https://rocm.docs.amd.com/en/latest/en/7.2.1/about/license.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/about/release-notes.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/compatibility/compatibility-matrix.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/compatibility/ml-compatibility/dgl-compatibility.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/compatibility/ml-compatibility/jax-compatibility.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/compatibility/ml-compatibility/pytorch-compatibility.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/compatibility/ml-compatibility/tensorflow-compatibility.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/ai-pytorch-inception.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/cmake-packages.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/compiler-topics.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/file-reorg.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-arch.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-arch/mi100.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-arch/mi250.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-arch/mi300.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-arch/mi300-mi200-performance-counters.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-arch/mi350-performance-counters.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/conceptual/gpu-isolation.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/contribute/building.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/contribute/contributing.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/contribute/feedback.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/contribute/toolchain.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/Bar-Memory.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/build-rocm.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/deep-learning-rocm.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/gpu-performance/mi300x.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/programming_guide.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/fine-tuning/fine-tuning-and-inference.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/fine-tuning/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/fine-tuning/multi-gpu-fine-tuning-and-inference.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/fine-tuning/overview.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/fine-tuning/single-gpu-fine-tuning-and-inference.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/model-acceleration-libraries.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/model-quantization.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/optimizing-triton-kernel.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/optimizing-with-composable-kernel.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/profiling-and-debugging.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/vllm-optimization.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference-optimization/workload.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/sglang-history.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.10.0-20250812.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.10.1-20250909.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.10.2-20251006.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.11.1-20251103.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.4.3.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.6.4.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.6.6.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.7.3-20250325.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.8.3-20250415.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.8.5-20250513.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.8.5-20250521.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.9.0.1-20250605.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.9.1-20250702.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-0.9.1-20250715.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/vllm-history.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-25.10.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-25.11.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-25.12.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-25.13.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-26.1.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-26.2.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-26.3.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/previous-versions/xdit-history.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/pytorch-inference.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/sglang.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/sglang-distributed.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/sglang-mori-distributed.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/vllm.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/benchmark-docker/vllm-mori-distributed.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/deploy-your-model.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/hugging-face-models.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/llm-inference-frameworks.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/inference/xdit-diffusion-inference.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/install.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/system-setup/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/system-setup/multi-node-setup.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/system-setup/prerequisite-system-validation.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/system-setup/system-health-check.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/jax-maxtext.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/megatron-lm.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/mpt-llm-foundry.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-history.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.11.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.4.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.5.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.7.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.9.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v26.1.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-history.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-primus-migration-guide.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v24.12-dev.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.10.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.11.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.3.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.4.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.5.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.6.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.7.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.8.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.9.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.10.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.11.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.7.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.8.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.9.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v26.1.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.10.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.11.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.8.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.9.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v26.1.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-history.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.10.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.11.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.3.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.4.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.5.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.6.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.7.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.8.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.9.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/primus-megatron.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/primus-pytorch.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/benchmark-docker/pytorch-training.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-ai/training/scale-model-training.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/rocm-for-hpc/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/setting-cus.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/system-debugging.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/system-optimization/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/system-optimization/rdna3-5.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/system-optimization/w6000-v620.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/how-to/tuning-guides/mi300x/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/index.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/api-libraries.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/env-variables.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/glossary.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/glossary/device-hardware.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/glossary/device-software.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/glossary/host-software.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/glossary/performance.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/gpu-arch-specs.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/gpu-atomics-operation.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/graph-safe-support.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/precision-support.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/reference/rocm-tools.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/release/changelog.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/release/versions.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/what-is-rocm.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/genindex.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/404.htmlhttps://rocm.docs.amd.com/en/latest/en/7.2.1/search.html