| # |
| # Licensed to the Apache Software Foundation (ASF) under one or more |
| # contributor license agreements. See the NOTICE file distributed with |
| # this work for additional information regarding copyright ownership. |
| # The ASF licenses this file to You under the Apache License, Version 2.0 |
| # (the "License"); you may not use this file except in compliance with |
| # the License. You may obtain a copy of the License at |
| # |
| # http://www.apache.org/licenses/LICENSE-2.0 |
| # |
| # Unless required by applicable law or agreed to in writing, software |
| # distributed under the License is distributed on an "AS IS" BASIS, |
| # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| # See the License for the specific language governing permissions and |
| # limitations under the License. |
| # |
| |
| use t::APISIX 'no_plan'; |
| |
| log_level("info"); |
| repeat_each(1); |
| no_long_string(); |
| no_root_location(); |
| |
| |
| my $resp_file = 't/assets/ai-proxy-response.json'; |
| open(my $fh, '<', $resp_file) or die "Could not open file '$resp_file' $!"; |
| my $resp = do { local $/; <$fh> }; |
| close($fh); |
| |
| print "Hello, World!\n"; |
| print $resp; |
| |
| |
| add_block_preprocessor(sub { |
| my ($block) = @_; |
| |
| if (!defined $block->request) { |
| $block->set_value("request", "GET /t"); |
| } |
| |
| my $http_config = $block->http_config // <<_EOC_; |
| server { |
| server_name openai; |
| listen 6724; |
| |
| default_type 'application/json'; |
| |
| location /v1/chat/completions { |
| content_by_lua_block { |
| local json = require("cjson.safe") |
| |
| if ngx.req.get_method() ~= "POST" then |
| ngx.status = 400 |
| ngx.say("Unsupported request method: ", ngx.req.get_method()) |
| end |
| ngx.req.read_body() |
| local body, err = ngx.req.get_body_data() |
| body, err = json.decode(body) |
| |
| local query_auth = ngx.req.get_uri_args()["api_key"] |
| |
| if query_auth ~= "apikey" then |
| ngx.status = 401 |
| ngx.say("Unauthorized") |
| return |
| end |
| |
| local res = [[ |
| { |
| "id": "chatcmpl-12345", |
| "object": "chat.completion", |
| "created": 1691234567, |
| "model": "gpt-3.5-turbo", |
| "choices": [ |
| { |
| "index": 0, |
| "message": { |
| "role": "assistant", |
| "content": "这是一个示例回复。" |
| }, |
| "finish_reason": "stop" |
| } |
| ], |
| "usage": { |
| "prompt_tokens": 10, |
| "completion_tokens": 20, |
| "total_tokens": 30 |
| } |
| }]] |
| ngx.status = 200 |
| ngx.say(res) |
| } |
| } |
| |
| location /null-content { |
| content_by_lua_block { |
| local json = require("cjson.safe") |
| |
| local res = [[ |
| { |
| "model": "gpt-3.5-turbo", |
| "choices": [ |
| { |
| "index": 0, |
| "message": { |
| "role": "assistant", |
| "content": null |
| }, |
| "finish_reason": "stop" |
| } |
| ], |
| "usage": null |
| }]] |
| ngx.status = 200 |
| ngx.say(res) |
| } |
| } |
| } |
| _EOC_ |
| |
| $block->set_value("http_config", $http_config); |
| }); |
| |
| run_tests(); |
| |
| __DATA__ |
| |
| === TEST 1: set access log |
| --- config |
| location /t { |
| content_by_lua_block { |
| local t = require("lib.test_admin").test |
| local code, body = t('/apisix/admin/routes/1', |
| ngx.HTTP_PUT, |
| [[{ |
| "uri": "/anything", |
| "plugins": { |
| "ai-proxy": { |
| "provider": "openai", |
| "auth": { |
| "query": { |
| "api_key": "apikey" |
| } |
| }, |
| "options": { |
| "model": "gpt-3.5-turbo", |
| "max_tokens": 512, |
| "temperature": 1.0 |
| }, |
| "override": { |
| "endpoint": "http://localhost:6724/v1/chat/completions" |
| }, |
| "ssl_verify": false |
| } |
| } |
| }]] |
| ) |
| |
| if code >= 300 then |
| ngx.status = code |
| end |
| ngx.say(body) |
| } |
| } |
| --- response_body |
| passed |
| |
| |
| |
| === TEST 2: send request |
| --- request |
| POST /anything |
| {"messages":[{"role":"system","content":"You are a mathematician"},{"role":"user","content":"What is 1+1?"}], "model": "gpt-4"} |
| --- error_code: 200 |
| --- response_body eval |
| qr/.*completion_tokens.*/ |
| --- access_log eval |
| qr/.*[\d.]+ \"http:\/\/localhost\" gpt-4 gpt-3.5-turbo \d+ 10 20.*/ |
| |
| |
| |
| === TEST 3: proxy to /null-content ai endpoint |
| --- config |
| location /t { |
| content_by_lua_block { |
| local t = require("lib.test_admin").test |
| local code, body = t('/apisix/admin/routes/1', |
| ngx.HTTP_PUT, |
| [[{ |
| "uri": "/anything", |
| "plugins": { |
| "ai-proxy": { |
| "provider": "openai", |
| "auth": { |
| "header": { |
| "Authorization": "Bearer token" |
| } |
| }, |
| "override": { |
| "endpoint": "http://localhost:6724/null-content" |
| } |
| } |
| } |
| }]] |
| ) |
| |
| if code >= 300 then |
| ngx.status = code |
| end |
| ngx.say(body) |
| } |
| } |
| --- response_body |
| passed |
| |
| |
| |
| === TEST 4: send request |
| --- request |
| POST /anything |
| {"messages":[{"role":"user","content":"What is 1+1?"}], "model": "gpt-4"} |
| --- error_code: 200 |
| --- response_body eval |
| qr/.*assistant.*/ |
| --- no_error_log |
| |
| |
| |
| === TEST 5: create a ai-proxy-multi route with delay streaming ai endpoint(every event delay 200ms) |
| --- config |
| location /t { |
| content_by_lua_block { |
| local t = require("lib.test_admin").test |
| local code, body = t('/apisix/admin/routes/1', |
| ngx.HTTP_PUT, |
| [[{ |
| "uri": "/anything", |
| "plugins": { |
| "ai-proxy-multi": { |
| "instances": [ |
| { |
| "name": "self-hosted", |
| "provider": "openai-compatible", |
| "weight": 1, |
| "auth": { |
| "header": { |
| "Authorization": "Bearer token" |
| } |
| }, |
| "options": { |
| "model": "gpt-3.5-turbo", |
| "stream": true |
| }, |
| "override": { |
| "endpoint": "http://localhost:7737/v1/chat/completions?delay=true" |
| } |
| } |
| ], |
| "ssl_verify": false |
| } |
| } |
| }]] |
| ) |
| if code >= 300 then |
| ngx.status = code |
| end |
| ngx.say(body) |
| } |
| } |
| --- response_body |
| passed |
| |
| |
| |
| === TEST 6: assert access log contains right llm variable |
| --- config |
| location /t { |
| content_by_lua_block { |
| local http = require("resty.http") |
| local httpc = http.new() |
| local core = require("apisix.core") |
| local ok, err = httpc:connect({ |
| scheme = "http", |
| host = "localhost", |
| port = ngx.var.server_port, |
| }) |
| if not ok then |
| ngx.status = 500 |
| ngx.say(err) |
| return |
| end |
| local params = { |
| method = "POST", |
| headers = { |
| ["Content-Type"] = "application/json", |
| }, |
| path = "/anything", |
| body = [[{ |
| "messages": [ |
| { "role": "system", "content": "some content" } |
| ], |
| "model": "gpt-4" |
| }]], |
| } |
| local res, err = httpc:request(params) |
| if not res then |
| ngx.status = 500 |
| ngx.say(err) |
| return |
| end |
| local final_res = {} |
| local inspect = require("inspect") |
| while true do |
| local chunk, err = res.body_reader() -- will read chunk by chunk |
| if err then |
| core.log.error("failed to read response chunk: ", err) |
| break |
| end |
| if not chunk then |
| break |
| end |
| core.table.insert_tail(final_res, chunk) |
| end |
| ngx.print(#final_res .. final_res[6]) |
| } |
| } |
| --- response_body_like eval |
| qr/6data: \[DONE\]\n\n/ |
| --- access_log eval |
| qr/.*[\d.]+ \"http:\/\/localhost:1984\" gpt-4 gpt-3.5-turbo 2\d\d 15 20.*/ |
