unified queueing

This commit is contained in:
Yannic Kilcher
2023-02-11 01:31:25 +01:00
parent 76f7af0dfd
commit 48212079f4
7 changed files with 144 additions and 82 deletions
@@ -9,8 +9,9 @@ class MessageRequest(pydantic.BaseModel):
model_name: str = "distilgpt2"
max_new_tokens: int = 100
def compatible_with(self, worker_config: inference.WorkerConfig) -> bool:
return self.model_name == worker_config.model_name
@property
def worker_compat_hash(self) -> str:
return f"{self.model_name}"
class TokenResponseEvent(pydantic.BaseModel):