登录
注册
开源
企业版
高校版
搜索
帮助中心
使用条款
关于我们
开源
企业版
高校版
私有云
Gitee AI
NEW
我知道了
查看详情
登录
注册
代码拉取完成，页面将自动刷新
捐赠
捐赠前请先登录
取消
前往登录
扫描微信二维码支付
取消
支付完成
支付提示
将跳转至支付宝完成支付
确定
取消
Watch
不关注
关注所有动态
仅关注版本发行动态
关注但不提醒动态
1
Star
0
Fork
0
liyueyong
/
AI
代码
Issues
1
Pull Requests
0
Wiki
统计
流水线
服务
Gitee Pages
JavaDoc
PHPDoc
质量分析
Jenkins for Gitee
腾讯云托管
腾讯云 Serverless
悬镜安全
阿里云 SAE
Codeblitz
我知道了，不再自动展开
更新失败，请稍后重试！
Issues
/
详情
移除标识
内容风险标识
本任务被
标识为内容中包含有代码安全 Bug 、隐私泄露等敏感信息，仓库外成员不可访问
python 代码1
待办的
#I8S13C
liyueyong
拥有者
创建于
2023-12-28 13:58
``` import os from tqdm import tqdm import logging import argparse import requests from openpyxl import load_workbook import multiprocessing BAICHUAN_MODEL = 'baichuan2' ERNIE_MODEL = 'ernie_modle' RAG_QA_MODEL = 'rag_qa' RAG_SHELL_MODEL = 'rag_shell' START = 3 COLS = { RAG_QA_MODEL: {'answer': 8, 'rag_source': 12,'distances':19 }, RAG_SHELL_MODEL: 16, } RAG_SESSION_ID = None ROW_START = 3 def init_args(): parser = argparse.ArgumentParser() parser.add_argument("--excel_path", type=str, required=True, help="excel path to process") args = parser.parse_args() return args def chat_with_baichuan2(prompt): endpoint = "http://60.204.250.91:8000/v1/chat/completions" messages = list() messages.append({"role": "user", "content": prompt}) data = { "temperature": 0.8, "messages": messages, "stream": False, "model": "baichuan2" } response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() return result["choices"][0]["message"]["content"] def chat_with_RAG(prompt): global RAG_SESSION_ID if not RAG_SESSION_ID: endpoint_get = "http://60.204.250.91:8002/session/generate_session" response = requests.get(endpoint_get) RAG_SESSION_ID = response.text.strip('"') endpoint = "http://60.204.250.91:8002/kb/get_answer" data = { "session_id": RAG_SESSION_ID, "question": prompt, "kb_sn": "openEuler_c339b4fb", "fetch_source": True, "top_k": 5 } try: response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() except Exception as _: return {} return result def chat_with_ERNIE(prompt): endpoint = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro?access_token=" + \ "24.89a359c0a59d817a2cdb9ab08afc7eb4.2592000.1704456790.282335-44446921" messages = list() messages.append({"role": "user", "content": prompt}) data = {"messages": messages} response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() return result['result'] def chat_with_RAG_shell(prompt): endpoint = "https://rag.test.osinfra.cn/kb/shell" data = { "question": prompt } try: response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() except Exception as _: return {} return result def sol(sheet_name, args): workbook = load_workbook(args.excel_path) testcase_sheet = workbook[sheet_name] result_path = os.path.dirname(args.excel_path) result_path = os.path.join(result_path, f'{sheet_name}.xlsx') for index in tqdm(range(ROW_START, testcase_sheet.max_row + 1)): for model in (RAG_QA_MODEL, RAG_SHELL_MODEL): cell_value = testcase_sheet.cell(row=index, column=5).value cell_value = cell_value.strip() if not cell_value: continue if model == RAG_QA_MODEL: ret = chat_with_RAG(cell_value) testcase_sheet.cell(row=index, column=COLS[model]['answer'], value=ret.get('answer', None)) testcase_sheet.cell(row=index, column=COLS[model]['rag_source'], value=str(ret['source_contents'])) testcase_sheet.cell(row=index, column=COLS[model]['distances'],value=str(ret['distances'])) elif model == RAG_SHELL_MODEL: ret = chat_with_RAG_shell(cell_value) testcase_sheet.cell(row=index, column=COLS[model], value=ret.get('answer', None)) workbook.save(result_path) def main(): args = init_args() sol("测试用例", args) if __name__ == '__main__': main() ```
``` import os from tqdm import tqdm import logging import argparse import requests from openpyxl import load_workbook import multiprocessing BAICHUAN_MODEL = 'baichuan2' ERNIE_MODEL = 'ernie_modle' RAG_QA_MODEL = 'rag_qa' RAG_SHELL_MODEL = 'rag_shell' START = 3 COLS = { RAG_QA_MODEL: {'answer': 8, 'rag_source': 12,'distances':19 }, RAG_SHELL_MODEL: 16, } RAG_SESSION_ID = None ROW_START = 3 def init_args(): parser = argparse.ArgumentParser() parser.add_argument("--excel_path", type=str, required=True, help="excel path to process") args = parser.parse_args() return args def chat_with_baichuan2(prompt): endpoint = "http://60.204.250.91:8000/v1/chat/completions" messages = list() messages.append({"role": "user", "content": prompt}) data = { "temperature": 0.8, "messages": messages, "stream": False, "model": "baichuan2" } response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() return result["choices"][0]["message"]["content"] def chat_with_RAG(prompt): global RAG_SESSION_ID if not RAG_SESSION_ID: endpoint_get = "http://60.204.250.91:8002/session/generate_session" response = requests.get(endpoint_get) RAG_SESSION_ID = response.text.strip('"') endpoint = "http://60.204.250.91:8002/kb/get_answer" data = { "session_id": RAG_SESSION_ID, "question": prompt, "kb_sn": "openEuler_c339b4fb", "fetch_source": True, "top_k": 5 } try: response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() except Exception as _: return {} return result def chat_with_ERNIE(prompt): endpoint = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro?access_token=" + \ "24.89a359c0a59d817a2cdb9ab08afc7eb4.2592000.1704456790.282335-44446921" messages = list() messages.append({"role": "user", "content": prompt}) data = {"messages": messages} response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() return result['result'] def chat_with_RAG_shell(prompt): endpoint = "https://rag.test.osinfra.cn/kb/shell" data = { "question": prompt } try: response = requests.post( endpoint, headers={ "Content-Type": "application/json", }, json=data, stream=False, ) result = response.json() except Exception as _: return {} return result def sol(sheet_name, args): workbook = load_workbook(args.excel_path) testcase_sheet = workbook[sheet_name] result_path = os.path.dirname(args.excel_path) result_path = os.path.join(result_path, f'{sheet_name}.xlsx') for index in tqdm(range(ROW_START, testcase_sheet.max_row + 1)): for model in (RAG_QA_MODEL, RAG_SHELL_MODEL): cell_value = testcase_sheet.cell(row=index, column=5).value cell_value = cell_value.strip() if not cell_value: continue if model == RAG_QA_MODEL: ret = chat_with_RAG(cell_value) testcase_sheet.cell(row=index, column=COLS[model]['answer'], value=ret.get('answer', None)) testcase_sheet.cell(row=index, column=COLS[model]['rag_source'], value=str(ret['source_contents'])) testcase_sheet.cell(row=index, column=COLS[model]['distances'],value=str(ret['distances'])) elif model == RAG_SHELL_MODEL: ret = chat_with_RAG_shell(cell_value) testcase_sheet.cell(row=index, column=COLS[model], value=ret.get('answer', None)) workbook.save(result_path) def main(): args = init_args() sol("测试用例", args) if __name__ == '__main__': main() ```
评论 (
0
)
liyueyong
创建了
任务
liyueyong
修改了
描述
原值
import os
from tqdm import tqdm
import
logging
impor
t
argparse
import
requests
from openpyxl import load_workbook
import
multiprocessing
BAICHUAN_MODEL = 'baichuan2'
ERNIE_MODEL = 'ernie_modle'
R
A
G_QA_MODEL =
'rag_qa'
RAG_SHELL_MODEL = 'rag_sh
e
ll'
START = 3
COLS = {
RAG_QA_MODEL: {'answer': 8, 'rag_source': 12,'distances':19 },
RAG_SHELL_MODEL: 16,
}
RAG_SESSION_ID = None
ROW_START = 3
def init_args():
parser = argparse.ArgumentParser()
parser.add_argument("--excel_path", type=str, required=True, help="excel path to process")
arg
s
=
parser.parse_args()
return
a
rgs
def chat_with_baichuan2(prompt):
endpoint = "http://60.204.250.91:8000/v1/chat/completions"
messages = l
i
st()
messages.append({"role": "user", "content": prompt})
data = {
"t
e
mperature": 0.8,
"messages": messages,
"
stream": False,
"m
odel":
"
baichu
a
n2"
}
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
stream=False,
)
resul
t
= response.json()
return result["choices"][0]["message"]["content"]
def chat_with_RAG(prompt):
global RAG_SESSION_ID
if not RAG_SESSION_ID
:
endpoint_get = "http://60.204.250.91:8002/session/generate_session"
response = requests.get(endpoint_get)
RAG_SESSION_ID
= response
.
text.strip('"')
endpoint = "http://60.204.250
.
91:8002/kb/get_answer"
data = {
"session_id"
:
RAG_SESSION_ID,
"question": prompt,
"
kb_
s
n": "openEuler_c339b4fb",
"
fetch_source": True,
"
top_k
":
5
}
try:
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
stream=False,
)
resul
t
= response.json()
except Exception as _:
re
t
u
rn
{}
return
resul
t
def chat_with_ERNIE(prompt):
endpoint = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro?access_token=" + \
"24.89a359c0a59d817a2cdb9ab08afc7eb4.2592000.1704456790.282335-44446921"
messages
=
lis
t
()
messages.append({"role": "user", "content": prompt})
data = {"messages": messages}
r
es
ponse = r
e
quests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
stream=False,
)
resul
t
= response.json()
return result['result']
def chat_with_RAG_shell(prompt):
endpoint = "https://rag.test.osinfra.cn/kb/shell"
data = {
"question": prompt
}
try:
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
stream=False,
)
resul
t
= response.json()
except Exception as _:
re
t
u
rn
{}
return
resul
t
def sol(sheet_name, args):
workbook = load_workbook(args.excel_path)
testcas
e
_sheet = workbook[sheet_name]
result_path = os.path.dirname(args.excel_path)
r
es
ult_pat
h
= os.path.j
o
in(r
e
sult_path, f'{sheet_name}.xlsx')
for index in tqdm(range(ROW_START, testcase_sheet.max_row + 1)):
for model
in (RAG_QA_MODEL, RAG_SHELL_MODEL):
cell_value = t
e
stcase_sheet.cell(row=index, column=5).value
cell_value = cell_value.strip()
if not cell_value:
continue
if
m
o
del == RAG_QA_MODEL:
ret = chat_with_RAG(cell_value)
testcase_sheet.cell(row=index, column=COLS[model]['answer'], value=ret.get('answer', None))
t
e
stcase_shee
t
.cell
(
row=index, column=COLS[model]['rag_source'], value=str(ret['source_contents']))
testcase_sheet.cell(row=index, column=COLS[model]['
di
s
tances'],value=str(ret['distances']
))
elif model == RAG_SHELL_MODEL:
r
e
t = chat_wi
t
h_RAG_shell(c
e
ll_va
lu
e)
testcase_sheet.cell(row=index, column=COLS[model], value=ret.get('answer', None))
workbook.save(result_path)
def main():
args = init_args()
sol("测试用例", args)
if __name__ == '__main__':
main()
新值
```
import
os
from
t
qdm import tqdm
import
logging
import argparse
import
requests
from openpyxl import load_workbook
import multiprocessing
B
A
ICHUAN_MODEL
= 'baichuan2'
ERNIE_MODEL = 'ernie_modl
e
'
RAG_QA_MODEL = 'rag_qa'
RAG_SHELL_MODEL = 'rag_shell'
START = 3
COLS = {
RAG_QA_MODEL: {'answer': 8, 'rag_source': 12,'distances':19 },
RAG_SHELL_MODEL: 16,
}
RAG_SESSION_ID = None
ROW_START = 3
def
init_args():
par
s
er
= argparse.ArgumentParser()
parser.
a
dd_argument("--excel_path", type=str, required=True, help="excel path to process")
args = parser.parse_args()
return args
def
chat_with_ba
i
chuan2(prompt):
endpoint = "http://60.204.250.91:8000/v1/chat/completions"
messages = list()
messag
e
s.append({"role": "user", "content": prompt})
data = {
"
temperature": 0.8,
"m
essages
"
: mess
a
ges,
"stream": False,
"model": "baichuan2"
}
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
s
t
ream=False,
)
result = response.json()
return result["choices"][0]["message"]["content"]
def
chat_with_RAG(prompt)
:
global RAG_SESSION_ID
if
not RAG_SESSION_ID:
endpoint_get =
"http://60
.
204.250.91:8002/session/generate_session"
response = requests.get(endpoint_get)
RAG_SESSION_ID = response
.
text.strip('"')
endpoint = "http
:
//60.204.250.91:8002/kb/get_answer"
data = {
"
ses
s
ion_id": RAG_SESSION_ID,
"
question": prompt,
"
kb_sn
":
"openEuler_c339b4fb",
"fetch_source": True,
"top_k": 5
}
try:
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
s
t
ream=False,
)
re
s
u
lt
= response.json()
except
Excep
t
ion as _:
return {}
return result
def
chat_with_ERNIE(prompt):
endpoint
=
"ht
t
ps://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro?access_token=" + \
"24.89a359c0a59d817a2cdb9ab08afc7eb4.2592000.1704456790.282335-44446921"
messages = list()
m
es
sages.app
e
nd({"role": "user", "content": prompt})
data = {"messages": messages}
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
s
t
ream=False,
)
result = response.json()
return result['result']
def
chat_with_RAG_shell(prompt):
endpoint = "https://rag.test.osinfra.cn/kb/shell"
data = {
"question": prompt
}
try:
response = requests.post(
endpoint,
headers={
"Content-Type": "application/json",
},
json=data,
s
t
ream=False,
)
re
s
u
lt
= response.json()
except
Excep
t
ion as _:
return {}
return result
def
sol(she
e
t_name, args):
workbook = load_workbook(args.excel_path)
t
es
tcase_s
h
eet = workbo
o
k[sh
e
et_name]
result_path = os.path.dirname(args.excel_path)
result_path =
os.path.join(result_path, f'{sheet_name}.xlsx')
for
index in tqdm(rang
e
(ROW_START, testcase_sheet.max_row + 1)):
for
model in (RAG_QA_MODEL, RAG_SHELL_MODEL):
cell_value = testcase_sheet.cell(row=index, column=5).value
cell_value = cell_value.strip()
if
n
o
t cell_value:
continue
if
model == RAG_QA_MODEL:
r
e
t = chat_wi
t
h_RAG
(
cell_value)
testcase_sheet.cell(row=index, column=COLS[model]['
an
s
wer'], value=ret.get('answer', None
))
testcase_sheet.cell(row=index, column=COLS[model]['rag_source'], value=str(ret['source_contents']))
t
e
stcase_shee
t
.cell(row=ind
e
x, co
lu
mn=COLS[model]['distances'],value=str(ret['distances']))
elif model == RAG_SHELL_MODEL:
ret = chat_with_RAG_shell(cell_value)
testcase_sheet.cell(row=index, column=COLS[model], value=ret.get('answer', None))
workbook.save(result_path)
def
main():
args = init_args()
sol("测试用例", args)
if __name__ == '__main__':
main()
```
展开全部操作日志
折叠全部操作日志
登录
后才可以发表评论
状态
待办的
待办的
进行中
已完成
已关闭
负责人
未设置
标签
未设置
标签管理
里程碑
未关联里程碑
未关联里程碑
Pull Requests
未关联
未关联
关联的 Pull Requests 被合并后可能会关闭此 issue
分支
未关联
未关联
master
开始日期 - 截止日期
-
置顶选项
不置顶
置顶等级：高
置顶等级：中
置顶等级：低
优先级
不指定
严重
主要
次要
不重要
参与者（1）
1
https://gitee.com/lyy549745/ai.git
git@gitee.com:lyy549745/ai.git
lyy549745
ai
AI
点此查找更多帮助
搜索帮助
Git 命令在线学习
如何在 Gitee 导入 GitHub 仓库
Git 仓库基础操作
企业版和社区版功能对比
SSH 公钥设置
如何处理代码冲突
仓库体积过大，如何减小？
如何找回被删除的仓库数据
Gitee 产品配额说明
GitHub仓库快速导入Gitee及同步更新
什么是 Release（发行版）
将 PHP 项目自动发布到 packagist.org
评论
仓库举报
回到顶部
登录提示
该操作需登录 Gitee 帐号，请先登录后再操作。
立即登录
没有帐号，去注册
