MCPcopy Create free account

hub / github.com/FlashSampling/FlashSampling / types & classes

Types & classes25 in github.com/FlashSampling/FlashSampling

↓ 14 callersClassSimpleSampler
src/fused_mm_sampling/core.py:653
↓ 4 callersClassModalEnvConfig
Common env-var config shared by Modal benchmark scripts. Fields are read from same-named env vars (case-insensitive). Subclass to add script-
src/fused_mm_sampling/modal_lib/utils.py:10
↓ 3 callersClassArgs
src/fused_mm_sampling/modal_lib/modal_vllm_benchmark.py:48
↓ 2 callersClassTPInfo
Tensor parallel context.
src/fused_mm_sampling/tp_info.py:14
↓ 1 callersClassArgs
src/fused_mm_sampling/bench/triton_benchmark_lib.py:60
↓ 1 callersClassArgs
benchmarking/plot_tp_scaling.py:168
↓ 1 callersClassArgs
benchmarking/plot-triton-bench.py:604
↓ 1 callersClassBenchRun
benchmarking/vllm/parse_engine_stats.py:30
↓ 1 callersClassCase
src/fused_mm_sampling/bench/speed_test.py:63
↓ 1 callersClassCliArgs
src/fused_mm_sampling/bench/triton_benchmark_lib.py:97
↓ 1 callersClassCliArgs
src/fused_mm_sampling/bench/speed_test.py:27
↓ 1 callersClassCliArgs
benchmarking/proton_profile.py:54
↓ 1 callersClassCliArgs
benchmarking/hbm-access.py:35
↓ 1 callersClassConfig
src/fused_mm_sampling/modal_lib/modal_triton_benchmark.py:5
↓ 1 callersClassConfig
src/fused_mm_sampling/modal_lib/modal_ncu.py:17
↓ 1 callersClassConfig
src/fused_mm_sampling/modal_lib/modal_nsys.py:17
↓ 1 callersClassConfig
src/fused_mm_sampling/modal_lib/modal_vllm_benchmark.py:59
↓ 1 callersClassSyntheticInputs
In the tensor-parallel case V is sharded among ranks
src/fused_mm_sampling/testing.py:15
ClassArgs
benchmarking/proton_profile.py:48
ClassArgs
benchmarking/hbm-access.py:31
ClassFlashSamplingNames
src/fused_mm_sampling/alg_names.py:43
ClassJLSampler
src/fused_mm_sampling/core.py:664
ClassLongNames
src/fused_mm_sampling/alg_names.py:25
ClassSampler
src/fused_mm_sampling/core.py:644
ClassShortNames
src/fused_mm_sampling/alg_names.py:7