[Doc] Update docs to refer to pooling models (#11093)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2024-12-11 13:36:27 +00:00
committed by GitHub
parent 8f10d5e393
commit cad5c0a6ed
14 changed files with 26 additions and 21 deletions
+1 -1
View File
@@ -178,7 +178,7 @@ class CPUWorker(LoraNotSupportedWorkerBase, LocalOrDistributedWorkerBase):
# Uninitialized cache engine. Will be initialized by
# initialize_cache.
self.cache_engine: List[CPUCacheEngine]
# Initialize cpu_cache as embedding models don't initialize kv_caches
# Initialize cpu_cache as pooling models don't initialize kv_caches
self.cpu_cache: Optional[List[List[torch.Tensor]]] = None
# Torch profiler. Enabled and configured through env vars:
+2 -2
View File
@@ -65,8 +65,8 @@ class HPUWorker(LocalOrDistributedWorkerBase):
# Uninitialized cache engine. Will be initialized by
# initialize_cache.
self.cache_engine: List[HPUCacheEngine]
# Initialize gpu_cache as embedding models don't initialize kv_caches
self.hpu_cache: Optional[List[List[torch.tensor]]] = None
# Initialize gpu_cache as pooling models don't initialize kv_caches
self.hpu_cache: Optional[List[List[torch.Tensor]]] = None
# Torch profiler. Enabled and configured through env vars:
# VLLM_TORCH_PROFILER_DIR=/path/to/save/trace
if envs.VLLM_TORCH_PROFILER_DIR:
+1 -1
View File
@@ -91,7 +91,7 @@ class Worker(LocalOrDistributedWorkerBase):
# Uninitialized cache engine. Will be initialized by
# initialize_cache.
self.cache_engine: List[CacheEngine]
# Initialize gpu_cache as embedding models don't initialize kv_caches
# Initialize gpu_cache as pooling models don't initialize kv_caches
self.gpu_cache: Optional[List[List[torch.Tensor]]] = None
self._seq_group_metadata_cache: Dict[str, SequenceGroupMetadata] = {}