TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: Is there a ChatGPT “in a box” for private data?

3 点作者 highwayman47超过 2 年前
For example, say I want to upload a few gigabytes of proprietary text from business and have it be a knowledge database or private knowledge.<p>An example of this is: https:&#x2F;&#x2F;huberman.rile.yt&#x2F; which is for Huberman lab podcast.<p>The developer describes it here: https:&#x2F;&#x2F;twitter.com&#x2F;rileytomasek&#x2F;status&#x2F;1603854647575384067<p>As a non-programmer I&#x27;m not sure how this was done, but wondering if there is a open source or no-code &#x2F; turn key solution available.

1 comment

joshka超过 2 年前
ChatGPT gives a pretty good answer to this (whether it&#x27;s accurate is another story):<p>===<p>Yes, it is possible to use the ChatGPT model to generate text based on a private dataset. However, keep in mind that the ChatGPT model was trained on a large public dataset and may not be well-suited for generating text based on a small private dataset without further fine-tuning.<p>To use ChatGPT with a private dataset, you will need to follow these steps:<p><pre><code> Preprocess your data and split it into training and evaluation sets. Fine-tune the ChatGPT model on your training data using a library or framework like Hugging Face&#x27;s Transformers. This involves specifying the model architecture, defining a loss function and an optimizer, and training the model on your data. Evaluate the fine-tuned model on your evaluation data to see how well it performs. If the model&#x27;s performance is not satisfactory, you may need to adjust the model architecture, training parameters, or the data itself to improve the model&#x27;s performance. </code></pre> Fine-tuning the ChatGPT model on a private dataset can be a complex task that requires programming skills and some knowledge of machine learning. If you are a non-programmer, you may want to consider using a turn-key solution like Hugging Face&#x27;s API, which allows you to fine-tune and use language models like ChatGPT without having to write any code.<p>Alternatively, you could consider using a chatbot platform like Dialogflow or Botpress, which allow you to build chatbots without programming knowledge by providing a graphical user interface for defining chatbot interactions and integrating them with various messaging platforms. However, keep in mind that these platforms may not offer the same level of customization and flexibility as a self-hosted solution like ChatGPT.