Hongxin Liu 8241c0c054 [fp8] support gemini plugin (#5978) ...

* [fp8] refactor hook

* [fp8] support gemini plugin

* [example] add fp8 option for llama benchmark

2024-08-09 14:09:48 +08:00

..

test_all_to_all_single.py

[fp8]support all2all fp8 (#5953)

2024-08-06 16:58:23 +08:00

test_fp8_all_to_all_single.py

[Feature] llama shardformer fp8 support (#5938)

2024-08-05 10:05:47 +08:00

test_fp8_all_to_all.py

[Feature] llama shardformer fp8 support (#5938)

2024-08-05 10:05:47 +08:00

test_fp8_allgather_flat.py

[Feature] llama shardformer fp8 support (#5938)

2024-08-05 10:05:47 +08:00

test_fp8_allreduce.py

[Feature] llama shardformer fp8 support (#5938)

2024-08-05 10:05:47 +08:00

test_fp8_cast.py

[Feature]: support FP8 communication in DDP, FSDP, Gemini (#5928)

2024-08-08 15:55:01 +08:00

test_fp8_ddp_comm_hook.py

[Feature]: support FP8 communication in DDP, FSDP, Gemini (#5928)

2024-08-08 15:55:01 +08:00

test_fp8_fsdp_comm_hook.py

[Feature]: support FP8 communication in DDP, FSDP, Gemini (#5928)

2024-08-08 15:55:01 +08:00

test_fp8_gather.py

[Feature] llama shardformer fp8 support (#5938)

2024-08-05 10:05:47 +08:00

test_fp8_hook.py

[fp8] support gemini plugin (#5978)

2024-08-09 14:09:48 +08:00

test_fp8_linear.py

[fp8] add fp8 linear (#5967)

2024-08-07 15:41:49 +08:00

test_fp8_reduce_scatter.py

[Feature] llama shardformer fp8 support (#5938)

2024-08-05 10:05:47 +08:00