nohup: ignoring input [W Context.cpp:69] Warning: torch.set_deterministic is in beta, and its design and functionality may change in the future. (function operator()) [08 06:15:54 :39] WRN migrate all stand-alone args into a single task [08 06:15:54 :159] set task task_0's 0th input model path as /root/axera/axera-quan-hjj/model/efficientnet_size-160.onnx [08 06:15:54 :178] set task task_0's 0th output model path as /root/axera/axera-quan-hjj/joint/efficientnetb0-handpose.joint [08 06:15:54 :297] set task task_0's pulsar_conf.output_dir as /root/axera/axera-quan-hjj /opt/venv/lib/python3.6/site-packages/torch/cuda/__init__.py:52: UserWarning: CUDA initialization: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx (Triggered internally at /pytorch/c10/cuda/CUDAFunctions.cpp:100.) return torch._C._cuda_getDeviceCount() > 0 [08 06:15:55 :17] planning task task_0 [08 06:15:55 :459] WRN affine_preprocess at QAT model compiling is deprecated, insert enforce_integers at front, please use scale_to_integers instead. [08 06:15:55 :340] ################## Running task task_0 ################## [08 06:15:55 :30] python3 /root/python_modules/super_pulsar/super_pulsar/toolchain_wrappers/wrapper_neuwizard.py --config /tmp/tmpmy092f06.prototxt [08 06:15:55 :36] DBG [neuwizard] [W Context.cpp:69] Warning: torch.set_deterministic is in beta, and its design and functionality may change in the future. (function operator()) [08 06:17:06 :36] DBG [neuwizard] ONNX Model Version 12 for "/root/axera/axera-quan-hjj/model/efficientnet_size-160.onnx" [08 06:17:06 :36] DBG [neuwizard] Planning load step finished; elapsed time: 0.01s [08 06:17:06 :36] DBG [neuwizard] Planning transform step finished; elapsed time: 0.01s [08 06:17:06 :36] DBG [neuwizard] Planning transform step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning transform step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "native" finished; elapsed time: 0.16s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "native_no_bn" finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "pretransformed" finished; elapsed time: 1.77s [08 06:17:06 :36] DBG [neuwizard] Planning calibrate step finished; elapsed time: 0.01s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "transformed" finished; elapsed time: 47.78s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "posttransformed" finished; elapsed time: 6.22s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "magma" finished; elapsed time: 1.51s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "magma_validified" finished; elapsed time: 0.34s [08 06:17:06 :36] DBG [neuwizard] Planning transform step to "lava_with_rtv" finished; elapsed time: 10.77s [08 06:17:06 :36] DBG [neuwizard] Planning dump_joint_model step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning evaluate step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning ir_bit_macs step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning ir_bit_float_params step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Planning ir_bit_quantized_params step finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Loading model finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Transforming to "onnx_step_1" finished; elapsed time: 0.01s [08 06:17:06 :36] DBG [neuwizard] Transforming to "onnx_step_2" finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Transforming to "onnx" finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Transforming to "native" finished; elapsed time: 0.07s [08 06:17:06 :36] DBG [neuwizard] Transforming to "native_no_bn" finished; elapsed time: 0.00s [08 06:17:06 :36] DBG [neuwizard] Transforming to "pretransformed" finished; elapsed time: 0.66s [08 06:17:06 :36] DBG [neuwizard] /opt/venv/lib/python3.6/site-packages/torch/cuda/__init__.py:52: UserWarning: CUDA initialization: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx (Triggered internally at /pytorch/c10/cuda/CUDAFunctions.cpp:100.) [08 06:17:06 :36] DBG [neuwizard] return torch._C._cuda_getDeviceCount() > 0 [08 06:17:30 :36] DBG [neuwizard] Calibrating finished; elapsed time: 7.91s [08 06:17:30 :36] DBG [neuwizard] Transforming to "transformed" finished; elapsed time: 8.75s [08 06:17:30 :36] DBG [neuwizard] Transforming to "posttransformed" finished; elapsed time: 1.83s [08 06:17:30 :36] DBG [neuwizard] Transforming to "magma" finished; elapsed time: 0.68s [08 06:17:30 :36] DBG [neuwizard] Transforming to "magma_validified" finished; elapsed time: 0.06s [08 06:17:30 :36] DBG [neuwizard] Transforming to "lava_with_rtv" finished; elapsed time: 4.61s [08 06:17:30 :36] DBG [neuwizard] Dynamically planning transform step to "lava" finished; elapsed time: 0.02s [08 06:17:30 :36] DBG [neuwizard] Transforming to "lava" dynamically finished; elapsed time: 0.06s [08 06:17:30 :36] DBG [neuwizard] Dynamically planning transform step to "lava_onnx" finished; elapsed time: 0.55s [08 06:17:30 :36] DBG [neuwizard] Transforming to "lava_onnx" dynamically finished; elapsed time: 0.78s [08 06:17:30 :36] DBG [neuwizard] Dynamically planning transform step to "lava_onnx_axe" finished; elapsed time: 0.02s [08 06:17:30 :36] DBG [neuwizard] Transforming to "lava_onnx_axe" dynamically finished; elapsed time: 0.06s [08 06:17:36 :36] DBG [neuwizard] /root/python_modules/neuwizard-latest/neuwizard/operators/lava/AX620/Conv2d.py:141: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at /pytorch/torch/csrc/utils/tensor_numpy.cpp:141.) [08 06:17:36 :36] DBG [neuwizard] Joint model dumpped as "/root/axera/axera-quan-hjj/joint/model.lava_joint" [08 06:17:36 :36] DBG [neuwizard] Dumping Joint Model finished; elapsed time: 4.17s [08 06:17:36 :36] DBG [neuwizard] Evaluation is not performed. [08 06:17:36 :36] DBG [neuwizard] Evaluating finished; elapsed time: 0.00s [08 06:17:36 :36] DBG [neuwizard] Overview Table of Bit MACs [08 06:17:36 :36] DBG [neuwizard] | Domain | native | pretransformed | transformed | posttransformed | magma | lava | [08 06:17:36 :36] DBG [neuwizard] |----------|----------|------------------|---------------|-------------------|---------|--------| [08 06:17:36 :36] DBG [neuwizard] | Bit MACs | 13G | 52G | 53G | 61G | 59G | 60G | [08 06:17:36 :36] DBG [neuwizard] Bit MACs measurement for each domain finished; elapsed time: 0.37s [08 06:17:36 :36] DBG [neuwizard] Overview Table of parameter size [08 06:17:36 :36] DBG [neuwizard] | Domain | native | pretransformed | transformed | posttransformed | magma | lava | [08 06:17:36 :36] DBG [neuwizard] |----------------------|----------|------------------|---------------|-------------------|---------|--------| [08 06:17:36 :36] DBG [neuwizard] | Parameter Size(bits) | 127M | 326M | 329M | 339M | 339M | 339M | [08 06:17:36 :36] DBG [neuwizard] Float Parameter size measurement for each domain finished; elapsed time: 0.54s [08 06:17:36 :36] DBG [neuwizard] Overview Table of parameter size [08 06:17:36 :36] DBG [neuwizard] | Domain | native | pretransformed | transformed | posttransformed | magma | lava | [08 06:17:36 :36] DBG [neuwizard] |----------------------|----------|------------------|---------------|-------------------|---------|--------| [08 06:17:36 :36] DBG [neuwizard] | Parameter Size(bits) | 32M | 81M | 82M | 85M | 84M | 84M | [08 06:17:36 :36] DBG [neuwizard] Quantized Parameter size measurement for each domain finished; elapsed time: 0.63s [08 06:17:39 :509] set op_8341's shape as [1, 8] [08 06:17:39 :509] set op_8341's shape as [1, 8] [08 06:17:41 :535] DBG working in "/root/tmpo0o7ztk9" [08 06:17:41 :227] python3 pulsar.py gen /root/tmpo0o7ztk9/part_0.lava/part_0.lava env/ax620a_virtual_111_config.ini -b 1 -pe 16 --times_thres 0 --job_stealing 3 --checkall --param_compress --continuous_input --no_sim --hyper_params run_cf.wait_mode=True [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] inference_report.log: [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:-------------------------|:-----------------------------|:-------------|:-------------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:----------------|-------------:|---------------:| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:---------------|:-----------------|:-------------|:--------------|:------------|:---------------|:---------------|:-------------------|:---------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:-----------------------|:------------------|:------------------|:------------------|:--------------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:-----------------------|:-----------------|:--------------------|:--------------------|:--------------------|:------------------|:-----------------|:----------------|:----------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] profile stream EU: ld/st_ratio might include ringbuf/linebuf/feature_swap parts; mv_ratio migth have ringbuf part. [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:--------------------|-----------:|:--------------|:-----------------|:---------------|:----------------------|:--------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:---------------------------------------------|:----------------|:-----------------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:-----------|-----------:|----------:|:--------|-------:|----------:| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] inference: 23.6 ms, 42.35 fps [08 06:25:32 :250] DBG [pulsar] qps = fps * batch_size = 42.35 [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] simulated fps is based on DDR_BW: 1.59 GB/s [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] DDR IO stats: [08 06:25:32 :250] DBG [pulsar] ideal_input_data_size: 14116608 Byte, [08 06:25:32 :250] DBG [pulsar] ideal_output_data_size: 32 Byte, [08 06:25:32 :250] DBG [pulsar] extra_mid_io_data_size: 7109136 Byte, [08 06:25:32 :250] DBG [pulsar] total_io_data_size: 21225776 Byte [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] MAC per inference: 628214464 MAC@int8 [08 06:25:32 :250] DBG [pulsar] MAC utils: 2.89 % [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] commit_id: [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] |:------------------------------------|-----------:|:-------------| [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] subgraph num: 4 [08 06:25:32 :250] DBG [pulsar]  [08 06:25:32 :250] DBG [pulsar] pulsar.py totally used 470s [08 06:25:38 :582] File saved: /root/axera/axera-quan-hjj/joint/efficientnetb0-handpose.joint [08 06:25:38 :587] DBG cleared /root/tmpo0o7ztk9 | Pre_alloc OCM | linebuffer(may SWAP later) | ringbuffer | parameter | | size(ratio of whole OCM) | 0(0.0)% | 0(0.0)% | 116736(5.6)% | | range | (None, None) | (None, None) | (1980416, 2097152) | | Pre_alloc DDR | ringbuffer | feature_swap | | size(M) | 0.00 | 0.00 | | profile conv | work_cyc | linebuf | warmup_tail | core_idle | io_idle | stride2_idle | standalone_fetch | MAC | | ratio in conv | 2772180 (100.0%) | 32131 (1.2%) | 240800 (8.7%) | 0 (0.0%) | 710439 (25.6%) | 831608 (30.0%) | 210 (0.0%) | 545325 (19.7%) | | profile ideal DDR_IO | min_io_sum | min_params_read | min_inputs_read | min_outputs_write | | DDR IO size (Byte) | 14116640 (100.0%) | 14039808 (99.5%) | 76800 (0.5%) | 32 (0.0%) | | profile extra DDR_IO | extra_ddr_io | extra_params_read | extra_inputs_read | extra_outputs_wrt | extra_swap_read | extra_swap_wrt | ddr_rb_read | ddr_rb_wrt | | DDR IO size (Byte) | 7109136 (100.0%) | 0 (0.0%) | 1157384 (16.3%) | 1106184 (15.6%) | 0 (0.0%) | 0 (0.0%) | 2422784 (34.1%) | 2422784 (34.1%) | | profile stream EU | work_cyc | ld_ratio | ld_param_ratio | mv_ratio | mv_linebuffer_ratio | st_ratio | | teng | 18636864 | 1844691(9.9%) | 7460162(40.0%) | 7509589(40.3%) | 40684(0.2%) | 1781737(9.6%) | | breakdown of mv_ratio in profile stream EU | teng | all_eus with mv-cmds | | total_cyc_num | 7509589(100.0%) | 7509589(100.0%) | | weight0_mode,convNxM,mode23,nopad | 2404764(32.0%) | 2404764(32.0%) | | teng_binary_mul | 1056050(14.1%) | 1056050(14.1%) | | mv,affine,unpack_lsb | 565295(7.5%) | 565295(7.5%) | | mv,broadcast | 483148(6.4%) | 483148(6.4%) | | teng_binary_sub | 451312(6.0%) | 451312(6.0%) | | mv,resize | 439020(5.8%) | 439020(5.8%) | | mv,padding | 389020(5.2%) | 389020(5.2%) | | mv,subtensor | 365393(4.9%) | 365393(4.9%) | | mv,concat_c | 328182(4.4%) | 328182(4.4%) | | teng_binary_add | 308771(4.1%) | 308771(4.1%) | | mv,pack_lsb | 267240(3.6%) | 267240(3.6%) | | weight0_mode,convNxM,mode20,nopad | 251040(3.3%) | 251040(3.3%) | | conv_transform | 85377(1.1%) | 85377(1.1%) | | mv,padding_ch | 63717(0.8%) | 63717(0.8%) | | const | 44401(0.6%) | 44401(0.6%) | | sigmoid | 6129(0.1%) | 6129(0.1%) | | weight0_mode,mode26,fc | 501(0.0%) | 501(0.0%) | | revert_split | 229(0.0%) | 229(0.0%) | | EU | work_cyc | tot_cyc | ratio | fps | fps_bnd | | conv-1core | 2772181 | 18888839 | 14.0% | 42.350 | 288.580 | | teng | 18636864 | 18888839 | 98.0% | 42.350 | 42.930 | | breakdown of cmds_num for each op | cmds_num | percentage | | weight0_mode,convNxM,mode23,nopad | 4590 | 35.20% | | mv,resize | 1306 | 10.01% | | mv,padding | 929 | 7.12% | | mv,affine,unpack_lsb | 913 | 7.00% | | mv,concat_c | 863 | 6.62% | | mv,subtensor | 748 | 5.74% | | weight0_mode,mode23,nopad | 699 | 5.36% | | weight0_mode,convNxM,mode26,nopad | 585 | 4.49% | | weight0_mode,mode23 | 543 | 4.16% | | weight0_mode,convNxM,mode20,nopad | 492 | 3.77% | | weight0_mode,mode20,nopad | 359 | 2.75% | | const | 199 | 1.53% | | conv_transform | 152 | 1.17% | | revert_split | 124 | 0.95% | | teng_binary_mul | 118 | 0.90% | | mv,padding_ch | 87 | 0.67% | | teng_binary_sub | 77 | 0.59% | | mv,broadcast | 76 | 0.58% | | teng_binary_add | 66 | 0.51% | | mv,pack_lsb | 53 | 0.41% | | sigmoid | 35 | 0.27% | | weight0_mode,mode20 | 13 | 0.10% | | weight0_mode,mode26,fc | 8 | 0.06% | | bgr_pack | 6 | 0.05% | | total_num | 13041 | 100% | 06:25:37 [I]final_check: FINAL_CHECK: output op_8341 check successed!