...
... """
To create a new folder named "projects", the appropriate action command would be:
``` [WRITE_FILE] projects/hello.py ```
To list the contents of the current directory, the command used would be:
``` [EXEC_COMM] ls ```
To create a new file named "filename.txt" with content "content here", the command used would be:
``` [WRITE_FILE] filename.txt ```
To modify an existing file named "file1.py", the command used would be:
``` [MV_COMM] edit file1.py ```
To extract a specific text from a large file named "file.txt", the command used would be:
``` [GZIP_COMM] grep -i 'search term' file.txt ```
tried to give him a list of commands, :
1. **Executing Terminal Commands:**
... To execute terminal commands, use the action tag [EXEC_COMM] followed by the specific command you wish to execute. For example, to list the contents of the current directory, you would use the command:
... Example:
...
... [EXEC_COMM] ls
...
...
... a. **Available Commands:**
... ls, // list contents of current directory
... mkdir, // create a new directory
... cat, // print contents of a file
... cp, // copy a file
... mv, // move a file
... rm, // remove a file
... touch, // create a new file
... find, // find a file
... ping, // ping a server
... wget, // download a file
... curl, // download a file
... ssh, // connect to a server
... scp, // copy a file to a server
... tar, // archive a file
... gzip, // compress a file
... gunzip, // uncompress a file
... zip, // archive a file
... unzip // unarchive a file
commands
llama2 seem to do better in this case, is this related to the settings or the model itself?
I tried 7b and this problem disappear with gpt4, I can't run 70b on my machine but can try on google colab.
