| Name | Mode | Size | Raw | |
|---|---|---|---|---|
| .. | ||||
| d | feature-tests/ | 0040000 |
||
| f | CMakeLists.txt | 0100644 |
1.1 KB |
raw |
| f | abs.comp | 0100644 |
525 B |
raw |
| f | acc.comp | 0100644 |
946 B |
raw |
| f | add.comp | 0100644 |
2.0 KB |
raw |
| f | add1.comp | 0100644 |
702 B |
raw |
| f | add_id.comp | 0100644 |
1.0 KB |
raw |
| f | arange.comp | 0100644 |
422 B |
raw |
| f | argmax.comp | 0100644 |
1.4 KB |
raw |
| f | argsort.comp | 0100644 |
2.5 KB |
raw |
| f | argsort_large.comp | 0100644 |
3.9 KB |
raw |
| f | ceil.comp | 0100644 |
549 B |
raw |
| f | clamp.comp | 0100644 |
427 B |
raw |
| f | concat.comp | 0100644 |
1.5 KB |
raw |
| f | contig_copy.comp | 0100644 |
1.5 KB |
raw |
| f | conv2d_dw.comp | 0100644 |
3.1 KB |
raw |
| f | conv2d_mm.comp | 0100644 |
12.5 KB |
raw |
| f | conv_transpose_1d.comp | 0100644 |
3.0 KB |
raw |
| f | copy.comp | 0100644 |
619 B |
raw |
| f | copy_from_quant.comp | 0100644 |
1.4 KB |
raw |
| f | copy_to_quant.comp | 0100644 |
7.8 KB |
raw |
| f | copy_transpose.comp | 0100644 |
2.2 KB |
raw |
| f | cos.comp | 0100644 |
374 B |
raw |
| f | count_equal.comp | 0100644 |
796 B |
raw |
| f | count_experts.comp | 0100644 |
1.2 KB |
raw |
| f | cumsum.comp | 0100644 |
2.8 KB |
raw |
| f | cumsum_multipass1.comp | 0100644 |
1.9 KB |
raw |
| f | cumsum_multipass2.comp | 0100644 |
1.9 KB |
raw |
| f | dequant_f32.comp | 0100644 |
442 B |
raw |
| f | dequant_funcs.glsl | 0100644 |
24.3 KB |
raw |
| f | dequant_funcs_cm2.glsl | 0100644 |
23.8 KB |
raw |
| f | dequant_head.glsl | 0100644 |
249 B |
raw |
| f | dequant_iq1_m.comp | 0100644 |
1.5 KB |
raw |
| f | dequant_iq1_s.comp | 0100644 |
1.1 KB |
raw |
| f | dequant_iq2_s.comp | 0100644 |
1.9 KB |
raw |
| f | dequant_iq2_xs.comp | 0100644 |
1.8 KB |
raw |
| f | dequant_iq2_xxs.comp | 0100644 |
2.0 KB |
raw |
| f | dequant_iq3_s.comp | 0100644 |
1.5 KB |
raw |
| f | dequant_iq3_xxs.comp | 0100644 |
2.1 KB |
raw |
| f | dequant_iq4_nl.comp | 0100644 |
909 B |
raw |
| f | dequant_iq4_xs.comp | 0100644 |
1.1 KB |
raw |
| f | dequant_mxfp4.comp | 0100644 |
941 B |
raw |
| f | dequant_q2_k.comp | 0100644 |
1.4 KB |
raw |
| f | dequant_q3_k.comp | 0100644 |
1.7 KB |
raw |
| f | dequant_q4_0.comp | 0100644 |
861 B |
raw |
| f | dequant_q4_1.comp | 0100644 |
892 B |
raw |
| f | dequant_q4_k.comp | 0100644 |
2.6 KB |
raw |
| f | dequant_q5_0.comp | 0100644 |
1.0 KB |
raw |
| f | dequant_q5_1.comp | 0100644 |
1.0 KB |
raw |
| f | dequant_q5_k.comp | 0100644 |
3.1 KB |
raw |
| f | dequant_q6_k.comp | 0100644 |
1.4 KB |
raw |
| f | dequant_q8_0.comp | 0100644 |
839 B |
raw |
| f | diag.comp | 0100644 |
902 B |
raw |
| f | diag_mask_inf.comp | 0100644 |
799 B |
raw |
| f | div.comp | 0100644 |
776 B |
raw |
| f | exp.comp | 0100644 |
544 B |
raw |
| f | fill.comp | 0100644 |
363 B |
raw |
| f | flash_attn.comp | 0100644 |
13.9 KB |
raw |
| f | flash_attn_base.glsl | 0100644 |
7.8 KB |
raw |
| f | flash_attn_cm1.comp | 0100644 |
24.2 KB |
raw |
| f | flash_attn_cm2.comp | 0100644 |
14.3 KB |
raw |
| f | flash_attn_mask_opt.comp | 0100644 |
4.8 KB |
raw |
| f | flash_attn_split_k_reduce.comp | 0100644 |
3.3 KB |
raw |
| f | floor.comp | 0100644 |
550 B |
raw |
| f | geglu.comp | 0100644 |
327 B |
raw |
| f | geglu_erf.comp | 0100644 |
882 B |
raw |
| f | geglu_quick.comp | 0100644 |
200 B |
raw |
| f | gelu.comp | 0100644 |
767 B |
raw |
| f | gelu_erf.comp | 0100644 |
1.3 KB |
raw |
| f | gelu_quick.comp | 0100644 |
631 B |
raw |
| f | generic_binary_head.glsl | 0100644 |
2.2 KB |
raw |
| f | generic_head.glsl | 0100644 |
196 B |
raw |
| f | generic_unary_head.glsl | 0100644 |
3.3 KB |
raw |
| f | get_rows.comp | 0100644 |
1.2 KB |
raw |
| f | get_rows_quant.comp | 0100644 |
1.5 KB |
raw |
| f | glu_head.glsl | 0100644 |
457 B |
raw |
| f | glu_main.glsl | 0100644 |
859 B |
raw |
| f | group_norm.comp | 0100644 |
1.7 KB |
raw |
| f | hardsigmoid.comp | 0100644 |
581 B |
raw |
| f | hardswish.comp | 0100644 |
585 B |
raw |
| f | im2col.comp | 0100644 |
3.0 KB |
raw |
| f | im2col_3d.comp | 0100644 |
3.6 KB |
raw |
| f | l2_norm.comp | 0100644 |
1.2 KB |
raw |
| f | leaky_relu.comp | 0100644 |
586 B |
raw |
| f | log.comp | 0100644 |
384 B |
raw |
| f | mul.comp | 0100644 |
776 B |
raw |
| f | mul_mat_split_k_reduce.comp | 0100644 |
1.3 KB |
raw |
| f | mul_mat_vec.comp | 0100644 |
5.3 KB |
raw |
| f | mul_mat_vec_base.glsl | 0100644 |
8.0 KB |
raw |
| f | mul_mat_vec_iface.glsl | 0100644 |
1.1 KB |
raw |
| f | mul_mat_vec_iq1_m.comp | 0100644 |
5.1 KB |
raw |
| f | mul_mat_vec_iq1_s.comp | 0100644 |
3.6 KB |
raw |
| f | mul_mat_vec_iq2_s.comp | 0100644 |
3.8 KB |
raw |
| f | mul_mat_vec_iq2_xs.comp | 0100644 |
4.5 KB |
raw |
| f | mul_mat_vec_iq2_xxs.comp | 0100644 |
3.7 KB |
raw |
| f | mul_mat_vec_iq3_s.comp | 0100644 |
3.8 KB |
raw |
| f | mul_mat_vec_iq3_xxs.comp | 0100644 |
3.8 KB |
raw |
| f | mul_mat_vec_nc.comp | 0100644 |
3.6 KB |
raw |
| f | mul_mat_vec_p021.comp | 0100644 |
4.5 KB |
raw |
| f | mul_mat_vec_q2_k.comp | 0100644 |
6.2 KB |
raw |
| f | mul_mat_vec_q3_k.comp | 0100644 |
6.4 KB |
raw |
| f | mul_mat_vec_q4_k.comp | 0100644 |
6.4 KB |
raw |
| f | mul_mat_vec_q5_k.comp | 0100644 |
7.7 KB |
raw |
| f | mul_mat_vec_q6_k.comp | 0100644 |
5.6 KB |
raw |
| f | mul_mat_vecq.comp | 0100644 |
4.9 KB |
raw |
| f | mul_mat_vecq_funcs.glsl | 0100644 |
21.2 KB |
raw |
| f | mul_mm.comp | 0100644 |
16.8 KB |
raw |
| f | mul_mm_cm2.comp | 0100644 |
25.0 KB |
raw |
| f | mul_mm_funcs.glsl | 0100644 |
30.2 KB |
raw |
| f | mul_mm_id_funcs.glsl | 0100644 |
2.1 KB |
raw |
| f | mul_mmq.comp | 0100644 |
9.2 KB |
raw |
| f | mul_mmq_funcs.glsl | 0100644 |
17.3 KB |
raw |
| f | mul_mmq_shmem_types.glsl | 0100644 |
1.5 KB |
raw |
| f | multi_add.comp | 0100644 |
7.2 KB |
raw |
| f | neg.comp | 0100644 |
520 B |
raw |
| f | norm.comp | 0100644 |
1.3 KB |
raw |
| f | opt_step_adamw.comp | 0100644 |
1.2 KB |
raw |
| f | opt_step_sgd.comp | 0100644 |
612 B |
raw |
| f | pad.comp | 0100644 |
2.4 KB |
raw |
| f | pool2d.comp | 0100644 |
1.7 KB |
raw |
| f | quantize_q8_1.comp | 0100644 |
3.1 KB |
raw |
| f | reglu.comp | 0100644 |
126 B |
raw |
| f | relu.comp | 0100644 |
528 B |
raw |
| f | repeat.comp | 0100644 |
809 B |
raw |
| f | repeat_back.comp | 0100644 |
1.2 KB |
raw |
| f | rms_norm.comp | 0100644 |
4.8 KB |
raw |
| f | rms_norm_back.comp | 0100644 |
1.8 KB |
raw |
| f | rms_norm_partials.comp | 0100644 |
2.2 KB |
raw |
| f | roll.comp | 0100644 |
1.4 KB |
raw |
| f | rope_funcs.glsl | 0100644 |
7.4 KB |
raw |
| f | rope_head.glsl | 0100644 |
607 B |
raw |
| f | rope_multi.comp | 0100644 |
467 B |
raw |
| f | rope_neox.comp | 0100644 |
466 B |
raw |
| f | rope_norm.comp | 0100644 |
466 B |
raw |
| f | rope_params.glsl | 0100644 |
566 B |
raw |
| f | rope_vision.comp | 0100644 |
468 B |
raw |
| f | round.comp | 0100644 |
728 B |
raw |
| f | rte.glsl | 0100644 |
154 B |
raw |
| f | scale.comp | 0100644 |
625 B |
raw |
| f | sigmoid.comp | 0100644 |
541 B |
raw |
| f | silu.comp | 0100644 |
565 B |
raw |
| f | silu_back.comp | 0100644 |
761 B |
raw |
| f | sin.comp | 0100644 |
374 B |
raw |
| f | soft_max.comp | 0100644 |
5.2 KB |
raw |
| f | soft_max_back.comp | 0100644 |
1.5 KB |
raw |
| f | soft_max_large1.comp | 0100644 |
1.6 KB |
raw |
| f | soft_max_large2.comp | 0100644 |
2.1 KB |
raw |
| f | soft_max_large3.comp | 0100644 |
1.7 KB |
raw |
| f | soft_max_large_common.glsl | 0100644 |
1.3 KB |
raw |
| f | softplus.comp | 0100644 |
611 B |
raw |
| f | solve_tri.comp | 0100644 |
2.6 KB |
raw |
| f | sqrt.comp | 0100644 |
375 B |
raw |
| f | square.comp | 0100644 |
375 B |
raw |
| f | ssm_conv.comp | 0100644 |
1.2 KB |
raw |
| f | ssm_scan.comp | 0100644 |
3.9 KB |
raw |
| f | step.comp | 0100644 |
565 B |
raw |
| f | sub.comp | 0100644 |
826 B |
raw |
| f | sum_rows.comp | 0100644 |
1.4 KB |
raw |
| f | sum_rows.glsl | 0100644 |
592 B |
raw |
| f | swiglu.comp | 0100644 |
134 B |
raw |
| f | swiglu_oai.comp | 0100644 |
288 B |
raw |
| f | tanh.comp | 0100644 |
544 B |
raw |
| f | timestep_embedding.comp | 0100644 |
1.0 KB |
raw |
| f | topk_argsort.comp | 0100644 |
4.0 KB |
raw |
| f | topk_moe.comp | 0100644 |
6.7 KB |
raw |
| f | topk_nary_search.comp | 0100644 |
8.9 KB |
raw |
| f | tri.comp | 0100644 |
1.3 KB |
raw |
| f | trunc.comp | 0100644 |
550 B |
raw |
| f | types.glsl | 0100644 |
120.3 KB |
raw |
| f | upscale.comp | 0100644 |
5.9 KB |
raw |
| f | utils.glsl | 0100644 |
752 B |
raw |
| f | vulkan-shaders-gen.cpp | 0100644 |
63.4 KB |
raw |
| f | wkv6.comp | 0100644 |
2.7 KB |
raw |
| f | wkv7.comp | 0100644 |
2.9 KB |
raw |
| f | xielu.comp | 0100644 |
836 B |
raw |