Merge pull request #4043 from BerriAI/litellm_enforce_params

[Feat] Enterprise - Enforce Params in request to LiteLLM Proxy
This commit is contained in:
Ishaan Jaff 2024-06-07 08:52:46 -07:00 committed by GitHub
commit 0e7cda8c73
5 changed files with 204 additions and 1 deletions

View file

@ -14,6 +14,7 @@ Features:
- ✅ [SSO for Admin UI](./ui.md#✨-enterprise-features)
- ✅ [Audit Logs](#audit-logs)
- ✅ [Tracking Spend for Custom Tags](#tracking-spend-for-custom-tags)
- ✅ [Enforce Required Params for LLM Requests (ex. Reject requests missing ["metadata"]["generation_name"])](#enforce-required-params-for-llm-requests)
- ✅ [Content Moderation with LLM Guard, LlamaGuard, Google Text Moderations](#content-moderation)
- ✅ [Prompt Injection Detection (with LakeraAI API)](#prompt-injection-detection---lakeraai)
- ✅ [Custom Branding + Routes on Swagger Docs](#swagger-docs---custom-routes--branding)
@ -204,6 +205,109 @@ curl -X GET "http://0.0.0.0:4000/spend/tags" \
```
## Enforce Required Params for LLM Requests
Use this when you want to enforce all requests to include certain params. Example you need all requests to include the `user` and `["metadata]["generation_name"]` params.
**Step 1** Define all Params you want to enforce on config.yaml
This means `["user"]` and `["metadata]["generation_name"]` are required in all LLM Requests to LiteLLM
```yaml
general_settings:
master_key: sk-1234
enforced_params:
- user
- metadata.generation_name
```
Start LiteLLM Proxy
**Step 2 Verify if this works**
<Tabs>
<TabItem value="bad" label="Invalid Request (No `user` passed)">
```shell
curl --location 'http://localhost:4000/chat/completions' \
--header 'Authorization: Bearer sk-5fmYeaUEbAMpwBNT-QpxyA' \
--header 'Content-Type: application/json' \
--data '{
"model": "gpt-3.5-turbo",
"messages": [
{
"role": "user",
"content": "hi"
}
]
}'
```
Expected Response
```shell
{"error":{"message":"Authentication Error, BadRequest please pass param=user in request body. This is a required param","type":"auth_error","param":"None","code":401}}%
```
</TabItem>
<TabItem value="bad2" label="Invalid Request (No `metadata` passed)">
```shell
curl --location 'http://localhost:4000/chat/completions' \
--header 'Authorization: Bearer sk-5fmYeaUEbAMpwBNT-QpxyA' \
--header 'Content-Type: application/json' \
--data '{
"model": "gpt-3.5-turbo",
"user": "gm",
"messages": [
{
"role": "user",
"content": "hi"
}
],
"metadata": {}
}'
```
Expected Response
```shell
{"error":{"message":"Authentication Error, BadRequest please pass param=[metadata][generation_name] in request body. This is a required param","type":"auth_error","param":"None","code":401}}%
```
</TabItem>
<TabItem value="good" label="Valid Request">
```shell
curl --location 'http://localhost:4000/chat/completions' \
--header 'Authorization: Bearer sk-5fmYeaUEbAMpwBNT-QpxyA' \
--header 'Content-Type: application/json' \
--data '{
"model": "gpt-3.5-turbo",
"user": "gm",
"messages": [
{
"role": "user",
"content": "hi"
}
],
"metadata": {"generation_name": "prod-app"}
}'
```
Expected Response
```shell
{"id":"chatcmpl-9XALnHqkCBMBKrOx7Abg0hURHqYtY","choices":[{"finish_reason":"stop","index":0,"message":{"content":"Hello! How can I assist you today?","role":"assistant"}}],"created":1717691639,"model":"gpt-3.5-turbo-0125","object":"chat.completion","system_fingerprint":null,"usage":{"completion_tokens":9,"prompt_tokens":8,"total_tokens":17}}%
```
</TabItem>
</Tabs>

View file

@ -97,6 +97,40 @@ def common_checks(
raise Exception(
f"'user' param not passed in. 'enforce_user_param'={general_settings['enforce_user_param']}"
)
if general_settings.get("enforced_params", None) is not None:
# Enterprise ONLY Feature
# we already validate if user is premium_user when reading the config
# Add an extra premium_usercheck here too, just incase
from litellm.proxy.proxy_server import premium_user, CommonProxyErrors
if premium_user is not True:
raise ValueError(
"Trying to use `enforced_params`"
+ CommonProxyErrors.not_premium_user.value
)
if route in LiteLLMRoutes.openai_routes.value:
# loop through each enforced param
# example enforced_params ['user', 'metadata', 'metadata.generation_name']
for enforced_param in general_settings["enforced_params"]:
_enforced_params = enforced_param.split(".")
if len(_enforced_params) == 1:
if _enforced_params[0] not in request_body:
raise ValueError(
f"BadRequest please pass param={_enforced_params[0]} in request body. This is a required param"
)
elif len(_enforced_params) == 2:
# this is a scenario where user requires request['metadata']['generation_name'] to exist
if _enforced_params[0] not in request_body:
raise ValueError(
f"BadRequest please pass param={_enforced_params[0]} in request body. This is a required param"
)
if _enforced_params[1] not in request_body[_enforced_params[0]]:
raise ValueError(
f"BadRequest please pass param=[{_enforced_params[0]}][{_enforced_params[1]}] in request body. This is a required param"
)
pass
# 7. [OPTIONAL] If 'litellm.max_budget' is set (>0), is proxy under budget
if (
litellm.max_budget > 0

View file

@ -21,7 +21,10 @@ model_list:
general_settings:
master_key: sk-1234
enforced_params:
- user
- metadata
- metadata.generation_name
litellm_settings:
callbacks: ["otel"]
store_audit_logs: true

View file

@ -2878,6 +2878,16 @@ class ProxyConfig:
)
health_check_interval = general_settings.get("health_check_interval", 300)
## check if user has set a premium feature in general_settings
if (
general_settings.get("enforced_params") is not None
and premium_user is not True
):
raise ValueError(
"Trying to use `enforced_params`"
+ CommonProxyErrors.not_premium_user.value
)
router_params: dict = {
"cache_responses": litellm.cache
!= None, # cache if user passed in cache values

View file

@ -2248,3 +2248,55 @@ async def test_create_update_team(prisma_client):
assert _team_info["budget_reset_at"] is not None and isinstance(
_team_info["budget_reset_at"], datetime.datetime
)
@pytest.mark.asyncio()
async def test_enforced_params(prisma_client):
setattr(litellm.proxy.proxy_server, "prisma_client", prisma_client)
setattr(litellm.proxy.proxy_server, "master_key", "sk-1234")
from litellm.proxy.proxy_server import general_settings
general_settings["enforced_params"] = [
"user",
"metadata",
"metadata.generation_name",
]
await litellm.proxy.proxy_server.prisma_client.connect()
request = NewUserRequest()
key = await new_user(request)
print(key)
generated_key = key.key
bearer_token = "Bearer " + generated_key
request = Request(scope={"type": "http"})
request._url = URL(url="/chat/completions")
# Case 1: Missing user
async def return_body():
return b'{"model": "gemini-pro-vision"}'
request.body = return_body
try:
await user_api_key_auth(request=request, api_key=bearer_token)
pytest.fail(f"This should have failed!. IT's an invalid request")
except Exception as e:
assert (
"BadRequest please pass param=user in request body. This is a required param"
in e.message
)
# Case 2: Missing metadata["generation_name"]
async def return_body_2():
return b'{"model": "gemini-pro-vision", "user": "1234", "metadata": {}}'
request.body = return_body_2
try:
await user_api_key_auth(request=request, api_key=bearer_token)
pytest.fail(f"This should have failed!. IT's an invalid request")
except Exception as e:
assert (
"Authentication Error, BadRequest please pass param=[metadata][generation_name] in request body"
in e.message
)