Huggingface.js documentation
Interface: QuantizationConfig
Interface: QuantizationConfig
Hierarchy
QuantizationConfig
Properties
bits
• Optional bits: number
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:695
config _ groups
• Optional config_groups: Record<string, { format?: string ; targets?: string[] ; weights?: { num_bits?: number } }>
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:700
expert _ dtype
• Optional expert_dtype: string
Routed expert precision when it differs from the main quantizer (e.g. FP4 experts with FP8 attention).
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:690
format
• Optional format: string
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:699
group _ size
• Optional group_size: number
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:693
ignore
• Optional ignore: string[]
compressed-tensors names its exclusion list ignore rather than modules_to_not_convert,
using the same re:-prefixed target syntax as config_groups[].targets.
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:705
load _ in _ 4bit
• Optional load_in_4bit: boolean
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:696
load _ in _ 8bit
• Optional load_in_8bit: boolean
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:697
mode
• Optional mode: string
MLX quantization mode (e.g. affine); MLX configs do not declare quant_method.
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:692
modules _ to _ not _ convert
• Optional modules_to_not_convert: string[]
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:694
quant _ method
• Optional quant_method: string
Defined in
packages/hub/src/lib/parse-safetensors-metadata.ts:688
Update on GitHub