A B C D E F G I L M N O P Q R S T W misc
| ag_abs | Element-wise Unary Operations (autograd) |
| ag_add | Element-wise addition with broadcasting |
| ag_argmax | Column-wise argmax |
| ag_batch_norm | Create a Batch Normalisation layer |
| ag_buffer | Device buffer for column-indexed reads |
| ag_buffer_copy_from | Copy a device tensor into buffer columns |
| ag_buffer_free | Device buffer for column-indexed reads |
| ag_buffer_read | Device buffer for column-indexed reads |
| ag_buffer_write | Device buffer for column-indexed reads |
| ag_capture | Capture an ag_* computation as a reusable device graph |
| ag_capture_free | Release the device memory held by a captured function |
| ag_capture_repeat | Run a captured function several times in one go |
| ag_capture_set | Change a setting of a captured training step |
| ag_capture_step | Capture a whole training step as reusable device graphs |
| ag_checkpoint | Run part of the forward pass without storing its activations |
| ag_clamp | Element-wise clamp |
| ag_concat_rows | Concatenate two tensors by rows |
| ag_conv2d | 2-D convolution |
| ag_conv2d_layer | Convolution layer |
| ag_conv_transpose2d | 2-D transposed convolution |
| ag_conv_transpose2d_layer | Transposed convolution layer |
| ag_cos | Element-wise Unary Operations (autograd) |
| ag_cross_entropy_loss | Categorical Cross-Entropy loss |
| ag_dataloader | Create a mini-batch data loader |
| ag_default_device | Return the current default compute device |
| ag_default_dtype | Return the current default dtype for GPU operations |
| ag_detach | Stop gradient |
| ag_device | Set the default compute device for ag_* operations |
| ag_discounted_returns | Generalised advantage estimation and discounted returns |
| ag_div | Element-wise division |
| ag_dropout | Create a Dropout layer |
| ag_dtype | Set the default floating-point precision for ag_* GPU operations |
| ag_embedding | Create an Embedding layer |
| ag_entropy | Entropy of a categorical distribution given by logits |
| ag_estimate_training_memory | Estimate the memory a training run needs |
| ag_eval | Switch a layer or sequential model to eval mode |
| ag_exp | Element-wise Unary Operations (autograd) |
| ag_flash_attention | Multi-head attention in a single fused operation |
| ag_gae | Generalised advantage estimation and discounted returns |
| ag_gather | Pick one entry per column |
| ag_get_rows | Read buffer columns by index |
| ag_grad | Gradient of a tensor as an R matrix |
| ag_gradcheck | Numerical gradient check (like torch.autograd.gradcheck) |
| ag_graph_mode | Graph mode for ag_* operations on the GPU |
| ag_gru | LSTM and GRU over a sequence |
| ag_gru_cell | LSTM and GRU cells |
| ag_hard_update | Update a target network from a source network |
| ag_huber_loss | Huber and smooth-L1 losses |
| ag_layer_norm | Create a Layer Normalization layer |
| ag_linear | Create a dense layer with learnable parameters |
| ag_load_model | Load an autograd module from a saved state |
| ag_load_state_dict | Load weights into an existing model |
| ag_local_mode | Set ag_* GPU modes for the rest of the calling function |
| ag_log | Element-wise Unary Operations (autograd) |
| ag_logsumexp | Column-wise log-sum-exp |
| ag_log_softmax | Column-wise log-softmax |
| ag_lstm | LSTM and GRU over a sequence |
| ag_lstm_cell | LSTM and GRU cells |
| ag_matmul | Matrix multiplication |
| ag_matmul_precision | Matrix-multiply precision for ag_* operations on the GPU |
| ag_max | Column-wise maximum |
| ag_maximum | Element-wise minimum and maximum of two tensors |
| ag_mean | Mean of elements (or along a dim) |
| ag_minimum | Element-wise minimum and maximum of two tensors |
| ag_mse_loss | Mean Squared Error loss |
| ag_mul | Element-wise multiplication |
| ag_multihead_attention | Create a Multi-Head Attention layer |
| ag_normal_entropy | Diagonal normal distribution: log-density and entropy |
| ag_normal_log_prob | Diagonal normal distribution: log-density and entropy |
| ag_nstep_targets | n-step bootstrapped targets |
| ag_param | Create a parameter tensor (gradient tracked) |
| ag_pass_scope | Evaluate eager device work in a private, self-freeing pass pool |
| ag_per_sampler | Prioritized replay sampler |
| ag_pow | Element-wise power |
| ag_rand | Random numbers on the device |
| ag_relu | Element-wise Unary Operations (autograd) |
| ag_replay_buffer | Replay buffer |
| ag_replay_buffer_load | Load a replay buffer saved with '$save()' |
| ag_reshape | Reshape tensor |
| ag_rms_norm | Create an RMS Normalization layer |
| ag_sample_categorical | Sample one index per column from logits |
| ag_sample_normal | Sample from a diagonal Gaussian |
| ag_save_model | Save an autograd module's state to disk |
| ag_scale | Scale tensor by a scalar constant |
| ag_sequential | Create a sequential container of layers |
| ag_set_rows | Write buffer columns by index |
| ag_sigmoid | Element-wise Unary Operations (autograd) |
| ag_silu | Element-wise Unary Operations (autograd) |
| ag_sin | Element-wise Unary Operations (autograd) |
| ag_slice_rows | Slice a range of rows |
| ag_smooth_l1_loss | Huber and smooth-L1 losses |
| ag_softmax | Softmax activation (column-wise) |
| ag_softmax_cross_entropy_loss | Fused softmax + cross-entropy loss (numerically stable) |
| ag_softplus | Element-wise Unary Operations (autograd) |
| ag_soft_update | Update a target network from a source network |
| ag_sqrt | Element-wise Unary Operations (autograd) |
| ag_state_dict | Model weights as plain R data |
| ag_sub | Element-wise subtraction |
| ag_sum | Sum all elements (or along a dim): out = sum(x) |
| ag_tanh | Element-wise Unary Operations (autograd) |
| ag_tape_memory | Report what the gradient tape is holding |
| ag_tensor | Create a dynamic tensor (no gradient tracking) |
| ag_to_device | Move a tensor to the specified device |
| ag_train | Switch a layer or sequential model to training mode |
| ag_transpose | Transpose a tensor |
| ag_where | Element-wise selection by a mask |
| as.matrix.ag_tensor | Extract the value of an ag_tensor as an R matrix |
| as.matrix.ggml_matrix | Extract the underlying matrix from a ggml_matrix |
| as_gpu_matrix | Wrap a matrix so multiplies run on the GPU |
| augment.ggmlr_parsnip_model | Augment new data with predictions from a fitted ggml parsnip model |
| backward | Run backward pass from a scalar loss tensor |
| check_grad_anomaly | Detect NaN / Inf gradients |
| clip_grad_norm | Clip gradients by global L2 norm |
| clip_grad_value | Clip gradients element-wise by value |
| compile.ggml_functional_model | Compile a Model |
| compile.ggml_sequential_model | Compile a Model |
| crossprod-method | A GPU-backed matrix wrapper |
| evaluate | Evaluate a Model |
| evaluate.ggml_functional_model | Evaluate a Model |
| evaluate.ggml_sequential_model | Evaluate a Model |
| fit.ggml_functional_model | Train a Model |
| fit.ggml_sequential_model | Train a Model |
| ggml_abort_is_r_enabled | Logging and Abort Callbacks |
| ggml_abs | Absolute Value (Graph) |
| ggml_abs_inplace | Absolute Value (Graph) |
| ggml_add | Element-wise Addition (Graph) |
| ggml_add1 | Add Scalar to Tensor (Graph) |
| ggml_add_inplace | Element-wise Addition (Graph) |
| ggml_add_rel_pos | Add Relative Position Bias (Graph) |
| ggml_apply | Apply a Layer Object to a Tensor Node |
| ggml_arange | Arange (Graph) |
| ggml_are_same_layout | Check if Two Tensors Have the Same Layout |
| ggml_are_same_shape | Compare Tensor Shapes |
| ggml_are_same_stride | Compare Tensor Strides |
| ggml_argmax | Argmax (Graph) |
| ggml_argsort | Argsort - Get Sorting Indices (Graph) |
| ggml_attention | Create a Multi-Head Attention Layer Object |
| ggml_backend_alloc_ctx_tensors | Allocate Context Tensors to Backend |
| ggml_backend_buffer_clear | Backend Buffer Operations |
| ggml_backend_buffer_free | Backend Buffer Operations |
| ggml_backend_buffer_get_size | Backend Buffer Operations |
| ggml_backend_buffer_get_usage | Backend Buffer Operations |
| ggml_backend_buffer_is_host | Backend Buffer Operations |
| ggml_backend_buffer_is_multi_buffer | Backend Multi-buffers |
| ggml_backend_buffer_name | Backend Buffer Operations |
| ggml_backend_buffer_reset | Backend Buffer Operations |
| ggml_backend_buffer_set_usage | Backend Buffer Operations |
| ggml_backend_buffer_usage_any | Backend Buffer Usage Constants |
| ggml_backend_buffer_usage_compute | Backend Buffer Usage Constants |
| ggml_backend_buffer_usage_weights | Backend Buffer Usage Constants |
| ggml_backend_buft_get_alignment | Backend Buffer Type Information |
| ggml_backend_buft_get_max_size | Backend Buffer Type Information |
| ggml_backend_buft_is_host | Backend Buffer Type Information |
| ggml_backend_buft_name | Backend Buffer Type Information |
| ggml_backend_cpu_init | Initialize CPU Backend |
| ggml_backend_cpu_set_n_threads | Set CPU Backend Threads |
| ggml_backend_device_register | Register a device |
| ggml_backend_device_type_accel | Backend Device Type Constants |
| ggml_backend_device_type_cpu | Backend Device Type Constants |
| ggml_backend_device_type_gpu | Backend Device Type Constants |
| ggml_backend_device_type_igpu | Backend Device Type Constants |
| ggml_backend_dev_buffer_type | Backend Device Buffer Types |
| ggml_backend_dev_by_name | Backend Device Lookup |
| ggml_backend_dev_by_type | Backend Device Lookup |
| ggml_backend_dev_count | Backend Device Lookup |
| ggml_backend_dev_description | Backend Device Properties |
| ggml_backend_dev_get | Backend Device Lookup |
| ggml_backend_dev_get_props | Backend Device Properties |
| ggml_backend_dev_host_buffer_type | Backend Device Buffer Types |
| ggml_backend_dev_init | Backend Initialization |
| ggml_backend_dev_memory | Backend Device Properties |
| ggml_backend_dev_name | Backend Device Properties |
| ggml_backend_dev_offload_op | Backend Device Capability Queries |
| ggml_backend_dev_supports_buft | Backend Device Capability Queries |
| ggml_backend_dev_supports_op | Backend Device Capability Queries |
| ggml_backend_dev_type | Backend Device Properties |
| ggml_backend_event_free | Backend Events |
| ggml_backend_event_new | Backend Events |
| ggml_backend_event_record | Backend Events |
| ggml_backend_event_synchronize | Backend Events |
| ggml_backend_event_wait | Backend Events |
| ggml_backend_free | Free Backend |
| ggml_backend_get_device | Get device from backend |
| ggml_backend_graph_compute | Compute Graph with Backend |
| ggml_backend_graph_compute_async | Compute graph asynchronously |
| ggml_backend_graph_plan_compute | Backend Graph Plans |
| ggml_backend_graph_plan_create | Backend Graph Plans |
| ggml_backend_graph_plan_free | Backend Graph Plans |
| ggml_backend_init_best | Backend Initialization |
| ggml_backend_init_by_name | Backend Initialization |
| ggml_backend_init_by_type | Backend Initialization |
| ggml_backend_load | Dynamic Backend Loading |
| ggml_backend_load_all | Dynamic Backend Loading |
| ggml_backend_meta_device | Create a Meta Backend Device |
| ggml_backend_multi_buffer_alloc_buffer | Backend Multi-buffers |
| ggml_backend_multi_buffer_set_usage | Backend Multi-buffers |
| ggml_backend_name | Get Backend Name |
| ggml_backend_register | Register a backend |
| ggml_backend_reg_by_name | Backend Registry |
| ggml_backend_reg_count | Backend Registry |
| ggml_backend_reg_dev_count | Backend Registry |
| ggml_backend_reg_dev_get | Backend Registry |
| ggml_backend_reg_get | Backend Registry |
| ggml_backend_reg_name | Backend Registry |
| ggml_backend_sched_alloc_graph | Allocate graph on scheduler |
| ggml_backend_sched_free | Free backend scheduler |
| ggml_backend_sched_get_backend | Scheduler Information |
| ggml_backend_sched_get_n_backends | Scheduler Information |
| ggml_backend_sched_get_n_copies | Scheduler Information |
| ggml_backend_sched_get_n_splits | Scheduler Information |
| ggml_backend_sched_get_tensor_backend | Scheduler Tensor Backend Assignment |
| ggml_backend_sched_graph_compute | Scheduler Graph Compute |
| ggml_backend_sched_graph_compute_async | Scheduler Graph Compute |
| ggml_backend_sched_new | Create a new backend scheduler |
| ggml_backend_sched_reserve | Reserve memory for scheduler |
| ggml_backend_sched_reset | Reset scheduler |
| ggml_backend_sched_set_tensor_backend | Scheduler Tensor Backend Assignment |
| ggml_backend_sched_synchronize | Synchronize scheduler |
| ggml_backend_sched_trace | Trace Every Node a Scheduler Computes |
| ggml_backend_synchronize | Synchronize backend |
| ggml_backend_tensor_copy_async | Asynchronous Tensor Transfers |
| ggml_backend_tensor_get_and_sync | Backend Tensor Get and Sync |
| ggml_backend_tensor_get_async | Asynchronous Tensor Transfers |
| ggml_backend_tensor_get_data | Backend Tensor Data Access |
| ggml_backend_tensor_get_f32_first | Get First Float from Backend Tensor |
| ggml_backend_tensor_set_async | Asynchronous Tensor Transfers |
| ggml_backend_tensor_set_data | Backend Tensor Data Access |
| ggml_backend_unload | Dynamic Backend Loading |
| ggml_batch_norm | Create a Batch Normalization Layer Object |
| ggml_blck_size | Type Information |
| ggml_build_backward_expand | Add the Backward Pass to a Computation Graph |
| ggml_build_forward_expand | Build Forward Computation Graph |
| ggml_build_forward_expand_grads | Build a Forward Graph That Can Be Differentiated |
| ggml_callback_early_stopping | Early stopping callback |
| ggml_can_repeat | Check If Tensor Can Be Repeated |
| ggml_cast | Cast a Tensor to Another Type |
| ggml_cast_numeric | Numeric Type Conversion |
| ggml_ceil | Ceiling (Graph) |
| ggml_ceil_inplace | Ceiling (Graph) |
| ggml_clamp | Clamp (Graph) |
| ggml_clamp_inplace | Clamp (Graph) |
| ggml_col2im_1d | Column to 1D Signal (Graph) |
| ggml_compile | Compile a Sequential Model |
| ggml_compile.ggml_functional_model | Compile a Sequential Model |
| ggml_compile.ggml_sequential_model | Compile a Sequential Model |
| ggml_concat | Concatenate Tensors (Graph) |
| ggml_cont | Make Contiguous (Graph) |
| ggml_conv_1d | 1D Convolution (Graph) |
| ggml_conv_1d_dw | 1D Convolution (Graph) |
| ggml_conv_1d_layer | Create a Reusable Conv1D Layer Object |
| ggml_conv_2d | 2D Convolution (Graph) |
| ggml_conv_2d_direct | 2D Convolution (Graph) |
| ggml_conv_2d_dw | 2D Convolution (Graph) |
| ggml_conv_2d_dw_direct | 2D Convolution (Graph) |
| ggml_conv_2d_layer | Create a Reusable Conv2D Layer Object |
| ggml_conv_3d | 3D Convolution (Graph) |
| ggml_conv_3d_direct | 3D Convolution (Graph) |
| ggml_conv_transpose_1d | Transposed 1D Convolution (Graph) |
| ggml_conv_transpose_2d_p0 | Transposed 2D Convolution, zero padding (Graph) |
| ggml_cos | Cosine (Graph) |
| ggml_count_equal | Count Equal Elements (Graph) |
| ggml_cpu_add | Element-wise Addition (CPU Direct) |
| ggml_cpu_features | CPU Feature Detection |
| ggml_cpu_get_rvv_vlen | CPU Feature Detection |
| ggml_cpu_get_sve_cnt | CPU Feature Detection |
| ggml_cpu_has_amx_int8 | CPU Feature Detection |
| ggml_cpu_has_arm_fma | CPU Feature Detection |
| ggml_cpu_has_avx | CPU Feature Detection |
| ggml_cpu_has_avx2 | CPU Feature Detection |
| ggml_cpu_has_avx512 | CPU Feature Detection |
| ggml_cpu_has_avx512_bf16 | CPU Feature Detection |
| ggml_cpu_has_avx512_vbmi | CPU Feature Detection |
| ggml_cpu_has_avx512_vnni | CPU Feature Detection |
| ggml_cpu_has_avx_vnni | CPU Feature Detection |
| ggml_cpu_has_bmi2 | CPU Feature Detection |
| ggml_cpu_has_dotprod | CPU Feature Detection |
| ggml_cpu_has_f16c | CPU Feature Detection |
| ggml_cpu_has_fma | CPU Feature Detection |
| ggml_cpu_has_fp16_va | CPU Feature Detection |
| ggml_cpu_has_llamafile | CPU Feature Detection |
| ggml_cpu_has_matmul_int8 | CPU Feature Detection |
| ggml_cpu_has_neon | CPU Feature Detection |
| ggml_cpu_has_riscv_v | CPU Feature Detection |
| ggml_cpu_has_sme | CPU Feature Detection |
| ggml_cpu_has_sme2 | CPU Feature Detection |
| ggml_cpu_has_sse3 | CPU Feature Detection |
| ggml_cpu_has_ssse3 | CPU Feature Detection |
| ggml_cpu_has_sve | CPU Feature Detection |
| ggml_cpu_has_vsx | CPU Feature Detection |
| ggml_cpu_has_vxe | CPU Feature Detection |
| ggml_cpu_has_wasm_simd | CPU Feature Detection |
| ggml_cpu_mul | Element-wise Multiplication (CPU Direct) |
| ggml_cpy | Copy Tensor with Type Conversion (Graph) |
| ggml_crossprod | GPU cross product (drop-in for 'crossprod') |
| ggml_cross_entropy_loss | Cross-Entropy Loss (Graph) |
| ggml_cross_entropy_loss_back | Cross-Entropy Loss Backward (Graph) |
| ggml_cumsum | Cumulative Sum (Graph) |
| ggml_custom | Custom Operation Node |
| ggml_custom_inplace | In-Place Custom Operation Node |
| ggml_custom_ops | List Registered Custom Ops |
| ggml_cycles | Timing Utilities |
| ggml_cycles_per_ms | Timing Utilities |
| ggml_default_mlp | Default MLP builder for classification and regression |
| ggml_dense | Create a Dense Layer Object |
| ggml_diag | Diagonal Matrix (Graph) |
| ggml_diag_mask_inf | Diagonal Mask with -Inf (Graph) |
| ggml_diag_mask_inf_inplace | Diagonal Mask with -Inf (Graph) |
| ggml_diag_mask_zero | Diagonal Mask with Zero (Graph) |
| ggml_div | Element-wise Division (Graph) |
| ggml_div_inplace | Element-wise Division (Graph) |
| ggml_dsv4_hc_comb | DeepSeek V4 Hyper-Connection Combination Matrix (Graph) |
| ggml_dsv4_hc_post | DeepSeek V4 Hyper-Connection Post-Mix (Graph) |
| ggml_dsv4_hc_pre | DeepSeek V4 Hyper-Connection Pre-Mix (Graph) |
| ggml_dup | Duplicate Tensor (Graph) |
| ggml_dup_inplace | Duplicate Tensor (Graph) |
| ggml_dup_tensor | Duplicate Tensor |
| ggml_element_size | Get Element Size |
| ggml_elu | ELU Activation (Graph) |
| ggml_elu_inplace | ELU Activation (Graph) |
| ggml_embedding | Create an Embedding Layer Object |
| ggml_estimate_memory | Estimate Required Memory |
| ggml_evaluate | Evaluate a Trained Model |
| ggml_evaluate.ggml_functional_model | Evaluate a Trained Model |
| ggml_evaluate.ggml_sequential_model | Evaluate a Trained Model |
| ggml_exp | Exponential (Graph) |
| ggml_expm1 | Exponential Minus One (Graph) |
| ggml_exp_inplace | Exponential (Graph) |
| ggml_extract | Extract a feature-by-cell matrix from a single-cell container |
| ggml_extract.dgCMatrix | Extract a feature-by-cell matrix from a single-cell container |
| ggml_extract.matrix | Extract a feature-by-cell matrix from a single-cell container |
| ggml_extract.Seurat | Extract a feature-by-cell matrix from a single-cell container |
| ggml_extract.SingleCellExperiment | Extract a feature-by-cell matrix from a single-cell container |
| ggml_fill | Fill (Graph) |
| ggml_fit | Train a Model (dispatcher) |
| ggml_fit.default | Train a Model (dispatcher) |
| ggml_fit.ggml_functional_model | Train a Model (dispatcher) |
| ggml_fit.ggml_sequential_model | Train a Model (dispatcher) |
| ggml_fit_generator | Train a Sequential Model from a Generator |
| ggml_fit_opt | Fit model with R-side epoch loop and callbacks |
| ggml_fit_opt_gen | Train from a generator, one batch at a time |
| ggml_fit_opt_multi | Fit a multi-output model with an R-side epoch loop and callbacks |
| ggml_flash_attn_back | Flash Attention Backward (Graph) |
| ggml_flash_attn_ext | Flash Attention (Graph) |
| ggml_flash_attn_ext_add_sinks | Flash Attention Settings |
| ggml_flash_attn_ext_get_prec | Flash Attention Settings |
| ggml_flash_attn_ext_set_n_kv_max | Flash Attention Settings |
| ggml_flash_attn_ext_set_prec | Flash Attention Settings |
| ggml_floor | Floor (Graph) |
| ggml_floor_inplace | Floor (Graph) |
| ggml_free | Free GGML context |
| ggml_freeze_bn_stats | Fix or Release Batch-Norm Statistics |
| ggml_freeze_weights | Freeze Layer Weights |
| ggml_ftype_to_ggml_type | Type Information |
| ggml_gallocr_alloc_graph | Graph Allocator |
| ggml_gallocr_free | Graph Allocator |
| ggml_gallocr_get_buffer_size | Graph Allocator |
| ggml_gallocr_new | Graph Allocator |
| ggml_gallocr_new_buft | Graph Allocator |
| ggml_gallocr_reserve | Graph Allocator |
| ggml_gated_delta_net | Gated Delta Net (Linear Attention Recurrence) |
| ggml_gated_linear_attn | Gated Linear Attention |
| ggml_gated_linear_attn_back | Backward Pass of Gated Linear Attention |
| ggml_geglu | Gated Linear Unit Operations (Graph) |
| ggml_geglu_quick | Gated Linear Unit Operations (Graph) |
| ggml_geglu_split | Gated Linear Unit Operations (Graph) |
| ggml_gelu | GELU Activation (Graph) |
| ggml_gelu_back | GELU Backward (Graph) |
| ggml_gelu_erf | Exact GELU Activation (Graph) |
| ggml_gelu_inplace | GELU Activation (Graph) |
| ggml_gelu_quick | GELU Quick Activation (Graph) |
| ggml_get_f32 | Get F32 Data |
| ggml_get_f32_nd | N-dimensional Element Access |
| ggml_get_first_tensor | Get First Tensor from Context |
| ggml_get_i32 | Get I32 Data |
| ggml_get_i32_nd | N-dimensional Element Access |
| ggml_get_layer | Get a Layer from a Sequential Model |
| ggml_get_max_tensor_size | Get Maximum Tensor Size |
| ggml_get_mem_size | Get Context Memory Size |
| ggml_get_name | Get Tensor Name |
| ggml_get_next_tensor | Get Next Tensor from Context |
| ggml_get_no_alloc | Get No Allocation Mode |
| ggml_get_n_threads | Get Number of Threads |
| ggml_get_op_params | Operation Parameters |
| ggml_get_op_params_f32 | Operation Parameters |
| ggml_get_op_params_i32 | Operation Parameters |
| ggml_get_rel_pos | Get Relative Position (Graph) |
| ggml_get_rows | Get Rows by Indices (Graph) |
| ggml_get_rows_back | Get Rows Backward (Graph) |
| ggml_get_unary_op | Operation Information |
| ggml_glu | Gated Linear Unit Operations (Graph) |
| GGML_GLU_OP_GEGLU | GLU Operation Types |
| GGML_GLU_OP_GEGLU_ERF | GLU Operation Types |
| GGML_GLU_OP_GEGLU_QUICK | GLU Operation Types |
| GGML_GLU_OP_REGLU | GLU Operation Types |
| GGML_GLU_OP_SWIGLU | GLU Operation Types |
| GGML_GLU_OP_SWIGLU_OAI | GLU Operation Types |
| ggml_glu_split | Gated Linear Unit Operations (Graph) |
| ggml_graph_add_node | Append a Node to a Computation Graph |
| ggml_graph_clear | Clear a Computation Graph |
| ggml_graph_compute | Compute Computation Graph |
| ggml_graph_compute_with_ctx | Compute Graph with Context (Alternative Method) |
| ggml_graph_cpy | Graph Copies and Views |
| ggml_graph_dump_dot | Export Graph to DOT Format |
| ggml_graph_dup | Graph Copies and Views |
| ggml_graph_expand | Add Another Root to an Existing Computation Graph |
| ggml_graph_get_grad | Graph Inspection |
| ggml_graph_get_grad_acc | Graph Inspection |
| ggml_graph_get_tensor | Graph Inspection |
| ggml_graph_node | Graph Inspection |
| ggml_graph_n_nodes | Graph Inspection |
| ggml_graph_order | Add a Tensor to a Graph Without Computing It |
| ggml_graph_overhead | Graph Inspection |
| ggml_graph_print | Print Graph Information |
| ggml_graph_reset | Reset Graph (for backpropagation) |
| ggml_graph_view | Graph Copies and Views |
| ggml_group_norm | Group Normalization (Graph) |
| ggml_group_norm_inplace | Group Normalization (Graph) |
| ggml_gru | Create a GRU Layer Object |
| ggml_hardsigmoid | Hard Sigmoid Activation (Graph) |
| ggml_hardswish | Hard Swish Activation (Graph) |
| GGML_HINT_NONE | Matrix Multiplication Hints |
| GGML_HINT_SRC0_IS_HADAMARD | Matrix Multiplication Hints |
| ggml_im2col | Image to Column (Graph) |
| ggml_im2col_3d | 3D Image to Column (Graph) |
| ggml_im2col_back | Column to Image, the backward of im2col (Graph) |
| ggml_init | Initialize GGML context |
| ggml_init_auto | Create Context with Auto-sizing |
| ggml_inject | Inject a single-cell result back into its container |
| ggml_inject.Seurat | Inject a single-cell result back into its container |
| ggml_inject.SingleCellExperiment | Inject a single-cell result back into its container |
| ggml_input | Declare a Functional API Input Tensor |
| ggml_is_available | Check if GGML is available |
| ggml_is_contiguous | Tensor Contiguity Checks |
| ggml_is_contiguously_allocated | Tensor Contiguity Checks |
| ggml_is_contiguous_0 | Tensor Contiguity Checks |
| ggml_is_contiguous_1 | Tensor Contiguity Checks |
| ggml_is_contiguous_2 | Tensor Contiguity Checks |
| ggml_is_contiguous_channels | Tensor Contiguity Checks |
| ggml_is_contiguous_rows | Tensor Contiguity Checks |
| ggml_is_contiguous_to_1 | Tensor Contiguity Checks |
| ggml_is_contiguous_to_2 | Tensor Contiguity Checks |
| ggml_is_contiguous_to_3 | Tensor Contiguity Checks |
| ggml_is_permuted | Check if Tensor is Permuted |
| ggml_is_quantized | Type Information |
| ggml_is_transposed | Check if Tensor is Transposed |
| ggml_l2_norm | L2 Normalization (Graph) |
| ggml_l2_norm_inplace | L2 Normalization (Graph) |
| ggml_layer_add | Element-wise Addition of Two Tensor Nodes |
| ggml_layer_attention | Add a Multi-Head Attention Layer |
| ggml_layer_batch_norm | Add Batch Normalization Layer |
| ggml_layer_concatenate | Concatenate Tensor Nodes Along an Axis |
| ggml_layer_conv_1d | Add 1D Convolution Layer |
| ggml_layer_conv_2d | Add 2D Convolution Layer |
| ggml_layer_custom | Define a Custom Layer From an R Function |
| ggml_layer_dense | Add Dense (Fully Connected) Layer |
| ggml_layer_dropout | Add Dropout Layer |
| ggml_layer_embedding | Add Embedding Layer |
| ggml_layer_flatten | Add Flatten Layer |
| ggml_layer_global_average_pooling_2d | Global Average Pooling for 2D Feature Maps |
| ggml_layer_global_max_pooling_2d | Global Max Pooling for 2D Feature Maps |
| ggml_layer_gru | Add a GRU Layer |
| ggml_layer_layer_norm | Add Layer Normalization Layer |
| ggml_layer_lstm | Add an LSTM Layer |
| ggml_layer_max_pooling_2d | Add 2D Max Pooling Layer |
| ggml_layer_permute | Add Permute Layer |
| ggml_layer_positional_embedding | Add a Learned Positional Embedding |
| ggml_layer_reshape | Add Reshape Layer |
| ggml_layer_rms_norm | Add RMS Normalization Layer |
| ggml_layer_sequence_pooling | Pool a Sequence Down to One Vector |
| ggml_layer_transformer_block | Add a Transformer Encoder Block |
| ggml_leaky_relu | Leaky ReLU Activation (Graph) |
| ggml_lightning_indexer | DSA Lightning Indexer (Graph) |
| ggml_load_model | Load a Full Model (Architecture + Weights) |
| ggml_load_weights | Load Model Weights from File |
| ggml_log | Natural Logarithm (Graph) |
| ggml_log_inplace | Natural Logarithm (Graph) |
| ggml_log_is_r_enabled | Logging and Abort Callbacks |
| ggml_log_set_default | Logging and Abort Callbacks |
| ggml_log_set_r | Logging and Abort Callbacks |
| ggml_lstm | Create an LSTM Layer Object |
| ggml_marshal_model | Marshal a ggmlR model to an in-memory container |
| ggml_matmul | GPU matrix multiply (drop-in for '%*%') |
| ggml_matmul_f64 | GPU double-precision matrix multiply |
| ggml_matrix-class | A GPU-backed matrix wrapper |
| ggml_mean | Mean (Graph) |
| ggml_meta_backend | Tensor-parallel backend over several devices (meta backend) |
| ggml_model | Create a Functional Model |
| ggml_model_backend | Backend a fitted ggml model actually ran on |
| ggml_model_sequential | Create a Sequential Neural Network Model |
| ggml_mul | Element-wise Multiplication (Graph) |
| ggml_mul_inplace | Element-wise Multiplication (Graph) |
| ggml_mul_mat | Matrix Multiplication (Graph) |
| ggml_mul_mat_id | Matrix Multiplication with Expert Selection (Graph) |
| ggml_mul_mat_set_hint | Set a Matrix Multiplication Hint (Graph) |
| ggml_nbytes | Get Number of Bytes |
| ggml_neg | Negation (Graph) |
| ggml_neg_inplace | Negation (Graph) |
| ggml_nelements | Get Number of Elements |
| ggml_new_f32 | Create Scalar F32 Tensor |
| ggml_new_i32 | Create Scalar I32 Tensor |
| ggml_new_tensor | Create New Tensors |
| ggml_new_tensor_1d | Create New Tensors |
| ggml_new_tensor_2d | Create New Tensors |
| ggml_new_tensor_3d | Create New Tensors |
| ggml_new_tensor_4d | Create New Tensors |
| ggml_norm | Layer Normalization (Graph) |
| ggml_norm_back | Layer Norm Backward (Graph) |
| ggml_norm_inplace | Layer Normalization (Graph) |
| ggml_nrows | Get Number of Rows |
| ggml_n_dims | Get Number of Dimensions |
| GGML_N_TASKS_MAX | Maximum Task Count for Custom Ops |
| ggml_ops_registry | Supported single-cell operations |
| ggml_opt_alloc | Allocate graph for evaluation |
| ggml_opt_context_optimizer_type | Get optimizer type from context |
| ggml_opt_dataset | Optimization datasets |
| ggml_opt_dataset_data | Optimization datasets |
| ggml_opt_dataset_free | Optimization datasets |
| ggml_opt_dataset_get_batch | Optimization datasets |
| ggml_opt_dataset_get_batch_head | Optimization datasets |
| ggml_opt_dataset_init | Optimization datasets |
| ggml_opt_dataset_labels | Optimization datasets |
| ggml_opt_dataset_ndata | Optimization datasets |
| ggml_opt_dataset_shuffle | Optimization datasets |
| ggml_opt_dataset_weights | Optimization datasets |
| ggml_opt_default_params | Get default optimizer parameters |
| ggml_opt_epoch | Run one training epoch |
| ggml_opt_eval | Evaluate model |
| ggml_opt_fit | Fit model to dataset |
| ggml_opt_fit_multi | Fit a multi-output model |
| ggml_opt_free | Free optimizer context |
| ggml_opt_get_lr | Get current learning rate from optimizer context |
| ggml_opt_grad_acc | Get gradient accumulator for a tensor |
| ggml_opt_init | Initialize optimizer context |
| ggml_opt_init_for_fit | Initialize optimizer context for R-side epoch loop |
| ggml_opt_init_for_fit_multi | Initialize a multi-output optimizer context for an R-side epoch loop |
| ggml_opt_inputs | Get inputs tensor from optimizer context |
| ggml_opt_labels | Get labels tensor from optimizer context |
| ggml_opt_labels_i | Get one output head's tensors from an optimizer context |
| ggml_opt_loss | Get loss tensor from optimizer context |
| ggml_opt_loss_i | Get one head's loss scalar, before weighting |
| ggml_opt_loss_type | Loss and optimizer type constants |
| ggml_opt_loss_type_binary_cross_entropy | Loss and optimizer type constants |
| ggml_opt_loss_type_cross_entropy | Loss and optimizer type constants |
| ggml_opt_loss_type_huber | Loss and optimizer type constants |
| ggml_opt_loss_type_mae | Loss and optimizer type constants |
| ggml_opt_loss_type_mean | Loss and optimizer type constants |
| ggml_opt_loss_type_mse | Loss and optimizer type constants |
| ggml_opt_loss_type_sum | Loss and optimizer type constants |
| ggml_opt_loss_type_weighted_mse | Loss and optimizer type constants |
| ggml_opt_loss_weights_i | Get one output head's tensors from an optimizer context |
| ggml_opt_ncorrect | Get number of correct predictions tensor |
| ggml_opt_ncorrect_i | Get one output head's tensors from an optimizer context |
| ggml_opt_n_loss | Number of output heads in an optimizer context |
| ggml_opt_optimizer_name | Get optimizer name |
| ggml_opt_optimizer_type_adamw | Loss and optimizer type constants |
| ggml_opt_optimizer_type_sgd | Loss and optimizer type constants |
| ggml_opt_outputs | Get outputs tensor from optimizer context |
| ggml_opt_outputs_i | Get one output head's tensors from an optimizer context |
| ggml_opt_pred | Get predictions tensor from optimizer context |
| ggml_opt_pred_i | Get one output head's tensors from an optimizer context |
| ggml_opt_prepare_alloc | Prepare allocation for non-static graphs |
| ggml_opt_reset | Reset optimizer context |
| ggml_opt_result | Optimization results |
| ggml_opt_result_accuracy | Optimization results |
| ggml_opt_result_accuracy_i | Optimization results |
| ggml_opt_result_free | Optimization results |
| ggml_opt_result_init | Optimization results |
| ggml_opt_result_loss | Optimization results |
| ggml_opt_result_loss_i | Optimization results |
| ggml_opt_result_ndata | Optimization results |
| ggml_opt_result_n_loss | Optimization results |
| ggml_opt_result_pred | Optimization results |
| ggml_opt_result_reset | Optimization results |
| ggml_opt_set_lr | Set learning rate in optimizer context |
| ggml_opt_static_graphs | Check if using static graphs |
| ggml_opt_step_adamw | AdamW Optimizer Step (Graph Node) |
| ggml_op_can_inplace | Check if Operation Can Be Done In-place |
| ggml_op_desc | Operation Information |
| ggml_op_name | Operation Information |
| GGML_OP_POOL_AVG | 1D Pooling (Graph) |
| GGML_OP_POOL_MAX | 1D Pooling (Graph) |
| ggml_op_symbol | Operation Information |
| ggml_out_prod | Outer Product (Graph) |
| ggml_out_prod_rowsum | Outer Product with Row Sums (Graph) |
| ggml_pad | Pad Tensor with Zeros (Graph) |
| ggml_pad_ext | Pad a tensor with zeros on both sides |
| ggml_pad_reflect_1d | Reflective 1D Padding (Graph) |
| ggml_permute | Permute Tensor Dimensions (Graph) |
| ggml_pool_1d | 1D Pooling (Graph) |
| ggml_pool_2d | 2D Pooling (Graph) |
| ggml_pool_2d_back | 2D Pooling Backward (Graph) |
| ggml_pop_layer | Remove the Last Layer from a Sequential Model |
| ggml_pp_dp_forward | Pipeline + data parallel forward pass (PP x DP) |
| ggml_pp_forward | Pipeline-parallel forward pass across per-device layer stages |
| ggml_predict | Get Predictions from a Trained Model |
| ggml_predict.ggml_functional_model | Get Predictions from a Trained Model |
| ggml_predict.ggml_sequential_model | Get Predictions from a Trained Model |
| ggml_predict_classes | Predict Classes from a Trained Model |
| ggml_print_mem_status | Print Context Memory Status |
| ggml_print_objects | Print Objects in Context |
| ggml_quantize_chunk | Quantize Data Chunk |
| ggml_quantize_free | Free Quantization Resources |
| ggml_quantize_init | Initialize Quantization Tables |
| ggml_quantize_requires_imatrix | Check if Quantization Requires Importance Matrix |
| ggml_quant_block_info | Get Quantization Block Info |
| ggml_rand | Counter-Based Random Numbers (Graph) |
| ggml_reglu | Gated Linear Unit Operations (Graph) |
| ggml_reglu_split | Gated Linear Unit Operations (Graph) |
| ggml_relu | ReLU Activation (Graph) |
| ggml_relu_inplace | ReLU Activation (Graph) |
| ggml_repeat | Repeat (Graph) |
| ggml_repeat_back | Repeat Backward (Graph) |
| ggml_reset | Reset GGML Context |
| ggml_reshape_1d | Reshape Tensor (Graph) |
| ggml_reshape_2d | Reshape Tensor (Graph) |
| ggml_reshape_3d | Reshape Tensor (Graph) |
| ggml_reshape_4d | Reshape Tensor (Graph) |
| ggml_result | Construct a single-cell result |
| ggml_rms_norm | RMS Normalization (Graph) |
| ggml_rms_norm_back | RMS Norm Backward (Graph) |
| ggml_rms_norm_inplace | RMS Normalization (Graph) |
| ggml_roll | Roll (Graph) |
| ggml_rope | Rotary Position Embedding (Graph) |
| ggml_rope_ext | Rotary Position Embedding (Graph) |
| ggml_rope_ext_back | Rotary Position Embedding (Graph) |
| ggml_rope_ext_inplace | Rotary Position Embedding (Graph) |
| ggml_rope_inplace | Rotary Position Embedding (Graph) |
| ggml_rope_multi | Rotary Position Embedding (Graph) |
| ggml_rope_multi_back | Rotary Position Embedding (Graph) |
| ggml_rope_multi_inplace | Rotary Position Embedding (Graph) |
| ggml_rope_set_offset | Rotary Position Embedding (Graph) |
| GGML_ROPE_TYPE_MROPE | RoPE Mode Constants |
| GGML_ROPE_TYPE_NEOX | RoPE Mode Constants |
| GGML_ROPE_TYPE_NORM | RoPE Mode Constants |
| GGML_ROPE_TYPE_VISION | RoPE Mode Constants |
| ggml_round | Round (Graph) |
| ggml_round_even | Round Half to Even (Graph) |
| ggml_round_inplace | Round (Graph) |
| ggml_run | Run a single-cell task on the GGML backend |
| ggml_rwkv_output | Views Into a Packed RWKV / GLA Result |
| ggml_rwkv_state | Views Into a Packed RWKV / GLA Result |
| ggml_rwkv_wkv6 | RWKV-6 WKV Recurrence |
| ggml_rwkv_wkv6_back | Backward Pass of the RWKV-6 WKV Recurrence |
| ggml_rwkv_wkv7 | RWKV-7 WKV Recurrence |
| ggml_rwkv_wkv7_back | Backward Pass of the RWKV-7 WKV Recurrence |
| ggml_save_model | Save a Full Model (Architecture + Weights) |
| ggml_save_weights | Save Model Weights to File |
| ggml_scale | Scale (Graph) |
| ggml_scale_bias | Scale and Shift by Constants (Graph) |
| ggml_scale_inplace | Scale (Graph) |
| GGML_SCALE_MODE_BICUBIC | Upscale Tensor (Graph) |
| GGML_SCALE_MODE_BILINEAR | Upscale Tensor (Graph) |
| GGML_SCALE_MODE_NEAREST | Upscale Tensor (Graph) |
| ggml_scatter_elements | Scatter elements into a copy of a tensor (ONNX ScatterElements) |
| ggml_schedule_cosine_decay | Cosine annealing LR scheduler |
| ggml_schedule_reduce_on_plateau | Reduce on plateau LR scheduler |
| ggml_schedule_step_decay | Step decay LR scheduler |
| ggml_set | Set Tensor Region (Graph) |
| ggml_set_1d | Set 1D Tensor Region (Graph) |
| ggml_set_2d | Set 2D Tensor Region (Graph) |
| ggml_set_abort_callback_default | Logging and Abort Callbacks |
| ggml_set_abort_callback_r | Logging and Abort Callbacks |
| ggml_set_f32 | Set F32 Data |
| ggml_set_f32_nd | N-dimensional Element Access |
| ggml_set_i32 | Set I32 Data |
| ggml_set_i32_nd | N-dimensional Element Access |
| ggml_set_input | Tensor Role Flags |
| ggml_set_loss | Tensor Role Flags |
| ggml_set_name | Set Tensor Name |
| ggml_set_no_alloc | Set No Allocation Mode |
| ggml_set_n_threads | Set Number of Threads |
| ggml_set_omp_threads | Set OpenMP Thread Count |
| ggml_set_op_params | Operation Parameters |
| ggml_set_op_params_f32 | Operation Parameters |
| ggml_set_op_params_i32 | Operation Parameters |
| ggml_set_output | Tensor Role Flags |
| ggml_set_param | Tensor Role Flags |
| ggml_set_rows | Set Rows by Index (Graph) |
| ggml_set_seed | Set the random seed for reproducible ggmlR runs |
| ggml_set_zero | Set Tensor to Zero |
| ggml_sgn | Sign Function (Graph) |
| ggml_sigmoid | Sigmoid Activation (Graph) |
| ggml_sigmoid_inplace | Sigmoid Activation (Graph) |
| ggml_silu | SiLU Activation (Graph) |
| ggml_silu_back | SiLU Backward (Graph) |
| ggml_silu_inplace | SiLU Activation (Graph) |
| ggml_sin | Sine (Graph) |
| ggml_snapshot_free | Release a Snapshot Made by a Trainer |
| ggml_softplus | Softplus Activation (Graph) |
| ggml_softplus_inplace | Softplus Activation (Graph) |
| ggml_soft_max | Softmax Operations (Graph) |
| ggml_soft_max_add_sinks | Softmax Operations (Graph) |
| ggml_soft_max_ext | Softmax Operations (Graph) |
| ggml_soft_max_ext_back | Softmax Operations (Graph) |
| ggml_soft_max_ext_back_inplace | Softmax Operations (Graph) |
| ggml_soft_max_ext_inplace | Softmax Operations (Graph) |
| ggml_soft_max_inplace | Softmax Operations (Graph) |
| GGML_SORT_ORDER_ASC | Sort Order Constants |
| GGML_SORT_ORDER_DESC | Sort Order Constants |
| GGML_SPLIT_AXIS | Split-axis codes for the meta backend |
| GGML_SPLIT_AXIS_0 | Split-axis codes for the meta backend |
| GGML_SPLIT_AXIS_1 | Split-axis codes for the meta backend |
| GGML_SPLIT_AXIS_2 | Split-axis codes for the meta backend |
| GGML_SPLIT_AXIS_3 | Split-axis codes for the meta backend |
| GGML_SPLIT_AXIS_MIRRORED | Split-axis codes for the meta backend |
| GGML_SPLIT_AXIS_PARTIAL | Split-axis codes for the meta backend |
| ggml_sqr | Square (Graph) |
| ggml_sqrt | Square Root (Graph) |
| ggml_sqrt_inplace | Square Root (Graph) |
| ggml_sqr_inplace | Square (Graph) |
| ggml_ssm_conv | State-Space Convolution (Mamba) |
| ggml_ssm_conv_back | Backward Pass of the State-Space Convolution |
| ggml_ssm_scan | Selective State-Space Scan (Mamba / Mamba-2) |
| ggml_ssm_scan_back | Backward Pass of the Selective State-Space Scan |
| ggml_ssm_scan_output | Views Into a Packed ssm_scan Result |
| ggml_ssm_scan_state | Views Into a Packed ssm_scan Result |
| ggml_step | Step Function (Graph) |
| ggml_sub | Element-wise Subtraction (Graph) |
| ggml_sub_inplace | Element-wise Subtraction (Graph) |
| ggml_sum | Sum (Graph) |
| ggml_sum_rows | Sum Rows (Graph) |
| ggml_swiglu | Gated Linear Unit Operations (Graph) |
| ggml_swiglu_clamp | Gated Linear Unit Operations (Graph) |
| ggml_swiglu_split | Gated Linear Unit Operations (Graph) |
| ggml_tanh | Tanh Activation (Graph) |
| ggml_tanh_inplace | Tanh Activation (Graph) |
| ggml_task | Construct a single-cell compute task |
| ggml_tcrossprod | GPU transposed cross product (drop-in for 'tcrossprod') |
| ggml_tensor_copy | Copy Tensor Data |
| ggml_tensor_data_ptr | Raw Data Pointer of a Tensor |
| ggml_tensor_nb | Get Tensor Strides (nb) |
| ggml_tensor_num | Count Tensors in Context |
| ggml_tensor_overhead | Get Tensor Overhead |
| ggml_tensor_set_f32_scalar | Fill Tensor with Scalar |
| ggml_tensor_shape | Get Tensor Shape |
| ggml_tensor_type | Get Tensor Type |
| ggml_test | Test GGML |
| ggml_test_adamw_steps | Compare the AdamW Optimizer Step Across Backends |
| ggml_test_backend_ops | Compare Backend Operations Against the CPU |
| ggml_timestep_embedding | Timestep Embedding (Graph Operation) |
| ggml_time_init | Timing Utilities |
| ggml_time_ms | Timing Utilities |
| ggml_time_us | Timing Utilities |
| ggml_top_k | Top-K Indices (Graph) |
| ggml_tp_dp_forward | TPxDP hybrid matrix multiply across replicas of Vulkan device groups |
| ggml_tp_forward | Tensor-parallel forward pass of a whole graph (meta backend) |
| ggml_trainer | A Training Context You Step Yourself |
| ggml_training_history | Training history of a fitted ggml model |
| ggml_transpose | Transpose (Graph) |
| ggml_trunc | Truncate (Graph) |
| GGML_TYPE_BF16 | GGML Data Types |
| GGML_TYPE_F16 | GGML Data Types |
| GGML_TYPE_F32 | GGML Data Types |
| GGML_TYPE_F64 | GGML Data Types |
| GGML_TYPE_I32 | GGML Data Types |
| GGML_TYPE_I64 | GGML Data Types |
| ggml_type_name | Type Information |
| GGML_TYPE_Q2_K | GGML Data Types |
| GGML_TYPE_Q3_K | GGML Data Types |
| GGML_TYPE_Q4_0 | GGML Data Types |
| GGML_TYPE_Q4_1 | GGML Data Types |
| GGML_TYPE_Q4_K | GGML Data Types |
| GGML_TYPE_Q5_K | GGML Data Types |
| GGML_TYPE_Q6_K | GGML Data Types |
| GGML_TYPE_Q8_0 | GGML Data Types |
| ggml_type_size | Type Information |
| ggml_type_sizef | Type Information |
| ggml_unary_op_name | Operation Information |
| ggml_unfreeze_bn_stats | Fix or Release Batch-Norm Statistics |
| ggml_unfreeze_weights | Unfreeze Layer Weights |
| ggml_unmarshal_model | Unmarshal a ggmlR model from an in-memory container |
| ggml_upscale | Upscale Tensor (Graph) |
| ggml_used_mem | Get Used Memory |
| ggml_version | Get GGML version |
| ggml_view_1d | Tensor Views (Graph) |
| ggml_view_2d | Tensor Views (Graph) |
| ggml_view_3d | Tensor Views (Graph) |
| ggml_view_4d | Tensor Views (Graph) |
| ggml_view_5d | Tensor Views (Graph) |
| ggml_view_tensor | View Tensor |
| ggml_vulkan_available | Check if Vulkan support is available |
| ggml_vulkan_backend_name | Vulkan Backend Lifecycle |
| ggml_vulkan_device_caps | Vulkan Device Information |
| ggml_vulkan_device_count | Vulkan Device Information |
| ggml_vulkan_device_description | Vulkan Device Information |
| ggml_vulkan_device_groups | Probe Vulkan device groups (NVLink / multi-GPU peer access) |
| ggml_vulkan_device_memory | Vulkan Device Information |
| ggml_vulkan_free | Vulkan Backend Lifecycle |
| ggml_vulkan_hard_exit_available | Is the Vulkan hard-exit path compiled in? |
| ggml_vulkan_init | Vulkan Backend Lifecycle |
| ggml_vulkan_is_backend | Vulkan Backend Lifecycle |
| ggml_vulkan_list_devices | Vulkan Device Information |
| ggml_vulkan_p2p_selftest | Opaque-fd device-to-device P2P self-test |
| ggml_vulkan_shutdown | Explicitly tear down the Vulkan instance and devices |
| ggml_vulkan_split_buffer_type | Create a Vulkan tensor-split buffer type |
| ggml_vulkan_split_mul_mat | Tensor-parallel matrix multiply across Vulkan devices |
| ggml_vulkan_split_row_ranges | Compute tensor-parallel row-split ranges |
| ggml_vulkan_status | Print Vulkan status |
| ggml_win_part | Window Partition (Graph) |
| ggml_win_unpart | Window Un-partition (Graph) |
| ggml_with_temp_ctx | Execute with Temporary Context |
| ggml_xielu | xIELU Activation (Graph) |
| gguf_free | Free GGUF Resources |
| gguf_load | Load a GGUF File |
| gguf_load_raw | Load a GGUF Image from Memory |
| gguf_metadata | Get GGUF Metadata |
| gguf_tensor_data | Extract Tensor Data |
| gguf_tensor_info | Get Tensor Info |
| gguf_tensor_names | List Tensor Names in a GGUF File |
| gguf_tensor_ne | Get Tensor Shape |
| glance.ggmlr_parsnip_model | One-row summary of a fitted ggml parsnip model |
| iq2xs_free_impl | Free IQ2 Quantization Tables |
| iq2xs_init_impl | Initialize IQ2 Quantization Tables |
| iq3xs_free_impl | Free IQ3 Quantization Tables |
| iq3xs_init_impl | Initialize IQ3 Quantization Tables |
| lr_scheduler_cosine | Cosine-annealing learning rate scheduler |
| lr_scheduler_cyclic | Cyclical learning rate scheduler |
| lr_scheduler_onecycle | One-cycle learning rate scheduler |
| lr_scheduler_step | Step-decay learning rate scheduler |
| lr_scheduler_warmup_cosine | Linear warmup followed by cosine annealing |
| mul_mat_hints | Matrix Multiplication Hints |
| nn_topo_sort | Topologically sort nodes reachable from output nodes |
| onnx_device_info | ONNX model device/scheduler diagnostics |
| onnx_graph_info | ONNX graph structure |
| onnx_inputs | List ONNX model inputs |
| onnx_load | Load an ONNX model |
| onnx_run | Run ONNX model inference |
| onnx_summary | ONNX model summary |
| optimizer_adam | Autograd Optimizers |
| optimizer_adamw | Autograd Optimizers |
| optimizer_rmsprop | Create an RMSprop optimizer |
| optimizer_sgd | Autograd Optimizers |
| plot.ggml_history | Plot training history |
| predict.ggml_functional_model | Predict with a Trained Model |
| predict.ggml_sequential_model | Predict with a Trained Model |
| predict.onnx_model | Predict with an ONNX Model |
| print.ag_tensor | Print method for ag_tensor |
| print.ggml_functional_model | Print method for ggml_functional_model |
| print.ggml_history | Print method for ggml_history |
| print.ggml_sequential_model | Print method for ggml_sequential_model |
| print.onnx_model | Print ONNX model summary |
| pth_catalogue | List the tensors in a PyTorch checkpoint |
| pth_load | Read the tensors of a PyTorch checkpoint |
| pt_cube4_layout | The piece layout of a 4x4x4 cube |
| pt_forward | Run a piece transformer |
| pt_transformer | Build a piece transformer |
| pt_transformer_load | Build a piece transformer and load a checkpoint into it |
| quantize_data | Quantize data |
| quantize_iq1_m | Quantize data |
| quantize_iq1_s | Quantize data |
| quantize_iq2_s | Quantize data |
| quantize_iq2_xs | Quantize data |
| quantize_iq2_xxs | Quantize data |
| quantize_iq3_s | Quantize data |
| quantize_iq3_xxs | Quantize data |
| quantize_iq4_nl | Quantize data |
| quantize_iq4_xs | Quantize data |
| quantize_mxfp4 | Quantize data |
| quantize_nvfp4 | Quantize data |
| quantize_q1_0 | Quantize data |
| quantize_q2_0 | Quantize data |
| quantize_q2_K | Quantize data |
| quantize_q3_K | Quantize data |
| quantize_q4_0 | Quantize data |
| quantize_q4_1 | Quantize data |
| quantize_q4_K | Quantize data |
| quantize_q5_0 | Quantize data |
| quantize_q5_1 | Quantize data |
| quantize_q5_K | Quantize data |
| quantize_q6_K | Quantize data |
| quantize_q8_0 | Quantize data |
| quantize_row_iq2_s_ref | Quantize row (reference implementations) |
| quantize_row_iq3_s_ref | Quantize row (reference implementations) |
| quantize_row_iq3_xxs_ref | Quantize row (reference implementations) |
| quantize_row_iq4_nl_ref | Quantize row (reference implementations) |
| quantize_row_iq4_xs_ref | Quantize row (reference implementations) |
| quantize_row_mxfp4_ref | Quantize row (reference implementations) |
| quantize_row_nvfp4_ref | Quantize row (reference implementations) |
| quantize_row_q1_0_ref | Quantize row (reference implementations) |
| quantize_row_q2_0_ref | Quantize row (reference implementations) |
| quantize_row_q2_K_ref | Quantize row (reference implementations) |
| quantize_row_q3_K_ref | Quantize row (reference implementations) |
| quantize_row_q4_0_ref | Quantize row (reference implementations) |
| quantize_row_q4_1_ref | Quantize row (reference implementations) |
| quantize_row_q4_K_ref | Quantize row (reference implementations) |
| quantize_row_q5_0_ref | Quantize row (reference implementations) |
| quantize_row_q5_1_ref | Quantize row (reference implementations) |
| quantize_row_q5_K_ref | Quantize row (reference implementations) |
| quantize_row_q6_K_ref | Quantize row (reference implementations) |
| quantize_row_q8_0_ref | Quantize row (reference implementations) |
| quantize_row_q8_1_ref | Quantize row (reference implementations) |
| quantize_row_q8_K_ref | Quantize row (reference implementations) |
| quantize_row_ref | Quantize row (reference implementations) |
| quantize_row_tq1_0_ref | Quantize row (reference implementations) |
| quantize_row_tq2_0_ref | Quantize row (reference implementations) |
| quantize_tq1_0 | Quantize data |
| quantize_tq2_0 | Quantize data |
| rope_types | RoPE Mode Constants |
| RunGGML | Run a GGML GPU operation on a Seurat object |
| RunGGML.default | Run a GGML GPU operation on a Seurat object |
| RunGGML.Seurat | Run a GGML GPU operation on a Seurat object |
| RunGGML.SingleCellExperiment | Run a GGML GPU operation on a Seurat object |
| show-method | A GPU-backed matrix wrapper |
| summary.ggml_sequential_model | Summary method for ggml_sequential_model |
| tcrossprod-method | A GPU-backed matrix wrapper |
| tidy.ggmlr_parsnip_model | Tidy a fitted ggml parsnip model into a per-layer table |
| with_grad_tape | Run code with gradient tape enabled |
| %*%-method | A GPU-backed matrix wrapper |