pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

llm configuation refine and trievalTest API refine (#40)

Browse Source
This commit is contained in:
KevinHuSh
2024-01-19 19:51:57 +08:00
committed by GitHub
parent f3dd131403
commit 484e5abc1f
39 changed files with 160 additions and 121 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
api/db/__init__.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

11
api/db/db_models.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
@ -426,8 +426,8 @@ class LLMFactories(DataBaseModel):
class LLM(DataBaseModel):
# defautlt LLMs for every users
llm_name = CharField(max_length=128, null=False, help_text="LLM name", primary_key=True)
# LLMs dictionary
llm_name = CharField(max_length=128, null=False, help_text="LLM name", index=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)
@ -448,6 +448,7 @@ class TenantLLM(DataBaseModel):
llm_name = CharField(max_length=128, null=True, help_text="LLM name", default="")
api_key = CharField(max_length=255, null=True, help_text="API KEY")
api_base = CharField(max_length=255, null=True, help_text="API Base")
used_tokens = IntegerField(default=0)
def __str__(self):
return self.llm_name
@ -468,8 +469,8 @@ class Knowledgebase(DataBaseModel):
doc_num = IntegerField(default=0)
token_num = IntegerField(default=0)
chunk_num = IntegerField(default=0)
#similarity_threshold = FloatField(default=0.4)
#vector_similarity_weight = FloatField(default=0.3)
similarity_threshold = FloatField(default=0.4)
vector_similarity_weight = FloatField(default=0.3)
parser_id = CharField(max_length=32, null=False, help_text="default parser ID")
status = CharField(max_length=1, null=True, help_text="is it validate(0: wasted，1: validate)", default="1")

2
api/db/db_utils.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

31
api/db/init_data.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
@ -46,6 +46,11 @@ def init_llm_factory():
"logo": "",
"tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
"status": "1",
},{
"name": "Infiniflow",
"logo": "",
"tags": "LLM,TEXT EMBEDDING,SPEECH2TEXT,MODERATION",
"status": "1",
},{
"name": "智普AI",
"logo": "",
@ -130,6 +135,30 @@ def init_llm_factory():
"tags": "LLM,CHAT,IMAGE2TEXT",
"max_tokens": 765,
"model_type": LLMType.IMAGE2TEXT.value
},{
"fid": factory_infos[2]["name"],
"llm_name": "gpt-3.5-turbo",
"tags": "LLM,CHAT,4K",
"max_tokens": 4096,
"model_type": LLMType.CHAT.value
},{
"fid": factory_infos[2]["name"],
"llm_name": "text-embedding-ada-002",
"tags": "TEXT EMBEDDING,8K",
"max_tokens": 8191,
"model_type": LLMType.EMBEDDING.value
},{
"fid": factory_infos[2]["name"],
"llm_name": "whisper-1",
"tags": "SPEECH2TEXT",
"max_tokens": 25*1024*1024,
"model_type": LLMType.SPEECH2TEXT.value
},{
"fid": factory_infos[2]["name"],
"llm_name": "gpt-4-vision-preview",
"tags": "LLM,CHAT,IMAGE2TEXT",
"max_tokens": 765,
"model_type": LLMType.IMAGE2TEXT.value
},
]
for info in factory_infos:

2
api/db/operatioins.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/reload_config_base.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/runtime_config.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/services/__init__.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/services/common_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/services/dialog_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/services/document_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/services/kb_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
api/db/services/knowledgebase_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

46
api/db/services/llm_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
@ -13,6 +13,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.
#
from api.db.services.user_service import TenantService
from rag.llm import EmbeddingModel, CvModel
from api.db import LLMType
from api.db.db_models import DB, UserTenant
@ -34,40 +35,39 @@ class TenantLLMService(CommonService):
@classmethod
@DB.connection_context()
def get_api_key(cls, tenant_id, model_type, model_name=""):
objs = cls.query(tenant_id=tenant_id, model_type=model_type)
if objs and len(objs)>0 and objs[0].llm_name:
return objs[0]
fields = [LLM.llm_name, cls.model.llm_factory, cls.model.api_key]
objs = cls.model.select(*fields).join(LLM, on=(LLM.fid == cls.model.llm_factory)).where(
(cls.model.tenant_id == tenant_id),
((cls.model.model_type == model_type) | (cls.model.llm_name == model_name)),
(LLM.status == StatusEnum.VALID)
)
if not objs:return
def get_api_key(cls, tenant_id, model_name):
objs = cls.query(tenant_id=tenant_id, llm_name=model_name)
if not objs: return
return objs[0]
@classmethod
@DB.connection_context()
def get_my_llms(cls, tenant_id):
fields = [cls.model.llm_factory, LLMFactories.logo, LLMFactories.tags, cls.model.model_type, cls.model.llm_name]
objs = cls.model.select(*fields).join(LLMFactories, on=(cls.model.llm_factory==LLMFactories.name)).where(cls.model.tenant_id==tenant_id).dicts()
objs = cls.model.select(*fields).join(LLMFactories, on=(cls.model.llm_factory == LLMFactories.name)).where(
cls.model.tenant_id == tenant_id).dicts()
return list(objs)
@classmethod
@DB.connection_context()
def model_instance(cls, tenant_id, llm_type):
model_config = cls.get_api_key(tenant_id, model_type=LLMType.EMBEDDING.value)
if not model_config:
model_config = {"llm_factory": "local", "api_key": "", "llm_name": ""}
else:
model_config = model_config[0].to_dict()
if llm_type == LLMType.EMBEDDING:
e,tenant = TenantService.get_by_id(tenant_id)
if not e: raise LookupError("Tenant not found")
if llm_type == LLMType.EMBEDDING.value: mdlnm = tenant.embd_id
elif llm_type == LLMType.SPEECH2TEXT.value: mdlnm = tenant.asr_id
elif llm_type == LLMType.IMAGE2TEXT.value: mdlnm = tenant.img2txt_id
elif llm_type == LLMType.CHAT.value: mdlnm = tenant.llm_id
else: assert False, "LLM type error"
model_config = cls.get_api_key(tenant_id, mdlnm)
if not model_config: raise LookupError("Model({}) not found".format(mdlnm))
model_config = model_config[0].to_dict()
if llm_type == LLMType.EMBEDDING.value:
if model_config["llm_factory"] not in EmbeddingModel: return
return EmbeddingModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"])
if llm_type == LLMType.IMAGE2TEXT:
if llm_type == LLMType.IMAGE2TEXT.value:
if model_config["llm_factory"] not in CvModel: return
return CvModel[model_config.llm_factory](model_config["api_key"], model_config["llm_name"])
return CvModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"])

2
api/db/services/user_service.py
View File

@ -1,5 +1,5 @@
#
# Copyright 2019 The InfiniFlow Authors. All Rights Reserved.
# Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.