441 commits

Author	SHA1	Message	Date
ishaan-jaff	9f72ce9fc6	(fix) improve batch_completion_models + multiple deployments, if 1 model fails, return result from 2nd	2023-10-30 17:20:07 -07:00
ishaan-jaff	43b450319e	(docs) add docstring for batch_completion	2023-10-30 14:31:34 -07:00
ishaan-jaff	f52b36a338	(docs) docstring for completion_with_retries	2023-10-30 14:20:09 -07:00
ishaan-jaff	43b8387334	(docs) add doc string for aembedding	2023-10-30 13:59:56 -07:00
ishaan-jaff	7037913f9d	(fix) update acompletion docstring	2023-10-30 13:53:32 -07:00
ishaan-jaff	cac320f74e	(fix) remove bloat - ratelimitmanager	2023-10-27 18:11:39 -07:00
Krrish Dholakia	afe14c8a96	fix(utils.py/completion_with_fallbacks): accept azure deployment name in rotations	2023-10-27 16:00:42 -07:00
Krrish Dholakia	c1b2553827	fix(utils.py): adding support for anyscale models	2023-10-25 09:08:10 -07:00
Krrish Dholakia	f12dc5df21	fix(vertex_ai.py): fix output parsing	2023-10-24 12:08:22 -07:00
ishaan-jaff	6373f6bddd	(feat) add async embeddings	2023-10-23 13:59:37 -07:00
Krrish Dholakia	cd0e699bcf	fix(main.py): multiple deployments fix - run in parallel	2023-10-21 14:28:50 -07:00
ishaan-jaff	0b0564167c	(fix) embedding() using get_llm_provider	2023-10-20 15:00:08 -07:00
ishaan-jaff	114d8fda65	(feat) native perplexity support	2023-10-20 14:29:07 -07:00
Krrish Dholakia	1f1cf7a11c	feat(main.py): support multiple deployments in 1 completion call	2023-10-20 13:01:53 -07:00
Krrish Dholakia	4b48af7c3c	fix(anthropic.py-+-bedrock.py): anthropic prompt format	2023-10-20 10:56:15 -07:00
Krrish Dholakia	00993f3575	fix: allow api base to be set for all providers ... enables proxy use cases	2023-10-19 19:07:42 -07:00
Krrish Dholakia	76bf8c4be3	fix(anthropic.py): enable api base to be customized	2023-10-19 18:45:29 -07:00
Krrish Dholakia	dcb866b353	docs(proxy_server.md): update proxy server docs to include multi-agent autogen tutorial	2023-10-17 09:22:34 -07:00
Krrish Dholakia	036c50e2bf	refactor(main.py): clean up print statement	2023-10-16 17:53:09 -07:00
Krrish Dholakia	4424aaf69f	refactor(main.py): remove print statement	2023-10-16 07:33:15 -07:00
Krrish Dholakia	1a09bbd4a9	docs(main.py): adding docstring for text_completion	2023-10-16 07:31:48 -07:00
Zeeland	9f6138ef0e	fix: llm_provider add openai finetune compatibility	2023-10-16 18:44:45 +08:00
ishaan-jaff	6413285551	(fix) deepinfra/llama should go to deepinfra not to openrouter	2023-10-14 16:47:25 -07:00
Krrish Dholakia	7358d2e4ea	bump: version 0.8.4 → 0.8.5	2023-10-14 16:43:06 -07:00
ishaan-jaff	882ac46727	(feat) add doc string for embedding	2023-10-14 16:08:13 -07:00
Krrish Dholakia	9513d6b862	fix(utils.py): read env variables for known openai-compatible api's (e.g. perplexity), dynamically from th eenvironment	2023-10-13 22:43:32 -07:00
Krrish Dholakia	91c8e92e71	fix(openai.p): adding support for exception mapping for openai-compatible apis via http calls	2023-10-13 21:56:51 -07:00
ishaan-jaff	b72dbe61c0	(feat) set api_base, api_key, api_version for embedding()	2023-10-13 21:09:44 -07:00
Krrish Dholakia	4d4f8bfa5d	feat(proxy_server): adding model fallbacks and default model to toml	2023-10-13 15:31:17 -07:00
ishaan-jaff	fabad3dc42	(fix) Ollama use new streaming format	2023-10-11 17:00:39 -07:00
ishaan-jaff	689acb8a08	(feat) add CustomStreamWrapper for Ollama - match OpenAI streaming	2023-10-11 17:00:39 -07:00
Krrish Dholakia	d280a8c434	fix(proxy_cli-and-utils.py): fixing how config file is read + infering llm_provider for known openai endpoints	2023-10-10 20:53:02 -07:00
Krrish Dholakia	b50013386f	fix(openai.py): enable custom proxy to pass in ca_bundle_path	2023-10-10 13:23:27 -07:00
Krrish Dholakia	af2fd0e0de	fix: fix value error if model returns empty completion	2023-10-10 10:11:40 -07:00
Krrish Dholakia	db20cb84d4	fix(main.py): return n>1 response for openai text completion	2023-10-09 20:44:07 -07:00
Krrish Dholakia	689371949c	fix(main.py): read openai org from env	2023-10-09 16:49:22 -07:00
ishaan-jaff	4e64f123ef	(fix) api_base, api_version and api_key	2023-10-09 14:11:05 -07:00
ishaan-jaff	bf4ce08640	(fix) acompletion for ollama non streaing	2023-10-09 13:47:08 -07:00
Krrish Dholakia	704be9dcd1	feat(factory.py): option to add function details to prompt, if model doesn't support functions param	2023-10-09 09:53:53 -07:00
ishaan-jaff	f6f7c0b891	(feat) add api_key, api_base, api_version to completion	2023-10-09 08:13:12 -07:00
Krrish Dholakia	bcf0b0ac7b	style(main.py): clean up print statement	2023-10-07 15:43:40 -07:00
Krrish Dholakia	9cda24e1b2	fix(utils): adds complete streaming response to success handler	2023-10-07 15:42:00 -07:00
Krrish Dholakia	d69038883c	docs(completion-docs): adds more details on provider-specific params	2023-10-07 13:49:30 -07:00
Krrish Dholakia	306a38880d	feat(ollama.py): exposing ollama config	2023-10-06 15:52:58 -07:00
ishaan-jaff	47521c5a97	test commitizen bump	2023-10-06 15:41:38 -07:00
ishaan-jaff	4ae8a71aa3	bump: version 0.2.5 → 0.3.0	2023-10-06 15:40:01 -07:00
Krrish Dholakia	7e34736a38	fix(add-custom-success-callback-for-streaming): add custom success callback for streaming	2023-10-06 15:02:02 -07:00
Krrish Dholakia	a977e94a5d	style(main.py): adding spacing	2023-10-06 06:16:17 -07:00
Krrish Dholakia	e0f1cffa87	fix: azure flag check	2023-10-05 22:44:41 -07:00
Krrish Dholakia	060a2e40b2	fix: fixing mypy linting errors and being backwards compatible for azure=true flag	2023-10-05 22:36:32 -07:00