-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathdse-rf.py
165 lines (156 loc) · 6.66 KB
/
dse-rf.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
import os
import sys
working_dir = os.path.dirname(os.path.abspath(__file__))
os.chdir(working_dir) # change working directory to the location of this file
sys.path.append(".") # Adds higher directory to python modules path.
import argparse
import torch
from skopt import Optimizer, Space
from skopt.space import Integer
from util.common import fold_maker, set_seed
from analytic.get_alg_properties import get_alg_info
bench_path = os.path.join(working_dir, 'analytic/benchmarks')
alg = get_alg_info(os.path.join(bench_path, 'fft/fft-pisa.out'), os.path.join(bench_path, 'fft/fft_results.txt'))
def parse_args():
parser = argparse.ArgumentParser(description='Boom script')
parser.add_argument('--n_sample', default='10', type=int, help='episode number for the whole training')
parser.add_argument('--area_limit', default='8', type=float, help='limit of the area')
parser.add_argument('--n_init', default='5', type=int, help='init points where action is randomly chosen')
parser.add_argument('--seed', default='1', type=int, help='seed for torch and random')
parser.add_argument('--log_path', default=os.path.join(working_dir, 'logs/RF'), type=str, help='file path of log files')
return parser.parse_args()
def evaluate(params, benchmarks, log_path, area_limit):
from hf.vcs import get_cpi_vcs
from analytic.Analytic_model_torch import McPAT, CPI
ds = set_system(torch.tensor(params, dtype=torch.float64))
area = McPAT(ds, alg, 0.5, 0.5, 0.5)
if area > area_limit:
return -1.1111, 5
else:
# cpi_avg = CPI(ds, alg)
# reward = cpi_avg
cpi_total = 0
# with open(os.path.join(log_path, "hf-progress.txt"),"a") as f: f.write('\n{}\n'.format(params))
for bm in benchmarks:
cpi = get_cpi_vcs(params, bm, log_path)
cpi_total += cpi
cpi_avg = cpi_total / len(benchmarks)
reward = cpi_avg
return float(reward), float(cpi_avg)
def param_regulator(params):
params = [int(x) for x in params]
# print('params', params, end='')
design = []
design.append(2**(params[0]+4))
design.append(2**(params[1]+1))
design.append(2**(params[2]+7))
design.append(2**(params[3]+1))
design.append(2*(params[4]+1))
design.append(params[5]+1)
design.append(32*(params[6]+1))
design.append(params[7]+1)
design.append(params[8]+1)
design.append(params[9]+1)
design.append(min(2**(params[10]+1), 24))
max_FU = max(design[7], design[8], design[9])
design[5] = max(design[5], max_FU) # decode width should be larger than FU issue width
if design[5] == 5:
design[10] = max(design[10], 8)
elif design[5] >= 3:
design[10] = max(design[10], 4)
return design
def set_system(params):
system = {
"L1LineSize": 64, #byte
"L1sets": params[0],
"L1ways": params[1],
"L2LineSize": 64, #byte
"L2sets": params[2],
"L2ways": params[3],
"L1latency": 4,
"L2latency": 21,
"DRAMlatency": 274,
"issue_width": params[7]+params[8]+params[9],
"mshr": params[4],
"dispatch": params[5],
"FUint": params[7],
"FUfp": params[8],
"FUmem": params[9],
"FUcontrol": 1,
"Cycle_int_op": 1,
"Cycle_int_mul": 3,
"Cycle_int_div": 18,
"Cycle_fp_op": 3,
"Cycle_fp_mul": 5,
"Cycle_fp_div": 10,
"BWCoreL1": 139586437120,
"BWL1L2": 42949672960,
"BWL2DRAM": 85899345920,
"freq": 2e9,
"frontpipe": 5,
"ROB": params[6],
"IQ": params[10],
}
return system
if __name__ == '__main__':
args = parse_args()
args.log_path = fold_maker(args.log_path)
print('Logs stored in {}.'.format(args.log_path))
set_seed(args.seed)
with open(os.path.join(args.log_path, "details.txt"),"a") as f:
f.write('seed: {}\n'.format(args.seed))
transform = 'identity'
problem_space = [
Integer(low=0, high=2, prior='uniform', transform=transform, name="L1set"),
Integer(low=0, high=3, prior='uniform', transform=transform, name="L1way"),
Integer(low=0, high=4, prior='uniform', transform=transform, name="L2set"),
Integer(low=0, high=3, prior='uniform', transform=transform, name="L2way"),
Integer(low=0, high=4, prior='uniform', transform=transform, name="MSHR"),
Integer(low=0, high=4, prior='uniform', transform=transform, name="decode_width"),
Integer(low=0, high=4, prior='uniform', transform=transform, name="rob"),
Integer(low=0, high=4, prior='uniform', transform=transform, name="int"),
Integer(low=0, high=1, prior='uniform', transform=transform, name="fp"),
Integer(low=0, high=1, prior='uniform', transform=transform, name="mem"),
Integer(low=0, high=4, prior='uniform', transform=transform, name="IQ"),
]
benchmarks = ['dijkstra', 'mm-405060-456', 'vvadd1000', 'qsort8192', 'fft', 'stringsearch']
# gbrt = GradientBoostingRegressor(**HBO_params_cpi)
# gbrt_estimator = GradientBoostingQuantileRegressor(base_estimator=gbrt)
# base_estimator = BaggingRegressor_std(estimator=gbrt_estimator, **HBO_params_ada_cpi)
opt = Optimizer(
dimensions=problem_space,
base_estimator='RF',
n_initial_points=args.n_init,
initial_point_generator="TED",#orthogonal
acq_func="LCB", #minimization version of UCB
acq_optimizer="sampling", # "auto",
random_state=args.seed,
n_jobs=-1,
model_queue_size=1,
acq_func_kwargs=None, # {"xi": 0.000001, "kappa": 0.001} #favor exploitaton
acq_optimizer_kwargs={"n_points": 10},
)
valid_pool=[]
cpi_pool=[]
counter = 0
for iter in range(200):
next_x = opt.ask()
params = param_regulator(next_x)
print(iter, '-', params, end=' ')
reward, cpi = evaluate(params, benchmarks, args.log_path, args.area_limit)
print('reward:', round(reward,4), ' cpi:', round(cpi,4))
with open(os.path.join(args.log_path, "details.txt"),"a") as f:
f.write('{} - '.format(iter))
f.write('params:{} '.format(params))
f.write('reward:{} '.format(round(reward, 4)))
f.write('cpi:{}\n'.format(round(cpi, 4))) if cpi>0 else f.write('\n')
if reward > 0 and params not in valid_pool:
counter += 1
valid_pool.append(params)
cpi_pool.append(cpi)
if counter >= args.n_sample: break
res = opt.tell(next_x, cpi)
# print(valid_pool)
print('best cpi: {}'.format(min(cpi_pool)))
with open(os.path.join(args.log_path, "details.txt"),"a") as f:
f.write('best cpi: {}'.format(min(cpi_pool)))