diff --git a/docs/reference/examples.md b/docs/reference/examples.md index 1d54c490..7ff18529 100644 --- a/docs/reference/examples.md +++ b/docs/reference/examples.md @@ -24,6 +24,11 @@ We've curated a growing collection of practical examples that showcase the power - **[12_gemm_all_scatter_bulk_synchronous](https://github.com/ROCm/iris/tree/main/examples/12_gemm_all_scatter_bulk_synchronous)**: Matrix multiplication with all-scatter using the bulk synchronous parallel approach - **[13_flash_decode](https://github.com/ROCm/iris/tree/main/examples/13_flash_decode)**: Fused Flash Decode Attention for accelerating LLM inference - **[14_all_gather_gemm](https://github.com/ROCm/iris/tree/main/examples/14_all_gather_gemm)**: Fused All-Gather + GEMM with Pull and Push models +- **[15_gemm_all_reduce_ring_based](https://github.com/ROCm/iris/tree/main/examples/15_gemm_all_reduce_ring_based)**: Matrix multiplication with ring-based all-reduce +- **[16_all_reduce_ring_based](https://github.com/ROCm/iris/tree/main/examples/16_all_reduce_ring_based)**: Ring-based all-reduce operation +- **[17_gemm_one_shot_all_reduce_pc](https://github.com/ROCm/iris/tree/main/examples/17_gemm_one_shot_all_reduce_pc)**: Matrix multiplication with one-shot all-reduce using producer-consumer pattern with two distribution modes (striding and block) +- **[20_gemm_all_scatter_independent](https://github.com/ROCm/iris/tree/main/examples/20_gemm_all_scatter_independent)**: Independent GEMM and all-scatter operations with support for CSV input configurations +- **[21_gemm_one_shot_all_reduce_independent](https://github.com/ROCm/iris/tree/main/examples/21_gemm_one_shot_all_reduce_independent)**: Independent GEMM and all-reduce operations with support for CSV input configurations and selective execution ### Utilities - **[benchmark](https://github.com/ROCm/iris/tree/main/examples/benchmark)**: Benchmarking utilities and performance testing tools diff --git a/examples/README.md b/examples/README.md index f064d8e4..c7a0b223 100644 --- a/examples/README.md +++ b/examples/README.md @@ -86,6 +86,12 @@ python examples/14_all_gather_gemm/example_run_pull.py --num_ranks 8 # All-Gather + GEMM - Push model python examples/14_all_gather_gemm/example_run_push.py --num_ranks 8 +# Example command to run benchmark with ring-based all-reduce for GEMM +python examples/15_gemm_all_reduce_ring_based/benchmark.py --benchmark --validate --num_ranks 8 + +# Example command to run benchmark with ring-based all-reduce +python examples/16_all_reduce_ring_based/benchmark.py --benchmark --validate --num_ranks 8 + # Independent GEMM and all-scatter - single configuration python examples/20_gemm_all_scatter_independent/benchmark.py --benchmark --validate --num_ranks 8