bridge.perf_recipes.qwen_vl.b200.qwen3_vl#
B200 performance recipes for Qwen3-VL.
Module Contents#
Functions#
Qwen3-VL 235B-A22B pretrain: 64× B200, BF16, PP=8 VP=4 EP=8. |
|
Qwen3-VL 235B-A22B pretrain: 64× B200, FP8 current-scaling, PP=8 VP=4 EP=8. |
|
Qwen3-VL 235B-A22B pretrain: 64× B200, MXFP8, PP=8 VP=4 EP=8. |
|
Qwen3-VL 30B-A3B pretrain: 8× B200, BF16, EP=8. |
|
Qwen3-VL 30B-A3B pretrain: 8× B200, FP8 current-scaling, EP=8. |
|
Qwen3-VL 30B-A3B pretrain: 8× B200, MXFP8, EP=8. |
API#
- bridge.perf_recipes.qwen_vl.b200.qwen3_vl.qwen3_vl_235b_a22b_pretrain_64gpu_b200_bf16_config() megatron.bridge.perf_recipes.qwen_vl.common.ConfigContainer#
Qwen3-VL 235B-A22B pretrain: 64× B200, BF16, PP=8 VP=4 EP=8.
- bridge.perf_recipes.qwen_vl.b200.qwen3_vl.qwen3_vl_235b_a22b_pretrain_64gpu_b200_fp8cs_config() megatron.bridge.perf_recipes.qwen_vl.common.ConfigContainer#
Qwen3-VL 235B-A22B pretrain: 64× B200, FP8 current-scaling, PP=8 VP=4 EP=8.
- bridge.perf_recipes.qwen_vl.b200.qwen3_vl.qwen3_vl_235b_a22b_pretrain_64gpu_b200_fp8mx_config() megatron.bridge.perf_recipes.qwen_vl.common.ConfigContainer#
Qwen3-VL 235B-A22B pretrain: 64× B200, MXFP8, PP=8 VP=4 EP=8.
- bridge.perf_recipes.qwen_vl.b200.qwen3_vl.qwen3_vl_30b_a3b_pretrain_8gpu_b200_bf16_config() megatron.bridge.perf_recipes.qwen_vl.common.ConfigContainer#
Qwen3-VL 30B-A3B pretrain: 8× B200, BF16, EP=8.
- bridge.perf_recipes.qwen_vl.b200.qwen3_vl.qwen3_vl_30b_a3b_pretrain_8gpu_b200_fp8cs_config() megatron.bridge.perf_recipes.qwen_vl.common.ConfigContainer#
Qwen3-VL 30B-A3B pretrain: 8× B200, FP8 current-scaling, EP=8.
- bridge.perf_recipes.qwen_vl.b200.qwen3_vl.qwen3_vl_30b_a3b_pretrain_8gpu_b200_fp8mx_config() megatron.bridge.perf_recipes.qwen_vl.common.ConfigContainer#
Qwen3-VL 30B-A3B pretrain: 8× B200, MXFP8, EP=8.