Fix OOM in attention kernel test (#1223)

tests/kernels/test_attention.py

@@ -247,8 +247,11 @@ def test_multi_query_kv_attention(
     torch.random.manual_seed(seed)
     torch.cuda.manual_seed(seed)
 
-    seq_lens = random.sample(range(1, MAX_SEQ_LEN), num_seqs)
-    seq_lens[-1] = MAX_SEQ_LEN
+    # MAX_SEQ_LEN sometimes causes OOM in the reference implementation.
+    # As the xformers library is already tested with its own tests, we can use
+    # a smaller MAX_SEQ_LEN here.
+    max_len = min(MAX_SEQ_LEN, 4096)
+    seq_lens = random.sample(range(1, max_len), num_seqs)
     num_tokens = sum(seq_lens)
 
     scale = float(1.0 / (head_size**0.5))