Siyuan (Ryans) Zhuang e3cec88aa5 Memcpy kernel for flash attention (#29) ...

* optimize

* add benchmark

* add assert

* add test

2023-04-10 18:22:49 -07:00

..

activation_kernels.cu

Optimize data movement (#20)

2023-04-02 00:30:17 -07:00

activation.cpp

Optimize data movement (#20)

2023-04-02 00:30:17 -07:00

attention_kernels.cu

Support block size 32 (#35)

2023-04-09 23:07:18 -07:00

attention_utils.h

Add custom kernel for RMS normalization (#16)

2023-04-01 00:51:22 +08:00

attention.cpp

Basic attention kernel that supports cached KV + (multi-)prompts (#24)

2023-04-04 20:34:46 -07:00

cache_kernels.cu

Memcpy kernel for flash attention (#29)

2023-04-10 18:22:49 -07:00

cache.cpp

Memcpy kernel for flash attention (#29)

2023-04-10 18:22:49 -07:00

cuda_primitives.h

Implement single_query_cached_kv_attention kernel (#3)

2023-03-01 15:02:19 -08:00

layernorm_kernels.cu

Add custom kernel for RMS normalization (#16)

2023-04-01 00:51:22 +08:00

layernorm.cpp

Add custom kernel for RMS normalization (#16)

2023-04-01 00:51:22 +08:00

pos_encoding_kernels.cu

Optimize data movement (#20)

2023-04-02 00:30:17 -07:00

pos_encoding.cpp

Optimize data movement (#20)

2023-04-02 00:30:17 -07:00

reduction_utils.h

Add custom kernel for RMS normalization (#16)

2023-04-01 00:51:22 +08:00