cohere-ai · billytrend-cohere · Dec 20, 2024 · Dec 20, 2024 · Sep 19, 2024 · Dec 20, 2024
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -35,7 +35,9 @@ jobs:
           version: 1.5.1
           virtualenvs-in-project: false
       - name: Install dependencies
-        run: poetry install --extras aws
+        run: poetry install
+      - name: Install aws deps
+        run: poetry run pip install boto3 sagemaker botocore
       - name: Test
         run: poetry run pytest .
         env:

diff --git a/poetry.lock b/poetry.lock
diff --git a/pyproject.toml b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "cohere"
-version = "5.13.3"
+version = "5.13.4"
 description = ""
 readme = "README.md"
 authors = []
@@ -32,16 +32,13 @@ Repository = 'https://github.com/cohere-ai/cohere-python'
 
 [tool.poetry.dependencies]
 python = "^3.9"
-boto3 = { version="^1.34.0", optional = true}
 fastavro = "^1.9.4"
 httpx = ">=0.21.2"
 httpx-sse = "0.4.0"
 parameterized = "^0.9.0"
 pydantic = ">= 1.9.2"
 pydantic-core = "^2.18.2"
 requests = "^2.0.0"
-sagemaker = { version="^2.232.1", optional = true}
-numpy = { version="^1.26", optional = true}
 tokenizers = ">=0.15,<1"
 types-requests = "^2.0.0"
 typing_extensions = ">= 4.0.0"
@@ -69,6 +66,3 @@ line-length = 120
 [build-system]
 requires = ["poetry-core"]
 build-backend = "poetry.core.masonry.api"
-
-[tool.poetry.extras]
-aws=["sagemaker", "boto3"]
diff --git a/reference.md b/reference.md
@@ -223,7 +223,7 @@ With `prompt_truncation` set to "AUTO_PRESERVE_ORDER", some elements from `chat_
 
 With `prompt_truncation` set to "OFF", no elements will be dropped. If the sum of the inputs exceeds the model's context length limit, a `TooManyTokens` error will be returned.
 
-Compatible Deployments: 
+Compatible Deployments:
  - AUTO: Cohere Platform Only
  - AUTO_PRESERVE_ORDER: Azure, AWS Sagemaker/Bedrock, Private Deployments
 
@@ -537,6 +537,8 @@ Safety modes are not yet configurable in combination with `tools`, `tool_results
 
 **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/docs/command-r-plus#august-2024-release) and newer.
 
+**Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 Compatible Deployments: Cohere Platform, Azure, AWS Sagemaker/Bedrock, Private Deployments
 
 
@@ -708,7 +710,7 @@ With `prompt_truncation` set to "AUTO_PRESERVE_ORDER", some elements from `chat_
 
 With `prompt_truncation` set to "OFF", no elements will be dropped. If the sum of the inputs exceeds the model's context length limit, a `TooManyTokens` error will be returned.
 
-Compatible Deployments: 
+Compatible Deployments:
  - AUTO: Cohere Platform Only
  - AUTO_PRESERVE_ORDER: Azure, AWS Sagemaker/Bedrock, Private Deployments
 
@@ -1022,6 +1024,8 @@ Safety modes are not yet configurable in combination with `tools`, `tool_results
 
 **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/docs/command-r-plus#august-2024-release) and newer.
 
+**Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 Compatible Deployments: Cohere Platform, Azure, AWS Sagemaker/Bedrock, Private Deployments
 
 
@@ -2431,7 +2435,7 @@ When `tools` is passed (without `tool_results`), the `text` content in the respo
 
 **strict_tools:** `typing.Optional[bool]` 
 
-When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Strict Tools guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
+When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Structured Outputs (Tools) guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
 
 **Note**: The first few requests with a new set of tools will take longer to process.
 
@@ -2476,6 +2480,8 @@ Safety modes are not yet configurable in combination with `tools`, `tool_results
 
 **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/v2/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/v2/docs/command-r-plus#august-2024-release) and newer.
 
+**Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 
 </dd>
 </dl>
@@ -2701,7 +2707,7 @@ When `tools` is passed (without `tool_results`), the `text` content in the respo
 
 **strict_tools:** `typing.Optional[bool]` 
 
-When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Strict Tools guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
+When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Structured Outputs (Tools) guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
 
 **Note**: The first few requests with a new set of tools will take longer to process.
 
@@ -2746,6 +2752,8 @@ Safety modes are not yet configurable in combination with `tools`, `tool_results
 
 **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/v2/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/v2/docs/command-r-plus#august-2024-release) and newer.
 
+**Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 
 </dd>
 </dl>

diff --git a/src/cohere/base_client.py b/src/cohere/base_client.py
@@ -423,6 +423,8 @@ def chat_stream(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
             Compatible Deployments: Cohere Platform, Azure, AWS Sagemaker/Bedrock, Private Deployments
 
 
@@ -974,6 +976,8 @@ def chat(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
             Compatible Deployments: Cohere Platform, Azure, AWS Sagemaker/Bedrock, Private Deployments
 
 
@@ -3418,6 +3422,8 @@ async def chat_stream(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
             Compatible Deployments: Cohere Platform, Azure, AWS Sagemaker/Bedrock, Private Deployments
 
 
@@ -3977,6 +3983,8 @@ async def chat(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
             Compatible Deployments: Cohere Platform, Azure, AWS Sagemaker/Bedrock, Private Deployments
 
 

diff --git a/src/cohere/core/client_wrapper.py b/src/cohere/core/client_wrapper.py
@@ -24,7 +24,7 @@ def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "cohere",
-            "X-Fern-SDK-Version": "5.13.3",
+            "X-Fern-SDK-Version": "5.13.4",
         }
         if self._client_name is not None:
             headers["X-Client-Name"] = self._client_name

diff --git a/src/cohere/manually_maintained/lazy_aws_deps.py b/src/cohere/manually_maintained/lazy_aws_deps.py
@@ -1,23 +1,24 @@
 
+warning = "AWS dependencies are not installed. Please install boto3, botocore, and sagemaker."
 
 def lazy_sagemaker():
     try:
         import sagemaker as sage # type: ignore
         return sage
     except ImportError:
-        raise ImportError("Sagemaker not available. Please install sagemaker.")
+        raise ImportError(warning)
 
 def lazy_boto3():
     try:
         import boto3 # type: ignore
         return boto3
     except ImportError:
-        raise ImportError("Boto3 not available. Please install lazy_boto3().")
+        raise ImportError(warning)
 
 def lazy_botocore():
     try:
         import botocore # type: ignore
         return botocore
     except ImportError:
-        raise ImportError("Botocore not available. Please install botocore.")
+        raise ImportError(warning)
 
diff --git a/src/cohere/types/citation_options.py b/src/cohere/types/citation_options.py
@@ -16,6 +16,8 @@ class CitationOptions(UncheckedBaseModel):
     """
     Defaults to `"accurate"`.
     Dictates the approach taken to generating citations as part of the RAG flow by allowing the user to specify whether they want `"accurate"` results, `"fast"` results or no results.
+
+    **Note**: `command-r7b-12-2024` only supports `"fast"` and `"off"` modes. Its default is `"fast"`.
     """
 
     if IS_PYDANTIC_V2:

diff --git a/src/cohere/utils.py b/src/cohere/utils.py
@@ -274,7 +274,7 @@ def dataset_generator(dataset: Dataset):
         if not part.url:
             raise ValueError("Dataset part does not have a url")
         resp = requests.get(part.url, stream=True)
-        for record in reader(resp.raw):
+        for record in reader(resp.raw): # type: ignore
             yield record
 
 

diff --git a/src/cohere/v2/client.py b/src/cohere/v2/client.py
@@ -92,7 +92,7 @@ def chat_stream(
 
 
         strict_tools : typing.Optional[bool]
-            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Strict Tools guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
+            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Structured Outputs (Tools) guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
 
             **Note**: The first few requests with a new set of tools will take longer to process.
 
@@ -113,6 +113,8 @@ def chat_stream(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/v2/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/v2/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 
         max_tokens : typing.Optional[int]
             The maximum number of tokens the model will generate as part of the response.
@@ -436,7 +438,7 @@ def chat(
 
 
         strict_tools : typing.Optional[bool]
-            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Strict Tools guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
+            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Structured Outputs (Tools) guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
 
             **Note**: The first few requests with a new set of tools will take longer to process.
 
@@ -457,6 +459,8 @@ def chat(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/v2/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/v2/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 
         max_tokens : typing.Optional[int]
             The maximum number of tokens the model will generate as part of the response.
@@ -1166,7 +1170,7 @@ async def chat_stream(
 
 
         strict_tools : typing.Optional[bool]
-            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Strict Tools guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
+            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Structured Outputs (Tools) guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
 
             **Note**: The first few requests with a new set of tools will take longer to process.
 
@@ -1187,6 +1191,8 @@ async def chat_stream(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/v2/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/v2/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 
         max_tokens : typing.Optional[int]
             The maximum number of tokens the model will generate as part of the response.
@@ -1518,7 +1524,7 @@ async def chat(
 
 
         strict_tools : typing.Optional[bool]
-            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Strict Tools guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
+            When set to `true`, tool calls in the Assistant message will be forced to follow the tool definition strictly. Learn more in the [Structured Outputs (Tools) guide](https://docs.cohere.com/docs/structured-outputs-json#structured-outputs-tools).
 
             **Note**: The first few requests with a new set of tools will take longer to process.
 
@@ -1539,6 +1545,8 @@ async def chat(
 
             **Note**: This parameter is only compatible with models [Command R 08-2024](https://docs.cohere.com/v2/docs/command-r#august-2024-release), [Command R+ 08-2024](https://docs.cohere.com/v2/docs/command-r-plus#august-2024-release) and newer.
 
+            **Note**: `command-r7b-12-2024` only supports `"CONTEXTUAL"` and `"STRICT"` modes.
+
 
         max_tokens : typing.Optional[int]
             The maximum number of tokens the model will generate as part of the response.