support multi modal inputs

Author: learner0810

pkg/epp/scheduling/types/types.go

@@ -17,6 +17,7 @@ limitations under the License.
 package types
 
 import (
+	"encoding/json"
 	"fmt"
 
 	"sigs.k8s.io/gateway-api-inference-extension/pkg/epp/backend"
@@ -97,7 +98,8 @@ func (r *ChatCompletionsRequest) String() string {
 
 	messagesLen := 0
 	for _, msg := range r.Messages {
-		messagesLen += len(msg.Content)
+		data, _ := json.Marshal(msg.Content)
+		messagesLen += len(data)
 	}
 
 	return fmt.Sprintf("{MessagesLength: %d}", messagesLen)
@@ -106,7 +108,7 @@ func (r *ChatCompletionsRequest) String() string {
 // Message represents a single message in a chat-completions request.
 type Message struct {
 	Role    string
-	Content string // TODO: support multi-modal content
+	Content any
 }
 
 type Pod interface {

pkg/epp/util/request/body_test.go

@@ -64,6 +64,36 @@ func TestExtractRequestData(t *testing.T) {
 				},
 			},
 		},
+		{
+			name: "chat completions request body with multi-modal content",
+			body: map[string]any{
+				"model": "test",
+				"messages": []any{
+					map[string]any{
+						"role": "system",
+						"content": map[string]any{
+							"type": "text",
+							"text": "Describe this image in one sentence.",
+						},
+					},
+					map[string]any{
+						"role": "user",
+						"content": map[string]any{
+							"type":      "image_url",
+							"image_url": "https://example.com/images/dui.jpg.",
+						},
+					},
+				},
+			},
+			want: &types.LLMRequestBody{
+				ChatCompletions: &types.ChatCompletionsRequest{
+					Messages: []types.Message{
+						{Role: "system", Content: map[string]any{"type": "text", "text": "Describe this image in one sentence."}},
+						{Role: "user", Content: map[string]any{"type": "image_url", "image_url": "https://example.com/images/dui.jpg."}},
+					},
+				},
+			},
+		},
 		{
 			name: "chat completions with all optional fields",
 			body: map[string]any{