/api_support/cohere.mdx
@@ -26,10 +26,8 @@ Reasons Mode::Tools is not recommended:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Cohere\CohereClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -46,7 +44,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('cohere');
 
 $user = $instructor->respond(

diff --git a/docs/cookbook/examples/api_support/fireworks.mdx b/docs/cookbook/examples/api_support/fireworks.mdx
@@ -21,10 +21,8 @@ Mode compatibility:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\FireworksAI\FireworksAIClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -41,10 +39,8 @@ class User {
     public array $hobbies;
 }
 
-// Mistral instance params
-$yourApiKey = Env::get('FIREWORKS_API_KEY'); // set your own API key
-
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('fireworks');
 
 $user = $instructor

diff --git a/docs/cookbook/examples/api_support/google_gemini.mdx b/docs/cookbook/examples/api_support/google_gemini.mdx
@@ -19,10 +19,8 @@ Here's how you can use Instructor with Gemini API.
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Gemini\GeminiClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -39,7 +37,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('gemini');
 
 $user = $instructor

diff --git a/docs/cookbook/examples/api_support/groq.mdx b/docs/cookbook/examples/api_support/groq.mdx
@@ -22,10 +22,8 @@ Here's how you can use Instructor with Groq API.
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Groq\GroqClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -42,10 +40,8 @@ class User {
     public ?int $age;
 }
 
-// Mistral instance params
-$yourApiKey = Env::get('GROQ_API_KEY'); // set your own API key
-
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('groq');
 
 $user = $instructor

diff --git a/docs/cookbook/examples/api_support/mistralai.mdx b/docs/cookbook/examples/api_support/mistralai.mdx
@@ -24,10 +24,8 @@ Mode compatibility:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Mistral\MistralClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -44,10 +42,8 @@ class User {
     public array $hobbies;
 }
 
-// Mistral instance params
-$yourApiKey = Env::get('MISTRAL_API_KEY'); // set your own API key
-
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('mistral');
 
 $user = $instructor

diff --git a/docs/cookbook/examples/api_support/ollama.mdx b/docs/cookbook/examples/api_support/ollama.mdx
@@ -21,7 +21,6 @@ Supported modes:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\Ollama\OllamaClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
 
@@ -40,7 +39,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('ollama');
 
 $user = $instructor->respond(

diff --git a/docs/cookbook/examples/api_support/openai.mdx b/docs/cookbook/examples/api_support/openai.mdx
@@ -20,10 +20,8 @@ Mode compatibility:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\OpenAI\OpenAIClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -40,10 +38,8 @@ class User {
     public array $hobbies;
 }
 
-// OpenAI auth params
-$yourApiKey = Env::get('OPENAI_API_KEY'); // use your own API key
-
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('openai');
 
 $user = $instructor->respond(

diff --git a/docs/cookbook/examples/api_support/openrouter.mdx b/docs/cookbook/examples/api_support/openrouter.mdx
@@ -22,10 +22,8 @@ in case of extraction issues.
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\OpenRouter\OpenRouterClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -42,7 +40,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('openrouter');
 
 $user = $instructor->withDebug()->respond(

diff --git a/docs/cookbook/examples/api_support/togetherai.mdx b/docs/cookbook/examples/api_support/togetherai.mdx
@@ -25,10 +25,8 @@ Mode compatibility:
 $loader = require 'vendor/autoload.php';
 $loader->add('Cognesy\\Instructor\\', __DIR__ . '../../src/');
 
-use Cognesy\Instructor\Clients\TogetherAI\TogetherAIClient;
 use Cognesy\Instructor\Enums\Mode;
 use Cognesy\Instructor\Instructor;
-use Cognesy\Instructor\Utils\Env;
 
 enum UserType : string {
     case Guest = 'guest';
@@ -45,7 +43,8 @@ class User {
     public array $hobbies;
 }
 
-/// Get Instructor with the default client component overridden with your own
+// Get Instructor with specified LLM client connection
+// See: /config/llm.php to check or change LLM client connection configuration details
 $instructor = (new Instructor)->withClient('together');
 
 $user = $instructor

diff --git a/docs/cookbook/examples/basics/basic_use.mdx b/docs/cookbook/examples/basics/basic_use.mdx
@@ -1,6 +1,7 @@
 ---
 title: 'Basic use'
 docname: 'basic_use'
+path: ''
 ---
 
 ## Overview

diff --git a/docs/cookbook/examples/extras/complex_extraction.mdx b/docs/cookbook/examples/extras/complex_extraction.mdx
@@ -98,12 +98,7 @@ enum StakeholderRole: string {
     case Other = 'other';
 }
 
-$client = new OpenAIClient(
-    apiKey: Env::get('OPENAI_API_KEY'),
-    requestTimeout: 90,
-);
-
-$instructor = (new Instructor)->withClient($client);
+$instructor = new Instructor;
 
 echo "PROJECT EVENTS:\n\n";
 

diff --git a/docs/cookbook/examples/extras/complex_extraction_claude.mdx b/docs/cookbook/examples/extras/complex_extraction_claude.mdx
@@ -106,13 +106,6 @@ enum StakeholderRole: string {
     case Other = 'other';
 }
 
-// Create instance of client initialized with custom parameters
-$client = new AnthropicClient(
-    apiKey: Env::get('ANTHROPIC_API_KEY'),
-    requestTimeout: 90,
-);
-
-/// Get Instructor with the default client component overridden with your own
 $instructor = (new Instructor)->withClient('anthropic');
 
 echo "PROJECT EVENTS:\n\n";
@@ -125,7 +118,7 @@ $events = $instructor
         model: 'claude-3-haiku-20240307', //'claude-3-sonnet-20240229',
         prompt: 'Extract a list of project events with all the details from the provided input in JSON format using schema: <|json_schema|>',
         mode: Mode::Json,
-        examples: [['input' => 'Acme Insurance project to implement SalesTech CRM solution is currently in RED status due to delayed delivery of document production system, led by 3rd party vendor - Alfatech. Customer (Acme) is discussing the resolution with the vendor. Production deployment plan has been finalized on Aug 15th and awaiting customer approval.', 'output' => [["type" => "object", "title" => "sequenceOfProjectEvent", "description" => "A sequence of ProjectEvent", "properties" => ["list" => [["title" => "Absorbing delay by deploying extra resources", "description" => "System integrator (SysCorp) are working to absorb some of the delay by deploying extra resources to speed up development when the doc production is done.", "type" => "action", "status" => "open", "stakeholders" => [["name" => "SysCorp", "role" => "system integrator", "details" => "System integrator",],], "date" => "2021-09-01",], ["title" => "Finalization of production deployment plan", "description" => "Production deployment plan has been finalized on Aug 15th and awaiting customer approval.", "type" => "progress", "status" => "open", "stakeholders" => [["name" => "Acme", "role" => "customer", "details" => "Customer",],], "date" => "2021-08-15",],],]]]]],
+//        examples: [['input' => 'Acme Insurance project to implement SalesTech CRM solution is currently in RED status due to delayed delivery of document production system, led by 3rd party vendor - Alfatech. Customer (Acme) is discussing the resolution with the vendor. Production deployment plan has been finalized on Aug 15th and awaiting customer approval.', 'output' => [["type" => "object", "title" => "sequenceOfProjectEvent", "description" => "A sequence of ProjectEvent", "properties" => ["list" => [["title" => "Absorbing delay by deploying extra resources", "description" => "System integrator (SysCorp) are working to absorb some of the delay by deploying extra resources to speed up development when the doc production is done.", "type" => "action", "status" => "open", "stakeholders" => [["name" => "SysCorp", "role" => "system integrator", "details" => "System integrator",],], "date" => "2021-09-01",], ["title" => "Finalization of production deployment plan", "description" => "Production deployment plan has been finalized on Aug 15th and awaiting customer approval.", "type" => "progress", "status" => "open", "stakeholders" => [["name" => "Acme", "role" => "customer", "details" => "Customer",],], "date" => "2021-08-15",],],]]]]],
         options: [
             'max_tokens' => 4096,
             'stream' => true,

diff --git a/docs/cookbook/examples/extras/complex_extraction_cohere.mdx b/docs/cookbook/examples/extras/complex_extraction_cohere.mdx
@@ -98,10 +98,6 @@ enum StakeholderRole: string {
     case Other = 'other';
 }
 
-$client = new CohereClient(
-    apiKey: Env::get('COHERE_API_KEY'),
-);
-
 $instructor = (new Instructor)->withClient('cohere');
 
 echo "PROJECT EVENTS:\n\n";