Hylang Lisp 的语音识别: 先py跑起来,用hylangλ化,就像Clojureλ化Spark和安卓,前端一样,任何领域的Lispλ化
douda.py跑在CPU上正常,但是在GPU上跑,是报错的: douda_gpu_erro.txt
只要能描述清楚了(文学编程的最高境界),就能lispλ化=>只要特征能描述清楚了,就能hylisp可微分化
Clojure和Java的互操作,迁移到Hylisp和Python的互操作: test_hy.hy
编译生成pyctest_hy.pyc
测试的标准: anaconda3可以跑下面这段代码, 先安装anaconda3,再安装cuda(其实就是先有anaconda3的PATH变量,CUDA能找到就行了)
import tensorflow as tf
with tf .device ('/gpu:0' ):
a = tf .constant ([1.0 , 2.0 , 3.0 , 4.0 , 5.0 , 6.0 ], shape = [2 , 3 ], name = 'a' )
b = tf .constant ([1.0 , 2.0 , 3.0 , 4.0 , 5.0 , 6.0 ], shape = [3 , 2 ], name = 'b' )
c = tf .matmul (a , b )
# Creates a session with log_device_placement set to True.
sess = tf .Session (config = tf .ConfigProto (log_device_placement = True ))
# Runs the op.
print (sess .run (c ))
zshrc: 如果CUDA找不到PATH里面有anaconda3,就会去找系统自带的python,所以只有系统自带的python可以用GPU跑,而anaconda3不能(只能跑在CPU上)
export PATH=/home/hylisp/anaconda3/bin:$PATH
# ## anaconda3 死活都用不了GPU ==>> Ubuntu自带的Python可以
export PATH=/usr/local/cuda/bin${PATH: +: ${PATH} }
export LD_LIBRARY_PATH=/usr/local/cuda/lib64${LD_LIBRARY_PATH: +: ${LD_LIBRARY_PATH} }
export CUDA_HOME=/usr/local/cuda
alias vv=' vi ~/.zshrc ; source ~/.zshrc '
# # pip install --upgrade tensorflow-gpu
alias gd=' git diff '
alias gs=' git status '
# # https://github.com/globus/globus-jupyter-notebooks
alias pyweb=' jupyter notebook --ip="*" --no-browser '
# # .e.g: ➜ learn git:(master) ✗ jupyter notebook --ip="*" --no-browser
alias http=' python -m http.server 2222 '
alias e=' emacs -q -l ~/clojure_emacs/init.el '
alias gpu_t=' nvidia-smi -l '
# #export CUDA_VISIBLE_DEVICES=0
alias gch=' git checkout '