bug #563 [Ollama] Fix example (dollmetzer)

This PR was squashed before being merged into the main branch.

Discussion
----------

[Ollama] Fix example

As discussed in the hackathon a fix for using different models in ollama (e.g. mistral)
Extended .env variables

Commits
-------

2caf3de Apply suggestions from code review
30aed96 Update examples/ollama/chat-llama.php
80a7e03 catch error with wrong model
64edbc8 catch error with wrong model

Author: OskarStark

examples/.env

@@ -17,6 +17,7 @@ REPLICATE_API_KEY=
 
 # For using Ollama
 OLLAMA_HOST_URL=
+OLLAMA_MODEL=
 
 # For using GPT on Azure
 AZURE_OPENAI_BASEURL=

examples/ollama/chat-llama.php

@@ -18,13 +18,20 @@
 require_once dirname(__DIR__).'/bootstrap.php';
 
 $platform = PlatformFactory::create(env('OLLAMA_HOST_URL'), http_client());
-$model = new Ollama();
+$ollamaModel = $_SERVER['OLLAMA_MODEL'] ?? '';
+$model = new Ollama($ollamaModel);
 
 $agent = new Agent($platform, $model, logger: logger());
 $messages = new MessageBag(
     Message::forSystem('You are a helpful assistant.'),
     Message::ofUser('Tina has one brother and one sister. How many sisters do Tina\'s siblings have?'),
 );
-$result = $agent->call($messages);
 
-echo $result->getContent().\PHP_EOL;
+try {
+    $result = $agent->call($messages);
+    echo $result->getContent().\PHP_EOL;
+} catch(InvalidArgumentException $e) {
+    echo $e->getMessage() . "\nMaybe use a different model?\n";
+}
+
+echo $response;