How to customize typechat for my own model and personalized scenarios (exp. for Chinese users)? #86

Activity

1. DanielRosenwasser commented on Aug 3, 2023

   

   Daniel Rosenwasser (DanielRosenwasser)

   on Aug 3, 2023

   Member

   More actions

   I am modifying model.ts to fit my own model

   I'll just note that TypeChat has a bring-your-own-model approach. You don't need to fork the library, you just need to provide an object with a complete method.

   Here's an example

   const model: TypeChatLanguageModel = {
       complete: async function complete(prompt) {
           const response = await fetch("https://api.openai.com/v1/chat/completions", {
               method: "POST",
               headers: {
                   Authorization: `Bearer ${apiKey}`,
                   "Content-Type": "application/json",
               },
               body: JSON.stringify({
                   model: modelName,
                   temperature: 0,
                   n: 1,
                   "messages": [
                       { "role": "user", "content": prompt }
                   ],
               })
           });
   
           assert(response.body);
   
           try {
               return success(await response.text());
           }
           catch (e) {
               return error(e instanceof Error ? e.message : String(e));
           }
       }
   }

   About other languages like Chinese - I'm not so sure. Maybe others have tips and ideas?

   👍React with 👍3Reacted by qftie, Rafhael Oliveira and Alexandre Godard
2. Daniel Rosenwasser (DanielRosenwasser)

   added

   QuestionFurther information is requestedFurther information is requested

   DiscussionInformal discussion among communityInformal discussion among community

   on Aug 3, 2023
3. weykon commented on Aug 4, 2023

   

   Weykon (weykon)

   on Aug 4, 2023

   Last edited by weykon

   Contributor

   More actions

   Hi! I've completed the first level of the experiment, which involved making additional translation requests based on the returned results. It seems that we need to obtain its output before we can begin the transformation. I have tried adding new requirements for translation in the createRequestPrompt method in typechat, but I have attempted multiple prompts to construct a structured processing flow, even with GPT-4, but have not been successful. So, in order to ensure a good result, here are a few suggestions:

   1. Before executing typechat, you can translate your schema file (using GPT or other simple translation services), where all string types are translated into Chinese. I have tried this, and if the content of your schema is in Chinese or another language, the values in the output JSON data will be translated accordingly.

   2. After executing typechat, you can perform translation on the resulting output. I achieved this using a simple prompt.

   Prompt

   --

   const prompt =`${JSON.stringify(cart, undefined, 2)}\n above is json string all of the following JSON key and value, the values of string need to translate into Chinese, but the JSON key name not change, such like "items": [
           {
             "type": "lineitem",
             "product": {
               "type": "LatteDrinks",
               "name": "latte"
             },
             "quantity": 1
           }
         ] will translate to chinese: 
           [{
             "type": "行项目",
             "product": {
               "type": "拿铁饮料",
               "name": "拿铁"
             },
             "quantity": 1
           }]
        and keep JSON data structure`;

   This way, you can always translate the current output, reducing the amount of translation needed. However, please note that translating the data may not preserve the same types as defined in the schema file. Therefore, this method is more suitable for displaying text data rather than precise data.

   3. Another approach, which I haven't achieved yet, is to granularly parse each JSON key and value within typechat and handle the required actions. This method should also be feasible but requires more experimentation time. For example, in my project, there is a section for translating Markdown. A one-shot experiment similar to this prompt can be found in this https://github.com/weykon/translate-md-by-gpt/blob/main/pmt_set.ts . (the original code is quite long, so please check the link for details.)

   Using this method of enabling translation options may result in two completely different sets of prompts, which increases complexity.

   4. I believe a better approach would be to use the i18n (internationalization) concept and create multiple language schema files, such as schema.en.ts, schema.zh.ts, schema.it.ts, to obtain more accurate outputs.

   So, currently, I think the fourth method is a better choice.

   ❤️React with ❤️2Reacted by qftie and Rafhael Oliveira
4. ybsun0215 commented on Sep 1, 2023

   

   cody (ybsun0215)

   on Sep 1, 2023

   More actions

   Hello, have you encountered an error after mounting vpn? How was it resolved?
5. robgruen commented on Oct 15, 2026

   

   robgruen

   on Oct 15, 2026

   Contributor

   More actions

   Following up on cody (@ybsun0215)'s VPN/proxy question in this thread: built-in HTTP(S) proxy support was added in #382, merged on July 7, 2026.

   References:

   • Implementation: Add fetch-native proxy support (HTTPS_PROXY / HTTP_PROXY / NO_PROXY) #382 and the model options and proxy-handling code.
   • Setup: Using a proxy.
   • Related issue: A network error occurred while running the example #117 reported timeouts with a VPN and ignored proxy settings; it was closed as addressed by Add fetch-native proxy support (HTTPS_PROXY / HTTP_PROXY / NO_PROXY) #382.
   • Related discussion: feat: export createAxiosLanguageModel #94 requested custom HTTP-client configuration, including proxies. The current implementation uses native fetch, rather than exposing Axios internals.
   • Earlier proposals: Add HTTP_PROXY variable and proxy process when access endpoints' api services #55 and Support axios config #73 were superseded by Add fetch-native proxy support (HTTPS_PROXY / HTTP_PROXY / NO_PROXY) #382.

   createLanguageModel(process.env) now honors HTTPS_PROXY, HTTP_PROXY, and ALL_PROXY (including lowercase variants), along with NO_PROXY. Install the optional proxy dependency with npm install undici, then configure your actual proxy, for example:

   HTTPS_PROXY=http://127.0.0.1:7890

   When calling createOpenAILanguageModel or createAzureOpenAILanguageModel directly, pass { proxyUrl } through the options argument instead; the setup documentation has the signatures.

   This addresses proxy routing, not the original questions about MPT output quality or Chinese-language prompting. A VPN or proxy can still have separate connectivity problems.
6. robgruen commented on Oct 15, 2026

   

   robgruen

   on Oct 15, 2026

   Contributor

   More actions

   Hi! I've completed the first level of the experiment, which involved making additional translation requests based on the returned results. It seems that we need to obtain its output before we can begin the transformation. I have tried adding new requirements for translation in the createRequestPrompt method in typechat, but I have attempted multiple prompts to construct a structured processing flow, even with GPT-4, but have not been successful. So, in order to ensure a good result, here are a few suggestions:

   1. Before executing typechat, you can translate your schema file (using GPT or other simple translation services), where all string types are translated into Chinese. I have tried this, and if the content of your schema is in Chinese or another language, the values in the output JSON data will be translated accordingly.
   2. After executing typechat, you can perform translation on the resulting output. I achieved this using a simple prompt.

   Prompt

   const prompt =${JSON.stringify(cart, undefined, 2)}\n above is json string all of the following JSON key and value, the values of string need to translate into Chinese, but the JSON key name not change, such like "items": [ { "type": "lineitem", "product": { "type": "LatteDrinks", "name": "latte" }, "quantity": 1 } ] will translate to chinese: [{ "type": "行项目", "product": { "type": "拿铁饮料", "name": "拿铁" }, "quantity": 1 }] and keep JSON data structure;
   This way, you can always translate the current output, reducing the amount of translation needed. However, please note that translating the data may not preserve the same types as defined in the schema file. Therefore, this method is more suitable for displaying text data rather than precise data.

   3. Another approach, which I haven't achieved yet, is to granularly parse each JSON key and value within typechat and handle the required actions. This method should also be feasible but requires more experimentation time. For example, in my project, there is a section for translating Markdown. A one-shot experiment similar to this prompt can be found in this https://github.com/weykon/translate-md-by-gpt/blob/main/pmt_set.ts . (the original code is quite long, so please check the link for details.)

   Using this method of enabling translation options may result in two completely different sets of prompts, which increases complexity.

   4. I believe a better approach would be to use the i18n (internationalization) concept and create multiple language schema files, such as schema.en.ts, schema.zh.ts, schema.it.ts, to obtain more accurate outputs.

   So, currently, I think the fourth method is a better choice.

   Please see my response for #142
7. robgruen

   closed this as completedon Oct 15, 2026