feat(sdk): allow input text into the python sdk (#2738)

GitOrigin-RevId: 9c36a4ad2fa7bd831920145ecf1c29b16765cc11

README.md

@@ -266,6 +266,37 @@ print(result.response)
 
 </details>
 
+
+<details>
+  <summary>Use LeMUR to with Input Text</summary>
+
+```python
+import assemblyai as aai
+
+transcriber = aai.Transcriber()
+config = aai.TranscriptionConfig(
+  speaker_labels=True,
+)
+transcript = transcriber.transcribe("https://example.org/customer.mp3", config=config)
+
+# Example converting speaker label utterances into LeMUR input text
+text = ""
+
+for utt in transcript.utterances:
+    text += f"Speaker {utt.speaker}:\n{utt.text}\n"
+
+result = aai.Lemur().task(
+  "You are a helpful coach. Provide an analysis of the transcript "
+  "and offer areas to improve with exact quotes. Include no preamble. "
+  "Start with an overall summary then get into the examples with feedback.",
+  input_text=text
+)
+
+print(result.response)
+```
+
+</details>
+
 <details>
   <summary>Delete data previously sent to LeMUR</summary>
 

assemblyai/lemur.py

@@ -14,7 +14,11 @@ def __init__(
     ) -> None:
         self._client = client
 
-        self._sources = [types.LemurSourceRequest.from_lemur_source(s) for s in sources]
+        self._sources = (
+            [types.LemurSourceRequest.from_lemur_source(s) for s in sources]
+            if sources is not None
+            else []
+        )
 
     def question(
         self,
@@ -24,6 +28,7 @@ def question(
         final_model: Optional[types.LemurModel],
         max_output_size: Optional[int],
         temperature: Optional[float],
+        input_text: Optional[str],
     ) -> types.LemurQuestionResponse:
         response = api.lemur_question(
             client=self._client.http_client,
@@ -34,6 +39,7 @@ def question(
                 final_model=final_model,
                 max_output_size=max_output_size,
                 temperature=temperature,
+                input_text=input_text,
             ),
             http_timeout=timeout,
         )
@@ -48,6 +54,7 @@ def summarize(
         max_output_size: Optional[int],
         timeout: Optional[float],
         temperature: Optional[float],
+        input_text: Optional[str],
     ) -> types.LemurSummaryResponse:
         response = api.lemur_summarize(
             client=self._client.http_client,
@@ -58,6 +65,7 @@ def summarize(
                 final_model=final_model,
                 max_output_size=max_output_size,
                 temperature=temperature,
+                input_text=input_text,
             ),
             http_timeout=timeout,
         )
@@ -72,6 +80,7 @@ def action_items(
         max_output_size: Optional[int],
         timeout: Optional[float],
         temperature: Optional[float],
+        input_text: Optional[str],
     ) -> types.LemurActionItemsResponse:
         response = api.lemur_action_items(
             client=self._client.http_client,
@@ -82,6 +91,7 @@ def action_items(
                 final_model=final_model,
                 max_output_size=max_output_size,
                 temperature=temperature,
+                input_text=input_text,
             ),
             http_timeout=timeout,
         )
@@ -95,6 +105,7 @@ def task(
         max_output_size: Optional[int],
         timeout: Optional[float],
         temperature: Optional[float],
+        input_text: Optional[str],
     ):
         response = api.lemur_task(
             client=self._client.http_client,
@@ -104,6 +115,7 @@ def task(
                 final_model=final_model,
                 max_output_size=max_output_size,
                 temperature=temperature,
+                input_text=input_text,
             ),
             http_timeout=timeout,
         )
@@ -121,7 +133,7 @@ class Lemur:
 
     def __init__(
         self,
-        sources: List[types.LemurSource],
+        sources: Optional[List[types.LemurSource]] = None,
         client: Optional[_client.Client] = None,
     ) -> None:
         """
@@ -147,6 +159,7 @@ def question(
         max_output_size: Optional[int] = None,
         timeout: Optional[float] = None,
         temperature: Optional[float] = None,
+        input_text: Optional[str] = None,
     ) -> types.LemurQuestionResponse:
         """
         Question & Answer allows you to ask free form questions about one or many transcripts.
@@ -178,6 +191,7 @@ def question(
             max_output_size=max_output_size,
             timeout=timeout,
             temperature=temperature,
+            input_text=input_text,
         )
 
     def summarize(
@@ -188,6 +202,7 @@ def summarize(
         max_output_size: Optional[int] = None,
         timeout: Optional[float] = None,
         temperature: Optional[float] = None,
+        input_text: Optional[str] = None,
     ) -> types.LemurSummaryResponse:
         """
         Summary allows you to distill a piece of audio into a few impactful sentences.
@@ -214,6 +229,7 @@ def summarize(
             max_output_size=max_output_size,
             timeout=timeout,
             temperature=temperature,
+            input_text=input_text,
         )
 
     def action_items(
@@ -224,6 +240,7 @@ def action_items(
         max_output_size: Optional[int] = None,
         timeout: Optional[float] = None,
         temperature: Optional[float] = None,
+        input_text: Optional[str] = None,
     ) -> types.LemurActionItemsResponse:
         """
         Action Items allows you to generate action items from one or many transcripts.
@@ -251,6 +268,7 @@ def action_items(
             max_output_size=max_output_size,
             timeout=timeout,
             temperature=temperature,
+            input_text=input_text,
         )
 
     def task(
@@ -260,6 +278,7 @@ def task(
         max_output_size: Optional[int] = None,
         timeout: Optional[float] = None,
         temperature: Optional[float] = None,
+        input_text: Optional[str] = None,
     ) -> types.LemurTaskResponse:
         """
         Task feature allows you to submit a custom prompt to the model.
@@ -282,6 +301,7 @@ def task(
             max_output_size=max_output_size,
             timeout=timeout,
             temperature=temperature,
+            input_text=input_text,
         )
 
     @classmethod

assemblyai/types.py

@@ -1773,6 +1773,7 @@ class BaseLemurRequest(BaseModel):
     final_model: Optional[LemurModel]
     max_output_size: Optional[int]
     temperature: Optional[float]
+    input_text: Optional[str]
 
 
 class LemurTaskRequest(BaseLemurRequest):

tests/unit/test_lemur.py

@@ -14,7 +14,7 @@
 aai.settings.api_key = "test"
 
 
-def test_lemur_single_question_succeeds(httpx_mock: HTTPXMock):
+def test_lemur_single_question_succeeds_transcript(httpx_mock: HTTPXMock):
     """
     Tests whether asking a single question succeeds.
     """
@@ -64,7 +64,54 @@ def test_lemur_single_question_succeeds(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
-def test_lemur_multiple_question_succeeds(httpx_mock: HTTPXMock):
+def test_lemur_single_question_succeeds_input_text(httpx_mock: HTTPXMock):
+    """
+    Tests whether asking a single question succeeds with input text.
+    """
+
+    # create a mock response of a LemurQuestionResponse
+    mock_lemur_answer = factories.generate_dict_factory(
+        factories.LemurQuestionResponse
+    )()
+
+    # we only want to mock one answer
+    mock_lemur_answer["response"] = [mock_lemur_answer["response"][0]]
+
+    # mock the specific endpoints
+    httpx_mock.add_response(
+        url=f"{aai.settings.base_url}{ENDPOINT_LEMUR}/question-answer",
+        status_code=httpx.codes.OK,
+        method="POST",
+        json=mock_lemur_answer,
+    )
+
+    # prepare the question to be asked
+    question = aai.LemurQuestion(
+        question="Which cars do the callers want to buy?",
+        context="Callers are interested in buying cars",
+        answer_options=["Toyota", "Honda", "Ford", "Chevrolet"],
+    )
+    # test input_text input
+    # mimic the usage of the SDK
+    lemur = aai.Lemur()
+    result = lemur.question(
+        question, input_text="This transcript is a test transcript."
+    )
+
+    # check whether answer is not a list
+    assert isinstance(result, aai.LemurQuestionResponse)
+
+    answers = result.response
+
+    # check the response
+    assert answers[0].question == mock_lemur_answer["response"][0]["question"]
+    assert answers[0].answer == mock_lemur_answer["response"][0]["answer"]
+
+    # check whether we mocked everything
+    assert len(httpx_mock.get_requests()) == 1
+
+
+def test_lemur_multiple_question_succeeds_transcript(httpx_mock: HTTPXMock):
     """
     Tests whether asking multiple questions succeeds.
     """
@@ -117,6 +164,59 @@ def test_lemur_multiple_question_succeeds(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
+def test_lemur_multiple_question_succeeds_input_text(httpx_mock: HTTPXMock):
+    """
+    Tests whether asking multiple questions succeeds.
+    """
+
+    # create a mock response of a LemurQuestionResponse
+    mock_lemur_answer = factories.generate_dict_factory(
+        factories.LemurQuestionResponse
+    )()
+
+    # prepare the questions to be asked
+    questions = [
+        aai.LemurQuestion(
+            question="Which cars do the callers want to buy?",
+        ),
+        aai.LemurQuestion(
+            question="What price range are the callers looking for?",
+        ),
+    ]
+
+    # update the mock questions with the questions
+    mock_lemur_answer["response"][0]["question"] = questions[0].question
+    mock_lemur_answer["response"][1]["question"] = questions[1].question
+
+    # mock the specific endpoints
+    httpx_mock.add_response(
+        url=f"{aai.settings.base_url}{ENDPOINT_LEMUR}/question-answer",
+        status_code=httpx.codes.OK,
+        method="POST",
+        json=mock_lemur_answer,
+    )
+
+    # test input_text input
+    # mimic the usage of the SDK
+    lemur = aai.Lemur()
+    result = lemur.question(
+        questions, input_text="This transcript is a test transcript."
+    )
+    assert isinstance(result, aai.LemurQuestionResponse)
+
+    answers = result.response
+    # check whether answers is a list
+    assert isinstance(answers, list)
+
+    # check the response
+    for idx, answer in enumerate(answers):
+        assert answer.question == mock_lemur_answer["response"][idx]["question"]
+        assert answer.answer == mock_lemur_answer["response"][idx]["answer"]
+
+    # check whether we mocked everything
+    assert len(httpx_mock.get_requests()) == 1
+
+
 def test_lemur_question_fails(httpx_mock: HTTPXMock):
     """
     Tests whether asking a question fails.
@@ -149,7 +249,7 @@ def test_lemur_question_fails(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
-def test_lemur_summarize_succeeds(httpx_mock: HTTPXMock):
+def test_lemur_summarize_succeeds_transcript(httpx_mock: HTTPXMock):
     """
     Tests whether summarizing a transcript via LeMUR succeeds.
     """
@@ -184,6 +284,41 @@ def test_lemur_summarize_succeeds(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
+def test_lemur_summarize_succeeds_input_text(httpx_mock: HTTPXMock):
+    """
+    Tests whether summarizing a transcript via LeMUR succeeds with input text.
+    """
+
+    # create a mock response of a LemurSummaryResponse
+    mock_lemur_summary = factories.generate_dict_factory(
+        factories.LemurSummaryResponse
+    )()
+
+    # mock the specific endpoints
+    httpx_mock.add_response(
+        url=f"{aai.settings.base_url}{ENDPOINT_LEMUR}/summary",
+        status_code=httpx.codes.OK,
+        method="POST",
+        json=mock_lemur_summary,
+    )
+
+    # test input_text input
+    lemur = aai.Lemur()
+    result = lemur.summarize(
+        context="Callers asking for cars", answer_format="TLDR", input_text="Test test"
+    )
+
+    assert isinstance(result, aai.LemurSummaryResponse)
+
+    summary = result.response
+
+    # check the response
+    assert summary == mock_lemur_summary["response"]
+
+    # check whether we mocked everything
+    assert len(httpx_mock.get_requests()) == 1
+
+
 def test_lemur_summarize_fails(httpx_mock: HTTPXMock):
     """
     Tests whether summarizing a transcript via LeMUR fails.
@@ -209,7 +344,7 @@ def test_lemur_summarize_fails(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
-def test_lemur_action_items_succeeds(httpx_mock: HTTPXMock):
+def test_lemur_action_items_succeeds_transcript(httpx_mock: HTTPXMock):
     """
     Tests whether generating action items for a transcript via LeMUR succeeds.
     """
@@ -247,6 +382,43 @@ def test_lemur_action_items_succeeds(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
+def test_lemur_action_items_succeeds_input_text(httpx_mock: HTTPXMock):
+    """
+    Tests whether generating action items for a transcript via LeMUR succeeds.
+    """
+
+    # create a mock response of a LemurActionItemsResponse
+    mock_lemur_action_items = factories.generate_dict_factory(
+        factories.LemurActionItemsResponse
+    )()
+
+    # mock the specific endpoints
+    httpx_mock.add_response(
+        url=f"{aai.settings.base_url}{ENDPOINT_LEMUR}/action-items",
+        status_code=httpx.codes.OK,
+        method="POST",
+        json=mock_lemur_action_items,
+    )
+
+    # test input_text input
+    lemur = aai.Lemur()
+    result = lemur.action_items(
+        context="Customers asking for help with resolving their problem",
+        answer_format="Three bullet points",
+        input_text="Test test",
+    )
+
+    assert isinstance(result, aai.LemurActionItemsResponse)
+
+    action_items = result.response
+
+    # check the response
+    assert action_items == mock_lemur_action_items["response"]
+
+    # check whether we mocked everything
+    assert len(httpx_mock.get_requests()) == 1
+
+
 def test_lemur_action_items_fails(httpx_mock: HTTPXMock):
     """
     Tests whether generating action items for a transcript via LeMUR fails.
@@ -275,7 +447,7 @@ def test_lemur_action_items_fails(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
-def test_lemur_task_succeeds(httpx_mock: HTTPXMock):
+def test_lemur_task_succeeds_transcript(httpx_mock: HTTPXMock):
     """
     Tests whether creating a task request succeeds.
     """
@@ -310,6 +482,38 @@ def test_lemur_task_succeeds(httpx_mock: HTTPXMock):
     assert len(httpx_mock.get_requests()) == 1
 
 
+def test_lemur_task_succeeds_input_text(httpx_mock: HTTPXMock):
+    """
+    Tests whether creating a task request succeeds.
+    """
+
+    # create a mock response of a LemurSummaryResponse
+    mock_lemur_task_response = factories.generate_dict_factory(
+        factories.LemurTaskResponse
+    )()
+
+    # mock the specific endpoints
+    httpx_mock.add_response(
+        url=f"{aai.settings.base_url}{ENDPOINT_LEMUR}/task",
+        status_code=httpx.codes.OK,
+        method="POST",
+        json=mock_lemur_task_response,
+    )
+    # test input_text input
+    lemur = aai.Lemur()
+    result = lemur.task(
+        prompt="Create action items of the meeting", input_text="Test test"
+    )
+
+    # check the response
+    assert isinstance(result, aai.LemurTaskResponse)
+
+    assert result.response == mock_lemur_task_response["response"]
+
+    # check whether we mocked everything
+    assert len(httpx_mock.get_requests()) == 1
+
+
 def test_lemur_ask_coach_fails(httpx_mock: HTTPXMock):
     """
     Tests whether creating a task request fails.