569 commits

Author	SHA1	Message	Date
ishaan-jaff	da7c07574f	(test) cost calc on azure	2023-11-23 13:50:09 -08:00
Krrish Dholakia	4f183dc6a0	fix(utils.py): support reading api keys dynamically from the os environment	2023-11-23 13:41:56 -08:00
Krrish Dholakia	3e76d4b422	feat(router.py): add server cooldown logic	2023-11-22 15:59:48 -08:00
ishaan-jaff	e8ff4d5eca	(feat) clean out junk params from litellm embedding	2023-11-22 13:50:45 -08:00
Krrish Dholakia	b0801f61e6	test(test_caching.py): cleaning up tests	2023-11-22 13:43:48 -08:00
Krrish Dholakia	10fe16c965	fix(utils.py): add param mapping for perplexity, anyscale, deepinfra ... n n	2023-11-22 10:04:27 -08:00
Krrish Dholakia	381fdcd37b	fix(utils.py): add response ms for async calls	2023-11-21 19:59:00 -08:00
Krrish Dholakia	2a681c1874	fix(utils.py): fix pre call rules	2023-11-21 07:10:04 -08:00
Krrish Dholakia	b991a35ebe	fix(main.py): revert model alias map change	2023-11-20 21:07:52 -08:00
Krrish Dholakia	ab2986290e	fix(utils.py): fix rules calling	2023-11-20 21:06:36 -08:00
Krrish Dholakia	b2e52eedf8	fix(main.py): fix model alias map logic	2023-11-20 20:49:10 -08:00
Krrish Dholakia	1306addfe8	fix(openai.py-+-azure.py): fix linting issues	2023-11-20 19:29:23 -08:00
Krrish Dholakia	855964ed45	fix(utils.py): adding support for rules + mythomax/alpaca prompt template	2023-11-20 18:58:15 -08:00
ishaan-jaff	4f46ac4ab5	(fix) linting error	2023-11-20 18:32:43 -08:00
ishaan-jaff	3f30f93516	(fix) pydantic errors openai usage	2023-11-20 18:28:19 -08:00
ishaan-jaff	980a3502de	(fix) completion - always map finish_reason	2023-11-20 17:24:16 -08:00
ishaan-jaff	11ec2710c6	(fix) completion: max_retries using OpenAI client	2023-11-20 16:57:37 -08:00
Krrish Dholakia	1976d0f7d6	fix(routing.py): update token usage on streaming	2023-11-20 14:19:25 -08:00
Krrish Dholakia	85932ac247	fix(utils.py): expanding exception mapping coverage for vertex ai	2023-11-18 20:05:40 -08:00
ishaan-jaff	69b6712707	(fix) streaming completion azure	2023-11-18 19:04:41 -08:00
ishaan-jaff	33c3665922	(fix) streaming ensure response obj is initialized	2023-11-18 17:31:58 -08:00
ishaan-jaff	ff8668aad5	(feat) print_verbose Raw openai chunk	2023-11-18 17:12:49 -08:00
ishaan-jaff	1927167bde	(fix) streaming openai + function calling	2023-11-18 17:01:46 -08:00
ishaan-jaff	70fc5afb5d	(fix) streaming + function / tool calling	2023-11-18 16:23:29 -08:00
Krrish Dholakia	478bd7def6	fix(router.py): introducing usage-based-routing	2023-11-17 17:56:26 -08:00
ishaan-jaff	88200432b0	(feat) support parallel function calling	2023-11-17 15:51:27 -08:00
ishaan-jaff	7de87c845b	(feat) improve logging - show model_call_details	2023-11-17 15:51:27 -08:00
Krrish Dholakia	0ab6b2451d	fix(acompletion): support client side timeouts + raise exceptions correctly for async calls	2023-11-17 15:39:47 -08:00
ishaan-jaff	7abb65d53f	(feat) completion: add response_format, seed, tools, tool_choice	2023-11-17 13:59:57 -08:00
ishaan-jaff	bd82559553	(v1.0+ breaking change) get_max_tokens -> return int	2023-11-17 10:38:50 -08:00
Krrish Dholakia	9bd1f4ebd0	fix(utils.py): improve exception mapping for vertex ai	2023-11-16 22:02:26 -08:00
Krrish Dholakia	7456c26940	docs(routing.md): update tutorial on deploying router	2023-11-16 21:46:43 -08:00
Krrish Dholakia	51bf637656	feat: global client for sync + async calls (openai + Azure only)	2023-11-16 14:44:13 -08:00
Krrish Dholakia	a94c09c13c	fix(openai.py): handling extra headers	2023-11-16 12:48:21 -08:00
ishaan-jaff	9e072f87bd	(fix) bedrock meta llama optional params	2023-11-16 12:38:27 -08:00
ishaan-jaff	04971674b4	(fix) only decode chunk when it's not a str	2023-11-16 12:24:31 -08:00
Krrish Dholakia	bf0f8b824c	fix(azure.py): use openai client sdk for handling sync+async calling	2023-11-16 12:08:12 -08:00
ishaan-jaff	da8c2f4a4a	(fix) HF api + streaming	2023-11-16 11:59:56 -08:00
ishaan-jaff	c84fe1af5e	(fix) langfuse logging + openai streaming when chunk = [DONE}	2023-11-16 10:45:35 -08:00
Krrish Dholakia	bb51216846	fix(openai.py): supporting openai client sdk for handling sync + async calls (incl. for openai-compatible apis)	2023-11-16 10:35:03 -08:00
Ishaan Jaff	d6ad62d793	Merge pull request #811 from dchristian3188/bedrock-llama ... Bedrock llama	2023-11-16 07:57:50 -08:00
Ishaan Jaff	3f95fd2da5	Merge pull request #826 from rodneyxr/ollama-fixes ... Fix typo for initial_prompt_value and too many values to unpack error	2023-11-16 07:55:53 -08:00
David Christian	45ae4a76e4	updated utils for bedrock.meta streaming	2023-11-16 07:12:27 -08:00
Krrish Dholakia	9fd7c5b343	test: set request timeout at request level	2023-11-15 17:42:31 -08:00
Rodney Rodriguez	5d39279ac0	bugfixes for ollama	2023-11-15 19:27:06 -06:00
Krrish Dholakia	bcea28e2e4	fix(utils): fixing exception mapping	2023-11-15 15:51:17 -08:00
Krrish Dholakia	f84db3ce14	feat(get_max_tokens): get max tokens for huggingface hub models	2023-11-15 15:25:40 -08:00
Krrish Dholakia	1a705bfbcb	refactor(huggingface_restapi.py): moving async completion + streaming to real async calls	2023-11-15 15:14:21 -08:00
Krrish Dholakia	c8b73d90f1	fix(utils.py): fix langfuse integration	2023-11-15 14:05:40 -08:00
Krrish Dholakia	60d2dff74a	fix(utils.py): check for none params	2023-11-15 13:39:09 -08:00