Merge remote-tracking branch 'origin/main' into dbgpt_api

docs/getting_started/faq/deploy/deploy_faq.md

@@ -2,26 +2,26 @@ Installation FAQ
 ==================================
 
 
-##### Q1: execute `pip install -r requirements.txt` error, found some package cannot find correct version.
+##### Q1: execute `pip install -e .` error, found some package cannot find correct version.
 change the pip source.
 
 ```bash
 # pypi
-$ pip install -r requirements.txt -i https://pypi.python.org/simple
+$ pip install -e . -i https://pypi.python.org/simple
 ```
 
 or
 
 ```bash
 # tsinghua
-$ pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple/
+$ pip install -e . -i https://pypi.tuna.tsinghua.edu.cn/simple/
 ```
 
 or
 
 ```bash
 # aliyun
-$ pip install -r requirements.txt -i http://mirrors.aliyun.com/pypi/simple/
+$ pip install -e . -i http://mirrors.aliyun.com/pypi/simple/
 ```
 
 ##### Q2: sqlalchemy.exc.OperationalError: (sqlite3.OperationalError) unable to open database file 
@@ -29,5 +29,20 @@ $ pip install -r requirements.txt -i http://mirrors.aliyun.com/pypi/simple/
 make sure you pull latest code or create directory with mkdir pilot/data
 
 ##### Q3: The model keeps getting killed.
+
 your GPU VRAM size is not enough, try replace your hardware or replace other llms.
 
+##### Q4: How to access website on the public network
+
+You can try to use gradio's [network](https://github.com/gradio-app/gradio/blob/main/gradio/networking.py) to achieve.
+```python
+import secrets
+from gradio import networking
+token=secrets.token_urlsafe(32)
+local_port=5000
+url = networking.setup_tunnel('0.0.0.0', local_port, token)
+print(f'Public url: {url}')
+time.sleep(60 * 60 * 24)
+```
+
+Open `url` with your browser to see the website.

docs/getting_started/install/deploy/deploy.md

@@ -49,7 +49,7 @@ For the entire installation process of DB-GPT, we use the miniconda3 virtual env
 python>=3.10
 conda create -n dbgpt_env python=3.10
 conda activate dbgpt_env
-pip install -r requirements.txt
+pip install -e .
 ```
 Before use DB-GPT Knowledge
 ```bash
@@ -97,15 +97,20 @@ You can configure basic parameters in the .env file, for example setting LLM_MOD
 
 ### 3. Run
 
+**(Optional) load examples into SQLlite**
+```bash
+bash ./scripts/examples/load_examples.sh
+```
+
 1.Run db-gpt server 
 
 ```bash
-$ python pilot/server/dbgpt_server.py
+python pilot/server/dbgpt_server.py
 ```
 
 Open http://localhost:5000 with your browser to see the product.
 
-```tip
+```{tip}
 If you want to access an external LLM service, you need to
 
 1.set the variables LLM_MODEL=YOUR_MODEL_NAME, MODEL_SERVER=YOUR_MODEL_SERVER（eg:http://localhost:5000） in the .env file.
@@ -116,7 +121,7 @@ If you want to access an external LLM service, you need to
 If you want to learn about dbgpt-webui, read https://github./csunny/DB-GPT/tree/new-page-framework/datacenter
 
 ```bash
-$ python pilot/server/dbgpt_server.py --light
+python pilot/server/dbgpt_server.py --light
 ```
 
 ### Multiple GPUs
@@ -141,6 +146,4 @@ DB-GPT supported 8-bit quantization and 4-bit quantization.
 
 You can modify the setting `QUANTIZE_8bit=True` or `QUANTIZE_4bit=True` in `.env` file to use quantization(8-bit quantization is enabled by default).
 
-Llama-2-70b with 8-bit quantization can run with 80 GB of VRAM, and 4-bit quantization can run with 48 GB of VRAM.
-
-Note: you need to install the latest dependencies according to [requirements.txt](https://github.com/eosphoros-ai/DB-GPT/blob/main/requirements.txt).
+Llama-2-70b with 8-bit quantization can run with 80 GB of VRAM, and 4-bit quantization can run with 48 GB of VRAM.

docs/getting_started/install/llm/llama/llama_cpp.md

@@ -1,6 +1,8 @@
-### llama.cpp
+llama.cpp
+==================================
 
-DB-GPT is now supported by [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) through [llama.cpp](https://github.com/ggerganov/llama.cpp).
+
+DB-GPT already supports [llama.cpp](https://github.com/ggerganov/llama.cpp) via [llama-cpp-python](https://github.com/abetlen/llama-cpp-python).
 
 ## Running llama.cpp
 

pilot/connections/rdbms/conn_sqlite.py

@@ -1,6 +1,7 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 
+import os
 from typing import Optional, Any, Iterable
 from sqlalchemy import create_engine, text
 
@@ -24,6 +25,9 @@ class SQLiteConnect(RDBMSDatabase):
         _engine_args = engine_args or {}
         _engine_args["connect_args"] = {"check_same_thread": False}
         # _engine_args["echo"] = True
+        directory = os.path.dirname(file_path)
+        if not os.path.exists(directory):
+            os.makedirs(directory)
         return cls(create_engine("sqlite:///" + file_path, **_engine_args), **kwargs)
 
     def get_indexes(self, table_name):

pilot/model/conversation.py

@@ -335,6 +335,20 @@ register_conv_template(
     )
 )
 
+# Alpaca default template
+register_conv_template(
+    Conversation(
+        name="alpaca",
+        system="Below is an instruction that describes a task. Write a response that appropriately completes the request.",
+        roles=("### Instruction", "### Response"),
+        messages=(),
+        offset=0,
+        sep_style=SeparatorStyle.ADD_COLON_TWO,
+        sep="\n\n",
+        sep2="</s>",
+    )
+)
+
 # Baichuan-13B-Chat template
 register_conv_template(
     # source: https://huggingface.co/baichuan-inc/Baichuan-13B-Chat/blob/f5f47be2adbbdceb784f334d6fa1ca2c73e65097/modeling_baichuan.py#L507

scripts/examples/load_examples.bat

@@ -0,0 +1,68 @@
+@echo off
+setlocal
+
+:: Get script location and set working directory
+for %%i in (%0) do set SCRIPT_LOCATION=%%~dpi
+cd %SCRIPT_LOCATION%
+cd ..
+cd ..
+set WORK_DIR=%CD%
+
+:: Check if sqlite3 is installed
+where sqlite3 >nul 2>nul
+if %ERRORLEVEL% neq 0 (
+    echo sqlite3 not found, please install sqlite3
+    exit /b 1
+)
+
+:: Default file paths
+set DEFAULT_DB_FILE=DB-GPT\pilot\data\default_sqlite.db
+set DEFAULT_SQL_FILE=DB-GPT\docker\examples\sqls\*_sqlite.sql
+set DB_FILE=%WORK_DIR%\pilot\data\default_sqlite.db
+set SQL_FILE=
+
+:argLoop
+if "%1"=="" goto argDone
+if "%1"=="-d" goto setDBFile
+if "%1"=="--db-file" goto setDBFile
+if "%1"=="-f" goto setSQLFile
+if "%1"=="--sql-file" goto setSQLFile
+if "%1"=="-h" goto printUsage
+if "%1"=="--help" goto printUsage
+goto argError
+
+:setDBFile
+shift
+set DB_FILE=%1
+shift
+goto argLoop
+
+:setSQLFile
+shift
+set SQL_FILE=%1
+shift
+goto argLoop
+
+:argError
+echo Invalid argument: %1
+goto printUsage
+
+:printUsage
+echo USAGE: %0 [--db-file sqlite db file] [--sql-file sql file path to run]
+echo   [-d|--db-file sqlite db file path] default: %DEFAULT_DB_FILE%
+echo   [-f|--sql-file sqlite file to run] default: %DEFAULT_SQL_FILE%
+echo   [-h|--help] Usage message
+exit /b 0
+
+:argDone
+
+if "%SQL_FILE%"=="" (
+    if not exist "%WORK_DIR%\pilot\data" mkdir "%WORK_DIR%\pilot\data"
+    for %%f in (%WORK_DIR%\docker\examples\sqls\*_sqlite.sql) do (
+        echo execute sql file: %%f
+        sqlite3 "%DB_FILE%" < "%%f"
+    )
+) else (
+    echo Execute SQL file %SQL_FILE%
+    sqlite3 "%DB_FILE%" < "%SQL_FILE%"
+)

scripts/examples/load_examples.sh

@@ -0,0 +1,69 @@
+#!/bin/bash
+# Only support SQLite now
+
+SCRIPT_LOCATION=$0
+cd "$(dirname "$SCRIPT_LOCATION")"
+WORK_DIR=$(pwd)
+WORK_DIR="$WORK_DIR/../.."
+
+if ! command -v sqlite3 > /dev/null 2>&1
+then
+  echo "sqlite3 not found, please install sqlite3"
+  exit 1
+fi
+
+DEFAULT_DB_FILE="DB-GPT/pilot/data/default_sqlite.db"
+DEFAULT_SQL_FILE="DB-GPT/docker/examples/sqls/*_sqlite.sql"
+DB_FILE="$WORK_DIR/pilot/data/default_sqlite.db"
+SQL_FILE=""
+
+usage () {
+    echo "USAGE: $0 [--db-file sqlite db file] [--sql-file sql file path to run]"
+    echo "  [-d|--db-file sqlite db file path] default: ${DEFAULT_DB_FILE}"
+    echo "  [-f|--sql-file sqlte file to run] default: ${DEFAULT_SQL_FILE}"
+    echo "  [-h|--help] Usage message"
+}
+
+while [[ $# -gt 0 ]]; do
+    key="$1"
+    case $key in
+        -d|--db-file)
+        DB_FILE="$2"
+        shift # past argument
+        shift # past value
+        ;;
+        -f|--sql-file)
+        SQL_FILE="$2"
+        shift
+        shift
+        ;;
+        -h|--help)
+        help="true"
+        shift
+        ;;
+        *)
+        usage
+        exit 1
+        ;;
+    esac
+done
+
+if [[ $help ]]; then
+    usage
+    exit 0
+fi
+
+if [ -n $SQL_FILE ];then
+    mkdir -p $WORK_DIR/pilot/data
+    for file in $WORK_DIR/docker/examples/sqls/*_sqlite.sql
+    do
+        echo "execute sql file: $file"
+        sqlite3 $DB_FILE  < "$file"
+    done
+
+else
+    echo "Execute SQL file ${SQL_FILE}"
+    sqlite3 $DB_FILE < $SQL_FILE
+fi
+
+

scripts/llama_cpp_install.sh

@@ -0,0 +1,47 @@
+#!/bin/bash
+
+# Get cuda version by torch
+CUDA_VERSION=`python -c "import torch;print(torch.version.cuda.replace('.', '') if torch.cuda.is_available() else '')"`
+
+DEVICE="cpu"
+CPU_OPT="basic"
+
+if [ "${CUDA_VERSION}" = "" ]; then
+    echo "CUDA not support, use cpu version"
+else
+    DEVICE="cu${CUDA_VERSION//./}"
+    echo "CUDA version: $CUDA_VERSION, download path: $DEVICE"
+fi
+
+echo "Checking CPU support:"
+CPU_SUPPORT=$(lscpu)
+
+echo "$CPU_SUPPORT" | grep -q "avx "
+if [ $? -eq 0 ]; then
+    echo "CPU supports AVX."
+    # CPU_OPT="AVX"
+    # TODO AVX will failed on my cpu
+else
+  echo "CPU does not support AVX."
+fi
+
+echo "$CPU_SUPPORT" | grep -q "avx2"
+if [ $? -eq 0 ]; then
+  echo "CPU supports AVX2."
+  CPU_OPT="AVX2"
+else
+  echo "CPU does not support AVX2."
+fi
+
+echo "$CPU_SUPPORT" | grep -q "avx512"
+if [ $? -eq 0 ]; then
+  echo "CPU supports AVX512."
+  CPU_OPT="AVX512"
+else
+  echo "CPU does not support AVX512."
+fi
+
+EXTRA_INDEX_URL="https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/$CPU_OPT/$DEVICE"
+
+echo "install llama-cpp-python from --extra-index-url ${EXTRA_INDEX_URL}"
+python -m pip install llama-cpp-python --force-reinstall --no-cache --prefer-binary --extra-index-url=$EXTRA_INDEX_URL