v1.16.6

Full Changelog: v1.16.18...v1.16.6

v1.16.16

What's Changed

• Litellm speed improvements by @krrishdholakia in #1344

Full Changelog: v1.16.15...v1.16.16

v1.16.15

litellm 1.16.5

What's Changed

• use s3 Buckets for caching /chat/completion, embedding responses. Proxy Caching: https://docs.litellm.ai/docs/proxy/caching, Caching with litellm.completion https://docs.litellm.ai/docs/caching/redis_cache
• litellm.completion_cost() Support for cost calculation for embedding responses - Azure embedding, and text-embedding-ada-002-v2 @jeromeroussin

async def _test():
      response = await litellm.aembedding(
          model="azure/azure-embedding-model",
          input=["good morning from litellm", "gm"],
      )
  
      print(response)
  
      return response
  
  response = asyncio.run(_test())
  
  cost = litellm.completion_cost(completion_response=response)

• litellm.completion_cost() raises exceptions (instead of swallowing exceptions) @jeromeroussin
• Improved token counting for azure streaming responses @langgg0511 #1304
• set os.environ/ variables for litellm proxy cache @Manouchehri

model_list:
  - model_name: gpt-3.5-turbo
    litellm_params:
      model: gpt-3.5-turbo
  - model_name: text-embedding-ada-002
    litellm_params:
      model: text-embedding-ada-002

litellm_settings:
  set_verbose: True
  cache: True          # set cache responses to True
  cache_params:        # set cache params for s3
    type: s3
    s3_bucket_name: cache-bucket-litellm   # AWS Bucket Name for S3
    s3_region_name: us-west-2              # AWS Region Name for S3
    s3_aws_access_key_id: os.environ/AWS_ACCESS_KEY_ID  # us os.environ/<variable name> to pass environment variables. This is AWS Access Key ID for S3
    s3_aws_secret_access_key: os.environ/AWS_SECRET_ACCESS_KEY  # AWS Secret Access Key for S3

• build(Dockerfile): moves prisma logic to dockerfile by @krrishdholakia in #1342

Full Changelog: 1.16.14...v1.16.15

1.16.14

Full Changelog: v1.16.13...1.16.14

1.16.12

New providers

Xinference Embeddings: https://docs.litellm.ai/docs/providers/xinference
Voyage AI: https://docs.litellm.ai/docs/providers/voyage
Cloudflare AI workers: https://docs.litellm.ai/docs/providers/cloudflare_workers

Fixes:

AWS region name error when passing user bedrock client: #1292
Azure OpenAI models - use correct context window in model_context_window_and_prices.json
Fixes for Azure OpenAI + Streaming - counting prompt tokens correctly: #1264

What's Changed

• Update mistral.md by @nanowell in #1157
• Use current Git folder for building Dockerfile by @Manouchehri in #1076
• (fix) curl example on proxy/quick_start by @ku-suke in #1163
• clarify the need to set an exporter by @nirga in #1162
• Update mistral.md by @ericmjl in #1174
• doc: updated langfuse ver 1.14 in pip install cmd by @speedyankur in #1178
• Fix for issue that occured when proxying to ollama by @clevcode in #1168
• Sample code to prevent logging API key in callback to Slack by @navidre in #1185
• Add partial support of VertexAI safety settings by @neubig in #1190
• docker build and push on release by @Rested in #1197
• Add a default for safety settings in vertex AI by @neubig in #1199
• Make vertex ai work with generate_content by @neubig in #1213
• fix: success_callback logic for cost_tracking by @sihyeonn in #1211
• Add aws_bedrock_runtime_endpoint support by @Manouchehri in #1203
• fix least_busy router by updating min_traffic by @AllentDan in #1195
• Improve langfuse integration by @maxdeichmann in #1183
• usage_based_routing_fix by @sumanth13131 in #1182
• add some GitHub workflows for flake8 and add black dependecy by @bufferoverflow in #1223
• fix: helicone logging by @evantancy in #1249
• update anyscale price link by @prateeksachan in #1246
• Bump aiohttp from 3.8.4 to 3.9.0 by @dependabot in #1180
• updated oobabooga to new api and support for embeddings by @danikhan632 in #1248
• add support for mistral json mode via anyscale by @marmikcfc in #1275
• Adds support for Vertex AI Unicorn by @AshGreh in #1277
• fix typos & add missing names for azure models by @fcakyon in #1290
• fix(proxy_server.py) Check when '_hidden_params' is None by @asedmammad in #1300

New Contributors

• @nanowell made their first contribution in #1157
• @ku-suke made their first contribution in #1163
• @ericmjl made their first contribution in #1174
• @speedyankur made their first contribution in #1178
• @clevcode made their first contribution in #1168
• @navidre made their first contribution in #1185
• @neubig made their first contribution in #1190
• @Rested made their first contribution in #1197
• @sihyeonn made their first contribution in #1211
• @AllentDan made their first contribution in #1195
• @sumanth13131 made their first contribution in #1182
• @bufferoverflow made their first contribution in #1223
• @evantancy made their first contribution in #1249
• @prateeksachan made their first contribution in #1246
• @danikhan632 made their first contribution in #1248
• @marmikcfc made their first contribution in #1275
• @AshGreh made their first contribution in #1277
• @fcakyon made their first contribution in #1290
• @asedmammad made their first contribution in #1300

Full Changelog: v1.15.0...1.16.12

v1.15.0

What's Changed

LiteLLM Proxy now maps exceptions for 100+ LLMs to the OpenAI format https://docs.litellm.ai/docs/proxy/quick_start
🧨 Log all LLM Input/Output to @dynamodb set litellm.success_callback = ["dynamodb"] https://docs.litellm.ai/docs/proxy/logging#logging-proxy-inputoutput---dynamodb
⭐️ Support for @MistralAI API, Gemini PRO
🔎 Set Aliases for model groups on LiteLLM Proxy
🔎 Exception mapping for openai.NotFoundError live now + testing for exception mapping on proxy added to LiteLLM ci/cd https://docs.litellm.ai/docs/exception_mapping
⚙️ Fixes for async + streaming caching https://docs.litellm.ai/docs/proxy/caching
👉 Support for using Async logging with @langfuse live on proxy

AI Generated Release Notes

• Enable setting default model value for LiteLLM, Chat, Completions by @estill01 in #985
• fix replicate system prompt: forgot to add **optional_params to input data by @nbaldwin98 in #1080
• Update factory.py to fix issue when calling from write-the -> langchain -> litellm served ollama by @James4Ever0 in #1054
• Update Dockerfile to preinstall Prisma CLI by @Manouchehri in #1039
• build(deps): bump aiohttp from 3.8.6 to 3.9.0 by @dependabot in #937
• multistage docker build by @wallies in #995
• fix: traceloop links by @nirga in #1123
• refactor: add CustomStreamWrapper return type for completion by @Undertone0809 in #1112
• fix langfuse tests by @maxdeichmann in #1097
• Fix #1119, no content when streaming. by @emsi in #1122
• docs(projects): add Docq to 'projects built on..' section by @janaka in #1142
• docs(projects): add Docq.AI to sidebar nav by @janaka in #1143

New Contributors

• @James4Ever0 made their first contribution in #1054
• @wallies made their first contribution in #995
• @maxdeichmann made their first contribution in #1097
• @emsi made their first contribution in #1122
• @janaka made their first contribution in #1142

Full Changelog: v1.11.1...v1.15.0

v1.11.1

Proxy

• Bug fix for non OpenAI LLMs on proxy
• Major stability improvements & Fixes + added test cases for proxy
• Async success/failure loggers
• Support for using custom loggers with aembedding()

What's Changed

• feat: add docker compose file and running guide by @geekyayush in #993
• (feat) Speedup health endpoint by @PSU3D0 in #1023
• (pricing) Add Claude v2.1 for Bedrock by @Manouchehri in #1042

New Contributors

• @geekyayush made their first contribution in #993

Full Changelog: v1.10.4...v1.11.1

v1.10.4

Note: Proxy Server on 1.10.4 has a bug for non OpenAI LLMs - Fixed on 1.10.11

Updates Proxy Server

• Use custom callbacks on the proxy https://docs.litellm.ai/docs/proxy/logging
• Set timeout and stream_timeout per model https://docs.litellm.ai/docs/proxy/load_balancing#custom-timeouts-stream-timeouts---per-model
• Stability: Added testing for reading config.yaml on the proxy
• NEW /model/new + /model/info endpoints - Add new models + Get model info without restarting proxy.
• Custom user auth - #898 (comment)
• Key Security -> keys now stored as just hashes in the db
• user id accepted + passed to OpenAI/Azure

litellm Package

• Specify kwargs for Redis Cache 9ba1765
• Fixes for Sagemaker + Palm Streaming
• Support for async custom callbacks - https://docs.litellm.ai/docs/observability/custom_callback#async-callback-functions
• Major improvements to stream chunk builder - support for parallel tool calling, system fingerprints, etc.
• Fixes for azure / openai streaming (return complete response object)
• Support for loading keys from azure key vault - https://docs.litellm.ai/docs/secret#azure-key-vault

What's Changed

• docs: adds gpt-3.5-turbo-1106 in supported models by @rishabgit in #958
• (feat) Allow installing proxy dependencies explicitly with pip install litellm[proxy] by @PSU3D0 in #966
• Mention Neon as a database option in docs by @Manouchehri in #977
• fix system prompts for replicate by @nbaldwin98 in #970

New Contributors

• @rishabgit made their first contribution in #958
• @PSU3D0 made their first contribution in #966
• @nbaldwin98 made their first contribution in #970

Full Changelog: v1.7.11...v1.10.4

v1.7.11

💥 LiteLLM Router + Proxy handles 500+ requests/second

💥LiteLLM Proxy - Now handles 500+ requests/second, Load Balance Azure + OpenAI deployments, Track spend per user 💥
Try it here: https://docs.litellm.ai/docs/simple_proxy
🔑 Support for AZURE_OPENAI_API_KEY on Azure https://docs.litellm.ai/docs/providers/azure
h/t
@solyarisoftware
⚡️ LiteLLM Router can now handle 20% more throughput https://docs.litellm.ai/docs/routing
📖Improvement to litellm debugging docs h/t
@solyarisoftware
https://docs.litellm.ai/docs/debugging/local_debugging

Full Changelog: v1.7.1...v1.7.11

v1.7.1

What's Changed

• 🚨 LiteLLM Proxy uses Async completion/embedding calls on this release onwards - this led to 30x more throughput for embedding/completion calls

New Contributors

• @guspan-tanadi made their first contribution in #851
• @Manouchehri made their first contribution in #880
• @maqsoodshaik made their first contribution in #884
• @okotek made their first contribution in #885
• @kumaranvpl made their first contribution in #902

Full Changelog: v1.1.0...v1.7.1