feat: Update Leva gem to the latest version

Update the Leva gem to the latest stable version to ensure compatibility with the latest dependencies and take advantage of any bug fixes or new features. This update will improve the overall performance and functionality of the application.

README.md

@@ -35,7 +35,7 @@ dataset.records << TextContent.create(text: "I's ok", expected_label: "Neutral")
 Create a run class to handle the execution of your inference logic:
 
 ```bash
-$ rails generate leva:runner sentiment
+rails generate leva:runner sentiment
 ```
 
 ```ruby
@@ -53,7 +53,7 @@ end
 Create one or more eval classes to evaluate the model's output:
 
 ```bash
-$ rails generate leva:eval sentiment_accuracy
+rails generate leva:eval sentiment_accuracy
 ```
 
 ```ruby

lib/generators/leva/templates/eval.rb.erb

@@ -1,8 +1,6 @@
 # frozen_string_literal: true
 
 class <%= class_name %>Eval < Leva::BaseEval
-  leva_dataset_record_class "YourRecordClass"
-
   # @param prediction [String] The prediction to evaluate
   # @param record [YourRecordClass] The record to evaluate
   # @return [Leva::Result] The result of the evaluation
@@ -11,7 +9,6 @@ class <%= class_name %>Eval < Leva::BaseEval
 
     Leva::Result.new(
       label: "<%= file_name.underscore %>",
-      prediction: prediction,
       score: score
     )
   end

test/dummy/app/evals/sentiment_accuracy_eval.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+class SentimentAccuracyEval < Leva::BaseEval
+  # @param prediction [String] The prediction to evaluate
+  # @param record [TextContent] The record to evaluate
+  # @return [Leva::Result] The result of the evaluation
+  def evaluate(prediction, text_content)
+    score = prediction == text_content.expected_label ? 1.0 : 0.0
+
+    Leva::Result.new(
+      label: "sentiment_accuracy",
+      score: score
+    )
+  end
+end

test/dummy/app/runners/sentiment_run.rb

@@ -1,11 +1,20 @@
 # frozen_string_literal: true
 
 class SentimentRun < Leva::BaseRun
-  # @param text [String] The text to analyze
-  # @return [String] The sentiment analysis result
-  def execute(record)
-    # Your model execution logic here
-    # This could involve calling an API, running a local model, etc.
-    # Return the model's output
+  # Executes sentiment analysis on the given text content.
+  #
+  # @param text_content [TextContent] The text to analyze
+  # @return [String] The sentiment analysis result (Positive, Neutral, or Negative)
+  def execute(text_content)
+    text = text_content.text.downcase
+
+    case
+    when text.match?(/\b(love|great|excellent|awesome|fantastic)\b/)
+      "Positive"
+    when text.match?(/\b(hate|terrible|awful|horrible|bad)\b/)
+      "Negative"
+    else
+      "Neutral"
+    end
   end
 end