search

siliconflow / onediff

OneDiff: An out-of-the-box acceleration library for diffusion models.
https://github.com/siliconflow/onediff/wiki
Apache License 2.0
1.4k stars 85 forks source link

Enterprise quant has problem #969

Open CuddleSabe opened 1 week ago

CuddleSabe commented 1 week ago

InferDataType Failed. Expected kFloat16, but got kInt8

ERROR [2024-06-20 17:46:58] /root/picasso/韩昊天/onediff/src/onediff/infer_compiler/oneflow/utils.py:23 - Exception in forward: e=RuntimeError('InferDataType Failed. Expected kFloat16, but got kInt8\n File "oneflow/core/framework/op_interpreter/op_interpreter_util.cpp", line 144, in Dispatch\n Dispatch(op_expr, inputs, ctx)\n File "oneflow/core/framework/op_interpreter/op_interpreter_util.cpp", line 135, in Dispatch\n Dispatch(op_expr, inputs, outputs.get(), ctx)\n File "oneflow/core/framework/op_interpreter/opinterpreter.cpp", line 103, in Apply\n internal->Apply(op_expr, inputs, outputs, ctx)\n File "oneflow/core/framework/op_interpreter/lazy_op_interpreter.cpp", line 856, in ApplyImpl\n infer_ctx->AddAndInferGlobalOp(op_conf)\n File "oneflow/core/job/job_build_and_infer_ctx.cpp", line 573, in AddAndInferOp\n op->InferLogicalOutBlobDescsIf()\n File "oneflow/core/operator/operator.cpp", line 329, in InferLogicalOutBlobDescsIf\n InferLogicalOutBlobDescs(BlobDesc4BnInOp, JUST(GetOpParallelDesc()))\n File "oneflow/core/operator/userop.cpp", line 772, in InferLogicalOutBlobDescs\n val->data_type_infer_fn(&infer_ctx)\n File "oneflow/user/ops/fused_matmul_bias_op.cpp", line 66, in InferDataType4MatmulBias\n CHECK_EQ_OR_RETURN(in_desc.data_type(), first_in_desc.data_type())\nError Type: oneflow.ErrorProto.check_failed_error')