refine error response, add set api-key MD (#178)

api/apps/conversation_app.py

@@ -253,6 +253,8 @@ def chat(dialog, messages, **kwargs):
     for c in kbinfos["chunks"]:
         if c.get("vector"):
             del c["vector"]
+    if answer.lower().find("invalid key") >= 0 or answer.lower().find("invalid api")>=0:
+        answer += " Please set LLM API-Key in 'User Setting -> Model Providers -> API-Key'"
     return {"answer": answer, "reference": kbinfos}
 
 

api/apps/kb_app.py

@@ -13,10 +13,12 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+from elasticsearch_dsl import Q
 from flask import request
 from flask_login import login_required, current_user
 
 from api.db.services import duplicate_name
+from api.db.services.document_service import DocumentService
 from api.db.services.user_service import TenantService, UserTenantService
 from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
 from api.utils import get_uuid, get_format_time
@@ -25,6 +27,8 @@ from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.db_models import Knowledgebase
 from api.settings import stat_logger, RetCode
 from api.utils.api_utils import get_json_result
+from rag.nlp import search
+from rag.utils import ELASTICSEARCH
 
 
 @manager.route('/create', methods=['post'])
@@ -125,11 +129,22 @@ def list():
 def rm():
     req = request.json
     try:
-        if not KnowledgebaseService.query(
+        kbs = KnowledgebaseService.query(
-                created_by=current_user.id, id=req["kb_id"]):
+                created_by=current_user.id, id=req["kb_id"])
+        if not kbs:
             return get_json_result(
                 data=False, retmsg=f'Only owner of knowledgebase authorized for this operation.', retcode=RetCode.OPERATING_ERROR)
 
+        for doc in DocumentService.query(kb_id=req["kb_id"]):
+            ELASTICSEARCH.deleteByQuery(
+                Q("match", doc_id=doc.id), idxnm=search.index_name(kbs[0].tenant_id))
+
+            DocumentService.increment_chunk_num(
+                doc.id, doc.kb_id, doc.token_num * -1, doc.chunk_num * -1, 0)
+            if not DocumentService.delete(doc):
+                return get_data_error_result(
+                    retmsg="Database error (Document removal)!")
+
         if not KnowledgebaseService.update_by_id(
                 req["kb_id"], {"status": StatusEnum.INVALID.value}):
             return get_data_error_result(

api/db/services/knowledgebase_service.py

@@ -62,7 +62,7 @@ class KnowledgebaseService(CommonService):
         if not kbs:
             return
         d = kbs[0].to_dict()
-        d["embd_id"] = kbs[0].tenant.embd_id
+        #d["embd_id"] = kbs[0].tenant.embd_id
         return d
 
     @classmethod

api/utils/api_utils.py

@@ -149,6 +149,9 @@ def server_error_response(e):
     if len(e.args) > 1:
         return get_json_result(
             retcode=RetCode.EXCEPTION_ERROR, retmsg=repr(e.args[0]), data=e.args[1])
+    if repr(e).find("index_not_found_exception") >=0:
+        return get_json_result(retcode=RetCode.EXCEPTION_ERROR, retmsg="No chunk found, please upload file and parse it.")
+
     return get_json_result(retcode=RetCode.EXCEPTION_ERROR, retmsg=repr(e))
 
 

docker/docker-compose-CN.yml

@@ -0,0 +1,133 @@
+version: '2.2'
+services:
+  es01:
+    container_name: ragflow-es-01
+    image: docker.elastic.co/elasticsearch/elasticsearch:${STACK_VERSION}
+    volumes:
+      - esdata01:/usr/share/elasticsearch/data
+    ports:
+      - ${ES_PORT}:9200
+    environment:
+      - node.name=es01
+      - cluster.name=${CLUSTER_NAME}
+      - cluster.initial_master_nodes=es01
+      - ELASTIC_PASSWORD=${ELASTIC_PASSWORD}
+      - bootstrap.memory_lock=false
+      - xpack.security.enabled=false
+      - TZ=${TIMEZONE}
+    mem_limit: ${MEM_LIMIT}
+    ulimits:
+      memlock:
+        soft: -1
+        hard: -1
+    healthcheck:
+      test: ["CMD-SHELL", "curl http://localhost:9200"]
+      interval: 10s
+      timeout: 10s
+      retries: 120
+    networks:
+      - ragflow
+    restart: always
+
+  kibana:
+    depends_on:
+        es01:
+          condition: service_healthy
+    image: docker.elastic.co/kibana/kibana:${STACK_VERSION}
+    container_name: ragflow-kibana
+    volumes:
+      - kibanadata:/usr/share/kibana/data
+    ports:
+      - ${KIBANA_PORT}:5601
+    environment:
+      - SERVERNAME=kibana
+      - ELASTICSEARCH_HOSTS=http://es01:9200
+      - TZ=${TIMEZONE}
+    mem_limit: ${MEM_LIMIT}
+    networks:
+      - ragflow
+
+  mysql:
+    image: mysql:5.7.18
+    container_name: ragflow-mysql
+    environment:
+      - MYSQL_ROOT_PASSWORD=${MYSQL_PASSWORD}
+      - TZ=${TIMEZONE}
+    command:
+      --max_connections=1000
+      --character-set-server=utf8mb4
+      --collation-server=utf8mb4_general_ci
+      --default-authentication-plugin=mysql_native_password
+      --tls_version="TLSv1.2,TLSv1.3"
+      --init-file /data/application/init.sql
+    ports:
+      - ${MYSQL_PORT}:3306
+    volumes:
+      - mysql_data:/var/lib/mysql
+      - ./init.sql:/data/application/init.sql
+    networks:
+      - ragflow
+    healthcheck:
+      test: ["CMD", "mysqladmin" ,"ping", "-uroot", "-p${MYSQL_PASSWORD}"]
+      interval: 10s
+      timeout: 10s
+      retries: 3
+    restart: always
+
+
+  minio:
+    image: quay.io/minio/minio:RELEASE.2023-12-20T01-00-02Z
+    container_name: ragflow-minio
+    command: server --console-address ":9001" /data
+    ports:
+      - 9000:9000
+      - 9001:9001
+    environment:
+      - MINIO_ROOT_USER=${MINIO_USER}
+      - MINIO_ROOT_PASSWORD=${MINIO_PASSWORD}
+      - TZ=${TIMEZONE}
+    volumes:
+      - minio_data:/data
+    networks:
+      - ragflow
+    restart: always
+
+
+  ragflow:
+    depends_on:
+      mysql:
+        condition: service_healthy
+      es01:
+        condition: service_healthy
+    image: swr.cn-north-4.myhuaweicloud.com/infiniflow/ragflow:v1.0
+    container_name: ragflow-server
+    ports:
+      - ${SVR_HTTP_PORT}:9380
+      - 80:80
+      - 443:443
+    volumes:
+      - ./service_conf.yaml:/ragflow/conf/service_conf.yaml
+      - ./ragflow-logs:/ragflow/logs
+      - ./nginx/ragflow.conf:/etc/nginx/conf.d/ragflow.conf
+      - ./nginx/proxy.conf:/etc/nginx/proxy.conf
+      - ./nginx/nginx.conf:/etc/nginx/nginx.conf
+    environment:
+      - TZ=${TIMEZONE}
+    networks:
+      - ragflow
+    restart: always
+
+
+volumes:
+  esdata01:
+    driver: local
+  kibanadata:
+    driver: local
+  mysql_data:
+    driver: local
+  minio_data:
+    driver: local
+
+networks:
+  ragflow:
+    driver: bridge

docs/llm_api_key_setup.md

@@ -0,0 +1,19 @@
+
+## Set Before Starting The System
+
+In **user_default_llm** of [service_conf.yaml](./docker/service_conf.yaml), you need to specify LLM factory and your own _API_KEY_. 
+RagFlow supports the flowing LLM factory, and with more coming in the pipeline:
+
+> [OpenAI](https://platform.openai.com/login?launch), [Tongyi-Qianwen](https://dashscope.console.aliyun.com/model), 
+> [ZHIPU-AI](https://open.bigmodel.cn/), [Moonshot](https://platform.moonshot.cn/docs/docs)
+
+After sign in these LLM suppliers, create your own API-Key, they all have a certain amount of free quota.
+
+## After Starting The System
+
+You can also set API-Key in **User Setting** as following:
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/infiniflow/ragflow/assets/12318111/e4e4066c-e964-45ff-bd56-c3fc7fb18bd3" width="1000"/>
+</div>
+

rag/llm/chat_model.py

@@ -156,7 +156,7 @@ class LocalLLM(Base):
 
             return do_rpc
 
-    def __init__(self, **kwargs):
+    def __init__(self, *args, **kwargs):
         self.client = LocalLLM.RPCProxy("127.0.0.1", 7860)
 
     def chat(self, system, history, gen_conf):

rag/llm/rpc_server.py

@@ -7,6 +7,23 @@ from threading import Thread
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
 
+def torch_gc():
+    try:
+        import torch
+        if torch.cuda.is_available():
+            # with torch.cuda.device(DEVICE):
+            torch.cuda.empty_cache()
+            torch.cuda.ipc_collect()
+        elif torch.backends.mps.is_available():
+            try:
+                from torch.mps import empty_cache
+                empty_cache()
+            except Exception as e:
+                pass
+    except Exception:
+        pass
+
+
 class RPCHandler:
     def __init__(self):
         self._functions = {}
@@ -49,6 +66,7 @@ def chat(messages, gen_conf):
     global tokenizer
     model = Model()
     try:
+        torch_gc()
         conf = {
             "max_new_tokens": int(
                 gen_conf.get(