d3x Commands¶
d3x commands are used for managing the user workspace, apikeys, ray clusters, user apps and deployments. This page provides information about all of the d3x commands that are currently available, along with their functions, usage syntax and available options.
Usage¶
d3x [OPTIONS] COMMAND [ARGS]...
d3x --help
Function¶
Group for d3x commands.
Options¶
Option |
Description |
---|---|
|
enable requests debugging. |
|
cluster profile |
|
Show help for the command.. |
Commands¶
Command |
Description |
---|---|
|
Details about d3x CLI configuration. |
|
Commands to generate and get DKubeX API keys. |
|
Commands regarding creating and managing user-defined applications. |
|
Commands for backing up datasets. |
|
Start local VSCode and connect to DKubeX workspace. |
|
Commands for creating, managing and using datasets. |
|
Commands regarding using foundation models. |
|
List DKubeX workspaces. |
|
Commands regarding using LLMs. |
|
Login to DKubeX with particular user profile in the CLI. |
|
Commands for using and managing experiments and models in MLFlow from CLI. |
|
List & Manage MLFlow models registry. |
|
Mount S3 bucket in read/write mode. |
|
Forward port from workspace to local computer. |
|
Commands regarding user profiles in DKubeX. |
|
Commands regarding Ray jobs and Ray clusters. |
|
Reset given workspace to default image |
|
Commands for restoring databases. |
|
Copy files(s)/directory from/to DKubeX workspace. |
|
Commands for creating model deployments. |
|
Commands for managing SkyPilot clusters, resources and jobs. |
|
SSH to DKubeX workspace. |
|
Copy ssh public key to DKubeX user’s home. |
|
Start given workspace. |
|
Start given workspace app. |
|
Stop given workspace. |
|
Stop given workspace app. |
|
Commands regarding using Tensorboard. |
|
Show workspace resource usage. |
|
Unmount bucket previously mounted using |
|
Show DKubeX version. |
|
VNC connection to DKubeX workspace. |
d3x about
¶
Usage
d3x about [OPTIONS]
d3x about --help
Function
Details about d3x CLI config
Options
Option
Description
-o
,--output [yaml|json]
output format (supported- json, yaml)
--help
Show help for the command..
d3x apikey
¶
Usage
d3x apikey [OPTIONS] COMMAND [ARGS]...
d3x apikey --help
Function
Commands to generate and get DKubeX API keys.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
generate
Generate new user API key
get
Get current user API key
Usage
d3x apikey generate [OPTIONS]
d3x apikey generate --help
Function
Generate new user API key.
Options
Option
Description
-o
,--output [yaml|json]
output format (supported- json, yaml)
--help
Show help for the command..
Usage
d3x apikey get [OPTIONS]
d3x apikey get --help
Function
Get current user API key.
Options
Option
Description
-o
,--output [yaml|json]
output format (supported- json, yaml)
--help
Show help for the command..
d3x apps
¶
Usage
d3x apps [OPTIONS] COMMAND [ARGS]...
d3x apps --help
Function
Commands regarding creating and managing user-defined applications.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
create
Launch user-defined application on DKubeX
delete
Delete user-defined application
list
List user-defined applications
open
Open user-defined application in browser
Usage
d3x apps create [OPTIONS] [COMMAND]...
d3x apps create --help
Function
Launch user-defined application on DKubeX.
Options
Option
Description
-c
,--config TEXT
Path to the YAML configuration file [required]
--help
Show help for the command..
Usage
d3x apps delete [OPTIONS] NAME
d3x apps delete --help
Function
Delete user-defined applications.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x apps list [OPTIONS]
d3x apps list --help
Function
List user apps.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x apps open [OPTIONS] NAME
d3x apps open --help
Function
Open user app in browser.
Options
Option
Description
--help
Show help for the command..
d3x backup
¶
Usage
d3x backup [OPTIONS] COMMAND [ARGS]...
d3x backup --help
Function
Commands for backing up datasets.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
weaviate
command for weaviate backup
Usage
d3x backup weaviate [OPTIONS]
d3x backup weaviate --help
Function
Command for dataset backup to Weaviate.
Options
Option
Description
-d
,--dataset TEXT
A name to represent the ingested docs [required]
-i
,--id TEXT
An ID for the backup [required]
-b
,--backend [filesystem|s3]
s3 or filesystem to save the backup [required]
--help
Show help for the command..
d3x code
¶
Usage
d3x code [OPTIONS] WORKSPACE_NAME
d3x code --help
Function
Start local VSCode and connect to DKubeX workspace
Options
Option
Description
--help
Show help for the command..
d3x dataset
¶
Usage
d3x dataset [OPTIONS] COMMAND [ARGS]...
d3x dataset --help
Function
Commands for creating, managing and using datasets.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
delete
Deletes specified dataset if present from the database
evaluate
Run evaluation pipeline over the given dataset.
ingest
Ingest the documents from provided source into the database.
list
lists all the datasets present in database
query
Query over documents using the selected pipeline.
show
shows documents present in the specified dataset
Usage
d3x dataset delete [OPTIONS] DATASET
d3x dataset delete --help
Function
Deletes specified dataset if present from the database
Options
Option
Description
--help
Show help for the command..
Usage
d3x dataset evaluate [OPTIONS]
d3x dataset evaluate --help
Function
Run evaluation pipeline over the given dataset.
Options
Option
Description
-d
,--dataset TEXT
Dataset on which the queries should be applied. [required]
-p
,--pipeline TEXT
RAG pipeline to be used.
-c
,--config TEXT
User configuration for pipeline stages
-s
,--remote-sky
If the ingestion should be scheduled on remote sky cluster.
-m
,--remote-command TEXT
sky/ray job command to run.
--dkubex-url TEXT
URL of the dkubex for the remote clusters to reach.
-k
,--dkubex-apikey TEXT
API key for remote clusters to reach dkubex.
--help
Show help for the command..
Usage
d3x dataset ingest [OPTIONS]
d3x dataset ingest --help
Function
Ingest the documents from provided source into the vectorstore identified by given dataset name.
Note
SUPPORTED EXTRA OPTIONS: Make use of the options below to overwrite the individual values without need to copy config file.
--embedder.class=[HuggingfaceEmbedding/OpenAIEmbedding] --embedder.model=<embedding-model-name-for-the-above-class>
Options
Option
Description
-d
,--dataset TEXT
A name to represent the ingested docs [required]
-p
,--pipeline TEXT
Ingestion pipeline to be used.
-c
,--config TEXT
User configuration for pipeline stages
-s
,--remote-sky
If the ingestion should be scheduled on remote sky cluster.
-r
,--remote-ray
If the ingestion should be scheduled on remote ray cluster.
-m
,--remote-command TEXT
sky/ray job command to run.
--dkubex-url TEXT
URL of the dkubex for the remote clusters to reach.
-k
,--dkubex-apikey TEXT
API key for remote clusters to reach dkubex.
-w
,--num-workers INTEGER
Number of process to use for parallelization.
--help
Show help for the command..
Usage
d3x dataset list [OPTIONS]
d3x dataset list --help
Function
Lists all the datasets present in database
Options
Option
Description
--help
Show help for the command..
Usage
d3x dataset query [OPTIONS]
d3x dataset query --help
Function
Query over documents using the selected pipeline.
Note
SUPPORTED EXTRA OPTIONS: Make use of the options below to overwrite the individual values without need to copy config file.
--vectorstore_retriever.embedding_class=[HuggingfaceEmbedding/OpenAIEmbedding] --vectorstore_retriever.embedding_model= --vectorstore_retriever.top_k= --chat_engine.llm=[openai/dkubex] --chat_engine.url=<dkubex-dep-url> --chat_engine.llmkey=<llm-access-key> --chat_engine.max_tokens
Options
Option
Description
-d
,--dataset TEXT
Dataset on which the question should be applied. [required]
-p
,--pipeline TEXT
RAG pipeline to be used.
-c
,--config TEXT
User configuration for pipeline stages
-q
,--question TEXT
User query to run through the pipeline
-b
,--batch TEXT
Batch of questions to run through the pipeline
-i
,--interactive
Run this tool in cli mode
-k
,--conversation TEXT
Name of the conversation to track all the questions for.
--help
Show help for the command..
Usage
d3x dataset show [OPTIONS]
d3x dataset show --help
Function
Shows documents present in the specified dataset
Options
Option
Description
-d
,--dataset TEXT
Dataset name to list all the docs present in it [required]
--help
Show help for the command..
d3x fm
¶
Usage
d3x fm [OPTIONS] COMMAND [ARGS]...
d3x fm --help
Function
Commands regarding using foundation models.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
delete
Delete FM job.
get
Get FM job details.
jobs
List FM jobs.
logs
Show logs of given FM job.
trainchunks
Create train and/or test chunks for finetuning.
tune
LLM Finetuning process related commands.
Usage
d3x fm delete [OPTIONS] JOB_NAME
d3x fm delete --help
Function
Delete FM job.
Options
Option
Description
--help
Show help for the command..
Usage
d3x fm get [OPTIONS] JOB_NAME
d3x fm get --help
Function
Get FM job details.
Options
Option
Description
--help
Show help for the command..
Usage
d3x fm jobs [OPTIONS]
d3x fm jobs --help
Function
List FM jobs.
Options
Option
Description
--help
Show help for the command..
Usage
d3x fm logs [OPTIONS] JOB_NAME
d3x fm logs --help
Function
Show logs of given FM job.
Options
Option
Description
--help
Show help for the command..
Usage
d3x fm trainchunks [OPTIONS]
d3x fm trainchunks --help
Function
Create train and/or test chunks for finetuning.
Options
Option
Description
-s
,--source TEXT
The path where finetune chunks are sourced from [required]
-d
,--destination TEXT
The path where chunks will be kept for training [required]
--split [train|test]
specify train or test. Default: both
-p
,--split-percent INTEGER
specify percentage of docs to go in train/test
-l
,--last
pick documents from backwards
--help
Show help for the command..
Usage
d3x fm tune [OPTIONS] COMMAND [ARGS]...
d3x fm tune --help
Function
LLM Finetuning process related commands.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
model
Commands regarding model finetuning.
Usage
d3x fm tune model [OPTIONS] COMMAND [ARGS]...
d3x fm tune model --help
Function
Commands regarding model finetuning.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
finetune
Command to finetune a model.
merge
Command to merge finetuned models.
quantize
Command to quantize a model.
Usage
d3x fm tune model finetune [OPTIONS]
d3x fm tune model finetune --help
Function
Command to finetune a model.
Options
Option
Description
-j
,--job-name TEXT
job name
-d
,--docker-image TEXT
docker image for the ray job
-e
,--epochs INTEGER
number of epochs
-b
,--batch-size INTEGER
batch size
-l
,--dataset-location TEXT
dataset location
-o
,--output-location TEXT
output location
-n
,--model-name TEXT
model name(should contain complete hf path meta-llama/Llama-2-13b-chat-hf)
-p
,--model-path TEXT
downloaded model location from huggingface
-h
,--hf-token TEXT
huggingface token
-c
,--cpu INTEGER
number of cpu
-g
,--gpu INTEGER
number of gpu
-t
,--type TEXT
instance type
-m
,--mem INTEGER
number of memories
--as-test
As test
--stop-perplexity FLOAT
Target perplexity to reach after which to stop training. Default is 0. If 0, training will not stop on perplexity.
--ctx-len INTEGER
Learning rate to use.
--experiment-name TEXT
mlflow experiment name
--help
Show help for the command..
Usage
d3x fm tune model merge [OPTIONS]
d3x fm tune model merge --help
Function
Command to merge finetuned models.
Options
Option
Description
-j
,--job-name TEXT
ray cluster name
-d
,--docker-image TEXT
docker image for the ray job
-o
,--output-path TEXT
Path to output directory. Defaults to the orginal checkpoint directory.
-n
,--model-name TEXT
model name(should contain complete hf path meta-llama/Llama-2-13b-chat-hf)
-p
,--model-path TEXT
downloaded model location from huggingface
-cp
,--checkpoint TEXT
Path to checkpoint containing the LoRA weights.
-c
,--cpu INTEGER
number of cpu
-m
,--mem INTEGER
cpu memory
-g
,--gpu INTEGER
number of gpus
-t
,--type TEXT
instance type
--help
Show help for the command..
Usage
d3x fm tune model quantize [OPTIONS]
d3x fm tune model quantize --help
Function
Command to quantize a model.
Options
Option
Description
-j
,--job-name TEXT
ray cluster name [required]
-d
,--docker-image TEXT
docker image for the ray job
-p
,--model-path TEXT
merged model path [required]
-o
,--output-path TEXT
Path to output directory [required]
-c
,--cpu INTEGER
number of cpu
-m
,--mem INTEGER
cpu memory
-g
,--gpu INTEGER
number of gpu
-t
,--type TEXT
instance type
--help
Show help for the command..
d3x list
¶
Usage
d3x list [OPTIONS]
d3x list --help
Function
List d3x workspaces
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x llms
¶
Usage
d3x llms [OPTIONS] COMMAND [ARGS]...
d3x llms --help
Function
Commands regarding using LLMs.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
deploy
Deploy a LLM on dkubex.
list
List the supported LLMs
Usage
d3x llms deploy [OPTIONS]
d3x llms deploy --help
Function
Deploy a LLM on dkubex.
Options
Option
Description
-n
,--name TEXT
name of the deployment [required]
-m
,--model TEXT
name or path of the model to deploy for serving
--base_model TEXT
base model name
--token TEXT
hugging face token
--config TEXT
github raw content url or local config path
--mlflow TEXT
name of the mlflow registered model along with version (example: diffusion:1)
--image TEXT
customized docker image for llm deployment
--type TEXT
instance type
--min_replicas TEXT
minimum replicas
--max_replicas TEXT
maximum replicas
--publish
publish the deployment
-o
,--output [yaml|json]
supported only json,yaml
-sky
,--remote-sky
If the model should be deployed on remote sky cluster.
--sky-accelerator TEXT
accelerator for remote sky cluster
-k
,--dkubex-apikey TEXT
API key for remote clusters to reach dkubex.
--sky-yaml TEXT
Path to custom yaml file for sky serve.
--help
Show help for the command..
Usage
d3x llms list [OPTIONS]
d3x llms list --help
Function
List the supported LLMs
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x login
¶
Usage
d3x login [OPTIONS]
d3x login --help
Function
Login to DKubeX with particular user profile in the CLI.
Options
Option
Description
--help
Show help for the command..
d3x mlflow
¶
Usage
d3x mlflow [OPTIONS] COMMAND [ARGS]...
d3x mlflow --help
Function
Commands for using and managing experiments and models in MLFlow from CLI.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
experiments
Commands for managing experiments in MLFlow.
models
Commands for managing models in MLFlow.
Usage
d3x mlflow experiments [OPTIONS] COMMAND [ARGS]...
d3x mlflow experiments --help
Function
Commands for managing experiments in MLFlow.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
archive
Archive experiment in MLFlow
delete
Delete experiment in MLFlow
list
List experiments in MLFlow
restore
Restore experiment in MLFlow
Usage
d3x mlflow experiments archive [OPTIONS] ID
d3x mlflow experiments archive --help
Function
Archive experiment in MLFlow.
Options
Option
Description
--help
Show help for the command..
Usage
d3x mlflow experiments delete [OPTIONS] ID
d3x mlflow experiments delete --help
Function
Delete experiment in MLFlow.
Options
Option
Description
--help
Show help for the command..
Usage
d3x mlflow experiments list [OPTIONS]
d3x mlflow experiments list --help
Function
List experiments in MLFlow.
Options
Option
Description
--help
Show help for the command..
Usage
d3x mlflow experiments restore [OPTIONS] ID
d3x mlflow experiments restore --help
Function
Restore experiment in MLFlow.
Options
Option
Description
--help
Show help for the command..
Usage
d3x mlflow models [OPTIONS] COMMAND [ARGS]...
d3x mlflow models --help
Function
Commands for managing models in MLFlow.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
archive
Archive model in MLFlow
delete
Delete model in MLFlow
Usage
d3x mlflow models archive [OPTIONS] NAME VERSION
d3x mlflow models archive --help
Function
Archive model in MLFlow.
Options
Option
Description
--help
Show help for the command..
Usage
d3x mlflow models delete [OPTIONS] NAME VERSION
d3x mlflow models delete --help
Function
Delete model in MLFlow.
Options
Option
Description
--help
Show help for the command..
d3x models
¶
Usage
d3x models [OPTIONS] COMMAND [ARGS]...
d3x models --help
Function
List & Manage MLFlow models registry.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
import
Import saved model to MLFlow.
list
List models registered in MLFlow.
Usage
d3x models import [OPTIONS] NAME {pytorch|tensorflow|sklearn|xgboost|custom_model} SAVED_MODEL_PATH
d3x models import --help
Function
Import saved model to MLFlow.
Options
Option
Description
--class_path TEXT
pytorch class file path
--class_instance TEXT
class instance
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x models list [OPTIONS]
d3x models list --help
Function
List models registered in MLFlow registry.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x mount
¶
Usage
d3x mount [OPTIONS] [BUCKET] [MOUNTPATH]
d3x mount --help
Function
Mount S3 bucket in read/write mode.
Options
Option
Description
-a
,--all
Mounts all the previous buckets.
-d
,--dry-run
Runs a dry run of mounting bucket, does not mount the buckets.
--help
Show help for the command..
d3x port-forward
¶
Usage
d3x port-forward [OPTIONS] WORKSPACE_NAME SERVICE TARGET_PORT LOCAL_PORT
d3x port-forward --help
Function
Forward port from workspace to local computer.
Options
Option
Description
--help
Show help for the command..
d3x profile
¶
Usage
d3x profile [OPTIONS] COMMAND [ARGS]...
d3x profile --help
Function
Commands regarding user profiles in DKubeX.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
configure
Configure user profile.
delete
Delete user profile.
list
List user profiles.
set
Set default user profile.
Usage
d3x profile configure [OPTIONS]
d3x profile configure --help
Function
Configure user profile.
Options
Option
Description
--help
Show help for the command..
Usage
d3x profile delete [OPTIONS] PROFILE
d3x profile delete --help
Function
Delete user profile.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x profile list [OPTIONS]
d3x profile list --help
Function
List user profiles.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x profile set [OPTIONS] PROFILE
d3x profile set --help
Function
Set default user profile.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x ray
¶
Usage
d3x ray [OPTIONS] COMMAND [ARGS]...
d3x ray --help
Function
Commands regarding Ray jobs and Ray clusters.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
activate
Set Ray cluster as active cluster.
address
List Ray clusters along with their cluster Ray address.
create
Create Ray cluster.
dashboard
Open Ray dashboard.
delete
Delete Ray cluster.
get
Get a Ray cluster.
job
Commands regarding managing Ray jobs in currently active Ray cluster.
list
List Ray clusters.
Usage
d3x ray activate [OPTIONS] CLUSTER
d3x ray activate --help
Function
Set Ray cluster as active cluster.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray address [OPTIONS]
d3x ray address --help
Function
List Ray clusters along with their cluster Ray address.
Options
Option
Description
--all
List all compute templates from all users
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray create [OPTIONS]
d3x ray create --help
Function
Create Ray cluster.
Options
Option
Description
-n
,--name TEXT
name of the ray cluster [required]
-v
,--version TEXT
version of the ray cluster (default 2.8.0)
-i
,--image TEXT
image of ray head/worker node
--cpu INTEGER
ray worker node CPU (default 2)
--gpu INTEGER
ray worker node GPU (default 0)
--memory INTEGER
ray worker node memory in GB (default 4)
--hcpu INTEGER
ray head node CPU (default 2)
--hgpu INTEGER
ray head node GPU (default 0)
--hmemory INTEGER
ray head node memory in GB (default 4)
--htype TEXT
ray worker head node instance type
--type TEXT
ray worker node instance type
--dockerserver TEXT
docker registry url
--dockeruser TEXT
docker registry username
--dockerpsw TEXT
docker registry password
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray dashboard [OPTIONS]
d3x ray dashboard --help
Function
Open Ray dashboard.
Options
Option
Description
--help
Show help for the command..
Usage
d3x ray delete [OPTIONS] NAME
d3x ray delete --help
Function
Delete Ray cluster.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray get [OPTIONS]
d3x ray get --help
Function
Get a Ray Cluster.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray job [OPTIONS] COMMAND [ARGS]...
d3x ray job --help
Function
Commands regarding managing Ray jobs in currently active Ray cluster.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
delete
Deletes a stopped job and its associated data from memory.
list
Lists all running jobs and their information.
logs
Gets the logs of a job.
show
Queries for show details of job.
status
Queries for the current status of a job.
stop
Attempts to stop a job.
submit
Submits a job to be run on the cluster.
Usage
d3x ray job delete [OPTIONS] JOB_ID
d3x ray job delete --help
Function
Deletes a stopped job and its associated data from memory.
Note
Only supported for jobs that are already in a terminal state. Fails with exit code 1 if the job is not already stopped. Does not delete job logs from disk. Submitting a job with the same submission ID as a previously deleted job is not supported and may lead to unexpected behavior.
PublicAPI (alpha): This API is in alpha and may change before becoming stable.
Options
Option
Description
--address TEXT
Address of the Ray cluster to connect to. Can also be specified using the RAY_ADDRESS environment variable.
--verify BOOL | TEXT
Boolean indication to verify the server’s TLS certificate or a path to a file or directory of trusted certificates. [default: True]
--headers TEXT
Used to pass headers through http/s to the Ray Cluster. Please follow JSON formatting.
--log-style [auto|record|pretty]
If ‘pretty’, outputs with formatting and color. If ‘record’, outputs record-style ‘pretty’, and disables pretty logging if stdin is not a TTY.
--log-color [auto|false|true]
Use color logging. Auto enables color logging if stdout is a TTY.
-v
,--verbose
Activates verbose logging.
--help
Show help for the command..
Usage
d3x ray job list [OPTIONS]
d3x ray job list --help
Function
Lists all running jobs and their information.
Options
Option
Description
--address TEXT
Address of the Ray cluster to connect to. Can also be specified using the RAY_ADDRESS environment variable.
--verify BOOL | TEXT
Boolean indication to verify the server’s TLS certificate or a path to a file or directory of trusted certificates. [default: True]
--headers TEXT
Used to pass headers through http/s to the Ray Cluster. Please follow JSON formatting.
--log-style [auto|record|pretty]
If ‘pretty’, outputs with formatting and color. If ‘record’, outputs record-style ‘pretty’, and disables pretty logging if stdin is not a TTY.
--log-color [auto|false|true]
Use color logging. Auto enables color logging if stdout is a TTY.
-v
,--verbose
Activates verbose logging.
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray job logs [OPTIONS] JOB_ID
d3x ray job logs --help
Function
Gets the logs of a job.
Options
Option
Description
--address TEXT
Address of the Ray cluster to connect to. Can also be specified using the RAY_ADDRESS environment variable.
-f
,--follow
If set, follow the logs (like tail -f).
--verify BOOL | TEXT
Boolean indication to verify the server’s TLS certificate or a path to a file or directory of trusted certificates. [default: True]
--headers TEXT
Used to pass headers through http/s to the Ray Cluster. Please follow JSON formatting.
--log-style [auto|record|pretty]
If ‘pretty’, outputs with formatting and color. If ‘record’, outputs record-style ‘pretty’, and disables pretty logging if stdin is not a TTY.
--log-color [auto|false|true]
Use color logging. Auto enables color logging if stdout is a TTY.
-v
,--verbose
Activates verbose logging.
--help
Show help for the command..
Usage
d3x ray job show [OPTIONS] JOB_ID
d3x ray job show --help
Function
Queries for show details of job.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x ray job status [OPTIONS] JOB_ID
d3x ray job status --help
Function
Queries for the current status of a job.
Options
Option
Description
--address TEXT
Address of the Ray cluster to connect to. Can also be specified using the RAY_ADDRESS environment variable.
--verify BOOL | TEXT
Boolean indication to verify the server’s TLS certificate or a path to a file or directory of trusted certificates. [default: True]
--headers TEXT
Used to pass headers through http/s to the Ray Cluster. Please follow JSON formatting.
--log-style [auto|record|pretty]
If ‘pretty’, outputs with formatting and color. If ‘record’, outputs record-style ‘pretty’, and disables pretty logging if stdin is not a TTY.
--log-color [auto|false|true]
Use color logging. Auto enables color logging if stdout is a TTY.
-v
,--verbose
Activates verbose logging.
--help
Show help for the command..
Usage
d3x ray job stop [OPTIONS] JOB_ID
d3x ray job stop --help
Function
Attempts to stop a job.
Options
Option
Description
--address TEXT
Address of the Ray cluster to connect to. Can also be specified using the RAY_ADDRESS environment variable.
--no-wait
If set, will not wait for the job to exit.
--verify BOOL | TEXT
Boolean indication to verify the server’s TLS certificate or a path to a file or directory of trusted certificates. [default: True]
--headers TEXT
Used to pass headers through http/s to the Ray Cluster. Please follow JSON formatting.
--log-style [auto|record|pretty]
If ‘pretty’, outputs with formatting and color. If ‘record’, outputs record-style ‘pretty’, and disables pretty logging if stdin is not a TTY.
--log-color [auto|false|true]
Use color logging. Auto enables color logging if stdout is a TTY.
-v
,--verbose
Activates verbose logging.
--help
Show help for the command..
Usage
d3x ray job submit [OPTIONS] ENTRYPOINT...
d3x ray job submit --help
Function
Submits a job to be run on the cluster.
Options
Option
Description
--address TEXT
Address of the Ray cluster to connect to. Can also be specified using the RAY_ADDRESS environment variable.
--job-id TEXT
DEPRECATED: Use –submission-id instead.
--submission-id TEXT
Submission ID to specify for the job. If not provided, one will be generated.
--runtime-env TEXT
Path to a local YAML file containing a runtime_env definition.
--runtime-env-json TEXT
JSON-serialized runtime_env dictionary.
--working-dir TEXT
Directory containing files that your job will run in. Can be a local directory or a remote URI to a .zip file (S3, GS, HTTP). If specified, this overrides the option in –runtime-env.
--metadata-json TEXT
JSON-serialized dictionary of metadata to attach to the job.
--entrypoint-num-cpus FLOAT
the quantity of CPU cores to reserve for the entrypoint command, separately from any tasks or actors that are launched by it
--entrypoint-num-gpus FLOAT
the quantity of GPUs to reserve for the entrypoint command, separately from any tasks or actors that are launched by it
--entrypoint-resources TEXT
a JSON-serialized dictionary mapping resource name to resource quantity describing resources to reserve for the entrypoint command, separately from any tasks or actors that are launched by it
--no-wait
If set, will not stream logs and wait for the job to exit.
--verify BOOL | TEXT
Boolean indication to verify the server’s TLS certificate or a path to a file or directory of trusted certificates. [default: True]
--headers TEXT
Used to pass headers through http/s to the Ray Cluster. Please follow JSON formatting
--log-style [auto|record|pretty]
If ‘pretty’, outputs with formatting and color. If ‘record’, outputs record-style without formatting. ‘auto’ defaults to ‘pretty’, and disables pretty logging if stdin is not a TTY.
--log-color [auto|false|true]
Use color logging. Auto enables color logging if stdout is a TTY.
-v
,--verbose
Activates verbose logging.
--help
Show help for the command..
Usage
d3x ray list [OPTIONS]
d3x ray list --help
Function
List Ray Clusters
Options
Option
Description
--all
List all compute templates from all users
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x reset
¶
Usage
d3x reset [OPTIONS]
d3x reset --help
Function
Reset given workspace to default image.
Options
Option
Description
--help
Show help for the command..
d3x restore
¶
Usage
d3x restore [OPTIONS] COMMAND [ARGS]...
d3x restore --help
Function
Commands for restoring databases.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
weaviate
Command for Weaviate restore.
Usage
d3x restore weaviate [OPTIONS]
d3x restore weaviate --help
Function
Command for Weaviate restore.
Options
Option
Description
-i
,--id TEXT
An ID for the backup [required]
-b
,--backend [filesystem|s3]
s3 or filesystem to save the backup [required]
--help
Show help for the command..
d3x scp
¶
Usage
d3x scp [OPTIONS] SOURCE TARGET
d3x scp --help
Function
Copy files(s)/directory from/to DKubeX workspace.
Options
Option
Description
--help
Show help for the command..
d3x serve
¶
Usage
d3x serve [OPTIONS] COMMAND [ARGS]...
d3x serve --help
Function
Commands for creating model deployments.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
create
Create deployment
delete
Delete deployment
get
Get Deployment
list
List Deployments
status
Check deployment Status
Usage
d3x serve create [OPTIONS]
d3x serve create --help
Function
Create deployment.
Options
Option
Description
-n
,--name TEXT
name of the ray serve [required]
-r
,--registry_type [hugging_face|mlflow|local|none]
registry type [required]
--hface_repoid TEXT
hugging face repo id
--hface_tokenizer TEXT
hugging face tokenizer
--hface_classifier TEXT
hugging face classifier
--hface_token TEXT
hugging face token
--model TEXT
mlflow model name
--model_version TEXT
mlflow model version
--depfilepath TEXT
deployment filepath
--working_dir TEXT
local serving directory path
--is_private_repo
private repo
--repo_name TEXT
github repo name
--repo_org TEXT
github personal/organization account
--access_token TEXT
github access token
--branch_name TEXT
github branch name
--commit_id TEXT
github commit id
--unprotected
serving without token
-e
,--env TEXT
environment variable
--ngpus TEXT
number of gpus
--min_cpu TEXT
min number of cpus for cluster creation
--max_cpu TEXT
max number of cpus per cluster creation
--min_memory TEXT
min memory for cluster cluster creation
--max_memory TEXT
max memory for cluster cluster creation
--min_replicas TEXT
min deployment replicas
--max_replicas TEXT
max deployment replicas
--max_concurrent_requests TEXT
max concurrent requests
--description TEXT
deployment description
--image TEXT
cluster image
--publish
publish the deployment
--no_rsrc_limits
no resource limits
--type TEXT
node selector
--timeout INTEGER
deployment timeout
--kserve
enable kserve deployment
--model_format [tensorflow|sklearn|pytorch|xgboost|pmml|lightgbm|paddle|mlflow|custom]
model format in case of kserve deployment
--model_path TEXT
local path of model for kserve deployment
--protocol_version [v1|v2]
dataplane protocol
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x serve delete [OPTIONS] NAME
d3x serve delete --help
Function
Delete deployment.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x serve get [OPTIONS] NAME
d3x serve get --help
Function
Get deployment.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x serve list [OPTIONS]
d3x serve list --help
Function
List deployments.
Options
Option
Description
--all
List all the deployments
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x serve status [OPTIONS] NAMESPACE ID
d3x serve status --help
Function
Check deployment Status.
Options
Option
Description
--help
Show help for the command..
d3x sky
¶
Usage
d3x sky [OPTIONS] COMMAND [ARGS]...
d3x sky --help
Function
Commands for managing SkyPilot clusters, resources and jobs.
Options
Option
Description
--install-shell-completion [bash|zsh|fish|auto]
Install shell completion for the specified shell.
--uninstall-shell-completion [bash|zsh|fish|auto]
Uninstall shell completion for the specified shell.
-v
,--version
Show the version and exit.
-c
,--commit
Show the commit hash and exit
-h
,--help
Show help for the command..
Commands
Command
Description
launch
Launch a cluster or task.
exec
Execute a task or command on an existing cluster.
status
Show clusters.
cost-report
Show estimated costs for launched clusters.
queue
Show the job queue for cluster(s).
logs
Tail the log of a job.
cancel
Cancel job(s).
stop
Stop cluster(s).
autostop
Schedule an autostop or autodown for cluster(s).
start
Restart cluster(s).
down
Tear down cluster(s).
check
Check which clouds are available to use.
show-gpus
Show supported GPU/TPU/accelerators and their prices.
storage
Commands for using SkyPilot Storage CLI.
bench
Commands for usingSkyPilot Benchmark CLI.
spot
Commands for using managed Spot CLI (spot instances with auto-recovery).
serve
Commands for using SkyServe CLI (multi-region, multi-cloud serving).
Usage
d3x sky launch [OPTIONS] [ENTRYPOINT]...
d3x sky launch --help
Function
Launch a cluster or task. If ENTRYPOINT points to a valid YAML file, it is read in as the task specification. Otherwise, it is interpreted as a bash command. In both cases, the commands are run under the task’s workdir (if specified) and they undergo job queue scheduling.
Options
Option
Description
-c
,--cluster TEXT
A cluster name. If provided, either reuse an existing cluster with that name or provision a new cluster with that name. Otherwise provision a new cluster with an autogenerated name.
--dryrun
If True, do not actually run the job.
-s
,--detach-setup
If True, run setup in non-interactive mode as part of the job itself. You can safely ctrl-c to detach from logging, and it will not interrupt the setup process. To see the logs again after detaching, use sky logs. To cancel setup, cancel the job via sky cancel. Useful for long-running setup commands.
-d
,--detach-run
If True, as soon as a job is submitted, return from this call and do not stream execution logs.
--docker
If used, runs locally inside a docker container.
-n
,--name TEXT
Task name. Overrides the ‘name’ config in the YAML if both are supplied.
--workdir DIRECTORY
If specified, sync this dir to the remote working directory, where the task will be invoked. Overrides the ‘workdir’ config in the YAML if both are supplied.
--cloud TEXT
The cloud to use. If specified, overrides the ‘resources.cloud’ config. Passing ‘none’ resets the config.
--region TEXT
The region to use. If specified, overrides the ‘resources.region’ config. Passing ‘none’ resets the config.
--zone TEXT
The zone to use. If specified, overrides the ‘resources.zone’ config. Passing ‘none’ resets the config.
--num-nodes INTEGER
Number of nodes to execute the task on. Overrides the ‘num_nodes’ config in the YAML if both are supplied.
--cpus TEXT
Number of vCPUs each instance must have (e.g.,
--cpus=4
(exactly 4) or--cpus=4+
(at least 4)). This is used to automatically select the instance type.--memory TEXT
Amount of memory each instance must have (e.g.,
--memory=4
(exactly 4 GB) or--memory=4+
(at least 4 GB)). This is used to automatically select the instance type.--hcpu INTEGER
Head node CPU. Overrides the ‘resources.head_cpu’ config in the YAML if both are supplied.
--hmemory INTEGER
Head node memory in GB. Overrides the ‘resources.head_memory’ config in the YAML if both are supplied.
--htype TEXT
Head node instance type. Overrides the ‘resources.head_instance_type’ config in the YAML if both are supplied.
--disk-size INTEGER
OS disk size in GBs.
--disk-tier [low|medium|high|best|none]
OS disk tier. Could be one of low, medium, high, best, none. If best is specified, use the best possible disk tier. If none is specified, enforce to use default value and override the option in task YAML. Default: medium
--use-spot / --no-use-spot
Whether to request spot instances. If specified, overrides the ‘resources.use_spot’ config.
--image-id TEXT
Custom image id for launching the instances. Passing ‘none’ resets the config.
--env-file DOTENV_VALUES
Path to a dotenv file with environment variables to set on the remote node.
--env _PARSE_ENV_VAR
Environment variable to set on the remote node. It can be specified multiple times. Examples: 1.
--env MY_ENV=1
: set$MY_ENV
on the cluster to be 1. 2.--env MY_ENV2=$HOME
: set$MY_ENV2
on the cluster to be the same value of$HOME
in the local environment where the CLI command is run. 3.--env MY_ENV3
: set$MY_ENV3
on the cluster to be the same value of$MY_ENV3
in the local environment.--gpus TEXT
Type and number of GPUs to use. Example values: ‘V100:8’, ‘V100’ (short for a count of 1), or ‘V100:0.5’ (fractional counts are supported by the scheduling framework). If a new cluster is being launched by this command, this is the resources to provision. If an existing cluster is being reused, this is seen as the task demand, which must fit the cluster’s total resources and is used for scheduling the task. Overrides the ‘accelerators’ config in the YAML if both are supplied. Passing ‘none’ resets the config.
-t
,--instance-type TEXT
The instance type to use. If specified, overrides the ‘resources.instance_type’ config. Passing ‘none’ resets the config.
--ports TEXT
Ports to open on the cluster. If specified, overrides the ‘ports’ config in the YAML.
-i
,--idle-minutes-to-autostop INTEGER
Automatically stop the cluster after this many minutes of idleness, i.e., no running or pending jobs in the cluster’s job queue. Idleness gets reset whenever setting-up/running/pending jobs are found in the job queue. Setting this flag is equivalent to running
sky launch -d ...
and thensky autostop -i <minutes>
. If not set, the cluster will not be autostopped.--down
Autodown the cluster: tear down the cluster after all jobs finish (successfully or abnormally). If –idle-minutes-to-autostop is also set, the cluster will be torn down after the specified idle time. Note that if errors occur during provisioning/data syncing/setting up, the cluster will not be torn down for debugging purposes.
-r
,--retry-until-up
Whether to retry provisioning infinitely until the cluster is up, if we fail to launch the cluster on any possible region/cloud due to unavailability errors.
-y
,--yes
Skip confirmation prompt.
--no-setup
Skip setup phase when (re-)launching cluster.
--clone-disk-from
,--clone TEXT
[Experimental] Clone disk from an existing cluster to launch a new one. This is useful when the new cluster needs to have the same data on the boot disk as an existing cluster.
-h
,--help
Show help for the command..
Usage
d3x sky exec [OPTIONS] CLUSTER ENTRYPOINT...
d3x sky exec --help
Function
Execute a task or command on an existing cluster. If ENTRYPOINT points to a valid YAML file, it is read in as the task specification. Otherwise, it is interpreted as a bash command.
Options
Option
Description
-d
,--detach-run
If True, as soon as a job is submitted, return from this call and do not stream execution logs.
-n
,--name TEXT
Task name. Overrides the ‘name’ config in the YAML if both are supplied.
--workdir DIRECTORY
If specified, sync this dir to the remote working directory, where the task will be invoked. Overrides the ‘workdir’ config in the YAML if both are supplied.
--cloud TEXT
The cloud to use. If specified, overrides the ‘resources.cloud’ config. Passing ‘none’ resets the config.
--region TEXT
The region to use. If specified, overrides the ‘resources.region’ config. Passing ‘none’ resets the config.
--zone TEXT
The zone to use. If specified, overrides the ‘resources.zone’ config. Passing ‘none’ resets the config.
--num-nodes INTEGER
Number of nodes to execute the task on. Overrides the ‘num_nodes’ config in the YAML if both are supplied.
--cpus TEXT
Number of vCPUs each instance must have (e.g.,
--cpus=4
(exactly 4) or--cpus=4+
(at least 4)). This is used to automatically select the instance type.--memory TEXT
Amount of memory each instance must have in GB (e.g.,
--memory=16
(exactly 16GB),--memory=16+
(at least 16GB))--disk-size INTEGER
OS disk size in GBs.
--disk-tier [low|medium|high|best|none]
OS disk tier. Could be one of low, medium, high, best, none. If best is specified, use the best possible disk tier. If none is specified, enforce to use default value and override the option in task YAML. Default: medium
--use-spot / --no-use-spot
Whether to request spot instances. If specified, overrides the ‘resources.use_spot’ config.
--image-id TEXT
Custom image id for launching the instances. Passing ‘none’ resets the config.
--env-file DOTENV_VALUES
Path to a dotenv file with environment variables to set on the remote node.
--env _PARSE_ENV_VAR
Environment variable to set on the remote node. It can be specified multiple times. Examples: 1.
--env MY_ENV=1
: set$MY_ENV
on the cluster to be 1. 2.--env MY_ENV2=$HOME
: set$MY_ENV2
on the cluster to be the same value of$HOME
in the local environment where the CLI command is run. 3.--env MY_ENV3
: set$MY_ENV3
on the cluster to be the same value of$MY_ENV3
in the local environment.--gpus TEXT
Type and number of GPUs to use. Example values: ‘V100:8’, ‘V100’ (short for a count of 1), or ‘V100:0.5’ (fractional counts are supported by the scheduling framework). If a new cluster is being launched by this command, this is the resources to provision. If an existing cluster is being reused, this is seen as the task demand, which must fit the cluster’s total resources and is used for scheduling the task. Overrides the ‘accelerators’ config in the YAML if both are supplied. Passing ‘none’ resets the config.
-t
,--instance-type TEXT
The instance type to use. If specified, overrides the ‘resources.instance_type’ config. Passing ‘none’ resets the config.
--ports TEXT
Ports to open on the cluster. If specified, overrides the ‘ports’ config in the YAML.
-h
,--help
Show help for the command..
Usage
d3x sky status [OPTIONS] [CLUSTERS]...
d3x sky status --help
Function
Show sky clusters. If CLUSTERS is given, show those clusters. Otherwise, show all clusters. If –ip is specified, show the IP address of the head node of the cluster. Only available when CLUSTERS contains exactly one cluster, e.g.
sky status --ip mycluster
. If –endpoints is specified, show all exposed endpoints in the cluster. Only available when CLUSTERS contains exactly one cluster, e.g.sky status --endpoints mycluster
. To query a single endpoint, you can usesky status mycluster --endpoint 8888
.Options
Option
Description
-a
,--all
Show all information in full.
-r
,--refresh
Query the latest cluster statuses from the cloud provider(s).
--ip
Get the IP address of the head node of a cluster. This option will override all other options. For Kubernetes clusters, the returned IP address is the internal IP of the head pod, and may not be accessible from outside the cluster.
--endpoints
Get all exposed endpoints and corresponding URLs for acluster. This option will override all other options.
--endpoint INTEGER
Get the endpoint URL for the specified port number on the cluster. This option will override all other options.
--show-spot-jobs / --no-show-spot-jobs
Also show recent in-progress spot jobs, if any.
--show-services / --no-show-services
Also show sky serve services, if any.
-h
,--help
Show help for the command..
Usage
d3x sky cost-report [OPTIONS]
d3x sky cost-report --help
Function
Show estimated costs for launched Skypilot clusters.
Options
Option
Description
-a
,--all
Show all information in full.
-h
,--help
Show help for the command..
Usage
d3x sky queue [OPTIONS] [CLUSTERS]...
d3x sky queue --help
Function
Show the job queue for cluster(s).
Options
Option
Description
-a
,--all-users
Show all users’ information in full.
-s
,--skip-finished
Show only pending/running jobs’ information.
-h
,--help
Show help for the command..
Usage
d3x sky logs [OPTIONS] CLUSTER [JOB_IDS]...
d3x sky logs --help
Function
Tail the log of a job.
Options
Option
Description
-s
,--sync-down
Sync down the logs of a job to the local machine. For a distributed job, a separate log file from each worker will be downloaded.
--status
If specified, do not show logs but exit with a status code for the job’s status: 0 for succeeded, or 1 for all other statuses.
--follow / --no-follow
Follow the logs of a job. If –no-follow is specified, print the log so far and exit. [default: –follow]
-h
,--help
Show help for the command..
Usage
d3x sky cancel [OPTIONS] CLUSTER [JOBS]...
d3x sky cancel --help
Function
Cancel Skypilot job(s).
Options
Option
Description
-a
,--all
Cancel all jobs on the specified cluster.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky stop [OPTIONS] [CLUSTERS]...
d3x sky stop --help
Function
Stop Skypilot cluster(s).
Options
Option
Description
-a
,--all
Stop all existing clusters.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky autostop [OPTIONS] [CLUSTERS]...
d3x sky autostop --help
Function
Schedule an autostop or autodown for cluster(s). Autostop/autodown will automatically stop or teardown a cluster when it becomes idle for a specified duration. Idleness means there are no in-progress (pending/running) jobs in a cluster’s job queue.
Options
Option
Description
-a
,--all
Apply this command to all existing clusters.
-i
,--idle-minutes INTEGER
Set the idle minutes before autostopping the cluster. See the doc above for detailed semantics.
--cancel
Cancel any currently active auto{stop,down} setting for the cluster. No-op if there is no active setting.
--down
Use autodown (tear down the cluster; non-restartable), instead of autostop (restartable).
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky start [OPTIONS] [CLUSTERS]...
d3x sky start --help
Function
Restart Skypilot cluster(s). If a cluster is previously stopped (status is STOPPED) or failed in provisioning/runtime installation (status is INIT), this command will attempt to start the cluster. In the latter case, provisioning and runtime installation will be retried.
Options
Option
Description
-a
,--all
Start all existing clusters.
-y
,--yes
Skip confirmation prompt.
-i
,--idle-minutes-to-autostop INTEGER
Automatically stop the cluster after this many minutes of idleness, i.e., no running or pending jobs in the cluster’s job queue. Idleness gets reset whenever setting-up/running/pending jobs are found in the job queue. Setting this flag is equivalent to running
sky launch -d ...
and thensky autostop -i <minutes>
. If not set, the cluster will not be autostopped.--down
Autodown the cluster: tear down the cluster after specified minutes of idle time after all jobs finish (successfully or abnormally). Requires –idle-minutes-to-autostop to be set.
-r
,--retry-until-up
Retry provisioning infinitely until the cluster is up, if we fail to start the cluster due to unavailability errors.
-f
,--force
Force start the cluster even if it is already UP. Useful for upgrading the SkyPilot runtime on the cluster.
-h
,--help
Show help for the command..
Usage
d3x sky down [OPTIONS] [CLUSTERS]...
d3x sky down --help
Function
Tear down Skypilot cluster(s).
Options
Option
Description
-a
,--all
Tear down all existing clusters.
-y
,--yes
Skip confirmation prompt.
-p
,--purge
Ignore cloud provider errors (if any). Useful for cleaning up manually deleted cluster(s).
-h
,--help
Show help for the command..
Usage
d3x sky check [OPTIONS]
d3x sky check --help
Function
Check which clouds are available to use by Skypilot.
Options
Option
Description
-v
,--verbose
Show the activated account for each cloud.
-h
,--help
Show help for the command..
Usage
d3x sky show-gpus [OPTIONS] [ACCELERATOR_STR]
d3x sky show-gpus --help
Function
Show GPU/TPU/accelerators supported by Skypilot and their prices.
Options
Option
Description
-a
,--all
Show details of all GPU/TPU/accelerator offerings.
--cloud TEXT
Cloud provider to query.
--region TEXT
The region to use. If not specified, shows accelerators from all regions.
--all-regions
Show pricing and instance details for a specified accelerator across all regions and clouds.
-h
,--help
Show help for the command..
Usage
d3x sky storage [OPTIONS] COMMAND [ARGS]...
d3x sky storage --help
Function
Commands for using SkyPilot storage CLI.
Options
Option
Description
-h
,--help
Show help for the command..
Commands
Command
Description
ls
List storage objects managed by SkyPilot.
delete
Delete storage objects.
Usage
d3x sky storage ls [OPTIONS]
d3x sky storage ls --help
Function
List storage objects managed by SkyPilot.
Options
Option
Description
-a
,--all
Show all information in full.
-h
,--help
Show help for the command..
Usage
d3x sky storage delete [OPTIONS] [NAMES]...
d3x sky storage delete --help
Function
Delete storage objects in Skypilot storage.
Options
Option
Description
-a
,--all
Delete all storage objects.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky bench [OPTIONS] COMMAND [ARGS]...
d3x sky bench --help
Function
Commands for using SkyPilot Benchmark CLI.
Options
Option
Description
-h
,--help
Show help for the command..
Commands
Command
Description
launch
Benchmark a task on different resources.
ls
List the benchmark history.
show
Show a benchmark report.
down
Tear down all clusters belonging to a benchmark.
delete
Delete benchmark reports from the history.
Usage
d3x sky bench launch [OPTIONS] ENTRYPOINT...
d3x sky bench launch --help
Function
Benchmark a task on different resources.
Options
Option
Description
-b
,--benchmark TEXT
Benchmark name. [required]
-n
,--name TEXT
Task name. Overrides the ‘name’ config in the YAML if both are supplied.
--workdir DIRECTORY
If specified, sync this dir to the remote working directory, where the task will be invoked. Overrides the ‘workdir’ config in the YAML if both are supplied.
--cloud TEXT
The cloud to use. If specified, overrides the ‘resources.cloud’ config. Passing ‘none’ resets the config.
--region TEXT
The region to use. If specified, overrides the ‘resources.region’ config. Passing ‘none’ resets the config.
--zone TEXT
The zone to use. If specified, overrides the ‘resources.zone’ config. Passing ‘none’ resets the config.
--num-nodes INTEGER
Number of nodes to execute the task on. Overrides the ‘num_nodes’ config in the YAML if both are supplied.
--cpus TEXT
Number of vCPUs each instance must have (e.g.,
--cpus=4
(exactly 4) or--cpus=4+
(at least 4)). This is used to automatically select the instance type.--memory TEXT
Amount of memory each instance must have in GB (e.g.,
--memory=16
(exactly 16GB),--memory=16+
(at least 16GB))--disk-size INTEGER
OS disk size in GBs.
--disk-tier [low|medium|high|best|none]
OS disk tier. Could be one of low, medium, high, best, none. If best is specified, use the best possible disk tier. If none is specified, enforce to use default value and override the option in task YAML. Default: medium
--use-spot / --no-use-spot
Whether to request spot instances. If specified, overrides the ‘resources.use_spot’ config.
--image-id TEXT
Custom image id for launching the instances. Passing ‘none’ resets the config.
--env-file DOTENV_VALUES
Path to a dotenv file with environment variables to set on the remote node.
--env _PARSE_ENV_VAR
Environment variable to set on the remote node. It can be specified multiple times. Examples: 1.
--env MY_ENV=1
: set$MY_ENV
on the cluster to be 1. 2.--env MY_ENV2=$HOME
: set$MY_ENV2
on the cluster to be the same value of$HOME
in the local environment where the CLI command is run. 3.--env MY_ENV3
: set$MY_ENV3
on the cluster to be the same value of$MY_ENV3
in the local environment.--gpus TEXT
Comma-separated list of GPUs to run benchmark on. Example values: ‘T4:4,V100:8’ (without blank spaces).
--ports TEXT
Ports to open on the cluster. If specified, overrides the ‘ports’ config in the YAML.
-i
,--idle-minutes-to-autostop INTEGER
Automatically stop the cluster after this many minutes of idleness after setup/file_mounts. This is equivalent to running sky launch -d … and then sky autostop -i <minutes>. If not set, the cluster will not be autostopped.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky bench ls [OPTIONS]
d3x sky bench ls --help
Function
List the benchmark history.
Options
Option
Description
-h
,--help
Show help for the command..
Usage
d3x sky bench show [OPTIONS] BENCHMARK
d3x sky bench show --help
Function
Show a benchmark report.
Options
Option
Description
-h
,--help
Show help for the command..
Usage
d3x sky bench down [OPTIONS] BENCHMARK
d3x sky bench down --help
Function
Tear down all clusters belonging to a benchmark.
Options
Option
Description
-e
,--exclude TEXT
Cluster name(s) to exclude from termination. Typically, you might want to see the benchmark results in sky bench show and exclude a ‘winner’ cluster from termination to finish the running task.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky bench delete [OPTIONS] [BENCHMARKS]...
d3x sky bench delete --help
Function
Delete benchmark reports from the history.
Options
Option
Description
-a
,--all
Delete all benchmark reports from the history.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky spot [OPTIONS] COMMAND [ARGS]...
d3x sky spot --help
Function
Commands for using managed Spot CLI (spot instances with auto-recovery).
Options
Option
Description
-h
,--help
Show help for the command..
Commands
Command
Description
launch
Launch a managed spot job from a YAML or a command.
queue
Show statuses of managed spot jobs.
cancel
Cancel managed spot jobs.
logs
Tail the log of a managed spot job.
dashboard
Opens a dashboard for spot jobs (needs controller to be UP).
Usage
d3x sky spot launch [OPTIONS] ENTRYPOINT...
d3x sky spot launch --help
Function
Launch a managed spot job from a YAML or a command.
Options
Option
Description
-n
,--name TEXT
Task name. Overrides the ‘name’ config in the YAML if both are supplied.
--workdir DIRECTORY
If specified, sync this dir to the remote working directory, where the task will be invoked. Overrides the ‘workdir’ config in the YAML if both are supplied.
--cloud TEXT
The cloud to use. If specified, overrides the ‘resources.cloud’ config. Passing ‘none’ resets the config.
--region TEXT
The region to use. If specified, overrides the ‘resources.region’ config. Passing ‘none’ resets the config.
--zone TEXT
The zone to use. If specified, overrides the ‘resources.zone’ config. Passing ‘none’ resets the config.
--num-nodes INTEGER
Number of nodes to execute the task on. Overrides the ‘num_nodes’ config in the YAML if both are supplied.
--cpus TEXT
Number of vCPUs each instance must have (e.g.,
--cpus=4
(exactly 4) or--cpus=4+
(at least 4)). This is used to automatically select the instance type.--memory TEXT
Amount of memory each instance must have in GB (e.g.,
--memory=16
(exactly 16GB),--memory=16+
(at least 16GB))--disk-size INTEGER
OS disk size in GBs.
--disk-tier [low|medium|high|best|none]
OS disk tier. Could be one of low, medium, high, best, none. If best is specified, use the best possible disk tier. If none is specified, enforce to use default value and override the option in task YAML. Default: medium
--use-spot / --no-use-spot
Whether to request spot instances. If specified, overrides the ‘resources.use_spot’ config.
--image-id TEXT
Custom image id for launching the instances. Passing ‘none’ resets the config.
--env-file DOTENV_VALUES
Path to a dotenv file with environment variables to set on the remote node.
--env _PARSE_ENV_VAR
Environment variable to set on the remote node. It can be specified multiple times. Examples: 1.
--env MY_ENV=1
: set$MY_ENV
on the cluster to be 1. 2.--env MY_ENV2=$HOME
: set$MY_ENV2
on the cluster to be the same value of$HOME
in the local environment where the CLI command is run. 3.--env MY_ENV3
: set$MY_ENV3
on the cluster to be the same value of$MY_ENV3
in the local environment.--gpus TEXT
Type and number of GPUs to use. Example values: ‘V100:8’, ‘V100’ (short for a count of 1), or ‘V100:0.5’ (fractional counts are supported by the scheduling framework). If a new cluster is being launched by this command, this is the resources to provision. If an existing cluster is being reused, this is seen as the task demand, which must fit the cluster’s total resources and is used for scheduling the task. Overrides the ‘accelerators’ config in the YAML if both are supplied. Passing ‘none’ resets the config.
-t
,--instance-type TEXT
The instance type to use. If specified, overrides the ‘resources.instance_type’ config. Passing ‘none’ resets the config.
--ports TEXT
Ports to open on the cluster. If specified, overrides the ‘ports’ config in the YAML.
--spot-recovery TEXT
Spot recovery strategy to use for the managed spot task.
-d
,--detach-run
If True, as soon as a job is submitted, return from this call and do not stream execution logs.
-r
,--retry-until-up / -no-retry-until-up
(Default: True; this flag is deprecated and will be removed in a future release.) Whether to retry provisioning infinitely until the cluster is up, if unavailability errors are encountered. This applies to launching the spot clusters (both the initial and any recovery attempts), not the spot controller.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky spot queue [OPTIONS]
d3x sky spot queue --help
Function
Show statuses of managed spot jobs.
Options
Option
Description
-a
,--all
Show all information in full.
-r
,--refresh
Query the latest statuses, restarting the spot controller if stopped.
-s
,--skip-finished
Show only pending/running jobs’ information.
-h
,--help
Show help for the command..
Usage
d3x sky spot cancel [OPTIONS] [JOB_IDS]...
d3x sky spot cancel --help
Function
Cancel managed spot jobs.
Options
Option
Description
-n
,--name TEXT
Managed spot job name to cancel.
-a
,--all
Cancel all managed spot jobs.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky spot logs [OPTIONS] [JOB_ID]
d3x sky spot logs --help
Function
Tail the log of a managed spot job.
Options
Option
Description
-n
,--name TEXT
Managed spot job name.
--follow / --no-follow
Follow the logs of the job. [default: –follow] If –no-follow is specified, print the log so far and exit.
--controller
Show the controller logs of this job; useful for debugging launching/recoveries, etc.
-h
,--help
Show help for the command..
Usage
d3x sky spot dashboard [OPTIONS]
d3x sky spot dashboard --help
Function
Opens a dashboard for spot jobs (needs controller to be UP).
Options
Option
Description
-p
,--port INTEGER
Local port to use for the dashboard. If None, a free port is automatically chosen.
-h
,--help
Show help for the command..
Usage
d3x sky serve [OPTIONS] COMMAND [ARGS]...
d3x sky serve --help
Function
Commands for using SkyServe CLI (multi-region, multi-cloud serving).
Options
Option
Description
-h
,--help
Show help for the command..
Commands
Command
Description
up
Launch a SkyServe service.
update
Update a SkyServe service.
status
Show statuses of SkyServe services.
down
Teardown service(s).
logs
Tail the log of a service.
Usage
d3x sky serve up [OPTIONS] SERVICE_YAML...
d3x sky serve up --help
Function
Launch a SkyServe service.
Options
Option
Description
-n
,--service-name TEXT
A service name. Unique for each service. If not provided, a unique name is autogenerated.
--workdir DIRECTORY
If specified, sync this dir to the remote working directory, where the task will be invoked. Overrides the ‘workdir’ config in the YAML if both are supplied.
--cloud TEXT
The cloud to use. If specified, overrides the ‘resources.cloud’ config. Passing ‘none’ resets the config.
--region TEXT
The region to use. If specified, overrides the ‘resources.region’ config. Passing ‘none’ resets the config.
--zone TEXT
The zone to use. If specified, overrides the ‘resources.zone’ config. Passing ‘none’ resets the config.
--num-nodes INTEGER
Number of nodes to execute the task on. Overrides the ‘num_nodes’ config in the YAML if both are supplied.
--cpus TEXT
Number of vCPUs each instance must have (e.g.,
--cpus=4
(exactly 4) or--cpus=4+
(at least 4)). This is used to automatically select the instance type.--memory TEXT
Amount of memory each instance must have in GB (e.g.,
--memory=16
(exactly 16GB),--memory=16+
(at least 16GB))--disk-size INTEGER
OS disk size in GBs.
--disk-tier [low|medium|high|best|none]
OS disk tier. Could be one of low, medium, high, best, none. If best is specified, use the best possible disk tier. If none is specified, enforce to use default value and override the option in task YAML. Default: medium
--use-spot / --no-use-spot
Whether to request spot instances. If specified, overrides the ‘resources.use_spot’ config.
--image-id TEXT
Custom image id for launching the instances. Passing ‘none’ resets the config.
--env-file DOTENV_VALUES
Path to a dotenv file with environment variables to set on the remote node.
--env _PARSE_ENV_VAR
Environment variable to set on the remote node. It can be specified multiple times. Examples: 1.
--env MY_ENV=1
: set$MY_ENV
on the cluster to be 1. 2.--env MY_ENV2=$HOME
: set$MY_ENV2
on the cluster to be the same value of$HOME
in the local environment where the CLI command is run. 3.--env MY_ENV3
: set$MY_ENV3
on the cluster to be the same value of$MY_ENV3
in the local environment.--gpus TEXT
Type and number of GPUs to use. Example values: ‘V100:8’, ‘V100’ (short for a count of 1), or ‘V100:0.5’ (fractional counts are supported by the scheduling framework). If a new cluster is being launched by this command, this is the resources to provision. If an existing cluster is being reused, this is seen as the task demand, which must fit the cluster’s total resources and is used for scheduling the task. Overrides the ‘accelerators’ config in the YAML if both are supplied. Passing ‘none’ resets the config.
-t
,--instance-type TEXT
The instance type to use. If specified, overrides the ‘resources.instance_type’ config. Passing ‘none’ resets the config.
--ports TEXT
Ports to open on the cluster. If specified, overrides the ‘ports’ config in the YAML.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky serve update [OPTIONS] SERVICE_NAME SERVICE_YAML...
d3x sky serve update --help
Function
Update a SkyServe service.
Options
Option
Description
--workdir DIRECTORY
If specified, sync this dir to the remote working directory, where the task will be invoked. Overrides the ‘workdir’ config in the YAML if both are supplied.
--cloud TEXT
The cloud to use. If specified, overrides the ‘resources.cloud’ config. Passing ‘none’ resets the config.
--region TEXT
The region to use. If specified, overrides the ‘resources.region’ config. Passing ‘none’ resets the config.
--zone TEXT
The zone to use. If specified, overrides the ‘resources.zone’ config. Passing ‘none’ resets the config.
--num-nodes INTEGER
Number of nodes to execute the task on. Overrides the ‘num_nodes’ config in the YAML if both are supplied.
--cpus TEXT
Number of vCPUs each instance must have (e.g.,
--cpus=4
(exactly 4) or--cpus=4+
(at least 4)). This is used to automatically select the instance type.--memory TEXT
Amount of memory each instance must have in GB (e.g.,
--memory=16
(exactly 16GB),--memory=16+
(at least 16GB))--disk-size INTEGER
OS disk size in GBs.
--disk-tier [low|medium|high|best|none]
OS disk tier. Could be one of low, medium, high, best, none. If best is specified, use the best possible disk tier. If none is specified, enforce to use default value and override the option in task YAML. Default: medium
--use-spot / --no-use-spot
Whether to request spot instances. If specified, overrides the ‘resources.use_spot’ config.
--image-id TEXT
Custom image id for launching the instances. Passing ‘none’ resets the config.
--env-file DOTENV_VALUES
Path to a dotenv file with environment variables to set on the remote node.
--env _PARSE_ENV_VAR
Environment variable to set on the remote node. It can be specified multiple times. Examples: 1.
--env MY_ENV=1
: set$MY_ENV
on the cluster to be 1. 2.--env MY_ENV2=$HOME
: set$MY_ENV2
on the cluster to be the same value of$HOME
in the local environment where the CLI command is run. 3.--env MY_ENV3
: set$MY_ENV3
on the cluster to be the same value of$MY_ENV3
in the local environment.--gpus TEXT
Type and number of GPUs to use. Example values: ‘V100:8’, ‘V100’ (short for a count of 1), or ‘V100:0.5’ (fractional counts are supported by the scheduling framework). If a new cluster is being launched by this command, this is the resources to provision. If an existing cluster is being reused, this is seen as the task demand, which must fit the cluster’s total resources and is used for scheduling the task. Overrides the ‘accelerators’ config in the YAML if both are supplied. Passing ‘none’ resets the config.
-t
,--instance-type TEXT
The instance type to use. If specified, overrides the ‘resources.instance_type’ config. Passing ‘none’ resets the config.
--ports TEXT
Ports to open on the cluster. If specified, overrides the ‘ports’ config in the YAML.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky serve status [OPTIONS] [SERVICE_NAMES]...
d3x sky serve status --help
Function
Show statuses of SkyServe services.
Options
Option
Description
-a
,--all
Show all information in full.
--endpoint
Show service endpoint.
-h
,--help
Show help for the command..
Usage
d3x sky serve down [OPTIONS] [SERVICE_NAMES]...
d3x sky serve down --help
Function
Teardown service(s).
Options
Option
Description
-a
,--all
Tear down all services.
-p
,--purge
Tear down services in failed status.
-y
,--yes
Skip confirmation prompt.
-h
,--help
Show help for the command..
Usage
d3x sky serve logs [OPTIONS] SERVICE_NAME [REPLICA_ID]
d3x sky serve logs --help
Function
Tail the log of a service.
Options
Option
Description
--follow / --no-follow
Follow the logs of the job. [default: –follow] If –no-follow is specified, print the log so far and exit.
--controller
Show the controller logs of this service.
--load-balancer
Show the load balancer logs of this service.
-h
,--help
Show help for the command..
d3x ssh
¶
Usage
d3x ssh [OPTIONS] [COMMAND]...
d3x ssh --help
Function
SSH to DKubeX workspace.
Options
Option
Description
-w TEXT
workspace name
--help
Show help for the command..
d3x ssh-copy-id
¶
Usage
d3x ssh-copy-id [OPTIONS] PUBLIC_KEY
d3x ssh-copy-id --help
Function
Copy ssh public key to DKubeX user’s home.
Options
Option
Description
--help
Show help for the command..
d3x start
¶
Usage
d3x start [OPTIONS]
d3x start --help
Function
Start given workspace.
Options
Option
Description
--help
Show help for the command..
d3x start-app
¶
Usage
d3x start-app [OPTIONS] APP_NAME
d3x start-app --help
Function
Start given workspace app.
Options
Option
Description
--help
Show help for the command..
d3x stop
¶
Usage
d3x stop [OPTIONS]
d3x stop --help
Function
Stop given workspace.
Options
Option
Description
--help
Show help for the command..
d3x stop-app
¶
Usage
d3x stop-app [OPTIONS] APP_NAME
d3x stop-app --help
Function
Stop given workspace App.
Options
Option
Description
--help
Show help for the command..
d3x tensorboard
¶
Usage
d3x tensorboard [OPTIONS] COMMAND [ARGS]...
d3x tensorboard --help
Function
Commands regarding using Tensorboard.
Options
Option
Description
--help
Show help for the command..
Commands
Command
Description
get-logdir
Getting log directory for tensorboard.
set-logdir
Setting log directory for tensorboard app inside workspace.
Usage
d3x tensorboard get-logdir [OPTIONS]
d3x tensorboard get-logdir --help
Function
Getting log directory for tensorboard.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
Usage
d3x tensorboard set-logdir [OPTIONS] LOGDIR
d3x tensorboard set-logdir --help
Function
Setting log directory for tensorboard app inside workspace.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x top
¶
Usage
d3x top [OPTIONS] [WORKSPACE]
d3x top --help
Function
Show workspace resource usage.
Options
Option
Description
--help
Show help for the command..
d3x umount
¶
Usage
d3x umount [OPTIONS] MOUNTPATH
d3x umount --help
Function
Unmount bucket previously mounted using
d3x mount
command.Options
Option
Description
--help
Show help for the command..
d3x version
¶
Usage
d3x version [OPTIONS]
d3x version --help
Function
Show DKubeX version.
Options
Option
Description
-o
,--output [yaml|json]
supported only json,yaml
--help
Show help for the command..
d3x vnc
¶
Usage
d3x vnc [OPTIONS] WORKSPACE_NAME
d3x vnc --help
Function
VNC connection to DKubeX workspace.
Options
Option
Description
--help
Show help for the command..