# Chat The "Chat" menu item includes tabs with all previously conducted dialogues with the neural network, as well as the ability to create a new dialogue and customize it according to the User's individual needs. It also implements the ability to attach files/folders and save dialogues, making the communication process between the neural network and the User more convenient and efficient. The "New Dialogue" button allows you to create a new dialogue with the neural network. If you hover the mouse cursor over this button, a chat settings icon will appear on the right: ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXd1S6lr2REB69BOCksWmkH_H7SdpwD-JUS_SU88fqbF1YLzPlalqRp-DeIyKn_4k9he6ETxwjTJVmCMzr8am5GuxJGtmcC_Y65lZz_q3sG1lrB-eyneklkT0W17I6D06bekxAVp8g?key=jXxpQJRYkQW6F4d0HoRgIxP1). Clicking on it will open the default Chat Settings popup:

This window contains the fields: * **Model** In this field, you can specify the name of one of the pre-installed large language models (LLM) available on the platform. * **System Message** In this field, you can specify a system message that will be sent to the neural network for processing requests. This message can contain instructions on how the neural network should behave, what style to respond in, or what information to use. * **Temperature** The slider allows you to adjust the creativity of the neural network's responses — the higher the value, the more random and less predictable the responses will be. This can be useful in situations where more variety in responses is required. In the upper right corner, a counter is displayed ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeWJ7KQKYNZriOkkYlociECWWgJMyf3r9EEZWlgmSWhR26tV8LTEMBEsg43lSeKW-HFPJIcZkw5T_E4cdeDYTA2oVV-OVbNIJBqwVEfDmAUfXitkIw6jBvua47vfVJ0ekn8zg1Avg?key=jXxpQJRYkQW6F4d0HoRgIxP1), where you can check the value in digital equivalent or enter it manually. The "Temperature" setting allows you to regulate the creativity of the neural network on a scale from 0 to 1, where: 0 – no creativity. All responses to the same questions will be almost identical, strict, and concise. This mode is suitable for situations where precise answers or solutions to coding tasks are required; 1 – maximum degree of creativity. All responses from the neural network will be different and unpredictable. Suitable for writing articles and creative texts. By default, the "Temperature" value is set at 0.7. If you are working with documents, it is recommended to set the value to 0.4. After setting the appropriate values in these fields, you need to click the OK button. The entered settings will be saved and applied to all newly created Chats, but will not affect existing ones. > It is not possible to change the settings of an already existing Chat after the dialogue has started. ### Create a New Dialogue After selecting "New Dialogue" in the Sherpa AI Server workspace, a new dialogue window will open:

After the first response to the User's message, the neural network will rename the Chat depending on the context of the dialogue. To send a text message, click the button ![](https://3237142148-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FscP4BXwl9ufpJr5mfVln%2Fuploads%2Fgit-blob-2c7ae3971935300a4973dba77601564528239cc6%2Fimage%20\(4\).png?alt=media) to the right of the "Message in chat..." field. To send a voice message, click the button ![](https://3237142148-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FscP4BXwl9ufpJr5mfVln%2Fuploads%2Fgit-blob-b2b0b26f2240df26c100110284736431bc8283fd%2Fimage%20\(1\)%20\(1\).png?alt=media) and allow microphone access in the popup window:

Finish recording by pressing the button

and the transcribed text will automatically appear in the message input field. If you want to set the Chat name yourself, hover the cursor over the name that needs to be changed and click the "Edit" icon: ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXc_Y1s8hS3aNd8p5kFuBziluJTCx4eLlpJN6TgemUTnTIJjTJtlkbDhWpKcBbRdVPkCVSywe_d5S3ELQsUsZSlFQCA_Ix03ufPQ-ImdOzGe2hmhzc74tNk8rl_UNuOs7iouXi74?key=jXxpQJRYkQW6F4d0HoRgIxP1). Enter the new name in the popup window and confirm the name change by clicking the "OK" button:

To delete the dialogue with the neural network, you need to click the button ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXc93vYLup8zm7QBN6MRKHej22ctCnSJDa4w76VYi25jErOEwEe2sBG4P3l0RvI0-UnerzjhxlbMX1bGWegOZYFyxpEcIyUAm0_26IkKetThyk8sylg2TEjIYqyMumaJM9zgRwtBDg?key=jXxpQJRYkQW6F4d0HoRgIxP1) and confirm the deletion by clicking the "Yes" button in the popup window:

### Attach Files and Folders In addition to the standard text dialogue, you can attach one or more folders with documents or files from your computer to your message. To do this, use the buttons to the left of the "Message in chat..." field:

To add a folder to the message, click the button ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXfEnfWOs_MdCBubivj3BXBXfS0PBJsanlWSG78sJaBdWMLhjZafUGgMKrH1942IaphdUXnmgR85QE-rHWTODB11Y6LzocgH3Z4neAX5zXez8q-he1ssLDRZ2RUqAJ71-eTYCMzheg?key=jXxpQJRYkQW6F4d0HoRgIxP1), check the checkboxes to the right of one or more folders that are needed in the dialogue:

If there are many folders, use the "Search" field to find the folders needed in the dialogue. Click the "Add to chat" button.

The added folder can be deleted by clicking the cross in the upper right corner:

> Folders are created using the "Documents" menu item on the left panel. To add a text file from your computer, click the button ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXdv94_HT3ZZm0hgsE0n8wPWnGdrmJe65cU8RtAGL640lN-VBCU6O22XwIYA5CRjx0i5gRmgY4v0HPPvdgjGKxXNwKV56MBdq9Z8i4Lt59Cp99T_DhtxQourF7l3safx1saCRgSZHQ?key=jXxpQJRYkQW6F4d0HoRgIxP1) and in the opened window select the desired file (any text file formats are supported: `.doc`, `.docx`, `.xls`, `.xlsx`, `.txt`, `.pdf`) and click the "Open" button:

You can also attach a file by dragging it into the Chat window. You can attach multiple files of different types. After attaching, the file is indexed. Once indexing is complete, the file can be sent to the neural network. The added file can be deleted by clicking the cross in the upper right corner:

In all created Chats, files and folders attached to user messages are saved. The neural network can use citations from files in its responses. Citations are marked with the icon ![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeEg77qKmYWmCvv_L8rN100kKP8jlILo0Xbo0y0CC5lZtP6CD07ubU_2zxOWLSNOtA4geaV80C-JxMv7ovke0QglCHJJ8awfQteiQoD0ju71jmmufBHPAMORNOOFa5pQ9SMkdI0HA?key=jXxpQJRYkQW6F4d0HoRgIxP1), and there can be several of them. To view the citation, click on the icon, after which a popup window will open with the citation and its source. For example:

To close the citation window, click the "OK" button in the lower right corner of the popup window. ### Additional Chat Features To open the dropdown list with additional Chat features, click the button ![](https://3237142148-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FscP4BXwl9ufpJr5mfVln%2Fuploads%2Fgit-blob-d8672003ba78371e43da58e88d6fba48ebcf0f2d%2Fimage%20\(6\).png?alt=media) to the right of the "Message in chat..." field:

The list includes the following buttons:


Button	Description
Reset Chat	Allows you to delete all messages in this Chat while keeping the tab with its name in the left panel for a new dialogue.
Request	Enables a mode where each message is treated by the neural network as a separate, isolated request. This means that the neural network does not consider previous messages and the context of communication. Users can use this mode to test the neural network's response to specific requests without worrying about how the previous message affected the neural network's answer.
Full-text Search	Searches for matches in the texts within files/folders, then sends the User chunks or links to the relevant data in the Chat. Before starting the Full-text search, the User must select the necessary files or folders that were previously uploaded in the "Documents" tab.
Code Interpreter	Enables a mode where the execution and debugging of code generated by the neural network occurs. Allows running scripts (Python, SQL, PowerShell, etc.) directly in the chat with output of execution results, error checking, and interactive debugging.
Dialogue	Enables a mode where the neural network considers the entire context of communication. The neural network can use information from previous messages to form more relevant and justified responses. This mode allows the neural network to remember user preferences and adjust its responses based on the style of communication or interests. Users can ask clarifying questions and develop the topic, making communication more dynamic and engaging.

--- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.sherparpa.ru/en/sherpa-ai/sherpa-ai-server/rabota-v-sherpa-ai-server/chat.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.