mirror of
https://github.com/comfyanonymous/ComfyUI.git
synced 2026-03-13 00:59:59 +00:00
Lower kv cache memory usage. (#12909)
This commit is contained in:
@@ -248,7 +248,7 @@ class KV_Attn_Input:
|
||||
self.set_cache = False
|
||||
return {"q": q, "k": torch.cat((k, kk), dim=2), "v": torch.cat((v, vv), dim=2)}
|
||||
|
||||
self.cache[cache_key] = (k[:, :, -ref_toks:], v[:, :, -ref_toks:])
|
||||
self.cache[cache_key] = (k[:, :, -ref_toks:].clone(), v[:, :, -ref_toks:].clone())
|
||||
self.set_cache = True
|
||||
return {"q": q, "k": k, "v": v}
|
||||
|
||||
|
||||
Reference in New Issue
Block a user