[Model][Misc] Add e5-mistral-7b-instruct and Embedding API (#3734)

This commit is contained in:
Chang Su
2024-05-11 11:30:37 -07:00
committed by GitHub
parent 4e12131089
commit e254497b66
38 changed files with 1627 additions and 160 deletions
+5 -5
View File
@@ -1,9 +1,9 @@
from typing import Any, Dict, List, Optional, Set, Tuple
from typing import Any, Dict, List, Optional, Set, Tuple, Union
from vllm.executor.executor_base import ExecutorAsyncBase, ExecutorBase
from vllm.logger import init_logger
from vllm.lora.request import LoRARequest
from vllm.sequence import ExecuteModelRequest, SamplerOutput
from vllm.sequence import ExecuteModelRequest, PoolerOutput, SamplerOutput
from vllm.utils import (get_distributed_init_method, get_ip, get_open_port,
make_async)
from vllm.worker.worker_base import WorkerWrapperBase
@@ -123,8 +123,8 @@ class GPUExecutor(ExecutorBase):
self.driver_worker.initialize_cache(num_gpu_blocks, num_cpu_blocks)
def execute_model(
self,
execute_model_req: ExecuteModelRequest) -> List[SamplerOutput]:
self, execute_model_req: ExecuteModelRequest
) -> List[Union[SamplerOutput, PoolerOutput]]:
output = self.driver_worker.execute_model(execute_model_req)
return output
@@ -150,7 +150,7 @@ class GPUExecutorAsync(GPUExecutor, ExecutorAsyncBase):
async def execute_model_async(
self,
execute_model_req: ExecuteModelRequest,
) -> List[SamplerOutput]:
) -> List[Union[SamplerOutput, PoolerOutput]]:
output = await make_async(self.driver_worker.execute_model
)(execute_model_req=execute_model_req, )
return output