phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-26 11:14:04 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

(test) router.py revert to stable version

Browse source
This commit is contained in:
ishaan-jaff 2024-01-23 09:05:19 -08:00
parent fa2c94aacb
commit 3f37cf0bec
1 changed files with 38 additions and 34 deletions
Download patch file Download diff file Expand all files Collapse all files

72
litellm/tests/test_router.py
View file

@ -783,6 +783,9 @@ def test_reading_keys_os_environ():
assert float(model["litellm_params"]["timeout"]) == float( assert float(model["litellm_params"]["timeout"]) == float(
os.environ["AZURE_TIMEOUT"] os.environ["AZURE_TIMEOUT"]
), f"{model['litellm_params']['timeout']} vs {os.environ['AZURE_TIMEOUT']}" ), f"{model['litellm_params']['timeout']} vs {os.environ['AZURE_TIMEOUT']}"
assert float(model["litellm_params"]["stream_timeout"]) == float(
os.environ["AZURE_STREAM_TIMEOUT"]
), f"{model['litellm_params']['stream_timeout']} vs {os.environ['AZURE_STREAM_TIMEOUT']}"
assert int(model["litellm_params"]["max_retries"]) == int( assert int(model["litellm_params"]["max_retries"]) == int(
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{model['litellm_params']['max_retries']} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{model['litellm_params']['max_retries']} vs {os.environ['AZURE_MAX_RETRIES']}"
@ -791,7 +794,7 @@ def test_reading_keys_os_environ():
async_client: openai.AsyncAzureOpenAI = router.cache.get_cache(f"{model_id}_async_client") # type: ignore async_client: openai.AsyncAzureOpenAI = router.cache.get_cache(f"{model_id}_async_client") # type: ignore
assert async_client.api_key == os.environ["AZURE_API_KEY"] assert async_client.api_key == os.environ["AZURE_API_KEY"]
assert async_client.base_url == os.environ["AZURE_API_BASE"] assert async_client.base_url == os.environ["AZURE_API_BASE"]
assert async_client.max_retries == int( assert async_client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert async_client.timeout == ( assert async_client.timeout == (
@ -804,7 +807,7 @@ def test_reading_keys_os_environ():
stream_async_client: openai.AsyncAzureOpenAI = router.cache.get_cache(f"{model_id}_stream_async_client") # type: ignore stream_async_client: openai.AsyncAzureOpenAI = router.cache.get_cache(f"{model_id}_stream_async_client") # type: ignore
assert stream_async_client.api_key == os.environ["AZURE_API_KEY"] assert stream_async_client.api_key == os.environ["AZURE_API_KEY"]
assert stream_async_client.base_url == os.environ["AZURE_API_BASE"] assert stream_async_client.base_url == os.environ["AZURE_API_BASE"]
assert stream_async_client.max_retries == int( assert stream_async_client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{stream_async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{stream_async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert stream_async_client.timeout == ( assert stream_async_client.timeout == (
@ -816,7 +819,7 @@ def test_reading_keys_os_environ():
client: openai.AzureOpenAI = router.cache.get_cache(f"{model_id}_client") # type: ignore client: openai.AzureOpenAI = router.cache.get_cache(f"{model_id}_client") # type: ignore
assert client.api_key == os.environ["AZURE_API_KEY"] assert client.api_key == os.environ["AZURE_API_KEY"]
assert client.base_url == os.environ["AZURE_API_BASE"] assert client.base_url == os.environ["AZURE_API_BASE"]
assert client.max_retries == int( assert client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert client.timeout == ( assert client.timeout == (
@ -828,7 +831,7 @@ def test_reading_keys_os_environ():
stream_client: openai.AzureOpenAI = router.cache.get_cache(f"{model_id}_stream_client") # type: ignore stream_client: openai.AzureOpenAI = router.cache.get_cache(f"{model_id}_stream_client") # type: ignore
assert stream_client.api_key == os.environ["AZURE_API_KEY"] assert stream_client.api_key == os.environ["AZURE_API_KEY"]
assert stream_client.base_url == os.environ["AZURE_API_BASE"] assert stream_client.base_url == os.environ["AZURE_API_BASE"]
assert stream_client.max_retries == int( assert stream_client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{stream_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{stream_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert stream_client.timeout == ( assert stream_client.timeout == (
@ -880,6 +883,9 @@ def test_reading_openai_keys_os_environ():
assert float(model["litellm_params"]["timeout"]) == float( assert float(model["litellm_params"]["timeout"]) == float(
os.environ["AZURE_TIMEOUT"] os.environ["AZURE_TIMEOUT"]
), f"{model['litellm_params']['timeout']} vs {os.environ['AZURE_TIMEOUT']}" ), f"{model['litellm_params']['timeout']} vs {os.environ['AZURE_TIMEOUT']}"
assert float(model["litellm_params"]["stream_timeout"]) == float(
os.environ["AZURE_STREAM_TIMEOUT"]
), f"{model['litellm_params']['stream_timeout']} vs {os.environ['AZURE_STREAM_TIMEOUT']}"
assert int(model["litellm_params"]["max_retries"]) == int( assert int(model["litellm_params"]["max_retries"]) == int(
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{model['litellm_params']['max_retries']} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{model['litellm_params']['max_retries']} vs {os.environ['AZURE_MAX_RETRIES']}"
@ -887,7 +893,7 @@ def test_reading_openai_keys_os_environ():
model_id = model["model_info"]["id"] model_id = model["model_info"]["id"]
async_client: openai.AsyncOpenAI = router.cache.get_cache(key=f"{model_id}_async_client") # type: ignore async_client: openai.AsyncOpenAI = router.cache.get_cache(key=f"{model_id}_async_client") # type: ignore
assert async_client.api_key == os.environ["OPENAI_API_KEY"] assert async_client.api_key == os.environ["OPENAI_API_KEY"]
assert async_client.max_retries == int( assert async_client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert async_client.timeout == ( assert async_client.timeout == (
@ -899,7 +905,7 @@ def test_reading_openai_keys_os_environ():
stream_async_client: openai.AsyncOpenAI = router.cache.get_cache(key=f"{model_id}_stream_async_client") # type: ignore stream_async_client: openai.AsyncOpenAI = router.cache.get_cache(key=f"{model_id}_stream_async_client") # type: ignore
assert stream_async_client.api_key == os.environ["OPENAI_API_KEY"] assert stream_async_client.api_key == os.environ["OPENAI_API_KEY"]
assert stream_async_client.max_retries == int( assert stream_async_client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{stream_async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{stream_async_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert stream_async_client.timeout == ( assert stream_async_client.timeout == (
@ -910,7 +916,7 @@ def test_reading_openai_keys_os_environ():
print("\n Testing sync client") print("\n Testing sync client")
client: openai.AzureOpenAI = router.cache.get_cache(key=f"{model_id}_client") # type: ignore client: openai.AzureOpenAI = router.cache.get_cache(key=f"{model_id}_client") # type: ignore
assert client.api_key == os.environ["OPENAI_API_KEY"] assert client.api_key == os.environ["OPENAI_API_KEY"]
assert client.max_retries == int( assert client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert client.timeout == ( assert client.timeout == (
@ -921,7 +927,7 @@ def test_reading_openai_keys_os_environ():
print("\n Testing sync stream client") print("\n Testing sync stream client")
stream_client: openai.AzureOpenAI = router.cache.get_cache(key=f"{model_id}_stream_client") # type: ignore stream_client: openai.AzureOpenAI = router.cache.get_cache(key=f"{model_id}_stream_client") # type: ignore
assert stream_client.api_key == os.environ["OPENAI_API_KEY"] assert stream_client.api_key == os.environ["OPENAI_API_KEY"]
assert stream_client.max_retries == int( assert stream_client.max_retries == (
os.environ["AZURE_MAX_RETRIES"] os.environ["AZURE_MAX_RETRIES"]
), f"{stream_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}" ), f"{stream_client.max_retries} vs {os.environ['AZURE_MAX_RETRIES']}"
assert stream_client.timeout == ( assert stream_client.timeout == (
@ -936,29 +942,27 @@ def test_reading_openai_keys_os_environ():
# test_reading_openai_keys_os_environ() # test_reading_openai_keys_os_environ()
# def test_router_timeout():
# model_list = [
def test_router_timeout(): # {
model_list = [ # "model_name": "gpt-3.5-turbo",
{ # "litellm_params": {
"model_name": "gpt-3.5-turbo", # "model": "gpt-3.5-turbo",
"litellm_params": { # "api_key": "os.environ/OPENAI_API_KEY",
"model": "gpt-3.5-turbo", # "timeout": "os.environ/AZURE_TIMEOUT",
"api_key": "os.environ/OPENAI_API_KEY", # "stream_timeout": "os.environ/AZURE_STREAM_TIMEOUT",
"timeout": "os.environ/AZURE_TIMEOUT", # "max_retries": "os.environ/AZURE_MAX_RETRIES",
"stream_timeout": "os.environ/AZURE_STREAM_TIMEOUT", # },
"max_retries": "os.environ/AZURE_MAX_RETRIES", # }
}, # ]
} # router = Router(model_list=model_list)
] # messages = [{"role": "user", "content": "Hey, how's it going?"}]
router = Router(model_list=model_list) # start_time = time.time()
messages = [{"role": "user", "content": "Hey, how's it going?"}] # try:
start_time = time.time() # router.completion(
try: # model="gpt-3.5-turbo", messages=messages, max_tokens=500, timeout=1
router.completion( # )
model="gpt-3.5-turbo", messages=messages, max_tokens=500, timeout=1 # except litellm.exceptions.Timeout as e:
) # pass
except litellm.exceptions.Timeout as e: # end_time = time.time()
pass # assert end_time - start_time < 1.1
end_time = time.time()
assert end_time - start_time < 1.1