Class: LlamaCpp::LlamaContextParams

Inherits:
Object
  • Object
show all
Defined in:
ext/llama_cpp/llama_cpp.c,
ext/llama_cpp/llama_cpp.c

Overview

“struct llama_context_params” wrapper class

Instance Method Summary collapse

Instance Method Details

#attention_type ⇒ Integer

Returns:

  • (Integer) 


699
700
701
702
 
# File 'ext/llama_cpp/llama_cpp.c', line 699

static VALUE llama_context_params_get_attention_type(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->attention_type);
}

#attention_type=(attention_type) ⇒ Integer

Parameters:

  • attention_type (Integer)

Returns:

  • (Integer) 


704
705
706
707
708
 
# File 'ext/llama_cpp/llama_cpp.c', line 704

static VALUE llama_context_params_set_attention_type(VALUE self, VALUE attention_type) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->attention_type = (enum llama_attention_type)NUM2INT(attention_type);
  return attention_type;
}

#defrag_thold ⇒ Float

Returns:

  • (Float) 


798
799
800
801
 
# File 'ext/llama_cpp/llama_cpp.c', line 798

static VALUE llama_context_params_get_defrag_thold(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->defrag_thold);
}

#defrag_thold=(defrag_thold) ⇒ Float

Parameters:

  • defrag_thold (Float)

Returns:

  • (Float) 


803
804
805
806
807
 
# File 'ext/llama_cpp/llama_cpp.c', line 803

static VALUE llama_context_params_set_defrag_thold(VALUE self, VALUE defrag_thold) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->defrag_thold = NUM2DBL(defrag_thold);
  return defrag_thold;
}

#embeddings ⇒ Boolean

Returns:

  • (Boolean) 


831
832
833
834
 
# File 'ext/llama_cpp/llama_cpp.c', line 831

static VALUE llama_context_params_get_embeddings(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return data->embeddings ? Qtrue : Qfalse;
}

#embeddings=(embeddings) ⇒ Boolean

Parameters:

  • embeddings (Boolean)

Returns:

  • (Boolean) 


836
837
838
839
840
 
# File 'ext/llama_cpp/llama_cpp.c', line 836

static VALUE llama_context_params_set_embeddings(VALUE self, VALUE embeddings) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->embeddings = RTEST(embeddings) ? true : false;
  return embeddings;
}

#flash_attn_type ⇒ Integer

Returns:

  • (Integer) 


710
711
712
713
 
# File 'ext/llama_cpp/llama_cpp.c', line 710

static VALUE llama_context_params_get_flash_attn_type(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->flash_attn_type);
}

#flash_attn_type=(flash_attn_type) ⇒ Integer

Parameters:

  • flash_attn_type (Integer)

Returns:

  • (Integer) 


715
716
717
718
719
 
# File 'ext/llama_cpp/llama_cpp.c', line 715

static VALUE llama_context_params_set_flash_attn_type(VALUE self, VALUE flash_attn_type) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->flash_attn_type = (enum llama_flash_attn_type)NUM2INT(flash_attn_type);
  return flash_attn_type;
}

#kv_unified ⇒ Boolean

Returns:

  • (Boolean) 


886
887
888
889
 
# File 'ext/llama_cpp/llama_cpp.c', line 886

static VALUE llama_context_params_get_kv_unified(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return data->kv_unified ? Qtrue : Qfalse;
}

#kv_unified=(kv_unified) ⇒ Boolean

Parameters:

  • kv_unified (Boolean)

Returns:

  • (Boolean) 


891
892
893
894
895
 
# File 'ext/llama_cpp/llama_cpp.c', line 891

static VALUE llama_context_params_set_kv_unified(VALUE self, VALUE kv_unified) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->kv_unified = RTEST(kv_unified) ? true : false;
  return kv_unified;
}

#n_batch ⇒ Integer

Returns:

  • (Integer) 


622
623
624
625
 
# File 'ext/llama_cpp/llama_cpp.c', line 622

static VALUE llama_context_params_get_n_batch(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return UINT2NUM(data->n_batch);
}

#n_batch=(n_batch) ⇒ Integer

Parameters:

  • n_batch (Integer)

Returns:

  • (Integer) 


627
628
629
630
631
 
# File 'ext/llama_cpp/llama_cpp.c', line 627

static VALUE llama_context_params_set_n_batch(VALUE self, VALUE n_batch) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->n_batch = NUM2UINT(n_batch);
  return n_batch;
}

#n_ctx ⇒ Integer

Returns:

  • (Integer) 


611
612
613
614
 
# File 'ext/llama_cpp/llama_cpp.c', line 611

static VALUE llama_context_params_get_n_ctx(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return UINT2NUM(data->n_ctx);
}

#n_ctx=(n_ctx) ⇒ Integer

Parameters:

  • n_ctx (Integer)

Returns:

  • (Integer) 


616
617
618
619
620
 
# File 'ext/llama_cpp/llama_cpp.c', line 616

static VALUE llama_context_params_set_n_ctx(VALUE self, VALUE n_ctx) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->n_ctx = NUM2UINT(n_ctx);
  return n_ctx;
}

#n_seq_max ⇒ Integer

Returns:

  • (Integer) 


644
645
646
647
 
# File 'ext/llama_cpp/llama_cpp.c', line 644

static VALUE llama_context_params_get_n_seq_max(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return UINT2NUM(data->n_seq_max);
}

#n_seq_max=(n_seq_max) ⇒ Integer

Parameters:

  • n_seq_max (Integer)

Returns:

  • (Integer) 


649
650
651
652
653
 
# File 'ext/llama_cpp/llama_cpp.c', line 649

static VALUE llama_context_params_set_n_seq_max(VALUE self, VALUE n_seq_max) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->n_seq_max = NUM2UINT(n_seq_max);
  return n_seq_max;
}

#n_threads ⇒ Integer

Returns:

  • (Integer) 


655
656
657
658
 
# File 'ext/llama_cpp/llama_cpp.c', line 655

static VALUE llama_context_params_get_n_threads(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->n_threads);
}

#n_threads=(n_threads) ⇒ Integer

Parameters:

  • n_threads (Integer)

Returns:

  • (Integer) 


660
661
662
663
664
 
# File 'ext/llama_cpp/llama_cpp.c', line 660

static VALUE llama_context_params_set_n_threads(VALUE self, VALUE n_threads) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->n_threads = NUM2INT(n_threads);
  return n_threads;
}

#n_threads_batch ⇒ Integer

Returns:

  • (Integer) 


666
667
668
669
 
# File 'ext/llama_cpp/llama_cpp.c', line 666

static VALUE llama_context_params_get_n_threads_batch(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->n_threads_batch);
}

#n_threads_batch=(n_threads_batch) ⇒ Integer

Parameters:

  • n_threads_batch (Integer)

Returns:

  • (Integer) 


671
672
673
674
675
 
# File 'ext/llama_cpp/llama_cpp.c', line 671

static VALUE llama_context_params_set_n_threads_batch(VALUE self, VALUE n_threads_batch) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->n_threads = NUM2INT(n_threads_batch);
  return n_threads_batch;
}

#n_ubatch ⇒ Integer

Returns:

  • (Integer) 


633
634
635
636
 
# File 'ext/llama_cpp/llama_cpp.c', line 633

static VALUE llama_context_params_get_n_ubatch(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return UINT2NUM(data->n_ubatch);
}

#n_ubatch=(n_ubatch) ⇒ Integer

Parameters:

  • n_ubatch (Integer)

Returns:

  • (Integer) 


638
639
640
641
642
 
# File 'ext/llama_cpp/llama_cpp.c', line 638

static VALUE llama_context_params_set_n_ubatch(VALUE self, VALUE n_ubatch) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->n_ubatch = NUM2UINT(n_ubatch);
  return n_ubatch;
}

#no_perf ⇒ Boolean

Returns:

  • (Boolean) 


853
854
855
856
 
# File 'ext/llama_cpp/llama_cpp.c', line 853

static VALUE llama_context_params_get_no_perf(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return data->no_perf ? Qtrue : Qfalse;
}

#no_perf=(no_perf) ⇒ Boolean

Parameters:

  • no_perf (Boolean)

Returns:

  • (Boolean) 


858
859
860
861
862
 
# File 'ext/llama_cpp/llama_cpp.c', line 858

static VALUE llama_context_params_set_no_perf(VALUE self, VALUE no_perf) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->no_perf = RTEST(no_perf) ? true : false;
  return no_perf;
}

#offload_kqv ⇒ Boolean

Returns:

  • (Boolean) 


842
843
844
845
 
# File 'ext/llama_cpp/llama_cpp.c', line 842

static VALUE llama_context_params_get_offload_kqv(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return data->offload_kqv ? Qtrue : Qfalse;
}

#offload_kqv=(offload_kqv) ⇒ Boolean

Parameters:

  • offload_kqv (Boolean)

Returns:

  • (Boolean) 


847
848
849
850
851
 
# File 'ext/llama_cpp/llama_cpp.c', line 847

static VALUE llama_context_params_set_offload_kqv(VALUE self, VALUE offload_kqv) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->offload_kqv = RTEST(offload_kqv) ? true : false;
  return offload_kqv;
}

#op_offload ⇒ Boolean

Returns:

  • (Boolean) 


864
865
866
867
 
# File 'ext/llama_cpp/llama_cpp.c', line 864

static VALUE llama_context_params_get_op_offload(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return data->op_offload ? Qtrue : Qfalse;
}

#op_offload=(op_offload) ⇒ Boolean

Parameters:

  • op_offload (Boolean)

Returns:

  • (Boolean) 


869
870
871
872
873
 
# File 'ext/llama_cpp/llama_cpp.c', line 869

static VALUE llama_context_params_set_op_offload(VALUE self, VALUE op_offload) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->op_offload = RTEST(op_offload) ? true : false;
  return op_offload;
}

#pooling_type ⇒ Integer

Returns:

  • (Integer) 


688
689
690
691
 
# File 'ext/llama_cpp/llama_cpp.c', line 688

static VALUE llama_context_params_get_pooling_type(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->pooling_type);
}

#pooling_type=(pooling_type) ⇒ Integer

Parameters:

  • pooling_type (Integer)

Returns:

  • (Integer) 


693
694
695
696
697
 
# File 'ext/llama_cpp/llama_cpp.c', line 693

static VALUE llama_context_params_set_pooling_type(VALUE self, VALUE pooling_type) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->pooling_type = (enum llama_pooling_type)NUM2INT(pooling_type);
  return pooling_type;
}

#rope_freq_base ⇒ Float

Returns:

  • (Float) 


721
722
723
724
 
# File 'ext/llama_cpp/llama_cpp.c', line 721

static VALUE llama_context_params_get_rope_freq_base(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->rope_freq_base);
}

#rope_freq_base=(rope_freq_base) ⇒ Float

Parameters:

  • rope_freq_base (Float)

Returns:

  • (Float) 


726
727
728
729
730
 
# File 'ext/llama_cpp/llama_cpp.c', line 726

static VALUE llama_context_params_set_rope_freq_base(VALUE self, VALUE rope_freq_base) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->rope_freq_base = NUM2DBL(rope_freq_base);
  return rope_freq_base;
}

#rope_freq_scale ⇒ Float

Returns:

  • (Float) 


732
733
734
735
 
# File 'ext/llama_cpp/llama_cpp.c', line 732

static VALUE llama_context_params_get_rope_freq_scale(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->rope_freq_scale);
}

#rope_freq_scale=(rope_freq_scale) ⇒ Float

Parameters:

  • rope_freq_scale (Float)

Returns:

  • (Float) 


737
738
739
740
741
 
# File 'ext/llama_cpp/llama_cpp.c', line 737

static VALUE llama_context_params_set_rope_freq_scale(VALUE self, VALUE rope_freq_scale) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->rope_freq_scale = NUM2DBL(rope_freq_scale);
  return rope_freq_scale;
}

#rope_scaling_type ⇒ Integer

Returns:

  • (Integer) 


677
678
679
680
 
# File 'ext/llama_cpp/llama_cpp.c', line 677

static VALUE llama_context_params_get_rope_scaling_type(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->rope_scaling_type);
}

#rope_scaling_type=(rope_scaling_type) ⇒ Integer

Parameters:

  • rope_scaling_type (Integer)

Returns:

  • (Integer) 


682
683
684
685
686
 
# File 'ext/llama_cpp/llama_cpp.c', line 682

static VALUE llama_context_params_set_rope_scaling_type(VALUE self, VALUE rope_scaling_type) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->rope_scaling_type = (enum llama_rope_scaling_type)NUM2INT(rope_scaling_type);
  return rope_scaling_type;
}

#swa_full ⇒ Boolean

Returns:

  • (Boolean) 


875
876
877
878
 
# File 'ext/llama_cpp/llama_cpp.c', line 875

static VALUE llama_context_params_get_swa_full(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return data->swa_full ? Qtrue : Qfalse;
}

#swa_full=(swa_full) ⇒ Boolean

Parameters:

  • swa_full (Boolean)

Returns:

  • (Boolean) 


880
881
882
883
884
 
# File 'ext/llama_cpp/llama_cpp.c', line 880

static VALUE llama_context_params_set_swa_full(VALUE self, VALUE swa_full) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->swa_full = RTEST(swa_full) ? true : false;
  return swa_full;
}

#type_k ⇒ Integer

Returns:

  • (Integer) 


809
810
811
812
 
# File 'ext/llama_cpp/llama_cpp.c', line 809

static VALUE llama_context_params_get_type_k(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->type_k);
}

#type_k=(type_k) ⇒ Integer

Parameters:

  • type_k (Integer)

Returns:

  • (Integer) 


814
815
816
817
818
 
# File 'ext/llama_cpp/llama_cpp.c', line 814

static VALUE llama_context_params_set_type_k(VALUE self, VALUE type_k) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->type_k = (enum ggml_type)NUM2INT(type_k);
  return type_k;
}

#type_v ⇒ Integer

Returns:

  • (Integer) 


820
821
822
823
 
# File 'ext/llama_cpp/llama_cpp.c', line 820

static VALUE llama_context_params_get_type_v(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return INT2NUM(data->type_v);
}

#type_v=(type_v) ⇒ Integer

Parameters:

  • type_v (Integer)

Returns:

  • (Integer) 


825
826
827
828
829
 
# File 'ext/llama_cpp/llama_cpp.c', line 825

static VALUE llama_context_params_set_type_v(VALUE self, VALUE type_v) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->type_v = (enum ggml_type)NUM2INT(type_v);
  return type_v;
}

#yarn_attn_factor ⇒ Float

Returns:

  • (Float) 


754
755
756
757
 
# File 'ext/llama_cpp/llama_cpp.c', line 754

static VALUE llama_context_params_get_yarn_attn_factor(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->yarn_attn_factor);
}

#yarn_attn_factor=(yarn_attn_factor) ⇒ Float

Parameters:

  • yarn_attn_factor (Float)

Returns:

  • (Float) 


759
760
761
762
763
 
# File 'ext/llama_cpp/llama_cpp.c', line 759

static VALUE llama_context_params_set_yarn_attn_factor(VALUE self, VALUE yarn_attn_factor) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->yarn_attn_factor = NUM2DBL(yarn_attn_factor);
  return yarn_attn_factor;
}

#yarn_beta_fast ⇒ Float

Returns:

  • (Float) 


765
766
767
768
 
# File 'ext/llama_cpp/llama_cpp.c', line 765

static VALUE llama_context_params_get_yarn_beta_fast(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->yarn_beta_fast);
}

#yarn_beta_fast=(yarn_beta_fast) ⇒ Float

Parameters:

  • yarn_beta_fast (Float)

Returns:

  • (Float) 


770
771
772
773
774
 
# File 'ext/llama_cpp/llama_cpp.c', line 770

static VALUE llama_context_params_set_yarn_beta_fast(VALUE self, VALUE yarn_beta_fast) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->yarn_beta_fast = NUM2DBL(yarn_beta_fast);
  return yarn_beta_fast;
}

#yarn_beta_slow ⇒ Float

Returns:

  • (Float) 


776
777
778
779
 
# File 'ext/llama_cpp/llama_cpp.c', line 776

static VALUE llama_context_params_get_yarn_beta_slow(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->yarn_beta_slow);
}

#yarn_beta_slow=(yarn_beta_slow) ⇒ Float

Parameters:

  • yarn_beta_slow (Float)

Returns:

  • (Float) 


781
782
783
784
785
 
# File 'ext/llama_cpp/llama_cpp.c', line 781

static VALUE llama_context_params_set_yarn_beta_slow(VALUE self, VALUE yarn_beta_slow) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->yarn_beta_slow = NUM2DBL(yarn_beta_slow);
  return yarn_beta_slow;
}

#yarn_ext_factor ⇒ Float

Returns:

  • (Float) 


743
744
745
746
 
# File 'ext/llama_cpp/llama_cpp.c', line 743

static VALUE llama_context_params_get_yarn_ext_factor(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return DBL2NUM(data->yarn_ext_factor);
}

#yarn_ext_factor=(yarn_ext_factor) ⇒ Float

Parameters:

  • yarn_ext_factor (Float)

Returns:

  • (Float) 


748
749
750
751
752
 
# File 'ext/llama_cpp/llama_cpp.c', line 748

static VALUE llama_context_params_set_yarn_ext_factor(VALUE self, VALUE yarn_ext_factor) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->yarn_ext_factor = NUM2DBL(yarn_ext_factor);
  return yarn_ext_factor;
}

#yarn_orig_ctx ⇒ Integer

Returns:

  • (Integer) 


787
788
789
790
 
# File 'ext/llama_cpp/llama_cpp.c', line 787

static VALUE llama_context_params_get_yarn_orig_ctx(VALUE self) {
  struct llama_context_params* data = get_llama_context_params(self);
  return UINT2NUM(data->yarn_orig_ctx);
}

#yarn_orig_ctx=(yarn_orig_ctx) ⇒ Integer

Parameters:

  • yarn_orig_ctx (Integer)

Returns:

  • (Integer) 


792
793
794
795
796
 
# File 'ext/llama_cpp/llama_cpp.c', line 792

static VALUE llama_context_params_set_yarn_orig_ctx(VALUE self, VALUE yarn_orig_ctx) {
  struct llama_context_params* data = get_llama_context_params(self);
  data->yarn_orig_ctx = NUM2UINT(yarn_orig_ctx);
  return yarn_orig_ctx;
}