kaggle-predict-ai-model-runtime
Kaggle contest
COMPETITION TITLE: Google - Fast or Slow? Predict AI Model Runtime
data set size 6 GB files 7,000
https://www.kaggle.com/competitions/predict-ai-model-runtime/data
my repo code https://github.com/timxor/kaggle-predict-ai-model-runtime
google-research-datasets/tpu_graphs repo https://github.com/google-research-datasets/tpu_graphs/tree/main
https://arxiv.org/pdf/2308.13490.pdf
local repo cd /Users/tim/code/kaggle-predict-ai-model-runtime
local dev env cd /Users/tim/data/tpugraphs cd /Users/tim/out
(tpugraphs) [email protected]😀=> python tiles_evaluate.py --dirs /Users/tim/out/tpugraphs_tiles/model_fc77809d9bd45c14f688a810e9e70318
WARNING:tensorflow:From /Users/tim/Library/Python/3.9/lib/python/site-packages/tensorflow/python/ops/distributions/distribution.py:259: ReparameterizationType.__init__ (from tensorflow.python.ops.distributions.distribution) is deprecated and will be removed after 2019-01-01.
Instructions for updating:
The TensorFlow Distributions library has moved to TensorFlow Probability (https://github.com/tensorflow/probability). You should update all references to use `tfp.distributions` instead of `tf.distributions`.
WARNING:tensorflow:From /Users/tim/Library/Python/3.9/lib/python/site-packages/tensorflow/python/ops/distributions/bernoulli.py:165: RegisterKL.__init__ (from tensorflow.python.ops.distributions.kullback_leibler) is deprecated and will be removed after 2019-01-01.
Instructions for updating:
The TensorFlow Distributions library has moved to TensorFlow Probability (https://github.com/tensorflow/probability). You should update all references to use `tfp.distributions` instead of `tf.distributions`.
dataset cache file: /Users/tim/data/tpugraphs/cache/tile/xla/9d2c87d76a2aa04cccb03e88d1a9811b-cache.npz
loaded from /Users/tim/data/tpugraphs/cache/tile/xla/9d2c87d76a2aa04cccb03e88d1a9811b-cache.npz
dataset cache file: /Users/tim/data/tpugraphs/cache/tile/xla/64b020c278749f00b1e5703e6c5f9a75-cache.npz
loaded from /Users/tim/data/tpugraphs/cache/tile/xla/64b020c278749f00b1e5703e6c5f9a75-cache.npz
dataset cache file: /Users/tim/data/tpugraphs/cache/tile/xla/fb70456c39b64044249d0737cbf34c47-cache.npz
loaded from /Users/tim/data/tpugraphs/cache/tile/xla/fb70456c39b64044249d0737cbf34c47-cache.npz
0%| | 0/1 [00:00<?, ?it/s]W0829 21:58:44.376159 8024514688 optimizer.py:70] At this time, the v2.11+ optimizer `tf.keras.optimizers.RestoredOptimizer` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.RestoredOptimizer`.
W0829 21:58:44.479212 8024514688 optimizer.py:70] At this time, the v2.11+ optimizer `tf.keras.optimizers.Adam` runs slowly on M1/M2 Macs, please use the legacy Keras optimizer instead, located at `tf.keras.optimizers.legacy.Adam`.
100%|████████████████████████████████████████████████████████████████████| 676/676 [01:51<00:00, 6.06it/s]
100%|███████████████████████████████████████████████████████████████████████| 1/1 [01:53<00:00, 113.74s/it]
{
"1": {
"bert_pretraining.4x4.fp16": 0.07383066415786743,
"inception_v3_batch_128_train": 0.07715363800525665,
"mlperf_bert_batch_24_2x2": 0.10472889244556427,
"resnet50.4x4.fp16": 0.1426999568939209,
"resnet_v1_50_official_batch_128_bf16": 0.10517748445272446,
"tf2_bert_pretrain_dynamic_batch_size": 0.06723014265298843,
"unet_3d.4x4.bf16": 0.37605753540992737
},
"5": {
"bert_pretraining.4x4.fp16": 0.01717139035463333,
"inception_v3_batch_128_train": 0.02828279696404934,
"mlperf_bert_batch_24_2x2": 0.04070887342095375,
"resnet50.4x4.fp16": 0.04174179583787918,
"resnet_v1_50_official_batch_128_bf16": 0.05139995366334915,
"tf2_bert_pretrain_dynamic_batch_size": 0.018168827518820763,
"unet_3d.4x4.bf16": 0.07754766941070557
}
}
(tpugraphs) [email protected]😀=> history
1 mvn clean install -P development
2 mvn clean install
3 c
cd /Users/tim/data/tpugraphs
1070 python3 tiles_train.py --model=EarlyJoinSAGE --toy_data=True
1071 pip install absl-py
1072 pip3 install absl-py
1073 python3 tiles_train.py --model=EarlyJoinSAGE --toy_data=True
1074 python3 tiles_train.py --model=EarlyJoinSAGE --toy_data=True
1075 pip3 install tensorflow tqdm tensorflow-ranking tensorflow_gnn --pre
1076 python3 tiles_train.py --model=EarlyJoinSAGE --toy_data=True
1077 pwd
1078 open .
1079 c
1080 python tiles_train.py --model=EarlyJoinSAGE
1081 atom /Users/tim/out/tpugraphs_tiles
1082 c
1083 python tiles_evaluate.py --dirs ./
1084 python3 tiles_evaluate.py --dirs ./
1085 python3 tiles_evaluate.py --dirs .
1086 python3 tiles_evaluate.py
1087 pwd
1088 ls
1089 c
1090 ls -la
1091 cd tpu_graphs
1092 ls
1093 -la
1094 ls -la
1095 pwd
1096 cd ..
1097 cd ..
1098 ls -la
1099 cd tpugraphs
1100 ls -la
1101 pwd
1102 python3 tiles_evaluate.py --dirs /Users/tim/data/tpugraphs
1103 python3 tiles_evaluate.py --dirs /Users/tim/out
1104 python tiles_evaluate.py --dirs /Users/tim/out
1105 python tiles_evaluate.py --dirs /Users/tim/out/tpugraphs_tiles/model_19f264c1c9ddfdf8ada61881bf824693
1106 python tiles_evaluate.py --dirs /Users/tim/out/tpugraphs_tiles/model_fc77809d9bd45c14f688a810e9e70318
(tpugraphs) [email protected]😀=>