search

apache / mxnet

Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
https://mxnet.apache.org
Apache License 2.0
20.78k stars 6.79k forks source link

quantization, excluded_sym_names doesn't work #16929

Open evi-Genius opened 4 years ago

evi-Genius commented 4 years ago

Description

I use the excluded_sym_nameswhen I test my custom model on imagenet_gen_qsym.py, but actually the symbols are not be excluded, that is the symbols are also be quantized. The model is convert from gluon.

leezu commented 4 years ago

Please specify the version of MXNet you are using. It's best to run the diagnosis script that you were requested to run when you opened this report..

evi-Genius commented 4 years ago

Please specify the version of MXNet you are using. It's best to run the diagnosis script that you were requested to run when you opened this report..

Environment

----------Python Info---------- Version : 3.5.2 Compiler : GCC 5.4.0 20160609 Build : ('default', 'Nov 23 2017 16:37:01') Arch : ('64bit', 'ELF') ------------Pip Info----------- Version : 19.1.1 Directory : /home/xiangyang/.virtualenvs/py35/lib/python3.5/site-packages/pip ----------MXNet Info----------- Version : 1.5.1 Directory : /home/xiangyang/.virtualenvs/py35/lib/python3.5/site-packages/mxnet Num GPUs : 4 Commit Hash : c9818480680f84daa6e281a974ab263691302ba8 ----------System Info---------- Platform : Linux-4.4.0-116-generic-x86_64-with-Ubuntu-16.04-xenial system : Linux node : jja-gpu034 release : 4.4.0-116-generic version : #140-Ubuntu SMP Mon Feb 12 21:23:04 UTC 2018 ----------Hardware Info---------- machine : x86_64 processor : x86_64 Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Byte Order: Little Endian CPU(s): 16 On-line CPU(s) list: 0-15 Thread(s) per core: 1 Core(s) per socket: 8 Socket(s): 2 NUMA node(s): 1 Vendor ID: GenuineIntel CPU family: 6 Model: 79 Model name: Intel(R) Xeon(R) CPU E5-2620 v4 @ 2.10GHz Stepping: 1 CPU MHz: 2089.582 CPU max MHz: 3000.0000 CPU min MHz: 1200.0000 BogoMIPS: 4201.45 Virtualization: VT-x L1d cache: 32K L1i cache: 32K L2 cache: 256K L3 cache: 20480K NUMA node0 CPU(s): 0-15 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch epb invpcid_single intel_pt retpoline kaiser tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm rdseed adx smap xsaveopt cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local dtherm arat pln pts ----------Network Test---------- Setting timeout: 10 Timing for GluonNLP GitHub: https://github.com/dmlc/gluon-nlp, DNS: 0.0015 sec, LOAD: 1.1211 sec. Timing for Conda: https://repo.continuum.io/pkgs/free/, DNS: 0.0008 sec, LOAD: 0.0412 sec. Timing for D2L (zh-cn): http://zh.d2l.ai, DNS: 0.0008 sec, LOAD: 0.0459 sec. Timing for FashionMNIST: https://repo.mxnet.io/gluon/dataset/fashion-mnist/train-labels-idx1-ubyte.gz, DNS: 0.0003 sec, LOAD: 0.6482 sec. Timing for PYPI: https://pypi.python.org/pypi/pip, DNS: 0.0035 sec, LOAD: 0.0439 sec. Timing for D2L: http://d2l.ai, DNS: 0.0008 sec, LOAD: 0.1977 sec. Timing for GluonNLP: http://gluon-nlp.mxnet.io, DNS: 0.0008 sec, LOAD: 0.0236 sec. Timing for MXNet: https://github.com/apache/incubator-mxnet, DNS: 0.0006 sec, LOAD: 1.2173 sec.

Steps to reproduce

(Paste the commands you ran that produced the error.)

1.python imagenet_gen_qsym_mkldnn.py --model=mobilenetv2_1.0 --calib-dataset=./data/val_256_q90.rec --num-calib-batches=5 --calib-mode=entropy 2.python imagenet_inference.py --symbol-file=./model/mobilenetv2_1.0-quantized-5batches-entropy-symbol.json --param-file=./model/mobilenetv2_1.0-0000.params --rgb-mean=123.68,116.779,103.939 --num-skipped-batches=50 --num-inference-batches=500 --dataset=./data/val_256_q90.rec --rgb-std=58.393,57.12,57.375

Error Message

Check failed: dshape[C] % 4 == 0U (3 vs. 0) : for 8bit cudnn conv, the number of channel must be multiple of 4

Description

the first layer is not be excluded, actually I have tested lots of excluded_sym_names but none of them work.

wuxun-zhang commented 4 years ago

@evi-Genius Are you running on GPU context? The models generated by imagenet_gen_qsym_mkldnn.py should be executed on CPU, please try to specify --ctx=cpu when you launched imagenet_inference.py.