v1.0.18

What's Changed

• Fix growing context on repeated requests by @MHHukiewitz in #17

Full Changelog: v1...v1.0.18

v1.0.17

What's Changed

• Add CTX_SIZE for CodeLlama max context size by @MHHukiewitz in #16

Full Changelog: v1.0.16...v1.0.17

v1: Add CTX_SIZE for CodeLlama max context size (#16)

* Add CTX_SIZE as an optional parameter with a default value of '64000' and use it in the main.py file.

* Change delay time from 2 seconds to 30 seconds in query_llama function.

* Remove URL from error message to maintain URL secret

* Update 'n_keep' value in dictionary to be an integer instead of a string.

v1.0.13

What's Changed

• Improve readability of comments with markdown formating by @MHHukiewitz in #11

Full Changelog: v1...v1.0.13

Add retry mechanism to Llama query

If your Llama is a bit weak on its knees, this release will give it more chances to answer correctly.

v1.0.11: Refactor main.py to add support for dry run mode. (#9)

* Refactor main.py to add support for dry run mode.

* Refactor code to handle different category values for comments in main.py.

v1.0.10

What's Changed

• Improve output robustness by @MHHukiewitz in #8

Full Changelog: v1.0.8...v1.0.10

v1.0.8

What's Changed

• Fine-tune system prompt by @MHHukiewitz in #7

Full Changelog: v1.0.7...v1.0.8

v1.0.7

What's Changed

• Fix checkout of correct repo by @MHHukiewitz in #6

Full Changelog: v1.0.6...v1.0.7

v1.0.6

What's Changed

• Improve input robustness and workflow efficiency by @MHHukiewitz in #5

Full Changelog: v1.0.5...v1.0.6