vectorvein 0.1.70__tar.gz → 0.1.71__tar.gz

{vectorvein-0.1.70 → vectorvein-0.1.71}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vectorvein
-Version: 0.1.70
+Version: 0.1.71
 Summary: VectorVein python SDK
 Author-Email: Anderson <andersonby@163.com>
 License: MIT

{vectorvein-0.1.70 → vectorvein-0.1.71}/pyproject.toml

@@ -17,7 +17,7 @@ description = "VectorVein python SDK"
 name = "vectorvein"
 readme = "README.md"
 requires-python = ">=3.10"
-version = "0.1.70"
+version = "0.1.71"
 
 [project.license]
 text = "MIT"

{vectorvein-0.1.70 → vectorvein-0.1.71}/src/vectorvein/chat_clients/utils.py

@@ -173,7 +173,7 @@ def get_token_counts(text: str | dict, model: str = "", use_token_server_first:
             return 1000
         result = response.json()
         return result["segments_num"]
-    elif model in ("moonshot-v1-8k", "moonshot-v1-32k", "moonshot-v1-128k"):
+    elif model.startswith("moonshot"):
         model_setting = settings.moonshot.models[model]
         if len(model_setting.endpoints) == 0:
             return len(get_gpt_35_encoding().encode(text))
@@ -294,6 +294,33 @@ def get_token_counts(text: str | dict, model: str = "", use_token_server_first:
             return 1000
         result = response.json()
         return result["data"]["total_tokens"]
+    elif model.startswith("glm"):
+        model_setting = settings.zhipuai.models[model]
+        if len(model_setting.endpoints) == 0:
+            return len(get_gpt_35_encoding().encode(text))
+        endpoint_id = model_setting.endpoints[0]
+        if isinstance(endpoint_id, dict):
+            endpoint_id = endpoint_id["endpoint_id"]
+        endpoint = settings.get_endpoint(endpoint_id)
+        tokenize_url = f"{endpoint.api_base}/tokenizer"
+        headers = {"Content-Type": "application/json", "Authorization": f"Bearer {endpoint.api_key}"}
+        request_body = {
+            "model": model,
+            "messages": [
+                {"role": "user", "content": text},
+            ],
+        }
+        _, response = (
+            Retry(httpx.post)
+            .args(url=tokenize_url, headers=headers, json=request_body, timeout=None)
+            .retry_times(5)
+            .sleep_time(10)
+            .run()
+        )
+        if response is None:
+            return 1000
+        result = response.json()
+        return result["usage"]["prompt_tokens"]
     else:
         return len(get_gpt_35_encoding().encode(text))
 

{vectorvein-0.1.70 → vectorvein-0.1.71}/src/vectorvein/types/defaults.py

@@ -44,13 +44,6 @@ DEEPSEEK_MODELS: Final[Dict[str, Dict[str, Any]]] = {
         "function_call_available": True,
         "response_format_available": True,
     },
-    "deepseek-coder": {
-        "id": "deepseek-chat",
-        "context_length": 128000,
-        "max_output_tokens": 4096,
-        "function_call_available": True,
-        "response_format_available": True,
-    },
 }
 DEEPSEEK_DEFAULT_MODEL: Final[str] = "deepseek-chat"
 
@@ -427,6 +420,22 @@ ZHIPUAI_MODELS: Final[Dict[str, Dict[str, Any]]] = {
         "max_output_tokens": 1024,
         "native_multimodal": True,
     },
+    "glm-4v-flash": {
+        "id": "glm-4v-flash",
+        "context_length": 2000,
+        "function_call_available": False,
+        "response_format_available": False,
+        "max_output_tokens": 1024,
+        "native_multimodal": True,
+    },
+    "glm-zero-preview": {
+        "id": "glm-zero-preview",
+        "context_length": 16000,
+        "function_call_available": False,
+        "response_format_available": False,
+        "max_output_tokens": 16000,
+        "native_multimodal": True,
+    },
 }
 
 # Mistral models
@@ -669,7 +678,7 @@ GEMINI_MODELS: Final[Dict[str, Dict[str, Any]]] = {
     },
     "gemini-2.0-flash-thinking-exp-1219": {
         "id": "gemini-2.0-flash-thinking-exp-1219",
-        "context_length": 1048576,
+        "context_length": 32767,
         "max_output_tokens": 8192,
         "function_call_available": True,
         "response_format_available": True,
@@ -677,7 +686,7 @@ GEMINI_MODELS: Final[Dict[str, Dict[str, Any]]] = {
     },
     "gemini-exp-1206": {
         "id": "gemini-exp-1206",
-        "context_length": 32767,
+        "context_length": 2097152,
         "function_call_available": True,
         "response_format_available": True,
         "native_multimodal": True,