Michael Mior
Michael Mior
```json { "chat_template": "{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% set system_message = false...
@carmocca It is true that there is no `tokenizer.model` and instead there is `tokenizer.json` and `tokenizer_config.json`. To clarify, this is just the output from litgpt and I didn't modify any...
@carmocca I downloaded `codellama/CodeLlama-7b-Instruct-hf` with `litgpt download` and then used that checkpoint.
This looks great @joscha! I might take this as an excuse to finally start using CoffeeScript :)
Thanks! I'll keep that in mind. Unfortunately I realized the project I was planning on isn't feasible (for non-technical reasons), but I'll definitely check it out if I think of...
I seem to be experiencing a similar. I copied a working configuration from one project to a new project and now I am unable to publish with an HTTP 422...
Here's a simple example that fails for me. ``` CREATE TABLE public.foo(a INT); COPY public.foo(a) FROM stdin; 1 2 3 \. ``` With no schema, it works fine.
It might be worth noting in the README that while the sed solution will probably work to get the data loaded, it has the potential to corrupt your data. If...
Would love to see this merged. I have some files I want to start automatically formatting but they have newlines at the end and I don't want to change that...
The example given by @maxfl doesn't seem to work for me. This would be very cool to have working somehow in any case.