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

add cerebras api

Browse source
This commit is contained in:
Ishaan Jaff 2024-08-31 08:35:23 -07:00
parent ccbbb4f3f8
commit 6aaa7a75cd
3 changed files with 35 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

3
litellm/__init__.py
View file

@ -452,6 +452,7 @@ openai_compatible_providers: List = [
"mistral", "mistral",
"groq", "groq",
"nvidia_nim", "nvidia_nim",
"cerebras",
"volcengine", "volcengine",
"codestral", "codestral",
"deepseek", "deepseek",
@ -690,6 +691,7 @@ provider_list: List = [
"mistral", "mistral",
"groq", "groq",
"nvidia_nim", "nvidia_nim",
"cerebras",
"volcengine", "volcengine",
"codestral", "codestral",
"text-completion-codestral", "text-completion-codestral",
@ -905,6 +907,7 @@ from .llms.openai import (
AzureAIStudioConfig, AzureAIStudioConfig,
) )
from .llms.nvidia_nim import NvidiaNimConfig from .llms.nvidia_nim import NvidiaNimConfig
from .llms.cerebras.chat import CerebrasConfig
from .llms.fireworks_ai import FireworksAIConfig from .llms.fireworks_ai import FireworksAIConfig
from .llms.volcengine import VolcEngineConfig from .llms.volcengine import VolcEngineConfig
from .llms.text_completion_codestral import MistralTextCompletionConfig from .llms.text_completion_codestral import MistralTextCompletionConfig

4
litellm/main.py
View file

@ -390,6 +390,7 @@ async def acompletion(
or custom_llm_provider == "perplexity" or custom_llm_provider == "perplexity"
or custom_llm_provider == "groq" or custom_llm_provider == "groq"
or custom_llm_provider == "nvidia_nim" or custom_llm_provider == "nvidia_nim"
or custom_llm_provider == "cerebras"
or custom_llm_provider == "volcengine" or custom_llm_provider == "volcengine"
or custom_llm_provider == "codestral" or custom_llm_provider == "codestral"
or custom_llm_provider == "text-completion-codestral" or custom_llm_provider == "text-completion-codestral"
@ -1295,6 +1296,7 @@ def completion(
or custom_llm_provider == "perplexity" or custom_llm_provider == "perplexity"
or custom_llm_provider == "groq" or custom_llm_provider == "groq"
or custom_llm_provider == "nvidia_nim" or custom_llm_provider == "nvidia_nim"
or custom_llm_provider == "cerebras"
or custom_llm_provider == "volcengine" or custom_llm_provider == "volcengine"
or custom_llm_provider == "codestral" or custom_llm_provider == "codestral"
or custom_llm_provider == "deepseek" or custom_llm_provider == "deepseek"
@ -3144,6 +3146,7 @@ async def aembedding(*args, **kwargs) -> EmbeddingResponse:
or custom_llm_provider == "perplexity" or custom_llm_provider == "perplexity"
or custom_llm_provider == "groq" or custom_llm_provider == "groq"
or custom_llm_provider == "nvidia_nim" or custom_llm_provider == "nvidia_nim"
or custom_llm_provider == "cerebras"
or custom_llm_provider == "volcengine" or custom_llm_provider == "volcengine"
or custom_llm_provider == "deepseek" or custom_llm_provider == "deepseek"
or custom_llm_provider == "fireworks_ai" or custom_llm_provider == "fireworks_ai"
@ -3795,6 +3798,7 @@ async def atext_completion(
or custom_llm_provider == "perplexity" or custom_llm_provider == "perplexity"
or custom_llm_provider == "groq" or custom_llm_provider == "groq"
or custom_llm_provider == "nvidia_nim" or custom_llm_provider == "nvidia_nim"
or custom_llm_provider == "cerebras"
or custom_llm_provider == "volcengine" or custom_llm_provider == "volcengine"
or custom_llm_provider == "text-completion-codestral" or custom_llm_provider == "text-completion-codestral"
or custom_llm_provider == "deepseek" or custom_llm_provider == "deepseek"

28
litellm/utils.py
View file

@ -2854,6 +2854,7 @@ def get_optional_params(
and custom_llm_provider != "together_ai" and custom_llm_provider != "together_ai"
and custom_llm_provider != "groq" and custom_llm_provider != "groq"
and custom_llm_provider != "nvidia_nim" and custom_llm_provider != "nvidia_nim"
and custom_llm_provider != "cerebras"
and custom_llm_provider != "volcengine" and custom_llm_provider != "volcengine"
and custom_llm_provider != "deepseek" and custom_llm_provider != "deepseek"
and custom_llm_provider != "codestral" and custom_llm_provider != "codestral"
@ -3613,6 +3614,16 @@ def get_optional_params(
non_default_params=non_default_params, non_default_params=non_default_params,
optional_params=optional_params, optional_params=optional_params,
) )
elif custom_llm_provider == "cerebras":
supported_params = get_supported_openai_params(
model=model, custom_llm_provider=custom_llm_provider
)
_check_valid_arg(supported_params=supported_params)
optional_params = litellm.CerebrasConfig().map_openai_params(
non_default_params=non_default_params,
optional_params=optional_params,
model=model,
)
elif custom_llm_provider == "fireworks_ai": elif custom_llm_provider == "fireworks_ai":
supported_params = get_supported_openai_params( supported_params = get_supported_openai_params(
model=model, custom_llm_provider=custom_llm_provider model=model, custom_llm_provider=custom_llm_provider
@ -4238,6 +4249,8 @@ def get_supported_openai_params(
return litellm.FireworksAIConfig().get_supported_openai_params() return litellm.FireworksAIConfig().get_supported_openai_params()
elif custom_llm_provider == "nvidia_nim": elif custom_llm_provider == "nvidia_nim":
return litellm.NvidiaNimConfig().get_supported_openai_params(model=model) return litellm.NvidiaNimConfig().get_supported_openai_params(model=model)
elif custom_llm_provider == "cerebras":
return litellm.CerebrasConfig().get_supported_openai_params(model=model)
elif custom_llm_provider == "volcengine": elif custom_llm_provider == "volcengine":
return litellm.VolcEngineConfig().get_supported_openai_params(model=model) return litellm.VolcEngineConfig().get_supported_openai_params(model=model)
elif custom_llm_provider == "groq": elif custom_llm_provider == "groq":
@ -4665,6 +4678,13 @@ def get_llm_provider(
or "https://integrate.api.nvidia.com/v1" or "https://integrate.api.nvidia.com/v1"
) # type: ignore ) # type: ignore
dynamic_api_key = api_key or get_secret("NVIDIA_NIM_API_KEY") dynamic_api_key = api_key or get_secret("NVIDIA_NIM_API_KEY")
elif custom_llm_provider == "cerebras":
api_base = (
api_base
or get_secret("CEREBRAS_API_BASE")
or "https://api.cerebras.ai/v1"
) # type: ignore
dynamic_api_key = api_key or get_secret("CEREBRAS_API_KEY")
elif custom_llm_provider == "volcengine": elif custom_llm_provider == "volcengine":
# volcengine is openai compatible, we just need to set this to custom_openai and have the api_base be https://api.endpoints.anyscale.com/v1 # volcengine is openai compatible, we just need to set this to custom_openai and have the api_base be https://api.endpoints.anyscale.com/v1
api_base = ( api_base = (
@ -4815,6 +4835,9 @@ def get_llm_provider(
elif endpoint == "https://integrate.api.nvidia.com/v1": elif endpoint == "https://integrate.api.nvidia.com/v1":
custom_llm_provider = "nvidia_nim" custom_llm_provider = "nvidia_nim"
dynamic_api_key = get_secret("NVIDIA_NIM_API_KEY") dynamic_api_key = get_secret("NVIDIA_NIM_API_KEY")
elif endpoint == "https://api.cerebras.ai/v1":
custom_llm_provider = "cerebras"
dynamic_api_key = get_secret("CEREBRAS_API_KEY")
elif endpoint == "https://codestral.mistral.ai/v1": elif endpoint == "https://codestral.mistral.ai/v1":
custom_llm_provider = "codestral" custom_llm_provider = "codestral"
dynamic_api_key = get_secret("CODESTRAL_API_KEY") dynamic_api_key = get_secret("CODESTRAL_API_KEY")
@ -5734,6 +5757,11 @@ def validate_environment(
keys_in_environment = True keys_in_environment = True
else: else:
missing_keys.append("NVIDIA_NIM_API_KEY") missing_keys.append("NVIDIA_NIM_API_KEY")
elif custom_llm_provider == "cerebras":
if "CEREBRAS_API_KEY" in os.environ:
keys_in_environment = True
else:
missing_keys.append("CEREBRAS_API_KEY")
elif custom_llm_provider == "volcengine": elif custom_llm_provider == "volcengine":
if "VOLCENGINE_API_KEY" in os.environ: if "VOLCENGINE_API_KEY" in os.environ:
keys_in_environment = True keys_in_environment = True