branch: master
dlc_info_2
97898 bytesRaw
DLC info for: /home/batman/xx/ml_tools/snpe/snpe-1.61.0.3358/mobilenetv2-7.dlc
Model Version: N/A
Model Copyright:N/A
-----------------------------------------------------------------------------------------------------------------------------------------
| Id | Name | Type | Inputs | Outputs | Out Dims | Runtimes | Parameters |
-----------------------------------------------------------------------------------------------------------------------------------------
| 0 | input | data | input | input | 1x224x224x3 | A D G C | input_preprocessing: passthrough |
| | | | | | | | input_type: image |
| 1 | Conv_0 | convolutional | input | 474 | 1x112x112x32 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 2 |
| | | | | | | | stride y: 2 |
| | | | | | | | num filters: 32 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | param count: 896 (0.0257%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 2 | Clip_1 | neuron | 474 | 317 | 1x112x112x32 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 3 | Conv_2 | convolutional | 317 | 477 | 1x112x112x32 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 32 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 32 |
| | | | | | | | param count: 320 (0.00917%) |
| | | | | | | | MACs per inference: 3M (1.2%) |
| 4 | Clip_3 | neuron | 477 | 320 | 1x112x112x32 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 5 | Conv_4 | convolutional | 320 | 480 | 1x112x112x16 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 16 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 528 (0.0151%) |
| | | | | | | | MACs per inference: 6M (2.13%) |
| 6 | Conv_5 | convolutional | 480 | 483 | 1x112x112x96 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 96 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 1k (0.0468%) |
| | | | | | | | MACs per inference: 19M (6.4%) |
| 7 | Clip_6 | neuron | 483 | 325 | 1x112x112x96 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 8 | Conv_7 | convolutional | 325 | 486 | 1x56x56x96 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 2 |
| | | | | | | | stride y: 2 |
| | | | | | | | num filters: 96 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 96 |
| | | | | | | | param count: 960 (0.0275%) |
| | | | | | | | MACs per inference: 2M (0.9%) |
| 9 | Clip_8 | neuron | 486 | 328 | 1x56x56x96 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 10 | Conv_9 | convolutional | 328 | 489 | 1x56x56x24 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 24 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 2k (0.0667%) |
| | | | | | | | MACs per inference: 7M (2.4%) |
| 11 | Conv_10 | convolutional | 489 | 492 | 1x56x56x144 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 144 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 3k (0.103%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 12 | Clip_11 | neuron | 492 | 333 | 1x56x56x144 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 13 | Conv_12 | convolutional | 333 | 495 | 1x56x56x144 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 144 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 144 |
| | | | | | | | param count: 1k (0.0413%) |
| | | | | | | | MACs per inference: 4M (1.35%) |
| 14 | Clip_13 | neuron | 495 | 336 | 1x56x56x144 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 15 | Conv_14 | convolutional | 336 | 498 | 1x56x56x24 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 24 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 3k (0.0998%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 16 | Add_15 | elementwise_binary_op | 489 | 339 | 1x56x56x24 | A D G C | operation: sum |
| | | | 498 | | | | MACs per inference: 75k (0.025%) |
| 17 | Conv_16 | convolutional | 339 | 501 | 1x56x56x144 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 144 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 3k (0.103%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 18 | Clip_17 | neuron | 501 | 342 | 1x56x56x144 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 19 | Conv_18 | convolutional | 342 | 504 | 1x28x28x144 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 2 |
| | | | | | | | stride y: 2 |
| | | | | | | | num filters: 144 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 144 |
| | | | | | | | param count: 1k (0.0413%) |
| | | | | | | | MACs per inference: 1M (0.338%) |
| 20 | Clip_19 | neuron | 504 | 345 | 1x28x28x144 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 21 | Conv_20 | convolutional | 345 | 507 | 1x28x28x32 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 32 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 4k (0.133%) |
| | | | | | | | MACs per inference: 3M (1.2%) |
| 22 | Conv_21 | convolutional | 507 | 510 | 1x28x28x192 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 192 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 6k (0.182%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 23 | Clip_22 | neuron | 510 | 350 | 1x28x28x192 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 24 | Conv_23 | convolutional | 350 | 513 | 1x28x28x192 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 192 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 192 |
| | | | | | | | param count: 1k (0.055%) |
| | | | | | | | MACs per inference: 1M (0.45%) |
| 25 | Clip_24 | neuron | 513 | 353 | 1x28x28x192 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 26 | Conv_25 | convolutional | 353 | 516 | 1x28x28x32 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 32 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 6k (0.177%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 27 | Add_26 | elementwise_binary_op | 507 | 356 | 1x28x28x32 | A D G C | operation: sum |
| | | | 516 | | | | MACs per inference: 25k (0.00833%) |
| 28 | Conv_27 | convolutional | 356 | 519 | 1x28x28x192 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 192 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 6k (0.182%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 29 | Clip_28 | neuron | 519 | 359 | 1x28x28x192 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 30 | Conv_29 | convolutional | 359 | 522 | 1x28x28x192 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 192 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 192 |
| | | | | | | | param count: 1k (0.055%) |
| | | | | | | | MACs per inference: 1M (0.45%) |
| 31 | Clip_30 | neuron | 522 | 362 | 1x28x28x192 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 32 | Conv_31 | convolutional | 362 | 525 | 1x28x28x32 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 32 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 6k (0.177%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 33 | Add_32 | elementwise_binary_op | 356 | 365 | 1x28x28x32 | A D G C | operation: sum |
| | | | 525 | | | | MACs per inference: 25k (0.00833%) |
| 34 | Conv_33 | convolutional | 365 | 528 | 1x28x28x192 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 192 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 6k (0.182%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 35 | Clip_34 | neuron | 528 | 368 | 1x28x28x192 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 36 | Conv_35 | convolutional | 368 | 531 | 1x14x14x192 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 2 |
| | | | | | | | stride y: 2 |
| | | | | | | | num filters: 192 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 192 |
| | | | | | | | param count: 1k (0.055%) |
| | | | | | | | MACs per inference: 338k (0.113%) |
| 37 | Clip_36 | neuron | 531 | 371 | 1x14x14x192 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 38 | Conv_37 | convolutional | 371 | 534 | 1x14x14x64 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 64 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 12k (0.354%) |
| | | | | | | | MACs per inference: 2M (0.8%) |
| 39 | Conv_38 | convolutional | 534 | 537 | 1x14x14x384 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.716%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 40 | Clip_39 | neuron | 537 | 376 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 41 | Conv_40 | convolutional | 376 | 540 | 1x14x14x384 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 384 |
| | | | | | | | param count: 3k (0.11%) |
| | | | | | | | MACs per inference: 677k (0.225%) |
| 42 | Clip_41 | neuron | 540 | 379 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 43 | Conv_42 | convolutional | 379 | 543 | 1x14x14x64 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 64 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.706%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 44 | Add_43 | elementwise_binary_op | 534 | 382 | 1x14x14x64 | A D G C | operation: sum |
| | | | 543 | | | | MACs per inference: 12k (0.00417%) |
| 45 | Conv_44 | convolutional | 382 | 546 | 1x14x14x384 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.716%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 46 | Clip_45 | neuron | 546 | 385 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 47 | Conv_46 | convolutional | 385 | 549 | 1x14x14x384 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 384 |
| | | | | | | | param count: 3k (0.11%) |
| | | | | | | | MACs per inference: 677k (0.225%) |
| 48 | Clip_47 | neuron | 549 | 388 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 49 | Conv_48 | convolutional | 388 | 552 | 1x14x14x64 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 64 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.706%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 50 | Add_49 | elementwise_binary_op | 382 | 391 | 1x14x14x64 | A D G C | operation: sum |
| | | | 552 | | | | MACs per inference: 12k (0.00417%) |
| 51 | Conv_50 | convolutional | 391 | 555 | 1x14x14x384 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.716%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 52 | Clip_51 | neuron | 555 | 394 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 53 | Conv_52 | convolutional | 394 | 558 | 1x14x14x384 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 384 |
| | | | | | | | param count: 3k (0.11%) |
| | | | | | | | MACs per inference: 677k (0.225%) |
| 54 | Clip_53 | neuron | 558 | 397 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 55 | Conv_54 | convolutional | 397 | 561 | 1x14x14x64 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 64 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.706%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 56 | Add_55 | elementwise_binary_op | 391 | 400 | 1x14x14x64 | A D G C | operation: sum |
| | | | 561 | | | | MACs per inference: 12k (0.00417%) |
| 57 | Conv_56 | convolutional | 400 | 564 | 1x14x14x384 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 24k (0.716%) |
| | | | | | | | MACs per inference: 4M (1.6%) |
| 58 | Clip_57 | neuron | 564 | 403 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 59 | Conv_58 | convolutional | 403 | 567 | 1x14x14x384 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 384 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 384 |
| | | | | | | | param count: 3k (0.11%) |
| | | | | | | | MACs per inference: 677k (0.225%) |
| 60 | Clip_59 | neuron | 567 | 406 | 1x14x14x384 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 61 | Conv_60 | convolutional | 406 | 570 | 1x14x14x96 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 96 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 36k (1.06%) |
| | | | | | | | MACs per inference: 7M (2.4%) |
| 62 | Conv_61 | convolutional | 570 | 573 | 1x14x14x576 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 576 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 55k (1.6%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 63 | Clip_62 | neuron | 573 | 411 | 1x14x14x576 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 64 | Conv_63 | convolutional | 411 | 576 | 1x14x14x576 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 576 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 576 |
| | | | | | | | param count: 5k (0.165%) |
| | | | | | | | MACs per inference: 1M (0.338%) |
| 65 | Clip_64 | neuron | 576 | 414 | 1x14x14x576 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 66 | Conv_65 | convolutional | 414 | 579 | 1x14x14x96 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 96 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 55k (1.59%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 67 | Add_66 | elementwise_binary_op | 570 | 417 | 1x14x14x96 | A D G C | operation: sum |
| | | | 579 | | | | MACs per inference: 18k (0.00625%) |
| 68 | Conv_67 | convolutional | 417 | 582 | 1x14x14x576 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 576 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 55k (1.6%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 69 | Clip_68 | neuron | 582 | 420 | 1x14x14x576 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 70 | Conv_69 | convolutional | 420 | 585 | 1x14x14x576 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 576 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 576 |
| | | | | | | | param count: 5k (0.165%) |
| | | | | | | | MACs per inference: 1M (0.338%) |
| 71 | Clip_70 | neuron | 585 | 423 | 1x14x14x576 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 72 | Conv_71 | convolutional | 423 | 588 | 1x14x14x96 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 96 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 55k (1.59%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 73 | Add_72 | elementwise_binary_op | 417 | 426 | 1x14x14x96 | A D G C | operation: sum |
| | | | 588 | | | | MACs per inference: 18k (0.00625%) |
| 74 | Conv_73 | convolutional | 426 | 591 | 1x14x14x576 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 576 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 55k (1.6%) |
| | | | | | | | MACs per inference: 10M (3.6%) |
| 75 | Clip_74 | neuron | 591 | 429 | 1x14x14x576 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 76 | Conv_75 | convolutional | 429 | 594 | 1x7x7x576 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 2 |
| | | | | | | | stride y: 2 |
| | | | | | | | num filters: 576 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 576 |
| | | | | | | | param count: 5k (0.165%) |
| | | | | | | | MACs per inference: 254k (0.0844%) |
| 77 | Clip_76 | neuron | 594 | 432 | 1x7x7x576 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 78 | Conv_77 | convolutional | 432 | 597 | 1x7x7x160 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 160 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 92k (2.65%) |
| | | | | | | | MACs per inference: 4M (1.5%) |
| 79 | Conv_78 | convolutional | 597 | 600 | 1x7x7x960 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 960 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 154k (4.43%) |
| | | | | | | | MACs per inference: 7M (2.5%) |
| 80 | Clip_79 | neuron | 600 | 437 | 1x7x7x960 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 81 | Conv_80 | convolutional | 437 | 603 | 1x7x7x960 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 960 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 960 |
| | | | | | | | param count: 9k (0.275%) |
| | | | | | | | MACs per inference: 423k (0.141%) |
| 82 | Clip_81 | neuron | 603 | 440 | 1x7x7x960 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 83 | Conv_82 | convolutional | 440 | 606 | 1x7x7x160 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 160 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 153k (4.41%) |
| | | | | | | | MACs per inference: 7M (2.5%) |
| 84 | Add_83 | elementwise_binary_op | 597 | 443 | 1x7x7x160 | A D G C | operation: sum |
| | | | 606 | | | | MACs per inference: 7k (0.0026%) |
| 85 | Conv_84 | convolutional | 443 | 609 | 1x7x7x960 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 960 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 154k (4.43%) |
| | | | | | | | MACs per inference: 7M (2.5%) |
| 86 | Clip_85 | neuron | 609 | 446 | 1x7x7x960 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 87 | Conv_86 | convolutional | 446 | 612 | 1x7x7x960 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 960 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 960 |
| | | | | | | | param count: 9k (0.275%) |
| | | | | | | | MACs per inference: 423k (0.141%) |
| 88 | Clip_87 | neuron | 612 | 449 | 1x7x7x960 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 89 | Conv_88 | convolutional | 449 | 615 | 1x7x7x160 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 160 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 153k (4.41%) |
| | | | | | | | MACs per inference: 7M (2.5%) |
| 90 | Add_89 | elementwise_binary_op | 443 | 452 | 1x7x7x160 | A D G C | operation: sum |
| | | | 615 | | | | MACs per inference: 7k (0.0026%) |
| 91 | Conv_90 | convolutional | 452 | 618 | 1x7x7x960 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 960 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 154k (4.43%) |
| | | | | | | | MACs per inference: 7M (2.5%) |
| 92 | Clip_91 | neuron | 618 | 455 | 1x7x7x960 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 93 | Conv_92 | convolutional | 455 | 621 | 1x7x7x960 | A D G C | padding x: 1 |
| | | | | | | | padding y: 1 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 960 |
| | | | | | | | kernel: 3x3 |
| | | | | | | | groups: 960 |
| | | | | | | | param count: 9k (0.275%) |
| | | | | | | | MACs per inference: 423k (0.141%) |
| 94 | Clip_93 | neuron | 621 | 458 | 1x7x7x960 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 95 | Conv_94 | convolutional | 458 | 624 | 1x7x7x320 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 320 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 307k (8.82%) |
| | | | | | | | MACs per inference: 15M (5%) |
| 96 | Conv_95 | convolutional | 624 | 627 | 1x7x7x1280 | A D G C | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | padding mode: zero |
| | | | | | | | stride x: 1 |
| | | | | | | | stride y: 1 |
| | | | | | | | num filters: 1280 |
| | | | | | | | kernel: 1x1 |
| | | | | | | | param count: 410k (11.8%) |
| | | | | | | | MACs per inference: 20M (6.67%) |
| 97 | Clip_96 | neuron | 627 | 463 | 1x7x7x1280 | A D G C | a: 0 |
| | | | | | | | b: 0 |
| | | | | | | | min_clamp: 0 |
| | | | | | | | max_clamp: 6 |
| | | | | | | | func: relu_min_max |
| 98 | GlobalAveragePool_97 | pooling | 463 | 464 | 1x1x1x1280 | A D G C | pool size x: 7 |
| | | | | | | | pool size y: 7 |
| | | | | | | | stride x: 7 |
| | | | | | | | stride y: 7 |
| | | | | | | | padding x: 0 |
| | | | | | | | padding y: 0 |
| | | | | | | | pool_type: POOL_AVG |
| | | | | | | | MACs per inference: 62k (0.0208%) |
| 99 | 464.ncs | permute | 464 | 464.ncs | 1x1280x1x1 | A D G C | permute_order: [0, 3, 1, 2] |
| 100 | Gemm_104 | fully_connected | 464.ncs | output | 1x1000 | A D G C | param count: 1M (36.7%) |
| | | | | | | | MACs per inference: 1M (0.425%) |
-----------------------------------------------------------------------------------------------------------------------------------------
Note: The supported runtimes column assumes a processor target of Snapdragon 835 (8998)
Key : A:AIP
D:DSP
G:GPU
C:CPU
Total parameters: 3487816 (13 MB assuming single precision float)
Total MACs per inference: 301M (100%)
Converter command: snpe-onnx-to-dlc adjust_nms_features_dims=False align_matmul_ranks=True copyright_file=None custom_op_config_paths=None debug=-1 disable_batchnorm_folding=False disable_chaining_eltwise_ops=False dry_run=None dumpIR=False dump_inferred_model=False dump_value_info=False enable_strict_validation=False extract_color_transform=False force_prune_cast_ops=True handle_gather_negative_indices=False inject_cast_for_gather=False input_dim=[['input', '1,3,224,224']] input_dtype=[] input_encoding=[] input_layout=[] input_type=[['input', 'image']] keep_disconnected_nodes=False keep_quant_nodes=False match_caffe_ssd_to_tf=False model_version=None no_simplification=False out_names=['output'] perform_axes_to_spatial_first_order=True prepare_inputs_as_params=True preprocess_lstm_ops=False preprocess_roi_pool_inputs=False quantization_overrides= squash_box_decoder=False unroll_lstm_time_steps=False use_convert_quantization_nodes=True validation_target=[]
Quantizer command: N/A
DLC created with converter version: 1.61.0.3358
Layers used by DLC: CONVOLUTIONAL, DATA, ELEMENTWISE_BINARY_OP_SUM, FULLY_CONNECTED, NEURON_RELU_MIN_MAX, PERMUTE, POOLING
Est. Steady-State Memory Needed to Run: 164.3 MiB
-----------------------------------------------------------------------------------------------------------------------------------------