Merge pull request #10266 from nero-dv/dev

Update sub_quadratic_attention.py

modules/sub_quadratic_attention.py

@@ -202,13 +202,22 @@ def efficient_dot_product_attention(
             value=value,
         )
     
-    # TODO: maybe we should use torch.empty_like(query) to allocate storage in-advance,
-    # and pass slices to be mutated, instead of torch.cat()ing the returned slices
-    res = torch.cat([
-        compute_query_chunk_attn(
+    # slices of res tensor are mutable, modifications made
+    # to the slices will affect the original tensor.
+    # if output of compute_query_chunk_attn function has same number of
+    # dimensions as input query tensor, we initialize tensor like this:
+    num_query_chunks = int(np.ceil(q_tokens / query_chunk_size))
+    query_shape = get_query_chunk(0).shape
+    res_shape = (query_shape[0], query_shape[1] * num_query_chunks, *query_shape[2:])
+    res_dtype = get_query_chunk(0).dtype
+    res = torch.zeros(res_shape, dtype=res_dtype)
+
+    for i in range(num_query_chunks):
+        attn_scores = compute_query_chunk_attn(
             query=get_query_chunk(i * query_chunk_size),
             key=key,
             value=value,
-        ) for i in range(math.ceil(q_tokens / query_chunk_size))
-    ], dim=1)
+        )
+        res[:, i * query_chunk_size:(i + 1) * query_chunk_size, :] = attn_scores
+
     return res