add Kimi-K2-Instruct from Tongyi-Qianwen API (#9125)

### What problem does this PR solve?

add Kimi-K2-Instruct from Tongyi-Qianwen API

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

conf/llm_factories.json

             "tags": "LLM,TEXT EMBEDDING,TEXT RE-RANK,TTS,SPEECH2TEXT,MODERATION",
             "status": "1",
             "llm": [
+                {
+                    "llm_name": "Moonshot-Kimi-K2-Instruct",
+                    "tags": "LLM,CHAT,128K",
+                    "max_tokens": 131072,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
                 {
                     "llm_name": "deepseek-r1",
                     "tags": "LLM,CHAT,64K",