FastDeploy

append_attn

support glm fa3 (#5586 )

2025-12-16 19:33:27 +08:00

common

…

custom_all_reduce

[BugFix] Fix custom_all_reduce overflow (#5662 )

2025-12-19 18:24:21 +08:00

cutlass_extensions

…

cutlass_kernels

[Feature] Support fusedmoe on Blackwell (#5325 )

2025-12-16 11:58:50 +08:00

flash_mask_attn

[Feature] support flash_mask_attention backend (#5134 )

2025-11-28 10:12:16 +08:00

fp8_gemm_with_cutlass

…

glog

…

int8_gemm_with_cutlass

…

machete

…

mla_attn

[PD Disaggregation] Support PD deployment of DeepSeekv3. (#5251 )

2025-12-02 14:11:50 +08:00

moba_attn

…

moe

[FIX BUG] fix bug in TP in permute_x_fp8_kernel (#5350 )

2025-12-03 05:17:37 -08:00

quantization

…

sample_kernels

[Metax] refactor cutlass moe and optimize flash attention (#5361 )

2025-12-10 17:15:17 +08:00

speculate_decoding

[BugFix] fix speculate_limit_thinking_content_length (#5590 )

2025-12-16 04:31:45 -08:00

w4afp8_gemm

fix w4afp8 (#5634 )

2025-12-22 13:39:41 +08:00

wfp8afp8_sparse_gemm

…

append_attention.cu

[Others]get_block_shape_and_split_kv_block clean code (#5123 )

2025-11-20 16:40:04 +08:00

beam_search_softmax.cu

…

cpp_extensions.cc

Revert "[Feature] add ue8m0 for per_token_quant_fp8 (#5563 )" (#5611 )

2025-12-17 13:59:06 +08:00

cuda_multiprocess.h

…

dequant_int8.cu

…

enforce_generation.cu

…

env.h

…

fused_get_rotary_embedding.cu

…

fused_hadamard_quant_fp8.cu

…

fused_neox_rope_embedding.cu

…

fused_rotary_position_encoding.cu

…

gather_idx.cu

…

gelu_tanh.cu

…

get_data_ptr_ipc.cu

…

get_img_boundaries.cc

…

get_mm_split_fuse.cc

…

get_output_ep.cc

[PD Disaggregation] Distinguish the pipelines for sending kv signal in different prefill (#5514 )

2025-12-12 14:05:36 +08:00

get_output_msg_with_topk.cc

…

get_output.cc

[PD Disaggregation] support DP via v1 router and decouple DP and EP (#5197 )

2025-12-04 15:38:43 +08:00

get_padding_offset_system.cu

…

get_padding_offset.cu

[Iluvatar] Support V1_KVCACHE_SCHEDULER and paddleocr-vl rope mode (#5555 )

2025-12-18 02:14:25 -08:00

get_position_ids_and_mask_encoder_batch.cu

…

helper.cu

[Feature][OP] Append Attn Support CUDA-PDL (#5072 )

2025-11-17 20:47:33 +08:00

helper.h

check METAX_GPU (#5114 )

2025-11-19 16:02:21 +08:00

init_signal_layerwise.cc

…

ipc_sent_key_value_cache_by_remote_ptr.cu

…

limit_thinking_content_length_v1.cu

[BugFix] Fix limit_thinking early return logic in CUDA kernels (#5471 )

2025-12-10 11:03:19 +08:00

limit_thinking_content_length_v2.cu

[BugFix] Fix limit_thinking early return logic in CUDA kernels (#5471 )

2025-12-10 11:03:19 +08:00

merge_prefill_decode_output.cu

…

msg_utils.h

…

multi_head_latent_attention.cu

…

ngram_mask.cu

…

noaux_tc_redundant.cu

[Feature] Support noaux for eplb (#5143 )

2025-11-21 14:10:32 +08:00

noaux_tc.cu

…

noauxtc_kernel.h

[Fix] fix eplb noaux (#5239 )

2025-11-26 17:50:51 +08:00

open_shm_and_get_meta_signal.cc

…

per_token_quant_fp8.cu

Revert "[Feature] add ue8m0 for per_token_quant_fp8 (#5563 )" (#5611 )

2025-12-17 13:59:06 +08:00

read_data_ipc.cu

…

read_ids.py

…

read_temp_ids.py

…

rebuild_padding.cu

…

recover_decode_task.cu

…

remote_cache_kv_ipc.cc

…

remote_cache_kv_ipc.h

[PD Disaggregation] Distinguish the pipelines for sending kv signal in different prefill (#5514 )

2025-12-12 14:05:36 +08:00

save_output_msg_with_topk.cc

…

save_with_output_msg.cc

…

save_with_output_msg.h

…

save_with_output.cc

…

scaled_gemm_f8_i4_f16_gemm.cu

…

scaled_gemm_f8_i4_f16_weight_quantize.cu

…

seqs2seqs.cu

…

set_data_ipc.cu

…

set_flags.cu

…

set_mask_value.cu

…

set_value_by_flags_and_idx.cu

…

share_external_data.cu

…

step_reschedule.cu

…

step_system_cache.cu

…

step.cu

…

stop_generation_multi_ends.cu

[Feature] support stop_token_ids (#5399 )

2025-12-09 17:49:12 +08:00

stop_generation.cu

…

swap_cache_batch.cu

[Feature] dyc8 support prefixcache (#5125 )

2025-11-21 19:46:26 +08:00

swap_cache.cu

…

system2group.cu

…

text_image_gather_scatter.cu

…

text_image_index_out.cu

…

token_penalty_multi_scores.cu

…

token_penalty_only_once.cu

…

token_transfer.hpp

…

transfer_output.cc

…

tune_cublaslt_gemm.cu

…

unset_data_ipc.cu

…

update_attn_mask_offsets.cu

[BugFix]Fix attention mask bug in D-Node of PD-split mode (#5245 )

2025-11-26 17:56:28 +08:00

update_inputs_beam.cu

…

update_inputs_v1.cu

…

update_inputs.cu

…

update_split_fuse_input.cu

…