[BugFix] test_mla_fp8.py fails on Cublas 12.9

python/sglang/srt/models/deepseek_v2.py

@@ -94,6 +94,7 @@
     block_quant_dequant,
     block_quant_to_tensor_quant,
     channel_quant_to_tensor_quant,
+    input_to_float8,
     normalize_e4m3fn_to_e4m3fnuz,
     requant_weight_ue8m0_inplace,
 )
@@ -131,6 +132,7 @@
     is_hip,
     is_non_idle_and_non_empty,
     is_npu,
+    is_nvidia_cublas_cu12_version_ge_12_9,
     is_sm100_supported,
     log_info_on_rank0,
     make_layers,
@@ -189,6 +191,7 @@
 
 _is_flashinfer_available = is_flashinfer_available()
 _is_sm100_supported = is_cuda() and is_sm100_supported()
+_is_cublas_ge_129 = is_nvidia_cublas_cu12_version_ge_12_9()
 
 
 logger = logging.getLogger(__name__)
@@ -1572,10 +1575,15 @@ def forward_absorb_prepare(
                     self.w_kc.to(torch.bfloat16) * self.w_scale,
                 )
         elif self.w_kc.dtype == torch.float8_e4m3fn:
-            q_nope_val, q_nope_scale = per_tensor_quant_mla_fp8(
-                q_nope.transpose(0, 1),
-                zero_allocator.allocate(1),
-            )
+            # TODO fix the per_tensor_quant_mla_fp8 for cublas 12.9
+            if _is_cublas_ge_129:
+                q_nope_val, q_nope_scale = input_to_float8(
+                    q_nope.transpose(0, 1), torch.float8_e4m3fn
+                )
+            else:
+                q_nope_val, q_nope_scale = per_tensor_quant_mla_fp8(
+                    q_nope.transpose(0, 1), zero_allocator.allocate(1)
+                )
             q_nope_out = bmm_fp8(
                 q_nope_val, self.w_kc, q_nope_scale, self.w_scale, torch.bfloat16
             )
@@ -1716,10 +1724,14 @@ def forward_absorb_core(
                 attn_bmm_output = attn_bmm_output.transpose(0, 1).flatten(1, 2)
 
         elif self.w_vc.dtype == torch.float8_e4m3fn:
-            attn_output_val, attn_output_scale = per_tensor_quant_mla_fp8(
-                attn_output.transpose(0, 1),
-                zero_allocator.allocate(1),
-            )
+            if _is_cublas_ge_129:
+                attn_output_val, attn_output_scale = input_to_float8(
+                    attn_output.transpose(0, 1), torch.float8_e4m3fn
+                )
+            else:
+                attn_output_val, attn_output_scale = per_tensor_quant_mla_fp8(
+                    attn_output.transpose(0, 1), zero_allocator.allocate(1)
+                )
             attn_bmm_output = bmm_fp8(
                 attn_output_val,
                 self.w_vc,

python/sglang/srt/utils/common.py

@@ -263,6 +263,17 @@ def is_flashinfer_available():
     return importlib.util.find_spec("flashinfer") is not None and is_cuda()
 
 
+def is_nvidia_cublas_cu12_version_ge_12_9():
+    """
+    temporary fix for issue #11272
+    """
+    try:
+        installed_version = version("nvidia-cublas-cu12")
+    except PackageNotFoundError:
+        return False
+    return pkg_version.parse(installed_version) >= pkg_version.parse("12.9")
+
+
 def random_uuid() -> str:
     return str(uuid.uuid4().hex)