[AUR-401] Disable Haystack telemetry with monkey patching (#1)
Sample Haystack log when running a pipeline. Note: the `pipeline.classname` can leak company information.
```json
{
"hardware.cpus": 16,
"hardware.gpus": 0,
"libraries.colab": false,
"libraries.cuda": false,
"libraries.haystack": "1.20.0rc0",
"libraries.ipython": false,
"libraries.pytest": false,
"libraries.ray": false,
"libraries.torch": false,
"libraries.transformers": "4.31.0",
"os.containerized": false,
"os.family": "Linux",
"os.machine": "x86_64",
"os.version": "6.2.0-26-generic",
"pipeline.classname": "TempPipeline",
"pipeline.config_hash": "07a8eddd5a6e512c0d898c6d9f445ed9",
"pipeline.nodes.PromptNode": 1,
"pipeline.nodes.Shaper": 1,
"pipeline.nodes.WebRetriever": 1,
"pipeline.run_parameters.debug": false,
"pipeline.run_parameters.documents": [
0
],
"pipeline.run_parameters.file_paths": 0,
"pipeline.run_parameters.labels": 0,
"pipeline.run_parameters.meta": 1,
"pipeline.run_parameters.params": false,
"pipeline.run_parameters.queries": true,
"pipeline.runs": 1,
"pipeline.type": "Query",
"python.version": "3.10.12"
}
```
Solution: Haystack telemetry uses the `telemetry` variable, `posthog` library and `HAYSTACK_TELEMETRY_ENABLED` envar. We set the envar to False and make sure the relevant objects are disabled.
2023-08-22 10:02:46 +07:00
|
|
|
import codecs
|
|
|
|
|
import re
|
|
|
|
|
from pathlib import Path
|
|
|
|
|
|
|
|
|
|
import setuptools
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def read(file_path: str) -> str:
|
|
|
|
|
return codecs.open(file_path, "r").read()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def get_version() -> str:
|
|
|
|
|
version_file = read(str(Path("kotaemon", "__init__.py")))
|
|
|
|
|
match = re.search(r"^__version__ = ['\"]([^'\"]*)['\"]", version_file, re.M)
|
|
|
|
|
if match:
|
|
|
|
|
return match.group(1)
|
|
|
|
|
raise RuntimeError("Cannot find verison string")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
setuptools.setup(
|
|
|
|
|
name="kotaemon",
|
|
|
|
|
version=get_version(),
|
|
|
|
|
author="john",
|
|
|
|
|
author_email="john@cinnamon.com",
|
|
|
|
|
description="Kotaemon core library for AI development",
|
|
|
|
|
long_description=read("README.md"),
|
|
|
|
|
long_description_content_type="text/markdown",
|
|
|
|
|
url="https://github.com/Cinnamon/kotaemon/",
|
|
|
|
|
packages=setuptools.find_packages(exclude=("tests", "tests.*")),
|
|
|
|
|
install_requires=[
|
[AUR-385, AUR-388] Declare BaseComponent and decide LLM call interface (#2)
- Use cases related to LLM call: https://cinnamon-ai.atlassian.net/browse/AUR-388?focusedCommentId=34873
- Sample usages: `test_llms_chat_models.py` and `test_llms_completion_models.py`:
```python
from kotaemon.llms.chats.openai import AzureChatOpenAI
model = AzureChatOpenAI(
openai_api_base="https://test.openai.azure.com/",
openai_api_key="some-key",
openai_api_version="2023-03-15-preview",
deployment_name="gpt35turbo",
temperature=0,
request_timeout=60,
)
output = model("hello world")
```
For the LLM-call component, I decide to wrap around Langchain's LLM models and Langchain's Chat models. And set the interface as follow:
- Completion LLM component:
```python
class CompletionLLM:
def run_raw(self, text: str) -> LLMInterface:
# Run text completion: str in -> LLMInterface out
def run_batch_raw(self, text: list[str]) -> list[LLMInterface]:
# Run text completion in batch: list[str] in -> list[LLMInterface] out
# run_document and run_batch_document just reuse run_raw and run_batch_raw, due to unclear use case
```
- Chat LLM component:
```python
class ChatLLM:
def run_raw(self, text: str) -> LLMInterface:
# Run chat completion (no chat history): str in -> LLMInterface out
def run_batch_raw(self, text: list[str]) -> list[LLMInterface]:
# Run chat completion in batch mode (no chat history): list[str] in -> list[LLMInterface] out
def run_document(self, text: list[BaseMessage]) -> LLMInterface:
# Run chat completion (with chat history): list[langchain's BaseMessage] in -> LLMInterface out
def run_batch_document(self, text: list[list[BaseMessage]]) -> list[LLMInterface]:
# Run chat completion in batch mode (with chat history): list[list[langchain's BaseMessage]] in -> list[LLMInterface] out
```
- The LLMInterface is as follow:
```python
@dataclass
class LLMInterface:
text: list[str]
completion_tokens: int = -1
total_tokens: int = -1
prompt_tokens: int = -1
logits: list[list[float]] = field(default_factory=list)
```
2023-08-29 15:47:12 +07:00
|
|
|
"farm-haystack==1.19.0",
|
|
|
|
|
"langchain",
|
|
|
|
|
"theflow",
|
[AUR-401] Disable Haystack telemetry with monkey patching (#1)
Sample Haystack log when running a pipeline. Note: the `pipeline.classname` can leak company information.
```json
{
"hardware.cpus": 16,
"hardware.gpus": 0,
"libraries.colab": false,
"libraries.cuda": false,
"libraries.haystack": "1.20.0rc0",
"libraries.ipython": false,
"libraries.pytest": false,
"libraries.ray": false,
"libraries.torch": false,
"libraries.transformers": "4.31.0",
"os.containerized": false,
"os.family": "Linux",
"os.machine": "x86_64",
"os.version": "6.2.0-26-generic",
"pipeline.classname": "TempPipeline",
"pipeline.config_hash": "07a8eddd5a6e512c0d898c6d9f445ed9",
"pipeline.nodes.PromptNode": 1,
"pipeline.nodes.Shaper": 1,
"pipeline.nodes.WebRetriever": 1,
"pipeline.run_parameters.debug": false,
"pipeline.run_parameters.documents": [
0
],
"pipeline.run_parameters.file_paths": 0,
"pipeline.run_parameters.labels": 0,
"pipeline.run_parameters.meta": 1,
"pipeline.run_parameters.params": false,
"pipeline.run_parameters.queries": true,
"pipeline.runs": 1,
"pipeline.type": "Query",
"python.version": "3.10.12"
}
```
Solution: Haystack telemetry uses the `telemetry` variable, `posthog` library and `HAYSTACK_TELEMETRY_ENABLED` envar. We set the envar to False and make sure the relevant objects are disabled.
2023-08-22 10:02:46 +07:00
|
|
|
],
|
|
|
|
|
extras_require={
|
|
|
|
|
"dev": [
|
[AUR-385, AUR-388] Declare BaseComponent and decide LLM call interface (#2)
- Use cases related to LLM call: https://cinnamon-ai.atlassian.net/browse/AUR-388?focusedCommentId=34873
- Sample usages: `test_llms_chat_models.py` and `test_llms_completion_models.py`:
```python
from kotaemon.llms.chats.openai import AzureChatOpenAI
model = AzureChatOpenAI(
openai_api_base="https://test.openai.azure.com/",
openai_api_key="some-key",
openai_api_version="2023-03-15-preview",
deployment_name="gpt35turbo",
temperature=0,
request_timeout=60,
)
output = model("hello world")
```
For the LLM-call component, I decide to wrap around Langchain's LLM models and Langchain's Chat models. And set the interface as follow:
- Completion LLM component:
```python
class CompletionLLM:
def run_raw(self, text: str) -> LLMInterface:
# Run text completion: str in -> LLMInterface out
def run_batch_raw(self, text: list[str]) -> list[LLMInterface]:
# Run text completion in batch: list[str] in -> list[LLMInterface] out
# run_document and run_batch_document just reuse run_raw and run_batch_raw, due to unclear use case
```
- Chat LLM component:
```python
class ChatLLM:
def run_raw(self, text: str) -> LLMInterface:
# Run chat completion (no chat history): str in -> LLMInterface out
def run_batch_raw(self, text: list[str]) -> list[LLMInterface]:
# Run chat completion in batch mode (no chat history): list[str] in -> list[LLMInterface] out
def run_document(self, text: list[BaseMessage]) -> LLMInterface:
# Run chat completion (with chat history): list[langchain's BaseMessage] in -> LLMInterface out
def run_batch_document(self, text: list[list[BaseMessage]]) -> list[LLMInterface]:
# Run chat completion in batch mode (with chat history): list[list[langchain's BaseMessage]] in -> list[LLMInterface] out
```
- The LLMInterface is as follow:
```python
@dataclass
class LLMInterface:
text: list[str]
completion_tokens: int = -1
total_tokens: int = -1
prompt_tokens: int = -1
logits: list[list[float]] = field(default_factory=list)
```
2023-08-29 15:47:12 +07:00
|
|
|
"ipython",
|
[AUR-401] Disable Haystack telemetry with monkey patching (#1)
Sample Haystack log when running a pipeline. Note: the `pipeline.classname` can leak company information.
```json
{
"hardware.cpus": 16,
"hardware.gpus": 0,
"libraries.colab": false,
"libraries.cuda": false,
"libraries.haystack": "1.20.0rc0",
"libraries.ipython": false,
"libraries.pytest": false,
"libraries.ray": false,
"libraries.torch": false,
"libraries.transformers": "4.31.0",
"os.containerized": false,
"os.family": "Linux",
"os.machine": "x86_64",
"os.version": "6.2.0-26-generic",
"pipeline.classname": "TempPipeline",
"pipeline.config_hash": "07a8eddd5a6e512c0d898c6d9f445ed9",
"pipeline.nodes.PromptNode": 1,
"pipeline.nodes.Shaper": 1,
"pipeline.nodes.WebRetriever": 1,
"pipeline.run_parameters.debug": false,
"pipeline.run_parameters.documents": [
0
],
"pipeline.run_parameters.file_paths": 0,
"pipeline.run_parameters.labels": 0,
"pipeline.run_parameters.meta": 1,
"pipeline.run_parameters.params": false,
"pipeline.run_parameters.queries": true,
"pipeline.runs": 1,
"pipeline.type": "Query",
"python.version": "3.10.12"
}
```
Solution: Haystack telemetry uses the `telemetry` variable, `posthog` library and `HAYSTACK_TELEMETRY_ENABLED` envar. We set the envar to False and make sure the relevant objects are disabled.
2023-08-22 10:02:46 +07:00
|
|
|
"pytest",
|
|
|
|
|
"pre-commit",
|
|
|
|
|
"black",
|
|
|
|
|
"flake8",
|
|
|
|
|
"sphinx",
|
|
|
|
|
"coverage",
|
[AUR-385, AUR-388] Declare BaseComponent and decide LLM call interface (#2)
- Use cases related to LLM call: https://cinnamon-ai.atlassian.net/browse/AUR-388?focusedCommentId=34873
- Sample usages: `test_llms_chat_models.py` and `test_llms_completion_models.py`:
```python
from kotaemon.llms.chats.openai import AzureChatOpenAI
model = AzureChatOpenAI(
openai_api_base="https://test.openai.azure.com/",
openai_api_key="some-key",
openai_api_version="2023-03-15-preview",
deployment_name="gpt35turbo",
temperature=0,
request_timeout=60,
)
output = model("hello world")
```
For the LLM-call component, I decide to wrap around Langchain's LLM models and Langchain's Chat models. And set the interface as follow:
- Completion LLM component:
```python
class CompletionLLM:
def run_raw(self, text: str) -> LLMInterface:
# Run text completion: str in -> LLMInterface out
def run_batch_raw(self, text: list[str]) -> list[LLMInterface]:
# Run text completion in batch: list[str] in -> list[LLMInterface] out
# run_document and run_batch_document just reuse run_raw and run_batch_raw, due to unclear use case
```
- Chat LLM component:
```python
class ChatLLM:
def run_raw(self, text: str) -> LLMInterface:
# Run chat completion (no chat history): str in -> LLMInterface out
def run_batch_raw(self, text: list[str]) -> list[LLMInterface]:
# Run chat completion in batch mode (no chat history): list[str] in -> list[LLMInterface] out
def run_document(self, text: list[BaseMessage]) -> LLMInterface:
# Run chat completion (with chat history): list[langchain's BaseMessage] in -> LLMInterface out
def run_batch_document(self, text: list[list[BaseMessage]]) -> list[LLMInterface]:
# Run chat completion in batch mode (with chat history): list[list[langchain's BaseMessage]] in -> list[LLMInterface] out
```
- The LLMInterface is as follow:
```python
@dataclass
class LLMInterface:
text: list[str]
completion_tokens: int = -1
total_tokens: int = -1
prompt_tokens: int = -1
logits: list[list[float]] = field(default_factory=list)
```
2023-08-29 15:47:12 +07:00
|
|
|
|
|
|
|
|
# optional dependency needed for test
|
|
|
|
|
"openai"
|
|
|
|
|
],
|
[AUR-401] Disable Haystack telemetry with monkey patching (#1)
Sample Haystack log when running a pipeline. Note: the `pipeline.classname` can leak company information.
```json
{
"hardware.cpus": 16,
"hardware.gpus": 0,
"libraries.colab": false,
"libraries.cuda": false,
"libraries.haystack": "1.20.0rc0",
"libraries.ipython": false,
"libraries.pytest": false,
"libraries.ray": false,
"libraries.torch": false,
"libraries.transformers": "4.31.0",
"os.containerized": false,
"os.family": "Linux",
"os.machine": "x86_64",
"os.version": "6.2.0-26-generic",
"pipeline.classname": "TempPipeline",
"pipeline.config_hash": "07a8eddd5a6e512c0d898c6d9f445ed9",
"pipeline.nodes.PromptNode": 1,
"pipeline.nodes.Shaper": 1,
"pipeline.nodes.WebRetriever": 1,
"pipeline.run_parameters.debug": false,
"pipeline.run_parameters.documents": [
0
],
"pipeline.run_parameters.file_paths": 0,
"pipeline.run_parameters.labels": 0,
"pipeline.run_parameters.meta": 1,
"pipeline.run_parameters.params": false,
"pipeline.run_parameters.queries": true,
"pipeline.runs": 1,
"pipeline.type": "Query",
"python.version": "3.10.12"
}
```
Solution: Haystack telemetry uses the `telemetry` variable, `posthog` library and `HAYSTACK_TELEMETRY_ENABLED` envar. We set the envar to False and make sure the relevant objects are disabled.
2023-08-22 10:02:46 +07:00
|
|
|
},
|
|
|
|
|
entry_points={"console_scripts": ["kh=kotaemon.cli:main"]},
|
|
|
|
|
python_requires=">=3",
|
|
|
|
|
classifiers=[
|
|
|
|
|
"Programming Language :: Python :: 3",
|
|
|
|
|
"License :: OSI Approved :: MIT License",
|
|
|
|
|
"Operating System :: OS Independent",
|
|
|
|
|
],
|
|
|
|
|
include_package_data=True,
|
|
|
|
|
)
|