Individual skill
ml-engineering
Not available
SKILL.md
Source commit b601d543c9989c825883c2a12c40da490bee5dac
What was checked
- Matches the observed repository HEAD
- Yes
- Scanner rules are current
- Yes
- Repository scan complete
- No
- Skill scan complete
- No
- Observed behavior
- Network or broad shell
- Last checked
- Nov 9, 2026
These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.
Install this skill
Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.
Standard SKILL.md format has not been attested at this commit. No installation command is available; inspect the source.
Repository context
- Repository stars (not skill ratings)
- 19,688
- Repository last updated
- Nov 4, 2026
- License
- CC-BY-SA-4.0
Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.
- pytorch Repository topic
- slurm Repository topic
- large-language-models Repository topic
- llm Repository topic
- machine-learning Repository topic
- scalability Repository topic
- transformers Repository topic
- machine-learning-engineering Repository topic
- mlops Repository topic
- ai Repository topic
Declared tags are partial or unavailable.
Files and excerpt
File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.
Browse files
.gitignoreasset · 1,984 bytesLICENSE-CC-BY-SAasset · 20,131 bytesMakefileruns · 3,625 bytesREADME.mddoc · 9,539 bytesSKILL.mdcore · 225 bytesbuild/README.mddoc · 1,493 bytesbuild/SESSION.mddoc · 88,899 bytesbuild/check-links.pyruns · 4,463 bytesbuild/check-new-links.shruns · 1,662 bytesbuild/check-programsruns · 2,545 bytesbuild/check-redirects.pyruns · 20,784 bytesbuild/check-style.pyruns · 4,444 bytesbuild/check-table-width.pyruns · 12,152 bytesbuild/consistency-checks.mddoc · 9,361 bytesbuild/cover/README.mddoc · 582 bytesbuild/cover/build_final_cover.pyruns · 8,594 bytesbuild/cover/sources/Machine-Learning-Engineering-art.pngasset · 416,042 bytesbuild/fix-tables.pyruns · 11,285 bytesbuild/linkcheckerrcasset · 211 bytesbuild/mdbook/md-to-html.pyruns · 3,480 bytesbuild/mdbook/mv-links.pyruns · 934 bytesbuild/mdbook/preprocess-html-for-epub.pyruns · 5,736 bytesbuild/mdbook/utils/build_utils.pyruns · 152 bytesbuild/mdbook/utils/github_md_utils.pyruns · 7,416 bytesbuild/prince_style.cssasset · 3,339 bytesbuild/requirements.txtdoc · 60 byteschapters-md.txtdoc · 1,299 bytescompute/README.mddoc · 240 bytescompute/accelerator/README.mddoc · 129,402 bytescompute/accelerator/amd/debug.mddoc · 11,491 bytescompute/accelerator/amd/performance.mddoc · 600 bytescompute/accelerator/benchmarks/README.mddoc · 37,251 bytescompute/accelerator/benchmarks/mamf-finder-all-gpus.pyruns · 17,524 bytescompute/accelerator/benchmarks/mamf-finder.pyruns · 115,762 bytescompute/accelerator/images/8x-H100-node-Dell-PowerEdge-XE9680.pngasset · 395,518 bytescompute/accelerator/images/maf-nvidia-amd-efficiency.pngasset · 78,662 bytescompute/accelerator/images/nvidia-a100-matmul-tflops.pngasset · 164,527 bytescompute/accelerator/images/nvidia-a100-spec.pngasset · 134,749 bytescompute/accelerator/nvidia/debug.mddoc · 28,705 bytescompute/accelerator/nvidia/images/dcgm-metrics.pngasset · 43,482 bytescompute/cpu-memory/README.mddoc · 2,523 bytescompute/cpu/README.mddoc · 6,424 bytescontributors.mddoc · 1,580 bytescourses/lesson-learned/README.mddoc · 41,679 bytescourses/lesson-learned/lesson-learned-cover.pngasset · 3,178,428 bytescourses/lesson-learned/lesson-learned-cover.promptasset · 2,328 bytesdebug/NicerTrace.pyruns · 9,469 bytesdebug/README.mddoc · 851 bytesdebug/code/collective_mismatch.pyruns · 495 bytesdebug/code/collective_mismatch_analyze.pyruns · 867 bytesdebug/code/kernel_oob.cuasset · 459 bytesdebug/code/kernel_oob.pyruns · 237 bytesdebug/code/kernel_oob_gdb.pyruns · 300 bytesdebug/code/kernel_oob_prebuilt.pyruns · 273 bytesdebug/code/kernel_race.cuasset · 513 bytesdebug/code/kernel_race.pyruns · 239 bytesdebug/code/see-mem-usage.pyruns · 4,950 bytesdebug/images/math-fp-discrepancy-outcome-lizard.pngasset · 689,964 bytesdebug/images/mem-profiler-llama-8b-ckpt-offload-cpu.pngasset · 427,834 bytesdebug/images/tensor-needle-in-a-hay-stack.pngasset · 32,098 bytesdebug/images/torch-mem-profile-mem-leak.pngasset · 189,562 bytesdebug/make-tiny-models-tokenizers-datasets.mddoc · 1,357 bytesdebug/nccl-performance-debug.mddoc · 79 bytesdebug/pytorch.mddoc · 239,060 bytesdebug/tiny-scripts/README.mddoc · 687 bytesdebug/tiny-scripts/c4-en-10k.pyruns · 2,622 bytesdebug/tiny-scripts/cm4-synthetic-testing.pyruns · 8,260 bytesdebug/tiny-scripts/fsmt-make-super-tiny-model.pyruns · 3,263 bytesdebug/tiny-scripts/general-pmd-ds-unpack.pyruns · 3,142 bytesdebug/tiny-scripts/general-pmd-synthetic-testing.pyruns · 6,837 bytesdebug/tiny-scripts/idefics-make-tiny-model.pyruns · 1,545 bytesdebug/tiny-scripts/m4-ds-unpack.pyruns · 3,553 bytesdebug/tiny-scripts/mt5-make-tiny-model.pyruns · 3,347 bytesdebug/tiny-scripts/openwebtext-10k.pyruns · 3,078 bytesdebug/tiny-scripts/oscar-en-10k.pyruns · 4,193 bytesdebug/tools.mddoc · 9,945 bytesdebug/torch-distributed-gpu-test.pyruns · 3,787 bytesdebug/torch-distributed-hanging-solutions.mddoc · 2,041 bytesdebug/underflow_overflow.mddoc · 288 bytesdebug/underflow_overflow.pyruns · 13,104 bytesimages/Machine-Learning-Engineering-SKILL.pngasset · 1,581,325 bytesimages/Machine-Learning-Engineering-book-cover-1275x1650.pngasset · 1,189,019 bytesimages/Machine-Learning-Engineering-book-cover-200x259.pngasset · 56,786 bytesimages/Machine-Learning-Engineering-book-cover.oraasset · 2,463,176 bytesimages/Machine-Learning-Engineering-book-cover.pdfasset · 2,002,489 bytesimages/Machine-Learning-Engineering-book-cover.pngasset · 318,841 bytesimages/Machine-Learning-Engineering-book-cover.svgasset · 1,754,255 bytesinference/README.mddoc · 55,205 bytesinference/images/github-vllm-stats-2024-08-24.pngasset · 87,581 bytesinference/images/infer-kv-cache.pngasset · 100,403 bytesinference/images/mha-gqa-mqa-mla.pngasset · 134,881 bytesinference/images/softmax-temperature.pngasset · 64,911 bytesinsights/ai-battlefield.mddoc · 30,919 bytesinsights/how-to-choose-cloud-provider.mddoc · 34,971 bytesinsights/images/640px-Baureihe52Heizer.jpgasset · 58,603 bytesinsights/when-to-upgrade-gpus/README.mddoc · 39,591 bytesinsights/when-to-upgrade-gpus/bench_decompose.pyruns · 8,830 bytesinsights/when-to-upgrade-gpus/bench_flash_attn.pyruns · 1,823 bytesinsights/when-to-upgrade-gpus/benchmark.pyruns · 9,587 bytesinsights/when-to-upgrade-gpus/install.shruns · 4,663 bytesinsights/when-to-upgrade-gpus/plot_results.pyruns · 2,975 bytesinsights/when-to-upgrade-gpus/results_plot.pngasset · 197,132 bytesinsights/when-to-upgrade-gpus/run_sweep.shruns · 1,124 bytesmodel-parallelism/README.mddoc · 87 bytesnetwork/README.mddoc · 153,026 bytesnetwork/benchmarks/README.mddoc · 57,048 bytesnetwork/benchmarks/all_gather_object_vs_all_gather.pyruns · 1,732 bytesnetwork/benchmarks/all_gather_object_vs_all_reduce.pyruns · 1,302 bytesnetwork/benchmarks/all_reduce_bench.pyruns · 898 bytesnetwork/benchmarks/all_reduce_latency_comp.pyruns · 3,365 bytesnetwork/benchmarks/images/all-reduce-bench-latency-8xh200.pngasset · 335,057 bytesnetwork/benchmarks/images/all-reduce-bench-plot-4n.pngasset · 330,720 bytesnetwork/benchmarks/images/all-reduce-bench-plot-8xh200.pngasset · 348,824 bytesnetwork/benchmarks/images/all-reduce-bench-profile-2gib.pngasset · 209,812 bytesnetwork/benchmarks/images/all-reduce-multi-node-bandwidth.pngasset · 137,296 bytesnetwork/benchmarks/images/torch-dist-bench-busbw-8xb200.pngasset · 608,562 bytesnetwork/benchmarks/images/torch-dist-bench-latency-8xb200.pngasset · 583,846 bytesnetwork/benchmarks/torch-dist-bench-pyxis.sbatchruns · 728 bytesnetwork/benchmarks/torch-dist-bench.pyruns · 48,493 bytesnetwork/comms.mddoc · 13,527 bytesnetwork/debug/README.mddoc · 20,080 bytesnetwork/images/all-reduce-bw-2025.pngasset · 157,333 bytesnetwork/images/all-reduce-collective.pngasset · 60,357 bytesnetwork/images/all-reduce-ring-chunk1.pngasset · 177,247 bytesnetwork/images/all-reduce-ring-chunk2.pngasset · 73,855 bytesnetwork/images/all-to-all-bw-2025.pngasset · 151,813 bytesnetwork/images/amd-infinity-arch-MI300X.pngasset · 264,498 bytesnetwork/images/broadcast-ring.pngasset · 64,457 bytesnetwork/images/ccgrid11-low-level-latency.pngasset · 251,943 bytesnetwork/images/ccgrid11-uni-direction-bandwidth.pngasset · 214,865 bytesnetwork/images/collective-all-gather-1.pngasset · 117,371 bytesnetwork/images/collective-all-gather-2.pngasset · 33,525 bytesnetwork/images/collective-all-reduce-1.pngasset · 66,403 bytesnetwork/images/collective-all-reduce-2.pngasset · 41,757 bytesnetwork/images/collective-all-to-all-1.pngasset · 62,088 bytesnetwork/images/collective-all-to-all.pngasset · 58,900 bytesnetwork/images/collective-broadcast-1.pngasset · 61,400 bytesnetwork/images/collective-broadcast-2.pngasset · 24,279 bytesnetwork/images/collective-gather-1.pngasset · 49,409 bytesnetwork/images/collective-gather-2.pngasset · 27,334 bytesnetwork/images/collective-reduce-1.pngasset · 43,953 bytesnetwork/images/collective-reduce-2.pngasset · 27,630 bytesnetwork/images/collective-reduce-scatter.pngasset · 68,898 bytesnetwork/images/collective-scatter-1.pngasset · 47,972 bytesnetwork/images/collective-scatter-2.pngasset · 28,292 bytesnetwork/images/nccl-all-reduce-scan-nvlstree.pngasset · 203,734 bytesnetwork/images/nccl-all-reduce-scan.pngasset · 413,813 bytesorchestration/README.mddoc · 2,469 bytesorchestration/kubernetes/README.mddoc · 2,582 bytesorchestration/slurm/README.mddoc · 1,277 bytesorchestration/slurm/admin.mddoc · 4,225 bytesorchestration/slurm/cron-daily.slurmruns · 690 bytesorchestration/slurm/cron-hourly.slurmruns · 692 bytesorchestration/slurm/example.slurmruns · 2,190 bytesorchestration/slurm/launchers/README.mddoc · 1,089 bytesorchestration/slurm/launchers/accelerate-launcher.slurmruns · 3,055 bytesorchestration/slurm/launchers/lightning-launcher.slurmruns · 2,152 bytesorchestration/slurm/launchers/srun-launcher.slurmruns · 2,650 bytesorchestration/slurm/launchers/torchrun-launcher.slurmruns · 2,775 bytesorchestration/slurm/performance.mddoc · 3,693 bytesorchestration/slurm/undrain-good-nodes.shruns · 1,880 bytesorchestration/slurm/users.mddoc · 46,867 bytesresources/README.mddoc · 7,808 bytesskills/README.mddoc · 1,069 bytesstabs/README.mddoc · 111 bytesstabs/incoming.mddoc · 8,594 bytesstorage/README.mddoc · 64,813 bytesstorage/fio-json-extract.pyruns · 1,055 bytesstorage/fio-scanruns · 2,822 bytesstorage/mmap-io-bench.pyruns · 3,704 bytestesting/README.mddoc · 32,783 bytestesting/testing_utils.pyruns · 38,139 bytestodo.mddoc · 11,781 bytestraining/README.mddoc · 1,101 bytestraining/checkpoints/README.mddoc · 577 bytestraining/checkpoints/torch-checkpoint-convert-to-bf16runs · 1,348 bytestraining/checkpoints/torch-checkpoint-shrink.pyruns · 3,312 bytestraining/datasets.mddoc · 2,557 bytestraining/dtype.mddoc · 12,293 bytestraining/emulate-multi-node.mddoc · 28,984 bytestraining/fault-tolerance/README.mddoc · 41,962 bytestraining/fault-tolerance/fs-watchdog.pyruns · 7,408 bytestraining/fault-tolerance/fs-watchdog.slurmruns · 614 bytestraining/fault-tolerance/slurm-status.pyruns · 6,213 bytestraining/fault-tolerance/slurm-status.slurmruns · 975 bytestraining/hparams.mddoc · 2,328 bytestraining/images/fp16-bf16-fp8.pngasset · 47,288 bytestraining/images/fp32-tf32-fp16-bf16.pngasset · 43,057 bytestraining/images/mixed-precision-fp16.pngasset · 140,509 bytestraining/instabilities/README.mddoc · 5,175 bytestraining/instabilities/images/bloom-176B-success.pngasset · 26,547 bytestraining/instabilities/images/idefics-80b-tr-190-01-image2text.pngasset · 52,589 bytestraining/instabilities/images/idefics-80b-tr-190-01-losses-2023-06-04.pngasset · 51,326 bytestraining/instabilities/images/idefics-80b-tr-190-01-spike-2023-05-27.pngasset · 19,478 bytestraining/instabilities/images/idefics-80b-tr-190-01-spike-recover-2023-05-30.pngasset · 25,808 bytestraining/instabilities/images/llama-7b-grokking-no-zoom.pngasset · 39,990 bytestraining/instabilities/images/llama-7b-grokking.pngasset · 52,129 bytestraining/instabilities/images/pre-bloom-104B-en-fail.pngasset · 35,744 bytestraining/instabilities/images/pre-bloom-tr1-13B-glitch-1-2.pngasset · 14,371 bytestraining/instabilities/images/pre-bloom-tr8-104B-glitch-1.pngasset · 18,601 bytestraining/instabilities/images/pre-bloom-tr8-104B-glitch-5.pngasset · 22,114 bytestraining/instabilities/images/pre-bloom-tr8-104B-glitch-7-10.pngasset · 22,063 bytestraining/instabilities/images/ptl-repeat-data-p1.pngasset · 39,661 bytestraining/instabilities/images/ptl-repeat-data-p2.pngasset · 26,141 bytestraining/instabilities/images/ptl-repeat-data-p3.pngasset · 43,889 bytestraining/instabilities/training-loss-patterns.mddoc · 10,937 bytestraining/model-parallelism/README.mddoc · 56,178 bytestraining/model-parallelism/images/all-reduce-reduce-scatter-all-gather.pngasset · 141,996 bytestraining/model-parallelism/images/deepspeed-ulysses-math.pngasset · 9,052 bytestraining/model-parallelism/images/deepspeed-ulysses.pngasset · 114,290 bytestraining/model-parallelism/images/dist-flash-attn.pngasset · 198,001 bytestraining/model-parallelism/images/parallelism-deepspeed-3d.pngasset · 22,806 bytestraining/model-parallelism/images/parallelism-flexflow.jpegasset · 350,725 bytestraining/model-parallelism/images/parallelism-gpipe-bubble.pngasset · 47,865 bytestraining/model-parallelism/images/parallelism-pp-dualpipe.pngasset · 104,951 bytestraining/model-parallelism/images/parallelism-sagemaker-interleaved-pipeline.pngasset · 20,982 bytestraining/model-parallelism/images/parallelism-tp-independent-gelu.pngasset · 5,846 bytestraining/model-parallelism/images/parallelism-tp-parallel_gemm.pngasset · 129,135 bytestraining/model-parallelism/images/parallelism-tp-parallel_self_attention.pngasset · 166,226 bytestraining/model-parallelism/images/parallelism-tp-parallel_shard_processing.pngasset · 101,490 bytestraining/model-parallelism/images/parallelism-zero-dp-pp.pngasset · 54,989 bytestraining/model-parallelism/images/parallelism-zero.pngasset · 162,554 bytestraining/performance/README.mddoc · 87,060 bytestraining/performance/benchmarks/activation-memory-per-layer.pyruns · 1,767 bytestraining/performance/benchmarks/dataloader/num-workers-bench.pyruns · 1,636 bytestraining/performance/benchmarks/dataloader/pin-memory-non-block-bench.pyruns · 2,156 bytestraining/performance/benchmarks/matrix-shape/swiglu-maf-bench.pyruns · 3,465 bytestraining/performance/benchmarks/numa/numa-set-pynvml.pyruns · 1,967 bytestraining/performance/benchmarks/numa/numa-set.shruns · 1,011 bytestraining/performance/benchmarks/torch-compile-bench.pyruns · 2,348 bytestraining/performance/distributed/torch-dist-mem-usage.pyruns · 3,395 bytestraining/performance/images/a100-server-hwloc.pngasset · 68,514 bytestraining/performance/images/attention-less-heads.pngasset · 107,777 bytestraining/performance/images/flash-attention.pngasset · 60,566 bytestraining/performance/images/tiling.pngasset · 122,838 bytestraining/performance/images/wave-quant.pngasset · 311,618 bytestraining/re-train-hub-models.mddoc · 2,541 bytestraining/reproducibility/README.mddoc · 13,926 bytestraining/tools/main_process_first.pyruns · 6,386 bytestraining/tools/multi-gpu-non-interleaved-print.pyruns · 1,202 bytestraining/tools/printflock.pyruns · 1,903 bytes
No licensed excerpt is available. You can inspect the pinned source.
