Releases: epam/ai-dial-adapter-openai
Releases · epam/ai-dial-adapter-openai
0.23.1
Other
- bumps xz-libs from 5.6.3-r0 to 5.6.3-r1 (#212)
0.24.0
Features
- support auto-caching for GPT o-series model family (#209)
Other
- bump epam/ai-dial-ci from 1.11.1 to 1.12.1 (#208)
0.23.0
Features
- migrate Dall-E-3 from making class via HTTP client to
openai
library (#205)
0.22.0
Features
- add configuration for Dall-E-3 via
/configuration
endpoint (#202)
Fixes
- allow user-defined
model
request field in embeddings requests (#200)
- predownload
tiktoken
tokenizer during Docker image build time (#201)
0.21.0
Fixes
- fix error handling for multi-modal GPT4 models (#198)
- allow user-defined
model
field in chat completion requests (#197)
Other
- bump cryptography from 43.0.1 to 44.0.1 (#196)
- add GitHub issue and pull request templates
0.20.0
Fixes
- remove the invalidated
content-length
header from the error response (#192)
Other
- bump epam/ai-dial-ci from 1.10.2 to 1.11.0 (#191)
- bump epam/ai-dial-ci from 1.11.0 to 1.11.1 (#194)
0.19.0
Other
- add integration tests (#185)
- pin poetry to 1.8.5 (#189)
- bump epam/ai-dial-ci from 1.10.0 to 1.10.2 (#187)
- bump virtualenv from 20.24.5 to 20.26.6 (#190)
0.18.0
Features
- introduce application config that simplifies handling of env variables and eases testing (#178)
Fixes
- tolerate tokenizer failures (#183)
- improve error handling (#182)
Other
- bump aidial-sdk from 0.15.0 to 0.16.0; bump protobuf from 5.29.0 to 5.29.1 (#181)
0.17.2
Fixes
- Supported approximate tokenization of tools and functions (#170)
0.17.1
Fixes
- gpt-4o: fixed failing tool/function calling (#175, #176)
- removed
Content-Encoding
header from response to avoid decoding errors from gzip (#174, #176)
- truncated long error messages in the plain-text tokenizer (#172, #176)
Other
- bump aiohttp from 3.10.2 to 3.10.11 (#173, #176)