[Perf] Avoid more GPU<->CPU syncs on the model execution path (#51738)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
This commit is contained in:
@@ -248,13 +248,13 @@ def _audioflamingo3_field_config(hf_inputs: Mapping[str, torch.Tensor]):
|
||||
feature_attention_mask=MultiModalFieldConfig.flat_from_sizes(
|
||||
"audio", chunk_counts, dim=0
|
||||
),
|
||||
chunk_counts=MultiModalFieldConfig.batched("audio"),
|
||||
chunk_counts=MultiModalFieldConfig.batched("audio", keep_on_cpu=True),
|
||||
)
|
||||
return dict(
|
||||
audio_embeds=MultiModalFieldConfig.batched("audio"),
|
||||
input_features=MultiModalFieldConfig.batched("audio"),
|
||||
feature_attention_mask=MultiModalFieldConfig.batched("audio"),
|
||||
chunk_counts=MultiModalFieldConfig.batched("audio"),
|
||||
chunk_counts=MultiModalFieldConfig.batched("audio", keep_on_cpu=True),
|
||||
)
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user