[LLM] support transformer int4 + amx int4 (#8838)

python/llm/src/bigdl/llm/transformers/linear_quant.py

@@ -51,10 +51,11 @@ from torch import Tensor, device, dtype, nn
 T = TypeVar("T", bound="torch.nn.Module")
 
 import bigdl.llm.ggml.model.llama.llama_cpp as ggml
-from bigdl.llm.utils.isa_checker import is_server
+from bigdl.llm.utils.isa_checker import is_server, is_spr
 import ctypes
 from bigdl.llm.ggml.quantize import ggml_tensor_qtype
 IS_SERVER = is_server()
+IS_SPR = is_spr()
 TORCH_LINEAR_THRESHOLD = 96
 SYM_INT4 = ggml_tensor_qtype["sym_int4"]
 
@@ -256,7 +257,8 @@ class LinearQuant(nn.Linear):
         else:
             # CPU logic
             # todo may need to set a different number on different platforms
-            if IS_SERVER and self.qtype == SYM_INT4 and x_2d.shape[0] >= TORCH_LINEAR_THRESHOLD:
+            if IS_SERVER and (not IS_SPR) and \
+                    self.qtype == SYM_INT4 and x_2d.shape[0] >= TORCH_LINEAR_THRESHOLD:
                 x0_fp32 = ggml_int4_convert_fp32(x0, self.weight_shape, self.weight_length)
                 result = F.linear(x, x0_fp32, self.bias)
             else:

python/llm/src/bigdl/llm/utils/isa_checker.py

@@ -79,3 +79,9 @@ def check_avx512_vnni():
 
 def is_server():
     return check_avx512_vnni()
+
+
+# todo: use cpuid to check SPR
+# note: now only SPR supports both avxvnni and avx512vnni
+def is_spr():
+    return check_avx_vnni() and check_avx512_vnni()