1532 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
louistiti	844af50008	feat(server): LLM provider architecture (WIP)	2024-05-07 02:51:18 +08:00
louistiti	b26d924890	feat(server): improve Paraphrase duty accuracy	2024-05-06 11:22:28 +08:00
louistiti	da1b469f8d	chore(server): remove unused countTokens method	2024-05-06 01:00:31 +08:00
louistiti	0189c74a0e	feat(server): finalize Leon's personality and optimize LLM duties	2024-05-06 00:57:20 +08:00
louistiti	a0a4f9d7b0	feat(server): emit socket message when a new update is available	2024-05-05 11:08:16 +08:00
louistiti	eeb1b07898	feat(server): chit-chat duty and skill and more	2024-05-05 00:20:59 +08:00
louistiti	78e5f79bce	feat(server): make Leon's personality more lively	2024-05-04 00:24:21 +08:00
louistiti	764e815204	feat(server): switch LLM duties back to using ChatWrapper instead of completions	2024-05-03 23:23:46 +08:00
louistiti	03d25f2a48	feat(server): Leon's personality done	2024-05-03 23:00:32 +08:00
louistiti	aa36c8cb2a	feat(server): personality support kick off; answer queue	2024-05-02 16:40:05 +08:00
louistiti	2146637022	feat(server): add paraphrase LLM duty to kick off personality attribution	2024-05-01 01:04:29 +08:00
louistiti	0a4e0ed90d	feat(server): add confidence value in logs	2024-04-30 11:48:09 +08:00
louistiti	ceeb5bd70e	feat(server): use Lexi-Llama-3-8B-Uncensored-Q5_K_S as default LLM	2024-04-30 09:58:30 +08:00
louistiti	f77343918d	chore: upgrade socket.io-client to latest	2024-04-29 21:28:20 +08:00
louistiti	e2984ea87e	chore: upgrade socket.io to latest	2024-04-29 21:27:16 +08:00
louistiti	694d398d7b	chore: upgrade fastify to latest	2024-04-29 21:25:25 +08:00
louistiti	a2bca433f6	chore: upgrade dotenv to latest	2024-04-29 21:24:21 +08:00
louistiti	60698704ca	chore: uninstall async npm package	2024-04-29 21:15:47 +08:00
louistiti	7d4e64d575	chore: upgrade tsc-watch to latest	2024-04-29 21:08:49 +08:00
louistiti	2c03efc167	refactor(server): warning message on explicit deactivation of LLM	2024-04-29 19:33:30 +08:00
louistiti	539281110a	feat(server): allow explicit deactivation of LLM	2024-04-29 19:31:57 +08:00
louistiti	3ed88544f8	fix(server): specify correct minimem total/free RAM for LLM	2024-04-29 19:18:27 +08:00
louistiti	9ee73ea5f9	feat(server): map resolved slots to dialog answers	2024-04-29 18:37:20 +08:00
louistiti	c934d0e30d	feat(server): support dialog type action after slots filled	2024-04-29 18:02:41 +08:00
louistiti	d9f0144e62	feat(server): action loop support	2024-04-29 11:25:18 +08:00
louistiti	57923f83ee	fix(scripts): always update manifest on LLM setup	2024-04-28 10:10:57 +08:00
louistiti	a38eee71f0	feat(server): set Phi-3 as default LLM	2024-04-26 15:14:28 +08:00
louistiti	c8e03ec401	fix(server): utterance as expected_item loop indefinitely	2024-04-22 23:29:30 +08:00
louistiti	4ca32b5070	feat(scripts): fallback to mirror in case of error to download LLM	2024-04-22 00:00:46 +08:00
louistiti	d8660251af	feat: use mistral-7b-instruct-v0.2.Q4_K_S as final choice	2024-04-20 14:33:37 +08:00
louistiti	26af271edc	feat(server): add utterance as expected_item type	2024-04-20 14:29:00 +08:00
louistiti	da49a4dd82	feat(scripts): llama.cpp compatible build	2024-04-18 20:38:02 +08:00
louistiti	cf02f0f91d	feat(server): final LLM setup	2024-04-17 00:10:41 +08:00
louistiti	2de95c4ef9	feat(server): Gemma support (prototype)	2024-04-16 20:18:18 +08:00
louistiti	2434d36564	feat(skill/translator-poc): short translate action kick off	2024-04-15 20:42:11 +08:00
louistiti	73d919868b	feat(server): LLM entities support	2024-04-14 17:41:39 +08:00
Théo LUDWIG	39cbd7114e	fix(bridge/python): usage of dict instead of empty TypedDict to avoid types errors ... We might in the future generate automatically from the TypeScript types, generate the Python types. For the moment, it is not typed, so we use a dict.	2024-02-25 20:57:02 +01:00
louistiti	a16ab4abfa	feat(server): add response data to LLM inference endpoint	2024-02-19 18:31:56 +08:00
louistiti	6e12d2638d	feat(server): verify whether LLM is enabled on inference	2024-02-19 18:30:10 +08:00
louistiti	0d3c42beb2	feat(server): implement PoC skill to validate LLM execution	2024-02-19 18:01:41 +08:00
louistiti	66f65b25b4	feat(server): expose LLM duties behind HTTP endpoint	2024-02-19 17:00:14 +08:00
louistiti	3f85a93a03	feat(server): LLM duties architecture + custom NER, summarization and translation duties	2024-02-18 23:43:25 +08:00
Louis Grenard	e9e9155366	feat(server): bootstrap LLM TCP server <> LLM duties <> TCP client (tmp)	2024-02-15 00:13:16 +08:00
Louis Grenard	30c9d3bce5	feat(server): create updater	2024-02-14 17:37:21 +08:00
Louis Grenard	dba5d90b94	chore: add LLM TCP server to Git commit message	2024-02-14 09:45:26 +08:00
Louis Grenard	0705659f3e	fix(scripts): Python TCP server setup	2024-02-14 09:44:50 +08:00
louistiti	4815651a1d	feat(server): bootstrap LLM TCP server (tmp)	2024-02-14 09:27:22 +08:00
louistiti	71ebdc5d80	refactor(server): prepare LLM TCP client connection	2024-02-13 18:51:45 +08:00
louistiti	49ac3218f7	feat(server): preparing LLM TCP server loading	2024-02-13 16:22:32 +08:00
louistiti	1c0c8080bf	feat(scripts): download and compile llama.cpp	2024-02-08 21:05:06 +08:00