Releases: langchain-ai/langsmith-sdk
Releases · langchain-ai/langsmith-sdk
v0.4.20
What's Changed
- feat: update SDK for org-scoped keys by @keshivtandon in #1920
- feat(pythong): allow nested openai agent tracing by @angus-langchain in #1971
- fix(python): Avoid resetting operation.name to "chain" during update_run by @qnnn in #1861
- fix(js): specify "Content-Type" for getRunStats to avoid parsing errors by @szedelenyi in #1598
- feat(py): Add support system qwen to OTEL attributes by @dubin555 in #1717
- feat(js): add datasetVersion parameter to listProjects by @tendev-eoghan in #1391
- Pass multiple run ids to listFeedbacks correctly by @airhorns in #1465
- feat(py): Fix list_projects typing, add dataset_version param by @jacoblee93 in #1973
- feat(js): Allow manually specifying ids for examples in Jestlike runners by @jacoblee93 in #1972
- feat(js): Add current repetition property to jestlike test wrapper by @tendev-liam in #1856
- release(py): 0.4.20 by @jacoblee93 in #1975
New Contributors
- @keshivtandon made their first contribution in #1920
- @qnnn made their first contribution in #1861
- @szedelenyi made their first contribution in #1598
- @dubin555 made their first contribution in #1717
- @airhorns made their first contribution in #1465
Full Changelog: v0.4.19...v0.4.20
v0.4.19
What's Changed
- release(js): 0.3.64 by @jacoblee93 in #1964
- fix(py): Disable API key warnings if tracing is disabled by @jacoblee93 in #1965
- fix(python): Alleviate memory pressue from zstd flush by @angus-langchain in #1966
Full Changelog: v0.4.18...v0.4.19
v0.4.18
What's Changed
- fix(py,js): Trace AzureOpenAI clients with proper provider for playground by @jacoblee93 in #1961
- fix(js): Fix multipart tracing for Bun by @jacoblee93 in #1960
- release(py): 0.4.18 by @jacoblee93 in #1963
Full Changelog: v0.4.17...v0.4.18
v0.4.17
What's Changed
- fix(js): Allow null to be passed as a first arg to traceable by @jacoblee93 in #1959
- fix: move otel check to client init time to avoid timing issue by @EugeneJinXin in #1958
Full Changelog: v0.4.16...v0.4.17
v0.4.16
What's Changed
- fix(py): adjust traceable reduce_fn type to allow returning str by @xl0 in #1950
- fix(js): Fix retry issues by @jacoblee93 in #1947
- fix(js): Turn down default max batch size by @jacoblee93 in #1954
- release(js): 0.3.63 by @jacoblee93 in #1955
- fix(python): Correctly set multipart operatios in replica mode by @angus-langchain in #1956
New Contributors
Full Changelog: v0.4.15...v0.4.16
v0.4.15
What's Changed
- feat(js): Allow async processOutputs methods for traceables by @jacoblee93 in #1928
- fix(py): Pass 'project_name' to _get_parent_run method by @yakir-nevona in #1927
- feat(js): Adds time to first token for new AI SDK wrapper by @jacoblee93 in #1930
- fix(js): Only inherit specific properties for child runs in AI SDK wrapper by @jacoblee93 in #1931
- fix(js): Copy tool def instead of mutating by @jacoblee93 in #1937
- feat(js): Improve tracing for multimodal AI SDK runs by @jacoblee93 in #1938
- feat(js): Trace AI SDK request/response fields in LLM response by @jacoblee93 in #1939
- chore(py): correct references to
langchain-core
and hyperlink by @mdrxy in #1936 - feat(python): Pass api config through background thread by @angus-langchain in #1940
- fix(js): Avoid double wrapping AI SDK tool functions if already wrapped by @jacoblee93 in #1942
- fix(js): Fix typing for AI SDK stream methods by @jacoblee93 in #1941
- feat(js): Allow passing runtime LangSmith config to AI SDK wrapper by @jacoblee93 in #1944
- feat(js): Compress multipart batches before sending, avoid sending inputs in patch if not necessary by @jacoblee93 in #1945
- feat(py): Adds process_buffered_run_ops method to Python SDK by @jacoblee93 in #1932
New Contributors
- @yakir-nevona made their first contribution in #1927
- @mdrxy made their first contribution in #1936
Full Changelog: v0.4.14...v0.4.15
v0.4.14
What's Changed
- fix(js): Revert legacy AI SDK dotted order changes by @jacoblee93 in #1907
- fix(js): Revert "feat(js): Adds dotted order to traced OTEL spans (#1876)" by @jacoblee93 in #1909
- release(js): 0.3.53 by @jacoblee93 in #1910
- chore: Replace ts-expect-error directives with casts by @jacoblee93 in #1913
- fix(js): Use start time from dotted order as canonical for legacy AI SDK exporter by @jacoblee93 in #1912
- feat(js): Allow use of new AI SDK processor with old exporter by @jacoblee93 in #1914
- release(js): 0.3.55 by @jacoblee93 in #1915
- fix(js): Allow direct wrapping of Jest/vitest methods, fix module resolution by @jacoblee93 in #1919
- chore(python): Cleanup replica setting by @angus-langchain in #1921
- perf(python, js): Pop all allowed oob keys by @angus-langchain in #1923
- chore(python): fix tests by @angus-langchain in #1925
- feat(js): Non-OTEL version of AI SDK tracing by @jacoblee93 in #1926
- fix(py): attachments url construction by @baskaryan in #1929
Full Changelog: v0.4.13...v0.4.14
v0.4.13
What's Changed
- Fix OpenAI Agents SDK wrapper token count extraction by @siviter-t in #1906
- fix(python): legacy replicas conversion by @lc-arjun in #1908
New Contributors
- @siviter-t made their first contribution in #1906
Full Changelog: v0.4.12...v0.4.13
v0.4.12
What's Changed
- fix(ci): Allow 429s for markdown link checks in CI by @jacoblee93 in #1895
- feat(py): add Run.latency by @baskaryan in #1745
- fix(js): Ensure dotted order and start time is always greater than parent start time for legacy AI SDK exporter by @jacoblee93 in #1903
- chore(python): Remove langsmith dotted order attribute from otel integration by @angus-langchain in #1904
Full Changelog: v0.4.11...v0.4.12
v0.4.11
What's Changed
- chore: Run ruff by @jacoblee93 in #1894
- Adopt pagination to list_shared_examples to ensure all data returned by @EugeneJinXin in #1893
- feat(python, js): Support tracing to multiple orgs/workspaces by @angus-langchain in #1899
Full Changelog: v0.4.10...v0.4.11