25 Commits

Author	SHA1	Message	Date
Simon Gardling	ffc079fb21	llama-cpp: use gpt-oss-20b-mxfp4	2025-08-05 19:53:20 -07:00
Simon Gardling	7f7dc03a20	extend nixpkgs's lib instead	2025-07-11 20:40:27 -07:00
Simon Gardling	ce7801795f	DeepSeek-R1-0528-Qwen3-8B	2025-06-06 20:05:42 -07:00
Simon Gardling	7c217d6ead	llama-cpp: use q8 quantization instead of q4	2025-05-28 21:20:42 -07:00
Simon Gardling	4dc577fdcb	llama-cpp: disable gpu	2025-05-28 21:09:45 -07:00
Simon Gardling	109e132497	llama-cpp: vulkan broken	2025-05-28 21:04:34 -07:00
Simon Gardling	d0da2591a3	llama-cpp: disable flash attn	2025-05-28 21:00:08 -07:00
Simon Gardling	a292c2fc75	llama-cpp: nvidia-acereason-7b	2025-05-28 20:59:45 -07:00
Simon Gardling	fb4043712e	llm: use vulkan	2025-04-30 12:59:07 -04:00
Simon Gardling	843b64a644	llm: use xiomo model	2025-04-30 11:22:52 -04:00
Simon Gardling	b00ad9a33a	deepcoder 14b	2025-04-14 13:11:40 -04:00
Simon Gardling	db78740db3	change llm model	2025-04-10 11:15:18 -04:00
Simon Gardling	fe85083810	llm: model stuff	2025-04-08 00:22:12 -04:00
Simon Gardling	3653e06c7d	create single function to optimize for system	2025-04-07 14:33:34 -04:00
Simon Gardling	b764d2de45	move optimizeWithFlags	2025-04-07 14:31:56 -04:00
Simon Gardling	a688d9e264	fmt	2025-04-02 23:19:06 -04:00
Simon Gardling	11164f0859	llm: use finetuned model	2025-04-02 10:11:41 -04:00
Simon Gardling	06feb4e1e2	gemma-3 27b	2025-03-31 21:52:14 -04:00
Simon Gardling	2d47c441fe	llm: use Q4_0 quants (faster)	2025-03-31 18:33:24 -04:00
Simon Gardling	c31635bdd7	format	2025-03-31 17:04:41 -04:00
Simon Gardling	1482429a00	llm: enable AVX2	2025-03-31 12:02:38 -04:00
Simon Gardling	6cc3d96362	llama-cpp: compiler optimizations	2025-03-31 11:17:56 -04:00
Simon Gardling	d5ac5c8cd8	gemma-3 12b	2025-03-31 10:31:29 -04:00
Simon Gardling	d774568e01	auth for llm	2025-03-31 10:29:36 -04:00
Simon Gardling	d34793c18f	add llama-server	2025-03-31 03:59:54 -04:00