lmnr 0.2.3__py3-none-any.whl → 0.2.4__py3-none-any.whl

lmnr/sdk/remote_debugger.py

@@ -1,6 +1,6 @@
 from typing import Callable, Optional
 from websockets.sync.client import connect
-from lmnr.types import NodeInput, RegisterDebuggerRequest, SDKError, ToolCall
+from lmnr.types import DeregisterDebuggerRequest, NodeInput, RegisterDebuggerRequest, SDKError, ToolCall
 import uuid
 import json
 from threading import Thread
@@ -32,7 +32,7 @@ class RemoteDebugger:
         return self.session
     
     def _run(self):
-        request = RegisterDebuggerRequest(debugger_session_id=self.session)
+        request = RegisterDebuggerRequest(debuggerSessionId=self.session)
         with connect(
             self.url,
             additional_headers={
@@ -71,6 +71,7 @@ class RemoteDebugger:
                         pass
                     response = tool(**arguments)
                     websocket.send(json.dumps(response))
+            websocket.send(DeregisterDebuggerRequest(debuggerSessionId=self.session, deregister=True).model_dump_json())
 
     def _generate_session_id(self) -> str:
         return uuid.uuid4().urn[9:]
@@ -80,7 +81,7 @@ class RemoteDebugger:
 f"""
 ========================================
 Debugger Session ID:
-{self.session_id}
+{self.session}
 ========================================
 """
     

lmnr/types.py

@@ -51,5 +51,10 @@ class ToolCall(pydantic.BaseModel):
     type: Optional[str]
     function: ToolCallRequest
 
+# TODO: allow snake_case and manually convert to camelCase
 class RegisterDebuggerRequest(pydantic.BaseModel):
-    debugger_session_id: str
+    debuggerSessionId: str
+
+class DeregisterDebuggerRequest(pydantic.BaseModel):
+    debuggerSessionId: str
+    deregister: bool

lmnr-0.2.4.dist-info/METADATA

@@ -0,0 +1,186 @@
+Metadata-Version: 2.1
+Name: lmnr
+Version: 0.2.4
+Summary: Python SDK for Laminar AI
+License: Apache-2.0
+Author: lmnr.ai
+Requires-Python: >=3.9,<4.0
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Requires-Dist: black (>=24.4.2,<25.0.0)
+Requires-Dist: click (>=8.1.7,<9.0.0)
+Requires-Dist: cookiecutter (>=2.6.0,<3.0.0)
+Requires-Dist: pydantic (>=2.7.4,<3.0.0)
+Requires-Dist: python-dotenv (>=1.0.1,<2.0.0)
+Requires-Dist: requests (>=2.32.3,<3.0.0)
+Requires-Dist: urllib3 (==1.26.6)
+Requires-Dist: websockets (>=12.0,<13.0)
+Description-Content-Type: text/markdown
+
+# Python SDK for Laminar AI
+
+## Quickstart
+```sh
+python3 -m venv .myenv
+source .myenv/bin/activate  # or use your favorite env management tool
+
+pip install lmnr
+```
+
+## Features
+
+- Make Laminar endpoint calls from your Python code
+- Make Laminar endpoint calls that can run your own functions as tools
+- CLI to generate code from pipelines you build on Laminar
+- `LaminarRemoteDebugger` to execute your own functions while you test your flows in workshop
+
+## Making Laminar endpoint calls
+
+After you are ready to use your pipeline in your code, deploy it in Laminar following the [docs](https://docs.lmnr.ai/pipeline/run-save-deploy#deploying-a-pipeline-version).
+
+Once your pipeline is deployed, you can call it from Python in just a few lines.
+
+Example use:
+
+```python
+from lmnr import Laminar
+
+l = Laminar('<YOUR_PROJECT_API_KEY>')
+result = l.run(
+    endpoint = 'my_endpoint_name',
+    inputs = {'input_node_name': 'some_value'},
+    # all environment variables
+    env = {'OPENAI_API_KEY': 'sk-some-key'},
+    # any metadata to attach to this run's trace
+    metadata = {'session_id': 'your_custom_session_id'}
+)
+```
+
+Resulting in:
+
+```python
+>>> result
+EndpointRunResponse(
+    outputs={'output': {'value': [ChatMessage(role='user', content='hello')]}},
+    # useful to locate your trace
+    run_id='53b012d5-5759-48a6-a9c5-0011610e3669'
+)
+```
+
+## Making calls to pipelines that run your own logic
+
+If your pipeline contains tool call nodes, they will be able to call your local code.
+The only difference is that you need to pass references
+to the functions you want to call right into our SDK.
+
+Example use:
+
+```python
+from lmnr import Laminar, NodeInput
+
+# adding **kwargs is safer, in case an LLM produces more arguments than needed
+def my_tool(arg1: string, arg2: string, **kwargs) -> NodeInput {
+    return f'{arg1}&{arg2}'
+}
+
+l = Laminar('<YOUR_PROJECT_API_KEY>')
+result = l.run(
+    endpoint = 'my_endpoint_name',
+    inputs = {'input_node_name': 'some_value'},
+    # all environment variables
+    env = {'OPENAI_API_KEY': '<YOUR_MODEL_PROVIDER_KEY>'},
+    # any metadata to attach to this run's trace
+    metadata = {'session_id': 'your_custom_session_id'},
+    # specify as many tools as needed.
+    # Each tool name must match tool node name in the pipeline
+    tools=[my_tool]
+)
+```
+
+## LaminarRemoteDebugger
+
+If your pipeline contains tool call nodes, they will be able to call your local code.
+If you want to test them from the Laminar workshop in your browser, you can attach to your
+locally running debugger.
+
+### Step by step instructions to use `LaminarRemoteDebugger`:
+
+#### 1. Create your pipeline with tool call nodes
+
+Add tool calls to your pipeline; node names must match the functions you want to call.
+
+#### 2. Start LaminarRemoteDebugger in your code
+
+Example:
+
+```python
+from lmnr import LaminarRemoteDebugger, NodeInput
+
+# adding **kwargs is safer, in case an LLM produces more arguments than needed
+def my_tool(arg1: string, arg2: string, **kwargs) -> NodeInput {
+    return f'{arg1}&{arg2}'
+}
+
+debugger = LaminarRemoteDebugger('<YOUR_PROJECT_API_KEY>', [my_tool])
+session_id = debugger.start()  # the session id will also be printed to console
+```
+
+This will establish a connection with Laminar API and allow for the pipeline execution
+to call your local functions.
+
+#### 3. Link lmnr.ai workshop to your debugger
+
+Set up `DEBUGGER_SESSION_ID` environment variable in your pipeline.
+
+#### 4. Run and experiment
+
+You can run as many sessions as you need, experimenting with your flows.
+
+#### 5. Stop the debugger
+
+In order to stop the session, do
+
+```python
+debugger.stop()
+```
+
+## CLI for code generation
+
+### Basic usage
+
+```
+lmnr pull <pipeline_name> <pipeline_version_name> --project-api-key <PROJECT_API_KEY>
+```
+
+Note that `lmnr` CLI command will only be available from within the virtual environment
+where you have installed the package.
+
+To import your pipeline
+```python
+# submodule with the name of your pipeline will be generated in lmnr_engine.pipelines
+from lmnr_engine.pipelines.my_custom_pipeline import MyCustomPipeline
+
+
+pipeline = MyCustomPipeline()
+res = pipeline.run(
+    inputs={
+        "instruction": "Write me a short linkedin post about a dev tool for LLM developers"
+    },
+    env={
+        "OPENAI_API_KEY": <OPENAI_API_KEY>,
+    }
+)
+print(f"RESULT:\n{res}")
+```
+
+### Current functionality
+- Supports graph generation for graphs with Input, Output, and LLM nodes only
+- For LLM nodes, it only supports OpenAI and Anthropic models and doesn't support structured output
+
+## PROJECT_API_KEY
+
+Read more [here](https://docs.lmnr.ai/api-reference/introduction#authentication) on how to get `PROJECT_API_KEY`.

{lmnr-0.2.3.dist-info → lmnr-0.2.4.dist-info}/RECORD

@@ -19,10 +19,10 @@ lmnr/cli/{{cookiecutter.lmnr_pipelines_dir_name}}/pipelines/{{cookiecutter.pipel
 lmnr/cli/{{cookiecutter.lmnr_pipelines_dir_name}}/pipelines/{{cookiecutter.pipeline_dir_name}}/{{cookiecutter.pipeline_dir_name}}.py,sha256=WG-ZMofPpGXCx5jdWVry3_XBzcKjqn8ZycFSiWEOBPg,2858
 lmnr/cli/{{cookiecutter.lmnr_pipelines_dir_name}}/types.py,sha256=iWuflMV7TiaBPs6-B-BlrovvWpZgHGGHK0v8rSqER7A,997
 lmnr/sdk/endpoint.py,sha256=tT6-w-mwbh4BAwnj5G0pCVE_Sz8EUzZmpBtacm_T2pE,6359
-lmnr/sdk/remote_debugger.py,sha256=eXbfnl0RoBcGCViajIFsY5W28HQIf791HTffbWBNJY4,3156
-lmnr/types.py,sha256=f_46qjFK_jJziNFEzN-ip8bgPHxkohib2xGesNOdyxs,1526
-lmnr-0.2.3.dist-info/LICENSE,sha256=67b_wJHVV1CBaWkrKFWU1wyqTPSdzH77Ls-59631COg,10411
-lmnr-0.2.3.dist-info/METADATA,sha256=GdJL6YTZ3VnbuQvkz9kPN5QYrC4wBFdkpwq_Er79P28,2258
-lmnr-0.2.3.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-lmnr-0.2.3.dist-info/entry_points.txt,sha256=Qg7ZRax4k-rcQsZ26XRYQ8YFSBiyY2PNxYfq4a6PYXI,41
-lmnr-0.2.3.dist-info/RECORD,,
+lmnr/sdk/remote_debugger.py,sha256=lAL7t0DXADLMTLsh6AN3rqVYf_A-aeG_mMbyU-G2wTk,3299
+lmnr/types.py,sha256=Pi6R5qMmN4fiwZR0TzP4rxfSP6rFv4iouZ_OKnJsmhA,1685
+lmnr-0.2.4.dist-info/LICENSE,sha256=67b_wJHVV1CBaWkrKFWU1wyqTPSdzH77Ls-59631COg,10411
+lmnr-0.2.4.dist-info/METADATA,sha256=BUDTL9WKd18EieHeb47or0gwjGd1dkSz0nJsvx6AFiw,5494
+lmnr-0.2.4.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+lmnr-0.2.4.dist-info/entry_points.txt,sha256=Qg7ZRax4k-rcQsZ26XRYQ8YFSBiyY2PNxYfq4a6PYXI,41
+lmnr-0.2.4.dist-info/RECORD,,

lmnr-0.2.3.dist-info/METADATA

@@ -1,78 +0,0 @@
-Metadata-Version: 2.1
-Name: lmnr
-Version: 0.2.3
-Summary: Python SDK for Laminar AI
-License: Apache-2.0
-Author: lmnr.ai
-Requires-Python: >=3.9,<4.0
-Classifier: License :: OSI Approved :: Apache Software License
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Requires-Dist: black (>=24.4.2,<25.0.0)
-Requires-Dist: click (>=8.1.7,<9.0.0)
-Requires-Dist: cookiecutter (>=2.6.0,<3.0.0)
-Requires-Dist: pydantic (>=2.7.4,<3.0.0)
-Requires-Dist: python-dotenv (>=1.0.1,<2.0.0)
-Requires-Dist: requests (>=2.32.3,<3.0.0)
-Requires-Dist: urllib3 (==1.26.6)
-Requires-Dist: websockets (>=12.0,<13.0)
-Description-Content-Type: text/markdown
-
-# Python SDK for Laminar AI
-
-Example use:
-
-```python
-from lmnr import Laminar
-
-l = Laminar('<YOUR_PROJECT_API_KEY>')
-result = l.run(
-    endpoint = 'my_endpoint_name',
-    inputs = {'input_node_name': 'some_value'},
-    env = {'OPENAI_API_KEY': 'sk-some-key'},
-    metadata = {'session_id': 'your_custom_session_id'}
-)
-```
-
-Resulting in:
-
-```python
->>> result
-EndpointRunResponse(outputs={'output': {'value': [ChatMessage(role='user', content='hello')]}}, run_id='53b012d5-5759-48a6-a9c5-0011610e3669')
-```
-
-## CLI for code generation
-
-### Basic usage
-
-```
-lmnr pull <pipeline_name> <pipeline_version_name> --project-api-key <PROJECT_API_KEY>
-```
-
-Read more [here](https://docs.lmnr.ai/api-reference/introduction#authentication) on how to get `PROJECT_API_KEY`.
-
-To import your pipeline
-```python
-# submodule with the name of your pipeline will be generated in lmnr_engine.pipelines
-from lmnr_engine.pipelines.my_custom_pipeline import MyCustomPipeline
-
-
-pipeline = MyCustomPipeline()
-res = pipeline.run(
-    inputs={
-        "instruction": "Write me a short linked post about dev tool for LLM developers which they'll love"
-    },
-    env={
-        "OPENAI_API_KEY": <OPENAI_API_KEY>,
-    }
-)
-print(f"RESULT:\n{res}")
-```
-
-### Current functionality
-- Supports graph generation for graphs with Input, Output, and LLM nodes only
-- For LLM nodes, it only supports OpenAI and Anthropic models and doesn't support structured output
-