{"id":143466,"date":"2023-03-08T14:14:44","date_gmt":"2023-03-08T05:14:44","guid":{"rendered":"https:\/\/support.questetra.com\/?p=143466"},"modified":"2023-03-20T17:31:31","modified_gmt":"2023-03-20T08:31:31","slug":"openai-chat-interact-with-parameters-2023","status":"publish","type":"post","link":"https:\/\/support.questetra.com\/ja\/addons\/openai-chat-interact-with-parameters-2023\/","title":{"rendered":"OpenAI: Chat, \u30d1\u30e9\u30e1\u30fc\u30bf\u4ed8\u304d\u5bfe\u8a71"},"content":{"rendered":"\n
\n

warning<\/span> PAGE UPDATED<\/h3>\nhttps:\/\/support.questetra.com\/ja\/addons\/openai-chat-interact-with-parameters-202303\/<\/a>\n<\/div>\n\n\n\n

OpenAI: Chat, \u30d1\u30e9\u30e1\u30fc\u30bf\u4ed8\u304d\u5bfe\u8a71<\/strong><\/h4>\n\n\n\n

OpenAI: Chat, Interact with Parameters<\/a><\/p>\n\n\n\n

OpenAI API (ChatGPT) \u3068\u30b3\u30df\u30e5\u30cb\u30b1\u30fc\u30b7\u30e7\u30f3\u3057\u307e\u3059\u3002\u4f1a\u8a71\u5c65\u6b74\u3092\u3075\u307e\u3048\u305f\u518d\u8cea\u554f\u3084\u9ad8\u5ea6\u306a\u30d1\u30e9\u30e1\u30fc\u30bf\u3092\u4ed8\u3051\u305f\u6307\u793a\uff08PROMPT\uff09\u304c\u53ef\u80fd\u3067\u3059\u3002\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u6e29\u5ea6\u3001\u4e0a\u4f4d\uff05\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u3001\u518d\u51fa\u73fe\u7981\u6b62\u5ea6\u3001\u983b\u51fa\u7981\u6b62\u5ea6\u3001Logit\u30d0\u30a4\u30a2\u30b9\u306a\u3069\u3092\u8a2d\u5b9a\u3067\u304d\u307e\u3059\u3002\u4f7f\u7528\u3055\u308c\u308b\u30e2\u30c7\u30ebID\u306f “gpt-3.5-turbo” \u3067\u3059\u3002<\/p>\n\n\n\n

\"Auto<\/figure>\n\n\n\n
\n
Configs for this Auto Step<\/summary>
\n
\n
AuthzConfU<\/dt>
U: HTTP\u8a8d\u8a3c\u8a2d\u5b9a\u3092\u9078\u629e\u3057\u3066\u304f\u3060\u3055\u3044\uff08Secret API Key \uff20\u30c8\u30fc\u30af\u30f3\u76f4\u63a5\u6307\u5b9a\uff09 *<\/span><\/dd>\n
StrConfA0<\/dt>
A0: \u30ec\u30b9\u30dd\u30f3\u30b9\u8005\u306e\u5f79\u5272\uff08SYSTEM Role\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfPro1<\/dt>
Pro1: \u521d\u56de\u306e\u6307\u793a\u6587\uff08PROMPT\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfCom1<\/dt>
Com1: \u521d\u56de\u6307\u793a\u6587\u306e\u56de\u7b54\u6587\uff08COMPLETION\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfPro2<\/dt>
Pro2: \uff12\u56de\u76ee\u306e\u6307\u793a\u6587\uff08PROMPT\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfCom2<\/dt>
Com2: \u7b2c\uff12\u6307\u793a\u6587\u306e\u56de\u7b54\u6587\uff08COMPLETION\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfPro3<\/dt>
Pro3: \uff13\u56de\u76ee\u306e\u6307\u793a\u6587\uff08PROMPT\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfCom3<\/dt>
Com3: \u7b2c\uff13\u6307\u793a\u6587\u306e\u56de\u7b54\u6587\uff08COMPLETION\uff09\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfA1<\/dt>
A1: \u30ea\u30af\u30a8\u30b9\u30c8 PROMPT \u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044 *<\/span>#{EL}<\/sup><\/span><\/dd>\n
StrConfA2<\/dt>
A2: \u30d1\u30e9\u30e1\u30fc\u30bf\uff08\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u6e29\u5ea6\u3001\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\uff05\u3001\u518d\u51fa\u73fe\u7981\u6b62\u5ea6\u3001\u983b\u51fa\u7981\u6b62\u5ea6\uff09\u30924\u884c\u306b\u5206\u3051\u3066\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfA3<\/dt>
A3: Logit\u30d0\u30a4\u30a2\u30b9\uff08TokenID \u3068 \u30d0\u30a4\u30a2\u30b9\u5024\u306e\u30da\u30a2\uff09\u3092\u5404\u884c\u306b\u5206\u3051\u3066\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044#{EL}<\/sup><\/span><\/dd>\n
StrConfA4<\/dt>
A4: \u30ec\u30b9\u30dd\u30f3\u30b9\u6570\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044 (default 1)#{EL}<\/sup><\/span><\/dd>\n
StrConfA5<\/dt>
A5: \u30ec\u30b9\u30dd\u30f3\u30b9\u30c8\u30fc\u30af\u30f3\u306e\u8a31\u5bb9\u91cf\u3092\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044 (default 2048)#{EL}<\/sup><\/span><\/dd>\n
StrConfA6<\/dt>
A6: \u4e2d\u65ad\u6587\u5b57\u3092\u5404\u884c\u306b\u5206\u3051\u3066\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044 (eg “.” “\u3002”)#{EL}<\/sup><\/span><\/dd>\n
StrConfB1<\/dt>
B1: \u30ec\u30b9\u30dd\u30f3\u30b9 COMPLETION \u304c\u683c\u7d0d\u3055\u308c\u308b\u30d5\u30a3\u30fc\u30eb\u30c9\u540d\uff08\u6587\u5b57\u5217\u578b\u30c7\u30fc\u30bf\u9805\u76ee\uff09\u3092\u5404\u884c\u306b\u5206\u3051\u3066\u30bb\u30c3\u30c8\u3057\u3066\u304f\u3060\u3055\u3044\uff08\u66f4\u65b0\uff09#{EL}<\/sup><\/span><\/dd>\n
SelectConfB2<\/dt>
B2: \u30ec\u30b9\u30dd\u30f3\u30b9 Json \u5168\u4f53\u3092\u683c\u7d0d\u3057\u305f\u3044\u5834\u5408\u3001\u6587\u5b57\u5217\u578b\u30c7\u30fc\u30bf\u9805\u76ee\u3092\u9078\u629e\u3057\u3066\u304f\u3060\u3055\u3044\uff08\u66f4\u65b0\uff09<\/dd>\n
SelectConfC1<\/dt>
C1: PROMPT \u30c8\u30fc\u30af\u30f3\u6570\u304c\u683c\u7d0d\u3055\u308c\u308b\u6570\u5024\u578b\u30c7\u30fc\u30bf\u9805\u76ee\u3092\u9078\u629e\u3057\u3066\u304f\u3060\u3055\u3044\uff08\u66f4\u65b0\uff09<\/dd>\n
SelectConfC2<\/dt>
C2: COMPLETION \u30c8\u30fc\u30af\u30f3\u6570\u304c\u683c\u7d0d\u3055\u308c\u308b\u6570\u5024\u578b\u30c7\u30fc\u30bf\u9805\u76ee\u3092\u9078\u629e\u3057\u3066\u304f\u3060\u3055\u3044\uff08\u66f4\u65b0\uff09<\/dd>\n
SelectConfC3<\/dt>
C3: \u5408\u8a08\u30c8\u30fc\u30af\u30f3\u6570\u304c\u683c\u7d0d\u3055\u308c\u308b\u6570\u5024\u578b\u30c7\u30fc\u30bf\u9805\u76ee\u3092\u9078\u629e\u3057\u3066\u304f\u3060\u3055\u3044\uff08\u66f4\u65b0\uff09<\/dd>\n\n<\/dl>\n<\/div><\/details><\/div>\n<\/div>\n\n\n\n
\n
Script (click to open)<\/summary>
\n
\/\/ GraalJS Script (engine type: 2)\n\n\/\/\/\/\/\/\/\/ START "main()" \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\n\nmain();\nfunction main(){ \n\n\/\/\/\/\/\/ == Config Retrieving \/ \u5de5\u7a0b\u30b3\u30f3\u30d5\u30a3\u30b0\u306e\u53c2\u7167 ==\nconst strAuthzSetting = configs.get      ( "AuthzConfU" );   \/\/\/ REQUIRED\n  engine.log( " AutomatedTask Config: Authz Setting: " + strAuthzSetting );\n\nconst strSystemRole   = configs.get      ( "StrConfA0" );    \/\/ NotRequired\nconst strLogPro1      = configs.get      ( "StrConfPro1" );  \/\/ NotRequired\nconst strLogCom1      = configs.get      ( "StrConfCom1" );  \/\/ NotRequired\nconst strLogPro2      = configs.get      ( "StrConfPro2" );  \/\/ NotRequired\nconst strLogCom2      = configs.get      ( "StrConfCom2" );  \/\/ NotRequired\nconst strLogPro3      = configs.get      ( "StrConfPro3" );  \/\/ NotRequired\nconst strLogCom3      = configs.get      ( "StrConfCom3" );  \/\/ NotRequired\nconst strPrompt       = configs.get      ( "StrConfA1" );    \/\/\/ REQUIRED\n  if( strPrompt     === "" ){\n    throw new Error( "\\n AutomatedTask ConfigError:" +\n                     " Config {A1:Prompt} MUST NOT be empty \\n" );\n  }\n\nconst strParams       = configs.get      ( "StrConfA2" );    \/\/ NotRequired\nconst arrParams       = strParams !== "" ? strParams.split("\\n") : null;\nconst numTemperature  = isNaN(parseFloat(arrParams?.[0])) ? 1 : parseFloat( arrParams[0] );\nconst numTopP         = isNaN(parseFloat(arrParams?.[1])) ? 1 : parseFloat( arrParams[1] );\nconst numPresPenalty  = isNaN(parseFloat(arrParams?.[2])) ? 0 : parseFloat( arrParams[2] );\nconst numFreqPenalty  = isNaN(parseFloat(arrParams?.[3])) ? 0 : parseFloat( arrParams[3] );\n\/\/ const jsonLogitBias   = arrParams?.[4] ? JSON.stringify( arrParams[4] ) : null;\n  \/\/ Number(undefined)     \/\/ NaN\n  \/\/ Number(null)          \/\/ 0 \u2606\n  \/\/ Number('100a')        \/\/ NaN\n  \/\/ parseFloat(undefined) \/\/ NaN\n  \/\/ parseFloat(null)      \/\/ NaN\n\nconst strBias         = configs.get      ( "StrConfA3" );    \/\/ NotRequired\nconst arrBias         = strBias !== "" ? strBias.split("\\n") : null;\nconst strChoises      = configs.get      ( "StrConfA4" );    \/\/ NotRequired\nconst numChoises      = isNaN(parseInt(strChoises,10)) ? 1 : parseInt(strChoises,10);\nconst strLimit        = configs.get      ( "StrConfA5" );    \/\/ NotRequired\nconst numLimit        = isNaN(parseInt(strLimit,10)) ? 2048 : parseInt(strLimit,10);\nconst strStops        = configs.get      ( "StrConfA6" );    \/\/ NotRequired\nconst arrStops        = strStops !== "" ? strStops.split("\\n") : null;\nconst strQfields      = configs.get      ( "StrConfB1" );    \/\/ NotRequired\nconst arrQfields      = strQfields !== "" ? strQfields.split("\\n") : null;\n\nconst strPocketResponseJson     = configs.getObject( "SelectConfB2" ); \/\/ NotRequired\nconst numPocketPromptTokens     = configs.getObject( "SelectConfC1" ); \/\/ NotRequired\nconst numPocketCompletionTokens = configs.getObject( "SelectConfC2" ); \/\/ NotRequired\nconst numPocketTotalTokens      = configs.getObject( "SelectConfC3" ); \/\/ NotRequired\n\n\n\n\/\/\/\/\/\/ == Data Retrieving \/ \u30ef\u30fc\u30af\u30d5\u30ed\u30fc\u30c7\u30fc\u30bf\u306e\u53c2\u7167 ==\n\/\/ (Nothing. Retrieved via Expression Language in Config Retrieving)\n\n\n\/\/\/\/\/\/ == Calculating \/ \u6f14\u7b97 ==\n\n\/\/\/\/ OpenAI API > Documentation > API REFERENCE > CHAT\n\/\/\/\/ https:\/\/platform.openai.com\/docs\/api-reference\/chat\n\n\/\/\/ prepare json\nlet strJson = {};\n    strJson.model = "gpt-3.5-turbo";\n    strJson.messages = [];\n    if ( strSystemRole !=="" ) {\n      let objSystemRole = {};\n          objSystemRole.role = "system";\n          objSystemRole.content = strSystemRole;\n      strJson.messages.push ( objSystemRole );\n    }\n    if ( strLogPro1 !=="" && strLogCom1 !=="" ) {\n      let objLogPro = {};\n          objLogPro.role = "user";\n          objLogPro.content = strLogPro1;\n      strJson.messages.push ( objLogPro );\n      let objLogCom = {};\n          objLogCom.role = "assistant";\n          objLogCom.content = strLogCom1;\n      strJson.messages.push ( objLogCom );\n    }\n    if ( strLogPro2 !=="" && strLogCom2 !=="" ) {\n      let objLogPro = {};\n          objLogPro.role = "user";\n          objLogPro.content = strLogPro2;\n      strJson.messages.push ( objLogPro );\n      let objLogCom = {};\n          objLogCom.role = "assistant";\n          objLogCom.content = strLogCom2;\n      strJson.messages.push ( objLogCom );\n    }\n    if ( strLogPro3 !=="" && strLogCom3 !=="" ) {\n      let objLogPro = {};\n          objLogPro.role = "user";\n          objLogPro.content = strLogPro3;\n      strJson.messages.push ( objLogPro );\n      let objLogCom = {};\n          objLogCom.role = "assistant";\n          objLogCom.content = strLogCom3;\n      strJson.messages.push ( objLogCom );\n    }\n\n    let objNewMsg = {};\n        objNewMsg.role = "user";\n        objNewMsg.content = strPrompt;\n    strJson.messages.push ( objNewMsg );\n\n    if ( arrParams?.[0] !=="" ) {\n      strJson.temperature       = numTemperature;\n    }\n    if ( arrParams?.[1] !=="" ) {\n      strJson.top_p             = numTopP;\n    }\n    if ( arrParams?.[2] !=="" ) {\n      strJson.presence_penalty  = numPresPenalty;\n    }\n    if ( arrParams?.[3] !=="" ) {\n      strJson.frequency_penalty = numFreqPenalty;\n    }\n\n    strJson.n          = numChoises;\n    strJson.max_tokens = numLimit;\n    strJson.user       = "m" + processInstance.getProcessModelInfoId().toString();\n    if ( arrStops !== null ){\n      strJson.stop = [];\n      for ( let i = 0; i < arrStops.length; i++ ){\n        if ( arrStops[i] === "- - -" ){\n          strJson.stop.push ( "\\n" );\n        }else{\n          strJson.stop.push ( arrStops[i] );\n        }\n      }\n    }\n    if ( arrBias !== null ){\n      strJson.logit_bias = {};\n      for ( let i = 0; i < arrBias.length; i++ ){\n        let arrNumParts = arrBias[i].match( \/-?\\d+\/g ); \/\/ numbers (including with minus signs)\n        if (arrNumParts.length >= 2) {\n          strJson.logit_bias[arrNumParts[0]] = Number(arrNumParts[1]);\n        }\n      }\n    }\n\n\/* engine.log( JSON.stringify( strJson ) ); \/\/ debug\n{\n  "model":"gpt-3.5-turbo",\n  "messages":[{\n      "role":"system",\n      "content":"Start with '>>>'. End with '<<<'."\n    },{\n      "role":"user",\n      "content":"What destinations do you recommend?"\n    },{\n      "role":"assistant","content":">>> As an AI language model ... interests and budget. <<<\n    "},{\n      "role":"user","content":"What other Japanese destinations ...  the 4?"\n    }\n  ],\n  "top_p":0.8,\n  "presence_penalty":0.1,\n  "frequency_penalty":0.1,\n  "n":3,\n  "max_tokens":200,\n  "user":"m2933",\n  "logit_bias":{\n    "52":2,\n    "14053":10,\n    "16504":5\n  }\n}\n*\/\n\n\/\/\/ prepare request1\nlet request1Uri = "https:\/\/api.openai.com\/v1\/chat\/completions";\nlet request1 = httpClient.begin(); \/\/ HttpRequestWrapper\n    request1 = request1.authSetting( strAuthzSetting ); \/\/ with "Authorization: Bearer XX"\n    request1 = request1.body( JSON.stringify( strJson ), "application\/json" );\n\n\/\/\/ try request1\nconst response1     = request1.post( request1Uri ); \/\/ HttpResponseWrapper\nengine.log( " AutomatedTask ApiRequest1 Start: " + request1Uri );\nconst response1Code = response1.getStatusCode() + ""; \/\/ JavaNum to string\nconst response1Body = response1.getResponseAsString();\nengine.log( " AutomatedTask ApiResponse1 Status: " + response1Code );\nif( response1Code !== "200"){\n  throw new Error( "\\n AutomatedTask UnexpectedResponseError: " +\n                    response1Code + "\\n" + response1Body + "\\n" );\n}\n\n\n\/\/\/ parse response1\n\/* engine.log( response1Body ); \/\/ debug\n{\n  "id":"chatcmpl-6rcsicnNlNV13EhppZVSXAuB8DLDy",\n  "object":"chat.completion",\n  "created":1678238864,\n  "model":"gpt-3.5-turbo-0301",\n  "usage":{\n    "prompt_tokens":150,\n    "completion_tokens":560,\n    "total_tokens":710\n  },\n  "choices":[{\n    "message":{\n      "role":"assistant",\n      "content":">>> Japan is a beautiful ... to visit in Japan. <<<"\n    },\n    "finish_reason":"stop",\n    "index":0\n  },{\n    "message":{\n      "role":"assistant",\n      "content":">>> Japan has many beautiful ... many amazing destinations Japan has to offer. <<<"\n    },\n    "finish_reason":"stop",\n    "index":1\n  },{\n    "message":{\n      "role":"assistant",\n      "content":">>> There are many other ... festivals, and local cuisine.\\n\\nEach of these"\n  },\n  "finish_reason":"length",\n  "index":2\n  }]\n}\n*\/\nconst response1Obj = JSON.parse( response1Body );\n\n\n\/\/\/\/\/\/ == Data Updating \/ \u30ef\u30fc\u30af\u30d5\u30ed\u30fc\u30c7\u30fc\u30bf\u3078\u306e\u4ee3\u5165 ==\n\nif( strPocketResponseJson !== null ){\n  engine.setData( strPocketResponseJson, response1Body );\n}\n\nfor ( let i = 0; i < response1Obj.choices.length; i++ ) {\n  if( engine.findDataDefinitionByVarName ( arrQfields?.[i] ) !== null ){\n    engine.setDataByVarName( arrQfields[i], \n                  response1Obj.choices[i].message.content ?? ""\n                );\n  }\n}\n\nif( numPocketPromptTokens !== null ){\n  engine.setData( numPocketPromptTokens, new java.math.BigDecimal(\n                  response1Obj.usage.prompt_tokens ?? 0\n                ));\n}\nif( numPocketCompletionTokens !== null ){\n  engine.setData( numPocketCompletionTokens, new java.math.BigDecimal(\n                  response1Obj.usage.completion_tokens ?? 0\n                ));\n}\nif( numPocketTotalTokens !== null ){\n  engine.setData( numPocketTotalTokens, new java.math.BigDecimal(\n                  response1Obj.usage.total_tokens ?? 0\n                ));\n}\n\/\/ "??": Nullish coalescing operator (ES11)\n\/\/ https:\/\/developer.mozilla.org\/docs\/Web\/JavaScript\/Reference\/Operators\/Nullish_coalescing\n\n} \/\/\/\/\/\/\/\/ END "main()" \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\n\n\n\/*\nNotes:\n- About "OpenAI API"\n    - https:\/\/platform.openai.com\/docs\/introduction\/overview\n    - The OpenAI API can be applied to virtually any task that involves understanding or generating natural language or code.\n    - For example, if you give the API the prompt, "Write a tagline for an ice cream shop",\n    - it will return a completion like "We serve up smiles with every scoop!"\n- If you place this "Addon Automated Step" on the Workflow diagram, \n    - a response will be retrieved automatically when the token reaches the automated step.\n    - A request prompt is automatically sent to the OpenAI API server. (REST API communication)\n    - The response text sent back from the OpenAI API server is automatically parsed.\n    - For example, the auto-step obtaining "advice from AI" can be integrated into business processes.\n- API key is required to use the OpenAI API.\n    - Obtain the API key to be used in advance.\n    - Set the "Secret API Key" as the communication token.\uff3bHTTP Authz Settings] > [Token Fixed Value]\n- This Automatied Step is fully upward compatible with "OpenAI: Chat, Start".\n    - The same requests as in "OpenAI: Chat, Start" are possible.\n    - https:\/\/support.questetra.com\/addons\/openai-chat-start-2023\/\n- Contextual requests are possible.\n    - The oldest instruction should be set to "1st PROMPT".\n    - The response to that directive should be set to "1st COMPLETION".\n    - A history that does not contain both a PROMPT and a COMPLETION will be invalid.\n    - Sets of PROMPT and COMPLETION can be registered up to three times.\n- It is not necessary to set all parameters.\n    - Parameter values can be tried in the Playground.\n    - https:\/\/platform.openai.com\/playground?mode=chat\n    - The settable range may differ.\n- Parameter 1st line: Sampling temperature (TEMPERATURE)\n    - range:"[0,2]", default: "1"\n    - "Higher values like 0.8 will make the output more random"\n    - Sets the randomness level (creativity).\n    - For fact-finding or tabulation, set to zero.\n- Parameter 2nd line: Sampling % (top_p)\n    - range:"(0,1]", default: "1"\n    - "the model considers the results of the tokens with top_p probability mass"\n    - Sets the top percentage of word tokens to be considered.\n    - It is not recommended to set both "sampling temperature". (as of 202303)\n- Parameter 3rd line: Presence_penalty\n    - range:"[-2,2]", default: "0"\n- Parameter 4th line: Frequency_penalty\n    - range:"[-2,2]", default: "0"\n- Logit bias: Token adjustment (logit_bias)\n    - range:"[-100,100]"\n    - eg: "16504 10"\n    - eg: "14053 10"\n    - eg: "17013 -5"\n        - 16504(Japan), 14053(USA), 2937(US), 52(U), 13(.), 50(S), 41187(united) 17013(United)\n    - "values like -100 should result in a ban selection of the relevant token."\n    - You can ban or overuse certain word Tokens.\n        - https:\/\/platform.openai.com\/tokenizer\n- High temperatures (">1") and high biases (">10") may induce 500 errors and timeouts.\n    - Determine by trial and error in the light of your work what values are appropriate.\n\nAPPENDIX\n- The SYSTEM Role setting is used to set the role as the text author.\n    - e.g. "End with '<<<'"\n    - e.g. "Start with '>>>'. End with '<<<'."\n    - e.g. "You are a creative Questetra employee."\n    - e.g. "You are a helpful assistant."\n        - https:\/\/platform.openai.com\/docs\/guides\/chat\/introduction\n- The number of responses that can be set is up to 128.\n- Interrupted words can be set up to four (invalid after the fifth line)\n    - If set `\\t` or other characters, they will be escaped.\n    - To set a newline code (`\\n`), set `- - -`. (experimental)\n        - The response will always be a single line.\n        - It always results in a 500 error. (as of 202303)\n- Large amount of allowable response sentences to be generated may exceed the system limit.\n    - Generation will be aborted if the allowable amount is reached.\n        - `"finish_reason":"length",`\n        - If a large number is set for "Number of Responses," the response is more likely to be interrupted.\n    - Model usage limit: 4096 (total number of tokens)\n        - Set smaller than 4096 minus the expected length of the PROMPT tokens.\n    - For English, one word or symbol often counts as one token.\n        - For average English, it is about 1 token for 4 characters.\n        - In Japanese, a single character may be divided into multiple tokens.\n        - In the case of average Japanese, a single character is about one token.\n    - You can check the approximate number of word tokens at tokenizer.\n        - https:\/\/platform.openai.com\/tokenizer\n- In settings that refer to numeric type data, be careful not to mix in digit separators.\n    - The formatting function `#sformat` is useful. (Java String.format)\n    - e.g. `#{#sformat("%1.1f", #q_numeric)}` (rounded to one decimal place)\n    - "R2272: Output of Strings via EL"\n        - https:\/\/questetra.zendesk.com\/hc\/en-us\/articles\/360024292872-R2272-Output-of-Strings-via-EL-syntax\n- Headers for developers belonging to multiple organizations are not yet supported (as of 202303).\n    - `OpenAI-Organization`.\n- Note that the workflow app ID is automatically assigned to the `user` parameter of the OpenAI API. (experimental)\n    - `processInstance.getProcessModelInfoId()`\n\nNotes-ja:\n- "OpenAI API" \u3068\u306f\n    - https:\/\/platform.openai.com\/docs\/introduction\/overview\n    - OpenAI API \u306f\u3001\u81ea\u7136\u8a00\u8a9e\u3084\u30b3\u30fc\u30c9\u3092\u7406\u89e3\u3057\u3001\u81ea\u7136\u8a00\u8a9e\u3084\u30b3\u30fc\u30c9\u3092\u751f\u6210\u3057\u307e\u3059\u3002\u3042\u3089\u3086\u308b\u4ed5\u4e8b\u306b\u9069\u7528\u53ef\u80fd\u3067\u3059\u3002\n    - \u305f\u3068\u3048\u3070\u3001\u300c\u30a2\u30a4\u30b9\u30af\u30ea\u30fc\u30e0\u5c4b\u306e\u30ad\u30e3\u30c3\u30c1\u30d5\u30ec\u30fc\u30ba\u3092\u66f8\u3044\u3066\u300d\u3068\u3044\u3046\u30d7\u30ed\u30f3\u30d7\u30c8\uff08Prompt\uff09\u3092\u4e0e\u3048\u308b\u3068\u3001\n    - \u300c\u5168\u3066\u306e\u30b9\u30af\u30fc\u30d7\u3092\u7b11\u9854\u3067\uff01\u300d\u3068\u3044\u3063\u305f\u5fdc\u7b54\uff08Completion\uff09\u304c\u8fd4\u3055\u308c\u307e\u3059\u3002 (Questetra\u793e\u306e\u610f\u8a33)\n- \u3053\u306e\uff3b\u81ea\u52d5\u5de5\u7a0b\uff3d\u3092\u30ef\u30fc\u30af\u30d5\u30ed\u30fc\u56f3\u306b\u914d\u7f6e\u3059\u308c\u3070\u3001\u6848\u4ef6\u304c\u5230\u9054\u3059\u308b\u5ea6\u306b\u30ea\u30af\u30a8\u30b9\u30c8\u304c\u81ea\u52d5\u9001\u4fe1\u3055\u308c\u307e\u3059\u3002\n    - OpenAI API \u30b5\u30fc\u30d0\u306b\u5bfe\u3057\u3066\u30ea\u30af\u30a8\u30b9\u30c8\u6587\u304c\u81ea\u52d5\u7684\u306b\u9001\u4fe1\u3055\u308c\u307e\u3059\u3002\uff08REST API\u901a\u4fe1\uff09\n    - OpenAI API \u30b5\u30fc\u30d0\u304b\u3089\u8fd4\u4fe1\u3055\u308c\u305f\u30ec\u30b9\u30dd\u30f3\u30b9\u6587\u304c\u81ea\u52d5\u7684\u306b\u30d1\u30fc\u30b9\uff08\u89e3\u6790\uff09\u3055\u308c\u307e\u3059\u3002\n    - \u305f\u3068\u3048\u3070\u3001\u300cAI\u304b\u3089\u306e\u52a9\u8a00\u300d\u3092\u81ea\u52d5\u7684\u306b\u53d6\u5f97\u3059\u308b\u5de5\u7a0b\u3092\u696d\u52d9\u30d7\u30ed\u30bb\u30b9\u306b\u7d44\u307f\u8fbc\u3080\u3053\u3068\u304c\u51fa\u6765\u307e\u3059\u3002\n- OpenAI API \u306e\u5229\u7528\u306b\u306f API key \u304c\u5fc5\u8981\u3067\u3059\u3002\n    - \u3042\u3089\u304b\u3058\u3081 API Key \u3092\u53d6\u5f97\u3057\u3066\u304a\u3044\u3066\u304f\u3060\u3055\u3044\u3002\n    - "Secret API Key" \u3092\u901a\u4fe1\u30c8\u30fc\u30af\u30f3\u3068\u3057\u3066\u30bb\u30c3\u30c8\u3057\u307e\u3059\u3002\uff3bHTTP \u8a8d\u8a3c\u8a2d\u5b9a\uff3d\uff1e\uff3b\u30c8\u30fc\u30af\u30f3\u76f4\u63a5\u6307\u5b9a\uff3d\n- \u3053\u306e\uff3b\u81ea\u52d5\u5de5\u7a0b\uff3d\u306f\u300eOpenAI: Chat, \u958b\u59cb\u300f\u306e\u5b8c\u5168\u4e0a\u4f4d\u4e92\u63db\u3067\u3059\u3002\n    - \u3059\u306a\u308f\u3061\u300eOpenAI: Chat, \u958b\u59cb\u300f\u3068\u540c\u3058\u30ea\u30af\u30a8\u30b9\u30c8\u3082\u53ef\u80fd\u3067\u3059\u3002\n    - https:\/\/support.questetra.com\/ja\/addons\/openai-chat-start-2023\/\n- \u6587\u8108\u3092\u4f7f\u3063\u305f\u30ea\u30af\u30a8\u30b9\u30c8\u304c\u53ef\u80fd\u3067\u3059\u3002\n    - \u3082\u3063\u3068\u3082\u53e4\u3044\u6307\u793a\u306f\u300e\u521d\u56de\u6307\u793a\u6587\u300f\uff081st PROMPT\uff09\u306b\u8a2d\u5b9a\u3057\u3066\u304f\u3060\u3055\u3044\u3002\n    - \u305d\u306e\u6307\u793a\u306b\u5bfe\u3059\u308b\u56de\u7b54\u306f\u300e\u521d\u56de\u6307\u793a\u6587\u306e\u56de\u7b54\u6587\u300f\uff081st COMPLETION\uff09\u306b\u8a2d\u5b9a\u3057\u3066\u304f\u3060\u3055\u3044\u3002\n    - \u6307\u793a\u6587\uff08PROMPT\uff09\u3068\u56de\u7b54\u6587\uff08COMPLETION\uff09\u304c\u30bb\u30c3\u30c8\u306b\u306a\u3063\u3066\u3044\u306a\u3044\u5c65\u6b74\u306f\u7121\u52b9\u3068\u306a\u308a\u307e\u3059\u3002\n    - \u6307\u793a\u6587\uff08PROMPT\uff09\u3068\u56de\u7b54\u6587\uff08COMPLETION\uff09\u306e\u30bb\u30c3\u30c8\u306f3\u56de\u5206\u307e\u3067\u767b\u9332\u3067\u304d\u307e\u3059\u3002\n- \u3059\u3079\u3066\u306e\u30d1\u30e9\u30e1\u30fc\u30bf\u3092\u8a2d\u5b9a\u3059\u308b\u5fc5\u8981\u306f\u3042\u308a\u307e\u305b\u3093\u3002\n    - \u30d1\u30e9\u30e1\u30fc\u30bf\u306e\u5024\u306f\u3001Playground \u306b\u3066\u8a66\u884c\u3067\u304d\u307e\u3059\u3002\n    - https:\/\/platform.openai.com\/playground?mode=chat\n    - \u8a2d\u5b9a\u53ef\u80fd\u7bc4\u56f2\u304c\u7570\u306a\u308b\u5834\u5408\u304c\u3042\u308a\u307e\u3059\u3002\n- \u30d1\u30e9\u30e1\u30fc\u30bf1\u884c\u76ee\uff1a\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u6e29\u5ea6\uff08temperature\uff09\n    - range:"[0,2]", default:"1"\n    - "Higher values like 0.8 will make the output more random"\n    - \u30e9\u30f3\u30c0\u30e0\u30ec\u30d9\u30eb\uff08\u5275\u9020\u6027\u5ea6\uff09\u3092\u8a2d\u5b9a\u3057\u307e\u3059\u3002\n    - \u4e8b\u5b9f\u8abf\u67fb\u3084\u96c6\u8a08\u306e\u5834\u5408\u306f\u30bc\u30ed\u306b\u8a2d\u5b9a\u3057\u307e\u3059\u3002\n- \u30d1\u30e9\u30e1\u30fc\u30bf2\u884c\u76ee\uff1a\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\uff05\uff08top_p\uff09\n    - range:"(0,1]", default:"1"\n    - "the model considers the results of the tokens with top_p probability mass"\n    - \u4e0a\u4f4d\u4f55\u30d1\u30fc\u30bb\u30f3\u30c8\u307e\u3067\u306e\u30ef\u30fc\u30c9Token\u3092\u8003\u616e\u3059\u308b\u304b\u3092\u8a2d\u5b9a\u3057\u307e\u3059\u3002\n    - \u300c\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u6e29\u5ea6\u300d\u3068\u4e21\u65b9\u3092\u8a2d\u5b9a\u3059\u308b\u3053\u3068\u306f\u63a8\u5968\u3055\u308c\u3066\u3044\u307e\u305b\u3093\u3002\uff08202303\u73fe\u5728\uff09\n- \u30d1\u30e9\u30e1\u30fc\u30bf3\u884c\u76ee\uff1a\u518d\u73fe\u7981\u6b62\u5ea6\uff08presence_penalty\uff09\n    - range:"[-2,2]", default:"0"\n- \u30d1\u30e9\u30e1\u30fc\u30bf4\u884c\u76ee\uff1a\u983b\u51fa\u7981\u6b62\u5ea6\uff08frequency_penalty\uff09\n    - range:"[-2,2]", default:"0"\n- Logit\u30d0\u30a4\u30a2\u30b9\uff1a\u30ef\u30fc\u30c9Token\u8abf\u6574\uff08logit_bias\uff09\n    - range:"[-100,100]"\n    - eg: "16504 10"\n    - eg: "14053 10"\n    - eg: "17013 -5"\n        - 16504(Japan), 14053(USA), 2937(US), 52(U), 13(.), 50(S), 41187(united) 17013(United)\n    - "values like -100 should result in a ban selection of the relevant token."\n    - \u7279\u5b9a\u306e\u30ef\u30fc\u30c9Token\u306b\u3064\u3044\u3066\u3001\u4f7f\u7528\u7981\u6b62\u306b\u3057\u305f\u308a\u3001\u591a\u7528\u3055\u305b\u305f\u308a\u3067\u304d\u307e\u3059\u3002\n        - https:\/\/platform.openai.com\/tokenizer\n- \u9ad8\u3044\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u6e29\u5ea6\uff08">1"\uff09\u3084\u9ad8\u3044\u30d0\u30a4\u30a2\u30b9\uff08">10"\uff09\u306f500\u30a8\u30e9\u30fc\u3084\u30bf\u30a4\u30e0\u30a2\u30a6\u30c8\u3092\u8a98\u767a\u3057\u307e\u3059\u3002\n    - \u3069\u306e\u7a0b\u5ea6\u306e\u8a2d\u5b9a\u6570\u5024\u304c\u59a5\u5f53\u304b\u306f\u3001\u696d\u52d9\u306b\u7167\u3089\u3057\u8a66\u884c\u932f\u8aa4\u306e\u3046\u3048\u6c7a\u5b9a\u3057\u3066\u304f\u3060\u3055\u3044\u3002\n\n\nAPPENDIX-ja\n- SYSTEM\u30ed\u30fc\u30eb\u8a2d\u5b9a\u306b\u306f\u3001\u6587\u7ae0\u4f5c\u6210\u8005\u3068\u3057\u3066\u306e\u5f79\u5272\uff08\u30ed\u30fc\u30eb\uff09\u3092\u8a2d\u5b9a\u3057\u307e\u3059\u3002\n    - e.g. "\u56de\u7b54\u6587\u306e\u5148\u982d\u306b\u300c\uff1c\uff1c\uff1c\u300d\u3092\u3001\u56de\u7b54\u6587\u306e\u672b\u5c3e\u306b\u300c\uff1e\uff1e\uff1e\u300d\u3092\u4ed8\u3051\u3066\u56de\u7b54\u3057\u3066\u304f\u3060\u3055\u3044\u3002"\n    - e.g. "\u8a9e\u5c3e\u306b\u306f80%\u306e\u78ba\u7387\u3067\u300c\u77e5\u3089\u3093\u3051\u3069\u3002\u300d\u3092\u3064\u3051\u307e\u3059\u3002"\n    - e.g. "\u30af\u30ea\u30a8\u30a4\u30c6\u30a3\u30d6\u306aQuestetra\u793e\u793e\u54e1\u3067\u3059\u3002"\n    - e.g. "You are a helpful assistant."\n        - https:\/\/platform.openai.com\/docs\/guides\/chat\/introduction\n- "\u30ec\u30b9\u30dd\u30f3\u30b9\u6570" \u3068\u3057\u3066\u30bb\u30c3\u30c8\u3067\u304d\u308b\u56de\u7b54\u6570\u8a2d\u5b9a\u306f128\u307e\u3067\u3067\u3059\u3002\n- "\u4e2d\u65ad\u6587\u5b57" \u3068\u3057\u3066\u30bb\u30c3\u30c8\u3067\u304d\u308b\u6587\u7bc0\u306f\uff14\u3064\u307e\u3067\u3067\u3059\u3002\uff085\u884c\u76ee\u4ee5\u964d\u306f\u7121\u52b9\uff09\n    - "\u4e2d\u65ad\u6587\u5b57" \u306b `\\t` \u7b49\u3092\u8a2d\u5b9a\u3057\u305f\u5834\u5408\u3001\u30a8\u30b9\u30b1\u30fc\u30d7\u51e6\u7406\u3055\u308c\u307e\u3059\u3002\n    - "\u4e2d\u65ad\u6587\u5b57" \u306b\u6539\u884c\u30b3\u30fc\u30c9\uff08`\\n`\uff09\u3092\u30bb\u30c3\u30c8\u3057\u305f\u3044\u5834\u5408\u3001`- - -` \u3092\u30bb\u30c3\u30c8\u3057\u307e\u3059\u3002\uff08\u8a66\u9a13\u7684\uff09\n        - \u30ec\u30b9\u30dd\u30f3\u30b9\u306f\u5e38\u306b\u4e00\u884c\u306b\u306a\u308a\u307e\u3059\u3002\n        - \u5fc5\u305a 500 \u30a8\u30e9\u30fc\u306b\u306a\u3063\u3066\u3057\u307e\u3044\u307e\u3059\u3002\uff08202303\u6642\u70b9\uff09\n- \u30ec\u30b9\u30dd\u30f3\u30b9\u6587\u306e\u751f\u6210\u8a31\u5bb9\u91cf\u3092\u5927\u304d\u304f\u8a2d\u5b9a\u3057\u305f\u5834\u5408\u3001\u30b7\u30b9\u30c6\u30e0\u5236\u9650\u3092\u8d85\u3048\u308b\u53ef\u80fd\u6027\u304c\u3042\u308a\u307e\u3059\u3002\n    - \u6587\u7ae0\u306e\u9014\u4e2d\u3067\u3042\u3063\u3066\u3082\u8a31\u5bb9\u91cf\u306b\u5230\u9054\u3059\u308c\u3070\u751f\u6210\u304c\u4e2d\u65ad\u3055\u308c\u307e\u3059\u3002\n        - `"finish_reason":"length",`\n        - "\u30ec\u30b9\u30dd\u30f3\u30b9\u6570" \u306b\u5927\u304d\u306a\u6570\u5b57\u3092\u30bb\u30c3\u30c8\u3059\u308b\u3068\u3001\u30ec\u30b9\u30dd\u30f3\u30b9\u4e2d\u65ad\u306e\u53ef\u80fd\u6027\u304c\u9ad8\u304f\u306a\u308a\u307e\u3059\u3002\n    - \u5408\u8a08\u30c8\u30fc\u30af\u30f3\u6570\u306b\u306f4096\u306e\u30e2\u30c7\u30eb\u5229\u7528\u5236\u9650\u304c\u3042\u308a\u307e\u3059\u3002\n        - \u6307\u793a\u6587\u306e\u9577\u3055\uff08PROMPT\u30c8\u30fc\u30af\u30f3\u6570\uff09\u306e\u60f3\u5b9a\u5024\u3092\u5f15\u3044\u305f\u5024\u3088\u308a\u5c0f\u3055\u3044\u5024\u3067\u30bb\u30c3\u30c8\u3057\u307e\u3059\u3002\n    - \u82f1\u8a9e\u306e\u5834\u5408\u3001\u5358\u8a9e\u3084\u8a18\u53f7\u3054\u3068\u306b1\u30c8\u30fc\u30af\u30f3\u306b\u306a\u308a\u307e\u3059\u3002\n        - \u5e73\u5747\u7684\u306a\u82f1\u8a9e\u306e\u5834\u5408\u30014\u6587\u5b57\u30671\u30c8\u30fc\u30af\u30f3\u7a0b\u5ea6\u306b\u306a\u308a\u307e\u3059\u3002\n        - \u65e5\u672c\u8a9e\u306e\u5834\u5408\u30011\u6587\u5b57\u304c\u8907\u6570\u30c8\u30fc\u30af\u30f3\u306b\u5206\u5272\u3055\u308c\u308b\u3053\u3068\u3082\u3042\u308a\u307e\u3059\u3002\n        - \u5e73\u5747\u7684\u306a\u65e5\u672c\u8a9e\u306e\u5834\u5408\u30011\u6587\u5b57\u30671\u30c8\u30fc\u30af\u30f3\u7a0b\u5ea6\u306b\u306a\u308a\u307e\u3059\u3002\n    - \u30ef\u30fc\u30c9\u30c8\u30fc\u30af\u30f3\u6570\u306e\u76ee\u5b89\u306f tokenizer \u306b\u3066\u78ba\u8a8d\u3067\u304d\u307e\u3059\u3002\n        - https:\/\/platform.openai.com\/tokenizer\n- \u6570\u5024\u578b\u30c7\u30fc\u30bf\u3092\u53c2\u7167\u3059\u308b\u8a2d\u5b9a\u3067\u306f\u3001\u6841\u533a\u5207\u308a\u6587\u5b57\u304c\u6df7\u5165\u3057\u306a\u3044\u3088\u3046\u306b\u6ce8\u610f\u3057\u3066\u304f\u3060\u3055\u3044\u3002\n    - \u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u95a2\u6570 `#sformat` \u3092\u4f7f\u3046\u3068\u4fbf\u5229\u3067\u3059\u3002\uff08Java String.format\uff09\n    - e.g. `#{#sformat("%1.1f", #q_numeric)}` \uff08\u5c0f\u6570\u7b2c\u4e00\u4f4d\u307e\u3067\u56db\u6368\u4e94\u5165\uff09\n    - "R2272: \u6587\u5b57\u5217\u3068\u3057\u3066\u306e\u51fa\u529b"\n        - https:\/\/questetra.zendesk.com\/hc\/ja\/articles\/360024292872-R2272\n- \u8907\u6570\u7d44\u7e54\u306b\u6240\u5c5e\u3059\u308b\u958b\u767a\u8005\u5411\u3051\u306e\u30d8\u30c3\u30c0\u306b\u306f\u672a\u5bfe\u5fdc\u3067\u3059\uff08202303\u6642\u70b9\uff09\n    - `OpenAI-Organization`\n- \u306a\u304a\u3001\u30ef\u30fc\u30af\u30d5\u30ed\u30fc\u30a2\u30d7\u30eaID\u304c\u3001OpenAI API \u306e `user` \u30d1\u30e9\u30e1\u30fc\u30bf\u306b\u81ea\u52d5\u7684\u306b\u4ee3\u5165\u3055\u308c\u307e\u3059\u3002\uff08\u8a66\u9a13\u7684\uff09\n    - `processInstance.getProcessModelInfoId()`\n\n*\/\n\n<\/code><\/pre><\/div>\n<\/div><\/details><\/div>\n<\/div>\n\n\n\n
Download<\/span><\/h3>\n\n\n\n