fix mlp batch size check (#9718)
This commit is contained in:
		
							parent
							
								
									1fa7793fc0
								
							
						
					
					
						commit
						f2e6abb563
					
				
					 3 changed files with 8 additions and 6 deletions
				
			
		| 
						 | 
				
			
			@ -69,11 +69,11 @@ def baichuan_mlp_forward(
 | 
			
		|||
    self,
 | 
			
		||||
    x: torch.Tensor,
 | 
			
		||||
) -> torch.Tensor:
 | 
			
		||||
    if x.shape[1] == 1 and x.dtype == torch.float32 and x.device.type == 'xpu' \
 | 
			
		||||
    x_2d = x.view(-1, x.shape[-1])
 | 
			
		||||
    if x_2d.shape[0] == 1 and x.dtype == torch.float32 and x.device.type == 'xpu' \
 | 
			
		||||
            and self.gate_proj.qtype == ggml_tensor_qtype["sym_int4"] \
 | 
			
		||||
            and not (self.training and x.requires_grad):
 | 
			
		||||
        import linear_q4_0
 | 
			
		||||
        x_2d = x.view(-1, x.shape[-1])
 | 
			
		||||
        if not x_2d.is_contiguous():
 | 
			
		||||
            x_2d = x_2d.contiguous()
 | 
			
		||||
        return self.down_proj(linear_q4_0.mlp_forward_q4_0_xpu(
 | 
			
		||||
| 
						 | 
				
			
			
 | 
			
		|||
| 
						 | 
				
			
			@ -42,6 +42,7 @@ from bigdl.llm.transformers.models.utils import init_kv_cache, extend_kv_cache,
 | 
			
		|||
from bigdl.llm.transformers.models.utils import rotate_half, apply_rotary_pos_emb
 | 
			
		||||
from bigdl.llm.transformers.models.utils import apply_rotary_pos_emb_no_cache_xpu
 | 
			
		||||
from bigdl.llm.transformers.low_bit_linear import SYM_INT4
 | 
			
		||||
from bigdl.llm.ggml.quantize import ggml_tensor_qtype
 | 
			
		||||
 | 
			
		||||
 | 
			
		||||
def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:
 | 
			
		||||
| 
						 | 
				
			
			@ -96,10 +97,11 @@ def llama_mlp_forward(
 | 
			
		|||
    self,
 | 
			
		||||
    x: torch.Tensor,
 | 
			
		||||
) -> torch.Tensor:
 | 
			
		||||
    if x.shape[1] == 1 and x.dtype == torch.float32 and x.device.type == 'xpu' \
 | 
			
		||||
    x_2d = x.view(-1, x.shape[-1])
 | 
			
		||||
    if x_2d.shape[0] == 1 and x.dtype == torch.float32 and x.device.type == 'xpu' \
 | 
			
		||||
            and self.gate_proj.qtype == ggml_tensor_qtype["sym_int4"] \
 | 
			
		||||
            and not (self.training and x.requires_grad):
 | 
			
		||||
        import linear_q4_0
 | 
			
		||||
        x_2d = x.view(-1, x.shape[-1])
 | 
			
		||||
        if not x_2d.is_contiguous():
 | 
			
		||||
            x_2d = x_2d.contiguous()
 | 
			
		||||
        return self.down_proj(linear_q4_0.mlp_forward_q4_0_xpu(
 | 
			
		||||
| 
						 | 
				
			
			
 | 
			
		|||
| 
						 | 
				
			
			@ -240,11 +240,11 @@ def qwen_attention_forward(
 | 
			
		|||
 | 
			
		||||
 | 
			
		||||
def qwen_mlp_forward(self, x: torch.Tensor) -> torch.Tensor:
 | 
			
		||||
    if x.shape[1] == 1 and x.dtype == torch.float32 and x.device.type == 'xpu' \
 | 
			
		||||
    x_2d = x.view(-1, x.shape[-1])
 | 
			
		||||
    if x_2d.shape[0] == 1 and x.dtype == torch.float32 and x.device.type == 'xpu' \
 | 
			
		||||
            and self.w2.qtype == ggml_tensor_qtype["sym_int4"] \
 | 
			
		||||
            and not (self.training and x.requires_grad):
 | 
			
		||||
        import linear_q4_0
 | 
			
		||||
        x_2d = x.view(-1, x.shape[-1])
 | 
			
		||||
        if not x_2d.is_contiguous():
 | 
			
		||||
            x_2d = x_2d.contiguous()
 | 
			
		||||
        return self.c_proj(linear_q4_0.mlp_forward_q4_0_xpu(
 | 
			
		||||
| 
						 | 
				
			
			
 | 
			
		|||
		Loading…
	
		Reference in a new issue