README.md

metadata

title: Dog Breed ImageWoof
emoji: ⚡
colorFrom: blue
colorTo: indigo
sdk: gradio
sdk_version: 3.17.0
app_file: app.py
pinned: false
license: mit

ImageWoof Classification

Click to visit the Github Repo ## Problem Statement And Description A subset of 10 harder to classify classes from Imagenet (all dog breeds): Australian terrier, Border terrier, Samoyed, beagle, Shih-Tzu, English foxhound, Rhodesian ridgeback, dingo, golden retriever, Old English sheepdog. An EfficientNetB2 feature extractor computer vision model to classify images of Dog breeds was created. summary(eff_b2, (3,224,224),device='cpu')

    Layer (type)               Output Shape         Param #

================================================================ Conv2d-1 [-1, 32, 112, 112] 864 BatchNorm2d-2 [-1, 32, 112, 112] 64 SiLU-3 [-1, 32, 112, 112] 0 Conv2d-4 [-1, 32, 112, 112] 288 BatchNorm2d-5 [-1, 32, 112, 112] 64 SiLU-6 [-1, 32, 112, 112] 0 AdaptiveAvgPool2d-7 [-1, 32, 1, 1] 0 Conv2d-8 [-1, 8, 1, 1] 264 SiLU-9 [-1, 8, 1, 1] 0 Conv2d-10 [-1, 32, 1, 1] 288 Sigmoid-11 [-1, 32, 1, 1] 0 SqueezeExcitation-12 [-1, 32, 112, 112] 0 Conv2d-13 [-1, 16, 112, 112] 512 BatchNorm2d-14 [-1, 16, 112, 112] 32 MBConv-15 [-1, 16, 112, 112] 0 Conv2d-16 [-1, 16, 112, 112] 144 BatchNorm2d-17 [-1, 16, 112, 112] 32 SiLU-18 [-1, 16, 112, 112] 0 AdaptiveAvgPool2d-19 [-1, 16, 1, 1] 0 Conv2d-20 [-1, 4, 1, 1] 68 SiLU-21 [-1, 4, 1, 1] 0 Conv2d-22 [-1, 16, 1, 1] 80 Sigmoid-23 [-1, 16, 1, 1] 0 SqueezeExcitation-24 [-1, 16, 112, 112] 0 Conv2d-25 [-1, 16, 112, 112] 256 BatchNorm2d-26 [-1, 16, 112, 112] 32 StochasticDepth-27 [-1, 16, 112, 112] 0 MBConv-28 [-1, 16, 112, 112] 0 Conv2d-29 [-1, 96, 112, 112] 1,536 BatchNorm2d-30 [-1, 96, 112, 112] 192 SiLU-31 [-1, 96, 112, 112] 0 Conv2d-32 [-1, 96, 56, 56] 864 BatchNorm2d-33 [-1, 96, 56, 56] 192 SiLU-34 [-1, 96, 56, 56] 0 AdaptiveAvgPool2d-35 [-1, 96, 1, 1] 0 Conv2d-36 [-1, 4, 1, 1] 388 SiLU-37 [-1, 4, 1, 1] 0 Conv2d-38 [-1, 96, 1, 1] 480 Sigmoid-39 [-1, 96, 1, 1] 0 SqueezeExcitation-40 [-1, 96, 56, 56] 0 Conv2d-41 [-1, 24, 56, 56] 2,304 BatchNorm2d-42 [-1, 24, 56, 56] 48 MBConv-43 [-1, 24, 56, 56] 0 Conv2d-44 [-1, 144, 56, 56] 3,456 BatchNorm2d-45 [-1, 144, 56, 56] 288 SiLU-46 [-1, 144, 56, 56] 0 Conv2d-47 [-1, 144, 56, 56] 1,296 BatchNorm2d-48 [-1, 144, 56, 56] 288 SiLU-49 [-1, 144, 56, 56] 0 AdaptiveAvgPool2d-50 [-1, 144, 1, 1] 0 Conv2d-51 [-1, 6, 1, 1] 870 SiLU-52 [-1, 6, 1, 1] 0 Conv2d-53 [-1, 144, 1, 1] 1,008 Sigmoid-54 [-1, 144, 1, 1] 0 SqueezeExcitation-55 [-1, 144, 56, 56] 0 Conv2d-56 [-1, 24, 56, 56] 3,456 BatchNorm2d-57 [-1, 24, 56, 56] 48 StochasticDepth-58 [-1, 24, 56, 56] 0 MBConv-59 [-1, 24, 56, 56] 0 Conv2d-60 [-1, 144, 56, 56] 3,456 BatchNorm2d-61 [-1, 144, 56, 56] 288 SiLU-62 [-1, 144, 56, 56] 0 Conv2d-63 [-1, 144, 56, 56] 1,296 BatchNorm2d-64 [-1, 144, 56, 56] 288 SiLU-65 [-1, 144, 56, 56] 0 AdaptiveAvgPool2d-66 [-1, 144, 1, 1] 0 Conv2d-67 [-1, 6, 1, 1] 870 SiLU-68 [-1, 6, 1, 1] 0 Conv2d-69 [-1, 144, 1, 1] 1,008 Sigmoid-70 [-1, 144, 1, 1] 0 SqueezeExcitation-71 [-1, 144, 56, 56] 0 Conv2d-72 [-1, 24, 56, 56] 3,456 BatchNorm2d-73 [-1, 24, 56, 56] 48 StochasticDepth-74 [-1, 24, 56, 56] 0 MBConv-75 [-1, 24, 56, 56] 0 Conv2d-76 [-1, 144, 56, 56] 3,456 BatchNorm2d-77 [-1, 144, 56, 56] 288 SiLU-78 [-1, 144, 56, 56] 0 Conv2d-79 [-1, 144, 28, 28] 3,600 BatchNorm2d-80 [-1, 144, 28, 28] 288 SiLU-81 [-1, 144, 28, 28] 0 AdaptiveAvgPool2d-82 [-1, 144, 1, 1] 0 Conv2d-83 [-1, 6, 1, 1] 870 SiLU-84 [-1, 6, 1, 1] 0 Conv2d-85 [-1, 144, 1, 1] 1,008 Sigmoid-86 [-1, 144, 1, 1] 0 SqueezeExcitation-87 [-1, 144, 28, 28] 0 Conv2d-88 [-1, 48, 28, 28] 6,912 BatchNorm2d-89 [-1, 48, 28, 28] 96 MBConv-90 [-1, 48, 28, 28] 0 Conv2d-91 [-1, 288, 28, 28] 13,824 BatchNorm2d-92 [-1, 288, 28, 28] 576 SiLU-93 [-1, 288, 28, 28] 0 Conv2d-94 [-1, 288, 28, 28] 7,200 BatchNorm2d-95 [-1, 288, 28, 28] 576 SiLU-96 [-1, 288, 28, 28] 0 AdaptiveAvgPool2d-97 [-1, 288, 1, 1] 0 Conv2d-98 [-1, 12, 1, 1] 3,468 SiLU-99 [-1, 12, 1, 1] 0 Conv2d-100 [-1, 288, 1, 1] 3,744 Sigmoid-101 [-1, 288, 1, 1] 0 SqueezeExcitation-102 [-1, 288, 28, 28] 0 Conv2d-103 [-1, 48, 28, 28] 13,824 BatchNorm2d-104 [-1, 48, 28, 28] 96 StochasticDepth-105 [-1, 48, 28, 28] 0 MBConv-106 [-1, 48, 28, 28] 0 Conv2d-107 [-1, 288, 28, 28] 13,824 BatchNorm2d-108 [-1, 288, 28, 28] 576 SiLU-109 [-1, 288, 28, 28] 0 Conv2d-110 [-1, 288, 28, 28] 7,200 BatchNorm2d-111 [-1, 288, 28, 28] 576 SiLU-112 [-1, 288, 28, 28] 0 AdaptiveAvgPool2d-113 [-1, 288, 1, 1] 0 Conv2d-114 [-1, 12, 1, 1] 3,468 SiLU-115 [-1, 12, 1, 1] 0 Conv2d-116 [-1, 288, 1, 1] 3,744 Sigmoid-117 [-1, 288, 1, 1] 0 SqueezeExcitation-118 [-1, 288, 28, 28] 0 Conv2d-119 [-1, 48, 28, 28] 13,824 BatchNorm2d-120 [-1, 48, 28, 28] 96 StochasticDepth-121 [-1, 48, 28, 28] 0 MBConv-122 [-1, 48, 28, 28] 0 Conv2d-123 [-1, 288, 28, 28] 13,824 BatchNorm2d-124 [-1, 288, 28, 28] 576 SiLU-125 [-1, 288, 28, 28] 0 Conv2d-126 [-1, 288, 14, 14] 2,592 BatchNorm2d-127 [-1, 288, 14, 14] 576 SiLU-128 [-1, 288, 14, 14] 0 AdaptiveAvgPool2d-129 [-1, 288, 1, 1] 0 Conv2d-130 [-1, 12, 1, 1] 3,468 SiLU-131 [-1, 12, 1, 1] 0 Conv2d-132 [-1, 288, 1, 1] 3,744 Sigmoid-133 [-1, 288, 1, 1] 0 SqueezeExcitation-134 [-1, 288, 14, 14] 0 Conv2d-135 [-1, 88, 14, 14] 25,344 BatchNorm2d-136 [-1, 88, 14, 14] 176 MBConv-137 [-1, 88, 14, 14] 0 Conv2d-138 [-1, 528, 14, 14] 46,464 BatchNorm2d-139 [-1, 528, 14, 14] 1,056 SiLU-140 [-1, 528, 14, 14] 0 Conv2d-141 [-1, 528, 14, 14] 4,752 BatchNorm2d-142 [-1, 528, 14, 14] 1,056 SiLU-143 [-1, 528, 14, 14] 0 AdaptiveAvgPool2d-144 [-1, 528, 1, 1] 0 Conv2d-145 [-1, 22, 1, 1] 11,638 SiLU-146 [-1, 22, 1, 1] 0 Conv2d-147 [-1, 528, 1, 1] 12,144 Sigmoid-148 [-1, 528, 1, 1] 0 SqueezeExcitation-149 [-1, 528, 14, 14] 0 Conv2d-150 [-1, 88, 14, 14] 46,464 BatchNorm2d-151 [-1, 88, 14, 14] 176 StochasticDepth-152 [-1, 88, 14, 14] 0 MBConv-153 [-1, 88, 14, 14] 0 Conv2d-154 [-1, 528, 14, 14] 46,464 BatchNorm2d-155 [-1, 528, 14, 14] 1,056 SiLU-156 [-1, 528, 14, 14] 0 Conv2d-157 [-1, 528, 14, 14] 4,752 BatchNorm2d-158 [-1, 528, 14, 14] 1,056 SiLU-159 [-1, 528, 14, 14] 0 AdaptiveAvgPool2d-160 [-1, 528, 1, 1] 0 Conv2d-161 [-1, 22, 1, 1] 11,638 SiLU-162 [-1, 22, 1, 1] 0 Conv2d-163 [-1, 528, 1, 1] 12,144 Sigmoid-164 [-1, 528, 1, 1] 0 SqueezeExcitation-165 [-1, 528, 14, 14] 0 Conv2d-166 [-1, 88, 14, 14] 46,464 BatchNorm2d-167 [-1, 88, 14, 14] 176 StochasticDepth-168 [-1, 88, 14, 14] 0 MBConv-169 [-1, 88, 14, 14] 0 Conv2d-170 [-1, 528, 14, 14] 46,464 BatchNorm2d-171 [-1, 528, 14, 14] 1,056 SiLU-172 [-1, 528, 14, 14] 0 Conv2d-173 [-1, 528, 14, 14] 4,752 BatchNorm2d-174 [-1, 528, 14, 14] 1,056 SiLU-175 [-1, 528, 14, 14] 0 AdaptiveAvgPool2d-176 [-1, 528, 1, 1] 0 Conv2d-177 [-1, 22, 1, 1] 11,638 SiLU-178 [-1, 22, 1, 1] 0 Conv2d-179 [-1, 528, 1, 1] 12,144 Sigmoid-180 [-1, 528, 1, 1] 0 SqueezeExcitation-181 [-1, 528, 14, 14] 0 Conv2d-182 [-1, 88, 14, 14] 46,464 BatchNorm2d-183 [-1, 88, 14, 14] 176 StochasticDepth-184 [-1, 88, 14, 14] 0 MBConv-185 [-1, 88, 14, 14] 0 Conv2d-186 [-1, 528, 14, 14] 46,464 BatchNorm2d-187 [-1, 528, 14, 14] 1,056 SiLU-188 [-1, 528, 14, 14] 0 Conv2d-189 [-1, 528, 14, 14] 13,200 BatchNorm2d-190 [-1, 528, 14, 14] 1,056 SiLU-191 [-1, 528, 14, 14] 0 AdaptiveAvgPool2d-192 [-1, 528, 1, 1] 0 Conv2d-193 [-1, 22, 1, 1] 11,638 SiLU-194 [-1, 22, 1, 1] 0 Conv2d-195 [-1, 528, 1, 1] 12,144 Sigmoid-196 [-1, 528, 1, 1] 0 SqueezeExcitation-197 [-1, 528, 14, 14] 0 Conv2d-198 [-1, 120, 14, 14] 63,360 BatchNorm2d-199 [-1, 120, 14, 14] 240 MBConv-200 [-1, 120, 14, 14] 0 Conv2d-201 [-1, 720, 14, 14] 86,400 BatchNorm2d-202 [-1, 720, 14, 14] 1,440 SiLU-203 [-1, 720, 14, 14] 0 Conv2d-204 [-1, 720, 14, 14] 18,000 BatchNorm2d-205 [-1, 720, 14, 14] 1,440 SiLU-206 [-1, 720, 14, 14] 0 AdaptiveAvgPool2d-207 [-1, 720, 1, 1] 0 Conv2d-208 [-1, 30, 1, 1] 21,630 SiLU-209 [-1, 30, 1, 1] 0 Conv2d-210 [-1, 720, 1, 1] 22,320 Sigmoid-211 [-1, 720, 1, 1] 0 SqueezeExcitation-212 [-1, 720, 14, 14] 0 Conv2d-213 [-1, 120, 14, 14] 86,400 BatchNorm2d-214 [-1, 120, 14, 14] 240 StochasticDepth-215 [-1, 120, 14, 14] 0 MBConv-216 [-1, 120, 14, 14] 0 Conv2d-217 [-1, 720, 14, 14] 86,400 BatchNorm2d-218 [-1, 720, 14, 14] 1,440 SiLU-219 [-1, 720, 14, 14] 0 Conv2d-220 [-1, 720, 14, 14] 18,000 BatchNorm2d-221 [-1, 720, 14, 14] 1,440 SiLU-222 [-1, 720, 14, 14] 0 AdaptiveAvgPool2d-223 [-1, 720, 1, 1] 0 Conv2d-224 [-1, 30, 1, 1] 21,630 SiLU-225 [-1, 30, 1, 1] 0 Conv2d-226 [-1, 720, 1, 1] 22,320 Sigmoid-227 [-1, 720, 1, 1] 0 SqueezeExcitation-228 [-1, 720, 14, 14] 0 Conv2d-229 [-1, 120, 14, 14] 86,400 BatchNorm2d-230 [-1, 120, 14, 14] 240 StochasticDepth-231 [-1, 120, 14, 14] 0 MBConv-232 [-1, 120, 14, 14] 0 Conv2d-233 [-1, 720, 14, 14] 86,400 BatchNorm2d-234 [-1, 720, 14, 14] 1,440 SiLU-235 [-1, 720, 14, 14] 0 Conv2d-236 [-1, 720, 14, 14] 18,000 BatchNorm2d-237 [-1, 720, 14, 14] 1,440 SiLU-238 [-1, 720, 14, 14] 0 AdaptiveAvgPool2d-239 [-1, 720, 1, 1] 0 Conv2d-240 [-1, 30, 1, 1] 21,630 SiLU-241 [-1, 30, 1, 1] 0 Conv2d-242 [-1, 720, 1, 1] 22,320 Sigmoid-243 [-1, 720, 1, 1] 0 SqueezeExcitation-244 [-1, 720, 14, 14] 0 Conv2d-245 [-1, 120, 14, 14] 86,400 BatchNorm2d-246 [-1, 120, 14, 14] 240 StochasticDepth-247 [-1, 120, 14, 14] 0 MBConv-248 [-1, 120, 14, 14] 0 Conv2d-249 [-1, 720, 14, 14] 86,400 BatchNorm2d-250 [-1, 720, 14, 14] 1,440 SiLU-251 [-1, 720, 14, 14] 0 Conv2d-252 [-1, 720, 7, 7] 18,000 BatchNorm2d-253 [-1, 720, 7, 7] 1,440 SiLU-254 [-1, 720, 7, 7] 0 AdaptiveAvgPool2d-255 [-1, 720, 1, 1] 0 Conv2d-256 [-1, 30, 1, 1] 21,630 SiLU-257 [-1, 30, 1, 1] 0 Conv2d-258 [-1, 720, 1, 1] 22,320 Sigmoid-259 [-1, 720, 1, 1] 0 SqueezeExcitation-260 [-1, 720, 7, 7] 0 Conv2d-261 [-1, 208, 7, 7] 149,760 BatchNorm2d-262 [-1, 208, 7, 7] 416 MBConv-263 [-1, 208, 7, 7] 0 Conv2d-264 [-1, 1248, 7, 7] 259,584 BatchNorm2d-265 [-1, 1248, 7, 7] 2,496 SiLU-266 [-1, 1248, 7, 7] 0 Conv2d-267 [-1, 1248, 7, 7] 31,200 BatchNorm2d-268 [-1, 1248, 7, 7] 2,496 SiLU-269 [-1, 1248, 7, 7] 0 AdaptiveAvgPool2d-270 [-1, 1248, 1, 1] 0 Conv2d-271 [-1, 52, 1, 1] 64,948 SiLU-272 [-1, 52, 1, 1] 0 Conv2d-273 [-1, 1248, 1, 1] 66,144 Sigmoid-274 [-1, 1248, 1, 1] 0 SqueezeExcitation-275 [-1, 1248, 7, 7] 0 Conv2d-276 [-1, 208, 7, 7] 259,584 BatchNorm2d-277 [-1, 208, 7, 7] 416 StochasticDepth-278 [-1, 208, 7, 7] 0 MBConv-279 [-1, 208, 7, 7] 0 Conv2d-280 [-1, 1248, 7, 7] 259,584 BatchNorm2d-281 [-1, 1248, 7, 7] 2,496 SiLU-282 [-1, 1248, 7, 7] 0 Conv2d-283 [-1, 1248, 7, 7] 31,200 BatchNorm2d-284 [-1, 1248, 7, 7] 2,496 SiLU-285 [-1, 1248, 7, 7] 0 AdaptiveAvgPool2d-286 [-1, 1248, 1, 1] 0 Conv2d-287 [-1, 52, 1, 1] 64,948 SiLU-288 [-1, 52, 1, 1] 0 Conv2d-289 [-1, 1248, 1, 1] 66,144 Sigmoid-290 [-1, 1248, 1, 1] 0 SqueezeExcitation-291 [-1, 1248, 7, 7] 0 Conv2d-292 [-1, 208, 7, 7] 259,584 BatchNorm2d-293 [-1, 208, 7, 7] 416 StochasticDepth-294 [-1, 208, 7, 7] 0 MBConv-295 [-1, 208, 7, 7] 0 Conv2d-296 [-1, 1248, 7, 7] 259,584 BatchNorm2d-297 [-1, 1248, 7, 7] 2,496 SiLU-298 [-1, 1248, 7, 7] 0 Conv2d-299 [-1, 1248, 7, 7] 31,200 BatchNorm2d-300 [-1, 1248, 7, 7] 2,496 SiLU-301 [-1, 1248, 7, 7] 0 AdaptiveAvgPool2d-302 [-1, 1248, 1, 1] 0 Conv2d-303 [-1, 52, 1, 1] 64,948 SiLU-304 [-1, 52, 1, 1] 0 Conv2d-305 [-1, 1248, 1, 1] 66,144 Sigmoid-306 [-1, 1248, 1, 1] 0 SqueezeExcitation-307 [-1, 1248, 7, 7] 0 Conv2d-308 [-1, 208, 7, 7] 259,584 BatchNorm2d-309 [-1, 208, 7, 7] 416 StochasticDepth-310 [-1, 208, 7, 7] 0 MBConv-311 [-1, 208, 7, 7] 0 Conv2d-312 [-1, 1248, 7, 7] 259,584 BatchNorm2d-313 [-1, 1248, 7, 7] 2,496 SiLU-314 [-1, 1248, 7, 7] 0 Conv2d-315 [-1, 1248, 7, 7] 31,200 BatchNorm2d-316 [-1, 1248, 7, 7] 2,496 SiLU-317 [-1, 1248, 7, 7] 0 AdaptiveAvgPool2d-318 [-1, 1248, 1, 1] 0 Conv2d-319 [-1, 52, 1, 1] 64,948 SiLU-320 [-1, 52, 1, 1] 0 Conv2d-321 [-1, 1248, 1, 1] 66,144 Sigmoid-322 [-1, 1248, 1, 1] 0 SqueezeExcitation-323 [-1, 1248, 7, 7] 0 Conv2d-324 [-1, 208, 7, 7] 259,584 BatchNorm2d-325 [-1, 208, 7, 7] 416 StochasticDepth-326 [-1, 208, 7, 7] 0 MBConv-327 [-1, 208, 7, 7] 0 Conv2d-328 [-1, 1248, 7, 7] 259,584 BatchNorm2d-329 [-1, 1248, 7, 7] 2,496 SiLU-330 [-1, 1248, 7, 7] 0 Conv2d-331 [-1, 1248, 7, 7] 11,232 BatchNorm2d-332 [-1, 1248, 7, 7] 2,496 SiLU-333 [-1, 1248, 7, 7] 0 AdaptiveAvgPool2d-334 [-1, 1248, 1, 1] 0 Conv2d-335 [-1, 52, 1, 1] 64,948 SiLU-336 [-1, 52, 1, 1] 0 Conv2d-337 [-1, 1248, 1, 1] 66,144 Sigmoid-338 [-1, 1248, 1, 1] 0 SqueezeExcitation-339 [-1, 1248, 7, 7] 0 Conv2d-340 [-1, 352, 7, 7] 439,296 BatchNorm2d-341 [-1, 352, 7, 7] 704 MBConv-342 [-1, 352, 7, 7] 0 Conv2d-343 [-1, 2112, 7, 7] 743,424 BatchNorm2d-344 [-1, 2112, 7, 7] 4,224 SiLU-345 [-1, 2112, 7, 7] 0 Conv2d-346 [-1, 2112, 7, 7] 19,008 BatchNorm2d-347 [-1, 2112, 7, 7] 4,224 SiLU-348 [-1, 2112, 7, 7] 0 AdaptiveAvgPool2d-349 [-1, 2112, 1, 1] 0 Conv2d-350 [-1, 88, 1, 1] 185,944 SiLU-351 [-1, 88, 1, 1] 0 Conv2d-352 [-1, 2112, 1, 1] 187,968 Sigmoid-353 [-1, 2112, 1, 1] 0 SqueezeExcitation-354 [-1, 2112, 7, 7] 0 Conv2d-355 [-1, 352, 7, 7] 743,424 BatchNorm2d-356 [-1, 352, 7, 7] 704 StochasticDepth-357 [-1, 352, 7, 7] 0 MBConv-358 [-1, 352, 7, 7] 0 Conv2d-359 [-1, 1408, 7, 7] 495,616 BatchNorm2d-360 [-1, 1408, 7, 7] 2,816 SiLU-361 [-1, 1408, 7, 7] 0 AdaptiveAvgPool2d-362 [-1, 1408, 1, 1] 0 Dropout-363 [-1, 1408] 0 Linear-364 [-1, 10] 14,090 EfficientNet-365 [-1, 10] 0

Total params: 7,715,084 Trainable params: 14,090 Non-trainable params: 7,700,994

Input size (MB): 0.57 Forward/backward pass size (MB): 257.42 Params size (MB): 29.43 Estimated Total Size (MB): 287.43

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference