Struggling to run using only a local model

[electricazimuth]

Looks like open ai is hardcoded here, I think it should be referencing the selected model I've setup in my settings (a local server on my network)

loom/util/multiverse_util.py

Lines 8 to 18 in 91ca920

	def generate(prompt, engine, api_base, api_key):
	openai.base_url, openai.api_key = api_base + '/', api_key
	#print('calling engine', engine, 'at endpoint', openai.api_base)
	#print('prompt:', prompt)
	response = openai.completions.create(prompt=prompt,
	max_tokens=1,
	n=1,
	temperature=0,
	logprobs=100,
	model=engine).dict()
	return response