mirror of
https://github.com/hwchase17/langchain.git
synced 2025-05-01 13:26:15 +00:00
a0c2281540
```python
"""python scripts/update_mypy_ruff.py"""
import glob
import tomllib
from pathlib import Path
import toml
import subprocess
import re
ROOT_DIR = Path(__file__).parents[1]
def main():
for path in glob.glob(str(ROOT_DIR / "libs/**/pyproject.toml"), recursive=True):
print(path)
with open(path, "rb") as f:
pyproject = tomllib.load(f)
try:
pyproject["tool"]["poetry"]["group"]["typing"]["dependencies"]["mypy"] = (
"^1.10"
)
pyproject["tool"]["poetry"]["group"]["lint"]["dependencies"]["ruff"] = (
"^0.5"
)
except KeyError:
continue
with open(path, "w") as f:
toml.dump(pyproject, f)
cwd = "/".join(path.split("/")[:-1])
completed = subprocess.run(
"poetry lock --no-update; poetry install --with typing; poetry run mypy . --no-color",
cwd=cwd,
shell=True,
capture_output=True,
text=True,
)
logs = completed.stdout.split("\n")
to_ignore = {}
for l in logs:
if re.match("^(.*)\:(\d+)\: error:.*\[(.*)\]", l):
path, line_no, error_type = re.match(
"^(.*)\:(\d+)\: error:.*\[(.*)\]", l
).groups()
if (path, line_no) in to_ignore:
to_ignore[(path, line_no)].append(error_type)
else:
to_ignore[(path, line_no)] = [error_type]
print(len(to_ignore))
for (error_path, line_no), error_types in to_ignore.items():
all_errors = ", ".join(error_types)
full_path = f"{cwd}/{error_path}"
try:
with open(full_path, "r") as f:
file_lines = f.readlines()
except FileNotFoundError:
continue
file_lines[int(line_no) - 1] = (
file_lines[int(line_no) - 1][:-1] + f" # type: ignore[{all_errors}]\n"
)
with open(full_path, "w") as f:
f.write("".join(file_lines))
subprocess.run(
"poetry run ruff format .; poetry run ruff --select I --fix .",
cwd=cwd,
shell=True,
capture_output=True,
text=True,
)
if __name__ == "__main__":
main()
```
84 lines
2.7 KiB
Python
84 lines
2.7 KiB
Python
"""Test MosaicML API wrapper."""
|
|
|
|
import re
|
|
|
|
import pytest
|
|
|
|
from langchain_community.llms.mosaicml import PROMPT_FOR_GENERATION_FORMAT, MosaicML
|
|
|
|
|
|
def test_mosaicml_llm_call() -> None:
|
|
"""Test valid call to MosaicML."""
|
|
llm = MosaicML(model_kwargs={})
|
|
output = llm.invoke("Say foo:")
|
|
assert isinstance(output, str)
|
|
|
|
|
|
def test_mosaicml_endpoint_change() -> None:
|
|
"""Test valid call to MosaicML."""
|
|
new_url = "https://models.hosted-on.mosaicml.hosting/mpt-30b-instruct/v1/predict"
|
|
llm = MosaicML(endpoint_url=new_url)
|
|
assert llm.endpoint_url == new_url
|
|
output = llm.invoke("Say foo:")
|
|
assert isinstance(output, str)
|
|
|
|
|
|
def test_mosaicml_extra_kwargs() -> None:
|
|
llm = MosaicML(model_kwargs={"max_new_tokens": 1})
|
|
assert llm.model_kwargs == {"max_new_tokens": 1}
|
|
|
|
output = llm.invoke("Say foo:")
|
|
|
|
assert isinstance(output, str)
|
|
|
|
# should only generate one new token (which might be a new line or whitespace token)
|
|
assert len(output.split()) <= 1
|
|
|
|
|
|
def test_instruct_prompt() -> None:
|
|
"""Test instruct prompt."""
|
|
llm = MosaicML(inject_instruction_format=True, model_kwargs={"max_new_tokens": 10})
|
|
instruction = "Repeat the word foo"
|
|
prompt = llm._transform_prompt(instruction)
|
|
expected_prompt = PROMPT_FOR_GENERATION_FORMAT.format(instruction=instruction)
|
|
assert prompt == expected_prompt
|
|
output = llm.invoke(prompt)
|
|
assert isinstance(output, str)
|
|
|
|
|
|
def test_retry_logic() -> None:
|
|
"""Tests that two queries (which would usually exceed the rate limit) works"""
|
|
llm = MosaicML(inject_instruction_format=True, model_kwargs={"max_new_tokens": 10})
|
|
instruction = "Repeat the word foo"
|
|
prompt = llm._transform_prompt(instruction)
|
|
expected_prompt = PROMPT_FOR_GENERATION_FORMAT.format(instruction=instruction)
|
|
assert prompt == expected_prompt
|
|
output = llm.invoke(prompt)
|
|
assert isinstance(output, str)
|
|
output = llm.invoke(prompt)
|
|
assert isinstance(output, str)
|
|
|
|
|
|
def test_short_retry_does_not_loop() -> None:
|
|
"""Tests that two queries with a short retry sleep does not infinite loop"""
|
|
llm = MosaicML(
|
|
inject_instruction_format=True,
|
|
model_kwargs={"do_sample": False},
|
|
retry_sleep=0.1,
|
|
)
|
|
instruction = "Repeat the word foo"
|
|
prompt = llm._transform_prompt(instruction)
|
|
expected_prompt = PROMPT_FOR_GENERATION_FORMAT.format(instruction=instruction)
|
|
assert prompt == expected_prompt
|
|
|
|
with pytest.raises(
|
|
ValueError,
|
|
match=re.escape(
|
|
"Error raised by inference API: rate limit exceeded.\nResponse: You have "
|
|
"reached maximum request limit.\n"
|
|
),
|
|
):
|
|
for _ in range(10):
|
|
output = llm.invoke(prompt)
|
|
assert isinstance(output, str)
|