Check:

• disk space enotgh to download this model
• raw enough to start this model

Hello.

I am happy with this extension. It is the only one, that works without any dependencies and therefore it is instantly usable within VScode. Thank you for bringing us such an extension. I have some requests:

It would be nice, if we could have a option to select the *.gguf model file by our own. I can download it somewhere (f.e. huggingFace) and inside the options of FireCoder i can select the file. I recognized that the models are saved inside C:\ drive. inside the USERS subfolder: ".firecoder". If its possible it would also be very handy, if we could have an option where to store downloaded models.

In my case, the drive C: is reserved for the OS and stays clean. I put heavy stuff onto the D: partition. (Windows OS). I managed to use the MkLink command, to alias this folder to drive D:.

Based on: #22

After a little testing sha256sum for linux I saw improving in 10 time. 4s vs 0.4s for the small model.

Hello.

When i install FireCoder vsix file offline with VScode, i got these errors:

• Server file info not found (while starting VScode)
• fetch failed (when start chat)

It seems he can not find the server.exe file ? Maybe it would be a good idea to add a option for defining a own path to its file inside the FireCoder-Extension-Settings.

Surprisingly, if i am online and i install FireCoder vsix file with VScode, than it works. Only offline install seems not to work.
I will try to investigate a bit more into this.

https://stackoverflow.com/questions/73043106/how-to-get-highlighted-text-from-vscode-extension-api

FireCoder currently uses a tokenizer to determine the maximum length of the prompt for autocomplete. To achieve this, FireCoder sends the text to the llama.cpp tokenizer endpoint. However, this process takes time and cannot be used if the user is working on the cloud. It is important to provide as much context as possible, but the current method has some issues.

Firstly, to use the llama.cpp tokenizer, the user must download the server and model. However, this is not convenient for users who want to work with the cloud.
Secondly, preparing a prompt can take more than 2 seconds, which can be time-consuming.
Finally, FireCoder has a complex algorithm for selecting the maximum suitable length of the prompt with the minimum request to llama.cpp.

The solution is to use a local tokenizer that can be directly called from the extension. There are two possible options for this:

1. Use tokenizers, but it works poorly when combined with nodejs bindings, so further investigation is needed.
2. Use transformers.js, which should work well, but it still needs to be tested.

vscode.workspace.onDidChangeConfiguration

TODO:

• add chat models for downloading
• add a setting for using chat
• start the server with chat model if firecoder.experimental.chat is true
• add webview for showing chat
• • use this example as first step
• add a wraper for request to chat with supporting chat template
• add chat history class

Version: 1.88.1
Commit: e170252f762678dec6ca2cc69aba1570769a5d39
Date: 2024-04-10T17:34:12.840Z
Electron: 28.2.8
ElectronBuildId: 27744544
Chromium: 120.0.6099.291
Node.js: 18.18.2
V8: 12.0.267.19-electron.0
OS: Linux x64 6.7.10-200.fc39.x86_64 snap

Log:

llama> /home/gespispace/.firecoder/server: error while loading shared libraries: libcudart.so.11.0: cannot open shared object file: No such file or directory

2024-04-17 17:31:05.150 [trace] llama> child process exited with code 127

Hello.

As i read FireCoder is in its early stage. Maybe it is the perfect time, to give some suggestions what could improve the usability of FireCoder. These things would be nice:

• a copy button, to copy the generated code-text
• a cancel / stop button to stop the ai commenting more text, because the code is finished but the AI explaination has a too long text
• a drop-down menu where we can store the chats (with the possibility to delete them by a small X-button)
• a mouse-button rightclick context menu, when selecting code inside the VScode editor. Inside such a context menu
  could be something like: analyze code, find problem, reformat code, etc. This could be send back to FireCoder and it will
  automatically start the chat.
• Start / Stop Server Button: Because sometimes i dont need FireCoder. And if i am not able to deactivate it, my RAM is full. So deactivating the Server manually would free RAM space
  [EDIT: When starting VScode, the server should not start automatically. imho. But this could be handled as a own preference by the user inside the FireCoder Settings ]

I think these few things, by adding some "Buttons" and "Menus" could improve the usability of FireCoder.

https://github.com/microsoft/vscode-extension-samples/tree/main/getting-started-sample

This will work only for manual mode and should improve hits.

https://stackoverflow.com/questions/72404038/how-to-download-big-file-resume-download-after-interuption-in-node-js