unique_toolkit 0.5.26__py3-none-any.whl → 0.5.30__py3-none-any.whl

unique_toolkit/content/service.py

@@ -204,6 +204,13 @@ class ContentService(BaseService):
 
         return self._map_contents(contents)
 
+    def search_content_on_chat(
+        self,
+    ) -> list[Content]:
+        where = {"ownerId": {"equals": self.event.payload.chat_id}}
+
+        return self.search_contents(where)
+
     @staticmethod
     def _map_content_chunk(content_chunk: dict):
         return ContentChunk(

unique_toolkit/language_model/infos.py

@@ -20,6 +20,35 @@ class LanguageModelName(StrEnum):
     AZURE_GPT_4o_MINI_2024_0718 = "AZURE_GPT_4o_MINI_2024_0718"
 
 
+class EncoderName(StrEnum):
+    O200K_BASE = "o200k_base"
+    CL100K_BASE = "cl100k_base"
+
+
+def get_encoder_name(model_name: LanguageModelName) -> Optional[EncoderName]:
+    LMN = LanguageModelName
+    match model_name:
+        case (
+            LMN.AZURE_GPT_35_TURBO
+            | LMN.AZURE_GPT_35_TURBO_16K
+            | LMN.AZURE_GPT_35_TURBO_0613
+        ):
+            return EncoderName.CL100K_BASE
+        case (
+            LMN.AZURE_GPT_4_0613
+            | LMN.AZURE_GPT_4_TURBO_1106
+            | LMN.AZURE_GPT_4_VISION_PREVIEW
+            | LMN.AZURE_GPT_4_32K_0613
+            | LMN.AZURE_GPT_4_TURBO_2024_0409
+        ):
+            return EncoderName.CL100K_BASE
+        case LMN.AZURE_GPT_4o_2024_0513 | LMN.AZURE_GPT_4o_MINI_2024_0718:
+            return EncoderName.O200K_BASE
+        case _:
+            print(f"{model_name} is not supported. Please add encoder information.")
+            return None
+
+
 class LanguageModelProvider(StrEnum):
     AZURE = "AZURE"
     CUSTOM = "CUSTOM"
@@ -30,6 +59,7 @@ class LanguageModelInfo(BaseModel):
     version: str
     provider: LanguageModelProvider
 
+    encoder_name: Optional[EncoderName] = None
     token_limits: Optional[LanguageModelTokenLimits] = None
 
     info_cutoff_at: Optional[date] = None
@@ -53,6 +83,7 @@ class LanguageModel:
         - name
         - version
         - provider
+        - encoder_name
         - token_limits
         - info_cutoff_at
         - published_at
@@ -86,6 +117,13 @@ class LanguageModel:
         """
         return self._model_info.version
 
+    @property
+    def encoder_name(self) -> Optional[EncoderName]:
+        """
+        Returns the encoder_name used for the model.
+        """
+        return self._model_info.encoder_name
+
     @property
     def token_limit(self) -> Optional[int]:
         """
@@ -191,6 +229,7 @@ def create_language_model(
     provider: LanguageModelProvider,
     info_cutoff_at: date,
     published_at: date,
+    encoder_name: Optional[EncoderName] = None,
     token_limit: Optional[int] = None,
     token_limit_input: Optional[int] = None,
     token_limit_output: Optional[int] = None,
@@ -202,6 +241,7 @@ def create_language_model(
         name=name,
         version=version,
         provider=provider,
+        encoder_name=encoder_name,
         token_limits=LanguageModelTokenLimits(
             token_limit=token_limit,
             token_limit_input=token_limit_input,
@@ -229,6 +269,7 @@ AzureGpt35Turbo0613 = create_language_model(
     name=LanguageModelName.AZURE_GPT_35_TURBO_0613,
     provider=LanguageModelProvider.AZURE,
     version="0613",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_35_TURBO_0613),
     token_limit=8192,
     info_cutoff_at=date(2021, 9, 1),
     published_at=date(2023, 6, 13),
@@ -239,6 +280,7 @@ AzureGpt35Turbo = create_language_model(
     name=LanguageModelName.AZURE_GPT_35_TURBO,
     provider=LanguageModelProvider.AZURE,
     version="0301",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_35_TURBO),
     token_limit=4096,
     info_cutoff_at=date(2021, 9, 1),
     published_at=date(2023, 3, 1),
@@ -249,6 +291,7 @@ AzureGpt35Turbo16k = create_language_model(
     name=LanguageModelName.AZURE_GPT_35_TURBO_16K,
     provider=LanguageModelProvider.AZURE,
     version="0613",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_35_TURBO_16K),
     token_limit=16382,
     info_cutoff_at=date(2021, 9, 1),
     published_at=date(2023, 6, 13),
@@ -260,6 +303,7 @@ AzureGpt40613 = create_language_model(
     name=LanguageModelName.AZURE_GPT_4_0613,
     provider=LanguageModelProvider.AZURE,
     version="0613",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4_0613),
     token_limit=8192,
     info_cutoff_at=date(2021, 9, 1),
     published_at=date(2023, 6, 13),
@@ -272,6 +316,7 @@ AzureGpt4Turbo1106 = create_language_model(
     name=LanguageModelName.AZURE_GPT_4_TURBO_1106,
     provider=LanguageModelProvider.AZURE,
     version="1106-preview",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4_TURBO_1106),
     token_limit_input=128000,
     token_limit_output=4096,
     info_cutoff_at=date(2023, 4, 1),
@@ -283,6 +328,7 @@ AzureGpt4VisionPreview = create_language_model(
     name=LanguageModelName.AZURE_GPT_4_VISION_PREVIEW,
     provider=LanguageModelProvider.AZURE,
     version="vision-preview",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4_VISION_PREVIEW),
     token_limit_input=128000,
     token_limit_output=4096,
     info_cutoff_at=date(2023, 4, 1),
@@ -293,6 +339,7 @@ AzureGpt432k0613 = create_language_model(
     name=LanguageModelName.AZURE_GPT_4_32K_0613,
     provider=LanguageModelProvider.AZURE,
     version="1106-preview",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4_32K_0613),
     token_limit=32768,
     info_cutoff_at=date(2021, 9, 1),
     published_at=date(2023, 6, 13),
@@ -302,6 +349,7 @@ AzureGpt432k0613 = create_language_model(
 
 AzureGpt4Turbo20240409 = create_language_model(
     name=LanguageModelName.AZURE_GPT_4_TURBO_2024_0409,
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4_TURBO_2024_0409),
     provider=LanguageModelProvider.AZURE,
     version="turbo-2024-04-09",
     token_limit_input=128000,
@@ -312,6 +360,7 @@ AzureGpt4Turbo20240409 = create_language_model(
 
 AzureGpt4o20240513 = create_language_model(
     name=LanguageModelName.AZURE_GPT_4o_2024_0513,
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4o_2024_0513),
     provider=LanguageModelProvider.AZURE,
     version="2024-05-13",
     token_limit_input=128000,
@@ -324,6 +373,7 @@ AzureGpt4oMini20240718 = create_language_model(
     name=LanguageModelName.AZURE_GPT_4o_MINI_2024_0718,
     provider=LanguageModelProvider.AZURE,
     version="2024-07-18",
+    encoder_name=get_encoder_name(LanguageModelName.AZURE_GPT_4o_MINI_2024_0718),
     token_limit_input=128000,
     token_limit_output=16384,
     info_cutoff_at=date(2023, 10, 1),

unique_toolkit/language_model/schemas.py

@@ -3,7 +3,14 @@ from enum import StrEnum
 from typing import Any, Optional, Self
 
 from humps import camelize
-from pydantic import BaseModel, ConfigDict, RootModel, field_validator, model_validator
+from pydantic import (
+    BaseModel,
+    ConfigDict,
+    Field,
+    RootModel,
+    field_validator,
+    model_validator,
+)
 
 # set config to convert camelCase to snake_case
 model_config = ConfigDict(
@@ -173,7 +180,11 @@ class LanguageModelToolParameters(BaseModel):
 
 
 class LanguageModelTool(BaseModel):
-    name: str
+    name: str = Field(
+        ...,
+        pattern=r"^[a-zA-Z1-9_-]+$",
+        description="Name must adhere to the pattern ^[a-zA-Z_-]+$",
+    )
     description: str
     parameters: LanguageModelToolParameters
     returns: LanguageModelToolParameterProperty | LanguageModelToolParameters | None = (

unique_toolkit/language_model/service.py

@@ -52,6 +52,7 @@ class LanguageModelService(BaseService):
             LanguageModelResponse: The LanguageModelResponse object.
         """
         options = self._add_tools_to_options({}, tools)
+        options["temperature"] = temperature
         messages = messages.model_dump(exclude_none=True)
         model = (
             model_name.name if isinstance(model_name, LanguageModelName) else model_name
@@ -67,7 +68,6 @@ class LanguageModelService(BaseService):
                     messages,
                 ),
                 timeout=timeout,
-                temperature=temperature,
                 options=options,  # type: ignore
             )
             return LanguageModelResponse(**response)
@@ -109,6 +109,7 @@ class LanguageModelService(BaseService):
             Exception: If an error occurs during the request, an exception is raised and logged.
         """
         options = cls._add_tools_to_options({}, tools)
+        options["temperature"] = temperature
         messages = messages.model_dump(exclude_none=True, exclude={"tool_calls"})
         model = (
             model_name.name if isinstance(model_name, LanguageModelName) else model_name
@@ -122,7 +123,6 @@ class LanguageModelService(BaseService):
                     messages,
                 ),
                 timeout=timeout,
-                temperature=temperature,
                 options=options,  # type: ignore
             )
             return LanguageModelResponse(**response)
@@ -196,6 +196,7 @@ class LanguageModelService(BaseService):
             The LanguageModelStreamResponse object once the stream has finished.
         """
         options = self._add_tools_to_options({}, tools)
+        options["temperature"] = temperature
         search_context = self._to_search_context(content_chunks)
         messages = messages.model_dump(exclude_none=True)
         model = (
@@ -217,7 +218,6 @@ class LanguageModelService(BaseService):
                 # TODO change or extend types in unique_sdk
                 model=model,
                 timeout=timeout,
-                temperature=temperature,
                 assistantId=self.event.payload.assistant_id,
                 debugInfo=debug_info,
                 options=options,  # type: ignore
@@ -257,6 +257,7 @@ class LanguageModelService(BaseService):
         """
 
         options = self._add_tools_to_options({}, tools)
+        options["temperature"] = temperature
         search_context = self._to_search_context(content_chunks)
         messages = messages.model_dump(exclude_none=True, exclude=["tool_calls"])
         model = (
@@ -277,7 +278,6 @@ class LanguageModelService(BaseService):
                 searchContext=search_context,
                 model=model,
                 timeout=timeout,
-                temperature=temperature,
                 assistantId=self.event.payload.assistant_id,
                 debugInfo=debug_info,
                 # TODO change or extend types in unique_sdk

{unique_toolkit-0.5.26.dist-info → unique_toolkit-0.5.30.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: unique_toolkit
-Version: 0.5.26
+Version: 0.5.30
 Summary: 
 License: Proprietary
 Author: Martin Fadler
@@ -100,6 +100,20 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), 
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.5.30] - 2024-10-28
+- Correctly use `temperature` parameter in `LanguageModelService.complete` and `LanguageModelService.complete_async` methods
+
+## [0.5.29] - 2024-10-28
+- Allow numbers in `LanguageModelTool` name
+
+## [0.5.28] - 2024-10-23
+- Correctly use `temperature` parameter in `LanguageModelService.stream_complete` and `LanguageModelService.stream_complete_async` methods
+
+## [0.5.27] - 2024-10-22
+- Add encoder_name to to language model info
+- Verify tool name for `LanguageModelTool` to conform with frontent requirements.
+- Add `search_on_chat` to `ContentService`
+
 ## [0.5.26] - 2024-10-16
 - Bump `unique_sdk` version
 

{unique_toolkit-0.5.26.dist-info → unique_toolkit-0.5.30.dist-info}/RECORD

@@ -17,7 +17,7 @@ unique_toolkit/chat/state.py,sha256=Cjgwv_2vhDFbV69xxsn7SefhaoIAEqLx3ferdVFCnOg,
 unique_toolkit/chat/utils.py,sha256=ihm-wQykBWhB4liR3LnwPVPt_qGW6ETq21Mw4HY0THE,854
 unique_toolkit/content/__init__.py,sha256=MSH2sxjQyKD2Sef92fzE5Dt9SihdzivB6yliSwJfTmQ,890
 unique_toolkit/content/schemas.py,sha256=zks_Pkki2VhxICJJgHZyc-LPmRuj5dLbw3pgcUT7SW8,2362
-unique_toolkit/content/service.py,sha256=AHyMJTXm5IpYbg1uINzjGqvSL_5aJwEHwSH7Y5pkXBg,14028
+unique_toolkit/content/service.py,sha256=ZGYWYTphXpcByXyMqr1VOVUHdmdnsR-XIS_YRX0Wyv4,14211
 unique_toolkit/content/utils.py,sha256=Lake671plRsqNvO3pN_rmyVcpwbdED_KQpLcCnc4lv4,6902
 unique_toolkit/embedding/__init__.py,sha256=dr8M9jvslQTxPpxgaGwzxY0FildiWf-DidN_cahPAWw,191
 unique_toolkit/embedding/schemas.py,sha256=1GvKCaSk4jixzVQ2PKq8yDqwGEVY_hWclYtoAr6CC2g,96
@@ -36,11 +36,11 @@ unique_toolkit/evaluators/hallucination/utils.py,sha256=507BsX1mFTEne1-LdRCNMgBj
 unique_toolkit/evaluators/output_parser.py,sha256=eI72qkzK1dZyUvnfP2SOAQCGBj_-PwX5wy_aLPMsJMY,883
 unique_toolkit/evaluators/schemas.py,sha256=Jaue6Uhx75X1CyHKWj8sT3RE1JZXTqoLtfLt2xQNCX8,2507
 unique_toolkit/language_model/__init__.py,sha256=YuhyczGPj6w9xX-sOVUhmozvzIFxcckHFEkeMBecr5s,1784
-unique_toolkit/language_model/infos.py,sha256=ETAUV0YTs6BjwuiTdhKz247CtL0W8Jwo3-c0ZQ2HdXs,9962
-unique_toolkit/language_model/schemas.py,sha256=sLpE29Ks0zEfhZUQrYOt1Cak2xzQcr9fpTXFDHkfURA,4868
-unique_toolkit/language_model/service.py,sha256=CvVo5CBa5Ia_fQD3DtJRsVChybuUfGFV5ml2_78_p1I,13395
+unique_toolkit/language_model/infos.py,sha256=Oxkr9_6s8gFubxjox-iCm1GSs1RCAQQ5t8oh20izlC0,12002
+unique_toolkit/language_model/schemas.py,sha256=LYWsZcGEPNK9T-fTHC4V0VbtCTNd1Q16UVf1b6WQzrU,5038
+unique_toolkit/language_model/service.py,sha256=dtC4I5iDDAJXsOwd_b2COj8tkKIp9LPaxTpGkuV2o3I,13411
 unique_toolkit/language_model/utils.py,sha256=WBPj1XKkDgxy_-T8HCZvsfkkSzj_1w4UZzNmyvdbBLY,1081
-unique_toolkit-0.5.26.dist-info/LICENSE,sha256=GlN8wHNdh53xwOPg44URnwag6TEolCjoq3YD_KrWgss,193
-unique_toolkit-0.5.26.dist-info/METADATA,sha256=zvoroS-Ti8npZOU8JXVyOThuNsgx29Z9N4KvDlqWC9g,12088
-unique_toolkit-0.5.26.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-unique_toolkit-0.5.26.dist-info/RECORD,,
+unique_toolkit-0.5.30.dist-info/LICENSE,sha256=GlN8wHNdh53xwOPg44URnwag6TEolCjoq3YD_KrWgss,193
+unique_toolkit-0.5.30.dist-info/METADATA,sha256=Vd_xIuIoQ_lnZNQJUoPUT3XQ32zXeGuAtC7FJdZQXxQ,12670
+unique_toolkit-0.5.30.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+unique_toolkit-0.5.30.dist-info/RECORD,,