Index

_

__call__() (transformer_engine.jax.flax.DenseGeneral method) (transformer_engine.jax.flax.DotProductAttention method) (transformer_engine.jax.flax.LayerNorm method) (transformer_engine.jax.flax.LayerNormDenseGeneral method) (transformer_engine.jax.flax.LayerNormMLP method) (transformer_engine.jax.flax.MultiHeadAttention method) (transformer_engine.jax.flax.RelativePositionBiases method) (transformer_engine.jax.flax.TransformerLayer method)

_functional_backward() (transformer_engine.pytorch.ops.BasicLinear static method) _functional_forward() (transformer_engine.pytorch.ops.BasicLinear static method)

A

add() (transformer_engine.pytorch.CudaRNGStatesTracker method) AddExtraInput (class in transformer_engine.pytorch.ops) AllGather (class in transformer_engine.pytorch.ops) AllReduce (class in transformer_engine.pytorch.ops) AttentionShape (C++ class) AttentionShape::AttentionShape (C++ function) AttentionShape::b (C++ function) AttentionShape::canonical_ (C++ member)

AttentionShape::d (C++ function) AttentionShape::dim_order (C++ function) AttentionShape::h (C++ function) AttentionShape::s (C++ function) AttentionShape::t (C++ function) AttentionShape::to_format (C++ function) autocast (class in transformer_engine.pytorch) autocast() (in module transformer_engine.jax)

B

BasicLinear (class in transformer_engine.pytorch.ops)

BasicOperation (class in transformer_engine.pytorch.ops) Bias (class in transformer_engine.pytorch.ops)

C

checkpoint() (in module transformer_engine.pytorch) ClampedSwiGLU (class in transformer_engine.pytorch.ops) ConstantScale (class in transformer_engine.pytorch.ops)

cp_resource (transformer_engine.jax.MeshResource attribute) CudaRNGStatesTracker (class in transformer_engine.pytorch) CustomRecipe (class in transformer_engine.common.recipe)

D

deinterleave_glu_tensor() (in module transformer_engine.pytorch) DelayedScaling (class in transformer_engine.common.recipe) DenseGeneral (class in transformer_engine.jax.flax) dequantize() (transformer_engine.pytorch.QuantizedTensor method) destroy_ub() (in module transformer_engine.pytorch) DisableFP8GEMM (class in transformer_engine.debug.features.disable_fp8_gemm) DisableFP8Layer (class in transformer_engine.debug.features.disable_fp8_layer)

DisableQuantizationGEMM (class in transformer_engine.debug.features.disable_quantization_gemm) DisableQuantizationLayer (class in transformer_engine.debug.features.disable_quantization_layer) DotProductAttention (class in transformer_engine.jax.flax) (class in transformer_engine.pytorch) dp_resource (transformer_engine.jax.MeshResource attribute) Dropout (class in transformer_engine.pytorch.ops) DType (class in transformer_engine.pytorch) DumpTensors (class in transformer_engine.debug.features.dump_tensors)

E

environment variable NVTE_AG_P2P_MULTI_ATOMIC NVTE_ALLOW_NONDETERMINISTIC_ALGO NVTE_APPLY_QK_LAYER_SCALING NVTE_BATCH_MHA_P2P_COMM NVTE_BIAS_DROPOUT_FUSION NVTE_BIAS_GELU_NVFUSION NVTE_BUILD_ACTIVATION_WITH_FAST_MATH NVTE_BUILD_DEBUG NVTE_BUILD_MAX_JOBS NVTE_BUILD_NUM_PHILOX_ROUNDS NVTE_BUILD_THREADS_PER_JOB NVTE_BWD_LAYERNORM_SM_MARGIN NVTE_CCACHE_BIN NVTE_CMAKE_BUILD_DIR NVTE_CPU_OFFLOAD_V1 NVTE_CUDA_ARCHS NVTE_CUDA_INCLUDE_DIR NVTE_CUDNN_MXFP8_NORM_OUTPUT_IN_INPUT_DTYPE NVTE_CUTLASS_GROUPED_GEMM_WARN_FALLBACK NVTE_DEBUG, [1] NVTE_DEBUG_LEVEL NVTE_DEBUG_NUMERICS NVTE_DISABLE_NVRTC NVTE_DPA_FP8_FORMAT NVTE_DPA_FP8_RECIPE, [1], [2], [3], [4], [5], [6], [7], [8] NVTE_DPA_FP8_RECIPE_DPA NVTE_DPA_FP8_RECIPE_MHA NVTE_DPA_FP8CS_O_in_F16 NVTE_DPA_FP8DS_AMAX_ALGO NVTE_DPA_FP8DS_AMAX_HISTLEN NVTE_DPA_FP8DS_REDUCE_AMAX NVTE_ENABLE_NVSHMEM NVTE_EXT_MARGIN_SM NVTE_FLASH_ATTN NVTE_FLASH_ATTN_V2 NVTE_FLASH_ATTN_V3 NVTE_FLASH_ATTN_V4 NVTE_FP8_DPA_BWD NVTE_FRAMEWORK NVTE_FUSED_ATTN NVTE_FUSED_ATTN_BACKEND NVTE_FUSED_ATTN_USE_FAv2_BWD NVTE_FUSED_RING_ATTENTION_USE_SCAN NVTE_FWD_LAYERNORM_SM_MARGIN NVTE_GROUPED_TENSOR_HANDLE_POOL_SIZE_MB NVTE_INF_LAYERNORM_SM_MARGIN NVTE_JAX_CUSTOM_CALLS, [1] NVTE_JAX_CUSTOM_CALLS_RE NVTE_JAX_ENFORCE_TRITON_AUTOTUNING NVTE_JAX_UNITTEST_LEVEL NVTE_NVFP4_4OVER6, [1] NVTE_NVFP4_4OVER6_E4M3_USE_256 NVTE_NVFP4_4OVER6_ERR_MODE NVTE_NVFP4_4OVER6_ERR_USE_FAST_MATH NVTE_NVFP4_ROW_SCALED_ACTIVATION NVTE_NVTX_ENABLED NVTE_ONNX_KVCACHE_MAX_SEQ_LEN NVTE_PRINT_LAYER_NUMBER NVTE_PRINT_RANK NVTE_PROJECT_BUILDING NVTE_RELEASE_BUILD NVTE_TENSOR_HANDLE_POOL_SIZE_MB NVTE_TEST_ARTIFACTS_DIR NVTE_TEST_NVINSPECT_CONFIG_FILE NVTE_TEST_NVINSPECT_ENABLED, [1], [2] NVTE_TEST_NVINSPECT_FEATURE_DIRS NVTE_TORCH_COMPILE NVTE_UB_WITH_MPI NVTE_UNFUSED_ATTN NVTE_UNFUSED_FP8_UPDATE NVTE_UnfusedDPA_Emulate_FP8 NVTE_USE_CCACHE, [1] NVTE_USE_CUTLASS_GROUPED_GEMM NVTE_USE_FAST_MATH NVTE_USE_PYTORCH_TRITON

ep_resource (transformer_engine.jax.MeshResource attribute) extend_logical_axis_rules() (in module transformer_engine.jax.flax)

F

FakeQuant (class in transformer_engine.debug.features.fake_quant) Float8BlockQuantizer (class in transformer_engine.pytorch) Float8BlockScaling (class in transformer_engine.common.recipe) Float8BlockwiseQTensor (class in transformer_engine.pytorch) Float8BlockwiseQTensorStorage (class in transformer_engine.pytorch) Float8CurrentScaling (class in transformer_engine.common.recipe) Float8CurrentScalingQuantizer (class in transformer_engine.pytorch) Float8Quantizer (class in transformer_engine.pytorch) Float8Tensor (class in transformer_engine.pytorch) Float8TensorStorage (class in transformer_engine.pytorch) fork() (transformer_engine.pytorch.CudaRNGStatesTracker method) Format (class in transformer_engine.common.recipe) forward() (transformer_engine.pytorch.DotProductAttention method) (transformer_engine.pytorch.GroupedLinear method) (transformer_engine.pytorch.LayerNormLinear method) (transformer_engine.pytorch.LayerNormMLP method) (transformer_engine.pytorch.Linear method) (transformer_engine.pytorch.MultiheadAttention method) (transformer_engine.pytorch.ops.Sequential method) (transformer_engine.pytorch.TransformerLayer method)

FP8 (transformer_engine.pytorch.UserBufferQuantizationMode attribute) fp8_autocast() (in module transformer_engine.jax) (in module transformer_engine.pytorch) fp8_gemm_enabled() (in module transformer_engine.debug.features.api.TEDefaultFeatures) fp8_model_init() (in module transformer_engine.pytorch) fsdp_resource (transformer_engine.jax.MeshResource attribute) FusedOperation (class in transformer_engine.pytorch.ops) fuser_backward() (transformer_engine.pytorch.ops.FusedOperation method) (transformer_engine.pytorch.ops.FusibleOperation method) fuser_forward() (transformer_engine.pytorch.ops.FusedOperation method) (transformer_engine.pytorch.ops.FusibleOperation method) FusibleOperation (class in transformer_engine.pytorch.ops)

G

GEGLU (class in transformer_engine.pytorch.ops) GELU (class in transformer_engine.pytorch.ops) get_cpu_offload_context() (in module transformer_engine.pytorch) get_cudnn_version() (in module transformer_engine.pytorch) get_default_recipe() (in module transformer_engine.pytorch)

get_device_compute_capability() (in module transformer_engine.pytorch) get_states() (transformer_engine.pytorch.CudaRNGStatesTracker method) GLU (class in transformer_engine.pytorch.ops) GroupedLinear (class in transformer_engine.pytorch) (class in transformer_engine.pytorch.ops)

I

Identity (class in transformer_engine.pytorch.ops) initialize_ub() (in module transformer_engine.pytorch) inspect_tensor() (in module transformer_engine.debug.features.api.TEDefaultFeatures) inspect_tensor_enabled() (in module transformer_engine.debug.features.api.TEDefaultFeatures) inspect_tensor_postquantize() (in module transformer_engine.debug.features.api.TEDefaultFeatures) inspect_tensor_postquantize_enabled() (in module transformer_engine.debug.features.api.TEDefaultFeatures)

interleave_glu_tensor() (in module transformer_engine.pytorch) is_bf16_available() (in module transformer_engine.pytorch) is_fp8_available() (in module transformer_engine.pytorch) is_fp8_block_scaling_available() (in module transformer_engine.pytorch) is_mxfp8_available() (in module transformer_engine.pytorch) is_nvfp4_available() (in module transformer_engine.pytorch)

K

kBFloat16 (transformer_engine.pytorch.DType attribute) kByte (transformer_engine.pytorch.DType attribute) kFloat16 (transformer_engine.pytorch.DType attribute) kFloat32 (transformer_engine.pytorch.DType attribute)

kFloat4E2M1 (transformer_engine.pytorch.DType attribute) kFloat8E4M3 (transformer_engine.pytorch.DType attribute) kFloat8E5M2 (transformer_engine.pytorch.DType attribute) kInt32 (transformer_engine.pytorch.DType attribute)

L

L2Normalization (class in transformer_engine.pytorch.ops) LayerNorm (class in transformer_engine.jax.flax) (class in transformer_engine.pytorch) (class in transformer_engine.pytorch.ops) LayerNormDenseGeneral (class in transformer_engine.jax.flax) LayerNormLinear (class in transformer_engine.pytorch)

LayerNormMLP (class in transformer_engine.jax.flax) (class in transformer_engine.pytorch) Linear (class in transformer_engine.pytorch) (class in transformer_engine.pytorch.ops) LogFp8TensorStats (class in transformer_engine.debug.features.log_fp8_tensor_stats) LogNvfp4TensorStats (class in transformer_engine.debug.features.log_nvfp4_tensor_stats) LogTensorStats (class in transformer_engine.debug.features.log_tensor_stats)

M

make_graphed_callables() (in module transformer_engine.pytorch) MakeExtraOutput (class in transformer_engine.pytorch.ops) ManualOffloadSynchronizer (class in transformer_engine.pytorch) mark_not_offload() (in module transformer_engine.pytorch) MeshResource (class in transformer_engine.jax) mhc_fused_aggregate() (in module transformer_engine.pytorch.triton.mhc) mhc_fused_expand_combine() (in module transformer_engine.pytorch.triton.mhc) mhc_fused_projection() (in module transformer_engine.pytorch.triton.mhc) mhc_fused_scale() (in module transformer_engine.pytorch.triton.mhc) mhc_fused_sinkhorn() (in module transformer_engine.pytorch.triton.mhc) modify_tensor() (in module transformer_engine.debug.features.api.TEDefaultFeatures)

modify_tensor_enabled() (in module transformer_engine.debug.features.api.TEDefaultFeatures) moe_permute() (in module transformer_engine.pytorch) moe_permute_with_probs() (in module transformer_engine.pytorch) moe_sort_chunks_by_index() (in module transformer_engine.pytorch) moe_sort_chunks_by_index_with_probs() (in module transformer_engine.pytorch) moe_unpermute() (in module transformer_engine.pytorch) MultiHeadAttention (class in transformer_engine.jax.flax) MultiheadAttention (class in transformer_engine.pytorch) MXFP8BlockScaling (class in transformer_engine.common.recipe) MXFP8Quantizer (class in transformer_engine.pytorch) MXFP8Tensor (class in transformer_engine.pytorch) MXFP8TensorStorage (class in transformer_engine.pytorch)

N

NONE (transformer_engine.pytorch.UserBufferQuantizationMode attribute) NVFP4BlockScaling (class in transformer_engine.common.recipe) NVFP4Quantizer (class in transformer_engine.pytorch) NVFP4Tensor (class in transformer_engine.pytorch) NVFP4TensorStorage (class in transformer_engine.pytorch) NVTE_Activation_Type (C++ enum) NVTE_Activation_Type::CLAMPED_SWIGLU (C++ enumerator) NVTE_Activation_Type::GEGLU (C++ enumerator) NVTE_Activation_Type::GELU (C++ enumerator) NVTE_Activation_Type::GLU (C++ enumerator) NVTE_Activation_Type::QGEGLU (C++ enumerator) NVTE_Activation_Type::QGELU (C++ enumerator) NVTE_Activation_Type::REGLU (C++ enumerator) NVTE_Activation_Type::RELU (C++ enumerator) NVTE_Activation_Type::SILU (C++ enumerator) NVTE_Activation_Type::SREGLU (C++ enumerator) NVTE_Activation_Type::SRELU (C++ enumerator) NVTE_Activation_Type::SWIGLU (C++ enumerator) NVTE_Bias_Type (C++ enum) NVTE_Bias_Type::NVTE_ALIBI (C++ enumerator) NVTE_Bias_Type::NVTE_NO_BIAS (C++ enumerator) NVTE_Bias_Type::NVTE_POST_SCALE_BIAS (C++ enumerator) NVTE_Bias_Type::NVTE_PRE_SCALE_BIAS (C++ enumerator) nvte_cast_transpose (C++ function) nvte_cast_transpose_dbias (C++ function) nvte_cast_transpose_dbias_dgelu (C++ function) nvte_cast_transpose_dbias_dqgelu (C++ function) nvte_cast_transpose_dbias_drelu (C++ function) nvte_cast_transpose_dbias_dsilu (C++ function) nvte_cast_transpose_dbias_dsrelu (C++ function) nvte_cast_transpose_with_noop (C++ function) nvte_clamped_dswiglu (C++ function) nvte_clamped_dswiglu_v2 (C++ function) nvte_clamped_swiglu (C++ function) nvte_clamped_swiglu_v2 (C++ function) nvte_compute_amax (C++ function) nvte_compute_amax_with_config (C++ function) nvte_compute_grouped_tensor_offsets (C++ function) nvte_compute_scale_from_amax (C++ function) nvte_convert_bshd_to_thd (C++ function) nvte_convert_int32_to_int64 (C++ function) nvte_convert_int32_to_int64_with_multiplier (C++ function) nvte_convert_thd_to_bshd (C++ function) nvte_copy_to_kv_cache (C++ function) nvte_cp_thd_get_partitioned_indices (C++ function) nvte_cp_thd_grad_correction (C++ function) nvte_cp_thd_out_correction (C++ function) nvte_cp_thd_read_half_tensor (C++ function) nvte_cp_thd_read_second_half_lse (C++ function) nvte_cp_thd_second_half_lse_correction (C++ function) nvte_create_grouped_matmul_config (C++ function) nvte_create_grouped_tensor (C++ function) nvte_create_matmul_config (C++ function) nvte_create_quantization_config (C++ function) nvte_create_tensor (C++ function) nvte_create_tensors (C++ function) nvte_cublas_atomic_gemm (C++ function) nvte_cublas_gemm (C++ function) nvte_cublas_gemm_scaled (C++ function) nvte_cublas_gemm_v2 (C++ function) NVTE_DEBUG nvte_delayed_scaling_recipe_amax_and_scale_update (C++ function) nvte_delayed_scaling_recipe_amax_and_scale_update_after_reduction (C++ function) nvte_dequantize (C++ function) nvte_destroy_grouped_matmul_config (C++ function) nvte_destroy_grouped_tensor (C++ function) nvte_destroy_matmul_config (C++ function) nvte_destroy_quantization_config (C++ function) nvte_destroy_tensor (C++ function) nvte_destroy_tensors (C++ function) nvte_device_radix_sort_pairs (C++ function) nvte_dgeglu (C++ function) nvte_dgeglu_cast_transpose (C++ function) nvte_dgelu (C++ function) nvte_dglu (C++ function) NVTE_DPA_FP8_RECIPE, [1], [2], [3], [4], [5], [6], [7] nvte_dqgeglu (C++ function) nvte_dqgeglu_cast_transpose (C++ function) nvte_dqgelu (C++ function) nvte_dreglu (C++ function) nvte_dreglu_cast_transpose (C++ function) nvte_drelu (C++ function) nvte_dsilu (C++ function) nvte_dsituglu (C++ function) nvte_dsreglu (C++ function) nvte_dsreglu_cast_transpose (C++ function) nvte_dsrelu (C++ function) nvte_dswiglu (C++ function) nvte_dswiglu_cast_transpose (C++ function) nvte_enable_cudnn_norm_bwd (C++ function) nvte_enable_cudnn_norm_fwd (C++ function) nvte_enable_zero_centered_gamma_in_weight_dtype (C++ function) nvte_extract_seed_and_offset (C++ function) nvte_fp8_block_scaling_compute_partial_amax (C++ function) nvte_fp8_block_scaling_partial_cast (C++ function) nvte_fp8_transpose_dbias (C++ function) NVTE_Fused_Attn_Backend (C++ enum) NVTE_Fused_Attn_Backend::NVTE_F16_arbitrary_seqlen (C++ enumerator) NVTE_Fused_Attn_Backend::NVTE_FP8 (C++ enumerator) NVTE_Fused_Attn_Backend::NVTE_No_Backend (C++ enumerator) nvte_fused_attn_bwd (C++ function) nvte_fused_attn_fwd (C++ function) nvte_fused_qkv_rope_backward (C++ function) nvte_fused_qkv_rope_forward (C++ function) nvte_fused_rope_backward (C++ function) nvte_fused_rope_forward (C++ function) nvte_geglu (C++ function) nvte_gelu (C++ function) nvte_get_fused_attn_backend (C++ function) nvte_get_grouped_gemm_setup_workspace_size (C++ function) nvte_get_grouped_matmul_config_attribute (C++ function) nvte_get_grouped_tensor_logical_shape (C++ function) nvte_get_grouped_tensor_param (C++ function) nvte_get_kv_format (C++ function) nvte_get_matmul_config_attribute (C++ function) nvte_get_q_format (C++ function) nvte_get_qkv_format (C++ function) nvte_get_qkv_layout_group (C++ function) nvte_get_quantization_config_attribute (C++ function) nvte_get_runtime_num_segments (C++ function) nvte_get_tensor_param (C++ function) nvte_get_tensor_param_v2 (C++ function) nvte_glu (C++ function) nvte_group_amax (C++ function) nvte_group_amax_graph_safe (C++ function) nvte_group_compute_amax_with_config (C++ function) nvte_group_compute_scale_from_amax (C++ function) nvte_group_dequantize (C++ function) nvte_group_dgelu (C++ function) nvte_group_dqgelu (C++ function) nvte_group_drelu (C++ function) nvte_group_dsilu (C++ function) nvte_group_dsrelu (C++ function) nvte_group_gelu (C++ function) nvte_group_nvfp4_quantize_with_amax (C++ function) nvte_group_qgelu (C++ function) nvte_group_quantize (C++ function) nvte_group_quantize_dbias (C++ function) nvte_group_quantize_dbias_dgelu (C++ function) nvte_group_quantize_dbias_dqgelu (C++ function) nvte_group_quantize_dbias_drelu (C++ function) nvte_group_quantize_dbias_dsilu (C++ function) nvte_group_quantize_dbias_dsrelu (C++ function) nvte_group_relu (C++ function) nvte_group_requantize (C++ function) nvte_group_scaled_clamped_swiglu (C++ function) nvte_group_scaled_swiglu (C++ function) nvte_group_silu (C++ function) nvte_group_srelu (C++ function) nvte_grouped_bias_add (C++ function) nvte_grouped_gemm (C++ function) nvte_grouped_gemm_with_discrete_inputA (C++ function) nvte_grouped_gemm_with_discrete_out (C++ function) nvte_grouped_scaled_bias_add (C++ function) nvte_grouped_tensor_num_tensors (C++ function) nvte_grouped_tensor_scaling_mode (C++ function) nvte_grouped_tensor_type (C++ function) nvte_is_non_tn_fp8_gemm_supported (C++ function) NVTE_JAX_CUSTOM_CALLS nvte_layernorm_bwd (C++ function) nvte_layernorm_fwd (C++ function) nvte_make_shape (C++ function) NVTE_Mask_Type (C++ enum) NVTE_Mask_Type::NVTE_CAUSAL_BOTTOM_RIGHT_MASK (C++ enumerator) NVTE_Mask_Type::NVTE_CAUSAL_MASK (C++ enumerator) NVTE_Mask_Type::NVTE_NO_MASK (C++ enumerator) NVTE_Mask_Type::NVTE_PADDING_CAUSAL_BOTTOM_RIGHT_MASK (C++ enumerator) NVTE_Mask_Type::NVTE_PADDING_CAUSAL_MASK (C++ enumerator) NVTE_Mask_Type::NVTE_PADDING_MASK (C++ enumerator) nvte_memset (C++ function) nvte_multi_cast_transpose (C++ function) nvte_multi_padding (C++ function) nvte_multi_tensor_adam_capturable_cuda (C++ function) nvte_multi_tensor_adam_capturable_master_cuda (C++ function) nvte_multi_tensor_adam_cuda (C++ function) nvte_multi_tensor_adam_fp8_cuda (C++ function) nvte_multi_tensor_adam_param_remainder_cuda (C++ function) nvte_multi_tensor_compute_scale_and_scale_inv_cuda (C++ function) nvte_multi_tensor_compute_scale_inv_e8m0_cuda (C++ function) nvte_multi_tensor_gemm (C++ function) nvte_multi_tensor_l2norm_cuda (C++ function) nvte_multi_tensor_pad_last_dim (C++ function) nvte_multi_tensor_quantize (C++ function) nvte_multi_tensor_scale_cuda (C++ function) nvte_multi_tensor_scale_tensor_cuda (C++ function) nvte_multi_tensor_sgd_cuda (C++ function) nvte_multi_tensor_swizzle_scaling_factors (C++ function) nvte_multi_tensor_swizzle_scaling_factors_unchecked (C++ function) nvte_multi_tensor_transpose_to_bhsd (C++ function) nvte_multi_tensor_unscale_l2norm_cuda (C++ function) nvte_multi_tensor_unswizzle_scaling_factors (C++ function) nvte_multi_unpadding (C++ function) nvte_mxfp8_scaling_compute_partial_amax (C++ function) nvte_mxfp8_scaling_partial_cast (C++ function) NVTE_Norm_Type (C++ enum) NVTE_Norm_Type::LayerNorm (C++ enumerator) NVTE_Norm_Type::RMSNorm (C++ enumerator) nvte_nvfp4_2d_compute_partial_amax (C++ function) nvte_nvfp4_2d_partial_cast (C++ function) NVTE_NVFP4_4OVER6 nvte_nvfp4_compute_global_scale (C++ function) nvte_nvfp4_compute_per_block_scale (C++ function) nvte_nvfp4_compute_per_tensor_scale (C++ function) nvte_nvfp4_data_transpose (C++ function) nvte_nvfp4_expand_scale_to_fp8 (C++ function) nvte_nvfp4_fused_scale (C++ function) nvte_nvfp4_scale_transpose (C++ function) nvte_permute (C++ function) nvte_populate_rng_state_async (C++ function) nvte_prepare_flash_attn_bwd (C++ function) nvte_prepare_flash_attn_fwd (C++ function) nvte_qgeglu (C++ function) nvte_qgelu (C++ function) NVTE_QKV_Format (C++ enum) NVTE_QKV_Format::NVTE_BHSD (C++ enumerator) NVTE_QKV_Format::NVTE_BSHD (C++ enumerator) NVTE_QKV_Format::NVTE_BSHD_2SBHD (C++ enumerator)

NVTE_QKV_Format::NVTE_QKV_Format_NOT_SET (C++ enumerator) NVTE_QKV_Format::NVTE_SBHD (C++ enumerator) NVTE_QKV_Format::NVTE_SBHD_2BSHD (C++ enumerator) NVTE_QKV_Format::NVTE_THD (C++ enumerator) NVTE_QKV_Format::NVTE_THD_2BSHD (C++ enumerator) NVTE_QKV_Format::NVTE_THD_2SBHD (C++ enumerator) NVTE_QKV_Layout (C++ enum) NVTE_QKV_Layout::NVTE_BHSD_BHSD_BHSD (C++ enumerator) NVTE_QKV_Layout::NVTE_BS3HD (C++ enumerator) NVTE_QKV_Layout::NVTE_BSH3D (C++ enumerator) NVTE_QKV_Layout::NVTE_BSHD_BS2HD (C++ enumerator) NVTE_QKV_Layout::NVTE_BSHD_BSH2D (C++ enumerator) NVTE_QKV_Layout::NVTE_BSHD_BSHD_BSHD (C++ enumerator) NVTE_QKV_Layout::NVTE_BSHD_SBHD_SBHD (C++ enumerator) NVTE_QKV_Layout::NVTE_Paged_KV_BSHD_BSHD_BSHD (C++ enumerator) NVTE_QKV_Layout::NVTE_Paged_KV_BSHD_SBHD_SBHD (C++ enumerator) NVTE_QKV_Layout::NVTE_Paged_KV_SBHD_BSHD_BSHD (C++ enumerator) NVTE_QKV_Layout::NVTE_Paged_KV_SBHD_SBHD_SBHD (C++ enumerator) NVTE_QKV_Layout::NVTE_Paged_KV_THD_BSHD_BSHD (C++ enumerator) NVTE_QKV_Layout::NVTE_Paged_KV_THD_SBHD_SBHD (C++ enumerator) NVTE_QKV_Layout::NVTE_QKV_Layout_NOT_SET (C++ enumerator) NVTE_QKV_Layout::NVTE_SB3HD (C++ enumerator) NVTE_QKV_Layout::NVTE_SBH3D (C++ enumerator) NVTE_QKV_Layout::NVTE_SBHD_BSHD_BSHD (C++ enumerator) NVTE_QKV_Layout::NVTE_SBHD_SB2HD (C++ enumerator) NVTE_QKV_Layout::NVTE_SBHD_SBH2D (C++ enumerator) NVTE_QKV_Layout::NVTE_SBHD_SBHD_SBHD (C++ enumerator) NVTE_QKV_Layout::NVTE_T3HD (C++ enumerator) NVTE_QKV_Layout::NVTE_TH3D (C++ enumerator) NVTE_QKV_Layout::NVTE_THD_BSHD_BSHD (C++ enumerator) NVTE_QKV_Layout::NVTE_THD_SBHD_SBHD (C++ enumerator) NVTE_QKV_Layout::NVTE_THD_T2HD (C++ enumerator) NVTE_QKV_Layout::NVTE_THD_TH2D (C++ enumerator) NVTE_QKV_Layout::NVTE_THD_THD_THD (C++ enumerator) NVTE_QKV_Layout_Group (C++ enum) NVTE_QKV_Layout_Group::NVTE_3HD (C++ enumerator) NVTE_QKV_Layout_Group::NVTE_H3D (C++ enumerator) NVTE_QKV_Layout_Group::NVTE_HD_2HD (C++ enumerator) NVTE_QKV_Layout_Group::NVTE_HD_H2D (C++ enumerator) NVTE_QKV_Layout_Group::NVTE_HD_HD_HD (C++ enumerator) NVTE_QKV_Layout_Group::NVTE_Paged_KV_HD_HD_HD (C++ enumerator) NVTE_QKV_Layout_Group::NVTE_SD_SD_SD (C++ enumerator) nvte_quantize (C++ function) nvte_quantize_dbias (C++ function) nvte_quantize_dbias_dgelu (C++ function) nvte_quantize_dbias_dqgelu (C++ function) nvte_quantize_dbias_drelu (C++ function) nvte_quantize_dbias_dsilu (C++ function) nvte_quantize_dbias_dsrelu (C++ function) nvte_quantize_noop (C++ function) nvte_quantize_v2 (C++ function) nvte_reglu (C++ function) nvte_relu (C++ function) nvte_rmsnorm_bwd (C++ function) nvte_rmsnorm_bwd_add (C++ function) nvte_rmsnorm_fwd (C++ function) nvte_scaled_aligned_causal_masked_softmax_backward (C++ function) nvte_scaled_aligned_causal_masked_softmax_forward (C++ function) nvte_scaled_clamped_dswiglu (C++ function) nvte_scaled_clamped_swiglu (C++ function) nvte_scaled_dsituglu (C++ function) nvte_scaled_dsrelu (C++ function) nvte_scaled_dswiglu (C++ function) nvte_scaled_masked_softmax_backward (C++ function) nvte_scaled_masked_softmax_forward (C++ function) nvte_scaled_situglu (C++ function) nvte_scaled_softmax_backward (C++ function) nvte_scaled_softmax_forward (C++ function) nvte_scaled_srelu (C++ function) nvte_scaled_swiglu (C++ function) nvte_scaled_upper_triang_masked_softmax_backward (C++ function) nvte_scaled_upper_triang_masked_softmax_forward (C++ function) nvte_set_grouped_matmul_config_attribute (C++ function) nvte_set_grouped_tensor_param (C++ function) nvte_set_matmul_config_attribute (C++ function) nvte_set_quantization_config_attribute (C++ function) nvte_set_tensor_param (C++ function) nvte_set_tensor_param_v2 (C++ function) nvte_silu (C++ function) nvte_situglu (C++ function) NVTE_Softmax_Type (C++ enum) NVTE_Softmax_Type::NVTE_LEARNABLE_SOFTMAX (C++ enumerator) NVTE_Softmax_Type::NVTE_OFF_BY_ONE_SOFTMAX (C++ enumerator) NVTE_Softmax_Type::NVTE_VANILLA_SOFTMAX (C++ enumerator) nvte_splits_to_offsets (C++ function) nvte_splits_to_offsets_2d (C++ function) nvte_splits_to_offsets_multi (C++ function) nvte_sreglu (C++ function) nvte_srelu (C++ function) nvte_swap_first_dims (C++ function) nvte_swiglu (C++ function) nvte_swizzle_block_scaling_to_mxfp8_scaling_factors (C++ function) nvte_swizzle_grouped_scaling_factors (C++ function) nvte_swizzle_scaling_factors (C++ function) nvte_tensor_amax (C++ function) nvte_tensor_columnwise_data (C++ function) nvte_tensor_columnwise_shape (C++ function) nvte_tensor_data (C++ function) nvte_tensor_element_size (C++ function) nvte_tensor_element_size_bits (C++ function) nvte_tensor_ndims (C++ function) nvte_tensor_numel (C++ function) nvte_tensor_pack_create (C++ function) nvte_tensor_pack_destroy (C++ function) nvte_tensor_scale (C++ function) nvte_tensor_scale_inv (C++ function) nvte_tensor_scale_inv_shape (C++ function) nvte_tensor_scaling_mode (C++ function) nvte_tensor_shape (C++ function) nvte_tensor_size (C++ function) nvte_tensor_size_bytes (C++ function) nvte_tensor_type (C++ function) NVTE_TEST_NVINSPECT_ENABLED, [1] nvte_thd_copy_valid_tokens_from_per_split_to_rank_local (C++ function) nvte_thd_cp_rank_order_to_sequence_order (C++ function) nvte_thd_sequence_order_to_cp_rank_order (C++ function) nvte_transpose (C++ function) nvte_transpose_with_noop (C++ function) nvte_unpermute (C++ function) nvte_unswizzle_grouped_scaling_factors (C++ function) nvte_unswizzle_scaling_factors (C++ function) NVTE_USE_CCACHE nvte_zero_tensor (C++ function) NVTEBasicTensor (C++ struct) NVTEBasicTensor::data_ptr (C++ member) NVTEBasicTensor::dtype (C++ member) NVTEBasicTensor::shape (C++ member) NVTEDType (C++ enum) NVTEDType::kNVTEBFloat16 (C++ enumerator) NVTEDType::kNVTEByte (C++ enumerator) NVTEDType::kNVTEFloat16 (C++ enumerator) NVTEDType::kNVTEFloat32 (C++ enumerator) NVTEDType::kNVTEFloat4E2M1 (C++ enumerator) NVTEDType::kNVTEFloat8E4M3 (C++ enumerator) NVTEDType::kNVTEFloat8E5M2 (C++ enumerator) NVTEDType::kNVTEFloat8E8M0 (C++ enumerator) NVTEDType::kNVTEInt16 (C++ enumerator) NVTEDType::kNVTEInt32 (C++ enumerator) NVTEDType::kNVTEInt64 (C++ enumerator) NVTEDType::kNVTENumTypes (C++ enumerator) NVTEGroupedMatmulConfig (C++ type) NVTEGroupedMatmulConfigAttribute (C++ enum) NVTEGroupedMatmulConfigAttribute::kNVTEGroupedMatmulConfigAvgK (C++ enumerator) NVTEGroupedMatmulConfigAttribute::kNVTEGroupedMatmulConfigAvgM (C++ enumerator) NVTEGroupedMatmulConfigAttribute::kNVTEGroupedMatmulConfigAvgN (C++ enumerator) NVTEGroupedMatmulConfigAttribute::kNVTEGroupedMatmulConfigNumAttributes (C++ enumerator) NVTEGroupedMatmulConfigAttribute::kNVTEGroupedMatmulConfigSMCount (C++ enumerator) NVTEGroupedMatmulConfigAttribute::kNVTEGroupedMatmulConfigUseSplitAccumulator (C++ enumerator) NVTEGroupedTensor (C++ type) NVTEGroupedTensorParam (C++ enum) NVTEGroupedTensorParam::kNVTEGroupedAmax (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedColumnwiseAmax (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedColumnwiseData (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedColumnwiseScaleInv (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedFirstDims (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedLastDims (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedRowwiseData (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedRowwiseScaleInv (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedScale (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedTensorOffsets (C++ enumerator) NVTEGroupedTensorParam::kNVTEGroupedWithGEMMSwizzledScales (C++ enumerator) NVTEGroupedTensorParam::kNVTENumGroupedTensorParams (C++ enumerator) NVTEMatmulConfig (C++ type) NVTEMatmulConfigAttribute (C++ enum) NVTEMatmulConfigAttribute::kNVTEMatmulConfigBiasTensor (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigDBiasTensor (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigEpilogueAuxTensor (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigNumAttributes (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigSMCount (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigUseSplitAccumulator (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigWithDGELUEpilogue (C++ enumerator) NVTEMatmulConfigAttribute::kNVTEMatmulConfigWithGELUEpilogue (C++ enumerator) NVTENVFP44Over6Mode (C++ enum) NVTENVFP44Over6Mode::kNVTENVFP44Over6Disabled (C++ enumerator) NVTENVFP44Over6Mode::kNVTENVFP44Over6MinMAE (C++ enumerator) NVTENVFP44Over6Mode::kNVTENVFP44Over6MinMSE (C++ enumerator) NVTEQuantizationConfig (C++ type) NVTEQuantizationConfigAttribute (C++ enum) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigAmaxEpsilon (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigFloat8BlockScaleTensorFormat (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigForcePow2Scales (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigMXFP82DQuantization (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigNoopTensor (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigNumAttributes (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigNVFP42DQuantization (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigNVFP44Over6ErrUseFastMath (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigNVFP44Over6Mode (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigRNGState (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigStochasticRounding (C++ enumerator) NVTEQuantizationConfigAttribute::kNVTEQuantizationConfigUseFastMath (C++ enumerator) NVTEScalingMode (C++ enum) NVTEScalingMode::NVTE_BLOCK_SCALING_1D (C++ enumerator) NVTEScalingMode::NVTE_BLOCK_SCALING_2D (C++ enumerator) NVTEScalingMode::NVTE_DELAYED_TENSOR_SCALING (C++ enumerator) NVTEScalingMode::NVTE_INVALID_SCALING (C++ enumerator) NVTEScalingMode::NVTE_MXFP8_1D_SCALING (C++ enumerator) NVTEScalingMode::NVTE_NVFP4_1D_SCALING (C++ enumerator) NVTEShape (C++ struct) NVTEShape::data (C++ member) NVTEShape::ndim (C++ member) NVTETensor (C++ type) NVTETensorPack (C++ struct) NVTETensorPack::MAX_SIZE (C++ member) NVTETensorPack::size (C++ member) NVTETensorPack::tensors (C++ member) NVTETensorParam (C++ enum) NVTETensorParam::kNVTEAmax (C++ enumerator) NVTETensorParam::kNVTEColumnwiseAmax (C++ enumerator) NVTETensorParam::kNVTEColumnwiseData (C++ enumerator) NVTETensorParam::kNVTEColumnwiseScaleInv (C++ enumerator) NVTETensorParam::kNVTENumTensorParams (C++ enumerator) NVTETensorParam::kNVTENVFP4E4M3Max (C++ enumerator) NVTETensorParam::kNVTERowScaledNVFP4 (C++ enumerator) NVTETensorParam::kNVTERowwiseData (C++ enumerator) NVTETensorParam::kNVTERowwiseScaleInv (C++ enumerator) NVTETensorParam::kNVTEScale (C++ enumerator) NVTETensorParam::kNVTEWithGEMMSwizzledScales (C++ enumerator)

O

op_backward() (transformer_engine.pytorch.ops.BasicOperation method)

op_forward() (transformer_engine.pytorch.ops.BasicOperation method)

P

parallel_cross_entropy() (in module transformer_engine.pytorch) PerTensorScaling (class in transformer_engine.debug.features.per_tensor_scaling)

pp_resource (transformer_engine.jax.MeshResource attribute) prepare_for_saving() (in module transformer_engine.pytorch) (transformer_engine.pytorch.QuantizedTensorStorage method)

Q

QGEGLU (class in transformer_engine.pytorch.ops) QGELU (class in transformer_engine.pytorch.ops) Quantize (class in transformer_engine.pytorch.ops) quantize() (transformer_engine.pytorch.Quantizer method)

quantize_() (transformer_engine.pytorch.QuantizedTensor method) quantized_model_init() (in module transformer_engine.pytorch) QuantizedTensor (class in transformer_engine.pytorch) QuantizedTensorStorage (class in transformer_engine.pytorch) Quantizer (class in transformer_engine.pytorch)

R

ReduceScatter (class in transformer_engine.pytorch.ops) register_backward_fusion() (in module transformer_engine.pytorch.ops) register_forward_backward_fusion() (in module transformer_engine.pytorch.ops) register_forward_fusion() (in module transformer_engine.pytorch.ops) ReGLU (class in transformer_engine.pytorch.ops) RelativePositionBiases (class in transformer_engine.jax.flax)

ReLU (class in transformer_engine.pytorch.ops) reset() (transformer_engine.pytorch.CudaRNGStatesTracker method) Reshape (class in transformer_engine.pytorch.ops) restore_from_saved() (in module transformer_engine.pytorch) (transformer_engine.pytorch.QuantizedTensorStorage method) RMSNorm (class in transformer_engine.pytorch) (class in transformer_engine.pytorch.ops)

S

ScaledClampedQGeGLU (class in transformer_engine.pytorch.ops) ScaledSwiGLU (class in transformer_engine.pytorch.ops) Sequential (class in transformer_engine.pytorch.ops) set_context_parallel_group() (transformer_engine.pytorch.DotProductAttention method) (transformer_engine.pytorch.MultiheadAttention method) (transformer_engine.pytorch.TransformerLayer method)

set_states() (transformer_engine.pytorch.CudaRNGStatesTracker method) set_tensor_parallel_group() (transformer_engine.pytorch.MultiheadAttention method) (transformer_engine.pytorch.TransformerLayer method) SiLU (class in transformer_engine.pytorch.ops) SReGLU (class in transformer_engine.pytorch.ops) SReLU (class in transformer_engine.pytorch.ops) SwiGLU (class in transformer_engine.pytorch.ops)

T

tp_resource (transformer_engine.jax.MeshResource attribute) tpsp_resource (transformer_engine.jax.MeshResource attribute) transformer_engine (C++ type), [1], [2] transformer_engine::DType (C++ enum) transformer_engine::DType::kBFloat16 (C++ enumerator) transformer_engine::DType::kByte (C++ enumerator) transformer_engine::DType::kFloat16 (C++ enumerator) transformer_engine::DType::kFloat32 (C++ enumerator) transformer_engine::DType::kFloat4E2M1 (C++ enumerator) transformer_engine::DType::kFloat8E4M3 (C++ enumerator) transformer_engine::DType::kFloat8E5M2 (C++ enumerator) transformer_engine::DType::kFloat8E8M0 (C++ enumerator) transformer_engine::DType::kInt16 (C++ enumerator) transformer_engine::DType::kInt32 (C++ enumerator) transformer_engine::DType::kInt64 (C++ enumerator) transformer_engine::DType::kNumTypes (C++ enumerator) transformer_engine::Float8BlockScaleTensorFormat (C++ enum) transformer_engine::Float8BlockScaleTensorFormat::COMPACT (C++ enumerator) transformer_engine::Float8BlockScaleTensorFormat::GEMM_READY (C++ enumerator) transformer_engine::Float8BlockScaleTensorFormat::INVALID (C++ enumerator) transformer_engine::GroupedMatmulConfigWrapper (C++ struct) transformer_engine::GroupedMatmulConfigWrapper::config_ (C++ member) transformer_engine::GroupedMatmulConfigWrapper::GroupedMatmulConfigWrapper (C++ function), [1], [2] transformer_engine::GroupedMatmulConfigWrapper::operator NVTEGroupedMatmulConfig (C++ function) transformer_engine::GroupedMatmulConfigWrapper::operator= (C++ function), [1] transformer_engine::GroupedMatmulConfigWrapper::set_avg_k (C++ function) transformer_engine::GroupedMatmulConfigWrapper::set_avg_m (C++ function) transformer_engine::GroupedMatmulConfigWrapper::set_avg_n (C++ function) transformer_engine::GroupedMatmulConfigWrapper::set_sm_count (C++ function) transformer_engine::GroupedMatmulConfigWrapper::set_use_split_accumulator (C++ function) transformer_engine::GroupedMatmulConfigWrapper::~GroupedMatmulConfigWrapper (C++ function) transformer_engine::GroupedTensorWrapper (C++ struct) transformer_engine::GroupedTensorWrapper::convertShape (C++ function), [1] transformer_engine::GroupedTensorWrapper::data (C++ function) transformer_engine::GroupedTensorWrapper::defaultData (C++ member) transformer_engine::GroupedTensorWrapper::defaultShape (C++ member) transformer_engine::GroupedTensorWrapper::dtype (C++ function) transformer_engine::GroupedTensorWrapper::emptyShape (C++ member) transformer_engine::GroupedTensorWrapper::get_amax (C++ function) transformer_engine::GroupedTensorWrapper::get_columnwise_amax (C++ function) transformer_engine::GroupedTensorWrapper::get_columnwise_data (C++ function) transformer_engine::GroupedTensorWrapper::get_columnwise_scale_inv (C++ function) transformer_engine::GroupedTensorWrapper::get_first_dims (C++ function) transformer_engine::GroupedTensorWrapper::get_last_dims (C++ function) transformer_engine::GroupedTensorWrapper::get_parameter (C++ function) transformer_engine::GroupedTensorWrapper::get_rowwise_data (C++ function) transformer_engine::GroupedTensorWrapper::get_rowwise_scale_inv (C++ function) transformer_engine::GroupedTensorWrapper::get_scale (C++ function) transformer_engine::GroupedTensorWrapper::get_tensor_offsets (C++ function) transformer_engine::GroupedTensorWrapper::get_with_gemm_swizzled_scales (C++ function) transformer_engine::GroupedTensorWrapper::GroupedTensorWrapper (C++ function), [1], [2], [3] transformer_engine::GroupedTensorWrapper::logical_shape (C++ function) transformer_engine::GroupedTensorWrapper::num_tensors (C++ function) transformer_engine::GroupedTensorWrapper::operator= (C++ function), [1] transformer_engine::GroupedTensorWrapper::scaling_mode (C++ function) transformer_engine::GroupedTensorWrapper::set_amax (C++ function) transformer_engine::GroupedTensorWrapper::set_columnwise_amax (C++ function) transformer_engine::GroupedTensorWrapper::set_columnwise_data (C++ function) transformer_engine::GroupedTensorWrapper::set_columnwise_scale_inv (C++ function) transformer_engine::GroupedTensorWrapper::set_first_dims (C++ function) transformer_engine::GroupedTensorWrapper::set_last_dims (C++ function) transformer_engine::GroupedTensorWrapper::set_parameter (C++ function) transformer_engine::GroupedTensorWrapper::set_rowwise_data (C++ function) transformer_engine::GroupedTensorWrapper::set_rowwise_scale_inv (C++ function) transformer_engine::GroupedTensorWrapper::set_scale (C++ function) transformer_engine::GroupedTensorWrapper::set_tensor_offsets (C++ function) transformer_engine::GroupedTensorWrapper::set_with_gemm_swizzled_scales (C++ function) transformer_engine::GroupedTensorWrapper::tensor_ (C++ member) transformer_engine::GroupedTensorWrapper::~GroupedTensorWrapper (C++ function) transformer_engine::is_fp4_dtype (C++ function) transformer_engine::is_fp8_dtype (C++ function) transformer_engine::is_high_precision_dtype (C++ function) transformer_engine::MatmulConfigWrapper (C++ struct) transformer_engine::MatmulConfigWrapper::config_ (C++ member) transformer_engine::MatmulConfigWrapper::MatmulConfigWrapper (C++ function), [1], [2] transformer_engine::MatmulConfigWrapper::operator NVTEMatmulConfig (C++ function) transformer_engine::MatmulConfigWrapper::operator= (C++ function), [1] transformer_engine::MatmulConfigWrapper::set_bias_tensor (C++ function) transformer_engine::MatmulConfigWrapper::set_dbias_tensor (C++ function) transformer_engine::MatmulConfigWrapper::set_epilogue_aux_tensor (C++ function) transformer_engine::MatmulConfigWrapper::set_sm_count (C++ function) transformer_engine::MatmulConfigWrapper::set_use_split_accumulator (C++ function) transformer_engine::MatmulConfigWrapper::set_with_dgelu_epilogue (C++ function) transformer_engine::MatmulConfigWrapper::set_with_gelu_epilogue (C++ function)

transformer_engine::MatmulConfigWrapper::~MatmulConfigWrapper (C++ function) transformer_engine::MultiTensorWrapper (C++ struct) transformer_engine::MultiTensorWrapper::begin (C++ function), [1] transformer_engine::MultiTensorWrapper::data (C++ function), [1] transformer_engine::MultiTensorWrapper::empty (C++ function) transformer_engine::MultiTensorWrapper::end (C++ function), [1] transformer_engine::MultiTensorWrapper::MultiTensorWrapper (C++ function), [1], [2], [3] transformer_engine::MultiTensorWrapper::operator NVTETensor* (C++ function) transformer_engine::MultiTensorWrapper::operator= (C++ function), [1] transformer_engine::MultiTensorWrapper::operator[] (C++ function) transformer_engine::MultiTensorWrapper::size (C++ function) transformer_engine::MultiTensorWrapper::tensors_ (C++ member) transformer_engine::MultiTensorWrapper::~MultiTensorWrapper (C++ function) transformer_engine::nvte_cublas_handle_init (C++ function) transformer_engine::nvte_cudnn_handle_init (C++ function) transformer_engine::QuantizationConfigWrapper (C++ struct) transformer_engine::QuantizationConfigWrapper::config_ (C++ member) transformer_engine::QuantizationConfigWrapper::operator NVTEQuantizationConfig (C++ function) transformer_engine::QuantizationConfigWrapper::operator= (C++ function), [1] transformer_engine::QuantizationConfigWrapper::QuantizationConfigWrapper (C++ function), [1], [2] transformer_engine::QuantizationConfigWrapper::set_amax_epsilon (C++ function) transformer_engine::QuantizationConfigWrapper::set_float8_block_scale_tensor_format (C++ function) transformer_engine::QuantizationConfigWrapper::set_force_pow_2_scales (C++ function) transformer_engine::QuantizationConfigWrapper::set_mxfp8_2d_quantization (C++ function) transformer_engine::QuantizationConfigWrapper::set_noop_tensor (C++ function) transformer_engine::QuantizationConfigWrapper::set_nvfp4_2d_quantization (C++ function) transformer_engine::QuantizationConfigWrapper::set_nvfp4_4over6_err_use_fast_math (C++ function) transformer_engine::QuantizationConfigWrapper::set_nvfp4_4over6_mode (C++ function) transformer_engine::QuantizationConfigWrapper::set_rng_state (C++ function) transformer_engine::QuantizationConfigWrapper::set_stochastic_rounding (C++ function) transformer_engine::QuantizationConfigWrapper::set_use_fast_math (C++ function) transformer_engine::QuantizationConfigWrapper::~QuantizationConfigWrapper (C++ function) transformer_engine::TensorWrapper (C++ struct) transformer_engine::TensorWrapper::amax (C++ function) transformer_engine::TensorWrapper::bytes (C++ function) transformer_engine::TensorWrapper::columnwise_dptr (C++ function) transformer_engine::TensorWrapper::columnwise_shape (C++ function) transformer_engine::TensorWrapper::convertShape (C++ function), [1] transformer_engine::TensorWrapper::data (C++ function) transformer_engine::TensorWrapper::defaultData (C++ member) transformer_engine::TensorWrapper::defaultShape (C++ member) transformer_engine::TensorWrapper::dptr (C++ function) transformer_engine::TensorWrapper::dtype (C++ function) transformer_engine::TensorWrapper::element_size (C++ function) transformer_engine::TensorWrapper::element_size_bits (C++ function) transformer_engine::TensorWrapper::emptyShape (C++ member) transformer_engine::TensorWrapper::get_amax (C++ function) transformer_engine::TensorWrapper::get_columnwise_amax (C++ function) transformer_engine::TensorWrapper::get_columnwise_data (C++ function) transformer_engine::TensorWrapper::get_columnwise_scale_inv (C++ function) transformer_engine::TensorWrapper::get_nvfp4_e4m3_max (C++ function) transformer_engine::TensorWrapper::get_parameter (C++ function) transformer_engine::TensorWrapper::get_row_scaled_nvfp4 (C++ function) transformer_engine::TensorWrapper::get_rowwise_data (C++ function) transformer_engine::TensorWrapper::get_rowwise_scale_inv (C++ function) transformer_engine::TensorWrapper::get_scale (C++ function) transformer_engine::TensorWrapper::get_with_gemm_swizzled_scales (C++ function) transformer_engine::TensorWrapper::ndim (C++ function) transformer_engine::TensorWrapper::numel (C++ function) transformer_engine::TensorWrapper::operator= (C++ function), [1] transformer_engine::TensorWrapper::scale (C++ function) transformer_engine::TensorWrapper::scale_inv (C++ function) transformer_engine::TensorWrapper::scale_inv_shape (C++ function) transformer_engine::TensorWrapper::scaling_mode (C++ function) transformer_engine::TensorWrapper::set_amax (C++ function) transformer_engine::TensorWrapper::set_columnwise_amax (C++ function) transformer_engine::TensorWrapper::set_columnwise_data (C++ function) transformer_engine::TensorWrapper::set_columnwise_scale_inv (C++ function) transformer_engine::TensorWrapper::set_nvfp4_e4m3_max (C++ function) transformer_engine::TensorWrapper::set_parameter (C++ function) transformer_engine::TensorWrapper::set_row_scaled_nvfp4 (C++ function) transformer_engine::TensorWrapper::set_rowwise_data (C++ function) transformer_engine::TensorWrapper::set_rowwise_scale_inv (C++ function) transformer_engine::TensorWrapper::set_scale (C++ function) transformer_engine::TensorWrapper::set_with_gemm_swizzled_scales (C++ function) transformer_engine::TensorWrapper::shape (C++ function) transformer_engine::TensorWrapper::size (C++ function) transformer_engine::TensorWrapper::tensor_ (C++ member) transformer_engine::TensorWrapper::TensorWrapper (C++ function), [1], [2], [3], [4] transformer_engine::TensorWrapper::zero_ (C++ function) transformer_engine::TensorWrapper::~TensorWrapper (C++ function) TransformerLayer (class in transformer_engine.jax.flax) (class in transformer_engine.pytorch) TransformerLayerType (class in transformer_engine.jax.flax)

U

update_collections() (in module transformer_engine.jax) update_quantized() (transformer_engine.pytorch.Quantizer method)

update_usage() (transformer_engine.pytorch.QuantizedTensorStorage method) UserBufferQuantizationMode (class in transformer_engine.pytorch)