From ba761289c157fc151c7f06aa37b812d8100561db Mon Sep 17 00:00:00 2001 From: Andy Brock Date: Tue, 16 Feb 2021 12:50:59 +0000 Subject: [PATCH] Upload links to pre-trained NFNet weights and add utility functions for loading them, as well as demo colab notebook. PiperOrigin-RevId: 357692801 --- nfnets/README.md | 54 +- nfnets/agc_optax.py | 2 +- nfnets/base.py | 2 +- nfnets/dataset.py | 2 +- nfnets/experiment.py | 2 +- nfnets/experiment_nf_regnets.py | 2 +- nfnets/experiment_nfnets.py | 2 +- nfnets/fixup_resnet.py | 2 +- nfnets/nf_regnet.py | 2 +- nfnets/nf_resnet.py | 2 +- nfnets/nfnet.py | 2 +- nfnets/nfnet_demo_colab.ipynb | 1295 +++++++++++++++++++++++++++++++ nfnets/optim.py | 2 +- nfnets/requirements.txt | 1 + nfnets/resnet.py | 2 +- nfnets/run.sh | 2 +- nfnets/skipinit_resnet.py | 2 +- nfnets/test.py | 2 +- nfnets/utils.py | 23 +- 19 files changed, 1378 insertions(+), 25 deletions(-) create mode 100644 nfnets/nfnet_demo_colab.ipynb diff --git a/nfnets/README.md b/nfnets/README.md index d316760e..926af948 100644 --- a/nfnets/README.md +++ b/nfnets/README.md @@ -1,21 +1,57 @@ # Code for Normalizer-Free Networks This repository contains code for the ICLR 2021 paper -"Characterizing signal propagation to close the performance gap in unnormalized -ResNets," by Andrew Brock, Soham De, and Samuel L. Smith, and the arXiv preprint -"High-Performance Large-Scale Image Recognition Without Normalization" by +["Characterizing signal propagation to close the performance gap in unnormalized +ResNets,"](https://arxiv.org/abs/2102.06171) by Andrew Brock, Soham De, and +Samuel L. Smith, and the arXiv preprint ["High-Performance Large-Scale Image +Recognition Without Normalization"](http://dpmd.ai/06171) by Andrew Brock, Soham De, Samuel L. Smith, and Karen Simonyan. ## Running this code -Install using pip install -r requirements.txt and use one of the experiment.py -files in combination with [JAXline](https://github.com/deepmind/jaxline) to -train models. Optionally copy test.py into -a dir one level up and run it to ensure you can take a single experiment step -with fake data. +Using `run.sh` will create and activate a virtualenv, install all necessary +dependencies and run a test program to ensure that you can import all the +modules and take a single experiment step. To train with this code, use this +virtualenv and use one of the experiment.py files in combination with +[JAXline](https://github.com/deepmind/jaxline). The provided +demo Colab can be run online, or by starting a jupyter notebook within this +virtualenv. Note that you will need a local copy of ImageNet compatible with the TFDS format used in dataset.py in order to train on ImageNet. + +## Pre-Trained Weights + +We provide pre-trained weights for NFNet-F0 through F5 (trained without SAM), +and for NFNet-F6 trained with SAM. All models are pre-trained on ImageNet for +360 epochs at batch size 4096, and are provided as numpy files containing +parameter trees compatible with haiku. In utils.py we provide a +`load_haiku_file` function which loads these parameter trees, and +`flatten_haiku_tree` to convert these to flat dictionaries +which may prove easier to port to other frameworks. Note that we do not provide +model `states`, as these models, lacking batchnorm, do not have running stats. +Note also that the conv layer weights are in the format HWIO, so for frameworks +like PyTorch which use OIHW you'll need to swap the axes appropriately to the +layout you use. + + +| Model | #FLOPs | #Params | Top-1 | Top-5 | TPUv3 Train | GPU Train | link | +|---|:---:|:---:|:---:|:---:|:---:|:---:|:---:| +F0 | 12.38B | 71.5M | 83.6 | 96.8 | 73.3ms | 56.7ms | [haiku](https://storage.googleapis.com/dm-nfnets/F0_haiku.npz) +F1 | 35.54B | 132.6M | 84.7 | 97.1 | 158.5ms | 133.9ms | [haiku](https://storage.googleapis.com/dm-nfnets/F1_haiku.npz) +F2 | 62.59B | 193.8M | 85.1 | 97.3 | 295.8ms | 226.3ms | [haiku](https://storage.googleapis.com/dm-nfnets/F2_haiku.npz) +F3 | 114.76B | 254.9M | 85.7 | 97.5 | 532.2ms | 524.5ms | [haiku](https://storage.googleapis.com/dm-nfnets/F3_haiku.npz) +F4 | 215.24B | 316.1M | 85.9 | 97.6 | 1033.3ms | 1190.6ms | [haiku](https://storage.googleapis.com/dm-nfnets/F4_haiku.npz) +F5 | 289.76B | 377.2M | 86.0 | 97.6 | 1398.5ms | 2177.1ms | [haiku](https://storage.googleapis.com/dm-nfnets/F5_haiku.npz) +F6+SAM | 377.28B | 438.4M | 86.5 | 97.9 | 2774.1ms | - | [haiku](https://storage.googleapis.com/dm-nfnets/F6_haiku.npz) + + +## Demo Colab [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/deepmind/deepmind-research/blob/master/nfnets/nfnet_demo_colab.ipynb) + +We also include a Colab notebook with a demo showing how to run an NFNet to +classify an image. + + ## Giving Credit If you use this code in your work, we ask you to please cite one or both of the @@ -39,7 +75,7 @@ The reference for Adaptive Gradient Clipping (AGC) and the NFNets models: @article{brock2021high, author={Andrew Brock and Soham De and Samuel L. Smith and Karen Simonyan}, title={High-Performance Large-Scale Image Recognition Without Normalization}, - journal={arXiv preprint arXiv:}, + journal={arXiv preprint arXiv:2102.06171}, year={2021} } ``` diff --git a/nfnets/agc_optax.py b/nfnets/agc_optax.py index 532f4d1e..2b435c02 100644 --- a/nfnets/agc_optax.py +++ b/nfnets/agc_optax.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/base.py b/nfnets/base.py index 1baaa8de..65a62984 100644 --- a/nfnets/base.py +++ b/nfnets/base.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/dataset.py b/nfnets/dataset.py index e6ca3e34..ea244bd6 100644 --- a/nfnets/dataset.py +++ b/nfnets/dataset.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/experiment.py b/nfnets/experiment.py index 3fd36c6b..724e5eab 100644 --- a/nfnets/experiment.py +++ b/nfnets/experiment.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/experiment_nf_regnets.py b/nfnets/experiment_nf_regnets.py index c1b1463e..5be057e6 100644 --- a/nfnets/experiment_nf_regnets.py +++ b/nfnets/experiment_nf_regnets.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/experiment_nfnets.py b/nfnets/experiment_nfnets.py index 1dad04c4..8de77590 100644 --- a/nfnets/experiment_nfnets.py +++ b/nfnets/experiment_nfnets.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/fixup_resnet.py b/nfnets/fixup_resnet.py index baca19b3..291933ca 100644 --- a/nfnets/fixup_resnet.py +++ b/nfnets/fixup_resnet.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/nf_regnet.py b/nfnets/nf_regnet.py index 8672a5b6..b93358d7 100644 --- a/nfnets/nf_regnet.py +++ b/nfnets/nf_regnet.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/nf_resnet.py b/nfnets/nf_resnet.py index 2861d088..c179b1ec 100644 --- a/nfnets/nf_resnet.py +++ b/nfnets/nf_resnet.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/nfnet.py b/nfnets/nfnet.py index 30ade039..f51d22bd 100644 --- a/nfnets/nfnet.py +++ b/nfnets/nfnet.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/nfnet_demo_colab.ipynb b/nfnets/nfnet_demo_colab.ipynb new file mode 100644 index 00000000..09a0fe33 --- /dev/null +++ b/nfnets/nfnet_demo_colab.ipynb @@ -0,0 +1,1295 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "id": "jev9nsjEePyx" + }, + "source": [ + "# Download and run an NFNet-F0 model pre-trained on ImageNet\n", + "This demo shows how to run a pre-trained NFNet classifier, from the paper \n", + "[High-Performance Large-Scale Image Recognition Without Normalization](http://dpmd.ai/06171) (Brock, De, Smith, Simonyan, 2021). It uses code from [the official JAX + Haiku implementation](http://dpmd.ai/nfnets).\n", + "\n", + "\n", + "It's recommended to use `Runtime-\u003eChange Runtime Type` to pick a GPU for speed." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "executionInfo": { + "elapsed": 16619, + "status": "ok", + "timestamp": 1613472332810, + "user": { + "displayName": "Andy Brock", + "photoUrl": "", + "userId": "04378600802759613630" + }, + "user_tz": 0 + }, + "id": "bvEKqIQAZjxo", + "outputId": "846ba4c5-08cb-4d14-de38-8426de8d564e" + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Collecting dm-haiku\n", + "\u001b[?25l Downloading https://files.pythonhosted.org/packages/61/dc/6da61c7f96bffd7ebc2888510e5ec82ee260cbc1d7a3f4f0a862914646f8/dm_haiku-0.0.3-py3-none-any.whl (260kB)\n", + "\r\u001b[K |█▎ | 10kB 18.5MB/s eta 0:00:01\r\u001b[K |██▌ | 20kB 22.7MB/s eta 0:00:01\r\u001b[K |███▊ | 30kB 22.3MB/s eta 0:00:01\r\u001b[K |█████ | 40kB 18.3MB/s eta 0:00:01\r\u001b[K |██████▎ | 51kB 15.6MB/s eta 0:00:01\r\u001b[K |███████▌ | 61kB 12.2MB/s eta 0:00:01\r\u001b[K |████████▉ | 71kB 12.9MB/s eta 0:00:01\r\u001b[K |██████████ | 81kB 12.7MB/s eta 0:00:01\r\u001b[K |███████████▎ | 92kB 12.9MB/s eta 0:00:01\r\u001b[K |████████████▋ | 102kB 12.7MB/s eta 0:00:01\r\u001b[K |█████████████▉ | 112kB 12.7MB/s eta 0:00:01\r\u001b[K |███████████████ | 122kB 12.7MB/s eta 0:00:01\r\u001b[K |████████████████▍ | 133kB 12.7MB/s eta 0:00:01\r\u001b[K |█████████████████▋ | 143kB 12.7MB/s eta 0:00:01\r\u001b[K |██████████████████▉ | 153kB 12.7MB/s eta 0:00:01\r\u001b[K |████████████████████▏ | 163kB 12.7MB/s eta 0:00:01\r\u001b[K |█████████████████████▍ | 174kB 12.7MB/s eta 0:00:01\r\u001b[K |██████████████████████▋ | 184kB 12.7MB/s eta 0:00:01\r\u001b[K |████████████████████████ | 194kB 12.7MB/s eta 0:00:01\r\u001b[K |█████████████████████████▏ | 204kB 12.7MB/s eta 0:00:01\r\u001b[K |██████████████████████████▍ | 215kB 12.7MB/s eta 0:00:01\r\u001b[K |███████████████████████████▊ | 225kB 12.7MB/s eta 0:00:01\r\u001b[K |█████████████████████████████ | 235kB 12.7MB/s eta 0:00:01\r\u001b[K |██████████████████████████████▏ | 245kB 12.7MB/s eta 0:00:01\r\u001b[K |███████████████████████████████▌| 256kB 12.7MB/s eta 0:00:01\r\u001b[K |████████████████████████████████| 266kB 12.7MB/s \n", + "\u001b[?25hRequirement already satisfied: absl-py\u003e=0.7.1 in /usr/local/lib/python3.6/dist-packages (from dm-haiku) (0.10.0)\n", + "Requirement already satisfied: numpy\u003e=1.18.0 in /usr/local/lib/python3.6/dist-packages (from dm-haiku) (1.19.5)\n", + "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from absl-py\u003e=0.7.1-\u003edm-haiku) (1.15.0)\n", + "Installing collected packages: dm-haiku\n", + "Successfully installed dm-haiku-0.0.3\n", + "Requirement already satisfied: dill in /usr/local/lib/python3.6/dist-packages (0.3.3)\n", + "Cloning into 'deepmind-research'...\n", + "remote: Enumerating objects: 10, done.\u001b[K\n", + "remote: Counting objects: 100% (10/10), done.\u001b[K\n", + "remote: Compressing objects: 100% (10/10), done.\u001b[K\n", + "remote: Total 1307 (delta 0), reused 4 (delta 0), pack-reused 1297\u001b[K\n", + "Receiving objects: 100% (1307/1307), 74.41 MiB | 20.20 MiB/s, done.\n", + "Resolving deltas: 100% (656/656), done.\n" + ] + } + ], + "source": [ + "# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved.\n", + "#\n", + "# Licensed under the Apache License, Version 2.0 (the \"License\");\n", + "# you may not use this file except in compliance with the License.\n", + "# You may obtain a copy of the License at\n", + "#\n", + "# http://www.apache.org/licenses/LICENSE-2.0\n", + "#\n", + "# Unless required by applicable law or agreed to in writing, software\n", + "# distributed under the License is distributed on an \"AS IS\" BASIS,\n", + "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n", + "# See the License for the specific language governing permissions and\n", + "# limitations under the License.\n", + "# ==============================================================================\n", + "!pip install dm-haiku\n", + "!pip install dill\n", + "!git clone https://github.com/deepmind/deepmind-research/\n", + "import dill\n", + "import haiku as hk\n", + "import jax\n", + "import jax.numpy as jnp\n", + "import os\n", + "import numpy as np\n", + "from PIL import Image\n", + "os.chdir('deepmind-research')\n", + "from nfnets import nfnet, base" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "d3oIj5O1dyVB" + }, + "source": [ + "# ImageNet Class List" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "A9pD88rU5wsw" + }, + "outputs": [], + "source": [ + "# Get ImageNet class list\n", + "imagenet_classlist = {0: 'tench, Tinca tinca',\n", + " 1: 'goldfish, Carassius auratus',\n", + " 2: 'great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias',\n", + " 3: 'tiger shark, Galeocerdo cuvieri',\n", + " 4: 'hammerhead, hammerhead shark',\n", + " 5: 'electric ray, crampfish, numbfish, torpedo',\n", + " 6: 'stingray',\n", + " 7: 'rooster',\n", + " 8: 'hen',\n", + " 9: 'ostrich, Struthio camelus',\n", + " 10: 'brambling, Fringilla montifringilla',\n", + " 11: 'goldfinch, Carduelis carduelis',\n", + " 12: 'house finch, linnet, Carpodacus mexicanus',\n", + " 13: 'junco, snowbird',\n", + " 14: 'indigo bunting, indigo finch, indigo bird, Passerina cyanea',\n", + " 15: 'robin, American robin, Turdus migratorius',\n", + " 16: 'bulbul',\n", + " 17: 'jay',\n", + " 18: 'magpie',\n", + " 19: 'chickadee',\n", + " 20: 'water ouzel, dipper',\n", + " 21: 'kite',\n", + " 22: 'bald eagle, American eagle, Haliaeetus leucocephalus',\n", + " 23: 'vulture',\n", + " 24: 'great grey owl, great gray owl, Strix nebulosa',\n", + " 25: 'European fire salamander, Salamandra salamandra',\n", + " 26: 'common newt, Triturus vulgaris',\n", + " 27: 'eft',\n", + " 28: 'spotted salamander, Ambystoma maculatum',\n", + " 29: 'axolotl, mud puppy, Ambystoma mexicanum',\n", + " 30: 'bullfrog, Rana catesbeiana',\n", + " 31: 'tree frog, tree-frog',\n", + " 32: 'tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui',\n", + " 33: 'loggerhead, loggerhead turtle, Caretta caretta',\n", + " 34: 'leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea',\n", + " 35: 'mud turtle',\n", + " 36: 'terrapin',\n", + " 37: 'box turtle, box tortoise',\n", + " 38: 'banded gecko',\n", + " 39: 'common iguana, iguana, Iguana iguana',\n", + " 40: 'American chameleon, anole, Anolis carolinensis',\n", + " 41: 'whiptail, whiptail lizard',\n", + " 42: 'agama',\n", + " 43: 'frilled lizard, Chlamydosaurus kingi',\n", + " 44: 'alligator lizard',\n", + " 45: 'Gila monster, Heloderma suspectum',\n", + " 46: 'green lizard, Lacerta viridis',\n", + " 47: 'African chameleon, Chamaeleo chamaeleon',\n", + " 48: 'Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis',\n", + " 49: 'African crocodile, Nile crocodile, Crocodylus niloticus',\n", + " 50: 'American alligator, Alligator mississipiensis',\n", + " 51: 'triceratops',\n", + " 52: 'thunder snake, worm snake, Carphophis amoenus',\n", + " 53: 'ringneck snake, ring-necked snake, ring snake',\n", + " 54: 'hognose snake, puff adder, sand viper',\n", + " 55: 'green snake, grass snake',\n", + " 56: 'king snake, kingsnake',\n", + " 57: 'garter snake, grass snake',\n", + " 58: 'water snake',\n", + " 59: 'vine snake',\n", + " 60: 'night snake, Hypsiglena torquata',\n", + " 61: 'boa constrictor, Constrictor constrictor',\n", + " 62: 'rock python, rock snake, Python sebae',\n", + " 63: 'Indian cobra, Naja naja',\n", + " 64: 'green mamba',\n", + " 65: 'sea snake',\n", + " 66: 'horned viper, cerastes, sand viper, horned asp, Cerastes cornutus',\n", + " 67: 'diamondback, diamondback rattlesnake, Crotalus adamanteus',\n", + " 68: 'sidewinder, horned rattlesnake, Crotalus cerastes',\n", + " 69: 'trilobite',\n", + " 70: 'harvestman, daddy longlegs, Phalangium opilio',\n", + " 71: 'scorpion',\n", + " 72: 'black and gold garden spider, Argiope aurantia',\n", + " 73: 'barn spider, Araneus cavaticus',\n", + " 74: 'garden spider, Aranea diademata',\n", + " 75: 'black widow, Latrodectus mactans',\n", + " 76: 'tarantula',\n", + " 77: 'wolf spider, hunting spider',\n", + " 78: 'tick',\n", + " 79: 'centipede',\n", + " 80: 'black grouse',\n", + " 81: 'ptarmigan',\n", + " 82: 'ruffed grouse, partridge, Bonasa umbellus',\n", + " 83: 'prairie chicken, prairie grouse, prairie fowl',\n", + " 84: 'peacock',\n", + " 85: 'quail',\n", + " 86: 'partridge',\n", + " 87: 'African grey, African gray, Psittacus erithacus',\n", + " 88: 'macaw',\n", + " 89: 'sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita',\n", + " 90: 'lorikeet',\n", + " 91: 'coucal',\n", + " 92: 'bee eater',\n", + " 93: 'hornbill',\n", + " 94: 'hummingbird',\n", + " 95: 'jacamar',\n", + " 96: 'toucan',\n", + " 97: 'drake',\n", + " 98: 'red-breasted merganser, Mergus serrator',\n", + " 99: 'goose',\n", + " 100: 'black swan, Cygnus atratus',\n", + " 101: 'tusker',\n", + " 102: 'echidna, spiny anteater, anteater',\n", + " 103: 'platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus',\n", + " 104: 'wallaby, brush kangaroo',\n", + " 105: 'koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus',\n", + " 106: 'wombat',\n", + " 107: 'jellyfish',\n", + " 108: 'sea anemone, anemone',\n", + " 109: 'brain coral',\n", + " 110: 'flatworm, platyhelminth',\n", + " 111: 'nematode, nematode worm, roundworm',\n", + " 112: 'conch',\n", + " 113: 'snail',\n", + " 114: 'slug',\n", + " 115: 'sea slug, nudibranch',\n", + " 116: 'chiton, coat-of-mail shell, sea cradle, polyplacophore',\n", + " 117: 'chambered nautilus, pearly nautilus, nautilus',\n", + " 118: 'Dungeness crab, Cancer magister',\n", + " 119: 'rock crab, Cancer irroratus',\n", + " 120: 'fiddler crab',\n", + " 121: 'king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica',\n", + " 122: 'American lobster, Northern lobster, Maine lobster, Homarus americanus',\n", + " 123: 'spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish',\n", + " 124: 'crayfish, crawfish, crawdad, crawdaddy',\n", + " 125: 'hermit crab',\n", + " 126: 'isopod',\n", + " 127: 'white stork, Ciconia ciconia',\n", + " 128: 'black stork, Ciconia nigra',\n", + " 129: 'spoonbill',\n", + " 130: 'flamingo',\n", + " 131: 'little blue heron, Egretta caerulea',\n", + " 132: 'American egret, great white heron, Egretta albus',\n", + " 133: 'bittern',\n", + " 134: 'crane',\n", + " 135: 'limpkin, Aramus pictus',\n", + " 136: 'European gallinule, Porphyrio porphyrio',\n", + " 137: 'American coot, marsh hen, mud hen, water hen, Fulica americana',\n", + " 138: 'bustard',\n", + " 139: 'ruddy turnstone, Arenaria interpres',\n", + " 140: 'red-backed sandpiper, dunlin, Erolia alpina',\n", + " 141: 'redshank, Tringa totanus',\n", + " 142: 'dowitcher',\n", + " 143: 'oystercatcher, oyster catcher',\n", + " 144: 'pelican',\n", + " 145: 'king penguin, Aptenodytes patagonica',\n", + " 146: 'albatross, mollymawk',\n", + " 147: 'grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus',\n", + " 148: 'killer whale, killer, orca, grampus, sea wolf, Orcinus orca',\n", + " 149: 'dugong, Dugong dugon',\n", + " 150: 'sea lion',\n", + " 151: 'Chihuahua',\n", + " 152: 'Japanese spaniel',\n", + " 153: 'Maltese dog, Maltese terrier, Maltese',\n", + " 154: 'Pekinese, Pekingese, Peke',\n", + " 155: 'Shih-Tzu',\n", + " 156: 'Blenheim spaniel',\n", + " 157: 'papillon',\n", + " 158: 'toy terrier',\n", + " 159: 'Rhodesian ridgeback',\n", + " 160: 'Afghan hound, Afghan',\n", + " 161: 'basset, basset hound',\n", + " 162: 'beagle',\n", + " 163: 'bloodhound, sleuthhound',\n", + " 164: 'bluetick',\n", + " 165: 'black-and-tan coonhound',\n", + " 166: 'Walker hound, Walker foxhound',\n", + " 167: 'English foxhound',\n", + " 168: 'redbone',\n", + " 169: 'borzoi, Russian wolfhound',\n", + " 170: 'Irish wolfhound',\n", + " 171: 'Italian greyhound',\n", + " 172: 'whippet',\n", + " 173: 'Ibizan hound, Ibizan Podenco',\n", + " 174: 'Norwegian elkhound, elkhound',\n", + " 175: 'otterhound, otter hound',\n", + " 176: 'Saluki, gazelle hound',\n", + " 177: 'Scottish deerhound, deerhound',\n", + " 178: 'Weimaraner',\n", + " 179: 'Staffordshire bullterrier, Staffordshire bull terrier',\n", + " 180: 'American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier',\n", + " 181: 'Bedlington terrier',\n", + " 182: 'Border terrier',\n", + " 183: 'Kerry blue terrier',\n", + " 184: 'Irish terrier',\n", + " 185: 'Norfolk terrier',\n", + " 186: 'Norwich terrier',\n", + " 187: 'Yorkshire terrier',\n", + " 188: 'wire-haired fox terrier',\n", + " 189: 'Lakeland terrier',\n", + " 190: 'Sealyham terrier, Sealyham',\n", + " 191: 'Airedale, Airedale terrier',\n", + " 192: 'cairn, cairn terrier',\n", + " 193: 'Australian terrier',\n", + " 194: 'Dandie Dinmont, Dandie Dinmont terrier',\n", + " 195: 'Boston bull, Boston terrier',\n", + " 196: 'miniature schnauzer',\n", + " 197: 'giant schnauzer',\n", + " 198: 'standard schnauzer',\n", + " 199: 'Scotch terrier, Scottish terrier, Scottie',\n", + " 200: 'Tibetan terrier, chrysanthemum dog',\n", + " 201: 'silky terrier, Sydney silky',\n", + " 202: 'soft-coated wheaten terrier',\n", + " 203: 'West Highland white terrier',\n", + " 204: 'Lhasa, Lhasa apso',\n", + " 205: 'flat-coated retriever',\n", + " 206: 'curly-coated retriever',\n", + " 207: 'golden retriever',\n", + " 208: 'Labrador retriever',\n", + " 209: 'Chesapeake Bay retriever',\n", + " 210: 'German short-haired pointer',\n", + " 211: 'vizsla, Hungarian pointer',\n", + " 212: 'English setter',\n", + " 213: 'Irish setter, red setter',\n", + " 214: 'Gordon setter',\n", + " 215: 'Brittany spaniel',\n", + " 216: 'clumber, clumber spaniel',\n", + " 217: 'English springer, English springer spaniel',\n", + " 218: 'Welsh springer spaniel',\n", + " 219: 'cocker spaniel, English cocker spaniel, cocker',\n", + " 220: 'Sussex spaniel',\n", + " 221: 'Irish water spaniel',\n", + " 222: 'kuvasz',\n", + " 223: 'schipperke',\n", + " 224: 'groenendael',\n", + " 225: 'malinois',\n", + " 226: 'briard',\n", + " 227: 'kelpie',\n", + " 228: 'komondor',\n", + " 229: 'Old English sheepdog, bobtail',\n", + " 230: 'Shetland sheepdog, Shetland sheep dog, Shetland',\n", + " 231: 'collie',\n", + " 232: 'Border collie',\n", + " 233: 'Bouvier des Flandres, Bouviers des Flandres',\n", + " 234: 'Rottweiler',\n", + " 235: 'German shepherd, German shepherd dog, German police dog, alsatian',\n", + " 236: 'Doberman, Doberman pinscher',\n", + " 237: 'miniature pinscher',\n", + " 238: 'Greater Swiss Mountain dog',\n", + " 239: 'Bernese mountain dog',\n", + " 240: 'Appenzeller',\n", + " 241: 'EntleBucher',\n", + " 242: 'boxer',\n", + " 243: 'bull mastiff',\n", + " 244: 'Tibetan mastiff',\n", + " 245: 'French bulldog',\n", + " 246: 'Great Dane',\n", + " 247: 'Saint Bernard, St Bernard',\n", + " 248: 'Inuit dog, husky',\n", + " 249: 'malamute, malemute, Alaskan malamute',\n", + " 250: 'Siberian husky',\n", + " 251: 'dalmatian, coach dog, carriage dog',\n", + " 252: 'affenpinscher, monkey pinscher, monkey dog',\n", + " 253: 'basenji',\n", + " 254: 'pug, pug-dog',\n", + " 255: 'Leonberg',\n", + " 256: 'Newfoundland, Newfoundland dog',\n", + " 257: 'Great Pyrenees',\n", + " 258: 'Samoyed, Samoyede',\n", + " 259: 'Pomeranian',\n", + " 260: 'chow, chow chow',\n", + " 261: 'keeshond',\n", + " 262: 'Brabancon griffon',\n", + " 263: 'Pembroke, Pembroke Welsh corgi',\n", + " 264: 'Cardigan, Cardigan Welsh corgi',\n", + " 265: 'toy poodle',\n", + " 266: 'miniature poodle',\n", + " 267: 'standard poodle',\n", + " 268: 'Mexican hairless',\n", + " 269: 'timber wolf, grey wolf, gray wolf, Canis lupus',\n", + " 270: 'white wolf, Arctic wolf, Canis lupus tundrarum',\n", + " 271: 'red wolf, maned wolf, Canis rufus, Canis niger',\n", + " 272: 'coyote, prairie wolf, brush wolf, Canis latrans',\n", + " 273: 'dingo, warrigal, warragal, Canis dingo',\n", + " 274: 'dhole, Cuon alpinus',\n", + " 275: 'African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus',\n", + " 276: 'hyena, hyaena',\n", + " 277: 'red fox, Vulpes vulpes',\n", + " 278: 'kit fox, Vulpes macrotis',\n", + " 279: 'Arctic fox, white fox, Alopex lagopus',\n", + " 280: 'grey fox, gray fox, Urocyon cinereoargenteus',\n", + " 281: 'tabby, tabby cat',\n", + " 282: 'tiger cat',\n", + " 283: 'Persian cat',\n", + " 284: 'Siamese cat, Siamese',\n", + " 285: 'Egyptian cat',\n", + " 286: 'cougar, puma, catamount, mountain lion, painter, panther, Felis concolor',\n", + " 287: 'lynx, catamount',\n", + " 288: 'leopard, Panthera pardus',\n", + " 289: 'snow leopard, ounce, Panthera uncia',\n", + " 290: 'jaguar, panther, Panthera onca, Felis onca',\n", + " 291: 'lion, king of beasts, Panthera leo',\n", + " 292: 'tiger, Panthera tigris',\n", + " 293: 'cheetah, chetah, Acinonyx jubatus',\n", + " 294: 'brown bear, bruin, Ursus arctos',\n", + " 295: 'American black bear, black bear, Ursus americanus, Euarctos americanus',\n", + " 296: 'ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus',\n", + " 297: 'sloth bear, Melursus ursinus, Ursus ursinus',\n", + " 298: 'mongoose',\n", + " 299: 'meerkat, mierkat',\n", + " 300: 'tiger beetle',\n", + " 301: 'ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle',\n", + " 302: 'ground beetle, carabid beetle',\n", + " 303: 'long-horned beetle, longicorn, longicorn beetle',\n", + " 304: 'leaf beetle, chrysomelid',\n", + " 305: 'dung beetle',\n", + " 306: 'rhinoceros beetle',\n", + " 307: 'weevil',\n", + " 308: 'fly',\n", + " 309: 'bee',\n", + " 310: 'ant, emmet, pismire',\n", + " 311: 'grasshopper, hopper',\n", + " 312: 'cricket',\n", + " 313: 'walking stick, walkingstick, stick insect',\n", + " 314: 'cockroach, roach',\n", + " 315: 'mantis, mantid',\n", + " 316: 'cicada, cicala',\n", + " 317: 'leafhopper',\n", + " 318: 'lacewing, lacewing fly',\n", + " 319: \"dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk\",\n", + " 320: 'damselfly',\n", + " 321: 'admiral',\n", + " 322: 'ringlet, ringlet butterfly',\n", + " 323: 'monarch, monarch butterfly, milkweed butterfly, Danaus plexippus',\n", + " 324: 'cabbage butterfly',\n", + " 325: 'sulphur butterfly, sulfur butterfly',\n", + " 326: 'lycaenid, lycaenid butterfly',\n", + " 327: 'starfish, sea star',\n", + " 328: 'sea urchin',\n", + " 329: 'sea cucumber, holothurian',\n", + " 330: 'wood rabbit, cottontail, cottontail rabbit',\n", + " 331: 'hare',\n", + " 332: 'Angora, Angora rabbit',\n", + " 333: 'hamster',\n", + " 334: 'porcupine, hedgehog',\n", + " 335: 'fox squirrel, eastern fox squirrel, Sciurus niger',\n", + " 336: 'marmot',\n", + " 337: 'beaver',\n", + " 338: 'guinea pig, Cavia cobaya',\n", + " 339: 'sorrel',\n", + " 340: 'zebra',\n", + " 341: 'hog, pig, grunter, squealer, Sus scrofa',\n", + " 342: 'wild boar, boar, Sus scrofa',\n", + " 343: 'warthog',\n", + " 344: 'hippopotamus, hippo, river horse, Hippopotamus amphibius',\n", + " 345: 'ox',\n", + " 346: 'water buffalo, water ox, Asiatic buffalo, Bubalus bubalis',\n", + " 347: 'bison',\n", + " 348: 'ram, tup',\n", + " 349: 'bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis',\n", + " 350: 'ibex, Capra ibex',\n", + " 351: 'hartebeest',\n", + " 352: 'impala, Aepyceros melampus',\n", + " 353: 'gazelle',\n", + " 354: 'Arabian camel, dromedary, Camelus dromedarius',\n", + " 355: 'llama',\n", + " 356: 'weasel',\n", + " 357: 'mink',\n", + " 358: 'polecat, fitch, foulmart, foumart, Mustela putorius',\n", + " 359: 'black-footed ferret, ferret, Mustela nigripes',\n", + " 360: 'otter',\n", + " 361: 'skunk, polecat, wood pussy',\n", + " 362: 'badger',\n", + " 363: 'armadillo',\n", + " 364: 'three-toed sloth, ai, Bradypus tridactylus',\n", + " 365: 'orangutan, orang, orangutang, Pongo pygmaeus',\n", + " 366: 'gorilla, Gorilla gorilla',\n", + " 367: 'chimpanzee, chimp, Pan troglodytes',\n", + " 368: 'gibbon, Hylobates lar',\n", + " 369: 'siamang, Hylobates syndactylus, Symphalangus syndactylus',\n", + " 370: 'guenon, guenon monkey',\n", + " 371: 'patas, hussar monkey, Erythrocebus patas',\n", + " 372: 'baboon',\n", + " 373: 'macaque',\n", + " 374: 'langur',\n", + " 375: 'colobus, colobus monkey',\n", + " 376: 'proboscis monkey, Nasalis larvatus',\n", + " 377: 'marmoset',\n", + " 378: 'capuchin, ringtail, Cebus capucinus',\n", + " 379: 'howler monkey, howler',\n", + " 380: 'titi, titi monkey',\n", + " 381: 'spider monkey, Ateles geoffroyi',\n", + " 382: 'squirrel monkey, Saimiri sciureus',\n", + " 383: 'Madagascar cat, ring-tailed lemur, Lemur catta',\n", + " 384: 'indri, indris, Indri indri, Indri brevicaudatus',\n", + " 385: 'Indian elephant, Elephas maximus',\n", + " 386: 'African elephant, Loxodonta africana',\n", + " 387: 'lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens',\n", + " 388: 'giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca',\n", + " 389: 'barracouta, snoek',\n", + " 390: 'eel',\n", + " 391: 'coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch',\n", + " 392: 'rock beauty, Holocanthus tricolor',\n", + " 393: 'anemone fish',\n", + " 394: 'sturgeon',\n", + " 395: 'gar, garfish, garpike, billfish, Lepisosteus osseus',\n", + " 396: 'lionfish',\n", + " 397: 'puffer, pufferfish, blowfish, globefish',\n", + " 398: 'abacus',\n", + " 399: 'abaya',\n", + " 400: \"academic gown, academic robe, judge's robe\",\n", + " 401: 'accordion, piano accordion, squeeze box',\n", + " 402: 'acoustic guitar',\n", + " 403: 'aircraft carrier, carrier, flattop, attack aircraft carrier',\n", + " 404: 'airliner',\n", + " 405: 'airship, dirigible',\n", + " 406: 'altar',\n", + " 407: 'ambulance',\n", + " 408: 'amphibian, amphibious vehicle',\n", + " 409: 'analog clock',\n", + " 410: 'apiary, bee house',\n", + " 411: 'apron',\n", + " 412: 'ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin',\n", + " 413: 'assault rifle, assault gun',\n", + " 414: 'backpack, back pack, knapsack, packsack, rucksack, haversack',\n", + " 415: 'bakery, bakeshop, bakehouse',\n", + " 416: 'balance beam, beam',\n", + " 417: 'balloon',\n", + " 418: 'ballpoint, ballpoint pen, ballpen, Biro',\n", + " 419: 'Band Aid',\n", + " 420: 'banjo',\n", + " 421: 'bannister, banister, balustrade, balusters, handrail',\n", + " 422: 'barbell',\n", + " 423: 'barber chair',\n", + " 424: 'barbershop',\n", + " 425: 'barn',\n", + " 426: 'barometer',\n", + " 427: 'barrel, cask',\n", + " 428: 'barrow, garden cart, lawn cart, wheelbarrow',\n", + " 429: 'baseball',\n", + " 430: 'basketball',\n", + " 431: 'bassinet',\n", + " 432: 'bassoon',\n", + " 433: 'bathing cap, swimming cap',\n", + " 434: 'bath towel',\n", + " 435: 'bathtub, bathing tub, bath, tub',\n", + " 436: 'beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon',\n", + " 437: 'beacon, lighthouse, beacon light, pharos',\n", + " 438: 'beaker',\n", + " 439: 'bearskin, busby, shako',\n", + " 440: 'beer bottle',\n", + " 441: 'beer glass',\n", + " 442: 'bell cote, bell cot',\n", + " 443: 'bib',\n", + " 444: 'bicycle-built-for-two, tandem bicycle, tandem',\n", + " 445: 'bikini, two-piece',\n", + " 446: 'binder, ring-binder',\n", + " 447: 'binoculars, field glasses, opera glasses',\n", + " 448: 'birdhouse',\n", + " 449: 'boathouse',\n", + " 450: 'bobsled, bobsleigh, bob',\n", + " 451: 'bolo tie, bolo, bola tie, bola',\n", + " 452: 'bonnet, poke bonnet',\n", + " 453: 'bookcase',\n", + " 454: 'bookshop, bookstore, bookstall',\n", + " 455: 'bottlecap',\n", + " 456: 'bow',\n", + " 457: 'bow tie, bow-tie, bowtie',\n", + " 458: 'brass, memorial tablet, plaque',\n", + " 459: 'brassiere, bra, bandeau',\n", + " 460: 'breakwater, groin, groyne, mole, bulwark, seawall, jetty',\n", + " 461: 'breastplate, aegis, egis',\n", + " 462: 'broom',\n", + " 463: 'bucket, pail',\n", + " 464: 'buckle',\n", + " 465: 'bulletproof vest',\n", + " 466: 'bullet train, bullet',\n", + " 467: 'butcher shop, meat market',\n", + " 468: 'cab, hack, taxi, taxicab',\n", + " 469: 'caldron, cauldron',\n", + " 470: 'candle, taper, wax light',\n", + " 471: 'cannon',\n", + " 472: 'canoe',\n", + " 473: 'can opener, tin opener',\n", + " 474: 'cardigan',\n", + " 475: 'car mirror',\n", + " 476: 'carousel, carrousel, merry-go-round, roundabout, whirligig',\n", + " 477: \"carpenter's kit, tool kit\",\n", + " 478: 'carton',\n", + " 479: 'car wheel',\n", + " 480: 'cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM',\n", + " 481: 'cassette',\n", + " 482: 'cassette player',\n", + " 483: 'castle',\n", + " 484: 'catamaran',\n", + " 485: 'CD player',\n", + " 486: 'cello, violoncello',\n", + " 487: 'cellular telephone, cellular phone, cellphone, cell, mobile phone',\n", + " 488: 'chain',\n", + " 489: 'chainlink fence',\n", + " 490: 'chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour',\n", + " 491: 'chain saw, chainsaw',\n", + " 492: 'chest',\n", + " 493: 'chiffonier, commode',\n", + " 494: 'chime, bell, gong',\n", + " 495: 'china cabinet, china closet',\n", + " 496: 'Christmas stocking',\n", + " 497: 'church, church building',\n", + " 498: 'cinema, movie theater, movie theatre, movie house, picture palace',\n", + " 499: 'cleaver, meat cleaver, chopper',\n", + " 500: 'cliff dwelling',\n", + " 501: 'cloak',\n", + " 502: 'clog, geta, patten, sabot',\n", + " 503: 'cocktail shaker',\n", + " 504: 'coffee mug',\n", + " 505: 'coffeepot',\n", + " 506: 'coil, spiral, volute, whorl, helix',\n", + " 507: 'combination lock',\n", + " 508: 'computer keyboard, keypad',\n", + " 509: 'confectionery, confectionary, candy store',\n", + " 510: 'container ship, containership, container vessel',\n", + " 511: 'convertible',\n", + " 512: 'corkscrew, bottle screw',\n", + " 513: 'cornet, horn, trumpet, trump',\n", + " 514: 'cowboy boot',\n", + " 515: 'cowboy hat, ten-gallon hat',\n", + " 516: 'cradle',\n", + " 517: 'crane',\n", + " 518: 'crash helmet',\n", + " 519: 'crate',\n", + " 520: 'crib, cot',\n", + " 521: 'Crock Pot',\n", + " 522: 'croquet ball',\n", + " 523: 'crutch',\n", + " 524: 'cuirass',\n", + " 525: 'dam, dike, dyke',\n", + " 526: 'desk',\n", + " 527: 'desktop computer',\n", + " 528: 'dial telephone, dial phone',\n", + " 529: 'diaper, nappy, napkin',\n", + " 530: 'digital clock',\n", + " 531: 'digital watch',\n", + " 532: 'dining table, board',\n", + " 533: 'dishrag, dishcloth',\n", + " 534: 'dishwasher, dish washer, dishwashing machine',\n", + " 535: 'disk brake, disc brake',\n", + " 536: 'dock, dockage, docking facility',\n", + " 537: 'dogsled, dog sled, dog sleigh',\n", + " 538: 'dome',\n", + " 539: 'doormat, welcome mat',\n", + " 540: 'drilling platform, offshore rig',\n", + " 541: 'drum, membranophone, tympan',\n", + " 542: 'drumstick',\n", + " 543: 'dumbbell',\n", + " 544: 'Dutch oven',\n", + " 545: 'electric fan, blower',\n", + " 546: 'electric guitar',\n", + " 547: 'electric locomotive',\n", + " 548: 'entertainment center',\n", + " 549: 'envelope',\n", + " 550: 'espresso maker',\n", + " 551: 'face powder',\n", + " 552: 'feather boa, boa',\n", + " 553: 'file, file cabinet, filing cabinet',\n", + " 554: 'fireboat',\n", + " 555: 'fire engine, fire truck',\n", + " 556: 'fire screen, fireguard',\n", + " 557: 'flagpole, flagstaff',\n", + " 558: 'flute, transverse flute',\n", + " 559: 'folding chair',\n", + " 560: 'football helmet',\n", + " 561: 'forklift',\n", + " 562: 'fountain',\n", + " 563: 'fountain pen',\n", + " 564: 'four-poster',\n", + " 565: 'freight car',\n", + " 566: 'French horn, horn',\n", + " 567: 'frying pan, frypan, skillet',\n", + " 568: 'fur coat',\n", + " 569: 'garbage truck, dustcart',\n", + " 570: 'gasmask, respirator, gas helmet',\n", + " 571: 'gas pump, gasoline pump, petrol pump, island dispenser',\n", + " 572: 'goblet',\n", + " 573: 'go-kart',\n", + " 574: 'golf ball',\n", + " 575: 'golfcart, golf cart',\n", + " 576: 'gondola',\n", + " 577: 'gong, tam-tam',\n", + " 578: 'gown',\n", + " 579: 'grand piano, grand',\n", + " 580: 'greenhouse, nursery, glasshouse',\n", + " 581: 'grille, radiator grille',\n", + " 582: 'grocery store, grocery, food market, market',\n", + " 583: 'guillotine',\n", + " 584: 'hair slide',\n", + " 585: 'hair spray',\n", + " 586: 'half track',\n", + " 587: 'hammer',\n", + " 588: 'hamper',\n", + " 589: 'hand blower, blow dryer, blow drier, hair dryer, hair drier',\n", + " 590: 'hand-held computer, hand-held microcomputer',\n", + " 591: 'handkerchief, hankie, hanky, hankey',\n", + " 592: 'hard disc, hard disk, fixed disk',\n", + " 593: 'harmonica, mouth organ, harp, mouth harp',\n", + " 594: 'harp',\n", + " 595: 'harvester, reaper',\n", + " 596: 'hatchet',\n", + " 597: 'holster',\n", + " 598: 'home theater, home theatre',\n", + " 599: 'honeycomb',\n", + " 600: 'hook, claw',\n", + " 601: 'hoopskirt, crinoline',\n", + " 602: 'horizontal bar, high bar',\n", + " 603: 'horse cart, horse-cart',\n", + " 604: 'hourglass',\n", + " 605: 'iPod',\n", + " 606: 'iron, smoothing iron',\n", + " 607: \"jack-o'-lantern\",\n", + " 608: 'jean, blue jean, denim',\n", + " 609: 'jeep, landrover',\n", + " 610: 'jersey, T-shirt, tee shirt',\n", + " 611: 'jigsaw puzzle',\n", + " 612: 'jinrikisha, ricksha, rickshaw',\n", + " 613: 'joystick',\n", + " 614: 'kimono',\n", + " 615: 'knee pad',\n", + " 616: 'knot',\n", + " 617: 'lab coat, laboratory coat',\n", + " 618: 'ladle',\n", + " 619: 'lampshade, lamp shade',\n", + " 620: 'laptop, laptop computer',\n", + " 621: 'lawn mower, mower',\n", + " 622: 'lens cap, lens cover',\n", + " 623: 'letter opener, paper knife, paperknife',\n", + " 624: 'library',\n", + " 625: 'lifeboat',\n", + " 626: 'lighter, light, igniter, ignitor',\n", + " 627: 'limousine, limo',\n", + " 628: 'liner, ocean liner',\n", + " 629: 'lipstick, lip rouge',\n", + " 630: 'Loafer',\n", + " 631: 'lotion',\n", + " 632: 'loudspeaker, speaker, speaker unit, loudspeaker system, speaker system',\n", + " 633: \"loupe, jeweler's loupe\",\n", + " 634: 'lumbermill, sawmill',\n", + " 635: 'magnetic compass',\n", + " 636: 'mailbag, postbag',\n", + " 637: 'mailbox, letter box',\n", + " 638: 'maillot',\n", + " 639: 'maillot, tank suit',\n", + " 640: 'manhole cover',\n", + " 641: 'maraca',\n", + " 642: 'marimba, xylophone',\n", + " 643: 'mask',\n", + " 644: 'matchstick',\n", + " 645: 'maypole',\n", + " 646: 'maze, labyrinth',\n", + " 647: 'measuring cup',\n", + " 648: 'medicine chest, medicine cabinet',\n", + " 649: 'megalith, megalithic structure',\n", + " 650: 'microphone, mike',\n", + " 651: 'microwave, microwave oven',\n", + " 652: 'military uniform',\n", + " 653: 'milk can',\n", + " 654: 'minibus',\n", + " 655: 'miniskirt, mini',\n", + " 656: 'minivan',\n", + " 657: 'missile',\n", + " 658: 'mitten',\n", + " 659: 'mixing bowl',\n", + " 660: 'mobile home, manufactured home',\n", + " 661: 'Model T',\n", + " 662: 'modem',\n", + " 663: 'monastery',\n", + " 664: 'monitor',\n", + " 665: 'moped',\n", + " 666: 'mortar',\n", + " 667: 'mortarboard',\n", + " 668: 'mosque',\n", + " 669: 'mosquito net',\n", + " 670: 'motor scooter, scooter',\n", + " 671: 'mountain bike, all-terrain bike, off-roader',\n", + " 672: 'mountain tent',\n", + " 673: 'mouse, computer mouse',\n", + " 674: 'mousetrap',\n", + " 675: 'moving van',\n", + " 676: 'muzzle',\n", + " 677: 'nail',\n", + " 678: 'neck brace',\n", + " 679: 'necklace',\n", + " 680: 'nipple',\n", + " 681: 'notebook, notebook computer',\n", + " 682: 'obelisk',\n", + " 683: 'oboe, hautboy, hautbois',\n", + " 684: 'ocarina, sweet potato',\n", + " 685: 'odometer, hodometer, mileometer, milometer',\n", + " 686: 'oil filter',\n", + " 687: 'organ, pipe organ',\n", + " 688: 'oscilloscope, scope, cathode-ray oscilloscope, CRO',\n", + " 689: 'overskirt',\n", + " 690: 'oxcart',\n", + " 691: 'oxygen mask',\n", + " 692: 'packet',\n", + " 693: 'paddle, boat paddle',\n", + " 694: 'paddlewheel, paddle wheel',\n", + " 695: 'padlock',\n", + " 696: 'paintbrush',\n", + " 697: \"pajama, pyjama, pj's, jammies\",\n", + " 698: 'palace',\n", + " 699: 'panpipe, pandean pipe, syrinx',\n", + " 700: 'paper towel',\n", + " 701: 'parachute, chute',\n", + " 702: 'parallel bars, bars',\n", + " 703: 'park bench',\n", + " 704: 'parking meter',\n", + " 705: 'passenger car, coach, carriage',\n", + " 706: 'patio, terrace',\n", + " 707: 'pay-phone, pay-station',\n", + " 708: 'pedestal, plinth, footstall',\n", + " 709: 'pencil box, pencil case',\n", + " 710: 'pencil sharpener',\n", + " 711: 'perfume, essence',\n", + " 712: 'Petri dish',\n", + " 713: 'photocopier',\n", + " 714: 'pick, plectrum, plectron',\n", + " 715: 'pickelhaube',\n", + " 716: 'picket fence, paling',\n", + " 717: 'pickup, pickup truck',\n", + " 718: 'pier',\n", + " 719: 'piggy bank, penny bank',\n", + " 720: 'pill bottle',\n", + " 721: 'pillow',\n", + " 722: 'ping-pong ball',\n", + " 723: 'pinwheel',\n", + " 724: 'pirate, pirate ship',\n", + " 725: 'pitcher, ewer',\n", + " 726: \"plane, carpenter's plane, woodworking plane\",\n", + " 727: 'planetarium',\n", + " 728: 'plastic bag',\n", + " 729: 'plate rack',\n", + " 730: 'plow, plough',\n", + " 731: \"plunger, plumber's helper\",\n", + " 732: 'Polaroid camera, Polaroid Land camera',\n", + " 733: 'pole',\n", + " 734: 'police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria',\n", + " 735: 'poncho',\n", + " 736: 'pool table, billiard table, snooker table',\n", + " 737: 'pop bottle, soda bottle',\n", + " 738: 'pot, flowerpot',\n", + " 739: \"potter's wheel\",\n", + " 740: 'power drill',\n", + " 741: 'prayer rug, prayer mat',\n", + " 742: 'printer',\n", + " 743: 'prison, prison house',\n", + " 744: 'projectile, missile',\n", + " 745: 'projector',\n", + " 746: 'puck, hockey puck',\n", + " 747: 'punching bag, punch bag, punching ball, punchball',\n", + " 748: 'purse',\n", + " 749: 'quill, quill pen',\n", + " 750: 'quilt, comforter, comfort, puff',\n", + " 751: 'racer, race car, racing car',\n", + " 752: 'racket, racquet',\n", + " 753: 'radiator',\n", + " 754: 'radio, wireless',\n", + " 755: 'radio telescope, radio reflector',\n", + " 756: 'rain barrel',\n", + " 757: 'recreational vehicle, RV, R.V.',\n", + " 758: 'reel',\n", + " 759: 'reflex camera',\n", + " 760: 'refrigerator, icebox',\n", + " 761: 'remote control, remote',\n", + " 762: 'restaurant, eating house, eating place, eatery',\n", + " 763: 'revolver, six-gun, six-shooter',\n", + " 764: 'rifle',\n", + " 765: 'rocking chair, rocker',\n", + " 766: 'rotisserie',\n", + " 767: 'rubber eraser, rubber, pencil eraser',\n", + " 768: 'rugby ball',\n", + " 769: 'rule, ruler',\n", + " 770: 'running shoe',\n", + " 771: 'safe',\n", + " 772: 'safety pin',\n", + " 773: 'saltshaker, salt shaker',\n", + " 774: 'sandal',\n", + " 775: 'sarong',\n", + " 776: 'sax, saxophone',\n", + " 777: 'scabbard',\n", + " 778: 'scale, weighing machine',\n", + " 779: 'school bus',\n", + " 780: 'schooner',\n", + " 781: 'scoreboard',\n", + " 782: 'screen, CRT screen',\n", + " 783: 'screw',\n", + " 784: 'screwdriver',\n", + " 785: 'seat belt, seatbelt',\n", + " 786: 'sewing machine',\n", + " 787: 'shield, buckler',\n", + " 788: 'shoe shop, shoe-shop, shoe store',\n", + " 789: 'shoji',\n", + " 790: 'shopping basket',\n", + " 791: 'shopping cart',\n", + " 792: 'shovel',\n", + " 793: 'shower cap',\n", + " 794: 'shower curtain',\n", + " 795: 'ski',\n", + " 796: 'ski mask',\n", + " 797: 'sleeping bag',\n", + " 798: 'slide rule, slipstick',\n", + " 799: 'sliding door',\n", + " 800: 'slot, one-armed bandit',\n", + " 801: 'snorkel',\n", + " 802: 'snowmobile',\n", + " 803: 'snowplow, snowplough',\n", + " 804: 'soap dispenser',\n", + " 805: 'soccer ball',\n", + " 806: 'sock',\n", + " 807: 'solar dish, solar collector, solar furnace',\n", + " 808: 'sombrero',\n", + " 809: 'soup bowl',\n", + " 810: 'space bar',\n", + " 811: 'space heater',\n", + " 812: 'space shuttle',\n", + " 813: 'spatula',\n", + " 814: 'speedboat',\n", + " 815: \"spider web, spider's web\",\n", + " 816: 'spindle',\n", + " 817: 'sports car, sport car',\n", + " 818: 'spotlight, spot',\n", + " 819: 'stage',\n", + " 820: 'steam locomotive',\n", + " 821: 'steel arch bridge',\n", + " 822: 'steel drum',\n", + " 823: 'stethoscope',\n", + " 824: 'stole',\n", + " 825: 'stone wall',\n", + " 826: 'stopwatch, stop watch',\n", + " 827: 'stove',\n", + " 828: 'strainer',\n", + " 829: 'streetcar, tram, tramcar, trolley, trolley car',\n", + " 830: 'stretcher',\n", + " 831: 'studio couch, day bed',\n", + " 832: 'stupa, tope',\n", + " 833: 'submarine, pigboat, sub, U-boat',\n", + " 834: 'suit, suit of clothes',\n", + " 835: 'sundial',\n", + " 836: 'sunglass',\n", + " 837: 'sunglasses, dark glasses, shades',\n", + " 838: 'sunscreen, sunblock, sun blocker',\n", + " 839: 'suspension bridge',\n", + " 840: 'swab, swob, mop',\n", + " 841: 'sweatshirt',\n", + " 842: 'swimming trunks, bathing trunks',\n", + " 843: 'swing',\n", + " 844: 'switch, electric switch, electrical switch',\n", + " 845: 'syringe',\n", + " 846: 'table lamp',\n", + " 847: 'tank, army tank, armored combat vehicle, armoured combat vehicle',\n", + " 848: 'tape player',\n", + " 849: 'teapot',\n", + " 850: 'teddy, teddy bear',\n", + " 851: 'television, television system',\n", + " 852: 'tennis ball',\n", + " 853: 'thatch, thatched roof',\n", + " 854: 'theater curtain, theatre curtain',\n", + " 855: 'thimble',\n", + " 856: 'thresher, thrasher, threshing machine',\n", + " 857: 'throne',\n", + " 858: 'tile roof',\n", + " 859: 'toaster',\n", + " 860: 'tobacco shop, tobacconist shop, tobacconist',\n", + " 861: 'toilet seat',\n", + " 862: 'torch',\n", + " 863: 'totem pole',\n", + " 864: 'tow truck, tow car, wrecker',\n", + " 865: 'toyshop',\n", + " 866: 'tractor',\n", + " 867: 'trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi',\n", + " 868: 'tray',\n", + " 869: 'trench coat',\n", + " 870: 'tricycle, trike, velocipede',\n", + " 871: 'trimaran',\n", + " 872: 'tripod',\n", + " 873: 'triumphal arch',\n", + " 874: 'trolleybus, trolley coach, trackless trolley',\n", + " 875: 'trombone',\n", + " 876: 'tub, vat',\n", + " 877: 'turnstile',\n", + " 878: 'typewriter keyboard',\n", + " 879: 'umbrella',\n", + " 880: 'unicycle, monocycle',\n", + " 881: 'upright, upright piano',\n", + " 882: 'vacuum, vacuum cleaner',\n", + " 883: 'vase',\n", + " 884: 'vault',\n", + " 885: 'velvet',\n", + " 886: 'vending machine',\n", + " 887: 'vestment',\n", + " 888: 'viaduct',\n", + " 889: 'violin, fiddle',\n", + " 890: 'volleyball',\n", + " 891: 'waffle iron',\n", + " 892: 'wall clock',\n", + " 893: 'wallet, billfold, notecase, pocketbook',\n", + " 894: 'wardrobe, closet, press',\n", + " 895: 'warplane, military plane',\n", + " 896: 'washbasin, handbasin, washbowl, lavabo, wash-hand basin',\n", + " 897: 'washer, automatic washer, washing machine',\n", + " 898: 'water bottle',\n", + " 899: 'water jug',\n", + " 900: 'water tower',\n", + " 901: 'whiskey jug',\n", + " 902: 'whistle',\n", + " 903: 'wig',\n", + " 904: 'window screen',\n", + " 905: 'window shade',\n", + " 906: 'Windsor tie',\n", + " 907: 'wine bottle',\n", + " 908: 'wing',\n", + " 909: 'wok',\n", + " 910: 'wooden spoon',\n", + " 911: 'wool, woolen, woollen',\n", + " 912: 'worm fence, snake fence, snake-rail fence, Virginia fence',\n", + " 913: 'wreck',\n", + " 914: 'yawl',\n", + " 915: 'yurt',\n", + " 916: 'web site, website, internet site, site',\n", + " 917: 'comic book',\n", + " 918: 'crossword puzzle, crossword',\n", + " 919: 'street sign',\n", + " 920: 'traffic light, traffic signal, stoplight',\n", + " 921: 'book jacket, dust cover, dust jacket, dust wrapper',\n", + " 922: 'menu',\n", + " 923: 'plate',\n", + " 924: 'guacamole',\n", + " 925: 'consomme',\n", + " 926: 'hot pot, hotpot',\n", + " 927: 'trifle',\n", + " 928: 'ice cream, icecream',\n", + " 929: 'ice lolly, lolly, lollipop, popsicle',\n", + " 930: 'French loaf',\n", + " 931: 'bagel, beigel',\n", + " 932: 'pretzel',\n", + " 933: 'cheeseburger',\n", + " 934: 'hotdog, hot dog, red hot',\n", + " 935: 'mashed potato',\n", + " 936: 'head cabbage',\n", + " 937: 'broccoli',\n", + " 938: 'cauliflower',\n", + " 939: 'zucchini, courgette',\n", + " 940: 'spaghetti squash',\n", + " 941: 'acorn squash',\n", + " 942: 'butternut squash',\n", + " 943: 'cucumber, cuke',\n", + " 944: 'artichoke, globe artichoke',\n", + " 945: 'bell pepper',\n", + " 946: 'cardoon',\n", + " 947: 'mushroom',\n", + " 948: 'Granny Smith',\n", + " 949: 'strawberry',\n", + " 950: 'orange',\n", + " 951: 'lemon',\n", + " 952: 'fig',\n", + " 953: 'pineapple, ananas',\n", + " 954: 'banana',\n", + " 955: 'jackfruit, jak, jack',\n", + " 956: 'custard apple',\n", + " 957: 'pomegranate',\n", + " 958: 'hay',\n", + " 959: 'carbonara',\n", + " 960: 'chocolate sauce, chocolate syrup',\n", + " 961: 'dough',\n", + " 962: 'meat loaf, meatloaf',\n", + " 963: 'pizza, pizza pie',\n", + " 964: 'potpie',\n", + " 965: 'burrito',\n", + " 966: 'red wine',\n", + " 967: 'espresso',\n", + " 968: 'cup',\n", + " 969: 'eggnog',\n", + " 970: 'alp',\n", + " 971: 'bubble',\n", + " 972: 'cliff, drop, drop-off',\n", + " 973: 'coral reef',\n", + " 974: 'geyser',\n", + " 975: 'lakeside, lakeshore',\n", + " 976: 'promontory, headland, head, foreland',\n", + " 977: 'sandbar, sand bar',\n", + " 978: 'seashore, coast, seacoast, sea-coast',\n", + " 979: 'valley, vale',\n", + " 980: 'volcano',\n", + " 981: 'ballplayer, baseball player',\n", + " 982: 'groom, bridegroom',\n", + " 983: 'scuba diver',\n", + " 984: 'rapeseed',\n", + " 985: 'daisy',\n", + " 986: \"yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum\",\n", + " 987: 'corn',\n", + " 988: 'acorn',\n", + " 989: 'hip, rose hip, rosehip',\n", + " 990: 'buckeye, horse chestnut, conker',\n", + " 991: 'coral fungus',\n", + " 992: 'agaric',\n", + " 993: 'gyromitra',\n", + " 994: 'stinkhorn, carrion fungus',\n", + " 995: 'earthstar',\n", + " 996: 'hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa',\n", + " 997: 'bolete',\n", + " 998: 'ear, spike, capitulum',\n", + " 999: 'toilet tissue, toilet paper, bathroom tissue'}" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "wqHZIbjyZXoY" + }, + "source": [ + "# Load F0 weights and grab an image\n", + "The chosen image is released to the public domain and was taken in 2019 so we can be sure it isn't a part of the ImageNet training set." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "executionInfo": { + "elapsed": 25019, + "status": "ok", + "timestamp": 1613472346442, + "user": { + "displayName": "Andy Brock", + "photoUrl": "", + "userId": "04378600802759613630" + }, + "user_tz": 0 + }, + "id": "KmA2FCPS6VkA", + "outputId": "1cfd1ff2-9adc-47c7-f3e8-af87f8bc2f26" + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "--2021-02-16 10:45:40-- https://storage.googleapis.com/dm-nfnets/F0_haiku.npz\n", + "Resolving storage.googleapis.com (storage.googleapis.com)... 108.177.127.128, 173.194.69.128, 173.194.79.128, ...\n", + "Connecting to storage.googleapis.com (storage.googleapis.com)|108.177.127.128|:443... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 285976842 (273M) [application/octet-stream]\n", + "Saving to: ‘F0_haiku.npz’\n", + "\n", + "F0_haiku.npz 100%[===================\u003e] 272.73M 64.7MB/s in 4.2s \n", + "\n", + "2021-02-16 10:45:45 (64.7 MB/s) - ‘F0_haiku.npz’ saved [285976842/285976842]\n", + "\n", + "Model loaded w/ 71.49M Params\n", + "--2021-02-16 10:45:45-- https://live.staticflickr.com/65535/50594927526_f6c3b2a5d4_b.jpg\n", + "Resolving live.staticflickr.com (live.staticflickr.com)... 13.35.250.20, 2600:9000:2057:7200:0:5a51:64c9:c681, 2600:9000:2057:4e00:0:5a51:64c9:c681, ...\n", + "Connecting to live.staticflickr.com (live.staticflickr.com)|13.35.250.20|:443... connected.\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: unspecified [image/jpeg]\n", + "Saving to: ‘peppers.jpg’\n", + "\n", + "peppers.jpg [ \u003c=\u003e ] 139.76K --.-KB/s in 0.02s \n", + "\n", + "2021-02-16 10:45:45 (7.95 MB/s) - ‘peppers.jpg’ saved [143110]\n", + "\n" + ] + } + ], + "source": [ + "# Load F0 weights\n", + "variant = 'F0'\n", + "os.environ['VARIANT'] = variant\n", + "!wget https://storage.googleapis.com/dm-nfnets/${VARIANT}_haiku.npz\n", + "with open(f'{variant}_haiku.npz', 'rb') as in_file:\n", + " params = dill.load(in_file)\n", + "print(f'Model loaded w/ {hk.data_structures.tree_size(params)/1e6:.2f}M Params')\n", + "# public domain image from https://www.flickr.com/photos/alabama_extension/50594927526\n", + "!wget https://live.staticflickr.com/65535/50594927526_f6c3b2a5d4_b.jpg -O peppers.jpg\n", + "im = Image.open('peppers.jpg')\n", + "# Resize and crop to variant test size\n", + "imsize = base.nfnet_params[variant]['test_imsize']\n", + "im = im.resize((imsize + 32, imsize + 32))\n", + "im = im.crop((16, 16, 16+imsize, 16+imsize))\n", + "# Convert im to tensor and normalize with channel-wise RGB\n", + "MEAN_RGB = (0.485 * 255, 0.456 * 255, 0.406 * 255)\n", + "STDDEV_RGB = (0.229 * 255, 0.224 * 255, 0.225 * 255)\n", + "x = (np.float32(im) - MEAN_RGB) / STDDEV_RGB\n" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "8pxUnQbC4MPj" + }, + "source": [ + "# Run NFNet-F0 to classify a single image.\n", + "You can either run it in eager mode by not jitting the forward function, which will be relatively slow but not incur any compilation cost, or JIT the forward\n", + "function, which will incur a compilation cost but yield faster inference. If you JIT (the default) then the first time you call the function will trigger the compilation, and all subsequent evaluations will be fast." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "Gr8tIBEwYOtb" + }, + "outputs": [], + "source": [ + "# Prepare the forward fn\n", + "def forward(inputs, is_training): \n", + " model = nfnet.NFNet(num_classes=1000, variant=variant)\n", + " return model(inputs, is_training=is_training)['logits']\n", + "net = hk.without_apply_rng(hk.transform(forward))\n", + "fwd = jax.jit(lambda inputs: net.apply(params, inputs, is_training=False))\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 290 + }, + "executionInfo": { + "elapsed": 27901, + "status": "ok", + "timestamp": 1613472375251, + "user": { + "displayName": "Andy Brock", + "photoUrl": "", + "userId": "04378600802759613630" + }, + "user_tz": 0 + }, + "id": "qeotZfkBYrIg", + "outputId": "84aa4055-e224-4838-9574-f483b6369079" + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "ImageNet class: bell pepper.\n" + ] + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "\u003cPIL.Image.Image image mode=RGB size=256x256 at 0x7F474593F6D8\u003e" + ] + }, + "execution_count": 5, + "metadata": { + "tags": [] + }, + "output_type": "execute_result" + } + ], + "source": [ + "# We split this into two cells so that we don't repeatedly jit the fwd fn.\n", + "logits = fwd(x[None]) # Give X a newaxis to make it batch-size-1\n", + "which_class = imagenet_classlist[int(logits.argmax())]\n", + "print(f'ImageNet class: {which_class}.')\n", + "im" + ] + } + ], + "metadata": { + "accelerator": "GPU", + "colab": { + "collapsed_sections": [ + "d3oIj5O1dyVB" + ], + "name": "Public: Load and Test NFNet Models.ipynb", + "provenance": [ + { + "file_id": "1PvnPRhmYywGYGsEf-UYRSt1y8SpBhiwv", + "timestamp": 1613472601457 + } + ] + }, + "kernelspec": { + "display_name": "Python 3", + "name": "python3" + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} diff --git a/nfnets/optim.py b/nfnets/optim.py index 0f1e4527..84f81298 100644 --- a/nfnets/optim.py +++ b/nfnets/optim.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/requirements.txt b/nfnets/requirements.txt index e591295f..217e984b 100644 --- a/nfnets/requirements.txt +++ b/nfnets/requirements.txt @@ -1,5 +1,6 @@ absl-py==0.10.0 chex>=0.0.2 +dill>=0.3.3 dm-haiku>=0.0.3 jax>=0.2.8 jaxlib>=0.1.58 diff --git a/nfnets/resnet.py b/nfnets/resnet.py index 59765269..85413760 100644 --- a/nfnets/resnet.py +++ b/nfnets/resnet.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/run.sh b/nfnets/run.sh index dade19fb..e40b157d 100644 --- a/nfnets/run.sh +++ b/nfnets/run.sh @@ -1,5 +1,5 @@ #!/bin/sh -# Copyright 2020 Deepmind Technologies Limited. +# Copyright 2021 Deepmind Technologies Limited. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/skipinit_resnet.py b/nfnets/skipinit_resnet.py index 51993153..06d6ba7c 100644 --- a/nfnets/skipinit_resnet.py +++ b/nfnets/skipinit_resnet.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/test.py b/nfnets/test.py index 4860a799..0e526886 100644 --- a/nfnets/test.py +++ b/nfnets/test.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. diff --git a/nfnets/utils.py b/nfnets/utils.py index 4a1599c7..f344a76d 100644 --- a/nfnets/utils.py +++ b/nfnets/utils.py @@ -1,4 +1,4 @@ -# Copyright 2020 DeepMind Technologies Limited. All Rights Reserved. +# Copyright 2021 DeepMind Technologies Limited. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. @@ -13,6 +13,7 @@ # limitations under the License. # ============================================================================== """Utils.""" +import dill import jax import jax.numpy as jnp import tree @@ -105,3 +106,23 @@ def split_tree(tuple_tree, base_tree, n): """Splits tuple_tree with n-tuple leaves into n trees.""" return [tree.map_structure_up_to(base_tree, lambda x: x[i], tuple_tree) # pylint: disable=cell-var-from-loop for i in range(n)] + + +def load_haiku_file(filename): + """Loads a haiku parameter tree, using dill.""" + with open(filename, 'rb') as in_file: + output = dill.load(in_file) + return output + + +def flatten_haiku_tree(haiku_dict): + """Flattens a haiku parameter tree into a flat dictionary.""" + out = {} + for module in haiku_dict.keys(): + out_module = module.replace('/~/', '.').replace('/', '.') + for key in haiku_dict[module]: + out_key = f'{out_module}.{key}' + out[out_key] = haiku_dict[module][key] + return out + +