MobileNetV3 appears correct based on paper update, cleaned up comments and compacted last block def

pull/2/head
Ross Wightman 5 years ago
parent 20d66beead
commit db8ad25a23

@ -892,15 +892,11 @@ def _gen_mobilenet_v3(depth_multiplier, num_classes=1000, **kwargs):
# stage 2, 56x56 in
['ir_r3_k5_s2_e3_c40_se0.25_are'], # relu
# stage 3, 28x28 in
# FIXME are expansions here correct?
['ir_r1_k3_s2_e6_c80', 'ir_r1_k3_s1_e2.5_c80', 'ir_r2_k3_s1_e2.3_c80'], # hard-swish
# stage 4, 14x14in
['ir_r2_k3_s1_e6_c112_se0.25'], # hard-swish
# stage 5, 14x14in
# FIXME paper has a mistaken block-stride pattern 1-2-1 that doesn't fit the usual 2-1-..., ignoring
# The paper numbers result in an exp factor of 4.2 in the middle of this block, but keeping at 6
# results in a param count closer to 5.4m
['ir_r1_k5_s2_e6_c160_se0.25', 'ir_r1_k5_s1_e6_c160_se0.25', 'ir_r1_k5_s1_e6_c160_se0.25'], # hard-swish
['ir_r3_k5_s2_e6_c160_se0.25'], # hard-swish
# stage 6, 7x7 in
['cn_r1_k1_s1_c960'], # hard-swish
]

Loading…
Cancel
Save