add Moonshot, debug my_llm (#126)

api/apps/conversation_app.py

@@ -309,13 +309,13 @@ def use_sql(question, field_map, tenant_id, chat_mdl):
     # compose markdown table
     clmns = "|"+"|".join([re.sub(r"(/.*|（[^（）]+）)", "", field_map.get(tbl["columns"][i]["name"], tbl["columns"][i]["name"])) for i in clmn_idx]) + ("|原文|" if docid_idx and docid_idx else "|")
     line = "|"+"|".join(["------" for _ in range(len(clmn_idx))]) + ("|------|" if docid_idx and docid_idx else "")
-    line = re.sub(r"T[0-9]{2}:[0-9]{2}:[0-9]{2}\|", "|", line)
     rows = ["|"+"|".join([rmSpace(str(r[i])) for i in clmn_idx]).replace("None", " ") + "|" for r in tbl["rows"]]
     if not docid_idx or not docnm_idx:
         chat_logger.warning("SQL missing field: " + sql)
         return "\n".join([clmns, line, "\n".join(rows)]), []
 
     rows = "\n".join([r + f" ##{ii}$$ |" for ii, r in enumerate(rows)])
+    rows = re.sub(r"T[0-9]{2}:[0-9]{2}:[0-9]{2}(\.[0-9]+Z)?\|", "|", rows)
     docid_idx = list(docid_idx)[0]
     docnm_idx = list(docnm_idx)[0]
     return "\n".join([clmns, line, rows]), [{"doc_id": r[docid_idx], "docnm_kwd": r[docnm_idx]} for r in tbl["rows"]]

api/apps/llm_app.py

@@ -39,36 +39,40 @@ def factories():
 def set_api_key():
     req = request.json
     # test if api key works
+    chat_passed = False
+    factory = req["llm_factory"]
     msg = ""
-    for llm in LLMService.query(fid=req["llm_factory"]):
+    for llm in LLMService.query(fid=factory):
         if llm.model_type == LLMType.EMBEDDING.value:
-            mdl = EmbeddingModel[req["llm_factory"]](
+            mdl = EmbeddingModel[factory](
                 req["api_key"], llm.llm_name)
             try:
                 arr, tc = mdl.encode(["Test if the api key is available"])
                 if len(arr[0]) == 0 or tc ==0: raise Exception("Fail")
             except Exception as e:
                 msg += f"\nFail to access embedding model({llm.llm_name}) using this api key."
-        elif llm.model_type == LLMType.CHAT.value:
-            mdl = ChatModel[req["llm_factory"]](
+        elif not chat_passed and llm.model_type == LLMType.CHAT.value:
+            mdl = ChatModel[factory](
                 req["api_key"], llm.llm_name)
             try:
                 m, tc = mdl.chat(None, [{"role": "user", "content": "Hello! How are you doing!"}], {"temperature": 0.9})
                 if not tc: raise Exception(m)
+                chat_passed = True
             except Exception as e:
                 msg += f"\nFail to access model({llm.llm_name}) using this api key." + str(e)
 
     if msg: return get_data_error_result(retmsg=msg)
 
     llm = {
-        "tenant_id": current_user.id,
-        "llm_factory": req["llm_factory"],
         "api_key": req["api_key"]
     }
     for n in ["model_type", "llm_name"]:
         if n in req: llm[n] = req[n]
 
-    TenantLLMService.filter_update([TenantLLM.tenant_id==llm["tenant_id"], TenantLLM.llm_factory==llm["llm_factory"]], llm)
+    if not TenantLLMService.filter_update([TenantLLM.tenant_id==current_user.id, TenantLLM.llm_factory==factory], llm):
+        for llm in LLMService.query(fid=factory):
+            TenantLLMService.save(tenant_id=current_user.id, llm_factory=factory, llm_name=llm.llm_name, model_type=llm.model_type, api_key=req["api_key"])
+
     return get_json_result(data=True)
 
 

api/db/db_models.py

@@ -429,7 +429,7 @@ class LLMFactories(DataBaseModel):
 
 class LLM(DataBaseModel):
     # LLMs dictionary
-    llm_name = CharField(max_length=128, null=False, help_text="LLM name", index=True)
+    llm_name = CharField(max_length=128, null=False, help_text="LLM name", index=True, primary_key=True)
     model_type = CharField(max_length=128, null=False, help_text="LLM, Text Embedding, Image2Text, ASR")
     fid = CharField(max_length=128, null=False, help_text="LLM factory id")
     max_tokens = IntegerField(default=0)

api/db/init_data.py

@@ -73,41 +73,41 @@ def init_superuser():
         print("\33[91m【ERROR】\33[0m:", " '{}' dosen't work!".format(tenant["embd_id"]))
 
 
+factory_infos = [{
+        "name": "OpenAI",
+        "logo": "",
+        "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
+        "status": "1",
+    },{
+        "name": "通义千问",
+        "logo": "",
+        "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
+        "status": "1",
+    },{
+        "name": "智谱AI",
+        "logo": "",
+        "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
+        "status": "1",
+    },
+     {
+         "name": "Local",
+         "logo": "",
+         "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
+        "status": "1",
+    },{
+        "name": "Moonshot",
+         "logo": "",
+         "tags": "LLM,TEXT EMBEDDING",
+         "status": "1",
+    }
+    # {
+    #     "name": "文心一言",
+    #     "logo": "",
+    #     "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
+    #     "status": "1",
+    # },
+]
 def init_llm_factory():
-    factory_infos = [{
-            "name": "OpenAI",
-            "logo": "",
-            "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
-            "status": "1",
-        },{
-            "name": "通义千问",
-            "logo": "",
-            "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
-            "status": "1",
-        },{
-            "name": "智谱AI",
-            "logo": "",
-            "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
-            "status": "1",
-        },
-         {
-             "name": "Local",
-             "logo": "",
-             "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
-            "status": "1",
-        },{
-            "name": "Moonshot",
-             "logo": "",
-             "tags": "LLM,TEXT EMBEDDING",
-             "status": "1",
-        }
-        # {
-        #     "name": "文心一言",
-        #     "logo": "",
-        #     "tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
-        #     "status": "1",
-        # },
-    ]
     llm_infos = [
         # ---------------------- OpenAI ------------------------
         {
@@ -260,21 +260,30 @@ def init_llm_factory():
         },
     ]
     for info in factory_infos:
-        LLMFactoriesService.save(**info)
+        try:
+            LLMFactoriesService.save(**info)
+        except Exception as e:
+            pass
     for info in llm_infos:
-        LLMService.save(**info)
+        try:
+            LLMService.save(**info)
+        except Exception as e:
+            pass
 
 
 def init_web_data():
     start_time = time.time()
 
-    if not LLMService.get_all().count():init_llm_factory()
+    if LLMFactoriesService.get_all().count() != len(factory_infos):
+        init_llm_factory()
     if not UserService.get_all().count():
         init_superuser()
 
     print("init web data success:{}".format(time.time() - start_time))
 
 
+
 if __name__ == '__main__':
     init_web_db()
-    init_web_data()
+    init_web_data()
+    add_tenant_llm()

api/db/services/llm_service.py

@@ -53,7 +53,7 @@ class TenantLLMService(CommonService):
             cls.model.used_tokens
         ]
         objs = cls.model.select(*fields).join(LLMFactories, on=(cls.model.llm_factory == LLMFactories.name)).where(
-            cls.model.tenant_id == tenant_id).dicts()
+            cls.model.tenant_id == tenant_id, ~cls.model.api_key.is_null()).dicts()
 
         return list(objs)
 

rag/llm/chat_model.py

@@ -54,6 +54,21 @@ class MoonshotChat(GptTurbo):
         self.client = OpenAI(api_key=key, base_url="https://api.moonshot.cn/v1",)
         self.model_name = model_name
 
+    def chat(self, system, history, gen_conf):
+        if system: history.insert(0, {"role": "system", "content": system})
+        try:
+            response = self.client.chat.completions.create(
+                model=self.model_name,
+                messages=history,
+                **gen_conf)
+            ans = response.choices[0].message.content.strip()
+            if response.choices[0].finish_reason == "length":
+                ans += "...\nFor the content length reason, it stopped, continue?" if is_english(
+                    [ans]) else "······\n由于长度的原因，回答被截断了，要继续吗？"
+            return ans, response.usage.completion_tokens
+        except openai.APIError as e:
+            return "**ERROR**: "+str(e), 0
+
 
 from dashscope import Generation
 class QWenChat(Base):