跳到主要内容

Trisol Command: trisol

bohr trisol forwards arguments to the Trisol training-and-inference platform CLI.

bohr trisol [args...]

Commands

CommandDescription
login / logoutAuthenticate via Vouch unified identity (browser or device code) / remove the local session
initInteractively configure the API URL, login, and default team
whoamiShow the active identity, API URL, team, and credential source
modelBrowse, create, update, and import models
trainManage training jobs (submit SFT runs, list, inspect)
datasetManage SFT datasets (create, upload, list)
imageBrowse and manage container images in the catalog
inferenceBrowse inference services and runtime status
frameworkBrowse framework builds (read-only)
quotaBrowse GPU quota allocations
usageQuery GPU resource usage (card-hour accounting)
leaderboardRank users by training jobs or inference services in a period
preflightCheck whether a deploy intent can be satisfied (advisory)
configManage the CLI configuration (~/.trisol/config.yaml)
api-keysManage inference API keys (issue / list / revoke)
teamBrowse and create teams
userUser lookup and management
monitorPlatform monitoring views (mostly platform admin)
notifyInbox: list and acknowledge notifications
direct-accessManage business-line direct kubectl access grants
skillManage the Trisol agent skill embedded in the binary
updateDownload and install the latest CLI binary from the Trisol server
completionGenerate shell completion scripts
versionPrint the CLI version, commit, and build date

Run bohr trisol <command> --help for details of each command.

Examples

bohr trisol --help
bohr trisol init
bohr trisol whoami
bohr trisol version
bohr extension install trisol

Trisol requires account entitlement and a locally installed companion executable. The command may be hidden when the entitlement is unavailable. Install the executable with bohr extension install trisol; actual simulation commands and billing behavior depend on the installed Trisol version.

Enterprise scenarios

Trisol serves enterprise (B-side) R&D scenarios, from dedicated model access to model training and inference integrated into team pipelines. If you are interested, contact bohrium@dp.tech.