4.3.0 (2024-06-07)
- Adds timing info to llm_classify (#3377) (3e2785f)
- Serializable execution details (#3358) (fc74513)
- ui: display input and output for tool spans (if available) (#3396) (73312dc)
- add separate package installations to notebooks (#3393) (914e3fe)
- filter out undefined (#3383) (e3a2d31)
- percentage sign for alembic configparser (#3403) (87bcd59)
4.2.4 (2024-05-28)
4.2.3 (2024-05-23)
4.2.2 (2024-05-23)
4.2.1 (2024-05-23)
4.2.0 (2024-05-23)
4.1.3 (2024-05-22)
4.1.2 (2024-05-20)
4.1.1 (2024-05-17)
4.1.0 (2024-05-17)
- bump base image in kustomize (#3193) (5e8bc3d)
- PHOENIX_WORKING_DIR default value documentation (#3190) (6957bd9)
4.0.3 (2024-05-13)
4.0.2 (2024-05-11)
- Bulk inserter begins first insert immediately (#3151) (7e17cb2)
- unflatten attributes when loading spans from
trace_dataset
(#3170) (a165023)
4.0.1 (2024-05-09)
4.0.0 (2024-05-09)
⚠ BREAKING CHANGES
- Remove experimental module (#2945)
- Add log_traces method that sends TraceDataset traces to Phoenix (#2897) (c8f9ed2)
- add a last N time range selector on project / projects pages (#2907) (3c115f8)
- add bedrock claude tracing tutorial (#2919) (b8b5240)
- add default limit to /v1/spans and corresponding client methods (#3026) (e5698d7)
- add gradient start/end to projects table (#2956) (5b6b217)
- add grpc endpoint (#2232) (8bbd136)
- Add indexes on Annotation tables (#3082) (682ecee)
- Add indexes on spans table (#3098) (12d2574)
- add opentelemetry trace instrumentation for Phoenix server (#2990) (6ed494e)
- Add SQL and Code Functionality Eval Templates (#2861) (c7d776a)
- add trace and document evals to GET v1/evaluations (#2910) (79229f2)
- Add user frustration eval (#2928) (406938b)
- Added support for default_headers for azure_openai. (#2917) (6ee5f24)
- convert graphql api to pull trace evaluations from db (#2867) (11aa455)
- Deprecate datasets module, rename to inferences (#2785) (4987ea3)
- experimental: postgres support (a2657d4)
- fetch annotation names (#2964) (6c5d25d)
- fetch document retrieval metrics per span using SQL (#2960) (9fdb765)
- graphql api pulls from db for document evaluations (#2865) (e4b667d)
- grpc interceptor for prometheus (#3056) (610c8fa)
- ingest document evals (#2847) (f3fde50)
- ingest pyarrow span evals into sqlite (#2837) (3a6666c)
- ingest trace annotations (#2852) (792f674)
- make graphql api for span evaluations read from database (#2860) (5adf750)
- move document evaluation summary to pull from db (#2888) (73ca2d7)
- openapi ui for api exploration (#3041) (5b22961)
- persistence: add support for sorting by eval scores and labels (#2977) (44c3068)
- persistence: bulk inserter for spans (#2808) (9ce841e)
- persistence: clear project (#2976) (665c166)
- persistence: clear traces UI (#2988) (a717ff6)
- persistence: dataloader for document retrieval metrics (#2978) (f55c458)
- persistence: dataloader for span descendants (#2980) (d8e10d4)
- persistence: ensure migrations run for TreadSession (#2855) (ec4fea7)
- persistence: fetch latency_ms percentiles using sql with dataloaders (#2818) (48d4643)
- persistence: fetch streaming_last_updated_at (#2819) (d665e49)
- persistence: get or delete projects using sql (#2839) (527b9a9)
- persistence: json binary for postgres (#2849) (29351bf)
- persistence: launch app with persist (#2817) (add6103)
- persistence: make launch_app runnable on tmp directory (#2851) (f41e922)
- persistence: span annotation tables (#2788) (874c61e)
- persistence: span query DSL with SQL (#2911) (7c01420)
- persistence: sql sorting for spans (#2823) (eeafb64)
- persistence: use sqlean v3.45.1 as sqlite engine (#2947) (3b202d7)
- Remove experimental module (#2945) (01758cf)
- restrict project metrics to be last 7 days (#2896) (066bc16)
- span filtering by span evaluations (#2923) (4458ec4)
- Support basic auth (#3061) (3202256)
- support for span evaluations to get evaluations endpoint (#2900) (379e336)
- support pagination on spans resolver (#3046) (2113c5c)
- Update API for OpenAPI compliance (#2866) (0db65d8)
- Update eval summaries to use persistence (#2920) (06eb320)
- add the remainder of the sentence (#2903) (64874b8)
- backward compatible truthiness for query from dict parsing (#3124) (b425f9d)
- cartesian product in sql join (#2959) (c96092d)
- cartesian products in get_evaluations (#3081) (64ebec8)
- check payload for legacy project_name (#3125) (d7eae60)
- close delete modal on delete (#3069) (083a467)
- commit insert into alembic_version (#3115) (93a144f)
- disable client-side sorting on trace/span tables (#2958) (139dc3e)
- disable grpc when readonly (#3105) (71ceba9)
- Dockerfile launches Phoenix that listens on IPv6 (#3047) (75cc979)
- eliminate interference on global tracer provider (#2998) (5d7b843)
- Enable listening on IPv6 (#3037) (dee6681)
- ensure recent version of opentelemetry-proto is used (#2948) (33647f5)
- evals: incorrect wording in hallucinations (#3085) (7aa0292)
- fix docker build for sql (b6d508d)
- forbid blank or empty evaluation names (#2962) (cb87977)
- improve error handling and logging for eval insertions (#2854) (d04694b)
- include migration files (#2887) (b0a772e)
- Invalidate cache on project reset (#3113) (2944ae5)
- normalize datetime for phoenix client (#3088) (94a25ae)
- normalize telemetry url before setup (#3001) (28389e8)
- persistence: db race condition between spans and evals (#2905) (2666464)
- persistence: import asert_never from typing_extensions (#2850) (62644cb)
- persistence: postgres down migration and url support (#2915) (4b4a776)
- persistence: postgres json calculations (#2848) (45f084d)
- persistence: postgres timestamp insertion (#2844) (3477bb9)
- preserve loggers across migrations (#2835) (2821bb4)
- prometheus transaction timers for bulkloader (#3066) (e0cc58d)
- Propagate migration errors and show an informative message (#2994) (3718e10)
- remove broken non-asyncio prometheus grpc server interceptor (#3065) (af75151)
- round down time points to facilitate caching (#3079) (42b03c9)
- run docker as nonroot user (#3100) (c640678)
- safely unpack Evaluations proto in bulk inserter (#2869) (50517f7)
- span and trace evaluation summaries (#3013) (088e6c2)
- span event to dict conversion (#3009) (3c73f03)
- switch license format in toml (5c6f345)
- typo in SpanAnnotation (#2967) (f41044e)
- typo in trace annotation table name (#2946) (344b858) Documentation
- Add log_traces tutorial (#2902) (e583f03)
- development: make it explicit that you need to run pnpm build (#3035) (672cbed)
- dockerize manual instrumentation example (#2797) (651efbe)
- manually instrumented chatbot (#2730) (46be32b)
- remove experimental tags in code (4c4a832)
3.25.0 (2024-05-06)
- evals: incorrect wording in hallucinations (#3085) (7aa0292)
- run docker as nonroot user (#3100) (c640678)
3.24.0 (2024-04-22)
3.23.0 (2024-04-19)
3.22.0 (2024-04-16)
3.21.0 (2024-04-12)
3.20.0 (2024-04-10)
- dockerize manual instrumentation example (#2797) (651efbe)
- remove experimental tags in code (4c4a832)
3.19.4 (2024-04-04)
- switch license format in toml (5c6f345)
- fix qa with reference tutorial (e1db1ce)
- fix qa with reference tutorial (ba24950)
- make dockerhub URL go to public (6650f67)
- manually instrumented chatbot (#2730) (46be32b)
3.19.3 (2024-03-30)
3.19.2 (2024-03-29)
- ui: broken context for markdown (556e901)
3.19.1 (2024-03-29)
- UI: color rotation for markdown (3184359)
3.19.0 (2024-03-29)
3.18.1 (2024-03-28)
- ignore docs/ directory when formatting (#2714) (1340f74)
- repair frontend build step in release pipeline (#2716) (796eb6a)
3.18.0 (2024-03-28)
3.17.1 (2024-03-24)
- Add mistral (GITBOOK-594) (78676af)
- add mistral instrumentation to notebook (#2681) (54dc47d)
- add mistral instrumentor to mistral tutorial (#2682) (13fc1f8)
- Evals Structure! (GITBOOK-547) (ac23311)
- fix missing parentheses (GITBOOK-571) (2353953)
- Mistral (GITBOOK-595) (f245844)
- No subject (GITBOOK-597) (b6196ac)
- No subject (GITBOOK-598) (f6a2bd6)
- Remove pinecone notebook (#2665) (9f1c1d4)
- trace a deployed app (GITBOOK-593) (08623ea)
3.17.0 (2024-03-21)
- Add
response_format
argument toMistralAIModel
(#2660) (7da51af) - evals: Add Mistral as an eval model (#2640) (c13ab6b)
3.16.3 (2024-03-20)
3.16.2 (2024-03-20)
- trace: redefine root span (#2632) (7940c9d)
- ui: increase pagination size for TracePage (#2642) (6cd456f)
3.16.1 (2024-03-19)
3.16.0 (2024-03-15)
3.15.1 (2024-03-15)
3.15.0 (2024-03-14)
- launch_app() with experimental span storage using environment variables for storage path and storage type enums (#2564) (8a0b572)
- project archiving and deletion (#2585) (121f904)
- projects: the home page should direct you to the projects page if there are multiple projects with data (#2586) (ced4e75)
- use environment variable for project name (#2590) (e2ace76)
3.14.2 (2024-03-14)
- increase attributes limit on spans (#2575) (94b1930)
- support numpy arrays in span to json encoder (#2583) (3a297d5)
3.14.1 (2024-03-14)
3.14.0 (2024-03-14)
- experimental span storage with append-only text files (909672b)
- experimental span storage with append-only text files (#2553) (909672b)
3.13.1 (2024-03-13)
3.13.0 (2024-03-13)
3.12.0 (2024-03-13)
3.11.1 (2024-03-12)
3.11.0 (2024-03-11)
- graphql: embed project inside graphql span as private attribute (#2522) (9be1afa)
- trace: context manager to pause tracing (#2520) (6bf7232)
- Update pyproject.toml with proper biline (4fdf710)
3.10.0 (2024-03-09)
- projects: add support for the PHOENIX_PROJECT_NAME param (#2515) (6f24786)
- show first non-empty project (#2508) (54a2834)
3.9.0 (2024-03-08)
3.8.0 (2024-03-07)
The Phoenix evals
module is graduating out of experimental
! You can now install Phoenix evals as a standalone package with pip install arize-phoenix-evals
or you can include the new version of phoenix.evals
along with the Phoenix install with pip install -U arize-phoenix[evals]
. Swapping to the new evals
module includes a few small breaking changes which might require some migration work. Details can be found in MIGRATION.md
.
phoenix.experimental.evals
is being deprecated and will remain in Phoenix for about a month before being removed.
- gql: add trace count to gql project (#2484) (91b4ae1)
- Integrate
phoenix.evals
intophoenix
(#2420) (dd3e7b4)
3.7.0 (2024-03-07)
3.6.0 (2024-03-06)
- traces: store and query spans by project name (#2433) (b8ef923)
- ui: auto-expand side nav on hover (#2458) (da83f69)
3.5.0 (2024-03-05)
- add metadata to spans and traces table (#2339) (e9725a2)
- Removes token processing module from
phoenix.evals
(#2421) (fbd4961) - ui: new side nav with projects (#2359) (d8c423e)
- Properly define
BedrockModel
(#2425) (81a720c) - remove computed atributes from exported dataframe (#2366) (1de1415)
- turn span_kind enums into string because it's not serializable by pyarrow (#2438) (50c7eb0)
- update rag and llm ops notebooks (#2442) (adf1b2b)
- evals: update tracing tutorials with arize-phoenix-evals (#2386) (1af8187)
- log information about the server at startup (#2445) (6d410c1)
- update readme for phoenix.evals, fix llama-index example (#2435) (dfffaad)
3.4.1 (2024-02-29)
- remove symbolic links for docker build (#2408) (b57abe9)
- source distribution build (#2407) (1e67d7e)
3.4.0 (2024-02-28)
- remove run_relevance_evals and fix import issues (#2375) (9a97e62)
- traces: add y scroll on trace tree (#2399) (9c4f6b9)
- evals: add README (#2363) (47842da)
- evals: migrate evaluation notebooks (#2388) (3dedc6e)
- update ragas integration (#2400) (7bebe98)
3.3.0 (2024-02-23)
- display status description under trace info (#2334) (aed925f)
- show span as soon as they arrive (#2353) (88397a5)
3.2.1 (2024-02-16)
3.2.0 (2024-02-16)
- px.Client
log_evaluations
(#2308) (69a4b2b) - trace: display metadata in the trace page UI (#2304) (fce2d63)
3.1.2 (2024-02-15)
- allow json string for
metadata
span attribute (#2301) (ec7fbe2) - ui: safely parse JSON and fallback to string for span attributes (#2293) (e43cdbb)
3.1.1 (2024-02-15)
- fix: cast message to string in vertexai model (86947a2)
3.1.0 (2024-02-15)
- set global session to None if it fails to start (#2286) (6752fd2)
- trace: Make dataset IDs unique by instance for TraceDataset (#2254) (1ac170f)
3.0.3 (2024-02-13)
3.0.2 (2024-02-13)
3.0.1 (2024-02-09)
3.0.0 (2024-02-09)
- replace Phoenix tracers with OpenInference instrumentors (#2190)
2.11.1 (2024-02-09)
2.11.0 (2024-02-08)
2.10.0 (2024-02-07)
- endpoint for client inside ProcessSession (#2211) (82e279e)
- trace: return to /tracing url when dismissing trace slide over (#2222) (ee4ced3)
- traces: warn if collector endpoint is set but launch app is called (#2209) (eb97b8d)
2.9.4 (2024-02-06)
2.9.3 (2024-02-05)
2.9.2 (2024-02-05)
2.9.1 (2024-02-05)
2.9.0 (2024-02-05)
- phoenix client
get_evaluations()
andget_trace_dataset()
(#2154) (29800e4) - phoenix client
get_spans_dataframe()
andquery_spans()
(#2151) (e44b948)
2.8.0 (2024-02-02)
- broken link and openinference links (#2144) (01fb046)
- databricks check crashes in python console (#2152) (5aeeeff)
- default collector endpoint breaks on windows (#2161) (f1a2007)
- Do not retry when context window has been exceeded (#2126) (ff6df1f)
- remove hyphens from span_id in legacy evaluation fixtures (#2153) (fae859d)
- add docker badge (e584ed8)
- Add terminal running steps (GITBOOK-441) (91c6b24)
- No subject (GITBOOK-442) (5c4eb6c)
- No subject (GITBOOK-443) (11f46cb)
- No subject (GITBOOK-444) (fcf2bc9)
- update badge (ddcecea)
- update prompt to reflect rails (GITBOOK-445) (dea6dd6)
2.7.0 (2024-01-24)
2.6.0 (2024-01-23)
- add ability to save and load TraceDatasets (#2082) (60c5e5e)
- add get_trace_dataset method to session (#2107) (9754b60)
- evals: Gpt 4 turbo context window size (#2112) (389c1a0)
- launch phoenix with evaluations (#2095) (9656d0c)
- support eval exports for session (#2094) (8757fa8)
- Clean up vertex clients after event loop closure (#2102) (202c7ea)
- Determine default async concurrency on a per-model basis (#2096) (b44d8aa)
- Resolves Bedrock model compatibility issues (#2114) (c4a5343)
- show localhost when the notebook is running locally (#2090) (095298d)
- evals: update RAG evaluations notebook (#2092) (9ad797a)
- evals: update ragas integration notebook (#2100) (66fb048)
2.5.0 (2024-01-16)
2.4.1 (2024-01-11)
- traces: prevent missing key exception when extracting invocation parameters in llama-index (#2076) (5cc9560)
2.4.0 (2024-01-10)
- add persistence for span evaluations (#2021) (589d482)
- ui: add filter condition snippets (#2049) (567fa54)
- Handle missing vertex candidates (#2055) (1d0475a)
- OpenAI clients are not cleaned up after calls to
llm_classify
(#2068) (3233d56) - traces: remove nan from log_evaluations (#2056) (df9ed5c)
2.3.0 (2024-01-08)
- Add demo link, examples getting started (GITBOOK-396) (e987315)
- Add Evaluating Traces Section (GITBOOK-386) (7d72029)
- Add evaluations section for results (GITBOOK-387) (2e74be0)
- Add final thoughts to evaluation (GITBOOK-405) (20eab16)
- add import statement (GITBOOK-408) (23247d7)
- add link (GITBOOK-403) (0be280a)
- eval concepts typo (GITBOOK-394) (7c80d4b)
- eval concepts typos (GITBOOK-393) (62bc99f)
- evaluation concepts typo fix (GITBOOK-390) (2cbc1dc)
- Extract Data from Spans (GITBOOK-383) (440f530)
- fix broken section link (GITBOOK-409) (fee537b)
- fix typos (GITBOOK-391) (c8f5a55)
- fix typos (GITBOOK-402) (3cd973d)
- fix typos (GITBOOK-406) (eaa9bea)
- fix typos (GITBOOK-407) (cad4820)
- Initial draft of evaluation core concept (GITBOOK-385) (67369cf)
- Log Evaluations (GITBOOK-389) (369d79d)
- No subject (GITBOOK-399) (94df884)
- Re-arrange nav (GITBOOK-398) (54a87eb)
- Remove the word golden, simplify title (GITBOOK-395) (a2233b2)
- simplify conceps (GITBOOK-384) (c38f6c2)
- Simplify examples page (GITBOOK-400) (6144158)
- Trace Evaluations Section (GITBOOK-388) (2ffa800)
- Update SECURITY.md (#2029) (363e891)
2.2.1 (2023-12-28)
- Do not retry if eval was successful when using SyncExecutor (#2016) (a869190)
- ensure float values are properly encoded by otel tracer (#2024) (b12a894)
- ensure llamaindex spans are correctly encoded (#2023) (3ca6262)
- Use separate versioning file (#2020) (f38eedf)
2.2.0 (2023-12-22)
- Add support for Google's Gemini models via Vertex python sdk (#2008) (caf826c)
- Support first-party Anthropic python SDK (#2004) (a323283)
2.1.0 (2023-12-21)
- instantiate evaluators by criteria (#1983) (9c72616)
- support function calling for run_evals (#1978) (8be325c)
- traces: add
v1/traces
HTTP endpoint to handleExportTraceServiceRequest
(3c94dea) - traces: add
v1/traces
HTTP endpoint to handleExportTraceServiceRequest
(#1968) (3c94dea) - traces: add retrieval summary to header (#2006) (8af0582)
- traces: evaluation summary on the header (#2000) (965beb0)
2.0.0 (2023-12-20)
- Update
llm_classify
andllm_generate
interfaces (#1974)
- Add async submission to
llm_generate
(#1965) (5999133) - add support for explanations to run_evals (#1975) (5143529)
- evaluation column selectors (#1932) (ed07809)
- openai streaming tool calls (#1936) (6dd14cf)
- support running multiple evals at once (#1742) (79d4473)
- Update
llm_classify
andllm_generate
interfaces (#1974) (9fd35a1)
- Add lock failsafe (#1956) (9ddbd9c)
- llama-index extra (#1958) (d9b68eb)
- LlamaIndex compatibility fix (#1940) (052349d)
- Model stability enhancements (#1939) (dca42e0)
- traces: span summary root span filter (#1981) (d286f07)
- Add anyscale tutorial (#1941) (e47c8d0)
- autogen link (#1946) (c3fb4ce)
- Clear anyscale tutorial outputs (#1942) (63580a6)
- RAG Evaluation (GITBOOK-378) (429f537)
- sync (#1947) (c72bbac)
- traces: autogen tracing tutorial (#1945) (0fd02ff)
- update rag eval notebook (#1950) (d06b8b7)
- update rag evals docs (#1954) (aa6f36a)
- Using phoenix with HuggingFace LLMs- getting started (#1916) (b446972)
1.9.0 (2023-12-11)
1.8.0 (2023-12-10)
- embeddings: audio support (#1920) (61cc550)
- openai streaming function call message support (#1914) (25279ca)
1.7.0 (2023-12-09)
- Instrument LlamaIndex streaming responses (#1901) (f46396e)
- openai async streaming instrumentation (#1900) (06d643b)
- traces: query spans into dataframes (#1910) (6b51435)
1.6.0 (2023-12-08)
- openai streaming spans show up in the ui (#1888) (ffa1d41)
- support instrumentation for openai synchronous streaming (#1879) (b6e8c73)
- traces: display document retrieval metrics on trace details (#1902) (0c35229)
- traces: filterable span and document evaluation summaries (#1880) (f90919c)
- traces: graphql query for document evaluation summary (#1874) (8a6a063)
1.5.1 (2023-12-06)
1.5.0 (2023-12-06)
- evals: Human vs AI Evals (#1850) (e96bd27)
- semantic conventions for
tool_calls
array in OpenAI ChatCompletion messages (#1837) (c079f00) - support asynchronous chat completions for openai instrumentation (#1849) (f066e10)
- traces: document retrieval metrics based on document evaluation scores (#1826) (3dfb7bd)
- traces: document retrieval metrics on trace / span tables (#1873) (733d233)
- traces: evaluation annotations on traces for associating spans with eval metrics (#1693) (a218a65)
- traces: server-side span filter by evaluation result values (#1858) (6b05f96)
- traces: span evaluation summary (aggregation metrics of scores and labels) (#1846) (5c5c3d6)
- allow streaming response to be iterated by user (#1862) (76a2443)
- trace dataset to disc (#1798) (278d344)
1.4.0 (2023-11-30)
- propagate error status codes to parent spans for improved visibility into trace exceptions (#1824) (1a234e9)
1.3.0 (2023-11-30)
- Add OpenAI Rate limiting (#1805) (115e044)
- evals: show span evaluations in trace details slideout (#1810) (4f0e4dc)
- evaluation ingestion (no user-facing feature is added) (#1764) (7c4039b)
- feature flags context (#1802) (a2732cd)
- Implement asynchronous submission for OpenAI evals (#1754) (30c011d)
- reference link correctness evaluation prompt template (#1771) (bf731df)
- traces: configurable endpoint for the exporter (#1795) (8515763)
- traces: display document evaluations alongside the document (#1823) (2ca3613)
- traces: server-side sort of spans by evaluation result (score or label) (#1812) (d139693)
- traces: show all evaluations in the table" (#1819) (2b27333)
- traces: Trace page header with latency, status, and evaluations (#1831) (1d88efd)
- enhance llama-index callback support for exception events (#1814) (8db01df)
- pin llama-index temporarily (#1806) (d6aa76e)
- remove sklearn metrics not available in sagemaker (#1791) (20ab6e5)
- traces: convert (non-list) iterables to lists during protobuf construction due to potential presence of ndarray when reading from parquet files (#1801) (ca72747)
- traces: make column selector sync'd between tabs (#1816) (125431a)
- Environment documentation (GITBOOK-370) (dbbb0a7)
- Explanations (GITBOOK-371) (5f33da3)
- No subject (GITBOOK-369) (656b5c0)
- sync for 1.3 (#1833) (4d01e83)
- update default value of variable in run_relevance_eval (GITBOOK-368) (d5bcaf8)
1.2.1 (2023-11-18)
- make the app launchable when nest_asyncio is applied (#1783) (f9d5085)
- restore process session (#1781) (34a32c3)
1.2.0 (2023-11-17)
- Add dockerfile (#1761) (4fa8929)
- evals: return partial results when llm function is interrupted (#1755) (1fb0849)
- LiteLLM model support for evals (#1675) (5f2a999)
- sagemaker nobebook support (#1772) (2c0ffbc)
1.1.1 (2023-11-16)
1.1.0 (2023-11-14)
- Evals with explanations (#1699) (2db8141)
- evals: add an output_parser to llm_generate (#1736) (6408dda)
1.0.0 (2023-11-10)
- models: openAI 1.0 (#1716)
0.1.1 (2023-11-09)
0.1.0 (2023-11-08)
- add long-context evaluators, including map reduce and refine patterns (#1710) (0c3b105)
- traces: span table column visibility controls (#1687) (559852f)