add konko chat_model files (#10267)
_Thank you to the LangChain team for the great project and in advance
for your review. Let me know if I can provide any other additional
information or do things differently in the future to make your lives
easier 🙏 _
@hwchase17 please let me know if you're not the right person to review 😄
This PR enables LangChain to access the Konko API via the chat_models
API wrapper.
Konko API is a fully managed API designed to help application
developers:
1. Select the right LLM(s) for their application
2. Prototype with various open-source and proprietary LLMs
3. Move to production in-line with their security, privacy, throughput,
latency SLAs without infrastructure set-up or administration using Konko
AI's SOC 2 compliant infrastructure
_Note on integration tests:_
We added 14 integration tests. They will all fail unless you export the
right API keys. 13 will pass with a KONKO_API_KEY provided and the other
one will pass with a OPENAI_API_KEY provided. When both are provided,
all 14 integration tests pass. If you would like to test this yourself,
please let me know and I can provide some temporary keys.
### Installation and Setup
1. **First you'll need an API key**
2. **Install Konko AI's Python SDK**
1. Enable a Python3.8+ environment
`pip install konko`
3. **Set API Keys**
**Option 1:** Set Environment Variables
You can set environment variables for
1. KONKO_API_KEY (Required)
2. OPENAI_API_KEY (Optional)
In your current shell session, use the export command:
`export KONKO_API_KEY={your_KONKO_API_KEY_here}`
`export OPENAI_API_KEY={your_OPENAI_API_KEY_here} #Optional`
Alternatively, you can add the above lines directly to your shell
startup script (such as .bashrc or .bash_profile for Bash shell and
.zshrc for Zsh shell) to have them set automatically every time a new
shell session starts.
**Option 2:** Set API Keys Programmatically
If you prefer to set your API keys directly within your Python script or
Jupyter notebook, you can use the following commands:
```python
konko.set_api_key('your_KONKO_API_KEY_here')
konko.set_openai_api_key('your_OPENAI_API_KEY_here') # Optional
```
### Calling a model
Find a model on the [[Konko Introduction
page](https://docs.konko.ai/docs#available-models)](https://docs.konko.ai/docs#available-models)
For example, for this [[LLama 2
model](https://docs.konko.ai/docs/meta-llama-2-13b-chat)](https://docs.konko.ai/docs/meta-llama-2-13b-chat).
The model id would be: `"meta-llama/Llama-2-13b-chat-hf"`
Another way to find the list of models running on the Konko instance is
through this
[[endpoint](https://docs.konko.ai/reference/listmodels)](https://docs.konko.ai/reference/listmodels).
From here, we can initialize our model:
```python
chat_instance = ChatKonko(max_tokens=10, model = 'meta-llama/Llama-2-13b-chat-hf')
```
And run it:
```python
msg = HumanMessage(content="Hi")
chat_response = chat_instance([msg])
```
2023-09-08 17:00:55 +00:00
|
|
|
"""Evaluate ChatKonko Interface."""
|
|
|
|
from typing import Any
|
|
|
|
|
|
|
|
import pytest
|
2023-12-11 21:53:30 +00:00
|
|
|
from langchain_core.callbacks import CallbackManager
|
2023-11-21 16:35:29 +00:00
|
|
|
from langchain_core.messages import BaseMessage, HumanMessage, SystemMessage
|
|
|
|
from langchain_core.outputs import ChatGeneration, ChatResult, LLMResult
|
2023-11-20 21:09:30 +00:00
|
|
|
|
2023-12-11 21:53:30 +00:00
|
|
|
from langchain_community.chat_models.konko import ChatKonko
|
add konko chat_model files (#10267)
_Thank you to the LangChain team for the great project and in advance
for your review. Let me know if I can provide any other additional
information or do things differently in the future to make your lives
easier 🙏 _
@hwchase17 please let me know if you're not the right person to review 😄
This PR enables LangChain to access the Konko API via the chat_models
API wrapper.
Konko API is a fully managed API designed to help application
developers:
1. Select the right LLM(s) for their application
2. Prototype with various open-source and proprietary LLMs
3. Move to production in-line with their security, privacy, throughput,
latency SLAs without infrastructure set-up or administration using Konko
AI's SOC 2 compliant infrastructure
_Note on integration tests:_
We added 14 integration tests. They will all fail unless you export the
right API keys. 13 will pass with a KONKO_API_KEY provided and the other
one will pass with a OPENAI_API_KEY provided. When both are provided,
all 14 integration tests pass. If you would like to test this yourself,
please let me know and I can provide some temporary keys.
### Installation and Setup
1. **First you'll need an API key**
2. **Install Konko AI's Python SDK**
1. Enable a Python3.8+ environment
`pip install konko`
3. **Set API Keys**
**Option 1:** Set Environment Variables
You can set environment variables for
1. KONKO_API_KEY (Required)
2. OPENAI_API_KEY (Optional)
In your current shell session, use the export command:
`export KONKO_API_KEY={your_KONKO_API_KEY_here}`
`export OPENAI_API_KEY={your_OPENAI_API_KEY_here} #Optional`
Alternatively, you can add the above lines directly to your shell
startup script (such as .bashrc or .bash_profile for Bash shell and
.zshrc for Zsh shell) to have them set automatically every time a new
shell session starts.
**Option 2:** Set API Keys Programmatically
If you prefer to set your API keys directly within your Python script or
Jupyter notebook, you can use the following commands:
```python
konko.set_api_key('your_KONKO_API_KEY_here')
konko.set_openai_api_key('your_OPENAI_API_KEY_here') # Optional
```
### Calling a model
Find a model on the [[Konko Introduction
page](https://docs.konko.ai/docs#available-models)](https://docs.konko.ai/docs#available-models)
For example, for this [[LLama 2
model](https://docs.konko.ai/docs/meta-llama-2-13b-chat)](https://docs.konko.ai/docs/meta-llama-2-13b-chat).
The model id would be: `"meta-llama/Llama-2-13b-chat-hf"`
Another way to find the list of models running on the Konko instance is
through this
[[endpoint](https://docs.konko.ai/reference/listmodels)](https://docs.konko.ai/reference/listmodels).
From here, we can initialize our model:
```python
chat_instance = ChatKonko(max_tokens=10, model = 'meta-llama/Llama-2-13b-chat-hf')
```
And run it:
```python
msg = HumanMessage(content="Hi")
chat_response = chat_instance([msg])
```
2023-09-08 17:00:55 +00:00
|
|
|
from tests.unit_tests.callbacks.fake_callback_handler import FakeCallbackHandler
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_chat_test() -> None:
|
|
|
|
"""Evaluate basic ChatKonko functionality."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10)
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
chat_response = chat_instance([msg])
|
|
|
|
assert isinstance(chat_response, BaseMessage)
|
|
|
|
assert isinstance(chat_response.content, str)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_chat_test_openai() -> None:
|
|
|
|
"""Evaluate basic ChatKonko functionality."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10, model="gpt-3.5-turbo")
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
chat_response = chat_instance([msg])
|
|
|
|
assert isinstance(chat_response, BaseMessage)
|
|
|
|
assert isinstance(chat_response.content, str)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_model_test() -> None:
|
|
|
|
"""Check how ChatKonko manages model_name."""
|
|
|
|
chat_instance = ChatKonko(model="alpha")
|
|
|
|
assert chat_instance.model == "alpha"
|
|
|
|
chat_instance = ChatKonko(model="beta")
|
|
|
|
assert chat_instance.model == "beta"
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_available_model_test() -> None:
|
|
|
|
"""Check how ChatKonko manages model_name."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10, n=2)
|
|
|
|
res = chat_instance.get_available_models()
|
|
|
|
assert isinstance(res, set)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_system_msg_test() -> None:
|
|
|
|
"""Evaluate ChatKonko's handling of system messages."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10)
|
|
|
|
sys_msg = SystemMessage(content="Initiate user chat.")
|
|
|
|
user_msg = HumanMessage(content="Hi there")
|
|
|
|
chat_response = chat_instance([sys_msg, user_msg])
|
|
|
|
assert isinstance(chat_response, BaseMessage)
|
|
|
|
assert isinstance(chat_response.content, str)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_generation_test() -> None:
|
|
|
|
"""Check ChatKonko's generation ability."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10, n=2)
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
gen_response = chat_instance.generate([[msg], [msg]])
|
|
|
|
assert isinstance(gen_response, LLMResult)
|
|
|
|
assert len(gen_response.generations) == 2
|
|
|
|
for gen_list in gen_response.generations:
|
|
|
|
assert len(gen_list) == 2
|
|
|
|
for gen in gen_list:
|
|
|
|
assert isinstance(gen, ChatGeneration)
|
|
|
|
assert isinstance(gen.text, str)
|
|
|
|
assert gen.text == gen.message.content
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_multiple_outputs_test() -> None:
|
|
|
|
"""Test multiple completions with ChatKonko."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10, n=5)
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
gen_response = chat_instance._generate([msg])
|
|
|
|
assert isinstance(gen_response, ChatResult)
|
|
|
|
assert len(gen_response.generations) == 5
|
|
|
|
for gen in gen_response.generations:
|
|
|
|
assert isinstance(gen.message, BaseMessage)
|
|
|
|
assert isinstance(gen.message.content, str)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_streaming_callback_test() -> None:
|
|
|
|
"""Evaluate streaming's token callback functionality."""
|
|
|
|
callback_instance = FakeCallbackHandler()
|
|
|
|
callback_mgr = CallbackManager([callback_instance])
|
|
|
|
chat_instance = ChatKonko(
|
|
|
|
max_tokens=10,
|
|
|
|
streaming=True,
|
|
|
|
temperature=0,
|
|
|
|
callback_manager=callback_mgr,
|
|
|
|
verbose=True,
|
|
|
|
)
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
chat_response = chat_instance([msg])
|
|
|
|
assert callback_instance.llm_streams > 0
|
|
|
|
assert isinstance(chat_response, BaseMessage)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_streaming_info_test() -> None:
|
|
|
|
"""Ensure generation details are retained during streaming."""
|
|
|
|
|
|
|
|
class TestCallback(FakeCallbackHandler):
|
|
|
|
data_store: dict = {}
|
|
|
|
|
|
|
|
def on_llm_end(self, *args: Any, **kwargs: Any) -> Any:
|
|
|
|
self.data_store["generation"] = args[0]
|
|
|
|
|
|
|
|
callback_instance = TestCallback()
|
|
|
|
callback_mgr = CallbackManager([callback_instance])
|
|
|
|
chat_instance = ChatKonko(
|
|
|
|
max_tokens=2,
|
|
|
|
temperature=0,
|
|
|
|
callback_manager=callback_mgr,
|
|
|
|
)
|
|
|
|
list(chat_instance.stream("hey"))
|
|
|
|
gen_data = callback_instance.data_store["generation"]
|
|
|
|
assert gen_data.generations[0][0].text == " Hey"
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_llm_model_name_test() -> None:
|
|
|
|
"""Check if llm_output has model info."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10)
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
llm_data = chat_instance.generate([[msg]])
|
|
|
|
assert llm_data.llm_output is not None
|
|
|
|
assert llm_data.llm_output["model_name"] == chat_instance.model
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_streaming_model_name_test() -> None:
|
|
|
|
"""Check model info during streaming."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10, streaming=True)
|
|
|
|
msg = HumanMessage(content="Hi")
|
|
|
|
llm_data = chat_instance.generate([[msg]])
|
|
|
|
assert llm_data.llm_output is not None
|
|
|
|
assert llm_data.llm_output["model_name"] == chat_instance.model
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_streaming_param_validation_test() -> None:
|
|
|
|
"""Ensure correct token callback during streaming."""
|
|
|
|
with pytest.raises(ValueError):
|
|
|
|
ChatKonko(
|
|
|
|
max_tokens=10,
|
|
|
|
streaming=True,
|
|
|
|
temperature=0,
|
|
|
|
n=5,
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
def test_konko_additional_args_test() -> None:
|
|
|
|
"""Evaluate extra arguments for ChatKonko."""
|
|
|
|
chat_instance = ChatKonko(extra=3, max_tokens=10)
|
|
|
|
assert chat_instance.max_tokens == 10
|
|
|
|
assert chat_instance.model_kwargs == {"extra": 3}
|
|
|
|
|
|
|
|
chat_instance = ChatKonko(extra=3, model_kwargs={"addition": 2})
|
|
|
|
assert chat_instance.model_kwargs == {"extra": 3, "addition": 2}
|
|
|
|
|
|
|
|
with pytest.raises(ValueError):
|
|
|
|
ChatKonko(extra=3, model_kwargs={"extra": 2})
|
|
|
|
|
|
|
|
with pytest.raises(ValueError):
|
|
|
|
ChatKonko(model_kwargs={"temperature": 0.2})
|
|
|
|
|
|
|
|
with pytest.raises(ValueError):
|
2023-12-18 21:49:46 +00:00
|
|
|
ChatKonko(model_kwargs={"model": "gpt-3.5-turbo-instruct"})
|
add konko chat_model files (#10267)
_Thank you to the LangChain team for the great project and in advance
for your review. Let me know if I can provide any other additional
information or do things differently in the future to make your lives
easier 🙏 _
@hwchase17 please let me know if you're not the right person to review 😄
This PR enables LangChain to access the Konko API via the chat_models
API wrapper.
Konko API is a fully managed API designed to help application
developers:
1. Select the right LLM(s) for their application
2. Prototype with various open-source and proprietary LLMs
3. Move to production in-line with their security, privacy, throughput,
latency SLAs without infrastructure set-up or administration using Konko
AI's SOC 2 compliant infrastructure
_Note on integration tests:_
We added 14 integration tests. They will all fail unless you export the
right API keys. 13 will pass with a KONKO_API_KEY provided and the other
one will pass with a OPENAI_API_KEY provided. When both are provided,
all 14 integration tests pass. If you would like to test this yourself,
please let me know and I can provide some temporary keys.
### Installation and Setup
1. **First you'll need an API key**
2. **Install Konko AI's Python SDK**
1. Enable a Python3.8+ environment
`pip install konko`
3. **Set API Keys**
**Option 1:** Set Environment Variables
You can set environment variables for
1. KONKO_API_KEY (Required)
2. OPENAI_API_KEY (Optional)
In your current shell session, use the export command:
`export KONKO_API_KEY={your_KONKO_API_KEY_here}`
`export OPENAI_API_KEY={your_OPENAI_API_KEY_here} #Optional`
Alternatively, you can add the above lines directly to your shell
startup script (such as .bashrc or .bash_profile for Bash shell and
.zshrc for Zsh shell) to have them set automatically every time a new
shell session starts.
**Option 2:** Set API Keys Programmatically
If you prefer to set your API keys directly within your Python script or
Jupyter notebook, you can use the following commands:
```python
konko.set_api_key('your_KONKO_API_KEY_here')
konko.set_openai_api_key('your_OPENAI_API_KEY_here') # Optional
```
### Calling a model
Find a model on the [[Konko Introduction
page](https://docs.konko.ai/docs#available-models)](https://docs.konko.ai/docs#available-models)
For example, for this [[LLama 2
model](https://docs.konko.ai/docs/meta-llama-2-13b-chat)](https://docs.konko.ai/docs/meta-llama-2-13b-chat).
The model id would be: `"meta-llama/Llama-2-13b-chat-hf"`
Another way to find the list of models running on the Konko instance is
through this
[[endpoint](https://docs.konko.ai/reference/listmodels)](https://docs.konko.ai/reference/listmodels).
From here, we can initialize our model:
```python
chat_instance = ChatKonko(max_tokens=10, model = 'meta-llama/Llama-2-13b-chat-hf')
```
And run it:
```python
msg = HumanMessage(content="Hi")
chat_response = chat_instance([msg])
```
2023-09-08 17:00:55 +00:00
|
|
|
|
|
|
|
|
|
|
|
def test_konko_token_streaming_test() -> None:
|
|
|
|
"""Check token streaming for ChatKonko."""
|
|
|
|
chat_instance = ChatKonko(max_tokens=10)
|
|
|
|
|
|
|
|
for token in chat_instance.stream("Just a test"):
|
|
|
|
assert isinstance(token.content, str)
|