-
Notifications
You must be signed in to change notification settings - Fork 147
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Finish soft quantize training script
- Loading branch information
Showing
4 changed files
with
186 additions
and
33 deletions.
There are no files selected for viewing
Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.
Oops, something went wrong.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,156 @@ | ||
| """ | ||
| Attempt to perform soft quantizing on face 12 net (with params -q, 0, q) | ||
| Must be executed under caffe-master folder | ||
| Training results will be saved as described in solver.prototxt | ||
| """ | ||
|
|
||
| import numpy as np | ||
| import sys | ||
| sys.path.append('/home/anson/PycharmProjects/face_net_surgery') | ||
| from quantize_functions import * | ||
|
|
||
| netKind = 24 | ||
| quantizeBitNum = 5 | ||
| stochasticRounding = False | ||
|
|
||
| # ================== caffe ====================================== | ||
| caffe_root = '/home/anson/caffe-master/' # this file is expected to be in {caffe_root}/examples | ||
| sys.path.insert(0, caffe_root + 'python') | ||
| import caffe | ||
|
|
||
| # ================== load trained high precision params ====================================== | ||
| MODEL_FILE = '/home/anson/caffe-master/models/face_' + str(netKind) + '_cal/train_val.prototxt' | ||
| PRETRAINED = '/home/anson/caffe-master/models/face_' + str(netKind) \ | ||
| + '_cal/face_' + str(netKind) + '_cal_train_iter_400000.caffemodel' | ||
| caffe.set_mode_gpu() | ||
| net = caffe.Net(MODEL_FILE, PRETRAINED, caffe.TRAIN) | ||
| # ============ should be modified for different files ================ | ||
| if netKind == 12: | ||
| params = ['conv1', 'fc2', 'fc3'] | ||
| elif netKind == 24: | ||
| params = ['conv1', 'fc2', 'fc3'] | ||
| elif netKind == 48: | ||
| params = ['conv1', 'conv2', 'fc3', 'fc4'] | ||
| # ===================================================================== | ||
| # fc_params = {name: (weights, biases)} | ||
| original_params = {pr: (net.params[pr][0].data, net.params[pr][1].data) for pr in params} | ||
|
|
||
| # ================== load file to save quantized parameters ======================= | ||
| MODEL_FILE = '/home/anson/caffe-master/models/face_' + str(netKind) + '_cal/train_val.prototxt' | ||
| if stochasticRounding: | ||
| PRETRAINED = '/home/anson/caffe-master/models/face_' + str(netKind) + '_cal/face_' \ | ||
| + str(netKind) + '_cal_soft_SRquantize_' + str(quantizeBitNum) +'.caffemodel' | ||
| else: | ||
| PRETRAINED = '/home/anson/caffe-master/models/face_' + str(netKind) + '_cal/face_' \ | ||
| + str(netKind) + '_cal_soft_quantize_' + str(quantizeBitNum) +'.caffemodel' | ||
| quantized_model = open(PRETRAINED, 'w') | ||
| net_quantized = caffe.Net(MODEL_FILE, PRETRAINED, caffe.TRAIN) | ||
| params_quantized = params | ||
| # conv_params = {name: (weights, biases)} | ||
| quantized_params = {pr: (net_quantized.params[pr][0].data, net_quantized.params[pr][1].data) for pr in params_quantized} | ||
|
|
||
| # transplant params | ||
| for pr, pr_quantized in zip(params, params_quantized): | ||
| quantized_params[pr_quantized][0].flat = original_params[pr][0].flat # flat unrolls the arrays | ||
| quantized_params[pr_quantized][1][...] = original_params[pr][1] | ||
|
|
||
| # =============== training ============================= | ||
| # Training net is called net_training | ||
| # At each iteration: | ||
| # 1. Assign high precision params to net_training, and quantize | ||
| # 2. Perform 1 step in net_training | ||
| # 3. Update params diff to high precision params in net_quantized | ||
| # At last iteration, quantize high precision params | ||
| # ====================================================== | ||
|
|
||
| solver = caffe.SGDSolver('/home/anson/caffe-master/models/face_' + str(netKind) + '_cal/solver.prototxt') | ||
| net_training = solver.net | ||
| params_train = params | ||
|
|
||
| # training_params[pr_train][0].flat = original_params[pr][0].flat # flat unrolls the arrays | ||
| # training_params[pr_train][1][...] = original_params[pr][1] | ||
| for numOfIterations in range(400000): | ||
| # if numOfIterations % 250 == 0: | ||
| # print "Current iteration : " + str(numOfIterations) | ||
| # =========== 1. Assign high precision params to net_training, and quantize =========== | ||
| quantized_params = {pr: (net_quantized.params[pr][0].data, net_quantized.params[pr][1].data) for pr in params_quantized} | ||
| training_params = {pr: (net_training.params[pr][0].data, net_training.params[pr][1].data) for pr in params_train} | ||
| # transplant params | ||
| for pr_quantized, pr_train in zip(params_quantized, params_train): | ||
| training_params[pr_train][0].flat = quantized_params[pr_quantized][0].flat # flat unrolls the arrays | ||
| training_params[pr_train][1][...] = quantized_params[pr_quantized][1] | ||
|
|
||
| for k, v in net_training.params.items(): | ||
| filters_weights = net_training.params[k][0].data | ||
| filters_bias = net_training.params[k][1].data | ||
|
|
||
| # ============ should be modified for different files ================ | ||
| if netKind == 12: | ||
| if k == 'conv1': | ||
| a_weight = 0 | ||
| a_bias = -4 | ||
| elif k == 'fc2': | ||
| a_weight = -2 | ||
| a_bias = 1 | ||
| elif k == 'fc3': | ||
| a_weight = -5 | ||
| a_bias = 0 | ||
|
|
||
| elif netKind == 24: | ||
| if k == 'conv1': | ||
| a_weight = -1 | ||
| a_bias = -3 | ||
| elif k == 'fc2': | ||
| a_weight = -3 | ||
| a_bias = 0 | ||
| elif k == 'fc3': | ||
| a_weight = -5 | ||
| a_bias = 0 | ||
|
|
||
| elif netKind == 48: | ||
| if k == 'conv1': | ||
| a_weight = -1 | ||
| a_bias = -6 | ||
| elif k == 'conv2': | ||
| a_weight = -1 | ||
| a_bias = -6 | ||
| elif k == 'fc3': | ||
| a_weight = -4 | ||
| a_bias = 1 | ||
| elif k == 'fc4': | ||
| a_weight = -6 | ||
| a_bias = 0 | ||
|
|
||
| # ===================================================================== | ||
| b_weight = quantizeBitNum - 1 - a_weight | ||
| b_bias = quantizeBitNum - 1 - a_bias | ||
| # lists of all possible values under current quantized bit num | ||
| if quantizeBitNum == 2: | ||
| weightFixedPointList = tri_section_points(a_weight) | ||
| biasFixedPointList = tri_section_points(a_bias) | ||
| else: | ||
| weightFixedPointList = fixed_point_list(a_weight, b_weight) | ||
| biasFixedPointList = fixed_point_list(a_bias, b_bias) | ||
|
|
||
| for currentNum in np.nditer(filters_weights, op_flags=['readwrite']): | ||
| currentNum[...] = round_number(currentNum[...], weightFixedPointList, stochasticRounding) | ||
|
|
||
| for currentNum in np.nditer(filters_bias, op_flags=['readwrite']): | ||
| currentNum[...] = round_number(currentNum[...], biasFixedPointList, stochasticRounding) | ||
| # ============ End of rounding params ================ | ||
| # =========== 2. Perform 1 step in net_training =========== | ||
| solver.step(1) | ||
| # =========== 3. Update params diff to high precision params in net_quantized =========== | ||
| for k, v in net_training.params.items(): | ||
| # print (k, v[0].data.shape) | ||
| net_quantized.params[k][0].data[...] -= net_training.params[k][0].diff | ||
| net_quantized.params[k][1].data[...] -= net_training.params[k][1].diff | ||
| filters_weights = net_quantized.params[k][0].data | ||
| filters_bias = net_quantized.params[k][1].data | ||
|
|
||
| # if numOfIterations % 10000 == 0: | ||
| # net_quantized.save('/home/anson/caffe-master/models/face_12c/face_12c_soft_quantize_' \ | ||
| # + str(quantizeBitNum) + '_' + str(numOfIterations) + '.caffemodel') | ||
|
|
||
|
|
||
|
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters