arize-phoenix 3.0.1__py3-none-any.whl → 3.0.3__py3-none-any.whl

{arize_phoenix-3.0.1.dist-info → arize_phoenix-3.0.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: arize-phoenix
-Version: 3.0.1
+Version: 3.0.3
 Summary: ML Observability in your notebook
 Project-URL: Documentation, https://docs.arize.com/phoenix/
 Project-URL: Issues, https://github.com/Arize-ai/phoenix/issues
@@ -23,6 +23,7 @@ Requires-Dist: numpy
 Requires-Dist: openinference-instrumentation-langchain
 Requires-Dist: openinference-instrumentation-llama-index
 Requires-Dist: openinference-instrumentation-openai
+Requires-Dist: openinference-semantic-conventions
 Requires-Dist: opentelemetry-exporter-otlp
 Requires-Dist: opentelemetry-proto
 Requires-Dist: opentelemetry-sdk
@@ -50,7 +51,7 @@ Requires-Dist: hatch; extra == 'dev'
 Requires-Dist: jupyter; extra == 'dev'
 Requires-Dist: langchain>=0.0.334; extra == 'dev'
 Requires-Dist: litellm>=1.0.3; extra == 'dev'
-Requires-Dist: llama-index>=0.9.14; extra == 'dev'
+Requires-Dist: llama-index<0.10.0; extra == 'dev'
 Requires-Dist: nbqa; extra == 'dev'
 Requires-Dist: pandas-stubs<=2.0.2.230605; extra == 'dev'
 Requires-Dist: pre-commit; extra == 'dev'
@@ -64,6 +65,7 @@ Provides-Extra: experimental
 Requires-Dist: tenacity; extra == 'experimental'
 Provides-Extra: llama-index
 Requires-Dist: llama-index==0.9.45; extra == 'llama-index'
+Requires-Dist: openinference-instrumentation-llama-index==0.1.3; extra == 'llama-index'
 Description-Content-Type: text/markdown
 
 <p align="center">

{arize_phoenix-3.0.1.dist-info → arize_phoenix-3.0.3.dist-info}/RECORD

@@ -4,14 +4,14 @@ phoenix/datetime_utils.py,sha256=D955QLrkgrrSdUM6NyqbCeAu2SMsjhR5rHVQEsVUdng,277
 phoenix/exceptions.py,sha256=X5k9ipUDfwSCwZB-H5zFJLas86Gf9tAx0W4l5TZxp5k,108
 phoenix/py.typed,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
 phoenix/services.py,sha256=f6AeyKTuOpy9RCcTCjVH3gx5nYZhbTMFOuv1WSUOB5o,4992
-phoenix/version.py,sha256=E3P6AbnCwaWk6ndR1zNqlOTVebX9z5rv9voltc71dos,22
+phoenix/version.py,sha256=3PslnGRHLeT8kAWbhtBM110cQkzH_QzfQO5_B6lHOuU,22
 phoenix/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 phoenix/core/embedding_dimension.py,sha256=zKGbcvwOXgLf-yrJBpQyKtd-LEOPRKHnUToyAU8Owis,87
 phoenix/core/evals.py,sha256=gJyqQzpud5YjtoY8h4pgXvHDsdubGfqmEewLuZHPPmQ,10224
 phoenix/core/model.py,sha256=vQ6RxpUPlncezJvur5u6xBN0Lkrk2gW0cTyb-qqaSqA,4713
 phoenix/core/model_schema.py,sha256=rR9VdhL_oXxbprDTPQJBXs5hw5sMPQmzx__m6Kwsxug,50394
 phoenix/core/model_schema_adapter.py,sha256=3GkyzqUST4fYi-Bgs8qAam5hwMCdQRZTDLjZ9Bnzdm4,8268
-phoenix/core/traces.py,sha256=0-CApYIFoppwKGss70_-xX_s8FPI2ABfTyT7pyeI_eE,14007
+phoenix/core/traces.py,sha256=I9xsQvEd4s7fZa1lquzBfnY7ZQoLcWwnT51wuuglnyM,14065
 phoenix/datasets/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 phoenix/datasets/dataset.py,sha256=scKVZ7zc6Dpc_ntt-pWhzY-KWqOJEwKePuyNnKSVTGE,30515
 phoenix/datasets/errors.py,sha256=cGp9vxnw4SewFoWBV3ZGMkhE0Kh73lPIv3Ppz_H_RoA,8261
@@ -23,7 +23,7 @@ phoenix/experimental/evals/__init__.py,sha256=q96YKLMt2GJD9zL8sjugvWx1INfw40Wa7E
 phoenix/experimental/evals/evaluators.py,sha256=r7fXrS-l4gn58SUhLAZSfY3P8lxysouSVJwHddrZJ_Q,15956
 phoenix/experimental/evals/retrievals.py,sha256=o3fqrsYbYZjyGj_jWkN_9VQVyXjLkDKDw5Ws7l8bwdI,3828
 phoenix/experimental/evals/functions/__init__.py,sha256=NNd0-_cmIopdV7vm3rspjfgM726qoQJ4DPq_vqbnaxQ,180
-phoenix/experimental/evals/functions/classify.py,sha256=6yCajPT9i98b4_2qYn9ZxGhdI3CLhfUSrEyUUcqQqmQ,19517
+phoenix/experimental/evals/functions/classify.py,sha256=rIQLf0GRXPNT2d5Oni0lrpFrubBorAscjN0ievjWx9A,19092
 phoenix/experimental/evals/functions/executor.py,sha256=bM7PI2rcPukQQzZ2rWqN_-Kfo_a935YJj0bh1Red8Ps,13406
 phoenix/experimental/evals/functions/generate.py,sha256=8LnnPAjBM9yxitdkaGZ67OabuDTOWBF3fvinJ_uCFRg,5584
 phoenix/experimental/evals/functions/processing.py,sha256=F4xtLsulLV4a8CkuLldRddsCim75dSTIShEJUYN6I6w,1823
@@ -39,7 +39,7 @@ phoenix/experimental/evals/models/vertexai.py,sha256=_txsOP2RHyR3AnugeJRFUNvYm3x
 phoenix/experimental/evals/templates/__init__.py,sha256=GSJSoWJ4jwyoUANniidmWMUtXQhNQYbTJbfFqCvuYuo,1470
 phoenix/experimental/evals/templates/default_templates.py,sha256=dVKmoLwqgAyGcRuezz9WKnXSHhw7-qk1R8j6wSmqh0s,20722
 phoenix/experimental/evals/templates/template.py,sha256=ImFSaTPo9oalPNwq7cNdOCndrvuwLuIyIFKsgDVcoJE,6715
-phoenix/experimental/evals/utils/__init__.py,sha256=608EX7sG0f5oDG__II16J8xnFJiNpY9dI9AC8vXwR00,5601
+phoenix/experimental/evals/utils/__init__.py,sha256=FE1tbb022t3pwn0CrprR7QXcIsgpsdsotV7-iB3khsI,5601
 phoenix/experimental/evals/utils/threads.py,sha256=ksI-egarPnlxit0qKKjtjZ2L82qGLxqxZ6s92O0eBA4,1005
 phoenix/metrics/README.md,sha256=5gekqTU-5gGdMwvcfNp2Wlu8p1ul9kGY_jq0XXQusoI,1964
 phoenix/metrics/__init__.py,sha256=sLp7td1GIt_0Z8dPUyP4L0-_4x9c871yAaGX30oMsvg,2433
@@ -70,7 +70,7 @@ phoenix/server/api/input_types/DimensionFilter.py,sha256=vcXgglSnZcB5pGh-6oEtRmG
 phoenix/server/api/input_types/DimensionInput.py,sha256=Vfx5FmiMKey4-EHDQsQRPzSAMRJMN5oVMLDUl4NKAa8,164
 phoenix/server/api/input_types/Granularity.py,sha256=6SVfZ5yTZYq1PI6vdpjfkBUc4YilLSkF-k6okuSNbbQ,2301
 phoenix/server/api/input_types/PerformanceMetricInput.py,sha256=fElsLTSEYYgGFGMYTEGcYid39tXUKFdV_JkdHavMcbA,591
-phoenix/server/api/input_types/SpanSort.py,sha256=9OxxFseY-9pwZ9rUJ71P-h_vpwYg5UAiy76x_pzHvPo,3537
+phoenix/server/api/input_types/SpanSort.py,sha256=72oWhYLLIZLy7K5_tFhMg7RcmDgatqMB-ZoRJhHGIR0,3527
 phoenix/server/api/input_types/TimeRange.py,sha256=yzx-gxj8mDeGLft1FzU_x1MVEgIG5Pt6-f8PUVDgipQ,522
 phoenix/server/api/input_types/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 phoenix/server/api/routers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -109,7 +109,7 @@ phoenix/server/api/types/Retrieval.py,sha256=OhMK2ncjoyp5h1yjKhjlKpoTbQrMHuxmgSF
 phoenix/server/api/types/ScalarDriftMetricEnum.py,sha256=IUAcRPpgL41WdoIgK6cNk2Te38SspXGyEs-S1fY23_A,232
 phoenix/server/api/types/Segments.py,sha256=zogJI9MdmctBL7J-fDSR_8tUJLvuISlVYgCLnTaigKE,2937
 phoenix/server/api/types/SortDir.py,sha256=OUpXhlCzCxPoXSDkJJygEs9Rw9pMymfaZUG5zPTrw4Y,152
-phoenix/server/api/types/Span.py,sha256=QWWW2MmWD0bwiR-jAUtJJZ-GvZJhTFYq7qhus7tsz3A,11304
+phoenix/server/api/types/Span.py,sha256=OvY4ffZa3KMnT3J9s0oLB0ghurYSJCWlK0YfoKkg-6s,11633
 phoenix/server/api/types/TimeSeries.py,sha256=QbLfxHnwYsMsirpq4tx9us6ha7YtAVzK4m8mAL3fMt0,5200
 phoenix/server/api/types/UMAPPoints.py,sha256=8l9RJXi308qty4MdHb2pBbiU6ZuLbrRRxXNbPhXoxKI,1639
 phoenix/server/api/types/ValidationResult.py,sha256=pHwdYk4J7SJ5xhlWWHg_6qWkfk4rjOx-bSkGHvkDE3Q,142
@@ -137,30 +137,28 @@ phoenix/session/data_extractor.py,sha256=V4ntT2FcmbV_-zobcbPR51HKWaDyMnXHj4Pmu7M
 phoenix/session/evaluation.py,sha256=DaAtA0XYJbXRJO_StGywa-9APlz2ORSmCXzxrtn3rvI,4997
 phoenix/session/session.py,sha256=mrTAQkUFMGjN65ApYqRjhGXExbfxu2ODJcJ14MSDOLE,21719
 phoenix/trace/__init__.py,sha256=4d_MqzUIFmlY9WWcFeTONJ4xL5mPGoWZaPM2TJ0ZDBQ,266
-phoenix/trace/errors.py,sha256=DbXSJnNErV7305tKv7pUWLD6jcVHJ6EBdSu4mZJ6IM4,112
+phoenix/trace/errors.py,sha256=wB1z8qdPckngdfU-TORToekvg3344oNFAA83_hC2yFY,180
 phoenix/trace/evaluation_conventions.py,sha256=t8jydM3U0-T5YpiQKRJ3tWdWGlHtzKyttYdw-ddvPOk,1048
 phoenix/trace/exporter.py,sha256=O-9Arn-S_B9Me-jy4Qa84y6lvxKNoa8pczrsamPl3_Q,4871
 phoenix/trace/fixtures.py,sha256=LokNedhbGYxpzXznteO4m5QehvNYjzvoh231-CMJQeY,7113
-phoenix/trace/otel.py,sha256=lJoTvkCSFiPT4dFRLMZZgZfsOoTSATctW9lQk9J30ZI,14821
-phoenix/trace/schemas.py,sha256=QDBlM94faceTnloY_ZVo2BHWWnE5UGymefC0jmfj4ew,6011
-phoenix/trace/semantic_conventions.py,sha256=u6NG85ZhbreriZr8cqJaddldM_jUcew7JilszY7JUk8,4652
+phoenix/trace/otel.py,sha256=naFWTwrqyIoeGJ5YMT9Fyhd44CEtQvpscVluDOXD6Fo,16744
+phoenix/trace/schemas.py,sha256=zlAY8v-PLgEDqTgbmDxY3NJlCAmzj_3ZCJFebQOBz3M,6028
 phoenix/trace/span_evaluations.py,sha256=FvY9YRnKuYIzCa-H9P5SuDaI2DeqGnVCjNgclC2v3HA,12909
-phoenix/trace/span_json_decoder.py,sha256=nrIPkcgbCcNML-0OSjWC6fxIfBEMiP0n67yM_m-vegg,3068
-phoenix/trace/span_json_encoder.py,sha256=C5y7rkyOcV08oJC5t8TZqVxsKCZMJKad7bBQzAgLoDs,1763
-phoenix/trace/trace_dataset.py,sha256=KW0TzmhlKuX8PUPLV172iTK08myYE0QXUC75KiIqJ7k,13204
+phoenix/trace/span_json_decoder.py,sha256=IAFakPRqSMYxTPKYFMiXYxm7U-FipdN8_xbvapDS0Qc,3131
+phoenix/trace/span_json_encoder.py,sha256=hIDd1I6xm01kaNmeKjHOHyxUGI3uTg5J_Os1kXtAb6g,1755
+phoenix/trace/trace_dataset.py,sha256=DF4JH3lq7ULsw6sGo6c4SULChoxNSRJA4knQXfSrTR8,13485
 phoenix/trace/tracer.py,sha256=JDKlyvjy6AsQmaA60ycJ1hKXoUQU61jqPx3nvYr8xUc,3647
 phoenix/trace/utils.py,sha256=7LurVGXn245cjj4MJsc7v6jq4DSJkpK6YGBfIaSywuw,1307
 phoenix/trace/dsl/__init__.py,sha256=WIQIjJg362XD3s50OsPJJ0xbDsGp41bSv7vDllLrPuA,144
-phoenix/trace/dsl/filter.py,sha256=VwlzL4KsIoyYjjCSKXlOSyVm49WyYRlkK5zedXqZlcc,13136
-phoenix/trace/dsl/helpers.py,sha256=qxRS65NFUFVpLDd8IZVrZUH891MamLeXyfIIN1BJJ_c,1552
+phoenix/trace/dsl/filter.py,sha256=7m1v9RtZsAzg-E8iXKLvVSlCJLPtnUapzEIb-nrygsM,13208
+phoenix/trace/dsl/helpers.py,sha256=U71HmCecJQ_zHq0g2mFKbxrRuiJDkopskWMpFoBNP-Y,1722
 phoenix/trace/dsl/missing.py,sha256=BWPOHr2_tBkPDgVeq8GVXXVbNbJiBelu4NtwHBg6mTE,1435
-phoenix/trace/dsl/query.py,sha256=W6_e1o_JbucGWWv6pSUQFyIHcTuj7QuCBSuKXodRfCE,14773
+phoenix/trace/dsl/query.py,sha256=XoFwKEALzGqUERy7B5fgD-n0s87zN6jRVrZgW6-jqRo,14819
 phoenix/trace/langchain/__init__.py,sha256=aTKMFmEOgjx_6dnyplalgYi7PQnetablwrwpcUZqcGE,764
 phoenix/trace/langchain/instrumentor.py,sha256=rmqdaEh2rwyZnaddns4RC2WQnj7TbCnJjETgoiwWvzs,1163
 phoenix/trace/langchain/tracer.py,sha256=fSjtPc5VxaZK63x3ob3DAW7eM2kfOZWgbSQ9dmhhtsw,1424
-phoenix/trace/llama_index/__init__.py,sha256=wCcQgD9CG5TA8i-1XsSed4ZzwHTUmqZwegQAV_FqEng,178
-phoenix/trace/llama_index/callback.py,sha256=MHgUNJSG6ecxZ8SZ_9LdcKnAAWmNI-U5pySFyXqubjM,1694
-phoenix/trace/llama_index/debug_callback.py,sha256=SKToD9q_QADSGTJ5lhilqRVKaUnUSRXUvURCzN4by2U,1367
+phoenix/trace/llama_index/__init__.py,sha256=4fpR5702Qh2t5TaXIx584EkA-BveCPftXPOKvI0Oi3I,105
+phoenix/trace/llama_index/callback.py,sha256=YKCZZtFwAwasZt9qNSJO6dj97Jt2eN_PBakUYWrCleE,3574
 phoenix/trace/openai/__init__.py,sha256=J3G0uqCxGdksUpaQVHds_Egv2drvh8UEqoLjiQAOveg,79
 phoenix/trace/openai/instrumentor.py,sha256=Lre96YfPxs5iM3frZfwdAh94LUpyYw_RZpSF0XmPtKw,1061
 phoenix/trace/v1/__init__.py,sha256=-IbAD0ruESMjvQLvGAg9CTfjBUATFDx1OXseDPis6-0,88
@@ -169,8 +167,8 @@ phoenix/trace/v1/evaluation_pb2.pyi,sha256=cCbbx06gwQmaH14s3J1X25TtaARh-k1abbxQd
 phoenix/utilities/__init__.py,sha256=3TVirVnjIGyaCFuJCqeZO4tjlzQ_chZgYM0itIwsEpE,656
 phoenix/utilities/error_handling.py,sha256=7b5rpGFj9EWZ8yrZK1IHvxB89suWk3lggDayUQcvZds,1946
 phoenix/utilities/logging.py,sha256=lDXd6EGaamBNcQxL4vP1au9-i_SXe0OraUDiJOcszSw,222
-arize_phoenix-3.0.1.dist-info/METADATA,sha256=MIwob2psMophui1KBYcjRPTmKinVvD9YIY0aDqRE5V0,28556
-arize_phoenix-3.0.1.dist-info/WHEEL,sha256=TJPnKdtrSue7xZ_AVGkp9YXcvDrobsjBds1du3Nx6dc,87
-arize_phoenix-3.0.1.dist-info/licenses/IP_NOTICE,sha256=JBqyyCYYxGDfzQ0TtsQgjts41IJoa-hiwDrBjCb9gHM,469
-arize_phoenix-3.0.1.dist-info/licenses/LICENSE,sha256=HFkW9REuMOkvKRACuwLPT0hRydHb3zNg-fdFt94td18,3794
-arize_phoenix-3.0.1.dist-info/RECORD,,
+arize_phoenix-3.0.3.dist-info/METADATA,sha256=LZ4BcSMrV3NFDcOfufwwutdTnKMYt6cBBB7t_BJ02b8,28693
+arize_phoenix-3.0.3.dist-info/WHEEL,sha256=TJPnKdtrSue7xZ_AVGkp9YXcvDrobsjBds1du3Nx6dc,87
+arize_phoenix-3.0.3.dist-info/licenses/IP_NOTICE,sha256=JBqyyCYYxGDfzQ0TtsQgjts41IJoa-hiwDrBjCb9gHM,469
+arize_phoenix-3.0.3.dist-info/licenses/LICENSE,sha256=HFkW9REuMOkvKRACuwLPT0hRydHb3zNg-fdFt94td18,3794
+arize_phoenix-3.0.3.dist-info/RECORD,,

phoenix/core/traces.py

@@ -20,12 +20,13 @@ from typing import (
 
 import opentelemetry.proto.trace.v1.trace_pb2 as otlp
 from ddsketch import DDSketch
+from openinference.semconv.trace import SpanAttributes
 from sortedcontainers import SortedKeyList
 from typing_extensions import TypeAlias
 from wrapt import ObjectProxy
 
+import phoenix.trace.schemas
 from phoenix.datetime_utils import right_open_time_range
-from phoenix.trace import semantic_conventions
 from phoenix.trace.otel import decode
 from phoenix.trace.schemas import (
     ATTRIBUTE_PREFIX,
@@ -33,12 +34,10 @@ from phoenix.trace.schemas import (
     CONTEXT_PREFIX,
     ComputedAttributes,
     Span,
-    SpanAttributes,
     SpanID,
     SpanStatusCode,
     TraceID,
 )
-from phoenix.trace.semantic_conventions import RETRIEVAL_DOCUMENTS
 
 END_OF_QUEUE = None  # sentinel value for queue termination
 
@@ -50,9 +49,9 @@ SPAN_ID = CONTEXT_PREFIX + "span_id"
 PARENT_ID = "parent_id"
 START_TIME = "start_time"
 END_TIME = "end_time"
-LLM_TOKEN_COUNT_TOTAL = ATTRIBUTE_PREFIX + semantic_conventions.LLM_TOKEN_COUNT_TOTAL
-LLM_TOKEN_COUNT_PROMPT = ATTRIBUTE_PREFIX + semantic_conventions.LLM_TOKEN_COUNT_PROMPT
-LLM_TOKEN_COUNT_COMPLETION = ATTRIBUTE_PREFIX + semantic_conventions.LLM_TOKEN_COUNT_COMPLETION
+LLM_TOKEN_COUNT_TOTAL = ATTRIBUTE_PREFIX + SpanAttributes.LLM_TOKEN_COUNT_TOTAL
+LLM_TOKEN_COUNT_PROMPT = ATTRIBUTE_PREFIX + SpanAttributes.LLM_TOKEN_COUNT_PROMPT
+LLM_TOKEN_COUNT_COMPLETION = ATTRIBUTE_PREFIX + SpanAttributes.LLM_TOKEN_COUNT_COMPLETION
 
 
 class ReadableSpan(ObjectProxy):  # type: ignore
@@ -73,7 +72,9 @@ class ReadableSpan(ObjectProxy):  # type: ignore
     @property
     def span(self) -> Span:
         span = decode(self._self_otlp_span)
-        span.attributes.update(cast(SpanAttributes, self._self_computed_values))
+        span.attributes.update(
+            cast(phoenix.trace.schemas.SpanAttributes, self._self_computed_values)
+        )
         # TODO: compute latency rank percent (which can change depending on how
         # many spans already ingested).
         return span
@@ -333,9 +334,13 @@ class Traces:
             self._token_count_total -= existing_span[LLM_TOKEN_COUNT_TOTAL] or 0
         self._token_count_total += new_span[LLM_TOKEN_COUNT_TOTAL] or 0
         # Update number of documents
-        num_documents_update = len(new_span.attributes.get(RETRIEVAL_DOCUMENTS) or ())
+        num_documents_update = len(
+            new_span.attributes.get(SpanAttributes.RETRIEVAL_DOCUMENTS) or ()
+        )
         if existing_span:
-            num_documents_update -= len(existing_span.attributes.get(RETRIEVAL_DOCUMENTS) or ())
+            num_documents_update -= len(
+                existing_span.attributes.get(SpanAttributes.RETRIEVAL_DOCUMENTS) or ()
+            )
         if num_documents_update:
             self._num_documents[span_id] += num_documents_update
         # Process previously orphaned spans, if any.

phoenix/experimental/evals/functions/classify.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import logging
 import warnings
 from collections import defaultdict
+from itertools import product
 from typing import (
     Any,
     DefaultDict,
@@ -18,6 +19,7 @@ from typing import (
 )
 
 import pandas as pd
+from openinference.semconv.trace import DocumentAttributes, SpanAttributes
 from pandas import DataFrame
 from typing_extensions import TypeAlias
 
@@ -40,9 +42,12 @@ from phoenix.experimental.evals.utils import (
     parse_openai_function_call,
     snap_to_rail,
 )
-from phoenix.trace.semantic_conventions import DOCUMENT_CONTENT, INPUT_VALUE, RETRIEVAL_DOCUMENTS
 from phoenix.utilities.logging import printif
 
+DOCUMENT_CONTENT = DocumentAttributes.DOCUMENT_CONTENT
+INPUT_VALUE = SpanAttributes.INPUT_VALUE
+RETRIEVAL_DOCUMENTS = SpanAttributes.RETRIEVAL_DOCUMENTS
+
 logger = logging.getLogger(__name__)
 
 
@@ -54,8 +59,7 @@ Label: TypeAlias = str
 Score: TypeAlias = Optional[float]
 Explanation: TypeAlias = Optional[str]
 Record: TypeAlias = Mapping[str, Any]
-EvaluatorIndex: TypeAlias = int
-RowIndex: TypeAlias = Any
+Index: TypeAlias = int
 
 # snapped_response, explanation, response
 ParsedLLMResponse: TypeAlias = Tuple[Optional[str], Optional[str], str]
@@ -343,8 +347,6 @@ def _get_contents_from_openinference_documents(documents: Iterable[Any]) -> List
 
 
 class RunEvalsPayload(NamedTuple):
-    evaluator_index: EvaluatorIndex
-    row_index: RowIndex
     evaluator: LLMEvaluator
     record: Record
 
@@ -404,23 +406,21 @@ def run_evals(
 
     async def _arun_eval(
         payload: RunEvalsPayload,
-    ) -> Tuple[EvaluatorIndex, RowIndex, Label, Score, Explanation]:
-        label, score, explanation = await payload.evaluator.aevaluate(
+    ) -> Tuple[Label, Score, Explanation]:
+        return await payload.evaluator.aevaluate(
             payload.record,
             provide_explanation=provide_explanation,
             use_function_calling_if_available=use_function_calling_if_available,
         )
-        return payload.evaluator_index, payload.row_index, label, score, explanation
 
     def _run_eval(
         payload: RunEvalsPayload,
-    ) -> Tuple[EvaluatorIndex, RowIndex, Label, Score, Explanation]:
-        label, score, explanation = payload.evaluator.evaluate(
+    ) -> Tuple[Label, Score, Explanation]:
+        return payload.evaluator.evaluate(
             payload.record,
             provide_explanation=provide_explanation,
             use_function_calling_if_available=use_function_calling_if_available,
         )
-        return payload.evaluator_index, payload.row_index, label, score, explanation
 
     executor = get_executor_on_sync_context(
         _run_eval,
@@ -428,24 +428,20 @@ def run_evals(
         concurrency=concurrency,
         tqdm_bar_format=get_tqdm_progress_bar_formatter("run_evals"),
         exit_on_error=True,
-        fallback_return_value=(None, None),
+        fallback_return_value=(None, None, None),
     )
+
+    total_records = len(dataframe)
     payloads = [
-        RunEvalsPayload(
-            evaluator_index=evaluator_index,
-            row_index=row_index,
-            evaluator=evaluator,
-            record=row.to_dict(),
-        )
-        # use the position of the row rather than the dataframe index, which is used
-        # to ensure the output dataframe has the same row order as the input dataframe
-        for row_index, (_, row) in enumerate(dataframe.iterrows())
-        for evaluator_index, evaluator in enumerate(evaluators)
+        RunEvalsPayload(evaluator=evaluator, record=row)
+        for evaluator, (_, row) in product(evaluators, dataframe.iterrows())
     ]
-    eval_results: List[DefaultDict[RowIndex, Dict[ColumnName, Union[Label, Explanation]]]] = [
+    eval_results: List[DefaultDict[Index, Dict[ColumnName, Union[Label, Explanation]]]] = [
         defaultdict(dict) for _ in range(len(evaluators))
     ]
-    for evaluator_index, row_index, label, score, explanation in executor.run(payloads):
+    for index, (label, score, explanation) in enumerate(executor.run(payloads)):
+        evaluator_index = index // total_records
+        row_index = index % total_records
         eval_results[evaluator_index][row_index]["label"] = label
         eval_results[evaluator_index][row_index]["score"] = score
         if provide_explanation:

phoenix/experimental/evals/utils/__init__.py

@@ -32,7 +32,7 @@ def download_benchmark_dataset(task: str, dataset_name: str) -> pd.DataFrame:
         pandas.DataFrame: A pandas dataframe containing the data.
     """
     jsonl_file_name = f"{dataset_name}.jsonl"
-    url = f"http://storage.googleapis.com/arize-assets/phoenix/evals/{task}/{jsonl_file_name}.zip"
+    url = f"http://storage.googleapis.com/arize-phoenix-assets/evals/{task}/{jsonl_file_name}.zip"
     try:
         with urlopen(url) as response:
             zip_byte_stream = BytesIO(response.read())

phoenix/server/api/input_types/SpanSort.py

@@ -4,6 +4,7 @@ from typing import Any, Iterable, Iterator, Optional, Protocol
 
 import pandas as pd
 import strawberry
+from openinference.semconv.trace import SpanAttributes
 from strawberry import UNSET
 from typing_extensions import assert_never
 
@@ -13,7 +14,6 @@ from phoenix.core.traces import (
     START_TIME,
 )
 from phoenix.server.api.types.SortDir import SortDir
-from phoenix.trace import semantic_conventions
 from phoenix.trace.schemas import ComputedAttributes, Span, SpanID
 
 
@@ -22,9 +22,9 @@ class SpanColumn(Enum):
     startTime = START_TIME
     endTime = END_TIME
     latencyMs = ComputedAttributes.LATENCY_MS.value
-    tokenCountTotal = semantic_conventions.LLM_TOKEN_COUNT_TOTAL
-    tokenCountPrompt = semantic_conventions.LLM_TOKEN_COUNT_PROMPT
-    tokenCountCompletion = semantic_conventions.LLM_TOKEN_COUNT_COMPLETION
+    tokenCountTotal = SpanAttributes.LLM_TOKEN_COUNT_TOTAL
+    tokenCountPrompt = SpanAttributes.LLM_TOKEN_COUNT_PROMPT
+    tokenCountCompletion = SpanAttributes.LLM_TOKEN_COUNT_COMPLETION
     cumulativeTokenCountTotal = ComputedAttributes.CUMULATIVE_LLM_TOKEN_COUNT_TOTAL.value
     cumulativeTokenCountPrompt = ComputedAttributes.CUMULATIVE_LLM_TOKEN_COUNT_PROMPT.value
     cumulativeTokenCountCompletion = ComputedAttributes.CUMULATIVE_LLM_TOKEN_COUNT_COMPLETION.value

phoenix/server/api/types/Span.py

@@ -5,6 +5,7 @@ from enum import Enum
 from typing import Any, DefaultDict, Dict, List, Mapping, Optional, Sized, cast
 
 import strawberry
+from openinference.semconv.trace import EmbeddingAttributes, SpanAttributes
 from strawberry import ID, UNSET
 from strawberry.types import Info
 
@@ -15,19 +16,17 @@ from phoenix.server.api.types.DocumentRetrievalMetrics import DocumentRetrievalM
 from phoenix.server.api.types.Evaluation import DocumentEvaluation, SpanEvaluation
 from phoenix.server.api.types.MimeType import MimeType
 from phoenix.trace.schemas import ComputedAttributes, SpanID
-from phoenix.trace.semantic_conventions import (
-    EMBEDDING_EMBEDDINGS,
-    EMBEDDING_VECTOR,
-    EXCEPTION_MESSAGE,
-    INPUT_MIME_TYPE,
-    INPUT_VALUE,
-    LLM_TOKEN_COUNT_COMPLETION,
-    LLM_TOKEN_COUNT_PROMPT,
-    LLM_TOKEN_COUNT_TOTAL,
-    OUTPUT_MIME_TYPE,
-    OUTPUT_VALUE,
-    RETRIEVAL_DOCUMENTS,
-)
+
+EMBEDDING_EMBEDDINGS = SpanAttributes.EMBEDDING_EMBEDDINGS
+EMBEDDING_VECTOR = EmbeddingAttributes.EMBEDDING_VECTOR
+INPUT_MIME_TYPE = SpanAttributes.INPUT_MIME_TYPE
+INPUT_VALUE = SpanAttributes.INPUT_VALUE
+LLM_TOKEN_COUNT_COMPLETION = SpanAttributes.LLM_TOKEN_COUNT_COMPLETION
+LLM_TOKEN_COUNT_PROMPT = SpanAttributes.LLM_TOKEN_COUNT_PROMPT
+LLM_TOKEN_COUNT_TOTAL = SpanAttributes.LLM_TOKEN_COUNT_TOTAL
+OUTPUT_MIME_TYPE = SpanAttributes.OUTPUT_MIME_TYPE
+OUTPUT_VALUE = SpanAttributes.OUTPUT_VALUE
+RETRIEVAL_DOCUMENTS = SpanAttributes.RETRIEVAL_DOCUMENTS
 
 
 @strawberry.enum
@@ -87,7 +86,7 @@ class SpanEvent:
     ) -> "SpanEvent":
         return SpanEvent(
             name=event.name,
-            message=cast(str, event.attributes.get(EXCEPTION_MESSAGE) or ""),
+            message=cast(str, event.attributes.get(trace_schema.EXCEPTION_MESSAGE) or ""),
             timestamp=event.timestamp,
         )
 

phoenix/trace/dsl/filter.py

@@ -1,4 +1,5 @@
 import ast
+import inspect
 import sys
 from dataclasses import dataclass, field
 from difflib import SequenceMatcher
@@ -15,10 +16,10 @@ from typing import (
     cast,
 )
 
+from openinference.semconv import trace
 from typing_extensions import TypeGuard
 
 import phoenix.trace.v1 as pb
-from phoenix.trace import semantic_conventions
 from phoenix.trace.dsl.missing import MISSING
 from phoenix.trace.schemas import COMPUTED_PREFIX, ComputedAttributes, Span, SpanID
 
@@ -137,9 +138,11 @@ def _allowed_replacements() -> Iterator[Tuple[str, ast.expr]]:
         yield "span.context." + source_segment, ast_replacement
 
     for field_name in (
-        getattr(semantic_conventions, variable_name)
-        for variable_name in dir(semantic_conventions)
-        if variable_name.isupper()
+        getattr(klass, attr)
+        for name in dir(trace)
+        if name.endswith("Attributes") and inspect.isclass(klass := getattr(trace, name))
+        for attr in dir(klass)
+        if attr.isupper()
     ):
         source_segment = field_name
         ast_replacement = _ast_replacement(f"span.attributes.get('{field_name}')")

phoenix/trace/dsl/helpers.py

@@ -1,15 +1,15 @@
 from typing import List, Optional, Protocol, Union, cast
 
 import pandas as pd
+from openinference.semconv.trace import DocumentAttributes, SpanAttributes
 
 from phoenix.trace.dsl import SpanQuery
-from phoenix.trace.semantic_conventions import (
-    DOCUMENT_CONTENT,
-    DOCUMENT_SCORE,
-    INPUT_VALUE,
-    OUTPUT_VALUE,
-    RETRIEVAL_DOCUMENTS,
-)
+
+DOCUMENT_CONTENT = DocumentAttributes.DOCUMENT_CONTENT
+DOCUMENT_SCORE = DocumentAttributes.DOCUMENT_SCORE
+INPUT_VALUE = SpanAttributes.INPUT_VALUE
+OUTPUT_VALUE = SpanAttributes.OUTPUT_VALUE
+RETRIEVAL_DOCUMENTS = SpanAttributes.RETRIEVAL_DOCUMENTS
 
 INPUT = {"input": INPUT_VALUE}
 OUTPUT = {"output": OUTPUT_VALUE}

phoenix/trace/dsl/query.py

@@ -19,13 +19,15 @@ from typing import (
 )
 
 import pandas as pd
+from openinference.semconv.trace import SpanAttributes
 
 from phoenix.trace.dsl import SpanFilter
 from phoenix.trace.dsl.filter import SupportsGetSpanEvaluation
 from phoenix.trace.schemas import ATTRIBUTE_PREFIX, CONTEXT_PREFIX, Span
-from phoenix.trace.semantic_conventions import RETRIEVAL_DOCUMENTS
 from phoenix.trace.span_json_encoder import span_to_json
 
+RETRIEVAL_DOCUMENTS = SpanAttributes.RETRIEVAL_DOCUMENTS
+
 _SPAN_ID = "context.span_id"
 _PRESCRIBED_POSITION_PREFIXES = {
     RETRIEVAL_DOCUMENTS: "document_",

phoenix/trace/errors.py

@@ -3,3 +3,7 @@ from phoenix.exceptions import PhoenixException
 
 class InvalidParquetMetadataError(PhoenixException):
     pass
+
+
+class IncompatibleLibraryVersionError(PhoenixException):
+    pass

phoenix/trace/llama_index/__init__.py

@@ -1,4 +1,3 @@
 from .callback import OpenInferenceTraceCallbackHandler
-from .debug_callback import DebugCallbackHandler
 
-__all__ = ["OpenInferenceTraceCallbackHandler", "DebugCallbackHandler"]
+__all__ = ["OpenInferenceTraceCallbackHandler"]

phoenix/trace/llama_index/callback.py

@@ -1,25 +1,64 @@
 import logging
-from importlib.metadata import PackageNotFoundError
+from importlib.metadata import PackageNotFoundError, version
 from importlib.util import find_spec
-from typing import (
-    Any,
-)
-
-from openinference.instrumentation.llama_index._callback import (
-    OpenInferenceTraceCallbackHandler as _OpenInferenceTraceCallbackHandler,
-)
-from openinference.instrumentation.llama_index.version import (
-    __version__,
-)
+from typing import Any
+
 from opentelemetry import trace as trace_api
 from opentelemetry.sdk import trace as trace_sdk
 from opentelemetry.sdk.trace.export import SimpleSpanProcessor
 
+from phoenix.trace.errors import IncompatibleLibraryVersionError
 from phoenix.trace.exporter import _OpenInferenceExporter
 from phoenix.trace.tracer import _show_deprecation_warnings
 
 logger = logging.getLogger(__name__)
 
+LLAMA_INDEX_MODERN_VERSION = (0, 10, 0)
+INSTRUMENTATION_MODERN_VERSION = (1, 0, 0)
+
+
+def _check_instrumentation_compatibility() -> bool:
+    if find_spec("llama_index") is None:
+        raise PackageNotFoundError("Missing `llama-index`. Install with `pip install llama-index`.")
+    # split the version string into a tuple of integers
+    llama_index_version_str = version("llama-index")
+    llama_index_version = tuple(map(int, llama_index_version_str.split(".")[:3]))
+    instrumentation_version_str = version("openinference-instrumentation-llama-index")
+    instrumentation_version = tuple(map(int, instrumentation_version_str.split(".")[:3]))
+    # check if the llama_index version is compatible with the instrumentation version
+    if (
+        llama_index_version < LLAMA_INDEX_MODERN_VERSION
+        and instrumentation_version >= INSTRUMENTATION_MODERN_VERSION
+    ):
+        raise IncompatibleLibraryVersionError(
+            f"llama-index v{llama_index_version_str} is not compatible with "
+            f"openinference-instrumentation-llama-index v{instrumentation_version_str}."
+            "Please either migrate llama-index to at least 0.10.0 or downgrade "
+            "openinference-instrumentation-llama-index via "
+            "`pip install 'openinference-instrumentation-llama-index<1.0.0'`."
+        )
+    elif (
+        llama_index_version >= LLAMA_INDEX_MODERN_VERSION
+        and instrumentation_version < INSTRUMENTATION_MODERN_VERSION
+    ):
+        raise IncompatibleLibraryVersionError(
+            f"llama-index v{llama_index_version_str} is not compatible with "
+            f"openinference-instrumentation-llama-index v{instrumentation_version_str}."
+            "Please upgrade openinference-instrumentation-llama-index to at least 1.0.0"
+            "`pip install 'openinference-instrumentation-llama-index>=1.0.0'`."
+        )
+    # if the versions are compatible, return True
+    return True
+
+
+if _check_instrumentation_compatibility():
+    from openinference.instrumentation.llama_index._callback import (
+        OpenInferenceTraceCallbackHandler as _OpenInferenceTraceCallbackHandler,
+    )
+    from openinference.instrumentation.llama_index.version import (
+        __version__,
+    )
+
 
 class OpenInferenceTraceCallbackHandler(_OpenInferenceTraceCallbackHandler):
     """Callback handler for storing LLM application trace data in OpenInference format.
@@ -33,10 +72,6 @@ class OpenInferenceTraceCallbackHandler(_OpenInferenceTraceCallbackHandler):
 
     def __init__(self, *args: Any, **kwargs: Any) -> None:
         _show_deprecation_warnings(self, *args, **kwargs)
-        if find_spec("llama_index") is None:
-            raise PackageNotFoundError(
-                "Missing `llama-index`. Install with `pip install llama-index`."
-            )
         tracer_provider = trace_sdk.TracerProvider()
         tracer_provider.add_span_processor(SimpleSpanProcessor(_OpenInferenceExporter()))
         super().__init__(trace_api.get_tracer(__name__, __version__, tracer_provider))

phoenix/trace/otel.py

@@ -1,3 +1,4 @@
+import inspect
 import json
 from binascii import hexlify, unhexlify
 from datetime import datetime, timezone
@@ -21,12 +22,23 @@ from typing import (
 
 import numpy as np
 import opentelemetry.proto.trace.v1.trace_pb2 as otlp
+from openinference.semconv import trace
+from openinference.semconv.trace import (
+    DocumentAttributes,
+    EmbeddingAttributes,
+    MessageAttributes,
+    SpanAttributes,
+    ToolCallAttributes,
+)
 from opentelemetry.proto.common.v1.common_pb2 import AnyValue, ArrayValue, KeyValue
 from opentelemetry.util.types import Attributes, AttributeValue
 from typing_extensions import TypeAlias, assert_never
 
-import phoenix.trace.semantic_conventions as sem_conv
 from phoenix.trace.schemas import (
+    EXCEPTION_ESCAPED,
+    EXCEPTION_MESSAGE,
+    EXCEPTION_STACKTRACE,
+    EXCEPTION_TYPE,
     MimeType,
     Span,
     SpanContext,
@@ -37,18 +49,38 @@ from phoenix.trace.schemas import (
     SpanStatusCode,
     TraceID,
 )
-from phoenix.trace.semantic_conventions import (
-    DOCUMENT_METADATA,
-    EXCEPTION_ESCAPED,
-    EXCEPTION_MESSAGE,
-    EXCEPTION_STACKTRACE,
-    EXCEPTION_TYPE,
-    INPUT_MIME_TYPE,
-    LLM_PROMPT_TEMPLATE_VARIABLES,
-    OPENINFERENCE_SPAN_KIND,
-    OUTPUT_MIME_TYPE,
-    TOOL_PARAMETERS,
-)
+
+DOCUMENT_CONTENT = DocumentAttributes.DOCUMENT_CONTENT
+DOCUMENT_ID = DocumentAttributes.DOCUMENT_ID
+DOCUMENT_METADATA = DocumentAttributes.DOCUMENT_METADATA
+EMBEDDING_EMBEDDINGS = SpanAttributes.EMBEDDING_EMBEDDINGS
+EMBEDDING_MODEL_NAME = SpanAttributes.EMBEDDING_MODEL_NAME
+EMBEDDING_TEXT = EmbeddingAttributes.EMBEDDING_TEXT
+EMBEDDING_VECTOR = EmbeddingAttributes.EMBEDDING_VECTOR
+INPUT_MIME_TYPE = SpanAttributes.INPUT_MIME_TYPE
+INPUT_VALUE = SpanAttributes.INPUT_VALUE
+LLM_INPUT_MESSAGES = SpanAttributes.LLM_INPUT_MESSAGES
+LLM_INVOCATION_PARAMETERS = SpanAttributes.LLM_INVOCATION_PARAMETERS
+LLM_MODEL_NAME = SpanAttributes.LLM_MODEL_NAME
+LLM_OUTPUT_MESSAGES = SpanAttributes.LLM_OUTPUT_MESSAGES
+LLM_PROMPTS = SpanAttributes.LLM_PROMPTS
+LLM_TOKEN_COUNT_COMPLETION = SpanAttributes.LLM_TOKEN_COUNT_COMPLETION
+LLM_TOKEN_COUNT_PROMPT = SpanAttributes.LLM_TOKEN_COUNT_PROMPT
+LLM_TOKEN_COUNT_TOTAL = SpanAttributes.LLM_TOKEN_COUNT_TOTAL
+MESSAGE_CONTENT = MessageAttributes.MESSAGE_CONTENT
+MESSAGE_FUNCTION_CALL_ARGUMENTS_JSON = MessageAttributes.MESSAGE_FUNCTION_CALL_ARGUMENTS_JSON
+MESSAGE_FUNCTION_CALL_NAME = MessageAttributes.MESSAGE_FUNCTION_CALL_NAME
+MESSAGE_ROLE = MessageAttributes.MESSAGE_ROLE
+MESSAGE_TOOL_CALLS = MessageAttributes.MESSAGE_TOOL_CALLS
+OPENINFERENCE_SPAN_KIND = SpanAttributes.OPENINFERENCE_SPAN_KIND
+OUTPUT_MIME_TYPE = SpanAttributes.OUTPUT_MIME_TYPE
+OUTPUT_VALUE = SpanAttributes.OUTPUT_VALUE
+RETRIEVAL_DOCUMENTS = SpanAttributes.RETRIEVAL_DOCUMENTS
+TOOL_CALL_FUNCTION_ARGUMENTS_JSON = ToolCallAttributes.TOOL_CALL_FUNCTION_ARGUMENTS_JSON
+TOOL_CALL_FUNCTION_NAME = ToolCallAttributes.TOOL_CALL_FUNCTION_NAME
+TOOL_PARAMETERS = SpanAttributes.TOOL_PARAMETERS
+LLM_PROMPT_TEMPLATE = SpanAttributes.LLM_PROMPT_TEMPLATE
+LLM_PROMPT_TEMPLATE_VARIABLES = SpanAttributes.LLM_PROMPT_TEMPLATE_VARIABLES
 
 
 def decode(otlp_span: otlp.Span) -> Span:
@@ -186,7 +218,13 @@ def _decode_status(otlp_status: otlp.Status) -> Tuple[SpanStatusCode, StatusMess
 
 
 _SEMANTIC_CONVENTIONS: List[str] = sorted(
-    (getattr(sem_conv, name) for name in dir(sem_conv) if name.isupper()),
+    (
+        getattr(klass, attr)
+        for name in dir(trace)
+        if name.endswith("Attributes") and inspect.isclass(klass := getattr(trace, name))
+        for attr in dir(klass)
+        if attr.isupper()
+    ),
     reverse=True,
 )  # sorted so the longer strings go first
 

phoenix/trace/schemas.py

@@ -4,12 +4,10 @@ from enum import Enum
 from typing import Any, Dict, List, Optional, Union
 from uuid import UUID
 
-from phoenix.trace.semantic_conventions import (
-    EXCEPTION_ESCAPED,
-    EXCEPTION_MESSAGE,
-    EXCEPTION_STACKTRACE,
-    EXCEPTION_TYPE,
-)
+EXCEPTION_TYPE = "exception.type"
+EXCEPTION_MESSAGE = "exception.message"
+EXCEPTION_ESCAPED = "exception.escaped"
+EXCEPTION_STACKTRACE = "exception.stacktrace"
 
 
 class SpanStatusCode(Enum):

phoenix/trace/span_json_decoder.py

@@ -2,7 +2,10 @@ import json
 from datetime import datetime
 from typing import Any, Dict, Optional
 
+from openinference.semconv.trace import SpanAttributes
+
 from phoenix.trace.schemas import (
+    EXCEPTION_MESSAGE,
     MimeType,
     Span,
     SpanContext,
@@ -14,11 +17,9 @@ from phoenix.trace.schemas import (
     SpanStatusCode,
     TraceID,
 )
-from phoenix.trace.semantic_conventions import (
-    EXCEPTION_MESSAGE,
-    INPUT_MIME_TYPE,
-    OUTPUT_MIME_TYPE,
-)
+
+INPUT_MIME_TYPE = SpanAttributes.INPUT_MIME_TYPE
+OUTPUT_MIME_TYPE = SpanAttributes.OUTPUT_MIME_TYPE
 
 
 def json_to_attributes(obj: Optional[Dict[str, Any]]) -> Dict[str, Any]:

phoenix/trace/span_json_encoder.py

@@ -5,12 +5,7 @@ from enum import Enum
 from typing import Any, List
 from uuid import UUID
 
-from .schemas import (
-    Span,
-    SpanContext,
-    SpanConversationAttributes,
-    SpanEvent,
-)
+from phoenix.trace.schemas import Span, SpanContext, SpanConversationAttributes, SpanEvent
 
 
 class SpanJSONEncoder(json.JSONEncoder):

phoenix/trace/trace_dataset.py

@@ -6,23 +6,26 @@ from uuid import UUID, uuid4
 from warnings import warn
 
 import pandas as pd
+from openinference.semconv.trace import (
+    DocumentAttributes,
+    RerankerAttributes,
+    SpanAttributes,
+)
 from pandas import DataFrame, read_parquet
 from pyarrow import Schema, Table, parquet
 
+from phoenix.config import DATASET_DIR, GENERATED_DATASET_NAME_PREFIX, TRACE_DATASET_DIR
 from phoenix.datetime_utils import normalize_timestamps
 from phoenix.trace.errors import InvalidParquetMetadataError
-
-from ..config import DATASET_DIR, GENERATED_DATASET_NAME_PREFIX, TRACE_DATASET_DIR
-from .schemas import ATTRIBUTE_PREFIX, CONTEXT_PREFIX, Span
-from .semantic_conventions import (
-    DOCUMENT_METADATA,
-    RERANKER_INPUT_DOCUMENTS,
-    RERANKER_OUTPUT_DOCUMENTS,
-    RETRIEVAL_DOCUMENTS,
-)
-from .span_evaluations import Evaluations, SpanEvaluations
-from .span_json_decoder import json_to_span
-from .span_json_encoder import span_to_json
+from phoenix.trace.schemas import ATTRIBUTE_PREFIX, CONTEXT_PREFIX, Span
+from phoenix.trace.span_evaluations import Evaluations, SpanEvaluations
+from phoenix.trace.span_json_decoder import json_to_span
+from phoenix.trace.span_json_encoder import span_to_json
+
+DOCUMENT_METADATA = DocumentAttributes.DOCUMENT_METADATA
+RERANKER_INPUT_DOCUMENTS = RerankerAttributes.RERANKER_INPUT_DOCUMENTS
+RERANKER_OUTPUT_DOCUMENTS = RerankerAttributes.RERANKER_OUTPUT_DOCUMENTS
+RETRIEVAL_DOCUMENTS = SpanAttributes.RETRIEVAL_DOCUMENTS
 
 # A set of columns that is required
 REQUIRED_COLUMNS = [

phoenix/version.py

@@ -1 +1 @@
-__version__ = "3.0.1"
+__version__ = "3.0.3"

phoenix/trace/llama_index/debug_callback.py

@@ -1,50 +0,0 @@
-import logging
-from typing import Any, Dict, List, Optional
-
-from llama_index.callbacks.base_handler import BaseCallbackHandler
-from llama_index.callbacks.schema import CBEventType
-
-logger = logging.getLogger(__name__)
-logger.addHandler(logging.NullHandler())
-
-CBEventID = str
-
-
-class DebugCallbackHandler(BaseCallbackHandler):
-    def _print_event(self, payload: Dict[Any, Any]) -> None:
-        for k, v in payload.items():
-            print(f"**{k}: **\n{v}")
-            print("*" * 50)
-
-    def __init__(self) -> None:
-        super().__init__(event_starts_to_ignore=[], event_ends_to_ignore=[])
-
-    def on_event_start(
-        self,
-        event_type: CBEventType,
-        payload: Optional[Dict[str, Any]] = None,
-        event_id: CBEventID = "",
-        parent_id: CBEventID = "",
-        **kwargs: Any,
-    ) -> CBEventID:
-        return event_id
-
-    def on_event_end(
-        self,
-        event_type: CBEventType,
-        payload: Optional[Dict[str, Any]] = None,
-        event_id: CBEventID = "",
-        **kwargs: Any,
-    ) -> None:
-        if payload is not None:
-            self._print_event(payload)
-
-    def start_trace(self, trace_id: Optional[str] = None) -> None:
-        return
-
-    def end_trace(
-        self,
-        trace_id: Optional[str] = None,
-        trace_map: Optional[Dict[str, List[str]]] = None,
-    ) -> None:
-        return

phoenix/trace/semantic_conventions.py

@@ -1,172 +0,0 @@
-"""
-Semantic conventions for the attributes of a span
-https://github.com/Arize-ai/open-inference-spec/blob/main/trace/spec/semantic_conventions.md
-"""
-
-EXCEPTION_TYPE = "exception.type"
-EXCEPTION_MESSAGE = "exception.message"
-EXCEPTION_ESCAPED = "exception.escaped"
-EXCEPTION_STACKTRACE = "exception.stacktrace"
-
-
-OUTPUT_VALUE = "output.value"
-OUTPUT_MIME_TYPE = "output.mime_type"
-"""
-The type of output.value. If unspecified, the type is plain text by default.
-If type is JSON, the value is a string representing a JSON object.
-"""
-INPUT_VALUE = "input.value"
-INPUT_MIME_TYPE = "input.mime_type"
-"""
-The type of input.value. If unspecified, the type is plain text by default.
-If type is JSON, the value is a string representing a JSON object.
-"""
-
-EMBEDDING_EMBEDDINGS = "embedding.embeddings"
-"""
-A list of objects containing embedding data, including the vector and represented piece of text.
-"""
-EMBEDDING_MODEL_NAME = "embedding.model_name"
-"""
-The name of the embedding model.
-"""
-EMBEDDING_TEXT = "embedding.text"
-"""
-The text represented by the embedding.
-"""
-EMBEDDING_VECTOR = "embedding.vector"
-"""
-The embedding vector.
-"""
-
-MESSAGE_ROLE = "message.role"
-"""
-The role of the message, such as "user", "agent", "function".
-"""
-MESSAGE_NAME = "message.name"
-"""
-The name of the message, often used to identify the function
-that was used to generate the message.
-"""
-MESSAGE_TOOL_CALLS = "message.tool_calls"
-"""
-The tool calls generated by the model, such as function calls.
-"""
-TOOL_CALL_FUNCTION_NAME = "tool_call.function.name"
-"""
-The name of function that is being called during a tool call.
-"""
-TOOL_CALL_FUNCTION_ARGUMENTS_JSON = "tool_call.function.arguments"
-"""
-The JSON string representing the arguments passed to the function
-during a tool call.
-"""
-MESSAGE_FUNCTION_CALL_NAME = "message.function_call_name"
-"""
-The function name that is a part of the message list.
-This is populated for role 'function' or 'agent' as a mechanism to identify
-the function that was called during the execution of a tool
-"""
-MESSAGE_FUNCTION_CALL_ARGUMENTS_JSON = "message.function_call_arguments_json"
-"""
-The JSON string representing the arguments passed to the function
-during a function call
-"""
-MESSAGE_CONTENT = "message.content"
-"""
-The content of the message to the llm
-"""
-LLM_FUNCTION_CALL = "llm.function_call"
-"""
-For models and APIs that support function calling. Records attributes such as the function name and
-arguments to the called function.
-"""
-LLM_INVOCATION_PARAMETERS = "llm.invocation_parameters"
-"""
-Invocation parameters passed to the LLM or API, such as the model name, temperature, etc.
-"""
-LLM_INPUT_MESSAGES = "llm.input_messages"
-"""
-Messages provided to a chat API.
-"""
-LLM_OUTPUT_MESSAGES = "llm.output_messages"
-"""
-Messages received from a chat API.
-"""
-LLM_MODEL_NAME = "llm.model_name"
-"""
-The name of the model being used.
-"""
-LLM_PROMPTS = "llm.prompts"
-"""
-Prompts provided to a completions API.
-"""
-LLM_PROMPT_TEMPLATE = "llm.prompt_template.template"
-"""
-The prompt template as a Python f-string.
-"""
-LLM_PROMPT_TEMPLATE_VARIABLES = "llm.prompt_template.variables"
-"""
-A list of input variables to the prompt template.
-"""
-LLM_PROMPT_TEMPLATE_VERSION = "llm.prompt_template.version"
-"""
-The version of the prompt template being used.
-"""
-LLM_TOKEN_COUNT_PROMPT = "llm.token_count.prompt"
-"""
-Number of tokens in the prompt.
-"""
-LLM_TOKEN_COUNT_COMPLETION = "llm.token_count.completion"
-"""
-Number of tokens in the completion.
-"""
-LLM_TOKEN_COUNT_TOTAL = "llm.token_count.total"
-"""
-Total number of tokens, including both prompt and completion.
-"""
-
-TOOL_NAME = "tool.name"
-"""
-Name of the tool being used.
-"""
-TOOL_DESCRIPTION = "tool.description"
-"""
-Description of the tool's purpose, typically used to select the tool.
-"""
-TOOL_PARAMETERS = "tool.parameters"
-"""
-Parameters of the tool, e.g. see https://platform.openai.com/docs/guides/gpt/function-calling
-"""
-
-RETRIEVAL_DOCUMENTS = "retrieval.documents"
-DOCUMENT_ID = "document.id"
-DOCUMENT_SCORE = "document.score"
-DOCUMENT_CONTENT = "document.content"
-DOCUMENT_METADATA = "document.metadata"
-"""
-Document metadata as a string representing a JSON object
-"""
-
-RERANKER_INPUT_DOCUMENTS = "reranker.input_documents"
-"""
-List of documents as input to the reranker
-"""
-RERANKER_OUTPUT_DOCUMENTS = "reranker.output_documents"
-"""
-List of documents as output from the reranker
-"""
-RERANKER_QUERY = "reranker.query"
-"""
-Query string for the reranker
-"""
-RERANKER_MODEL_NAME = "reranker.model_name"
-"""
-Model name of the reranker
-"""
-RERANKER_TOP_K = "reranker.top_k"
-"""
-Top K parameter of the reranker
-"""
-
-OPENINFERENCE_SPAN_KIND = "openinference.span.kind"