Commit graph

477 commits

Author SHA1 Message Date
Krrish Dholakia
cc89aa7456 fix(bedrock.py): add support for sts based boto3 initialization
https://github.com/BerriAI/litellm/issues/1476
2024-01-17 12:08:59 -08:00
ishaan-jaff
118c53af26 (fix) using base_url Azure 2024-01-17 10:12:55 -08:00
ishaan-jaff
82954728a7 (feat) support base_url with /openai passed for Azure 2024-01-17 10:03:25 -08:00
ishaan-jaff
a49fa3abaf (v0) fixes for Azure GPT Vision enhancements 2024-01-17 09:57:16 -08:00
Krrish Dholakia
f7e8bdbb32 fix(vertex_ai.py): raise exception if vertex ai missing required dependendencies 2024-01-16 16:23:29 -08:00
ishaan-jaff
66c8eb582c (feat) sagemaker - map status code and message 2024-01-15 21:43:16 -08:00
ishaan-jaff
a5f3907334 (feat) provisioned throughput - bedrock embedding models 2024-01-13 21:07:38 -08:00
ishaan-jaff
a661149a39 (feat) bedrock support provisioned throughput 2024-01-13 15:39:54 -08:00
Krish Dholakia
84879c0318 Merge pull request #1429 from dleen/data
[bug] unbound variable in bedrock
2024-01-12 22:16:11 +05:30
David Leen
1660e4ab72 improve bedrock exception granularity 2024-01-12 16:38:55 +01:00
David Leen
d0e9c9dce9 [bug] unbound variable in bedrock
note: the code was written as `json.dumps({})` even though it is more verbose in order to facilitate easier refactoring in the future

fixes #1428
2024-01-12 12:33:00 +01:00
Krrish Dholakia
0e1ea4325c fix(azure.py): support health checks to text completion endpoints 2024-01-12 00:13:01 +05:30
ishaan-jaff
1f04446222 (fix) bedrock - embedding - support str input 2024-01-11 23:02:12 +05:30
ishaan-jaff
9aac1de191 v0 2024-01-11 22:56:18 +05:30
Krrish Dholakia
f32ec52673 build(pyproject.toml): drop certifi dependency (unused) 2024-01-10 08:09:03 +05:30
Krrish Dholakia
556e7d4e1a fix(openai.py): fix exception raising logic 2024-01-09 11:58:30 +05:30
Krrish Dholakia
d105751643 fix(azure.py,-openai.py): raise the correct exceptions for image generation calls 2024-01-09 11:55:38 +05:30
ishaan-jaff
3081dc525a (feat) litellm.completion - support ollama timeout 2024-01-09 10:34:41 +05:30
Krrish Dholakia
d89a58ec54 fix(ollama.py): use tiktoken as backup for prompt token counting 2024-01-09 09:47:18 +05:30
Krrish Dholakia
045ece4582 refactor(gemini.py): fix linting issue 2024-01-08 11:43:33 +05:30
Krrish Dholakia
e4a5a3395c fix(huggingface_restapi.py): support timeouts for huggingface + openai text completions
https://github.com/BerriAI/litellm/issues/1334
2024-01-08 11:40:56 +05:30
Krish Dholakia
a394eb12db Merge pull request #1315 from spdustin/feature_allow_claude_prefill
Adds "pre-fill" support for Claude
2024-01-08 10:48:15 +05:30
Krrish Dholakia
f300d17176 fix(gemini.py): better error handling 2024-01-08 07:32:26 +05:30
Krrish Dholakia
67ff7797c6 fix(factory.py): more logging around the image loading for gemini 2024-01-06 22:50:44 +05:30
Krish Dholakia
67ecab4b38 Merge pull request #1344 from BerriAI/litellm_speed_improvements
Litellm speed improvements
2024-01-06 22:38:10 +05:30
Krrish Dholakia
2d1871a1ae fix(factory.py): support gemini-pro-vision on google ai studio
https://github.com/BerriAI/litellm/issues/1329
2024-01-06 22:36:22 +05:30
Krrish Dholakia
35fd28073e fix(sagemaker.py): fix the post-call logging logic 2024-01-06 21:52:58 +05:30
Krrish Dholakia
4c7d530c2a fix(openai.py): fix image generation model dump 2024-01-06 17:55:32 +05:30
Krrish Dholakia
807b64e68e perf(azure+openai-files): use model_dump instead of json.loads + model_dump_json 2024-01-06 15:50:05 +05:30
spdustin@gmail.com
6520d153e7 Update factory (and tests) for Claude 2.1 via Bedrock 2024-01-05 23:32:32 +00:00
Dustin Miller
7172f83ef4 Merge branch 'BerriAI:main' into feature_allow_claude_prefill 2024-01-05 15:15:29 -06:00
ishaan-jaff
a36b1a4890 (fix) undo - model_dump_json() before logging 2024-01-05 11:47:16 +05:30
ishaan-jaff
6e1ea2c44c (fix) proxy - log response before model_dump_json 2024-01-05 11:00:02 +05:30
ishaan-jaff
70bbc2e446 (fix) azure+cf gateway, health check 2024-01-04 12:34:07 +05:30
Krrish Dholakia
62ea95c25b fix(proxy/rules.md): add docs on setting post-call rules on the proxy 2024-01-04 11:16:50 +05:30
Dustin Miller
5f54fc2383 Adds "pre-fill" support for Claude 2024-01-03 18:45:36 -06:00
ishaan-jaff
6672591198 (fix) init_bedrock_client 2024-01-01 22:48:56 +05:30
Krrish Dholakia
7be5f74b70 fix(aimage_generation): fix response type 2023-12-30 12:53:24 +05:30
Krrish Dholakia
4d239f1e65 fix(openai.py): fix async image gen call 2023-12-30 12:44:54 +05:30
Krrish Dholakia
b69ffb3738 fix: support dynamic timeouts for openai and azure 2023-12-30 12:14:02 +05:30
Krrish Dholakia
7d55a563ee fix(main.py): don't set timeout as an optional api param 2023-12-30 11:47:07 +05:30
ishaan-jaff
224d38ba48 (fix) vertex ai - use usage from response 2023-12-29 16:30:25 +05:30
ishaan-jaff
c69f4f17a5 (feat) cloudflare - add optional params 2023-12-29 11:50:09 +05:30
ishaan-jaff
b990fc8324 (feat) cloudflare ai workers - add completion support 2023-12-29 11:34:58 +05:30
ishaan-jaff
796e735881 (feat) v0 adding cloudflare 2023-12-29 09:32:29 +05:30
ishaan-jaff
362bed6ca3 (fix) together_ai cost tracking 2023-12-28 22:11:08 +05:30
Krrish Dholakia
5a48dac83f fix(vertex_ai.py): support function calling for gemini 2023-12-28 19:07:04 +05:30
Krrish Dholakia
8188475c16 feat(admin_ui.py): support creating keys on admin ui 2023-12-28 16:59:11 +05:30
Krrish Dholakia
507b6bf96e fix(utils.py): use local tiktoken copy 2023-12-28 11:22:33 +05:30
Krrish Dholakia
2285282ef8 feat(health_check.py): more detailed health check calls 2023-12-28 09:12:57 +05:30