The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Fix attention masks properly for multiple batches.
comfyanonymous6bcf57ff10f4
diff --git a/comfy/ldm/modules/attention.py b/comfy/ldm/modules/attention.pyindex f1dca2c28..48399bc07 100644--- a/comfy/ldm/modules/attention.py+++ b/comfy/ldm/modules/attention.py@@ -118,7 +118,7 @@ def attention_basic(q, k, v, heads, mask=None):bs = 1else:bs = mask.shape[0]- mask = mask.reshape(bs, -1, mask.shape[-2], mask.shape[-1]).expand(-1, heads, -1, -1).reshape(-1, mask.shape[-2], mask.shape[-1])+ mask = mask.reshape(bs, -1, mask.shape[-2], mask.shape[-1]).expand(b, heads, -1, -1).reshape(-1, mask.shape[-2], mask.shape[-1])sim.add_(mask)# attention, what we cannot get enough of@@ -175,7 +175,7 @@ def attention_sub_quad(query, key, value, heads, mask=None):bs = 1else:bs = mask.shape[0]- mask = mask.reshape(bs, -1, mask.shape[-2], mask.shape[-1]).expand(-1, heads, -1, -1).reshape(-1, mask.shape[-2], mask.shape[-1])+ mask = mask.reshape(bs, -1, mask.shape[-2], mask.shape[-1]).expand(b, heads, -1, -1).reshape(-1, mask.shape[-2], mask.shape[-1])hidden_states = efficient_dot_product_attention(query,@@ -240,7 +240,7 @@ def attention_split(q, k, v, heads, mask=None):bs = 1else:bs = mask.shape[0]- mask = mask.reshape(bs, -1, mask.shape[-2], mask.shape[-1]).expand(-1, heads, -1, -1).reshape(-1, mask.shape[-2], mask.shape[-1])+ mask = mask.reshape(bs, -1, mask.shape[-2], mask.shape[-1]).expand(b, heads, -1, -1).reshape(-1, mask.shape[-2], mask.shape[-1])# print("steps", steps, mem_required, mem_free_total, modifier, q.element_size(), tensor_size)first_op_done = False