phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-26 19:24:27 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
21024 commits 2490 branches 894 tags 548 MiB
Commit graph

513 commits

Author SHA1 Message Date
Ishaan Jaff
39488780e0 litellm_add_proxy_base_url in slack alerts 2024-04-17 17:42:28 -07:00
Krrish Dholakia
d75cfc5e32 fix(utils.py): return vertex api base for request hanging alerts 2024-04-16 17:53:28 -07:00
Krrish Dholakia
aa5da4346a fix(proxy_server.py): support tracking org spend 
currently works when org set for jwt auth
 2024-04-11 23:01:21 -07:00
Krrish Dholakia
07798af50d fix(proxy/utils.py): fix error message 2024-04-08 20:47:13 -07:00
Krrish Dholakia
da216c6915 fix(proxy_server.py): allow mapping a user to an org 2024-04-08 20:45:11 -07:00
Krrish Dholakia
0dad78b53c feat(proxy/utils.py): return api base for request hanging alerts 2024-04-06 15:58:53 -07:00
Krrish Dholakia
ece37a4b7f feat(ui): add models via ui 
adds ability to add models via ui to the proxy. also fixes additional bugs around new /model/new endpoint
 2024-04-04 18:56:20 -07:00
Krrish Dholakia
129bb52e9d fix(proxy_server.py): persist models added via /model/new to db 
allows models to be used across instances

https://github.com/BerriAI/litellm/issues/2319 , https://github.com/BerriAI/litellm/issues/2329
 2024-04-03 20:16:41 -07:00
Krrish Dholakia
029ee15951 perf(proxy_server.py): batch write spend logs 
reduces prisma client errors, by batch writing spend logs - max 1k logs at a time
 2024-04-02 18:46:55 -07:00
Krrish Dholakia
e06d43dc90 fix(tpm_rpm_limiter.py): fix cache init logic 2024-04-01 18:01:38 -07:00
Krrish Dholakia
8d35e659ad fix(proxy/utils.py): support redis caching for alerting 2024-04-01 16:13:59 -07:00
Krrish Dholakia
60b9e25e3c fix(proxy/utils.py): uncomment max parallel request limit check 2024-03-30 20:51:59 -07:00
Krrish Dholakia
7738107d49 fix(utils.py): set redis_usage_cache to none by default 2024-03-30 20:10:56 -07:00
Krrish Dholakia
555f0af027 fix(tpm_rpm_limiter.py): enable redis caching for tpm/rpm checks on keys/user/teams 
allows tpm/rpm checks to work across instances

https://github.com/BerriAI/litellm/issues/2730
 2024-03-30 20:01:36 -07:00
Krrish Dholakia
49e2624240 fix(proxy_server.py): enforce end user budgets with 'litellm.max_end_user_budget' param 2024-03-29 17:14:40 -07:00
Krrish Dholakia
6848e3b1d2 fix(proxy_server.py): enable spend tracking for team-based jwt auth 2024-03-28 20:16:22 -07:00
Krrish Dholakia
473bab8a19 refactor(proxy/utils.py): add more debug logs 2024-03-28 18:44:35 -07:00
Krish Dholakia
b828290c81 Merge pull request #2722 from BerriAI/litellm_db_perf_improvement 
feat(proxy/utils.py): enable updating db in a separate server
 2024-03-28 14:56:14 -07:00
Krrish Dholakia
eca5d04126 test(test_update_spend.py): allow db_client to be none 2024-03-28 13:44:40 -07:00
Krrish Dholakia
e87c5f5d6f fix(proxy_server.py): allow user to pass in spend logs collector url 2024-03-28 09:14:30 -07:00
Ishaan Jaff
c96e1af901 Merge pull request #2728 from BerriAI/litellm_reduce_deep_copies 
[FEAT] Proxy - reduce deep copies
 2024-03-27 21:26:09 -07:00
Ishaan Jaff
7c32955f64 (fix) remove deep copy from all responses 2024-03-27 20:36:53 -07:00
Krrish Dholakia
7fe02405e0 fix(proxy/utils.py): check cache before alerting user 2024-03-27 20:09:15 -07:00
Krrish Dholakia
0417ce6cbe feat(auth_checks.py): enable admin to enforce 'user' param for all openai endpoints 2024-03-27 17:36:27 -07:00
Krrish Dholakia
46937935d1 feat(proxy/utils.py): enable updating db in a separate server 2024-03-27 16:02:36 -07:00
Krrish Dholakia
7bc76ddbc3 feat(llm_guard.py): enable key-specific llm guard check 2024-03-26 17:21:51 -07:00
Ishaan Jaff
f0992c2dbd (fix) stop using f strings with logger 2024-03-25 10:47:18 -07:00
Ishaan Jaff
2c01457a4b (feat) stop eagerly evaluating fstring 2024-03-25 09:01:42 -07:00
Krrish Dholakia
e9cc6b4cc9 feat(proxy_server.py): enable llm api based prompt injection checks 
run user calls through an llm api to check for prompt injection attacks. This happens in parallel to th
e actual llm call using `async_moderation_hook`
 2024-03-20 22:43:42 -07:00
Krrish Dholakia
499087301c Revert "Merge pull request #2593 from BerriAI/litellm_reset_budget_fix" 
This reverts commit afd363129f, reversing
changes made to c94bc94ad5.
 2024-03-19 20:25:41 -07:00
Krrish Dholakia
f15d105ed2 bump: version 1.32.5 → 1.32.6 2024-03-19 20:17:50 -07:00
Krrish Dholakia
809124a698 fix: better debug logs 2024-03-19 19:28:26 -07:00
Krrish Dholakia
2f022920f1 fix(proxy/utils.py): fix reset budget logic 
uses fewer clients - prevents read timeouts
 2024-03-19 18:02:09 -07:00
Krrish Dholakia
4879265d7f build: build fixes 2024-03-19 16:59:59 -07:00
Krrish Dholakia
b2f8b1aa8a fix(proxy/utils.py): cleanup logging 2024-03-19 09:29:29 -07:00
Krrish Dholakia
1b10123f07 fix(proxy/utils.py): batch writing updates to db 2024-03-18 16:47:02 -07:00
Krrish Dholakia
1618751824 fix(proxy/utils.py): move to batch writing db updates 2024-03-16 22:32:00 -07:00
Krrish Dholakia
3232feb123 fix(proxy_server.py): fix key caching logic 2024-03-13 19:10:24 -07:00
Krrish Dholakia
60219254a9 fix(proxy_server.py): fix /user/info for non-existent user id 2024-03-12 11:05:41 -07:00
Krrish Dholakia
366aaf744e fix(proxy_server.py): bug fix 2024-03-11 13:54:58 -07:00
Krrish Dholakia
1dcaba83d2 fix(proxy_server.py): prevent user from deleting non-user owned keys when they use ui 2024-03-11 12:13:30 -07:00
Krrish Dholakia
b2ce963498 feat: add cost tracking + caching for transcription calls 2024-03-09 15:43:38 -08:00
Krrish Dholakia
8781f9caa4 fix(proxy/utils.py): add more logging for prisma client get_data error 2024-03-09 13:06:30 -08:00
ishaan-jaff
de8bc01c18 (fix) improve improve prisma alerting/debug 2024-03-07 15:44:03 -08:00
ishaan-jaff
b3e6280125 (fix) better prisma debug logs 2024-03-07 14:04:02 -08:00
ishaan-jaff
1a3a097dd4 (feat) better debugging when reading prisma 2024-03-07 13:42:54 -08:00
Krrish Dholakia
df0eb170e6 fix(proxy_server.py): fix sql query 2024-03-06 19:41:12 -08:00
Krrish Dholakia
7bfadc258e feat(proxy_server.py): team based model aliases 
allow setting model aliases at a team level (e.g. route all 'gpt-3.5-turbo' requests from team-1 to model-deployment-group-2)
 2024-03-06 17:42:08 -08:00
ishaan-jaff
48f6189760 (feat) circular ref error on prisa 2024-03-06 12:08:22 -08:00
ishaan-jaff
8a75c4c3a3 (fix) circular ref error h 2024-03-06 12:02:44 -08:00