Index

_

__init__() (ttnn.MemoryConfig method) (ttnn.Tensor method)

A

abs (in module ttnn) abs_bw (in module ttnn) acos (in module ttnn) acos_bw (in module ttnn) acosh (in module ttnn) acosh_bw (in module ttnn) act_block_h_override (ttnn.Conv2dConfig property) act_block_w_div (ttnn.Conv2dConfig property) activation (ttnn.Conv2dConfig property) AdaptiveAvgPool2d (class in tt_lib.fallback_ops) add (in module ttnn) add_bw (in module ttnn) addalpha (in module ttnn) addalpha_bw (in module ttnn) AddAndNorm() (in module tt_lib.fused_ops.add_and_norm) addcdiv (in module ttnn) addcdiv_bw (in module ttnn) addcmul (in module ttnn) addcmul_bw (in module ttnn) all_gather (in module ttnn)

all_reduce (in module ttnn.experimental) alt_complex_rotate90 (in module ttnn) always_preprocess_weights (ttnn.Conv2dConfig property) angle (in module ttnn) angle_bw (in module ttnn) arange (in module ttnn) argmax (in module ttnn) as_tensor (in module ttnn) asin (in module ttnn) asin_bw (in module ttnn) asinh (in module ttnn) asinh_bw (in module ttnn) assign_bw (in module ttnn) atan (in module ttnn) atan2 (in module ttnn) atan2_bw (in module ttnn) atan_bw (in module ttnn) atanh (in module ttnn) atanh_bw (in module ttnn) attention_softmax (in module ttnn.transformer) attention_softmax_ (in module ttnn.transformer)

B

batch_norm (in module ttnn) BatchNorm2d (class in tt_lib.fallback_ops) BcastOpDim (class in ttnn) BcastOpMath (class in ttnn) bias_gelu_bw (in module ttnn) binary_bitwise_and (class in tt_lib.fallback_ops) binary_bitwise_left_shift (class in tt_lib.fallback_ops) binary_bitwise_or (class in tt_lib.fallback_ops) binary_bitwise_right_shift (class in tt_lib.fallback_ops)

binary_bitwise_xor (class in tt_lib.fallback_ops) binary_fmod (class in tt_lib.fallback_ops) bitwise_and (in module ttnn) bitwise_left_shift (in module ttnn) bitwise_not (class in tt_lib.fallback_ops) (in module ttnn) bitwise_or (in module ttnn) bitwise_right_shift (in module ttnn) bitwise_xor (in module ttnn) buffer() (ttnn.Tensor method)

C

cbrt (in module ttnn) ceil (class in tt_lib.fallback_ops) (in module ttnn) ceil_bw (in module ttnn) celu (in module ttnn) celu_bw (in module ttnn) chunk() (in module tt_lib.fallback_ops) clamp (in module ttnn) clamp_bw (in module ttnn) clip (in module ttnn) clip_bw (in module ttnn) clone (in module ttnn) close_device() (in module ttnn) concat (in module ttnn) concat() (in module tt_lib.fallback_ops) concat_bw (in module ttnn) concatenate_heads (in module ttnn.transformer) conj (in module ttnn)

conj_bw (in module ttnn) conv1d (in module ttnn) Conv2d (class in tt_lib.fallback_ops) conv2d (in module ttnn) conv2d() (in module tt_lib.fallback_ops) Conv2dConfig (class in ttnn) Conv2dSliceConfig (class in ttnn) Conv2dSliceConfig.SliceTypeEnum (class in ttnn) conv3d (in module ttnn.experimental) conv_transpose2d (in module ttnn) core_grid (ttnn.Conv2dConfig property) cos (in module ttnn) cos_bw (in module ttnn) cosh (in module ttnn) cosh_bw (in module ttnn) cpu() (in module ttnn.Tensor) create_sharded_memory_config() (in module ttnn) cumprod (in module ttnn.experimental) cumsum (in module ttnn.experimental)

D

deallocate (in module ttnn) deallocate_activation (ttnn.Conv2dConfig property) deg2rad (in module ttnn) deg2rad_bw (in module ttnn) device() (ttnn.Tensor method) digamma (in module ttnn) digamma_bw (in module ttnn)

div (in module ttnn) div_bw (in module ttnn) div_no_nan (in module ttnn) div_no_nan_bw (in module ttnn) downsample (in module ttnn) dropout (in module ttnn.experimental) dtype (ttnn.Conv2dConfig property) dump_tensor (in module ttnn)

E

elu (in module ttnn) elu_bw (in module ttnn) embedding (in module ttnn) embedding_bw (in module ttnn) empty (in module ttnn) empty_like (in module ttnn) enable_act_double_buffer (ttnn.Conv2dConfig property) enable_split_reader (ttnn.Conv2dConfig property) enable_subblock_padding (ttnn.Conv2dConfig property) enable_weights_double_buffer (ttnn.Conv2dConfig property) eq (in module ttnn) eq_ (in module ttnn)

eqz (in module ttnn) erf (in module ttnn) erf_bw (in module ttnn) erfc (in module ttnn) erfc_bw (in module ttnn) erfinv (in module ttnn) erfinv_bw (in module ttnn) exp (in module ttnn) exp2 (in module ttnn) exp2_bw (in module ttnn) exp_bw (in module ttnn) expm1 (in module ttnn) expm1_bw (in module ttnn)

F

fill (in module ttnn) fill_bw (in module ttnn) fill_cache_for_user_ (in module ttnn.kv_cache) fill_ones_rm (in module ttnn) fill_rm (in module ttnn) fill_zero_bw (in module ttnn) floor (class in tt_lib.fallback_ops) (in module ttnn) floor_bw (in module ttnn) floor_div (in module ttnn)

fmod (in module ttnn) fmod_bw (in module ttnn) format_input_tensor() (in module ttnn) format_output_tensor() (in module ttnn) frac (in module ttnn) frac_bw (in module ttnn) from_device (in module ttnn) from_torch (in module ttnn) full (in module ttnn) full() (in module tt_lib.fallback_ops) full_like (in module ttnn)

G

gcd (in module ttnn) ge (in module ttnn) ge_ (in module ttnn) geglu (in module ttnn) gelu (in module ttnn) gelu_bw (in module ttnn) (in module ttnn.experimental) get_dtype() (ttnn.Tensor method) get_layout() (ttnn.Tensor method) GetDefaultDevice() (in module ttnn)

gez (in module ttnn) global_avg_pool2d (in module ttnn) glu (in module ttnn) group_norm (in module ttnn) group_norm() (in module tt_lib.fallback_ops) (in module ttnn.operations.moreh) group_norm_backward() (in module ttnn.operations.moreh) GroupNorm (class in tt_lib.fallback_ops) gt (in module ttnn) gt_ (in module ttnn) gtz (in module ttnn)

H

hardshrink (in module ttnn) hardshrink_bw (in module ttnn) hardsigmoid (in module ttnn) hardsigmoid_bw (in module ttnn) hardswish (in module ttnn)

hardswish_bw (in module ttnn) hardtanh (in module ttnn) hardtanh_bw (in module ttnn) heaviside (in module ttnn) hypot (in module ttnn) hypot_bw (in module ttnn)

I

i0 (in module ttnn) i0_bw (in module ttnn) identity (in module ttnn) imag (in module ttnn) imag_bw (in module ttnn) in_place (ttnn.Conv2dConfig property) indexed_fill (in module ttnn) interpolate() (in module tt_lib.fallback_ops)

is_imag (in module ttnn) is_real (in module ttnn) isclose (in module ttnn) isfinite (in module ttnn) isinf (in module ttnn) isnan (in module ttnn) isneginf (in module ttnn) isposinf (in module ttnn)

L

l1_loss (in module ttnn) layer_norm (in module ttnn) layer_norm() (in module tt_lib.fallback_ops) LayerNorm (class in tt_lib.fallback_ops) Layernorm() (in module tt_lib.fused_ops.layernorm) lcm (in module ttnn) ldexp (in module ttnn) ldexp_bw (in module ttnn) le (in module ttnn) le_ (in module ttnn) leaky_relu (in module ttnn) leaky_relu_bw (in module ttnn) lerp (in module ttnn) lerp_bw (in module ttnn) lez (in module ttnn) lgamma (in module ttnn) lgamma_bw (in module ttnn) linear (in module ttnn) Linear() (in module tt_lib.fused_ops.linear) load_tensor (in module ttnn) log (in module ttnn) log10 (in module ttnn) log10_bw (in module ttnn) log1p (in module ttnn) log1p_bw (in module ttnn)

log2 (in module ttnn) log2_bw (in module ttnn) log_bw (in module ttnn) log_sigmoid (in module ttnn) log_sigmoid_bw (in module ttnn) logaddexp (in module ttnn) logaddexp2 (in module ttnn) logaddexp2_bw (in module ttnn) logaddexp_bw (in module ttnn) logical_and (in module ttnn) logical_and_ (in module ttnn) logical_not (in module ttnn) logical_not_ (in module ttnn) logical_or (in module ttnn) logical_or_ (in module ttnn) logical_xor (in module ttnn) logical_xor_ (in module ttnn) logit (in module ttnn) logit_bw (in module ttnn) logiteps_bw (in module ttnn) logsoftmax() (in module ttnn.operations.moreh) logsoftmax_backward() (in module ttnn.operations.moreh) lt (in module ttnn) lt_ (in module ttnn) ltz (in module ttnn)

M

mac (in module ttnn) manage_device() (in module ttnn) matmul (in module ttnn) max (in module ttnn) max_bw (in module ttnn) max_pool2d (in module ttnn) maximum (in module ttnn) MaxPool2d (class in tt_lib.fallback_ops) mean (in module ttnn) mean() (in module ttnn.operations.moreh) mean_backward() (in module ttnn.operations.moreh)

MemoryConfig (class in ttnn) min (in module ttnn) min_bw (in module ttnn) minimum (in module ttnn) mish (in module ttnn) moreh_sum (in module ttnn) mse_loss (in module ttnn) mul_bw (in module ttnn) multigammaln (in module ttnn) multigammaln_bw (in module ttnn) multiply (in module ttnn)

N

name (ttnn.Conv2dSliceConfig.SliceTypeEnum property) ne (in module ttnn) ne_ (in module ttnn) neg (in module ttnn) neg_bw (in module ttnn) nextafter (in module ttnn)

nez (in module ttnn) nonzero (in module ttnn) norm() (in module ttnn.operations.moreh) norm_backward() (in module ttnn.operations.moreh) normalize_global (in module ttnn) normalize_hw (in module ttnn) num_slices (ttnn.Conv2dSliceConfig property)

O

ones (in module ttnn) ones_like (in module ttnn) open_device() (in module ttnn)

outer (in module ttnn) output_layout (ttnn.Conv2dConfig property) override_sharding_config (ttnn.Conv2dConfig property)

P

pad (in module ttnn) pad() (in module tt_lib.fallback_ops) (ttnn.Tensor method) pad_to_tile() (ttnn.Tensor method) pad_to_tile_shape() (in module ttnn) permute (in module ttnn) polar (in module ttnn) polar_bw (in module ttnn) polygamma (in module ttnn) polygamma_bw (in module ttnn)

polyval (in module ttnn) pow (in module ttnn) pow_bw (in module ttnn) prelu (in module ttnn) prepare_conv_bias() (in module ttnn) prepare_conv_weights() (in module ttnn) preprocess_model() (in module ttnn.model_preprocessing) preprocess_model_parameters() (in module ttnn.model_preprocessing) preprocess_weights_on_device (ttnn.Conv2dConfig property) prod (in module ttnn) prod_bw (in module ttnn)

R

rad2deg (in module ttnn) rad2deg_bw (in module ttnn) rank (ttnn.Shape property) rdiv (in module ttnn) rdiv_bw (in module ttnn) real (in module ttnn) real_bw (in module ttnn) reallocate (in module ttnn) reallocate_halo_output (ttnn.Conv2dConfig property) reciprocal (in module ttnn) reciprocal_bw (in module ttnn) reduce_scatter (in module ttnn) register_post_operation_hook() (in module ttnn) register_pre_operation_hook() (in module ttnn) reglu (in module ttnn) relu (in module ttnn) relu6 (in module ttnn) relu6_bw (in module ttnn) relu_bw (in module ttnn) relu_max (in module ttnn)

relu_min (in module ttnn) remainder (in module ttnn) remainder_bw (in module ttnn) repeat (in module ttnn) repeat() (in module tt_lib.fallback_ops) repeat_bw (in module ttnn) repeat_interleave (in module ttnn) repeat_interleave() (in module tt_lib.fallback_ops) reshape (in module ttnn) reshape() (in module tt_lib.fallback_ops) reshard_if_not_optimal (ttnn.Conv2dConfig property) rms_norm (in module ttnn) rotary_embedding (in module ttnn.experimental) round (in module ttnn) round_bw (in module ttnn) rpow (in module ttnn) rpow_bw (in module ttnn) rsqrt (in module ttnn) rsqrt_bw (in module ttnn) rsub (in module ttnn) rsub_bw (in module ttnn)

S

scaled_dot_product_attention (in module ttnn.transformer) scaled_dot_product_attention_decode (in module ttnn.transformer) scatter (in module ttnn) selu (in module ttnn) selu_bw (in module ttnn) set_printoptions() (in module ttnn) SetDefaultDevice() (in module ttnn) Shape (class in ttnn) shard_layout (ttnn.Conv2dConfig property) sigmoid (in module ttnn) sigmoid_accurate (in module ttnn) sigmoid_bw (in module ttnn) sign (in module ttnn) sign_bw (in module ttnn) signbit (in module ttnn) silu (in module ttnn) silu() (in module tt_lib.fallback_ops) silu_bw (in module ttnn) sin (in module ttnn) sin_bw (in module ttnn) sinh (in module ttnn) sinh_bw (in module ttnn) slice (in module ttnn) slice_type (ttnn.Conv2dSliceConfig property) SliceHeight (ttnn.Conv2dSliceConfig.SliceTypeEnum attribute) SliceWidth (ttnn.Conv2dSliceConfig.SliceTypeEnum attribute) softmax (in module ttnn) softmax() (in module tt_lib.fallback_ops) (in module ttnn.operations.moreh)

softmax_backward() (in module ttnn.operations.moreh) softmin() (in module ttnn.operations.moreh) softmin_backward() (in module ttnn.operations.moreh) softplus (in module ttnn) softplus_bw (in module ttnn) softshrink (in module ttnn) softshrink_bw (in module ttnn) softsign (in module ttnn) softsign_bw (in module ttnn) sort (in module ttnn.experimental) split_query_key_value_and_split_heads (in module ttnn.transformer) sqrt (in module ttnn) sqrt_bw (in module ttnn) square (in module ttnn) square_bw (in module ttnn) squared_difference (in module ttnn) squared_difference_bw (in module ttnn) std (in module ttnn) storage_type() (ttnn.Tensor method) sub_bw (in module ttnn) subalpha (in module ttnn) subalpha_bw (in module ttnn) subtract (in module ttnn) sum (in module ttnn) swiglu (in module ttnn) swish (in module ttnn) synchronize_device() (in module ttnn)

T

tan (in module ttnn) tan_bw (in module ttnn) tanh (in module ttnn) tanh_bw (in module ttnn) tanhshrink (in module ttnn) tanhshrink_bw (in module ttnn) Tensor (class in ttnn) tensor_slice() (in module tt_lib.fallback_ops) threshold (in module ttnn) threshold_bw (in module ttnn) tilize (in module ttnn) tilize_with_val_padding (in module ttnn) to() (ttnn.Tensor method)

to_device (in module ttnn) to_layout (in module ttnn) to_memory_config (in module ttnn) to_rank() (ttnn.Shape method) to_torch (in module ttnn) topk (in module ttnn) torch_argmax (class in tt_lib.fallback_ops) torch_argmin (class in tt_lib.fallback_ops) transpose_shards (ttnn.Conv2dConfig property) tril (in module ttnn) triu (in module ttnn) trunc (class in tt_lib.fallback_ops) (in module ttnn) trunc_bw (in module ttnn)

U

unary_bitwise_and (class in tt_lib.fallback_ops) unary_bitwise_left_shift (class in tt_lib.fallback_ops) unary_bitwise_or (class in tt_lib.fallback_ops) unary_bitwise_right_shift (class in tt_lib.fallback_ops) unary_bitwise_xor (class in tt_lib.fallback_ops) unary_chain (in module ttnn)

unary_fmod (class in tt_lib.fallback_ops) unpad() (ttnn.Tensor method) unpad_from_tile() (ttnn.Tensor method) untilize (in module ttnn) untilize_with_unpadding (in module ttnn) update_cache_for_token_ (in module ttnn.kv_cache) upsample (in module ttnn)

V

value (ttnn.Conv2dSliceConfig.SliceTypeEnum property)

var (in module ttnn)

W

weights_dtype (ttnn.Conv2dConfig property)

where (in module ttnn) where_bw (in module ttnn)

X

xlogy (in module ttnn)

xlogy_bw (in module ttnn)

Z

zeros (in module ttnn)

zeros_like (in module ttnn)