PaddleX
PaddleX copied to clipboard
PaddlePaddle
使用ppyolov2模型进行预测时报错
使用GUI模式，训练产生了ppyolov2的模型（832*832）并导出，然后使用model_infer进行调用时报错：
REGISTER_CLASS:paddlex init PaddleXModel,model_type=paddlex WARNING: Logging before InitGoogleLogging() is written to STDERR I0909 11:13:14.494064 29276 analysis_config.cc:424] use_dlnne_:0 I0909 11:13:14.494064 29276 analysis_config.cc:424] use_dlnne_:0 I0909 11:13:14.494064 29276 analysis_config.cc:424] use_dlnne_:0 I0909 11:13:16.817966 29276 analysis_config.cc:424] use_dlnne_:0 I0909 11:13:16.817966 29276 analysis_predictor.cc:155] Profiler is deactivated, and no profiling report will be generated. e[1me[35m--- Running analysis [ir_graph_build_pass]e[0m e[1me[35m--- Running analysis [ir_graph_clean_pass]e[0m e[1me[35m--- Running analysis [ir_analysis_pass]e[0m e[32m--- Running IR pass [is_test_pass]e[0m e[32m--- Running IR pass [simplify_with_basic_ops_pass]e[0m e[32m--- Running IR pass [conv_affine_channel_fuse_pass]e[0m e[32m--- Running IR pass [conv_eltwiseadd_affine_channel_fuse_pass]e[0m e[32m--- Running IR pass [conv_bn_fuse_pass]e[0m I0909 11:13:17.718261 29276 graph_pattern_detector.cc:91] --- detected 101 subgraphs e[32m--- Running IR pass [conv_eltwiseadd_bn_fuse_pass]e[0m e[32m--- Running IR pass [embedding_eltwise_layernorm_fuse_pass]e[0m e[32m--- Running IR pass [multihead_matmul_fuse_pass_v2]e[0m e[32m--- Running IR pass [squeeze2_matmul_fuse_pass]e[0m e[32m--- Running IR pass [reshape2_matmul_fuse_pass]e[0m e[32m--- Running IR pass [flatten2_matmul_fuse_pass]e[0m e[32m--- Running IR pass [map_matmul_to_mul_pass]e[0m e[32m--- Running IR pass [fc_fuse_pass]e[0m e[32m--- Running IR pass [fc_elementwise_layernorm_fuse_pass]e[0m e[32m--- Running IR pass [conv_elementwise_add_act_fuse_pass]e[0m I0909 11:13:18.036963 29276 graph_pattern_detector.cc:91] --- detected 32 subgraphs e[32m--- Running IR pass [conv_elementwise_add2_act_fuse_pass]e[0m I0909 11:13:18.096801 29276 graph_pattern_detector.cc:91] --- detected 4 subgraphs e[32m--- Running IR pass [conv_elementwise_add_fuse_pass]e[0m I0909 11:13:18.146697 29276 graph_pattern_detector.cc:91] --- detected 71 subgraphs e[32m--- Running IR pass [transpose_flatten_concat_fuse_pass]e[0m e[32m--- Running IR pass [runtime_context_cache_pass]e[0m e[1me[35m--- Running analysis [ir_params_sync_among_devices_pass]e[0m I0909 11:13:18.216498 29276 ir_params_sync_among_devices_pass.cc:45] Sync params from CPU to GPU e[1me[35m--- Running analysis [adjust_cudnn_workspace_size_pass]e[0m e[1me[35m--- Running analysis [inference_op_replace_pass]e[0m e[1me[35m--- Running analysis [memory_optimize_pass]e[0m I0909 11:13:18.709683 29276 memory_optimize_pass.cc:199] Cluster name : elementwise_add_2 size: 23658496 I0909 11:13:18.709683 29276 memory_optimize_pass.cc:199] Cluster name : tmp_18 size: 2957312 I0909 11:13:18.711669 29276 memory_optimize_pass.cc:199] Cluster name : image size: 4435968 I0909 11:13:18.715656 29276 memory_optimize_pass.cc:199] Cluster name : batch_norm_6.tmp_3 size: 23658496 I0909 11:13:18.715656 29276 memory_optimize_pass.cc:199] Cluster name : tmp_40 size: 739328 I0909 11:13:18.716653 29276 memory_optimize_pass.cc:199] Cluster name : elementwise_add_1 size: 23658496 I0909 11:13:18.717651 29276 memory_optimize_pass.cc:199] Cluster name : relu_17.tmp_0 size: 11829248 I0909 11:13:18.718649 29276 memory_optimize_pass.cc:199] Cluster name : tmp_8 size: 1478656 I0909 11:13:18.718649 29276 memory_optimize_pass.cc:199] Cluster name : tmp_20 size: 2957312 I0909 11:13:18.719645 29276 memory_optimize_pass.cc:199] Cluster name : im_shape size: 8 I0909 11:13:18.721639 29276 memory_optimize_pass.cc:199] Cluster name : scale_factor size: 8 I0909 11:13:18.722636 29276 memory_optimize_pass.cc:199] Cluster name : cast_0.tmp_0 size: 8 e[1me[35m--- Running analysis [ir_graph_to_program_pass]e[0m I0909 11:13:19.009511 29276 analysis_predictor.cc:595] ======= optimize end ======= I0909 11:13:19.009511 29276 naive_executor.cc:98] --- skip [feed], feed -> scale_factor I0909 11:13:19.010509 29276 naive_executor.cc:98] --- skip [feed], feed -> image I0909 11:13:19.013506 29276 naive_executor.cc:98] --- skip [feed], feed -> im_shape I0909 11:13:19.026468 29276 naive_executor.cc:98] --- skip [elementwise_add_2], fetch -> fetch I0909 11:13:19.026468 29276 naive_executor.cc:98] --- skip [batch_norm_6.tmp_3], fetch -> fetch W0909 11:13:19.067358 29276 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 5.0, Driver API Version: 11.0, Runtime API Version: 10.2 W0909 11:13:19.068353 29276 device_context.cc:422] device: 0, cuDNN Version: 7.6.
C++ Traceback (most recent call last):
Not support stack backtrace yet.
Error Message Summary:
InvalidArgumentError: The 'shape' attribute in ReshapeOp is invalid. The input tensor X'size must be divisible by known capacity of 'shape'. But received X's shape = [1, 18, 26, 26], X's size = 12168, 'shape' is [-1, 3, 6, 361], known capacity of 'shape' is -6498. [Hint: Expected output_shape[unk_dim_idx] * capacity == -in_size, but received output_shape[unk_dim_idx] * capacity:-6498 != -in_size:-12168.] (at C:/home/workspace/Paddle_release/paddle/fluid/operators/reshape_op.cc:208)
收到，谢谢！----- 原始邮件 @.>发送时间：2021-09-13 @.@.@.>主 题：Re: [PaddlePaddle/PaddleX] 使用ppyolov2模型进行预测时报错 (#1113)抱歉，这个是已知问题，该问题已修复，见pr https://github.com/PaddlePaddle/PaddleX/pull/1114/files目前可以安装develop版本的PaddleX，见https://github.com/PaddlePaddle/PaddleX/blob/release/2.0.0/docs/install.md#paddlex-develop%E5%AE%89%E8%A3%85也可以参考pr https://github.com/PaddlePaddle/PaddleX/pull/1114/files 直接修改安装目录下的paddlex:—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or unsubscribe.Triage notifications on the go with GitHub Mobile for iOS or Android.