pkg/cli/cmd.go

@@ -64,11 +64,12 @@ func LLMRequest(messages []Message) (string, error) {
 	go HandleDelayedContent(receiver)
 
 	response, err := CreateChatCompletionStream(model, messages, maxTokens, receiver)
+	if err != nil && response != "" {
+		Warn("Received partial response. Error: %v\n", err)
+		err = nil // ignore partial response error
+	}
+
 	if response != "" {
-		if err != nil {
-			Warn("Received partial response. Error: %v\n", err)
-			err = nil
-		}
 		// there was some content, so break to a new line after it
 		fmt.Println()
 	}

pkg/cli/openai.go

@@ -22,7 +22,6 @@ func CreateChatCompletionRequest(model string, messages []Message, maxTokens int
 		Model:     model,
 		Messages:  chatCompletionMessages,
 		MaxTokens: maxTokens,
-		N:         1, // limit responses to 1 "choice". we use choices[0] to reference it
 	}
 }