mirror of
https://github.com/hanzo-ml/katib.git
synced 2026-09-24 00:31:27 +00:00
Mahdi KhashanandGitHub
19db2d47d4
* move to ghcr Signed-off-by: mahdikhashan <[email protected]> * move images to ghcr Signed-off-by: mahdikhashan <[email protected]> * manifests Signed-off-by: mahdikhashan <[email protected]> * change registry in all path Signed-off-by: mahdikhashan <[email protected]> * update script Signed-off-by: mahdikhashan <[email protected]> * fix Signed-off-by: mahdikhashan <[email protected]> * fix Signed-off-by: mahdikhashan <[email protected]> * slight fix Signed-off-by: mahdikhashan <[email protected]> --------- Signed-off-by: mahdikhashan <[email protected]> Signed-off-by: Mahdi Khashan <[email protected]>
151 lines
4.4 KiB
YAML
151 lines
4.4 KiB
YAML
---
|
|
# This CPU example aims to show all the possible operations
|
|
# is not very likely to get good result due to the extensive search space
|
|
|
|
# In practice, setting up a limited search space with more common operations is more likely to get better performance.
|
|
# For example, Efficient Neural Architecture Search via Parameter Sharing (https://arxiv.org/abs/1802.03268)
|
|
# uses only 6 operations, 3x3/5x5 convolution, 3x3/5x5 separable_convolution and 3x3 max_pooling/avg_pooling.
|
|
|
|
# It uses only 1 layer of CNN and 1 train epoch to show CPU support and it has very bad results.
|
|
# In practice, if you increase number of layers, training process on CPU will take more time.
|
|
apiVersion: kubeflow.org/v1beta1
|
|
kind: Experiment
|
|
metadata:
|
|
namespace: kubeflow
|
|
name: enas-cpu
|
|
spec:
|
|
parallelTrialCount: 2
|
|
maxTrialCount: 3
|
|
maxFailedTrialCount: 2
|
|
objective:
|
|
type: maximize
|
|
goal: 0.99
|
|
objectiveMetricName: Validation-Accuracy
|
|
algorithm:
|
|
algorithmName: enas
|
|
nasConfig:
|
|
graphConfig:
|
|
numLayers: 1
|
|
inputSizes:
|
|
- 32
|
|
- 32
|
|
- 3
|
|
outputSizes:
|
|
- 10
|
|
operations:
|
|
- operationType: convolution
|
|
parameters:
|
|
- name: filter_size
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "3"
|
|
- "5"
|
|
- "7"
|
|
- name: num_filter
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "32"
|
|
- "48"
|
|
- "64"
|
|
- "96"
|
|
- "128"
|
|
- name: stride
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "1"
|
|
- "2"
|
|
- operationType: separable_convolution
|
|
parameters:
|
|
- name: filter_size
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "3"
|
|
- "5"
|
|
- "7"
|
|
- name: num_filter
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "32"
|
|
- "48"
|
|
- "64"
|
|
- "96"
|
|
- "128"
|
|
- name: stride
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "1"
|
|
- "2"
|
|
- name: depth_multiplier
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "1"
|
|
- "2"
|
|
- operationType: depthwise_convolution
|
|
parameters:
|
|
- name: filter_size
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "3"
|
|
- "5"
|
|
- "7"
|
|
- name: stride
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "1"
|
|
- "2"
|
|
- name: depth_multiplier
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- "1"
|
|
- "2"
|
|
- operationType: reduction
|
|
parameters:
|
|
- name: reduction_type
|
|
parameterType: categorical
|
|
feasibleSpace:
|
|
list:
|
|
- max_pooling
|
|
- avg_pooling
|
|
- name: pool_size
|
|
parameterType: int
|
|
feasibleSpace:
|
|
min: "2"
|
|
max: "3"
|
|
step: "1"
|
|
trialTemplate:
|
|
primaryContainerName: training-container
|
|
trialParameters:
|
|
- name: neuralNetworkArchitecture
|
|
description: NN architecture contains operations ID on each NN layer and skip connections between layers
|
|
reference: architecture
|
|
- name: neuralNetworkConfig
|
|
description: Configuration contains NN number of layers, input and output sizes, description what each operation ID means
|
|
reference: nn_config
|
|
trialSpec:
|
|
apiVersion: batch/v1
|
|
kind: Job
|
|
spec:
|
|
template:
|
|
spec:
|
|
containers:
|
|
- name: training-container
|
|
image: ghcr.io/kubeflow/katib/enas-cnn-cifar10-cpu:latest
|
|
command:
|
|
- python3
|
|
- -u
|
|
- RunTrial.py
|
|
- --num_epochs=1
|
|
- --architecture="${trialParameters.neuralNetworkArchitecture}"
|
|
- --nn_config="${trialParameters.neuralNetworkConfig}"
|
|
restartPolicy: Never
|