7 commits

Author	SHA1	Message	Date
Pat Wendorf	c947e08cdc	Allow for adjustment of temp, top_p and max tokens for models that have issues with repeating.	2025-02-08 08:54:29 -05:00
Pat Wendorf	933b9e24d4	Skip transcribe step if it's already done. Show each part being summarized.	2025-02-03 16:30:52 -05:00
Pat Wendorf	80637cc26c	Simplified the transcriber by using the openai library to access llama.cpp server or ollama server mode.	2024-11-29 09:06:22 -05:00
Pat Wendorf	3c05cd78dd	Used mistral large to re-write the summarize process.	2024-08-19 07:41:32 -04:00
Pat Wendorf	81e4e165bc	Added some silence trimming	2024-07-30 11:56:40 -04:00
Pat Wendorf	f5af4960fc	Extract some fact based summaries now as well	2024-06-24 09:52:38 -04:00
Pat Wendorf	69d269d2f0	first commit	2024-06-21 12:52:10 -04:00