Overview > klyn.llm.inference.hub > HubClient > downloadGGUF
Version 0.2.1 Overview Index Tree Deprecated
klyn.llm.inference.hub.HubClient.downloadGGUF
method
public static downloadGGUF(repository as String, quantization as ModelQuantization, revision as String = "main", offline as Boolean = false, token as String = "", verbose as Boolean = true) as HubSnapshot
Description

Downloads only the selected GGUF file group, using pinned revision URLs, bounded resumable transfers, byte progress and SHA-256 verification. Cache manifests are variant-specific; completed hits perform no weight scan. Currently executable formats are checked before any large download.

Parameters
ParameterDescription
repositoryHub owner/name.
quantizationExact variant, or AUTO for an unambiguous Q8_0 group.
revisionBranch, tag or commit.
offlineRequire an already complete local snapshot.
tokenOptional private-repository token.
verboseDisplay download progress.
Returns

Snapshot containing the selected weights only.

Throws
  • ValueException for unsupported or ambiguous variants.
  • IOException for missing assets, integrity or network errors.
Example

snapshot = HubClient.downloadGGUF("owner/model-GGUF", ModelQuantization.Q8_0)