IModelParams
Namespace: LLama.Abstractions
The parameters for initializing a LLama model.
public interface IModelParams
Properties
MainGpu
main_gpu interpretation depends on split_mode: NoneThe GPU that is used for the entire mode.RowThe GPU that is used for small tensors and intermediate results.LayerIgnored.
public abstract int MainGpu { get; set; }
Property Value
SplitMode
How to split the model across multiple GPUs
public abstract GPUSplitMode SplitMode { get; }
Property Value
GpuLayerCount
Number of layers to run in VRAM / GPU memory (n_gpu_layers)
public abstract int GpuLayerCount { get; }
Property Value
UseMemorymap
Use mmap for faster loads (use_mmap)
public abstract bool UseMemorymap { get; }
Property Value
UseMemoryLock
Use mlock to keep model in memory (use_mlock)
public abstract bool UseMemoryLock { get; }
Property Value
ModelPath
Model path (model)
public abstract string ModelPath { get; }
Property Value
TensorSplits
how split tensors should be distributed across GPUs
public abstract TensorSplitsCollection TensorSplits { get; }
Property Value
VocabOnly
Load vocab only (no weights)
public abstract bool VocabOnly { get; }
Property Value
LoraAdapters
List of LoRA adapters to apply
public abstract AdapterCollection LoraAdapters { get; }
Property Value
LoraBase
base model path for the lora adapter (lora_base)
public abstract string LoraBase { get; }
Property Value
MetadataOverrides
Override specific metadata items in the model
public abstract List<MetadataOverride> MetadataOverrides { get; }