[Bugfix][Quantization] Fix INT8 W8A8 MoE crash in TritonExperts (#51411)

Signed-off-by: Djordje Ramic <djoramic@amd.com>
2026-08-07 22:06:58 -05:00
parent f7ef489e93
commit 5eaa70dac5
1 changed files with 2 additions and 0 deletions
@@ -252,6 +252,7 @@ class TritonExperts(LoRAExpertsMixin, mk.FusedMoEExpertsModular):
            torch.bfloat16,
            torch.float8_e4m3fn,
            torch.float8_e4m3fnuz,
+            torch.int8,
        ]

        # We declared expects_unquantized_inputs (LoRA + DP/EP all2all), so the
@@ -298,6 +299,7 @@ class TritonExperts(LoRAExpertsMixin, mk.FusedMoEExpertsModular):
        elif (
            hidden_states.dtype == torch.float8_e4m3fn
            or hidden_states.dtype == torch.float8_e4m3fnuz
+            or hidden_states.dtype == torch.int8
        ):
            compute_type = tl.bfloat16
        else: