From 0da2f0c86a8558ae1bd3bc03f7c5439b0a53e95f Mon Sep 17 00:00:00 2001 From: Jianyu Huang Date: Tue, 30 Apr 2024 15:33:39 -0700 Subject: [PATCH] Refactor fbgemm / llama csrc code base Summary: Dedup and reuse the ops moved in fbgemm gpu. Reviewed By: sryap, xintwfb, jiawenliu64 Differential Revision: D56685973 fbshipit-source-id: e3e6e27e9f3f380460f4fbf01eb3075109e37d2f --- fbgemm_gpu/experimental/gen_ai/test/quantize/quantize_test.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fbgemm_gpu/experimental/gen_ai/test/quantize/quantize_test.py b/fbgemm_gpu/experimental/gen_ai/test/quantize/quantize_test.py index f126a34a6..82cb3001e 100644 --- a/fbgemm_gpu/experimental/gen_ai/test/quantize/quantize_test.py +++ b/fbgemm_gpu/experimental/gen_ai/test/quantize/quantize_test.py @@ -189,7 +189,7 @@ def test_tensor_with_nan(self, G_B: int, D: int, HD_L: int) -> None: w = wq.bfloat16() * w_scale zq_ref = (x @ w.T).to(torch.bfloat16) - torch.testing.assert_close(zq[:B, :], zq_ref, atol=1.0e-3, rtol=1.0e-3) + torch.testing.assert_close(zq[:B, :], zq_ref, atol=2.0e-3, rtol=2.0e-3) @settings(deadline=None) @given(