-
### Feature request
Now that the context window is variable (per #1668) it would be helpful to have models2.json updated to populate the n_ctx field along with the correct system and user prompts.
…
-
Determine the best chunk size for our application.
The application should be able to keep track of individuals, so that RAG can make an appropriate prompt to feed to LLM, and the application can gi…
-
If you upload a PNG, which can be OCRed fine with the new Document Intelligence, and then ask a question on it, you'll see this error:
Traceback (most recent call last):
File "/works…
-
## Dimensions
1. Complexity of PDFs structure => Single PDF, 2 Delinked PDFs, Linked PDFs, PDFs with Images, PDFs with Tables.
2. Complexity of Questions => Fact Parrot, Neural Coref, Reasoning, Calcu…
-
Make the following algorithm:
INPUT: genus 2 curve over Q
OUTPUT: the non-semistable reduction primes of its Jacobian
It's probably just the primes dividing the conductor with some condition on…
-
Example of UI style: sonnet 3.5 with the workspace area
Things done so far:
Notes:
- Unable to use an accordion for 'Advanced' options on front page. Breaks formatting. Probably a skill i…
-
Links about all kind of models, but the preference is for ones which can be practically applied with less resources etc.
# Images
* Free Stable Difussion online - but no fine control: https://s…
-
I'm playing around with MicroWebSrv2 on an esp32.
The following demo features are working just fine:
*Test a POST form
*Test the WebSockets page
*Test the multi-user chat page
But these links…
-
Might be useful to be able to do this:
```bash
git diff | llm chat -m gpt-4 --in
```
Where the content doesn't actually get sent to the model at all (yet), but is instead added to the history an…
-
ng