arize-phoenix 3.4.1__tar.gz → 3.5.0__tar.gz

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: arize-phoenix
-Version: 3.4.1
+Version: 3.5.0
 Summary: ML Observability in your notebook
 Project-URL: Documentation, https://docs.arize.com/phoenix/
 Project-URL: Issues, https://github.com/Arize-ai/phoenix/issues
@@ -59,7 +59,7 @@ Requires-Dist: pytest-asyncio; extra == 'dev'
 Requires-Dist: pytest-cov; extra == 'dev'
 Requires-Dist: pytest-lazy-fixture; extra == 'dev'
 Requires-Dist: pytest==7.4.4; extra == 'dev'
-Requires-Dist: ruff==0.1.5; extra == 'dev'
+Requires-Dist: ruff==0.3.0; extra == 'dev'
 Requires-Dist: strawberry-graphql[debug-server]==0.208.2; extra == 'dev'
 Provides-Extra: evals
 Requires-Dist: arize-phoenix-evals>=0.0.3; extra == 'evals'
@@ -134,14 +134,11 @@ Phoenix provides MLOps and LLMOps insights at lightning speed with zero-config o
 Install Phoenix via `pip` or or `conda` as well as any of its subpackages.
 
 ```shell
-pip install arize-phoenix
+pip install arize-phoenix[evals]
 ```
 
-Some functionality such as LLM evals are under the `experimental` subpackage.
-
-```shell
-pip install arize-phoenix[experimental]
-```
+> [!NOTE]
+> The above will install Phoenix and its `evals` subpackage. To just install phoenix's evaluation package, you can run `pip install arize-phoenix-evals` instead.
 
 ## LLM Traces
 
@@ -159,39 +156,44 @@ To extract traces from your LlamaIndex application, you will have to add Phoenix
 
 ```shell
 # Install phoenix as well as llama_index and your LLM of choice
-pip install arize-phoenix llama-index openai
-
+pip install "arize-phoenix[evals]" "openai>=1" "llama-index>=0.10.3" "openinference-instrumentation-llama-index>=1.0.0" "llama-index-callbacks-arize-phoenix>=0.1.2" llama-index-llms-openai
 ```
 
 Launch Phoenix in a notebook and view the traces of your LlamaIndex application in the Phoenix UI.
 
 ```python
+import os
 import phoenix as px
+from llama_index.core import (
+    Settings,
+    VectorStoreIndex,
+    SimpleDirectoryReader,
+    set_global_handler,
+)
+from llama_index.embeddings.openai import OpenAIEmbedding
+from llama_index.llms.openai import OpenAI
+
+os.environ["OPENAI_API_KEY"] = "YOUR_OPENAI_API_KEY"
 
 # To view traces in Phoenix, you will first have to start a Phoenix server. You can do this by running the following:
 session = px.launch_app()
 
 
-# Once you have started a Phoenix server, you can start your LlamaIndex application with the `OpenInferenceTraceCallback` as a callback. To do this, you will have to add the callback to the initialization of your LlamaIndex application:
+# Once you have started a Phoenix server, you can start your LlamaIndex application and configure it to send traces to Phoenix. To do this, you will have to add configure Phoenix as the global handler
 
-from phoenix.trace.llama_index import (
-    OpenInferenceTraceCallbackHandler,
-)
+set_global_handler("arize_phoenix")
 
-# Initialize the callback handler
-callback_handler = OpenInferenceTraceCallbackHandler()
 
 # LlamaIndex application initialization may vary
 # depending on your application
-service_context = ServiceContext.from_defaults(
-    llm_predictor=LLMPredictor(llm=ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0)),
-    embed_model=OpenAIEmbedding(model="text-embedding-ada-002"),
-    callback_manager=CallbackManager(handlers=[callback_handler]),
-)
-index = load_index_from_storage(
-    storage_context,
-    service_context=service_context,
-)
+Settings.llm = OpenAI(model="gpt-4-turbo-preview")
+Settings.embed_model = OpenAIEmbedding(model="text-embedding-ada-002")
+
+
+# Load your data and create an index. Note you usually want to store your index in a persistent store like a database or the file system
+documents = SimpleDirectoryReader("YOUR_DATA_DIRECTORY").load_data()
+index = VectorStoreIndex.from_documents(documents)
+
 query_engine = index.as_query_engine()
 
 # Query your LlamaIndex application
@@ -267,8 +269,6 @@ session.url
 
 ## LLM Evals
 
-🚧 LLM Evals is still under construction under a sub-module `arize-phoenix[experimental]`
-
 [![Open in Colab](https://img.shields.io/static/v1?message=Open%20in%20Colab&logo=googlecolab&labelColor=grey&color=blue&logoColor=orange&label=%20)](https://colab.research.google.com/github/Arize-ai/phoenix/blob/main/tutorials/evals/evaluate_relevance_classifications.ipynb) [![Open in GitHub](https://img.shields.io/static/v1?message=Open%20in%20GitHub&logo=github&labelColor=grey&color=blue&logoColor=white&label=%20)](https://github.com/Arize-ai/phoenix/blob/main/tutorials/evals/evaluate_relevance_classifications.ipynb)
 
 Phoenix provides tooling to evaluate LLM applications, including tools to determine the relevance or irrelevance of documents retrieved by retrieval-augmented generation (RAG) application, whether or not the response is toxic, and much more.
@@ -283,12 +283,12 @@ Phoenix's approach to LLM evals is notable for the following reasons:
 Here is an example of running the RAG relevance eval on a dataset of Wikipedia questions and answers:
 
 ```shell
-# Install phoenix as well as the experimental subpackage
-pip install arize-phoenix[experimental] ipython matplotlib openai pycm scikit-learn
+# Install phoenix as well as the evals subpackage
+pip install 'arize-phoenix[evals]' ipython matplotlib openai pycm scikit-learn
 ```
 
 ```python
-from phoenix.experimental.evals import (
+from phoenix.evals import (
     RAG_RELEVANCY_PROMPT_TEMPLATE,
     RAG_RELEVANCY_PROMPT_RAILS_MAP,
     OpenAIModel,
@@ -324,7 +324,7 @@ y_pred = df["eval_relevance"]
 precision, recall, f1, support = precision_recall_fscore_support(y_true, y_pred)
 ```
 
-To learn more about LLM Evals, see the [LLM Evals documentation](https://docs.arize.com/phoenix/concepts/llm-evals/).
+To learn more about LLM Evals, see the [Evals documentation](https://docs.arize.com/phoenix/concepts/llm-evals/).
 
 ## Embedding Analysis
 

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/README.md

@@ -61,14 +61,11 @@ Phoenix provides MLOps and LLMOps insights at lightning speed with zero-config o
 Install Phoenix via `pip` or or `conda` as well as any of its subpackages.
 
 ```shell
-pip install arize-phoenix
+pip install arize-phoenix[evals]
 ```
 
-Some functionality such as LLM evals are under the `experimental` subpackage.
-
-```shell
-pip install arize-phoenix[experimental]
-```
+> [!NOTE]
+> The above will install Phoenix and its `evals` subpackage. To just install phoenix's evaluation package, you can run `pip install arize-phoenix-evals` instead.
 
 ## LLM Traces
 
@@ -86,39 +83,44 @@ To extract traces from your LlamaIndex application, you will have to add Phoenix
 
 ```shell
 # Install phoenix as well as llama_index and your LLM of choice
-pip install arize-phoenix llama-index openai
-
+pip install "arize-phoenix[evals]" "openai>=1" "llama-index>=0.10.3" "openinference-instrumentation-llama-index>=1.0.0" "llama-index-callbacks-arize-phoenix>=0.1.2" llama-index-llms-openai
 ```
 
 Launch Phoenix in a notebook and view the traces of your LlamaIndex application in the Phoenix UI.
 
 ```python
+import os
 import phoenix as px
+from llama_index.core import (
+    Settings,
+    VectorStoreIndex,
+    SimpleDirectoryReader,
+    set_global_handler,
+)
+from llama_index.embeddings.openai import OpenAIEmbedding
+from llama_index.llms.openai import OpenAI
+
+os.environ["OPENAI_API_KEY"] = "YOUR_OPENAI_API_KEY"
 
 # To view traces in Phoenix, you will first have to start a Phoenix server. You can do this by running the following:
 session = px.launch_app()
 
 
-# Once you have started a Phoenix server, you can start your LlamaIndex application with the `OpenInferenceTraceCallback` as a callback. To do this, you will have to add the callback to the initialization of your LlamaIndex application:
+# Once you have started a Phoenix server, you can start your LlamaIndex application and configure it to send traces to Phoenix. To do this, you will have to add configure Phoenix as the global handler
 
-from phoenix.trace.llama_index import (
-    OpenInferenceTraceCallbackHandler,
-)
+set_global_handler("arize_phoenix")
 
-# Initialize the callback handler
-callback_handler = OpenInferenceTraceCallbackHandler()
 
 # LlamaIndex application initialization may vary
 # depending on your application
-service_context = ServiceContext.from_defaults(
-    llm_predictor=LLMPredictor(llm=ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0)),
-    embed_model=OpenAIEmbedding(model="text-embedding-ada-002"),
-    callback_manager=CallbackManager(handlers=[callback_handler]),
-)
-index = load_index_from_storage(
-    storage_context,
-    service_context=service_context,
-)
+Settings.llm = OpenAI(model="gpt-4-turbo-preview")
+Settings.embed_model = OpenAIEmbedding(model="text-embedding-ada-002")
+
+
+# Load your data and create an index. Note you usually want to store your index in a persistent store like a database or the file system
+documents = SimpleDirectoryReader("YOUR_DATA_DIRECTORY").load_data()
+index = VectorStoreIndex.from_documents(documents)
+
 query_engine = index.as_query_engine()
 
 # Query your LlamaIndex application
@@ -194,8 +196,6 @@ session.url
 
 ## LLM Evals
 
-🚧 LLM Evals is still under construction under a sub-module `arize-phoenix[experimental]`
-
 [![Open in Colab](https://img.shields.io/static/v1?message=Open%20in%20Colab&logo=googlecolab&labelColor=grey&color=blue&logoColor=orange&label=%20)](https://colab.research.google.com/github/Arize-ai/phoenix/blob/main/tutorials/evals/evaluate_relevance_classifications.ipynb) [![Open in GitHub](https://img.shields.io/static/v1?message=Open%20in%20GitHub&logo=github&labelColor=grey&color=blue&logoColor=white&label=%20)](https://github.com/Arize-ai/phoenix/blob/main/tutorials/evals/evaluate_relevance_classifications.ipynb)
 
 Phoenix provides tooling to evaluate LLM applications, including tools to determine the relevance or irrelevance of documents retrieved by retrieval-augmented generation (RAG) application, whether or not the response is toxic, and much more.
@@ -210,12 +210,12 @@ Phoenix's approach to LLM evals is notable for the following reasons:
 Here is an example of running the RAG relevance eval on a dataset of Wikipedia questions and answers:
 
 ```shell
-# Install phoenix as well as the experimental subpackage
-pip install arize-phoenix[experimental] ipython matplotlib openai pycm scikit-learn
+# Install phoenix as well as the evals subpackage
+pip install 'arize-phoenix[evals]' ipython matplotlib openai pycm scikit-learn
 ```
 
 ```python
-from phoenix.experimental.evals import (
+from phoenix.evals import (
     RAG_RELEVANCY_PROMPT_TEMPLATE,
     RAG_RELEVANCY_PROMPT_RAILS_MAP,
     OpenAIModel,
@@ -251,7 +251,7 @@ y_pred = df["eval_relevance"]
 precision, recall, f1, support = precision_recall_fscore_support(y_true, y_pred)
 ```
 
-To learn more about LLM Evals, see the [LLM Evals documentation](https://docs.arize.com/phoenix/concepts/llm-evals/).
+To learn more about LLM Evals, see the [Evals documentation](https://docs.arize.com/phoenix/concepts/llm-evals/).
 
 ## Embedding Analysis
 

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/pyproject.toml

@@ -56,7 +56,7 @@ dev = [
   "hatch",
   "jupyter",
   "nbqa",
-  "ruff==0.1.5",
+  "ruff==0.3.0",
   "pandas-stubs<=2.0.2.230605",  # version 2.0.3.230814 is causing a dependency conflict.
   "pytest==7.4.4",
   "pytest-asyncio",
@@ -115,14 +115,9 @@ dependencies = [
   "pytest-cov",
   "pytest-lazy-fixture",
   "arize",
-  "langchain>=0.0.334",
   "litellm>=1.0.3",
-  "llama-index>=0.10.3",
   "openai>=1.0.0",
   "tenacity",
-  "nltk==3.8.1",
-  "sentence-transformers==2.2.2",
-  "pydantic<2",  # for @root_validator in llama-index
   "requests",
   "protobuf==3.20",  # version minimum (for tests)
   "responses",
@@ -136,13 +131,13 @@ dependencies = [
 [tool.hatch.envs.type]
 dependencies = [
   "mypy==1.5.1",
-  "pydantic==v1.10.14",  # for mypy
-  "llama-index<0.10.0",
+  "tenacity",
   "pandas-stubs<=2.0.2.230605",  # version 2.0.3.230814 is causing a dependency conflict.
   "types-psutil",
   "types-tqdm",
   "types-requests",
   "types-protobuf",
+  "types-setuptools",
   "openai>=1.0.0",
   "litellm>=1.0.3",
 ]
@@ -150,7 +145,7 @@ dependencies = [
 [tool.hatch.envs.style]
 detached = true
 dependencies = [
-  "ruff~=0.1.5",
+  "ruff==0.3.0",
 ]
 
 [tool.hatch.envs.notebooks]
@@ -207,12 +202,12 @@ check = [
 
 [tool.hatch.envs.style.scripts]
 check = [
-  "ruff .",
   "ruff format --check --diff .",
+  "ruff check .",
 ]
 fix = [
   "ruff format .",
-  "ruff --fix .",
+  "ruff check --fix .",
 ]
 
 [tool.hatch.envs.notebooks.scripts]
@@ -315,6 +310,7 @@ ignore_missing_imports = true
 exclude = [
   "packages",
   "src/phoenix/evals/",
+  "dist/",
   ".git",
   "__pycache__",
   "docs/source/conf.py",
@@ -322,13 +318,15 @@ exclude = [
   "*.pyi",
 ]
 extend-include = ["*.ipynb"]
-ignore-init-module-imports = true
 line-length = 100
-select = ["E", "F", "W", "I"]
 target-version = "py38"
 
 [tool.ruff.lint.per-file-ignores]
 "*.ipynb" = ["E402", "E501"]
 
-[tool.ruff.isort]
+[tool.ruff.lint]
+ignore-init-module-imports = true
+select = ["E", "F", "W", "I"]
+
+[tool.ruff.lint.isort]
 force-single-line = false

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/core/evals.py

@@ -33,15 +33,15 @@ class Evals:
         self._trace_evaluations_by_name: DefaultDict[
             EvaluationName, Dict[TraceID, pb.Evaluation]
         ] = defaultdict(dict)
-        self._evaluations_by_trace_id: DefaultDict[
-            TraceID, Dict[EvaluationName, pb.Evaluation]
-        ] = defaultdict(dict)
-        self._span_evaluations_by_name: DefaultDict[
-            EvaluationName, Dict[SpanID, pb.Evaluation]
-        ] = defaultdict(dict)
-        self._evaluations_by_span_id: DefaultDict[
-            SpanID, Dict[EvaluationName, pb.Evaluation]
-        ] = defaultdict(dict)
+        self._evaluations_by_trace_id: DefaultDict[TraceID, Dict[EvaluationName, pb.Evaluation]] = (
+            defaultdict(dict)
+        )
+        self._span_evaluations_by_name: DefaultDict[EvaluationName, Dict[SpanID, pb.Evaluation]] = (
+            defaultdict(dict)
+        )
+        self._evaluations_by_span_id: DefaultDict[SpanID, Dict[EvaluationName, pb.Evaluation]] = (
+            defaultdict(dict)
+        )
         self._span_evaluation_labels: DefaultDict[EvaluationName, Set[str]] = defaultdict(set)
         self._document_evaluations_by_span_id: DefaultDict[
             SpanID, DefaultDict[EvaluationName, Dict[DocumentPosition, pb.Evaluation]]

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/core/model.py

@@ -12,36 +12,36 @@ def _get_embedding_dimensions(
     embedding_dimensions: List[EmbeddingDimension] = []
     embedding_features: EmbeddingFeatures = {}
 
-    primary_embedding_features: Optional[
-        EmbeddingFeatures
-    ] = primary_dataset.schema.embedding_feature_column_names
+    primary_embedding_features: Optional[EmbeddingFeatures] = (
+        primary_dataset.schema.embedding_feature_column_names
+    )
     if primary_embedding_features is not None:
         embedding_features.update(primary_embedding_features)
-    primary_prompt_column_names: Optional[
-        EmbeddingColumnNames
-    ] = primary_dataset.schema.prompt_column_names
+    primary_prompt_column_names: Optional[EmbeddingColumnNames] = (
+        primary_dataset.schema.prompt_column_names
+    )
     if primary_prompt_column_names is not None:
         embedding_features.update({"prompt": primary_prompt_column_names})
-    primary_response_column_names: Optional[
-        Union[str, EmbeddingColumnNames]
-    ] = primary_dataset.schema.response_column_names
+    primary_response_column_names: Optional[Union[str, EmbeddingColumnNames]] = (
+        primary_dataset.schema.response_column_names
+    )
     if isinstance(primary_response_column_names, EmbeddingColumnNames):
         embedding_features.update({"response": primary_response_column_names})
 
     if reference_dataset is not None:
-        reference_embedding_features: Optional[
-            EmbeddingFeatures
-        ] = reference_dataset.schema.embedding_feature_column_names
+        reference_embedding_features: Optional[EmbeddingFeatures] = (
+            reference_dataset.schema.embedding_feature_column_names
+        )
         if reference_embedding_features is not None:
             embedding_features.update(reference_embedding_features)
-        reference_prompt_column_names: Optional[
-            EmbeddingColumnNames
-        ] = reference_dataset.schema.prompt_column_names
+        reference_prompt_column_names: Optional[EmbeddingColumnNames] = (
+            reference_dataset.schema.prompt_column_names
+        )
         if reference_prompt_column_names is not None:
             embedding_features.update({"prompt": reference_prompt_column_names})
-        reference_response_column_names: Optional[
-            Union[str, EmbeddingColumnNames]
-        ] = reference_dataset.schema.response_column_names
+        reference_response_column_names: Optional[Union[str, EmbeddingColumnNames]] = (
+            reference_dataset.schema.response_column_names
+        )
         if isinstance(reference_response_column_names, EmbeddingColumnNames):
             embedding_features.update({"response": reference_response_column_names})
 

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/core/model_schema.py

@@ -52,8 +52,7 @@ from phoenix.config import GENERATED_DATASET_NAME_PREFIX
 from phoenix.datetime_utils import floor_to_minute
 
 
-class DimensionRole(IntEnum):
-    ...
+class DimensionRole(IntEnum): ...
 
 
 @unique
@@ -151,8 +150,7 @@ class CompositeDimensionSpec(SchemaSpec, ABC):
         ...
 
     @abstractmethod
-    def __iter__(self) -> Iterator[str]:
-        ...
+    def __iter__(self) -> Iterator[str]: ...
 
 
 @dataclass(frozen=True)
@@ -311,12 +309,10 @@ class Column:
             object.__setattr__(self, "name", _rand_str())
 
     @overload
-    def __call__(self, data: pd.DataFrame) -> "pd.Series[Any]":
-        ...
+    def __call__(self, data: pd.DataFrame) -> "pd.Series[Any]": ...
 
     @overload
-    def __call__(self, data: "pd.Series[Any]") -> Any:
-        ...
+    def __call__(self, data: "pd.Series[Any]") -> Any: ...
 
     def __call__(self, data: DataFrameOrSeries) -> Any:
         """Extracts a value from series, or a series from a dataframe. If
@@ -569,8 +565,7 @@ class ModelData(ObjectProxy, ABC):  # type: ignore
 
     @property
     @abstractmethod
-    def null_value(self) -> Any:
-        ...
+    def null_value(self) -> Any: ...
 
     def __getitem__(self, key: Any) -> Any:
         if _is_column_key(key):
@@ -614,12 +609,10 @@ class Event(ModelData):
         return np.nan
 
     @overload
-    def __getitem__(self, key: ColumnKey) -> Any:
-        ...
+    def __getitem__(self, key: ColumnKey) -> Any: ...
 
     @overload
-    def __getitem__(self, key: Any) -> Any:
-        ...
+    def __getitem__(self, key: Any) -> Any: ...
 
     def __getitem__(self, key: Any) -> Any:
         return super().__getitem__(key)
@@ -668,12 +661,10 @@ class Events(ModelData):
             )
 
     @overload
-    def __getitem__(self, key: ColumnKey) -> "pd.Series[Any]":
-        ...
+    def __getitem__(self, key: ColumnKey) -> "pd.Series[Any]": ...
 
     @overload
-    def __getitem__(self, key: List[RowId]) -> "Events":
-        ...
+    def __getitem__(self, key: List[RowId]) -> "Events": ...
 
     def __getitem__(self, key: Any) -> Any:
         if isinstance(key, list):
@@ -728,12 +719,10 @@ class Dataset(Events):
         return pd.Index(self[PREDICTION_ID])
 
     @overload
-    def __getitem__(self, key: ColumnKey) -> "pd.Series[Any]":
-        ...
+    def __getitem__(self, key: ColumnKey) -> "pd.Series[Any]": ...
 
     @overload
-    def __getitem__(self, key: List[RowId]) -> Events:
-        ...
+    def __getitem__(self, key: List[RowId]) -> Events: ...
 
     def __getitem__(self, key: Any) -> Any:
         if isinstance(key, list):
@@ -1030,32 +1019,25 @@ class Model:
         return ans
 
     @overload
-    def __getitem__(self, key: Type[Dataset]) -> Iterator[Dataset]:
-        ...
+    def __getitem__(self, key: Type[Dataset]) -> Iterator[Dataset]: ...
 
     @overload
-    def __getitem__(self, key: DatasetRole) -> Dataset:
-        ...
+    def __getitem__(self, key: DatasetRole) -> Dataset: ...
 
     @overload
-    def __getitem__(self, key: ColumnKey) -> Dimension:
-        ...
+    def __getitem__(self, key: ColumnKey) -> Dimension: ...
 
     @overload
-    def __getitem__(self, key: MultiDimensionKey) -> Iterator[Dimension]:
-        ...
+    def __getitem__(self, key: MultiDimensionKey) -> Iterator[Dimension]: ...
 
     @overload
-    def __getitem__(self, key: Type[ScalarDimension]) -> Iterator[ScalarDimension]:
-        ...
+    def __getitem__(self, key: Type[ScalarDimension]) -> Iterator[ScalarDimension]: ...
 
     @overload
-    def __getitem__(self, key: Type[EmbeddingDimension]) -> Iterator[EmbeddingDimension]:
-        ...
+    def __getitem__(self, key: Type[EmbeddingDimension]) -> Iterator[EmbeddingDimension]: ...
 
     @overload
-    def __getitem__(self, key: Type[Dimension]) -> Iterator[Dimension]:
-        ...
+    def __getitem__(self, key: Type[Dimension]) -> Iterator[Dimension]: ...
 
     @overload
     def __getitem__(
@@ -1064,8 +1046,7 @@ class Model:
             MultiDimensionKey,
             Union[Type[ScalarDimension], Type[EmbeddingDimension]],
         ],
-    ) -> Iterator[Dimension]:
-        ...
+    ) -> Iterator[Dimension]: ...
 
     def __getitem__(self, key: Any) -> Any:
         if key is Dataset:
@@ -1124,8 +1105,7 @@ class Model:
         obj: DimensionRole,
         cls: Type[Dimension] = ScalarDimension,
         **kwargs: Any,
-    ) -> Dimension:
-        ...
+    ) -> Dimension: ...
 
     @overload
     def _new_dimension(
@@ -1133,16 +1113,14 @@ class Model:
         obj: Name,
         cls: Type[Dimension] = ScalarDimension,
         **kwargs: Any,
-    ) -> Dimension:
-        ...
+    ) -> Dimension: ...
 
     @overload
     def _new_dimension(
         self,
         obj: Dimension,
         **kwargs: Any,
-    ) -> Dimension:
-        ...
+    ) -> Dimension: ...
 
     def _new_dimension(
         self, obj: Any, cls: Type[Dimension] = ScalarDimension, **kwargs: Any

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/experimental/evals/functions/executor.py

@@ -21,8 +21,7 @@ _unset = Unset()
 
 
 class Executor(Protocol):
-    def run(self, inputs: Sequence[Any]) -> List[Any]:
-        ...
+    def run(self, inputs: Sequence[Any]) -> List[Any]: ...
 
 
 class AsyncExecutor(Executor):

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/experimental/evals/functions/processing.py

@@ -1,7 +1,22 @@
-from typing import List
+"""
+Token processing functions for supported models. This module is being deprecated.
+"""
+
+import logging
+import sys
+from typing import Any, List
 
 from ..models import BaseEvalModel
 
+logger = logging.getLogger(__name__)
+
+_DEPRECATION_WARNING = (
+    "The processing module is being deprecated. For advanced token processing, please use the "
+    "encoding approach recommended by the model provider. For example, OpenAI models can use the "
+    "`tiktoken` library to encode and decode text. For other models, please refer to the model "
+    "provider's documentation."
+)
+
 
 def truncate_text_by_model(model: BaseEvalModel, text: str, token_buffer: int = 0) -> str:
     """Truncates text using a give model token limit.
@@ -42,3 +57,20 @@ def concatenate_and_truncate_chunks(
         str: _description_
     """
     return truncate_text_by_model(model=model, text=" ".join(chunks), token_buffer=token_buffer)
+
+
+class _DEPRECATED_MODULE:
+    __all__ = ("truncate_text_by_model", "concatenate_and_truncate_chunks")
+
+    def __getattr__(self, name: str) -> Any:
+        if name == "truncate_text_by_model":
+            logger.warning(_DEPRECATION_WARNING)
+            return truncate_text_by_model
+        if name == "concatenate_and_truncate_chunks":
+            logger.warning(_DEPRECATION_WARNING)
+            return concatenate_and_truncate_chunks
+        raise AttributeError(f"module {__name__} has no attribute {name}")
+
+
+# See e.g. https://stackoverflow.com/a/7668273
+sys.modules[__name__] = _DEPRECATED_MODULE()  # type: ignore

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/experimental/evals/models/base.py

@@ -158,17 +158,13 @@ class BaseEvalModel(ABC):
         raise ImportError(msg)
 
     @abstractmethod
-    def get_tokens_from_text(self, text: str) -> List[int]:
-        ...
+    def get_tokens_from_text(self, text: str) -> List[int]: ...
 
     @abstractmethod
-    def get_text_from_tokens(self, tokens: List[int]) -> str:
-        ...
+    def get_text_from_tokens(self, tokens: List[int]) -> str: ...
 
     @abstractproperty
-    def max_context_size(self) -> int:
-        ...
+    def max_context_size(self) -> int: ...
 
     @abstractproperty
-    def encoder(self) -> "Encoding":
-        ...
+    def encoder(self) -> "Encoding": ...

{arize_phoenix-3.4.1 → arize_phoenix-3.5.0}/src/phoenix/experimental/evals/models/litellm.py

@@ -67,7 +67,7 @@ class LiteLLMModel(BaseEvalModel):
             self._litellm = litellm
             env_info = validate_environment(self._litellm.utils.get_llm_provider(self.model))
 
-            if not env_info["keys_in_environment"]:
+            if not env_info["keys_in_environment"] and env_info["missing_keys"]:
                 raise RuntimeError(
                     f"Missing environment variable(s): '{str(env_info['missing_keys'])}', for "
                     f"model: {self.model}. \nFor additional information about the right "