Class: LLaMACpp::ModelParams
- Inherits:
-
Object
- Object
- LLaMACpp::ModelParams
- Defined in:
- ext/llama_cpp/dummy.rb
Overview
Class for model parameters.
Instance Method Summary collapse
-
#check_tensors ⇒ Boolean
Returns the flag to validate model tensor data.
-
#check_tensors=(flag) ⇒ Object
Sets the flag to validate model tensor data.
-
#main_gpu ⇒ Integer
Returns the GPU that is used.
-
#main_gpu=(main_gpu) ⇒ Object
Sets the GPU that is used.
-
#n_gpu_layers ⇒ Integer
Returns the number of layers to stor in VRAM.
-
#n_gpu_layers=(n_gpu_layers) ⇒ Object
Sets the number of layers to stor in VRAM.
-
#split_mode ⇒ Integer
Return the split mode.
-
#split_mode=(split_mode) ⇒ Object
Sets the split mode.
-
#tensor_split ⇒ Array<Float>
Returns the how layers are split across multi-GPUs.
-
#use_mlock ⇒ Boolean
Returns the flag to force system to keep model in memory.
-
#use_mlock=(flag) ⇒ Object
Sets the flag to force system to keep model in memory.
-
#use_mmap ⇒ Boolean
Returns the flag to use mmap.
-
#use_mmap=(flag) ⇒ Object
Sets the flag to use mmap.
-
#vocab_only ⇒ Boolean
Returns the flag to load only vocabulary.
-
#vocab_only=(flag) ⇒ Object
Sets the flag to load only vocabulary.
Instance Method Details
#check_tensors ⇒ Boolean
Returns the flag to validate model tensor data.
721 |
# File 'ext/llama_cpp/dummy.rb', line 721 def check_tensors; end |
#check_tensors=(flag) ⇒ Object
Sets the flag to validate model tensor data.
717 |
# File 'ext/llama_cpp/dummy.rb', line 717 def check_tensors=(flag); end |
#main_gpu ⇒ Integer
Returns the GPU that is used.
685 |
# File 'ext/llama_cpp/dummy.rb', line 685 def main_gpu; end |
#main_gpu=(main_gpu) ⇒ Object
Sets the GPU that is used.
681 |
# File 'ext/llama_cpp/dummy.rb', line 681 def main_gpu=(main_gpu); end |
#n_gpu_layers ⇒ Integer
Returns the number of layers to stor in VRAM.
669 |
# File 'ext/llama_cpp/dummy.rb', line 669 def n_gpu_layers; end |
#n_gpu_layers=(n_gpu_layers) ⇒ Object
Sets the number of layers to stor in VRAM.
665 |
# File 'ext/llama_cpp/dummy.rb', line 665 def n_gpu_layers=(n_gpu_layers); end |
#split_mode ⇒ Integer
Return the split mode.
677 |
# File 'ext/llama_cpp/dummy.rb', line 677 def split_mode; end |
#split_mode=(split_mode) ⇒ Object
Sets the split mode.
673 |
# File 'ext/llama_cpp/dummy.rb', line 673 def split_mode=(split_mode); end |
#tensor_split ⇒ Array<Float>
Returns the how layers are split across multi-GPUs.
689 |
# File 'ext/llama_cpp/dummy.rb', line 689 def tensor_split; end |
#use_mlock ⇒ Boolean
Returns the flag to force system to keep model in memory.
713 |
# File 'ext/llama_cpp/dummy.rb', line 713 def use_mlock; end |
#use_mlock=(flag) ⇒ Object
Sets the flag to force system to keep model in memory.
709 |
# File 'ext/llama_cpp/dummy.rb', line 709 def use_mlock=(flag); end |
#use_mmap ⇒ Boolean
Returns the flag to use mmap.
705 |
# File 'ext/llama_cpp/dummy.rb', line 705 def use_mmap; end |
#use_mmap=(flag) ⇒ Object
Sets the flag to use mmap.
701 |
# File 'ext/llama_cpp/dummy.rb', line 701 def use_mmap=(flag); end |
#vocab_only ⇒ Boolean
Returns the flag to load only vocabulary.
697 |
# File 'ext/llama_cpp/dummy.rb', line 697 def vocab_only; end |
#vocab_only=(flag) ⇒ Object
Sets the flag to load only vocabulary.
693 |
# File 'ext/llama_cpp/dummy.rb', line 693 def vocab_only=(flag); end |