ozgrozer / ai-renamer Goto Github PK

It should be pretty simple to add Llava support, since it just accepts the contents of an image as a base64 image url. Could add it to readFileContent.js as a special case like .pdf

Ai-renamer focused on the screenshot folder and processed some images with the same file name. This resulted in the photos being overwritten when the application was completed.

Docker support 🐳
It'd be super helpful if we could get a Docker setup for this.
Dry-run option 🏜️
Could we get a "dry-run" or "preview" mode?
This would be a lifesaver for checking things before making any changes.

Keep a backup and a rollback feature in case a picture gets renamed by accident.

for example, i only want to rename files that matchScreenshot* or only *.png or only ./directory/*.jp?g

Seems that as it's going through old photos and renaming them, it occasionally pops up this error message.

All media is JPG format, however I am suspecting the LLM (llava:13b) not able to open the container or view it due to an older format or implementation of JPEG vs JPG, even if JPEG is renamed.

Would suggest that you add better error correction such as:

1. Cannot change name of file "X"

• Reason is I can move the file out of the directory or...

2. Skips the file and continues without failing back to the prompt

• Reason is better error handling and not being pushed back to command prompt / failing out

3. Not sure if it's possible to recognize pictures already renamed?

• Reason is when I rerun the script to rename the pictures, it renames the renamed ones already. Not an huge issue but perhaps a better time savings?

Thank you for this fine implementation of ollama capabilites!

Any plans to support the OpenAI API standard to do exactly what you're doing now?

It'll help with batching large files that can't be processed locally / using a more capable model to perform the same task

Json data documents constructed as such
[{"Fecha":"05/09/2021 17:24:51","Resultado":251.05},{"Fecha":"05/09/2021 17:25:12","Resultado":270.09},....]

Fails with this message
ENOENT: no such file or directory, rename 'E:\Documentos\test\mediciones-potencia.json' -> 'E:\Documentos\test\Here is a concise, descriptive filename in kebab-case format:

2021-09-05-scores.json

This filename includes the date (YYYY-MM-DD) and the type of data being stored (scores). The .json extension indicates that the file contains JSON-formatted data..json'

Not sure why it fails but here it is.
And thanks for the work!

The current implementation only processes files in the top-level directory specified by the input path. i would like if code had the ability to recursively process all files in subdirectories as well based on user choice.
I made necessary changes in my local , if it's ok i can raise a pull request.