agentjido / req_llm / 487bb51258a45e343a1f9e8048c5ee331d7ccc8d-PR-206

Builds Branch Commit Type Ran Committer Via Coverage
487bb51258a45e343a1f9e8048c5ee331d7ccc8d-PR-206 idempotent-stream-response Fix `mix quality` issues Pull #206 vkryukov github
52.76
5c2674cd99687ab3f3fd03b17cb48c3983e38f8b-PR-202 add-vllm-support Fix quality issues Pull #202 meanderingstream github
52.82
d203939b2810cbf997f86ec38ec6380abba669a9 main Custom providers defined outside req_llm (#201) * Apps using req_llm as a dependency can specify additional provider implementation modules in config * documentation update * remove ai preamble * rename to :custom_providers push web-flow github
52.8
312582ec825240237fa4cfb87e379a506e96cfe6-PR-201 feature/custom_providers rename to :custom_providers Pull #201 bcxbb github
52.76
312d2a4f671f96c861194526cef36e39f51c7316-PR-194 vertex-gemini-rebased feat(vertex): Add Gemini support to Google Vertex AI provider - Add GoogleVertex.Gemini formatter for Gemini 2.5 models - Extract shared Google.Context module (like Anthropic.Context pattern) - Make Google helper functions public for shared use -... Pull #194 neilberkman github
51.57
f6a7d51abc748d8a74f26fbe9ca9a129db30ccc2-PR-194 vertex-gemini-rebased feat(vertex): Add Gemini support to Google Vertex AI provider - Add GoogleVertex.Gemini formatter for Gemini 2.5 models - Register gemini family in @model_families - Support for all Gemini-specific options (thinking, safety, grounding, caching) -... Pull #194 neilberkman github
51.74
20cb655c7e63ec10c089c22690d0b034dda10c47-PR-193 feat/google-context-caching feat: Add Google Context Caching support for Gemini models Adds explicit context caching API for Gemini models to reduce costs by up to 90% when reusing large amounts of content. ## Changes - Add ReqLLM.Providers.Google.CachedContent module for... Pull #193 neilberkman github
51.79
7177c7cf8704fb09797422dd712c269426699916-PR-192 fix/google-cached-tokens Fix: Extract cached token counts from Google API responses Fixes #191 Google's API returns cachedContentTokenCount in usageMetadata for both implicit (automatic) and explicit (CachedContent API) prompt caching. The provider was not extracting th... Pull #192 neilberkman github
52.83
8975f67f117f9461da65570207258acc4bd2e4e9-PR-192 fix/google-cached-tokens Fix: Extract cached token counts from Google API responses Fixes #191 Google's API returns cachedContentTokenCount in usageMetadata for both implicit (automatic) and explicit (CachedContent API) prompt caching. The provider was not extracting th... Pull #192 neilberkman github
52.79
6fe32b91cd07861e771cebc4b8f3115bcb8219a9 main Include Credo config file (#186) * Include Credo config file * Skip credo rule for Alias * Fix multiple format and credo issues - Trailing whitespaces - Alias loading alphabetical order - Remove apply/2 when known argument number push web-flow github
52.76
eece47f13fa73b16123f696956037a33e2aaeb4b-PR-186 credo-config Fix multiple format and credo issues - Trailing whitespaces - Alias loading alphabetical order - Remove apply/2 when known argument number Pull #186 3zcurdia github
52.8
4de19f4d1143d531f0162c3aee0f729e34f42c9d-PR-186 credo-config Skip credo rule for Alias Pull #186 3zcurdia github
52.8
fe2fc913f969b3d215c2b7d6d3c4ea28b4a8ac4d main feat: add StreamResponse.process_stream/2 for real-time callbacks (#178) Adds process_stream/2 function to StreamResponse that enables real-time stream processing with incremental callbacks, avoiding double-stream consumption bugs. Features: -... push web-flow github
52.8
f855392fadfb99176572d59812a15c15aea90e4d-PR-178 feature/stream-collectors Merge branch 'main' into feature/stream-collectors Pull #178 web-flow github
52.8
bf6ad8b690a17d3aeba9c768b988a29d2b4beb4c main Fix Cerebras strict mode handling and add zai-glm-4.6 (#180) Changes from model name string matching to metadata-based approach for determining strict mode support, following OpenAI provider pattern. - Add supports_strict_tools metadata to Cereb... push web-flow github
52.47