-
Hi,
What are the recommended settings to use BLIP as a paragraph captioner?
Does anyone have any experience with that?
I'm trying to modify `self.model.generate(frame, sample=False, num_beams=3, …
-
### Bug Description
`MarkdownNodeParser` extracts any MD headings into the respective metadata field, but also directly adds it to the `text`.
As a consequence, and since all metadata are by defau…
-
As pointed out by @TWiStErRob kindly in #3928 , we have some defect in our documentation generation infrastructure:
> While doing so I noticed that there are bigger issues at hand:
> 1. `
-
Related to #1038.
I've done some research concerning Markdown line lengths and XML indention.
Markdown line length limit:
- GitLab: [100 characters](https://docs.gitlab.com/ee/development/docum…
-
- [ ] [SELF-RAG: Learning to Retrieve, Generate and Critique through Self-reflection](https://github.com/AkariAsai/self-rag/blob/main/README.md?plain=1)
# SELF-RAG: Learning to Retrieve, Generate and…
-
This is an open-ended issue; I expect there will be more than one solution to this.
There have been a couple of solutions for constraining the output of generations:
- https://twitter.com/GrantS…
-
Don,
I’ve had a chance to use the plugin on a small project. If you never did another thing to it, this plugin is an excellent tool–exactly what I needed.
Along the line of feature requests for …
-
Trying to run the augment toolkit on MacOs M3, with ollama (ollama run llama3) on the following config.yaml
PATH:
INPUT: "./raw_text_input"
OUTPUT: "./output"
DEFAULT_PROMPTS: "./prompts" # …
-
Write in Python the pure iterative version of the function defined in the [previous exercise](#29).
-
### Steps to reproduce:
1. First and foremost, make sure the Screen Layout mode is turned off. It can be done via Insert+V when placed on a webpage.
2. The issue can be duplicated everywhere but…