refactor: image model get_num_tokens override

Author: liuruibin

apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/model/image.py

@@ -1,12 +1,19 @@
 # coding=utf-8
 
-from typing import Dict
+from typing import Dict, List
 
 from langchain_community.chat_models import ChatOpenAI
+from langchain_core.messages import BaseMessage, get_buffer_string
 
+from common.config.tokenizer_manage_config import TokenizerManage
 from setting.models_provider.base_model_provider import MaxKBBaseModel
 
 
+def custom_get_token_ids(text: str):
+    tokenizer = TokenizerManage.get_tokenizer()
+    return tokenizer.encode(text)
+
+
 class QwenVLChatModel(MaxKBBaseModel, ChatOpenAI):
 
     @staticmethod
@@ -21,3 +28,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             **optional_params,
         )
         return chat_tong_yi
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/azure_model_provider/model/image.py

@@ -1,7 +1,7 @@
-from typing import Dict
+from typing import Dict, List
 
+from langchain_core.messages import BaseMessage, get_buffer_string
 from langchain_openai import AzureChatOpenAI
-from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
 from setting.models_provider.base_model_provider import MaxKBBaseModel
@@ -26,3 +26,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/ollama_model_provider/model/image.py

@@ -1,6 +1,7 @@
-from typing import Dict
+from typing import Dict, List
 from urllib.parse import urlparse, ParseResult
 
+from langchain_core.messages import get_buffer_string, BaseMessage
 from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
@@ -36,3 +37,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/openai_model_provider/model/image.py

@@ -1,5 +1,6 @@
-from typing import Dict
+from typing import Dict, List
 
+from langchain_core.messages import BaseMessage, get_buffer_string
 from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
@@ -24,3 +25,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/qwen_model_provider/model/image.py

@@ -1,12 +1,19 @@
 # coding=utf-8
 
-from typing import Dict
+from typing import Dict, List
 
 from langchain_community.chat_models import ChatOpenAI
+from langchain_core.messages import BaseMessage, get_buffer_string
 
+from common.config.tokenizer_manage_config import TokenizerManage
 from setting.models_provider.base_model_provider import MaxKBBaseModel
 
 
+def custom_get_token_ids(text: str):
+    tokenizer = TokenizerManage.get_tokenizer()
+    return tokenizer.encode(text)
+
+
 class QwenVLChatModel(MaxKBBaseModel, ChatOpenAI):
 
     @staticmethod
@@ -21,3 +28,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             **optional_params,
         )
         return chat_tong_yi
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/tencent_model_provider/model/image.py

@@ -1,5 +1,6 @@
-from typing import Dict
+from typing import Dict, List
 
+from langchain_core.messages import BaseMessage, get_buffer_string
 from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
@@ -24,3 +25,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/volcanic_engine_model_provider/model/image.py

@@ -1,5 +1,6 @@
-from typing import Dict
+from typing import Dict, List
 
+from langchain_core.messages import BaseMessage, get_buffer_string
 from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
@@ -24,3 +25,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/xinference_model_provider/model/image.py

@@ -1,5 +1,6 @@
-from typing import Dict
+from typing import Dict, List
 
+from langchain_core.messages import BaseMessage, get_buffer_string
 from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
@@ -24,3 +25,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))

apps/setting/models_provider/impl/zhipu_model_provider/model/image.py

@@ -1,5 +1,6 @@
-from typing import Dict
+from typing import Dict, List
 
+from langchain_core.messages import get_buffer_string, BaseMessage
 from langchain_openai.chat_models import ChatOpenAI
 
 from common.config.tokenizer_manage_config import TokenizerManage
@@ -24,3 +25,17 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
             streaming=True,
             **optional_params,
         )
+
+    def get_num_tokens_from_messages(self, messages: List[BaseMessage]) -> int:
+        try:
+            return super().get_num_tokens_from_messages(messages)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return sum([len(tokenizer.encode(get_buffer_string([m]))) for m in messages])
+
+    def get_num_tokens(self, text: str) -> int:
+        try:
+            return super().get_num_tokens(text)
+        except Exception as e:
+            tokenizer = TokenizerManage.get_tokenizer()
+            return len(tokenizer.encode(text))