mirror of
https://github.com/ROCm/ROCm.git
synced 2026-02-11 06:55:06 -05:00
Compare commits
1 Commits
dependabot
...
hipdnn
| Author | SHA1 | Date | |
|---|---|---|---|
|
69360270f5 |
@@ -165,7 +165,6 @@ Flang
|
||||
FlashAttention
|
||||
FlashInfer’s
|
||||
FlashInfer
|
||||
flashinfer
|
||||
FluxBenchmark
|
||||
Fortran
|
||||
Fuyu
|
||||
|
||||
@@ -39,6 +39,7 @@ additional licenses. Please review individual repositories for more information.
|
||||
| [hipBLASLt](https://github.com/ROCm/rocm-libraries/tree/develop/projects/hipblaslt/) | [MIT](https://github.com/ROCm/rocm-libraries/blob/develop/projects/hipblaslt/LICENSE.md) |
|
||||
| [HIPCC](https://github.com/ROCm/llvm-project/tree/amd-staging/amd/hipcc) | [MIT](https://github.com/ROCm/llvm-project/blob/amd-staging/amd/hipcc/LICENSE.txt) |
|
||||
| [hipCUB](https://github.com/ROCm/rocm-libraries/tree/develop/projects/hipcub/) | [Custom](https://github.com/ROCm/rocm-libraries/blob/develop/projects/hipcub/LICENSE.txt) |
|
||||
| [hipDNN](https://github.com/ROCm/rocm-libraries/tree/develop/projects/hipdnn/) | [MIT](https://github.com/ROCm/rocm-libraries/blob/develop/projects/hipdnn/LICENSE.md) |
|
||||
| [hipFFT](https://github.com/ROCm/rocm-libraries/tree/develop/projects/hipfft/) | [MIT](https://github.com/ROCm/rocm-libraries/blob/develop/projects/hipfft/LICENSE.md) |
|
||||
| [hipfort](https://github.com/ROCm/hipfort/) | [MIT](https://github.com/ROCm/hipfort/blob/develop/LICENSE) |
|
||||
| [HIPIFY](https://github.com/ROCm/HIPIFY/) | [MIT](https://github.com/ROCm/HIPIFY/blob/amd-staging/LICENSE.txt) |
|
||||
|
||||
@@ -39,7 +39,7 @@ ROCm Version,7.2.0,7.1.1,7.1.0,7.0.2,7.0.1/7.0.0,6.4.3,6.4.2,6.4.1,6.4.0,6.3.3,6
|
||||
:doc:`Megablocks <../compatibility/ml-compatibility/megablocks-compatibility>` [#megablocks_compat-past-60]_,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,0.7.0,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
|
||||
:doc:`Ray <../compatibility/ml-compatibility/ray-compatibility>` [#ray_compat-past-60]_,N/A,N/A,N/A,N/A,N/A,N/A,N/A,2.48.0.post0,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
|
||||
:doc:`llama.cpp <../compatibility/ml-compatibility/llama-cpp-compatibility>` [#llama-cpp_compat-past-60]_,N/A,N/A,N/A,N/A,b6652,b6356,b6356,b6356,b5997,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
|
||||
:doc:`FlashInfer <../compatibility/ml-compatibility/flashinfer-compatibility>` [#flashinfer_compat-past-60]_,N/A,v0.2.5,N/A,N/A,N/A,N/A,N/A,v0.2.5,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
|
||||
:doc:`FlashInfer <../compatibility/ml-compatibility/flashinfer-compatibility>` [#flashinfer_compat-past-60]_,N/A,N/A,N/A,N/A,N/A,N/A,N/A,v0.2.5,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A,N/A
|
||||
`ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_,1.23.2,1.23.1,1.22.0,1.22.0,1.22.0,1.20.0,1.20.0,1.20.0,1.20.0,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.17.3,1.14.1,1.14.1
|
||||
,,,,,,,,,,,,,,,,,,,,,,,
|
||||
,,,,,,,,,,,,,,,,,,,,,,,
|
||||
|
||||
|
@@ -59,7 +59,6 @@ compatibility and system requirements.
|
||||
:doc:`JAX <../compatibility/ml-compatibility/jax-compatibility>`,0.8.0,0.7.1,0.4.35
|
||||
:doc:`DGL <../compatibility/ml-compatibility/dgl-compatibility>` [#dgl_compat]_,N/A,N/A,2.4.0
|
||||
:doc:`llama.cpp <../compatibility/ml-compatibility/llama-cpp-compatibility>` [#llama-cpp_compat]_,N/A,N/A,b5997
|
||||
:doc:`FlashInfer <../compatibility/ml-compatibility/flashinfer-compatibility>` [#flashinfer_compat]_,N/A,v0.2.5,N/A
|
||||
`ONNX Runtime <https://onnxruntime.ai/docs/build/eps.html#amd-migraphx>`_,1.23.2,1.23.1,1.20.0
|
||||
,,,
|
||||
THIRD PARTY COMMS,.. _thirdpartycomms-support-compatibility-matrix:,,
|
||||
@@ -159,9 +158,8 @@ compatibility and system requirements.
|
||||
|
||||
.. [#os-compatibility] Some operating systems are supported on specific GPUs. For detailed information about operating systems supported on ROCm 7.2.0, see the latest :ref:`supported_distributions`. For version specific information, see `ROCm 7.1.1 <https://rocm.docs.amd.com/projects/install-on-linux/en/docs-7.1.1/reference/system-requirements.html#supported-operating-systems>`__, and `ROCm 6.4.0 <https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.4.0/reference/system-requirements.html#supported-operating-systems>`__.
|
||||
.. [#gpu-compatibility] Some GPUs have limited operating system support. For detailed information about GPUs supporting ROCm 7.2.0, see the latest :ref:`supported_GPUs`. For version specific information, see `ROCm 7.1.1 <https://rocm.docs.amd.com/projects/install-on-linux/en/docs-7.1.1/reference/system-requirements.html#supported-gpus>`__, `ROCm 7.1.0 <https://rocm.docs.amd.com/projects/install-on-linux/en/docs-7.1.0/reference/system-requirements.html#supported-gpus>`__, and `ROCm 6.4.0 <https://rocm.docs.amd.com/projects/install-on-linux/en/docs-6.4.0/reference/system-requirements.html#supported-gpus>`__.
|
||||
.. [#dgl_compat] DGL is only supported on ROCm 7.0.0, 6.4.3 and 6.4.0.
|
||||
.. [#llama-cpp_compat] llama.cpp is only supported on ROCm 7.0.0 and 6.4.x.
|
||||
.. [#flashinfer_compat] FlashInfer is only supported on ROCm 7.1.1 and 6.4.1.
|
||||
.. [#dgl_compat] DGL is supported only on ROCm 7.0.0, ROCm 6.4.3 and ROCm 6.4.0.
|
||||
.. [#llama-cpp_compat] llama.cpp is supported only on ROCm 7.0.0 and ROCm 6.4.x.
|
||||
.. [#mi325x_KVM] For AMD Instinct MI325X KVM SR-IOV users, do not use AMD GPU Driver (amdgpu) 30.20.0.
|
||||
.. [#driver_patch] AMD GPU Driver (amdgpu) 30.10.1 is a quality release that resolves an issue identified in the 30.10 release. There are no other significant changes or feature additions in ROCm 7.0.1 from ROCm 7.0.0. AMD GPU Driver (amdgpu) 30.10.1 is compatible with ROCm 7.0.1 and ROCm 7.0.0.
|
||||
.. [#kfd_support] As of ROCm 6.4.0, forward and backward compatibility between the AMD GPU Driver (amdgpu) and its user space software is provided up to a year apart. For earlier ROCm releases, the compatibility is provided for +/- 2 releases. The supported user space versions on this page were accurate as of the time of initial ROCm release. For the most up-to-date information, see the latest version of this information at `User and AMD GPU Driver support matrix <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/user-kernel-space-compat-matrix.html>`_.
|
||||
@@ -207,13 +205,13 @@ Expand for full historical view of:
|
||||
.. [#os-compatibility-past-60] Some operating systems are supported on specific GPUs. For detailed information, see :ref:`supported_distributions` and select the required ROCm version for version specific support.
|
||||
.. [#gpu-compatibility-past-60] Some GPUs have limited operating system support. For detailed information, see :ref:`supported_GPUs` and select the required ROCm version for version specific support.
|
||||
.. [#tf-mi350-past-60] TensorFlow 2.17.1 is not supported on AMD Instinct MI350 Series GPUs. Use TensorFlow 2.19.1 or 2.18.1 with MI350 Series GPUs instead.
|
||||
.. [#verl_compat-past-60] verl is only supported on ROCm 7.0.0 and 6.2.0.
|
||||
.. [#stanford-megatron-lm_compat-past-60] Stanford Megatron-LM is only supported on ROCm 6.3.0.
|
||||
.. [#dgl_compat-past-60] DGL is only supported on ROCm 7.0.0, 6.4.3 and 6.4.0.
|
||||
.. [#megablocks_compat-past-60] Megablocks is only supported on ROCm 6.3.0.
|
||||
.. [#ray_compat-past-60] Ray is only supported on ROCm 7.0.0 and 6.4.1.
|
||||
.. [#llama-cpp_compat-past-60] llama.cpp is only supported on ROCm 7.0.0 and 6.4.x.
|
||||
.. [#flashinfer_compat-past-60] FlashInfer is only supported on ROCm 7.1.1 and 6.4.1.
|
||||
.. [#verl_compat-past-60] verl is supported only on ROCm 6.2.0.
|
||||
.. [#stanford-megatron-lm_compat-past-60] Stanford Megatron-LM is supported only on ROCm 6.3.0.
|
||||
.. [#dgl_compat-past-60] DGL is supported only on ROCm 7.0.0, ROCm 6.4.3 and ROCm 6.4.0.
|
||||
.. [#megablocks_compat-past-60] Megablocks is supported only on ROCm 6.3.0.
|
||||
.. [#ray_compat-past-60] Ray is supported only on ROCm 6.4.1.
|
||||
.. [#llama-cpp_compat-past-60] llama.cpp is supported only on ROCm 7.0.0 and 6.4.x.
|
||||
.. [#flashinfer_compat-past-60] FlashInfer is supported only on ROCm 6.4.1.
|
||||
.. [#mi325x_KVM-past-60] For AMD Instinct MI325X KVM SR-IOV users, do not use AMD GPU Driver (amdgpu) 30.20.0.
|
||||
.. [#driver_patch-past-60] AMD GPU Driver (amdgpu) 30.10.1 is a quality release that resolves an issue identified in the 30.10 release. There are no other significant changes or feature additions in ROCm 7.0.1 from ROCm 7.0.0. AMD GPU Driver (amdgpu) 30.10.1 is compatible with ROCm 7.0.1 and ROCm 7.0.0.
|
||||
.. [#kfd_support-past-60] As of ROCm 6.4.0, forward and backward compatibility between the AMD GPU Driver (amdgpu) and its user space software is provided up to a year apart. For earlier ROCm releases, the compatibility is provided for +/- 2 releases. The supported user space versions on this page were accurate as of the time of initial ROCm release. For the most up-to-date information, see the latest version of this information at `User and AMD GPU Driver support matrix <https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/user-kernel-space-compat-matrix.html>`_.
|
||||
|
||||
@@ -68,16 +68,6 @@ Click |docker-icon| to view the image on Docker Hub.
|
||||
- Python
|
||||
- GPU
|
||||
|
||||
* - .. raw:: html
|
||||
|
||||
<a href="https://hub.docker.com/layers/rocm/flashinfer/flashinfer-0.2.5.amd2_rocm7.1.1_ubuntu24.04_py3.12_pytorch2.8/images/sha256-9ab6426750a11dbab9bcddeaccaf492683bfd96a1d60b21dd9fc3a609a98175b"><i class="fab fa-docker fa-lg"></i> rocm/flashinfer</a>
|
||||
- `7.1.1 <https://repo.radeon.com/rocm/apt/7.1.1/>`__
|
||||
- `v0.2.5 <https://github.com/flashinfer-ai/flashinfer/releases/tag/v0.2.5>`__
|
||||
- `2.8.0 <https://github.com/ROCm/pytorch/releases/tag/v2.8.0>`__
|
||||
- 24.04
|
||||
- `3.12 <https://www.python.org/downloads/release/python-3129/>`__
|
||||
- MI325X, MI300X
|
||||
|
||||
* - .. raw:: html
|
||||
|
||||
<a href="https://hub.docker.com/layers/rocm/flashinfer/flashinfer-0.2.5_rocm6.4_ubuntu24.04_py3.12_pytorch2.7/images/sha256-558914838821c88c557fb6d42cfbc1bdb67d79d19759f37c764a9ee801f93313"><i class="fab fa-docker fa-lg"></i> rocm/flashinfer</a>
|
||||
@@ -93,21 +83,16 @@ Click |docker-icon| to view the image on Docker Hub.
|
||||
Use cases and recommendations
|
||||
================================================================================
|
||||
|
||||
FlashInfer on ROCm enables you to perform LLM inference for both prefill and decode:
|
||||
during prefill, your model efficiently processes input prompts to build KV caches
|
||||
and internal activations; during decode, it generates tokens sequentially based on
|
||||
prior outputs and context. Use the attention mode supported upstream (Multi-Head
|
||||
Attention, Grouped-Query Attention, or Multi-Query Attention) that matches your
|
||||
model configuration.
|
||||
The release of FlashInfer on ROCm provides the decode functionality for LLM inferencing.
|
||||
In the decode phase, tokens are generated sequentially, with the model predicting each new
|
||||
token based on the previously generated tokens and the input context.
|
||||
|
||||
FlashInfer on ROCm also includes capabilities such as load balancing,
|
||||
sparse and dense attention optimizations, and single and batch decode, alongside
|
||||
prefill for high‑performance execution on MI300X GPUs.
|
||||
FlashInfer on ROCm brings over upstream features such as load balancing, sparse and dense
|
||||
attention optimizations, and batching support, enabling efficient execution on AMD Instinct™ MI300X GPUs.
|
||||
|
||||
For currently supported use cases and recommendations, refer to the `AMD ROCm blog <https://rocm.blogs.amd.com/search.html?q=flashinfer>`__,
|
||||
Because large LLMs often require substantial KV caches or long context windows, FlashInfer on ROCm
|
||||
also implements cascade attention from upstream to reduce memory usage.
|
||||
|
||||
For currently supported use cases and recommendations, refer to the `AMD ROCm blog <https://rocm.blogs.amd.com/>`__,
|
||||
where you can search for examples and best practices to optimize your workloads on AMD GPUs.
|
||||
|
||||
Previous versions
|
||||
===============================================================================
|
||||
See :doc:`rocm-install-on-linux:install/3rd-party/previous-versions/flashinfer-history` to find documentation for previous releases
|
||||
of the ``ROCm/flashinfer`` Docker image.
|
||||
|
||||
@@ -138,14 +138,12 @@ article_pages = [
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.8", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.9", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.10", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-v25.11", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/megatron-lm-primus-migration-guide", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/primus-megatron", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.7", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.8", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.9", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.10", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-megatron-v25.11", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/pytorch-training", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-history", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.3", "os": ["linux"]},
|
||||
@@ -156,18 +154,14 @@ article_pages = [
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.8", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.9", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.10", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/pytorch-training-v25.11", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/primus-pytorch", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.8", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.9", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.10", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/primus-pytorch-v25.11", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/jax-maxtext", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-history", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.4", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.5", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.9", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/previous-versions/jax-maxtext-v25.11", "os": ["linux"]},
|
||||
{"file": "how-to/rocm-for-ai/training/benchmark-docker/mpt-llm-foundry", "os": ["linux"]},
|
||||
|
||||
{"file": "how-to/rocm-for-ai/fine-tuning/index", "os": ["linux"]},
|
||||
|
||||
@@ -1,11 +1,11 @@
|
||||
dockers:
|
||||
- pull_tag: rocm/jax-training:maxtext-v26.1
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/jax-training/maxtext-v26.1/images/sha256-901083bde353fe6362ada3036e452c792b2c96124e5900f4e9b5946c02ff9d6a
|
||||
- pull_tag: rocm/jax-training:maxtext-v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/jax-training/maxtext-v25.11/images/sha256-18e4d8f0b8ce7a7422c58046940dd5f32249960449fca09a562b65fb8eb1562a
|
||||
components:
|
||||
ROCm: 7.1.1
|
||||
JAX: 0.8.2
|
||||
ROCm: 7.1.0
|
||||
JAX: 0.7.1
|
||||
Python: 3.12
|
||||
Transformer Engine: 2.8.0.dev0+aec00a7f
|
||||
Transformer Engine: 2.4.0.dev0+281042de
|
||||
hipBLASLt: 1.2.x
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
@@ -15,29 +15,21 @@ model_groups:
|
||||
mad_tag: jax_maxtext_train_llama-2-7b
|
||||
model_repo: Llama-2-7B
|
||||
precision: bf16
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_llama2_7b.yml
|
||||
gfx942: env_scripts/llama2_7b.yml
|
||||
multinode_training_script: llama2_7b_multinode.sh
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
- model: Llama 2 70B
|
||||
mad_tag: jax_maxtext_train_llama-2-70b
|
||||
model_repo: Llama-2-70B
|
||||
precision: bf16
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_llama2_70b.yml
|
||||
gfx942: env_scripts/llama2_70b.yml
|
||||
multinode_training_script: llama2_70b_multinode.sh
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
- model: Llama 3 8B (multi-node)
|
||||
mad_tag: jax_maxtext_train_llama-3-8b
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_llama3_8b.yml
|
||||
gfx942: env_scripts/llama3_8b.yml
|
||||
multinode_training_script: llama3_8b_multinode.sh
|
||||
doc_options: ["multi-node"]
|
||||
- model: Llama 3 70B (multi-node)
|
||||
mad_tag: jax_maxtext_train_llama-3-70b
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_llama3_70b.yml
|
||||
gfx942: env_scripts/llama3_70b.yml
|
||||
multinode_training_script: llama3_70b_multinode.sh
|
||||
doc_options: ["multi-node"]
|
||||
- model: Llama 3.1 8B
|
||||
mad_tag: jax_maxtext_train_llama-3.1-8b
|
||||
@@ -49,21 +41,11 @@ model_groups:
|
||||
model_repo: Llama-3.1-70B
|
||||
precision: bf16
|
||||
doc_options: ["single-node"]
|
||||
- model: Llama 3.1 405B (multi-node)
|
||||
mad_tag: jax_maxtext_train_llama-3.1-405b
|
||||
model_repo: Llama-3.1-405B
|
||||
precision: bf16
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_llama3_405b.yml
|
||||
doc_options: ["multi-node"]
|
||||
- model: Llama 3.3 70B
|
||||
mad_tag: jax_maxtext_train_llama-3.3-70b
|
||||
model_repo: Llama-3.3-70B
|
||||
precision: bf16
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_llama3.3_70b.yml
|
||||
gfx942: env_scripts/llama3.3_70b.yml
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
doc_options: ["single-node"]
|
||||
- group: DeepSeek
|
||||
tag: deepseek
|
||||
models:
|
||||
@@ -71,10 +53,7 @@ model_groups:
|
||||
mad_tag: jax_maxtext_train_deepseek-v2-lite-16b
|
||||
model_repo: DeepSeek-V2-lite
|
||||
precision: bf16
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_deepseek2_16b.yml
|
||||
gfx942: env_scripts/deepseek2_16b.yml
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
doc_options: ["single-node"]
|
||||
- group: Mistral AI
|
||||
tag: mistral
|
||||
models:
|
||||
@@ -82,7 +61,4 @@ model_groups:
|
||||
mad_tag: jax_maxtext_train_mixtral-8x7b
|
||||
model_repo: Mixtral-8x7B
|
||||
precision: bf16
|
||||
multinode_config:
|
||||
gfx950: env_scripts/gfx950_mixtral_8x7b.yml
|
||||
gfx942: env_scripts/llama3_8x7b.yml
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
doc_options: ["single-node"]
|
||||
|
||||
@@ -1,13 +1,15 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v26.1
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v26.1/images/sha256-4fc8808bdb14117c6af7f38d79c809056e6fdbfd530c1fabbb61d097ddaf820d
|
||||
pull_tag: rocm/primus:v25.10
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
Primus: 0.3.0
|
||||
Primus Turbo: 0.1.1
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.6.0.dev0+f141f34b
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 34459f66ea
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
Triton: 3.4.0
|
||||
RCCL: 2.27.7
|
||||
model_groups:
|
||||
|
||||
@@ -1,64 +0,0 @@
|
||||
dockers:
|
||||
- pull_tag: rocm/jax-training:maxtext-v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/jax-training/maxtext-v25.11/images/sha256-18e4d8f0b8ce7a7422c58046940dd5f32249960449fca09a562b65fb8eb1562a
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
JAX: 0.7.1
|
||||
Python: 3.12
|
||||
Transformer Engine: 2.4.0.dev0+281042de
|
||||
hipBLASLt: 1.2.x
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
models:
|
||||
- model: Llama 2 7B
|
||||
mad_tag: jax_maxtext_train_llama-2-7b
|
||||
model_repo: Llama-2-7B
|
||||
precision: bf16
|
||||
multinode_training_script: llama2_7b_multinode.sh
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
- model: Llama 2 70B
|
||||
mad_tag: jax_maxtext_train_llama-2-70b
|
||||
model_repo: Llama-2-70B
|
||||
precision: bf16
|
||||
multinode_training_script: llama2_70b_multinode.sh
|
||||
doc_options: ["single-node", "multi-node"]
|
||||
- model: Llama 3 8B (multi-node)
|
||||
mad_tag: jax_maxtext_train_llama-3-8b
|
||||
multinode_training_script: llama3_8b_multinode.sh
|
||||
doc_options: ["multi-node"]
|
||||
- model: Llama 3 70B (multi-node)
|
||||
mad_tag: jax_maxtext_train_llama-3-70b
|
||||
multinode_training_script: llama3_70b_multinode.sh
|
||||
doc_options: ["multi-node"]
|
||||
- model: Llama 3.1 8B
|
||||
mad_tag: jax_maxtext_train_llama-3.1-8b
|
||||
model_repo: Llama-3.1-8B
|
||||
precision: bf16
|
||||
doc_options: ["single-node"]
|
||||
- model: Llama 3.1 70B
|
||||
mad_tag: jax_maxtext_train_llama-3.1-70b
|
||||
model_repo: Llama-3.1-70B
|
||||
precision: bf16
|
||||
doc_options: ["single-node"]
|
||||
- model: Llama 3.3 70B
|
||||
mad_tag: jax_maxtext_train_llama-3.3-70b
|
||||
model_repo: Llama-3.3-70B
|
||||
precision: bf16
|
||||
doc_options: ["single-node"]
|
||||
- group: DeepSeek
|
||||
tag: deepseek
|
||||
models:
|
||||
- model: DeepSeek-V2-Lite (16B)
|
||||
mad_tag: jax_maxtext_train_deepseek-v2-lite-16b
|
||||
model_repo: DeepSeek-V2-lite
|
||||
precision: bf16
|
||||
doc_options: ["single-node"]
|
||||
- group: Mistral AI
|
||||
tag: mistral
|
||||
models:
|
||||
- model: Mixtral 8x7B
|
||||
mad_tag: jax_maxtext_train_mixtral-8x7b
|
||||
model_repo: Mixtral-8x7B
|
||||
precision: bf16
|
||||
doc_options: ["single-node"]
|
||||
@@ -1,47 +0,0 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.11/images/sha256-71aa65a9bfc8e9dd18bce5b68c81caff864f223e9afa75dc1b719671a1f4a3c3
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
Triton: 3.4.0
|
||||
RCCL: 2.27.7
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
models:
|
||||
- model: Llama 3.3 70B
|
||||
mad_tag: pyt_megatron_lm_train_llama-3.3-70b
|
||||
- model: Llama 3.1 8B
|
||||
mad_tag: pyt_megatron_lm_train_llama-3.1-8b
|
||||
- model: Llama 3.1 70B
|
||||
mad_tag: pyt_megatron_lm_train_llama-3.1-70b
|
||||
- model: Llama 2 7B
|
||||
mad_tag: pyt_megatron_lm_train_llama-2-7b
|
||||
- model: Llama 2 70B
|
||||
mad_tag: pyt_megatron_lm_train_llama-2-70b
|
||||
- group: DeepSeek
|
||||
tag: deepseek
|
||||
models:
|
||||
- model: DeepSeek-V3 (proxy)
|
||||
mad_tag: pyt_megatron_lm_train_deepseek-v3-proxy
|
||||
- model: DeepSeek-V2-Lite
|
||||
mad_tag: pyt_megatron_lm_train_deepseek-v2-lite-16b
|
||||
- group: Mistral AI
|
||||
tag: mistral
|
||||
models:
|
||||
- model: Mixtral 8x7B
|
||||
mad_tag: pyt_megatron_lm_train_mixtral-8x7b
|
||||
- model: Mixtral 8x22B (proxy)
|
||||
mad_tag: pyt_megatron_lm_train_mixtral-8x22b-proxy
|
||||
- group: Qwen
|
||||
tag: qwen
|
||||
models:
|
||||
- model: Qwen 2.5 7B
|
||||
mad_tag: pyt_megatron_lm_train_qwen2.5-7b
|
||||
- model: Qwen 2.5 72B
|
||||
mad_tag: pyt_megatron_lm_train_qwen2.5-72b
|
||||
@@ -1,58 +0,0 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.11/images/sha256-71aa65a9bfc8e9dd18bce5b68c81caff864f223e9afa75dc1b719671a1f4a3c3
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
Triton: 3.4.0
|
||||
RCCL: 2.27.7
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
models:
|
||||
- model: Llama 3.3 70B
|
||||
mad_tag: primus_pyt_megatron_lm_train_llama-3.3-70b
|
||||
config_name: llama3.3_70B-pretrain.yaml
|
||||
- model: Llama 3.1 70B
|
||||
mad_tag: primus_pyt_megatron_lm_train_llama-3.1-70b
|
||||
config_name: llama3.1_70B-pretrain.yaml
|
||||
- model: Llama 3.1 8B
|
||||
mad_tag: primus_pyt_megatron_lm_train_llama-3.1-8b
|
||||
config_name: llama3.1_8B-pretrain.yaml
|
||||
- model: Llama 2 7B
|
||||
mad_tag: primus_pyt_megatron_lm_train_llama-2-7b
|
||||
config_name: llama2_7B-pretrain.yaml
|
||||
- model: Llama 2 70B
|
||||
mad_tag: primus_pyt_megatron_lm_train_llama-2-70b
|
||||
config_name: llama2_70B-pretrain.yaml
|
||||
- group: DeepSeek
|
||||
tag: deepseek
|
||||
models:
|
||||
- model: DeepSeek-V3 (proxy)
|
||||
mad_tag: primus_pyt_megatron_lm_train_deepseek-v3-proxy
|
||||
config_name: deepseek_v3-pretrain.yaml
|
||||
- model: DeepSeek-V2-Lite
|
||||
mad_tag: primus_pyt_megatron_lm_train_deepseek-v2-lite-16b
|
||||
config_name: deepseek_v2_lite-pretrain.yaml
|
||||
- group: Mistral AI
|
||||
tag: mistral
|
||||
models:
|
||||
- model: Mixtral 8x7B
|
||||
mad_tag: primus_pyt_megatron_lm_train_mixtral-8x7b
|
||||
config_name: mixtral_8x7B_v0.1-pretrain.yaml
|
||||
- model: Mixtral 8x22B (proxy)
|
||||
mad_tag: primus_pyt_megatron_lm_train_mixtral-8x22b-proxy
|
||||
config_name: mixtral_8x22B_v0.1-pretrain.yaml
|
||||
- group: Qwen
|
||||
tag: qwen
|
||||
models:
|
||||
- model: Qwen 2.5 7B
|
||||
mad_tag: primus_pyt_megatron_lm_train_qwen2.5-7b
|
||||
config_name: primus_qwen2.5_7B-pretrain.yaml
|
||||
- model: Qwen 2.5 72B
|
||||
mad_tag: primus_pyt_megatron_lm_train_qwen2.5-72b
|
||||
config_name: qwen2.5_72B-pretrain.yaml
|
||||
@@ -1,32 +0,0 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.11/images/sha256-71aa65a9bfc8e9dd18bce5b68c81caff864f223e9afa75dc1b719671a1f4a3c3
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
models:
|
||||
- model: Llama 3.1 8B
|
||||
mad_tag: primus_pyt_train_llama-3.1-8b
|
||||
model_repo: Llama-3.1-8B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.1-8B
|
||||
precision: BF16
|
||||
- model: Llama 3.1 70B
|
||||
mad_tag: primus_pyt_train_llama-3.1-70b
|
||||
model_repo: Llama-3.1-70B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.1-70B
|
||||
precision: BF16
|
||||
- group: DeepSeek
|
||||
tag: deepseek
|
||||
models:
|
||||
- model: DeepSeek V3 16B
|
||||
mad_tag: primus_pyt_train_deepseek-v3-16b
|
||||
model_repo: DeepSeek-V3
|
||||
url: https://huggingface.co/deepseek-ai/DeepSeek-V3
|
||||
precision: BF16
|
||||
@@ -1,195 +0,0 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.11/images/sha256-71aa65a9bfc8e9dd18bce5b68c81caff864f223e9afa75dc1b719671a1f4a3c3
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
models:
|
||||
- model: Llama 4 Scout 17B-16E
|
||||
mad_tag: pyt_train_llama-4-scout-17b-16e
|
||||
model_repo: Llama-4-17B_16E
|
||||
url: https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Llama 3.3 70B
|
||||
mad_tag: pyt_train_llama-3.3-70b
|
||||
model_repo: Llama-3.3-70B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora, finetune_qlora]
|
||||
- model: Llama 3.2 1B
|
||||
mad_tag: pyt_train_llama-3.2-1b
|
||||
model_repo: Llama-3.2-1B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.2-1B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Llama 3.2 3B
|
||||
mad_tag: pyt_train_llama-3.2-3b
|
||||
model_repo: Llama-3.2-3B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.2-3B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Llama 3.2 Vision 11B
|
||||
mad_tag: pyt_train_llama-3.2-vision-11b
|
||||
model_repo: Llama-3.2-Vision-11B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.2-11B-Vision
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw]
|
||||
- model: Llama 3.2 Vision 90B
|
||||
mad_tag: pyt_train_llama-3.2-vision-90b
|
||||
model_repo: Llama-3.2-Vision-90B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.2-90B-Vision
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw]
|
||||
- model: Llama 3.1 8B
|
||||
mad_tag: pyt_train_llama-3.1-8b
|
||||
model_repo: Llama-3.1-8B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.1-8B
|
||||
precision: BF16
|
||||
training_modes: [pretrain, finetune_fw, finetune_lora, HF_pretrain]
|
||||
- model: Llama 3.1 70B
|
||||
mad_tag: pyt_train_llama-3.1-70b
|
||||
model_repo: Llama-3.1-70B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct
|
||||
precision: BF16
|
||||
training_modes: [pretrain, finetune_fw, finetune_lora]
|
||||
- model: Llama 3.1 405B
|
||||
mad_tag: pyt_train_llama-3.1-405b
|
||||
model_repo: Llama-3.1-405B
|
||||
url: https://huggingface.co/meta-llama/Llama-3.1-405B
|
||||
precision: BF16
|
||||
training_modes: [finetune_qlora]
|
||||
- model: Llama 3 8B
|
||||
mad_tag: pyt_train_llama-3-8b
|
||||
model_repo: Llama-3-8B
|
||||
url: https://huggingface.co/meta-llama/Meta-Llama-3-8B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Llama 3 70B
|
||||
mad_tag: pyt_train_llama-3-70b
|
||||
model_repo: Llama-3-70B
|
||||
url: https://huggingface.co/meta-llama/Meta-Llama-3-70B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Llama 2 7B
|
||||
mad_tag: pyt_train_llama-2-7b
|
||||
model_repo: Llama-2-7B
|
||||
url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora, finetune_qlora]
|
||||
- model: Llama 2 13B
|
||||
mad_tag: pyt_train_llama-2-13b
|
||||
model_repo: Llama-2-13B
|
||||
url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Llama 2 70B
|
||||
mad_tag: pyt_train_llama-2-70b
|
||||
model_repo: Llama-2-70B
|
||||
url: https://github.com/meta-llama/llama-models/tree/main/models/llama2
|
||||
precision: BF16
|
||||
training_modes: [finetune_lora, finetune_qlora]
|
||||
- group: OpenAI
|
||||
tag: openai
|
||||
models:
|
||||
- model: GPT OSS 20B
|
||||
mad_tag: pyt_train_gpt_oss_20b
|
||||
model_repo: GPT-OSS-20B
|
||||
url: https://huggingface.co/openai/gpt-oss-20b
|
||||
precision: BF16
|
||||
training_modes: [HF_finetune_lora]
|
||||
- model: GPT OSS 120B
|
||||
mad_tag: pyt_train_gpt_oss_120b
|
||||
model_repo: GPT-OSS-120B
|
||||
url: https://huggingface.co/openai/gpt-oss-120b
|
||||
precision: BF16
|
||||
training_modes: [HF_finetune_lora]
|
||||
- group: DeepSeek
|
||||
tag: deepseek
|
||||
models:
|
||||
- model: DeepSeek V2 16B
|
||||
mad_tag: primus_pyt_train_deepseek-v2
|
||||
model_repo: DeepSeek-V2
|
||||
url: https://huggingface.co/deepseek-ai/DeepSeek-V2
|
||||
precision: BF16
|
||||
training_modes: [pretrain]
|
||||
- group: Qwen
|
||||
tag: qwen
|
||||
models:
|
||||
- model: Qwen 3 8B
|
||||
mad_tag: pyt_train_qwen3-8b
|
||||
model_repo: Qwen3-8B
|
||||
url: https://huggingface.co/Qwen/Qwen3-8B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Qwen 3 32B
|
||||
mad_tag: pyt_train_qwen3-32b
|
||||
model_repo: Qwen3-32
|
||||
url: https://huggingface.co/Qwen/Qwen3-32B
|
||||
precision: BF16
|
||||
training_modes: [finetune_lora]
|
||||
- model: Qwen 2.5 32B
|
||||
mad_tag: pyt_train_qwen2.5-32b
|
||||
model_repo: Qwen2.5-32B
|
||||
url: https://huggingface.co/Qwen/Qwen2.5-32B
|
||||
precision: BF16
|
||||
training_modes: [finetune_lora]
|
||||
- model: Qwen 2.5 72B
|
||||
mad_tag: pyt_train_qwen2.5-72b
|
||||
model_repo: Qwen2.5-72B
|
||||
url: https://huggingface.co/Qwen/Qwen2.5-72B
|
||||
precision: BF16
|
||||
training_modes: [finetune_lora]
|
||||
- model: Qwen 2 1.5B
|
||||
mad_tag: pyt_train_qwen2-1.5b
|
||||
model_repo: Qwen2-1.5B
|
||||
url: https://huggingface.co/Qwen/Qwen2-1.5B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- model: Qwen 2 7B
|
||||
mad_tag: pyt_train_qwen2-7b
|
||||
model_repo: Qwen2-7B
|
||||
url: https://huggingface.co/Qwen/Qwen2-7B
|
||||
precision: BF16
|
||||
training_modes: [finetune_fw, finetune_lora]
|
||||
- group: Stable Diffusion
|
||||
tag: sd
|
||||
models:
|
||||
- model: Stable Diffusion XL
|
||||
mad_tag: pyt_huggingface_stable_diffusion_xl_2k_lora_finetuning
|
||||
model_repo: SDXL
|
||||
url: https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0
|
||||
precision: BF16
|
||||
training_modes: [posttrain]
|
||||
- group: Flux
|
||||
tag: flux
|
||||
models:
|
||||
- model: FLUX.1-dev
|
||||
mad_tag: pyt_train_flux
|
||||
model_repo: Flux
|
||||
url: https://huggingface.co/black-forest-labs/FLUX.1-dev
|
||||
precision: BF16
|
||||
training_modes: [posttrain]
|
||||
- group: NCF
|
||||
tag: ncf
|
||||
models:
|
||||
- model: NCF
|
||||
mad_tag: pyt_ncf_training
|
||||
model_repo:
|
||||
url: https://github.com/ROCm/FluxBenchmark
|
||||
precision: FP32
|
||||
- group: DLRM
|
||||
tag: dlrm
|
||||
models:
|
||||
- model: DLRM v2
|
||||
mad_tag: pyt_train_dlrm
|
||||
model_repo: DLRM
|
||||
url: https://github.com/AMD-AGI/DLRMBenchmark
|
||||
training_modes: [pretrain]
|
||||
@@ -1,13 +1,13 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v26.1
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v26.1/images/sha256-4fc8808bdb14117c6af7f38d79c809056e6fdbfd530c1fabbb61d097ddaf820d
|
||||
pull_tag: rocm/primus:v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.6.0.dev0+f141f34b
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 34459f66ea
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
Triton: 3.4.0
|
||||
RCCL: 2.27.7
|
||||
model_groups:
|
||||
|
||||
@@ -1,13 +1,13 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v26.1
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v26.1/images/sha256-4fc8808bdb14117c6af7f38d79c809056e6fdbfd530c1fabbb61d097ddaf820d
|
||||
pull_tag: rocm/primus:v25.11
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.6.0.dev0+f141f34b
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 34459f66ea
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
|
||||
@@ -1,13 +1,15 @@
|
||||
docker:
|
||||
pull_tag: rocm/primus:v26.1
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v26.1/images/sha256-4fc8808bdb14117c6af7f38d79c809056e6fdbfd530c1fabbb61d097ddaf820d
|
||||
pull_tag: rocm/primus:v25.10
|
||||
docker_hub_url: https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197
|
||||
components:
|
||||
ROCm: 7.1.0
|
||||
Primus: 0.3.0
|
||||
Primus Turbo: 0.1.1
|
||||
PyTorch: 2.10.0.dev20251112+rocm7.1
|
||||
Python: "3.10"
|
||||
Transformer Engine: 2.6.0.dev0+f141f34b
|
||||
Transformer Engine: 2.4.0.dev0+32e2d1d4
|
||||
Flash Attention: 2.8.3
|
||||
hipBLASLt: 34459f66ea
|
||||
hipBLASLt: 1.2.0-09ab7153e2
|
||||
model_groups:
|
||||
- group: Meta Llama
|
||||
tag: llama
|
||||
|
||||
@@ -35,6 +35,13 @@ It includes the following software components:
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. note::
|
||||
|
||||
The ``rocm/jax-training:maxtext-v25.9`` has been updated to
|
||||
``rocm/jax-training:maxtext-v25.9.1``. This revision should include
|
||||
a fix to address segmentation fault issues during launch. See the
|
||||
:doc:`versioned documentation <previous-versions/jax-maxtext-v25.9>`.
|
||||
|
||||
MaxText with on ROCm provides the following key features to train large language models efficiently:
|
||||
|
||||
- Transformer Engine (TE)
|
||||
@@ -47,7 +54,7 @@ MaxText with on ROCm provides the following key features to train large language
|
||||
|
||||
- NANOO FP8 (for MI300X series GPUs) and FP8 (for MI355X and MI350X) quantization support
|
||||
|
||||
.. _amd-maxtext-model-support-v26.1:
|
||||
.. _amd-maxtext-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
@@ -129,7 +136,7 @@ Use the following command to pull the Docker image from Docker Hub.
|
||||
|
||||
docker pull {{ docker.pull_tag }}
|
||||
|
||||
.. _amd-maxtext-multi-node-setup-v26.1:
|
||||
.. _amd-maxtext-multi-node-setup-v25.11:
|
||||
|
||||
Multi-node configuration
|
||||
------------------------
|
||||
@@ -137,7 +144,7 @@ Multi-node configuration
|
||||
See :doc:`/how-to/rocm-for-ai/system-setup/multi-node-setup` to configure your
|
||||
environment for multi-node training.
|
||||
|
||||
.. _amd-maxtext-get-started-v26.1:
|
||||
.. _amd-maxtext-get-started-v25.11:
|
||||
|
||||
Benchmarking
|
||||
============
|
||||
@@ -162,7 +169,7 @@ benchmark results:
|
||||
.. tab-item:: MAD-integrated benchmarking
|
||||
|
||||
The following run command is tailored to {{ model.model }}.
|
||||
See :ref:`amd-maxtext-model-support-v26.1` to switch to another available model.
|
||||
See :ref:`amd-maxtext-model-support-v25.11` to switch to another available model.
|
||||
|
||||
1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
|
||||
directory and install the required packages on the host machine.
|
||||
@@ -193,7 +200,7 @@ benchmark results:
|
||||
.. tab-item:: Standalone benchmarking
|
||||
|
||||
The following commands are optimized for {{ model.model }}. See
|
||||
:ref:`amd-maxtext-model-support-v26.1` to switch to another
|
||||
:ref:`amd-maxtext-model-support-v25.11` to switch to another
|
||||
available model. Some instructions and resources might not be
|
||||
available for all models and configurations.
|
||||
|
||||
@@ -289,57 +296,56 @@ benchmark results:
|
||||
{% endif %}
|
||||
|
||||
{% endif %}
|
||||
{% if model.multinode_config and "multi-node" in model.doc_options %}
|
||||
{% if model.multinode_training_script and "multi-node" in model.doc_options %}
|
||||
.. rubric:: Multi-node training
|
||||
|
||||
The following SLURM scripts will launch the Docker container and
|
||||
run the benchmark. Run them outside of any Docker container. The
|
||||
unified multi-node benchmark script accepts a configuration file
|
||||
that specifies the model and training parameters.
|
||||
The following examples use SLURM to run on multiple nodes.
|
||||
|
||||
.. code-block:: shell
|
||||
.. note::
|
||||
|
||||
sbatch -N <NUM_NODES> jax_maxtext_multinode_benchmark.sh <config_file.yml> [docker_image]
|
||||
The following scripts will launch the Docker container and run the
|
||||
benchmark. Run them outside of any Docker container.
|
||||
|
||||
<NUM_NODES>
|
||||
The number of nodes to use for training (for example, 2, 4,
|
||||
8).
|
||||
1. Make sure ``$HF_HOME`` is set before running the test. See
|
||||
`ROCm benchmarking <https://github.com/ROCm/MAD/blob/develop/scripts/jax-maxtext/gpu-rocm/readme.md>`__
|
||||
for more details on downloading the Llama models before running the
|
||||
benchmark.
|
||||
|
||||
<config_file.yml>
|
||||
Path to the YAML configuration file containing model and
|
||||
training parameters. Configuration files are available in the
|
||||
``scripts/jax-maxtext/env_scripts/`` directory for different
|
||||
models and GPU architectures.
|
||||
2. To run multi-node training for {{ model.model }},
|
||||
use the
|
||||
`multi-node training script <https://github.com/ROCm/MAD/blob/develop/scripts/jax-maxtext/gpu-rocm/{{ model.multinode_training_script }}>`__
|
||||
under the ``scripts/jax-maxtext/gpu-rocm/`` directory.
|
||||
|
||||
[docker_image] (optional)
|
||||
The Docker image to use. If not specified, it defaults to
|
||||
``rocm/jax-training:maxtext-v26.1``.
|
||||
3. Run the multi-node training benchmark script.
|
||||
|
||||
For example, to run a multi-node training benchmark on {{ model.model }}:
|
||||
.. code-block:: shell
|
||||
|
||||
.. tab-set::
|
||||
sbatch -N <num_nodes> {{ model.multinode_training_script }}
|
||||
|
||||
{% if model.multinode_config.gfx950 %}
|
||||
.. tab-item:: MI355X and MI350X (gfx950)
|
||||
.. rubric:: Profiling with rocprofv3
|
||||
|
||||
.. code-block:: bash
|
||||
If you need to collect a trace and the JAX profiler isn't working, use ``rocprofv3`` provided by the :doc:`ROCprofiler-SDK <rocprofiler-sdk:index>` as a workaround. For example:
|
||||
|
||||
sbatch -N 4 jax_maxtext_multinode_benchmark.sh {{ model.multinode_config.gfx950 }}
|
||||
{% endif %}
|
||||
.. code-block:: bash
|
||||
|
||||
{% if model.multinode_config.gfx942 %}
|
||||
.. tab-item:: MI325X and MI300X (gfx942)
|
||||
rocprofv3 \
|
||||
--hip-trace \
|
||||
--kernel-trace \
|
||||
--memory-copy-trace \
|
||||
--rccl-trace \
|
||||
--output-format pftrace \
|
||||
-d ./v3_traces \ # output directory
|
||||
-- ./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }} # or desired command
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
sbatch -N 4 jax_maxtext_multinode_benchmark.sh {{ model.multinode_config.gfx942 }}
|
||||
{% endif %}
|
||||
You can set the directory where you want the .json traces to be
|
||||
saved using ``-d <TRACE_DIRECTORY>``. The resulting traces can be
|
||||
opened in Perfetto: `<https://ui.perfetto.dev/>`__.
|
||||
|
||||
{% else %}
|
||||
.. rubric:: Multi-node training
|
||||
|
||||
For multi-node training examples, choose a model from :ref:`amd-maxtext-model-support-v26.1`
|
||||
with an available `multi-node training script <https://github.com/ROCm/MAD/tree/develop/scripts/jax-maxtext/env_scripts>`__.
|
||||
For multi-node training examples, choose a model from :ref:`amd-maxtext-model-support-v25.11`
|
||||
with an available `multi-node training script <https://github.com/ROCm/MAD/tree/develop/scripts/jax-maxtext/gpu-rocm>`__.
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
@@ -347,13 +353,35 @@ benchmark results:
|
||||
Known issues
|
||||
============
|
||||
|
||||
- You might see NaNs in the losses when setting ``packing=True``. As
|
||||
a workaround, turn off input sequence packing (``packing=False``).
|
||||
- Minor performance regression (< 4%) for BF16 quantization in Llama models and Mixtral 8x7b.
|
||||
|
||||
- You might see minor loss spikes, or loss curve may have slightly higher
|
||||
convergence end values compared to the previous ``jax-training`` image.
|
||||
|
||||
- For FP8 training on MI355, many models will display a warning message like:
|
||||
``Warning: Latency not found for MI_M=16, MI_N=16, MI_K=128,
|
||||
mi_input_type=BFloat8Float8_fnuz. Returning latency value of 32 (really
|
||||
slow).`` The compile step may take longer than usual, but training will run.
|
||||
This will be fixed in a future release.
|
||||
|
||||
- Docker ``rocm/jax-training:maxtext-v26.1`` does not include `Primus
|
||||
<https://github.com/AMD-AGI/Primus/tree/main>`__. It is planned to be
|
||||
supported in a future release.
|
||||
- The built-in JAX profiler isn't working. If you need to collect a trace and
|
||||
the JAX profiler isn't working, use ``rocprofv3`` provided by the
|
||||
:doc:`ROCprofiler-SDK <rocprofiler-sdk:index>` as a workaround. For example:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
rocprofv3 \
|
||||
--hip-trace \
|
||||
--kernel-trace \
|
||||
--memory-copy-trace \
|
||||
--rccl-trace \
|
||||
--output-format pftrace \
|
||||
-d ./v3_traces \ # output directory
|
||||
-- ./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }} # or desired command
|
||||
|
||||
You can set the directory where you want the .json traces to be
|
||||
saved using ``-d <TRACE_DIRECTORY>``. The resulting traces can be
|
||||
opened in Perfetto: `<https://ui.perfetto.dev/>`__.
|
||||
|
||||
Further reading
|
||||
===============
|
||||
|
||||
@@ -52,7 +52,7 @@ accelerate training workloads:
|
||||
- {{ component_version }}
|
||||
{% endfor %}
|
||||
|
||||
.. _amd-megatron-lm-model-support-v26.01:
|
||||
.. _amd-megatron-lm-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
@@ -97,7 +97,7 @@ accelerate training workloads:
|
||||
Some models, such as Llama, require an external license agreement through
|
||||
a third party (for example, Meta).
|
||||
|
||||
.. _amd-megatron-lm-performance-measurements-v26.01:
|
||||
.. _amd-megatron-lm-performance-measurements-v25.11:
|
||||
|
||||
Performance measurements
|
||||
========================
|
||||
@@ -129,7 +129,7 @@ To test for optimal performance, consult the recommended :ref:`System health ben
|
||||
<rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
|
||||
system's configuration.
|
||||
|
||||
.. _mi300x-amd-megatron-lm-training-v26.01:
|
||||
.. _mi300x-amd-megatron-lm-training-v25.11:
|
||||
|
||||
Environment setup
|
||||
=================
|
||||
@@ -138,7 +138,7 @@ Use the following instructions to set up the environment, configure the script t
|
||||
reproduce the benchmark results on MI300X Series GPUs with the AMD Megatron-LM Docker
|
||||
image.
|
||||
|
||||
.. _amd-megatron-lm-requirements-v26.01:
|
||||
.. _amd-megatron-lm-requirements-v25.11:
|
||||
|
||||
Download the Docker image
|
||||
-------------------------
|
||||
@@ -190,7 +190,7 @@ Download the Docker image
|
||||
The Docker container hosts a verified commit of
|
||||
`<https://github.com/ROCm/Megatron-LM/tree/rocm_dev>`__.
|
||||
|
||||
.. _amd-megatron-lm-environment-setup-v26.01:
|
||||
.. _amd-megatron-lm-environment-setup-v25.11:
|
||||
|
||||
Configuration
|
||||
=============
|
||||
@@ -200,39 +200,39 @@ Configuration
|
||||
Update the ``train_llama3.sh`` configuration script in the ``examples/llama``
|
||||
directory of
|
||||
`<https://github.com/ROCm/Megatron-LM/tree/rocm_dev/examples/llama>`__ to configure your training run.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v26.01>`.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v25.11>`.
|
||||
|
||||
.. container:: model-doc pyt_megatron_lm_train_llama-2-7b pyt_megatron_lm_train_llama-2-70b
|
||||
|
||||
Update the ``train_llama2.sh`` configuration script in the ``examples/llama``
|
||||
directory of
|
||||
`<https://github.com/ROCm/Megatron-LM/tree/rocm_dev/examples/llama>`__ to configure your training run.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v26.01>`.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v25.11>`.
|
||||
|
||||
.. container:: model-doc pyt_megatron_lm_train_deepseek-v3-proxy
|
||||
|
||||
Update the ``train_deepseekv3.sh`` configuration script in the ``examples/deepseek_v3``
|
||||
directory of
|
||||
`<https://github.com/ROCm/Megatron-LM/tree/rocm_dev/examples/deepseek_v3>`__ to configure your training run.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v26.01>`.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v25.11>`.
|
||||
|
||||
.. container:: model-doc pyt_megatron_lm_train_deepseek-v2-lite-16b
|
||||
|
||||
Update the ``train_deepseekv2.sh`` configuration script in the ``examples/deepseek_v2``
|
||||
directory of
|
||||
`<https://github.com/ROCm/Megatron-LM/tree/rocm_dev/examples/deepseek_v2>`__ to configure your training run.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v26.01>`.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v25.11>`.
|
||||
|
||||
.. container:: model-doc pyt_megatron_lm_train_mixtral-8x7b pyt_megatron_lm_train_mixtral-8x22b-proxy
|
||||
|
||||
Update the ``train_mixtral_moe.sh`` configuration script in the ``examples/mixtral``
|
||||
directory of
|
||||
`<https://github.com/ROCm/Megatron-LM/tree/rocm_dev/examples/mixtral>`__ to configure your training run.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v26.01>`.
|
||||
Options can also be passed as command line arguments as described in :ref:`Run training <amd-megatron-lm-run-training-v25.11>`.
|
||||
|
||||
.. note::
|
||||
|
||||
See :ref:`Key options <amd-megatron-lm-benchmark-test-vars-v26.01>` for more information on configuration options.
|
||||
See :ref:`Key options <amd-megatron-lm-benchmark-test-vars-v25.11>` for more information on configuration options.
|
||||
|
||||
Multi-node configuration
|
||||
------------------------
|
||||
@@ -240,7 +240,7 @@ Multi-node configuration
|
||||
Refer to :doc:`/how-to/rocm-for-ai/system-setup/multi-node-setup` to configure your environment for multi-node
|
||||
training. See :ref:`amd-megatron-lm-multi-node-examples` for example run commands.
|
||||
|
||||
.. _amd-megatron-lm-tokenizer-v26.01:
|
||||
.. _amd-megatron-lm-tokenizer-v25.11:
|
||||
|
||||
Tokenizer
|
||||
---------
|
||||
@@ -377,7 +377,7 @@ Download the dataset
|
||||
|
||||
``TOKENIZER_MODEL`` can be any accessible Hugging Face tokenizer.
|
||||
Remember to either pre-download the tokenizer or setup Hugging Face access
|
||||
otherwise when needed -- see the :ref:`Tokenizer <amd-megatron-lm-tokenizer-v26.01>` section.
|
||||
otherwise when needed -- see the :ref:`Tokenizer <amd-megatron-lm-tokenizer-v25.11>` section.
|
||||
|
||||
.. note::
|
||||
|
||||
@@ -479,13 +479,13 @@ Download the dataset
|
||||
|
||||
Ensure that the files are accessible inside the Docker container.
|
||||
|
||||
.. _amd-megatron-lm-run-training-v26.01:
|
||||
.. _amd-megatron-lm-run-training-v25.11:
|
||||
|
||||
Run training
|
||||
============
|
||||
|
||||
Use the following example commands to set up the environment, configure
|
||||
:ref:`key options <amd-megatron-lm-benchmark-test-vars-v26.01>`, and run training on
|
||||
:ref:`key options <amd-megatron-lm-benchmark-test-vars-v25.11>`, and run training on
|
||||
MI300X Series GPUs with the AMD Megatron-LM environment.
|
||||
|
||||
Before starting training, export the following environment variables.
|
||||
@@ -920,7 +920,7 @@ Single node training
|
||||
RECOMPUTE_ACTIVATIONS=full \
|
||||
CKPT_FORMAT=torch_dist
|
||||
|
||||
.. _amd-megatron-lm-multi-node-examples-v26.01:
|
||||
.. _amd-megatron-lm-multi-node-examples-v25.11:
|
||||
|
||||
Multi-node training examples
|
||||
----------------------------
|
||||
@@ -971,7 +971,7 @@ training on 16 nodes, try the following command:
|
||||
|
||||
sbatch examples/deepseek_v3/train_deepseek_v3_slurm.sh
|
||||
|
||||
.. _amd-megatron-lm-benchmark-test-vars-v26.01:
|
||||
.. _amd-megatron-lm-benchmark-test-vars-v25.11:
|
||||
|
||||
Key options
|
||||
-----------
|
||||
|
||||
@@ -22,7 +22,7 @@ previous releases of the ``ROCm/jax-training`` Docker image on `Docker Hub <http
|
||||
* ROCm 7.1.0
|
||||
* JAX 0.7.1
|
||||
-
|
||||
* :doc:`Documentation <jax-maxtext-v25.11>`
|
||||
* :doc:`Documentation <../jax-maxtext>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/jax-training/maxtext-v25.11/images/sha256-18e4d8f0b8ce7a7422c58046940dd5f32249960449fca09a562b65fb8eb1562a>`__
|
||||
|
||||
* - 25.9.1
|
||||
|
||||
@@ -1,408 +0,0 @@
|
||||
:orphan:
|
||||
|
||||
.. meta::
|
||||
:description: How to train a model using JAX MaxText for ROCm.
|
||||
:keywords: ROCm, AI, LLM, train, jax, torch, Llama, flux, tutorial, docker
|
||||
|
||||
******************************************
|
||||
Training a model with JAX MaxText on ROCm
|
||||
******************************************
|
||||
|
||||
.. caution::
|
||||
|
||||
This documentation does not reflect the latest version of ROCm JAX MaxText
|
||||
training performance documentation. See :doc:`../jax-maxtext` for the latest version.
|
||||
|
||||
The MaxText for ROCm training Docker image
|
||||
provides a prebuilt environment for training on AMD Instinct MI355X, MI350X, MI325X, and MI300X GPUs,
|
||||
including essential components like JAX, XLA, ROCm libraries, and MaxText utilities.
|
||||
It includes the following software components:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/jax-maxtext-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set dockers = data.dockers %}
|
||||
.. tab-set::
|
||||
|
||||
{% for docker in dockers %}
|
||||
{% set jax_version = docker.components["JAX"] %}
|
||||
|
||||
.. tab-item:: ``{{ docker.pull_tag }}``
|
||||
:sync: {{ docker.pull_tag }}
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Software component
|
||||
- Version
|
||||
|
||||
{% for component_name, component_version in docker.components.items() %}
|
||||
* - {{ component_name }}
|
||||
- {{ component_version }}
|
||||
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. note::
|
||||
|
||||
The ``rocm/jax-training:maxtext-v25.9`` has been updated to
|
||||
``rocm/jax-training:maxtext-v25.9.1``. This revision should include
|
||||
a fix to address segmentation fault issues during launch. See the
|
||||
:doc:`versioned documentation <jax-maxtext-v25.9>`.
|
||||
|
||||
MaxText with on ROCm provides the following key features to train large language models efficiently:
|
||||
|
||||
- Transformer Engine (TE)
|
||||
|
||||
- Flash Attention (FA) 3 -- with or without sequence input packing
|
||||
|
||||
- GEMM tuning
|
||||
|
||||
- Multi-node support
|
||||
|
||||
- NANOO FP8 (for MI300X series GPUs) and FP8 (for MI355X and MI350X) quantization support
|
||||
|
||||
.. _amd-maxtext-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
|
||||
The following models are pre-optimized for performance on AMD Instinct
|
||||
GPUs. Some instructions, commands, and available training
|
||||
configurations in this documentation might vary by model -- select one to get
|
||||
started.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/jax-maxtext-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. raw:: html
|
||||
|
||||
<div id="vllm-benchmark-ud-params-picker" class="container-fluid">
|
||||
<div class="row gx-0">
|
||||
<div class="col-2 me-1 px-2 model-param-head">Model</div>
|
||||
<div class="row col-10 pe-0">
|
||||
{% for model_group in model_groups %}
|
||||
<div class="col-4 px-2 model-param" data-param-k="model-group" data-param-v="{{ model_group.tag }}" tabindex="0">{{ model_group.group }}</div>
|
||||
{% endfor %}
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div class="row gx-0 pt-1">
|
||||
<div class="col-2 me-1 px-2 model-param-head">Variant</div>
|
||||
<div class="row col-10 pe-0">
|
||||
{% for model_group in model_groups %}
|
||||
{% set models = model_group.models %}
|
||||
{% for model in models %}
|
||||
{% if models|length % 3 == 0 %}
|
||||
<div class="col-4 px-2 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
|
||||
{% else %}
|
||||
<div class="col-6 px-2 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
.. note::
|
||||
|
||||
Some models, such as Llama 3, require an external license agreement through
|
||||
a third party (for example, Meta).
|
||||
|
||||
System validation
|
||||
=================
|
||||
|
||||
Before running AI workloads, it's important to validate that your AMD hardware is configured
|
||||
correctly and performing optimally.
|
||||
|
||||
If you have already validated your system settings, including aspects like NUMA auto-balancing, you
|
||||
can skip this step. Otherwise, complete the procedures in the :ref:`System validation and
|
||||
optimization <rocm-for-ai-system-optimization>` guide to properly configure your system settings
|
||||
before starting training.
|
||||
|
||||
To test for optimal performance, consult the recommended :ref:`System health benchmarks
|
||||
<rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
|
||||
system's configuration.
|
||||
|
||||
Environment setup
|
||||
=================
|
||||
|
||||
This Docker image is optimized for specific model configurations outlined
|
||||
as follows. Performance can vary for other training workloads, as AMD
|
||||
doesn’t validate configurations and run conditions outside those described.
|
||||
|
||||
Pull the Docker image
|
||||
---------------------
|
||||
|
||||
Use the following command to pull the Docker image from Docker Hub.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/jax-maxtext-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.dockers[0] %}
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker pull {{ docker.pull_tag }}
|
||||
|
||||
.. _amd-maxtext-multi-node-setup-v25.11:
|
||||
|
||||
Multi-node configuration
|
||||
------------------------
|
||||
|
||||
See :doc:`/how-to/rocm-for-ai/system-setup/multi-node-setup` to configure your
|
||||
environment for multi-node training.
|
||||
|
||||
.. _amd-maxtext-get-started-v25.11:
|
||||
|
||||
Benchmarking
|
||||
============
|
||||
|
||||
Once the setup is complete, choose between two options to reproduce the
|
||||
benchmark results:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/jax-maxtext-v25.11-benchmark-models.yaml
|
||||
|
||||
.. _vllm-benchmark-mad:
|
||||
|
||||
{% set docker = data.dockers[0] %}
|
||||
{% set model_groups = data.model_groups %}
|
||||
{% for model_group in model_groups %}
|
||||
{% for model in model_group.models %}
|
||||
|
||||
.. container:: model-doc {{model.mad_tag}}
|
||||
|
||||
.. tab-set::
|
||||
|
||||
{% if model.mad_tag and "single-node" in model.doc_options %}
|
||||
.. tab-item:: MAD-integrated benchmarking
|
||||
|
||||
The following run command is tailored to {{ model.model }}.
|
||||
See :ref:`amd-maxtext-model-support-v25.11` to switch to another available model.
|
||||
|
||||
1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
|
||||
directory and install the required packages on the host machine.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
git clone https://github.com/ROCm/MAD
|
||||
cd MAD
|
||||
pip install -r requirements.txt
|
||||
|
||||
2. Use this command to run the performance benchmark test on the {{ model.model }} model
|
||||
using one GPU with the :literal:`{{model.precision}}` data type on the host machine.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
export MAD_SECRETS_HFTOKEN="your personal Hugging Face token to access gated models"
|
||||
madengine run \
|
||||
--tags {{model.mad_tag}} \
|
||||
--keep-model-dir \
|
||||
--live-output \
|
||||
--timeout 28800
|
||||
|
||||
MAD launches a Docker container with the name
|
||||
``container_ci-{{model.mad_tag}}``. The latency and throughput reports of the
|
||||
model are collected in the following path: ``~/MAD/perf.csv/``.
|
||||
{% endif %}
|
||||
|
||||
.. tab-item:: Standalone benchmarking
|
||||
|
||||
The following commands are optimized for {{ model.model }}. See
|
||||
:ref:`amd-maxtext-model-support-v25.11` to switch to another
|
||||
available model. Some instructions and resources might not be
|
||||
available for all models and configurations.
|
||||
|
||||
.. rubric:: Download the Docker image and required scripts
|
||||
|
||||
Run the JAX MaxText benchmark tool independently by starting the
|
||||
Docker container as shown in the following snippet.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker pull {{ docker.pull_tag }}
|
||||
|
||||
{% if model.model_repo and "single-node" in model.doc_options %}
|
||||
.. rubric:: Single node training
|
||||
|
||||
1. Set up environment variables.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
export MAD_SECRETS_HFTOKEN=<Your Hugging Face token>
|
||||
export HF_HOME=<Location of saved/cached Hugging Face models>
|
||||
|
||||
``MAD_SECRETS_HFTOKEN`` is your Hugging Face access token to access models, tokenizers, and data.
|
||||
See `User access tokens <https://huggingface.co/docs/hub/en/security-tokens>`__.
|
||||
|
||||
``HF_HOME`` is where ``huggingface_hub`` will store local data. See `huggingface_hub CLI <https://huggingface.co/docs/huggingface_hub/main/en/guides/cli#huggingface-cli-download>`__.
|
||||
If you already have downloaded or cached Hugging Face artifacts, set this variable to that path.
|
||||
Downloaded files typically get cached to ``~/.cache/huggingface``.
|
||||
|
||||
2. Launch the Docker container.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker run -it \
|
||||
--device=/dev/dri \
|
||||
--device=/dev/kfd \
|
||||
--network host \
|
||||
--ipc host \
|
||||
--group-add video \
|
||||
--cap-add=SYS_PTRACE \
|
||||
--security-opt seccomp=unconfined \
|
||||
--privileged \
|
||||
-v $HOME:$HOME \
|
||||
-v $HOME/.ssh:/root/.ssh \
|
||||
-v $HF_HOME:/hf_cache \
|
||||
-e HF_HOME=/hf_cache \
|
||||
-e MAD_SECRETS_HFTOKEN=$MAD_SECRETS_HFTOKEN
|
||||
--shm-size 64G \
|
||||
--name training_env \
|
||||
{{ docker.pull_tag }}
|
||||
|
||||
3. In the Docker container, clone the ROCm MAD repository and navigate to the
|
||||
benchmark scripts directory at ``MAD/scripts/jax-maxtext``.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
git clone https://github.com/ROCm/MAD
|
||||
cd MAD/scripts/jax-maxtext
|
||||
|
||||
4. Run the setup scripts to install libraries and datasets needed
|
||||
for benchmarking.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./jax-maxtext_benchmark_setup.sh -m {{ model.model_repo }}
|
||||
|
||||
5. To run the training benchmark without quantization, use the following command:
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }}
|
||||
|
||||
For quantized training, run the script with the appropriate option for your Instinct GPU.
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MI355X and MI350X
|
||||
|
||||
For ``fp8`` quantized training on MI355X and MI350X GPUs, use the following command:
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }} -q fp8
|
||||
|
||||
{% if model.model_repo not in ["Llama-3.1-70B", "Llama-3.3-70B"] %}
|
||||
.. tab-item:: MI325X and MI300X
|
||||
|
||||
For ``nanoo_fp8`` quantized training on MI300X series GPUs, use the following command:
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }} -q nanoo_fp8
|
||||
{% endif %}
|
||||
|
||||
{% endif %}
|
||||
{% if model.multinode_training_script and "multi-node" in model.doc_options %}
|
||||
.. rubric:: Multi-node training
|
||||
|
||||
The following examples use SLURM to run on multiple nodes.
|
||||
|
||||
.. note::
|
||||
|
||||
The following scripts will launch the Docker container and run the
|
||||
benchmark. Run them outside of any Docker container.
|
||||
|
||||
1. Make sure ``$HF_HOME`` is set before running the test. See
|
||||
`ROCm benchmarking <https://github.com/ROCm/MAD/blob/develop/scripts/jax-maxtext/gpu-rocm/readme.md>`__
|
||||
for more details on downloading the Llama models before running the
|
||||
benchmark.
|
||||
|
||||
2. To run multi-node training for {{ model.model }},
|
||||
use the
|
||||
`multi-node training script <https://github.com/ROCm/MAD/blob/develop/scripts/jax-maxtext/gpu-rocm/{{ model.multinode_training_script }}>`__
|
||||
under the ``scripts/jax-maxtext/gpu-rocm/`` directory.
|
||||
|
||||
3. Run the multi-node training benchmark script.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
sbatch -N <num_nodes> {{ model.multinode_training_script }}
|
||||
|
||||
.. rubric:: Profiling with rocprofv3
|
||||
|
||||
If you need to collect a trace and the JAX profiler isn't working, use ``rocprofv3`` provided by the :doc:`ROCprofiler-SDK <rocprofiler-sdk:index>` as a workaround. For example:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
rocprofv3 \
|
||||
--hip-trace \
|
||||
--kernel-trace \
|
||||
--memory-copy-trace \
|
||||
--rccl-trace \
|
||||
--output-format pftrace \
|
||||
-d ./v3_traces \ # output directory
|
||||
-- ./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }} # or desired command
|
||||
|
||||
You can set the directory where you want the .json traces to be
|
||||
saved using ``-d <TRACE_DIRECTORY>``. The resulting traces can be
|
||||
opened in Perfetto: `<https://ui.perfetto.dev/>`__.
|
||||
|
||||
{% else %}
|
||||
.. rubric:: Multi-node training
|
||||
|
||||
For multi-node training examples, choose a model from :ref:`amd-maxtext-model-support-v25.11`
|
||||
with an available `multi-node training script <https://github.com/ROCm/MAD/tree/develop/scripts/jax-maxtext/gpu-rocm>`__.
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
Known issues
|
||||
============
|
||||
|
||||
- Minor performance regression (< 4%) for BF16 quantization in Llama models and Mixtral 8x7b.
|
||||
|
||||
- You might see minor loss spikes, or loss curve may have slightly higher
|
||||
convergence end values compared to the previous ``jax-training`` image.
|
||||
|
||||
- For FP8 training on MI355, many models will display a warning message like:
|
||||
``Warning: Latency not found for MI_M=16, MI_N=16, MI_K=128,
|
||||
mi_input_type=BFloat8Float8_fnuz. Returning latency value of 32 (really
|
||||
slow).`` The compile step may take longer than usual, but training will run.
|
||||
This will be fixed in a future release.
|
||||
|
||||
- The built-in JAX profiler isn't working. If you need to collect a trace and
|
||||
the JAX profiler isn't working, use ``rocprofv3`` provided by the
|
||||
:doc:`ROCprofiler-SDK <rocprofiler-sdk:index>` as a workaround. For example:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
rocprofv3 \
|
||||
--hip-trace \
|
||||
--kernel-trace \
|
||||
--memory-copy-trace \
|
||||
--rccl-trace \
|
||||
--output-format pftrace \
|
||||
-d ./v3_traces \ # output directory
|
||||
-- ./jax-maxtext_benchmark_report.sh -m {{ model.model_repo }} # or desired command
|
||||
|
||||
You can set the directory where you want the .json traces to be
|
||||
saved using ``-d <TRACE_DIRECTORY>``. The resulting traces can be
|
||||
opened in Perfetto: `<https://ui.perfetto.dev/>`__.
|
||||
|
||||
Further reading
|
||||
===============
|
||||
|
||||
- To learn more about MAD and the ``madengine`` CLI, see the `MAD usage guide <https://github.com/ROCm/MAD?tab=readme-ov-file#usage-guide>`__.
|
||||
|
||||
- To learn more about system settings and management practices to configure your system for
|
||||
AMD Instinct MI300X Series GPUs, see `AMD Instinct MI300X system optimization <https://instinct.docs.amd.com/projects/amdgpu-docs/en/latest/system-optimization/mi300x.html>`_.
|
||||
|
||||
- For a list of other ready-made Docker images for AI with ROCm, see
|
||||
`AMD Infinity Hub <https://www.amd.com/en/developer/resources/infinity-hub.html#f-amd_hub_category=AI%20%26%20ML%20Models>`_.
|
||||
|
||||
Previous versions
|
||||
=================
|
||||
|
||||
See :doc:`jax-maxtext-history` to find documentation for previous releases
|
||||
of the ``ROCm/jax-training`` Docker image.
|
||||
@@ -16,23 +16,14 @@ previous releases of the ``ROCm/megatron-lm`` Docker image on `Docker Hub <https
|
||||
- Components
|
||||
- Resources
|
||||
|
||||
* - v26.1 (latest)
|
||||
* - v25.11
|
||||
-
|
||||
* ROCm 7.1.0
|
||||
* PyTorch 2.10.0.dev20251112+rocm7.1
|
||||
-
|
||||
* :doc:`Primus Megatron documentation <../primus-megatron>`
|
||||
* :doc:`Megatron-LM (legacy) documentation <../megatron-lm>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v26.1/images/sha256-4fc8808bdb14117c6af7f38d79c809056e6fdbfd530c1fabbb61d097ddaf820d>`__
|
||||
|
||||
* - v25.11
|
||||
-
|
||||
* ROCm 7.1.0
|
||||
* PyTorch 2.10.0.dev20251112+rocm7.1
|
||||
-
|
||||
* :doc:`Primus Megatron documentation <primus-megatron-v25.11>`
|
||||
* :doc:`Megatron-LM (legacy) documentation <megatron-lm-v25.10>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v25.11/images/sha256-71aa65a9bfc8e9dd18bce5b68c81caff864f223e9afa75dc1b719671a1f4a3c3>`__
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197>`__
|
||||
|
||||
* - v25.10
|
||||
-
|
||||
|
||||
@@ -37,7 +37,7 @@ GPUs containing essential components, including PyTorch, ROCm libraries, and
|
||||
Megatron-LM utilities. It contains the following software components to
|
||||
accelerate training workloads:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/megatron-lm-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/megatron-lm-benchmark-models.yaml
|
||||
|
||||
.. tab-set::
|
||||
|
||||
@@ -146,7 +146,7 @@ image.
|
||||
Download the Docker image
|
||||
-------------------------
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/megatron-lm-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/megatron-lm-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
1. Use the following command to pull the Docker image from Docker Hub.
|
||||
@@ -811,7 +811,7 @@ Single node training
|
||||
Note that DeepSeek-V2-Lite is experiencing instability due to GPU memory access fault
|
||||
for large iterations.
|
||||
For stability, it's recommended to use Primus for this workload.
|
||||
See :doc:`../primus-megatron`.
|
||||
See :doc:`primus-megatron`.
|
||||
|
||||
.. container:: model-doc pyt_megatron_lm_train_mixtral-8x7b
|
||||
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -25,10 +25,10 @@ model training. Performance acceleration is powered by `Primus Turbo
|
||||
<https://hub.docker.com/r/rocm/megatron-lm/>`__ Docker Hub registry will be
|
||||
deprecated soon in favor of `rocm/primus <https://hub.docker.com/r/rocm/primus>`__.
|
||||
The ``rocm/primus`` Docker containers will cover PyTorch training ecosystem frameworks,
|
||||
including Megatron-LM and :doc:`torchtitan <../primus-pytorch>`.
|
||||
including Megatron-LM and :doc:`torchtitan <primus-pytorch>`.
|
||||
|
||||
Primus with Megatron is designed to replace the :doc:`ROCm Megatron-LM
|
||||
training <../megatron-lm>` workflow. To learn how to migrate workloads from
|
||||
training <megatron-lm>` workflow. To learn how to migrate workloads from
|
||||
Megatron-LM to Primus with Megatron, see
|
||||
:doc:`megatron-lm-primus-migration-guide`.
|
||||
|
||||
@@ -36,7 +36,7 @@ AMD provides a ready-to-use Docker images for MI355X, MI350X,
|
||||
MI325X, and MI300X GPUs containing essential components for Primus, ROCm, and
|
||||
Megatron-LM.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-megatron-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
.. tab-set::
|
||||
|
||||
@@ -63,7 +63,7 @@ The following models are pre-optimized for performance on AMD Instinct GPUs.
|
||||
Some instructions, commands, and training examples in this documentation
|
||||
might vary by model -- select one to get started.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-megatron-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. raw:: html
|
||||
@@ -120,7 +120,7 @@ system's configuration.
|
||||
Environment setup
|
||||
=================
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-megatron-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
Use the following instructions to set up the environment, configure the script to train models, and
|
||||
reproduce the benchmark results on AMD Instinct GPUs.
|
||||
@@ -129,7 +129,7 @@ Environment setup
|
||||
|
||||
Pull the Docker image
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-megatron-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
|
||||
@@ -175,7 +175,7 @@ Configuration
|
||||
Primus defines a training configuration in YAML for each model in
|
||||
`examples/megatron/configs <https://github.com/AMD-AGI/Primus/tree/e16b27bf6c1b2798f38848fc574fee60d9a9b902/examples/megatron/configs>`__.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-megatron-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
{% for model_group in model_groups %}
|
||||
@@ -805,7 +805,7 @@ To run training on multiple nodes, you can use the
|
||||
`run_slurm_pretrain.sh <https://github.com/AMD-AGI/Primus/blob/main/examples/run_slurm_pretrain.sh>`__
|
||||
to launch the multi-node workload. Use the following steps to setup your environment:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-megatron-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
.. code-block:: shell
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -24,17 +24,17 @@ Primus now supports the PyTorch torchtitan backend.
|
||||
<https://hub.docker.com/r/rocm/pytorch-training/>`__ Docker Hub registry will be
|
||||
deprecated soon in favor of `rocm/primus <https://hub.docker.com/r/rocm/primus>`__.
|
||||
The ``rocm/primus`` Docker containers will cover PyTorch training ecosystem frameworks,
|
||||
including torchtitan and :doc:`Megatron-LM <../primus-megatron>`.
|
||||
including torchtitan and :doc:`Megatron-LM <primus-megatron>`.
|
||||
|
||||
Primus with the PyTorch torchtitan backend is designed to replace the
|
||||
:doc:`ROCm PyTorch training <../pytorch-training>` workflow. See
|
||||
:doc:`../pytorch-training` to see steps to run workloads without Primus.
|
||||
:doc:`ROCm PyTorch training <pytorch-training>` workflow. See
|
||||
:doc:`pytorch-training` to see steps to run workloads without Primus.
|
||||
|
||||
AMD provides a ready-to-use Docker image for MI355X, MI350X, MI325X, and
|
||||
MI300X GPUs containing essential components for Primus and PyTorch training
|
||||
with Primus Turbo optimizations.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-pytorch-benchmark-models.yaml
|
||||
|
||||
.. tab-set::
|
||||
|
||||
@@ -61,7 +61,7 @@ The following models are pre-optimized for performance on the AMD Instinct MI325
|
||||
Some instructions, commands, and training recommendations in this documentation might
|
||||
vary by model -- select one to get started.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-pytorch-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. raw:: html
|
||||
@@ -96,7 +96,7 @@ vary by model -- select one to get started.
|
||||
.. seealso::
|
||||
|
||||
For additional workloads, including Llama 3.3, Llama 3.2, Llama 2, GPT OSS, Qwen, and Flux models,
|
||||
see the documentation :doc:`../pytorch-training` (without Primus)
|
||||
see the documentation :doc:`pytorch-training` (without Primus)
|
||||
|
||||
.. _amd-primus-pytorch-performance-measurements-v2510:
|
||||
|
||||
@@ -122,7 +122,7 @@ doesn’t test configurations and run conditions outside those described.
|
||||
Pull the Docker image
|
||||
=====================
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-pytorch-benchmark-models.yaml
|
||||
|
||||
Use the following command to pull the Docker image from Docker Hub.
|
||||
|
||||
@@ -134,11 +134,11 @@ Run training
|
||||
============
|
||||
|
||||
Once the setup is complete, choose between the following two workflows to start benchmarking training.
|
||||
For fine-tuning workloads and multi-node training examples, see :doc:`../pytorch-training` (without Primus).
|
||||
For fine-tuning workloads and multi-node training examples, see :doc:`pytorch-training` (without Primus).
|
||||
For best performance on MI325X, MI350X, and MI355X GPUs, you might need to
|
||||
tweak some configurations (such as batch sizes).
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-pytorch-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
{% set model_groups = data.model_groups %}
|
||||
|
||||
@@ -1,422 +0,0 @@
|
||||
:orphan:
|
||||
|
||||
.. meta::
|
||||
:description: How to train a model using PyTorch for ROCm.
|
||||
:keywords: ROCm, AI, LLM, train, PyTorch, torch, Llama, flux, tutorial, docker
|
||||
|
||||
****************************************
|
||||
Training a model with Primus and PyTorch
|
||||
****************************************
|
||||
|
||||
.. caution::
|
||||
|
||||
This documentation does not reflect the latest version of ROCm Primus PyTorch training
|
||||
performance benchmark documentation. See :doc:`../primus-pytorch` for the latest version.
|
||||
|
||||
`Primus <https://github.com/AMD-AGI/Primus>`__ is a unified and flexible
|
||||
LLM training framework designed to streamline training. It streamlines LLM
|
||||
training on AMD Instinct GPUs using a modular, reproducible configuration paradigm.
|
||||
Primus now supports the PyTorch torchtitan backend.
|
||||
|
||||
.. note::
|
||||
|
||||
For a unified training solution on AMD GPUs with ROCm, the `rocm/pytorch-training
|
||||
<https://hub.docker.com/r/rocm/pytorch-training/>`__ Docker Hub registry will be
|
||||
deprecated soon in favor of `rocm/primus <https://hub.docker.com/r/rocm/primus>`__.
|
||||
The ``rocm/primus`` Docker containers will cover PyTorch training ecosystem frameworks,
|
||||
including torchtitan and :doc:`Megatron-LM <../primus-megatron>`.
|
||||
|
||||
Primus with the PyTorch torchtitan backend is designed to replace the
|
||||
:doc:`ROCm PyTorch training <../pytorch-training>` workflow. See
|
||||
:doc:`../pytorch-training` to see steps to run workloads without Primus.
|
||||
|
||||
AMD provides a ready-to-use Docker image for MI355X, MI350X, MI325X, and
|
||||
MI300X GPUs containing essential components for Primus and PyTorch training
|
||||
with Primus Turbo optimizations.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.11-benchmark-models.yaml
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: {{ data.docker.pull_tag }}
|
||||
:sync: {{ data.docker.pull_tag }}
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Software component
|
||||
- Version
|
||||
|
||||
{% for component_name, component_version in data.docker.components.items() %}
|
||||
* - {{ component_name }}
|
||||
- {{ component_version }}
|
||||
{% endfor %}
|
||||
|
||||
.. _amd-primus-pytorch-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
|
||||
The following models are pre-optimized for performance on the AMD Instinct MI325X and MI300X GPUs.
|
||||
Some instructions, commands, and training recommendations in this documentation might
|
||||
vary by model -- select one to get started.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. raw:: html
|
||||
|
||||
<div id="vllm-benchmark-ud-params-picker" class="container-fluid">
|
||||
<div class="row gx-0">
|
||||
<div class="col-2 me-1 px-2 model-param-head">Model</div>
|
||||
<div class="row col-10 pe-0">
|
||||
{% for model_group in model_groups %}
|
||||
<div class="col-6 px-2 model-param" data-param-k="model-group" data-param-v="{{ model_group.tag }}" tabindex="0">{{ model_group.group }}</div>
|
||||
{% endfor %}
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div class="row gx-0 pt-1">
|
||||
<div class="col-2 me-1 px-2 model-param-head">Variant</div>
|
||||
<div class="row col-10 pe-0">
|
||||
{% for model_group in model_groups %}
|
||||
{% set models = model_group.models %}
|
||||
{% for model in models %}
|
||||
{% if models|length % 3 == 0 %}
|
||||
<div class="col-4 px-2 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
|
||||
{% else %}
|
||||
<div class="col-6 px-2 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
.. seealso::
|
||||
|
||||
For additional workloads, including Llama 3.3, Llama 3.2, Llama 2, GPT OSS, Qwen, and Flux models,
|
||||
see the documentation :doc:`../pytorch-training` (without Primus)
|
||||
|
||||
.. _amd-primus-pytorch-performance-measurements-v25.11:
|
||||
|
||||
System validation
|
||||
=================
|
||||
|
||||
Before running AI workloads, it's important to validate that your AMD hardware is configured
|
||||
correctly and performing optimally.
|
||||
|
||||
If you have already validated your system settings, including aspects like NUMA auto-balancing, you
|
||||
can skip this step. Otherwise, complete the procedures in the :ref:`System validation and
|
||||
optimization <rocm-for-ai-system-optimization>` guide to properly configure your system settings
|
||||
before starting training.
|
||||
|
||||
To test for optimal performance, consult the recommended :ref:`System health benchmarks
|
||||
<rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
|
||||
system's configuration.
|
||||
|
||||
This Docker image is optimized for specific model configurations outlined
|
||||
below. Performance can vary for other training workloads, as AMD
|
||||
doesn’t test configurations and run conditions outside those described.
|
||||
|
||||
Pull the Docker image
|
||||
=====================
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.11-benchmark-models.yaml
|
||||
|
||||
Use the following command to pull the Docker image from Docker Hub.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker pull {{ data.docker.pull_tag }}
|
||||
|
||||
Run training
|
||||
============
|
||||
|
||||
Once the setup is complete, choose between the following two workflows to start benchmarking training.
|
||||
For fine-tuning workloads and multi-node training examples, see :doc:`../pytorch-training` (without Primus).
|
||||
For best performance on MI325X, MI350X, and MI355X GPUs, you might need to
|
||||
tweak some configurations (such as batch sizes).
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/primus-pytorch-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
{% set model_groups = data.model_groups %}
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MAD-integrated benchmarking
|
||||
|
||||
{% for model_group in model_groups %}
|
||||
{% for model in model_group.models %}
|
||||
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following run command is tailored to {{ model.model }}.
|
||||
See :ref:`amd-primus-pytorch-model-support-v25.11` to switch to another available model.
|
||||
|
||||
1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
|
||||
directory and install the required packages on the host machine.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
git clone https://github.com/ROCm/MAD
|
||||
cd MAD
|
||||
pip install -r requirements.txt
|
||||
|
||||
2. For example, use this command to run the performance benchmark test on the {{ model.model }} model
|
||||
using one node with the {{ model.precision }} data type on the host machine.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
export MAD_SECRETS_HFTOKEN="your personal Hugging Face token to access gated models"
|
||||
madengine run \
|
||||
--tags {{ model.mad_tag }} \
|
||||
--keep-model-dir \
|
||||
--live-output \
|
||||
--timeout 28800
|
||||
|
||||
MAD launches a Docker container with the name
|
||||
``container_ci-{{ model.mad_tag }}``. The latency and throughput reports of the
|
||||
model are collected in ``~/MAD/perf.csv``.
|
||||
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. tab-item:: Primus benchmarking
|
||||
|
||||
{% for model_group in model_groups %}
|
||||
{% for model in model_group.models %}
|
||||
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following run commands are tailored to {{ model.model }}.
|
||||
See :ref:`amd-primus-pytorch-model-support-v25.11` to switch to another available model.
|
||||
|
||||
.. rubric:: Download the Docker image and required packages
|
||||
|
||||
1. Pull the ``{{ docker.pull_tag }}`` Docker image from Docker Hub.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker pull {{ docker.pull_tag }}
|
||||
|
||||
2. Run the Docker container.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker run -it \
|
||||
--device /dev/dri \
|
||||
--device /dev/kfd \
|
||||
--network host \
|
||||
--ipc host \
|
||||
--group-add video \
|
||||
--cap-add SYS_PTRACE \
|
||||
--security-opt seccomp=unconfined \
|
||||
--privileged \
|
||||
-v $HOME:$HOME \
|
||||
-v $HOME/.ssh:/root/.ssh \
|
||||
--shm-size 64G \
|
||||
--name training_env \
|
||||
{{ docker.pull_tag }}
|
||||
|
||||
Use these commands if you exit the ``training_env`` container and need to return to it.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker start training_env
|
||||
docker exec -it training_env bash
|
||||
|
||||
The Docker container hosts verified commit ``c4c083de`` of the `Primus
|
||||
<https://github.com/AMD-AGI/Primus/tree/c4c083de64ba3e8f19ccc9629411267108931f9e/>`__ repository.
|
||||
|
||||
.. rubric:: Prepare training datasets and dependencies
|
||||
|
||||
The following benchmarking examples require downloading models and datasets
|
||||
from Hugging Face. To ensure successful access to gated repos, set your
|
||||
``HF_TOKEN``.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
export HF_TOKEN=$your_personal_hugging_face_access_token
|
||||
|
||||
.. rubric:: Pretraining
|
||||
|
||||
To get started, navigate to the ``Primus`` directory in your container.
|
||||
|
||||
.. code-block::
|
||||
|
||||
cd /workspace/Primus
|
||||
|
||||
Now, to start the pretraining benchmark, use the ``run_pretrain.sh`` script
|
||||
included with Primus with the appropriate options.
|
||||
|
||||
.. rubric:: Benchmarking examples
|
||||
|
||||
.. container:: model-doc primus_pyt_train_llama-3.1-8b
|
||||
|
||||
Use the following command to run train Llama 3.1 8B with BF16 precision using Primus torchtitan.
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MI355X and MI350X
|
||||
:sync: MI355X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 6
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
To train Llama 3.1 8B with FP8 precision, use the following command.
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MI355X and MI350X
|
||||
:sync: MI355X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 7
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_train_llama-3.1-70b
|
||||
|
||||
Use the following command to run train Llama 3.1 70B with BF16 precision using Primus torchtitan.
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MI355X and MI350X
|
||||
:sync: MI355X and MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 6
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
To train Llama 3.1 70B with FP8 precision, use the following command.
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MI355X and MI350X
|
||||
:sync: MI355X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 5
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_train_deepseek-v3-16b
|
||||
|
||||
Use the following command to run train DeepSeek V3 16B with BF16 precision using Primus torchtitan.
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MI355X and MI350X
|
||||
:sync: MI355X and MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI355X/deepseek_v3_16b-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/deepseek_v3_16b-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 10
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
EXP=examples/torchtitan/configs/MI300X/deepseek_v3_16b-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
Further reading
|
||||
===============
|
||||
|
||||
- For an introduction to Primus, see `Primus: A Lightweight, Unified Training
|
||||
Framework for Large Models on AMD GPUs <https://rocm.blogs.amd.com/software-tools-optimization/primus/README.html>`__.
|
||||
|
||||
- To learn more about MAD and the ``madengine`` CLI, see the `MAD usage guide <https://github.com/ROCm/MAD?tab=readme-ov-file#usage-guide>`__.
|
||||
|
||||
- To learn more about system settings and management practices to configure your system for
|
||||
AMD Instinct MI300X Series GPUs, see `AMD Instinct MI300X system optimization <https://instinct.docs.amd.com/projects/amdgpu-docs/en/latest/system-optimization/mi300x.html>`_.
|
||||
|
||||
- For a list of other ready-made Docker images for AI with ROCm, see
|
||||
`AMD Infinity Hub <https://www.amd.com/en/developer/resources/infinity-hub.html#f-amd_hub_category=AI%20%26%20ML%20Models>`_.
|
||||
|
||||
Previous versions
|
||||
=================
|
||||
|
||||
See :doc:`pytorch-training-history` to find documentation for previous releases
|
||||
of the ``ROCm/pytorch-training`` Docker image.
|
||||
@@ -16,30 +16,21 @@ previous releases of the ``ROCm/pytorch-training`` Docker image on `Docker Hub <
|
||||
- Components
|
||||
- Resources
|
||||
|
||||
* - v26.1 (latest)
|
||||
-
|
||||
* ROCm 7.1.0
|
||||
* PyTorch 2.10.0.dev20251112+rocm7.1
|
||||
-
|
||||
* :doc:`Primus PyTorch training documentation <../primus-megatron>`
|
||||
* :doc:`PyTorch training (legacy) documentation <../megatron-lm>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v26.1/images/sha256-4fc8808bdb14117c6af7f38d79c809056e6fdbfd530c1fabbb61d097ddaf820d>`__
|
||||
|
||||
* - v25.11
|
||||
-
|
||||
* ROCm 7.1.0
|
||||
* PyTorch 2.10.0.dev20251112+rocm7.1
|
||||
-
|
||||
* :doc:`Primus PyTorch training documentation <primus-pytorch-v25.11>`
|
||||
* :doc:`PyTorch training (legacy) documentation <pytorch-training-v25.11>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v25.11/images/sha256-71aa65a9bfc8e9dd18bce5b68c81caff864f223e9afa75dc1b719671a1f4a3c3>`__
|
||||
* :doc:`Primus PyTorch Training documentation <../primus-pytorch>`
|
||||
* :doc:`PyTorch training (legacy) documentation <../pytorch-training>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197>`__
|
||||
|
||||
* - v25.10
|
||||
-
|
||||
* ROCm 7.1.0
|
||||
* PyTorch 2.10.0.dev20251112+rocm7.1
|
||||
-
|
||||
* :doc:`Primus PyTorch training documentation <primus-pytorch-v25.10>`
|
||||
* :doc:`Primus PyTorch Training documentation <primus-pytorch-v25.10>`
|
||||
* :doc:`PyTorch training (legacy) documentation <pytorch-training-v25.10>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/primus/v25.10/images/sha256-140c37cd2eeeb183759b9622543fc03cc210dc97cbfa18eeefdcbda84420c197>`__
|
||||
|
||||
@@ -49,7 +40,7 @@ previous releases of the ``ROCm/pytorch-training`` Docker image on `Docker Hub <
|
||||
* Primus 0.3.0
|
||||
* PyTorch 2.9.0.dev20250821+rocm7.0.0.lw.git125803b7
|
||||
-
|
||||
* :doc:`Primus PyTorch training documentation <primus-pytorch-v25.9>`
|
||||
* :doc:`Primus PyTorch Training documentation <primus-pytorch-v25.9>`
|
||||
* :doc:`PyTorch training (legacy) documentation <pytorch-training-v25.9>`
|
||||
* `Docker Hub (gfx950) <https://hub.docker.com/layers/rocm/primus/v25.9_gfx950/images/sha256-1a198be32f49efd66d0ff82066b44bd99b3e6b04c8e0e9b36b2c481e13bff7b6>`__
|
||||
* `Docker Hub (gfx942) <https://hub.docker.com/layers/rocm/primus/v25.9_gfx942/images/sha256-df6ab8f45b4b9ceb100fb24e19b2019a364e351ee3b324dbe54466a1d67f8357>`__
|
||||
@@ -59,7 +50,7 @@ previous releases of the ``ROCm/pytorch-training`` Docker image on `Docker Hub <
|
||||
* ROCm 6.4.3
|
||||
* PyTorch 2.8.0a0+gitd06a406
|
||||
-
|
||||
* :doc:`Primus PyTorch training documentation <primus-pytorch-v25.8>`
|
||||
* :doc:`Primus PyTorch Training documentation <primus-pytorch-v25.8>`
|
||||
* :doc:`PyTorch training (legacy) documentation <pytorch-training-v25.8>`
|
||||
* `Docker Hub <https://hub.docker.com/layers/rocm/pytorch-training/v25.8/images/sha256-5082ae01d73fec6972b0d84e5dad78c0926820dcf3c19f301d6c8eb892e573c5>`__
|
||||
|
||||
|
||||
@@ -30,7 +30,7 @@ environment for fine-tuning and pretraining a model on AMD Instinct MI325X
|
||||
and MI300X GPUs. It includes the following software components to accelerate
|
||||
training workloads:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml
|
||||
|
||||
.. tab-set::
|
||||
|
||||
@@ -58,7 +58,7 @@ MI355X, MI350X, MI325X, and MI300X GPUs. Some instructions, commands, and
|
||||
training recommendations in this documentation might vary by model -- select
|
||||
one to get started.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. raw:: html
|
||||
@@ -94,7 +94,7 @@ one to get started.
|
||||
|
||||
The following table lists supported training modes per model.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. dropdown:: Supported training modes
|
||||
@@ -164,7 +164,7 @@ doesn’t test configurations and run conditions outside those described.
|
||||
Run training
|
||||
============
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.10-benchmark-models.yaml
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/pytorch-training-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
{% set model_groups = data.model_groups %}
|
||||
|
||||
@@ -1,669 +0,0 @@
|
||||
:orphan:
|
||||
|
||||
.. meta::
|
||||
:description: How to train a model using PyTorch for ROCm.
|
||||
:keywords: ROCm, AI, LLM, train, PyTorch, torch, Llama, flux, tutorial, docker
|
||||
|
||||
**************************************
|
||||
Training a model with PyTorch on ROCm
|
||||
**************************************
|
||||
|
||||
.. caution::
|
||||
|
||||
This documentation does not reflect the latest version of ROCm PyTorch training
|
||||
performance benchmark documentation. See :doc:`../pytorch-training` for the latest version.
|
||||
|
||||
.. note::
|
||||
|
||||
For a unified training solution on AMD GPUs with ROCm, the `rocm/pytorch-training
|
||||
<https://hub.docker.com/r/rocm/pytorch-training/>`__ Docker Hub registry will be
|
||||
deprecated soon in favor of `rocm/primus <https://hub.docker.com/r/rocm/primus>`__.
|
||||
The ``rocm/primus`` Docker containers will cover PyTorch training ecosystem frameworks,
|
||||
including torchtitan and :doc:`Megatron-LM <../primus-megatron>`.
|
||||
|
||||
See :doc:`../primus-pytorch` for details.
|
||||
|
||||
PyTorch is an open-source machine learning framework that is widely used for
|
||||
model training with GPU-optimized components for transformer-based models.
|
||||
The PyTorch for ROCm training Docker image provides a prebuilt optimized
|
||||
environment for fine-tuning and pretraining a model on AMD Instinct MI325X
|
||||
and MI300X GPUs. It includes the following software components to accelerate
|
||||
training workloads:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.11-benchmark-models.yaml
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: {{ data.docker.pull_tag }}
|
||||
:sync: {{ data.docker.pull_tag }}
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Software component
|
||||
- Version
|
||||
|
||||
{% for component_name, component_version in data.docker.components.items() %}
|
||||
* - {{ component_name }}
|
||||
- {{ component_version }}
|
||||
{% endfor %}
|
||||
|
||||
.. _amd-pytorch-training-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
|
||||
The following models are pre-optimized for performance on the AMD Instinct
|
||||
MI355X, MI350X, MI325X, and MI300X GPUs. Some instructions, commands, and
|
||||
training recommendations in this documentation might vary by model -- select
|
||||
one to get started.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. raw:: html
|
||||
|
||||
<div id="vllm-benchmark-ud-params-picker" class="container-fluid">
|
||||
<div class="row gx-0">
|
||||
<div class="col-2 me-1 px-2 model-param-head">Model</div>
|
||||
<div class="row col-10 pe-0">
|
||||
{% for model_group in model_groups %}
|
||||
<div class="col-4 px-2 model-param" data-param-k="model-group" data-param-v="{{ model_group.tag }}" tabindex="0">{{ model_group.group }}</div>
|
||||
{% endfor %}
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div class="row gx-0 pt-1">
|
||||
<div class="col-2 me-1 px-2 model-param-head">Variant</div>
|
||||
<div class="row col-10 pe-0">
|
||||
{% for model_group in model_groups %}
|
||||
{% set models = model_group.models %}
|
||||
{% for model in models %}
|
||||
{% if models|length % 3 == 0 %}
|
||||
<div class="col-4 px-2 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
|
||||
{% else %}
|
||||
<div class="col-6 px-2 model-param" data-param-k="model" data-param-v="{{ model.mad_tag }}" data-param-group="{{ model_group.tag }}" tabindex="0">{{ model.model }}</div>
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
.. _amd-pytorch-training-supported-training-modes-v25.11:
|
||||
|
||||
The following table lists supported training modes per model.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set model_groups = data.model_groups %}
|
||||
.. dropdown:: Supported training modes
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Model
|
||||
- Supported training modes
|
||||
|
||||
{% for model_group in model_groups %}
|
||||
{% set models = model_group.models %}
|
||||
{% for model in models %}
|
||||
{% if model.training_modes %}
|
||||
* - {{ model.model }}
|
||||
- ``{{ model.training_modes | join('``, ``') }}``
|
||||
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. note::
|
||||
|
||||
Some model and fine-tuning combinations are not listed. This is
|
||||
because the `upstream torchtune repository <https://github.com/pytorch/torchtune>`__
|
||||
doesn't provide default YAML configurations for them.
|
||||
For advanced usage, you can create a custom configuration to enable
|
||||
unlisted fine-tuning methods by using an existing file in the
|
||||
``/workspace/torchtune/recipes/configs`` directory as a template.
|
||||
|
||||
.. _amd-pytorch-training-performance-measurements-v25.11:
|
||||
|
||||
Performance measurements
|
||||
========================
|
||||
|
||||
To evaluate performance, the
|
||||
`Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
|
||||
page provides reference throughput and latency measurements for training
|
||||
popular AI models.
|
||||
|
||||
.. note::
|
||||
|
||||
The performance data presented in
|
||||
`Performance results with AMD ROCm software <https://www.amd.com/en/developer/resources/rocm-hub/dev-ai/performance-results.html#tabs-a8deaeb413-item-21cea50186-tab>`_
|
||||
should not be interpreted as the peak performance achievable by AMD
|
||||
Instinct MI325X and MI300X GPUs or ROCm software.
|
||||
|
||||
System validation
|
||||
=================
|
||||
|
||||
Before running AI workloads, it's important to validate that your AMD hardware is configured
|
||||
correctly and performing optimally.
|
||||
|
||||
If you have already validated your system settings, including aspects like NUMA auto-balancing, you
|
||||
can skip this step. Otherwise, complete the procedures in the :ref:`System validation and
|
||||
optimization <rocm-for-ai-system-optimization>` guide to properly configure your system settings
|
||||
before starting training.
|
||||
|
||||
To test for optimal performance, consult the recommended :ref:`System health benchmarks
|
||||
<rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
|
||||
system's configuration.
|
||||
|
||||
This Docker image is optimized for specific model configurations outlined
|
||||
below. Performance can vary for other training workloads, as AMD
|
||||
doesn’t test configurations and run conditions outside those described.
|
||||
|
||||
Run training
|
||||
============
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/previous-versions/pytorch-training-v25.11-benchmark-models.yaml
|
||||
|
||||
{% set docker = data.docker %}
|
||||
{% set model_groups = data.model_groups %}
|
||||
|
||||
Once the setup is complete, choose between two options to start benchmarking training:
|
||||
|
||||
.. tab-set::
|
||||
|
||||
.. tab-item:: MAD-integrated benchmarking
|
||||
|
||||
{% for model_group in model_groups %}
|
||||
{% for model in model_group.models %}
|
||||
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following run command is tailored to {{ model.model }}.
|
||||
See :ref:`amd-pytorch-training-model-support-v25.11` to switch to another available model.
|
||||
|
||||
1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
|
||||
directory and install the required packages on the host machine.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
git clone https://github.com/ROCm/MAD
|
||||
cd MAD
|
||||
pip install -r requirements.txt
|
||||
|
||||
2. For example, use this command to run the performance benchmark test on the {{ model.model }} model
|
||||
using one node with the {{ model.precision }} data type on the host machine.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
export MAD_SECRETS_HFTOKEN="your personal Hugging Face token to access gated models"
|
||||
madengine run \
|
||||
--tags {{ model.mad_tag }} \
|
||||
--keep-model-dir \
|
||||
--live-output \
|
||||
--timeout 28800
|
||||
|
||||
MAD launches a Docker container with the name
|
||||
``container_ci-{{ model.mad_tag }}``. The latency and throughput reports of the
|
||||
model are collected in ``~/MAD/perf.csv``.
|
||||
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. tab-item:: Standalone benchmarking
|
||||
|
||||
{% for model_group in model_groups %}
|
||||
{% for model in model_group.models %}
|
||||
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following commands are tailored to {{ model.model }}.
|
||||
See :ref:`amd-pytorch-training-model-support-v25.11` to switch to another available model.
|
||||
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. rubric:: Download the Docker image and required packages
|
||||
|
||||
1. Use the following command to pull the Docker image from Docker Hub.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker pull {{ docker.pull_tag }}
|
||||
|
||||
2. Launch the Docker container.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker run -it \
|
||||
--device /dev/dri \
|
||||
--device /dev/kfd \
|
||||
--network host \
|
||||
--ipc host \
|
||||
--group-add video \
|
||||
--cap-add SYS_PTRACE \
|
||||
--security-opt seccomp=unconfined \
|
||||
--privileged \
|
||||
-v $HOME:$HOME \
|
||||
-v $HOME/.ssh:/root/.ssh \
|
||||
--shm-size 64G \
|
||||
--name training_env \
|
||||
{{ docker.pull_tag }}
|
||||
|
||||
Use these commands if you exit the ``training_env`` container and need to return to it.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
docker start training_env
|
||||
docker exec -it training_env bash
|
||||
|
||||
3. In the Docker container, clone the `<https://github.com/ROCm/MAD>`__
|
||||
repository and navigate to the benchmark scripts directory
|
||||
``/workspace/MAD/scripts/pytorch_train``.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
git clone https://github.com/ROCm/MAD
|
||||
cd MAD/scripts/pytorch_train
|
||||
|
||||
.. rubric:: Prepare training datasets and dependencies
|
||||
|
||||
1. The following benchmarking examples require downloading models and datasets
|
||||
from Hugging Face. To ensure successful access to gated repos, set your
|
||||
``HF_TOKEN``.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
export HF_TOKEN=$your_personal_hugging_face_access_token
|
||||
|
||||
2. Run the setup script to install libraries and datasets needed for benchmarking.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./pytorch_benchmark_setup.sh
|
||||
|
||||
.. container:: model-doc pyt_train_llama-3.1-8b
|
||||
|
||||
``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 8B:
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Library
|
||||
- Reference
|
||||
|
||||
* - ``accelerate``
|
||||
- `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_
|
||||
|
||||
* - ``datasets``
|
||||
- `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
|
||||
|
||||
.. container:: model-doc pyt_train_llama-3.1-70b
|
||||
|
||||
``pytorch_benchmark_setup.sh`` installs the following libraries for Llama 3.1 70B:
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Library
|
||||
- Reference
|
||||
|
||||
* - ``datasets``
|
||||
- `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`_ 3.2.0
|
||||
|
||||
* - ``torchdata``
|
||||
- `TorchData <https://meta-pytorch.org/data/beta/index.html#torchdata>`__
|
||||
|
||||
* - ``tomli``
|
||||
- `Tomli <https://pypi.org/project/tomli/>`__
|
||||
|
||||
* - ``tiktoken``
|
||||
- `tiktoken <https://github.com/openai/tiktoken>`__
|
||||
|
||||
* - ``blobfile``
|
||||
- `blobfile <https://pypi.org/project/blobfile/>`__
|
||||
|
||||
* - ``tabulate``
|
||||
- `tabulate <https://pypi.org/project/tabulate/>`__
|
||||
|
||||
* - ``wandb``
|
||||
- `Weights & Biases <https://github.com/wandb/wandb>`__
|
||||
|
||||
* - ``sentencepiece``
|
||||
- `SentencePiece <https://github.com/google/sentencepiece>`__ 0.2.0
|
||||
|
||||
* - ``tensorboard``
|
||||
- `TensorBoard <https://www.tensorflow.org/tensorboard>`__ 2.18.0
|
||||
|
||||
.. container:: model-doc pyt_train_flux
|
||||
|
||||
``pytorch_benchmark_setup.sh`` installs the following libraries for FLUX:
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Library
|
||||
- Reference
|
||||
|
||||
* - ``accelerate``
|
||||
- `Hugging Face Accelerate <https://huggingface.co/docs/accelerate/en/index>`_
|
||||
|
||||
* - ``datasets``
|
||||
- `Hugging Face Datasets <https://huggingface.co/docs/datasets/v3.2.0/en/index>`__ 3.2.0
|
||||
|
||||
* - ``sentencepiece``
|
||||
- `SentencePiece <https://github.com/google/sentencepiece>`__ 0.2.0
|
||||
|
||||
* - ``tensorboard``
|
||||
- `TensorBoard <https://www.tensorflow.org/tensorboard>`__ 2.18.0
|
||||
|
||||
* - ``csvkit``
|
||||
- `csvkit <https://csvkit.readthedocs.io/en/latest/>`__ 2.0.1
|
||||
|
||||
* - ``deepspeed``
|
||||
- `DeepSpeed <https://github.com/deepspeedai/DeepSpeed>`__ 0.16.2
|
||||
|
||||
* - ``diffusers``
|
||||
- `Hugging Face Diffusers <https://huggingface.co/docs/diffusers/en/index>`__ 0.31.0
|
||||
|
||||
* - ``GitPython``
|
||||
- `GitPython <https://github.com/gitpython-developers/GitPython>`__ 3.1.44
|
||||
|
||||
* - ``opencv-python-headless``
|
||||
- `opencv-python-headless <https://pypi.org/project/opencv-python-headless/>`__ 4.10.0.84
|
||||
|
||||
* - ``peft``
|
||||
- `PEFT <https://huggingface.co/docs/peft/en/index>`__ 0.14.0
|
||||
|
||||
* - ``protobuf``
|
||||
- `Protocol Buffers <https://github.com/protocolbuffers/protobuf>`__ 5.29.2
|
||||
|
||||
* - ``pytest``
|
||||
- `PyTest <https://docs.pytest.org/en/stable/>`__ 8.3.4
|
||||
|
||||
* - ``python-dotenv``
|
||||
- `python-dotenv <https://pypi.org/project/python-dotenv/>`__ 1.0.1
|
||||
|
||||
* - ``seaborn``
|
||||
- `Seaborn <https://seaborn.pydata.org/>`__ 0.13.2
|
||||
|
||||
* - ``transformers``
|
||||
- `Transformers <https://huggingface.co/docs/transformers/en/index>`__ 4.47.0
|
||||
|
||||
``pytorch_benchmark_setup.sh`` downloads the following datasets from Hugging Face:
|
||||
|
||||
* `frank-chieng/chinese_architecture_siheyuan <https://huggingface.co/datasets/frank-chieng/chinese_architecture_siheyuan>`__
|
||||
|
||||
{% for model_group in model_groups %}
|
||||
{% for model in model_group.models %}
|
||||
{% set training_modes = model.training_modes %}
|
||||
{% set training_mode_descs = {
|
||||
"pretrain": "Benchmark pre-training.",
|
||||
"HF_pretrain": "Llama 3.1 8B pre-training with FP8 precision."
|
||||
} %}
|
||||
{% set available_modes = training_modes | select("in", ["pretrain", "HF_pretrain"]) | list %}
|
||||
{% if available_modes %}
|
||||
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
.. rubric:: Pretraining
|
||||
|
||||
To start the pre-training benchmark, use the following command with the
|
||||
appropriate options. See the following list of options and their descriptions.
|
||||
|
||||
{% if model.mad_tag == "pyt_train_dlrm" %}
|
||||
|
||||
1. Go to the DLRM directory.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
cd /workspace/DLRMBenchmark
|
||||
|
||||
2. To run the single node training benchmark for DLRM-v2 with TF32 precision,
|
||||
run the following script.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./launch_training_single_node.sh
|
||||
|
||||
To run with MAD within the Docker container, use the following command.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./pytorch_benchmark_report.sh -t pretrain -m DLRM
|
||||
|
||||
{% else %}
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./pytorch_benchmark_report.sh -t {% if available_modes | length == 1 %}{{ available_modes[0] }}{% else %}$training_mode{% endif %} \
|
||||
-m {{ model.model_repo }} \
|
||||
-p $datatype \
|
||||
-s $sequence_length
|
||||
|
||||
{% if model.mad_tag == "pyt_train_flux" %}
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
.. note::
|
||||
|
||||
Currently, FLUX models are not supported out-of-the-box on this Docker.
|
||||
To use FLUX, refer to ``rocm/pytorch-training`` Docker: :doc:`pytorch-training-v25.6`
|
||||
|
||||
Occasionally, downloading the Flux dataset might fail. In the event of this
|
||||
error, manually download it from Hugging Face at
|
||||
`black-forest-labs/FLUX.1-dev <https://huggingface.co/black-forest-labs/FLUX.1-dev>`_
|
||||
and save it to `/workspace/FluxBenchmark`. This ensures that the test script can access
|
||||
the required dataset.
|
||||
{% endif %}
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Name
|
||||
- Options
|
||||
- Description
|
||||
|
||||
{% for mode in available_modes %}
|
||||
* - {% if loop.first %}``$training_mode``{% endif %}
|
||||
- ``{{ mode }}``
|
||||
- {{ training_mode_descs[mode] }}
|
||||
{% endfor %}
|
||||
|
||||
* - ``$datatype``
|
||||
- ``BF16``{% if model.mad_tag == "pyt_train_llama-3.1-8b" %} or ``FP8``{% endif %}
|
||||
- Only Llama 3.1 8B supports FP8 precision.
|
||||
|
||||
* - ``$sequence_length``
|
||||
- Sequence length for the language model.
|
||||
- Between 2048 and 8192. 8192 by default.
|
||||
{% endif %}
|
||||
{% endif %}
|
||||
|
||||
{% set training_modes = model.training_modes %}
|
||||
{% set training_mode_descs = {
|
||||
"posttrain": "Benchmark post-training.",
|
||||
} %}
|
||||
{% set available_modes = training_modes | select("in", ["posttrain"]) | list %}
|
||||
{% if available_modes %}
|
||||
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
.. rubric:: Post-training
|
||||
|
||||
To start the post-training benchmark, use the following command with the
|
||||
appropriate options. See the following list of options and their descriptions.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./pytorch_benchmark_report.sh -t {% if available_modes | length == 1 %}{{ available_modes[0] }}{% else %}$training_mode{% endif %} \
|
||||
-m {{ model.model_repo }} \
|
||||
-p $datatype \
|
||||
-s $sequence_length
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Name
|
||||
- Options
|
||||
- Description
|
||||
|
||||
{% for mode in available_modes %}
|
||||
* - {% if loop.first %}``$training_mode``{% endif %}
|
||||
- ``{{ mode }}``
|
||||
- {{ training_mode_descs[mode] }}
|
||||
{% endfor %}
|
||||
|
||||
* - ``$datatype``
|
||||
- ``BF16``{% if model.mad_tag == "pyt_train_llama-3.1-8b" %} or ``FP8``{% endif %}
|
||||
- Only Llama 3.1 8B supports FP8 precision.
|
||||
|
||||
* - ``$sequence_length``
|
||||
- Sequence length for the language model.
|
||||
- Between 2048 and 8192. 8192 by default.
|
||||
{% endif %}
|
||||
|
||||
{% set training_mode_descs = {
|
||||
"finetune_fw": "Full weight fine-tuning (BF16 and FP8 supported).",
|
||||
"finetune_lora": "LoRA fine-tuning (BF16 supported).",
|
||||
"finetune_qlora": "QLoRA fine-tuning (BF16 supported).",
|
||||
"HF_finetune_lora": "LoRA fine-tuning with Hugging Face PEFT.",
|
||||
} %}
|
||||
{% set available_modes = training_modes | select("in", ["finetune_fw", "finetune_lora", "finetune_qlora", "HF_finetune_lora"]) | list %}
|
||||
{% if available_modes %}
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
.. rubric:: Fine-tuning
|
||||
|
||||
To start the fine-tuning benchmark, use the following command with the
|
||||
appropriate options. See the following list of options and their descriptions.
|
||||
See :ref:`supported training modes <amd-pytorch-training-supported-training-modes-v25.11>`.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
./pytorch_benchmark_report.sh -t $training_mode \
|
||||
-m {{ model.model_repo }} \
|
||||
-p $datatype \
|
||||
-s $sequence_length
|
||||
|
||||
.. list-table::
|
||||
:header-rows: 1
|
||||
|
||||
* - Name
|
||||
- Options
|
||||
- Description
|
||||
|
||||
{% for mode in available_modes %}
|
||||
* - {% if loop.first %}``$training_mode``{% endif %}
|
||||
- ``{{ mode }}``
|
||||
- {{ training_mode_descs[mode] }}
|
||||
{% endfor %}
|
||||
|
||||
* - ``$datatype``
|
||||
- ``BF16``{% if "finetune_fw" in available_modes %} or ``FP8``{% endif %}
|
||||
- All models support BF16.{% if "finetune_fw" in available_modes %} FP8 is only available for full weight fine-tuning.{% endif %}
|
||||
|
||||
* - ``$sequence_length``
|
||||
- Between 2048 and 16384.
|
||||
- Sequence length for the language model.
|
||||
|
||||
{% if model.mad_tag in ["pyt_train_llama3.2-vision-11b", "pyt_train_llama-3.2-vision-90b"] %}
|
||||
.. note::
|
||||
|
||||
For LoRA and QLoRA support with vision models (Llama 3.2 11B and 90B),
|
||||
use the following torchtune commit for compatibility:
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
git checkout 48192e23188b1fc524dd6d127725ceb2348e7f0e
|
||||
|
||||
{% elif model.mad_tag in ["pyt_train_llama-2-7b", "pyt_train_llama-2-13b", "pyt_train_llama-2-70b"] %}
|
||||
.. note::
|
||||
|
||||
You might encounter the following error with Llama 2: ``ValueError: seq_len (16384) of
|
||||
input tensor should be smaller than max_seq_len (4096)``.
|
||||
This error indicates that an input sequence is longer than the model's maximum context window.
|
||||
|
||||
Ensure your tokenized input does not exceed the model's ``max_seq_len`` (4096
|
||||
tokens in this case). You can resolve this by truncating the input or splitting
|
||||
it into smaller chunks before passing it to the model.
|
||||
|
||||
Note on reproducibility: The results in this guide are based on
|
||||
commit ``b4c98ac`` from the upstream
|
||||
`<https://github.com/pytorch/torchtune>`__ repository. For the
|
||||
latest updates, you can use the main branch.
|
||||
|
||||
{% endif %}
|
||||
{% endif %}
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
.. rubric:: Benchmarking examples
|
||||
|
||||
For examples of benchmarking commands, see `<https://github.com/ROCm/MAD/tree/develop/benchmark/pytorch_train#benchmarking-examples>`__.
|
||||
|
||||
.. _amd-pytorch-training-multinode-examples-v25.11:
|
||||
|
||||
Multi-node training
|
||||
-------------------
|
||||
|
||||
Refer to :doc:`/how-to/rocm-for-ai/system-setup/multi-node-setup` to configure your environment for multi-node
|
||||
training. See :ref:`rocm-for-ai-multi-node-setup-pyt-train-example` for example Slurm run commands.
|
||||
|
||||
Pre-training
|
||||
~~~~~~~~~~~~
|
||||
|
||||
Multi-node training with torchtitan is supported. The provided SLURM script is pre-configured for Llama 3 70B.
|
||||
|
||||
To launch the training job on a SLURM cluster for Llama 3 70B, run the following commands from the MAD repository.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
# In the MAD repository
|
||||
cd scripts/pytorch_train
|
||||
sbatch run_slurm_train.sh
|
||||
|
||||
Fine-tuning
|
||||
~~~~~~~~~~~
|
||||
|
||||
Multi-node training with torchtune is supported. The provided SLURM script is pre-configured for Llama 3.3 70B.
|
||||
|
||||
To launch the training job on a SLURM cluster for Llama 3.3 70B, run the following commands from the MAD repository.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
huggingface-cli login # Get access to HF Llama model space
|
||||
huggingface-cli download meta-llama/Llama-3.3-70B-Instruct --local-dir ./models/Llama-3.3-70B-Instruct # Download the Llama 3.3 model locally
|
||||
# In the MAD repository
|
||||
cd scripts/pytorch_train
|
||||
sbatch Torchtune_Multinode.sh
|
||||
|
||||
.. note::
|
||||
|
||||
Information regarding benchmark setup:
|
||||
|
||||
* By default, Llama 3.3 70B is fine-tuned using ``alpaca_dataset``.
|
||||
* You can adjust the torchtune `YAML configuration file
|
||||
<https://github.com/pytorch/torchtune/blob/main/recipes/configs/llama3_3/70B_full_multinode.yaml>`__
|
||||
if you're using a different model.
|
||||
* The number of nodes and other parameters can be tuned in the SLURM script ``Torchtune_Multinode.sh``.
|
||||
* Set the ``mounting_paths`` inside the SLURM script.
|
||||
|
||||
Once the run is finished, you can find the log files in the ``result_torchtune/`` directory.
|
||||
|
||||
Further reading
|
||||
===============
|
||||
|
||||
- To learn more about MAD and the ``madengine`` CLI, see the `MAD usage guide <https://github.com/ROCm/MAD?tab=readme-ov-file#usage-guide>`__.
|
||||
|
||||
- To learn more about system settings and management practices to configure your system for
|
||||
AMD Instinct MI300X Series GPUs, see `AMD Instinct MI300X system optimization <https://instinct.docs.amd.com/projects/amdgpu-docs/en/latest/system-optimization/mi300x.html>`_.
|
||||
|
||||
- For a list of other ready-made Docker images for AI with ROCm, see
|
||||
`AMD Infinity Hub <https://www.amd.com/en/developer/resources/infinity-hub.html#f-amd_hub_category=AI%20%26%20ML%20Models>`_.
|
||||
|
||||
Previous versions
|
||||
=================
|
||||
|
||||
See :doc:`pytorch-training-history` to find documentation for previous releases
|
||||
of the ``ROCm/pytorch-training`` Docker image.
|
||||
@@ -47,7 +47,7 @@ Megatron-LM.
|
||||
- {{ component_version }}
|
||||
{% endfor %}
|
||||
|
||||
.. _amd-primus-megatron-lm-model-support-v26.01:
|
||||
.. _amd-primus-megatron-lm-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
@@ -108,7 +108,7 @@ To test for optimal performance, consult the recommended :ref:`System health ben
|
||||
<rocm-for-ai-system-health-bench>`. This suite of tests will help you verify and fine-tune your
|
||||
system's configuration.
|
||||
|
||||
.. _mi300x-amd-primus-megatron-lm-training-v26.01:
|
||||
.. _mi300x-amd-primus-megatron-lm-training-v25.11:
|
||||
|
||||
Environment setup
|
||||
=================
|
||||
@@ -118,7 +118,7 @@ Environment setup
|
||||
Use the following instructions to set up the environment, configure the script to train models, and
|
||||
reproduce the benchmark results on AMD Instinct GPUs.
|
||||
|
||||
.. _amd-primus-megatron-lm-requirements-v26.01:
|
||||
.. _amd-primus-megatron-lm-requirements-v25.11:
|
||||
|
||||
Pull the Docker image
|
||||
|
||||
@@ -157,16 +157,16 @@ Pull the Docker image
|
||||
docker start primus_training_env
|
||||
docker exec -it primus_training_env bash
|
||||
|
||||
The Docker container hosts verified commit ``9c529cd4`` of the `Primus
|
||||
<https://github.com/AMD-AGI/Primus/tree/9c529cd4a934a68a880ede036c3e97b792e38167>`__ repository.
|
||||
The Docker container hosts verified commit ``c4c083de`` of the `Primus
|
||||
<https://github.com/AMD-AGI/Primus/tree/c4c083de64ba3e8f19ccc9629411267108931f9e/>`__ repository.
|
||||
|
||||
.. _amd-primus-megatron-lm-environment-setup-v26.01:
|
||||
.. _amd-primus-megatron-lm-environment-setup-v25.11:
|
||||
|
||||
Configuration
|
||||
=============
|
||||
|
||||
Primus defines a training configuration in YAML for each model in
|
||||
`examples/megatron/configs <https://github.com/AMD-AGI/Primus/tree/9c529cd4a934a68a880ede036c3e97b792e38167/examples/megatron/configs>`__.
|
||||
`examples/megatron/configs <https://github.com/AMD-AGI/Primus/tree/c4c083de64ba3e8f19ccc9629411267108931f9e/examples/megatron/configs>`__.
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
|
||||
@@ -207,7 +207,7 @@ You can use either mock data or real data for training.
|
||||
|
||||
Ensure that the files are accessible inside the Docker container.
|
||||
|
||||
.. _amd-primus-megatron-lm-tokenizer-v26.01:
|
||||
.. _amd-primus-megatron-lm-tokenizer-v25.11:
|
||||
|
||||
Tokenizer
|
||||
---------
|
||||
@@ -220,7 +220,15 @@ right permissions to access the tokenizer for each model.
|
||||
# Export your HF_TOKEN in the workspace
|
||||
export HF_TOKEN=<your_hftoken>
|
||||
|
||||
.. _amd-primus-megatron-lm-run-training-v26.01:
|
||||
.. note::
|
||||
|
||||
In Primus, each model uses a tokenizer from Hugging Face. For example, Llama
|
||||
3.1 8B model uses ``tokenizer_model: meta-llama/Llama-3.1-8B`` and
|
||||
``tokenizer_type: Llama3Tokenizer`` defined in the `llama3.1-8B model
|
||||
<https://github.com/AMD-AGI/Primus/blob/e16b27bf6c1b2798f38848fc574fee60d9a9b902/examples/megatron/configs/llama3.1_8B-pretrain.yaml>`__
|
||||
definition.
|
||||
|
||||
.. _amd-primus-megatron-lm-run-training-v25.11:
|
||||
|
||||
Run training
|
||||
============
|
||||
@@ -244,7 +252,7 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 3.3 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run pre-training for Llama 3.3 70B BF16, run:
|
||||
|
||||
@@ -255,10 +263,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.3_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama3.3_70B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama3.3_70B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -270,16 +276,14 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.3_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama3.3_70B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama3.3_70B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_llama-3.1-8b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 3.1 8B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run pre-training for Llama 3.1 8B FP8, run:
|
||||
|
||||
@@ -290,10 +294,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama3.1_8B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -305,10 +307,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
For Llama 3.1 8B BF16, use the following command:
|
||||
|
||||
@@ -319,10 +319,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama3.1_8B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama3.1_BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -334,16 +332,14 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_llama-3.1-70b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 3.1 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run pre-training for Llama 3.1 70B BF16, run:
|
||||
|
||||
@@ -354,10 +350,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama3.1_70B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -369,10 +363,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
To run the training on a single node for Llama 3.1 70B FP8, use the following command.
|
||||
|
||||
@@ -389,10 +381,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama3.1_70B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -404,20 +394,18 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B_fp8_proxy.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
--train_iters 50 \
|
||||
--num_layers 40 \
|
||||
--fp8 hybrid \
|
||||
--no_fp8_weight_transpose_cache true
|
||||
EXP=examples/megatron/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh \
|
||||
--train_iters 50 \
|
||||
--num_layers 40 \
|
||||
--fp8 hybrid \
|
||||
--no_fp8_weight_transpose_cache true
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_llama-2-7b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 2 7B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run pre-training for Llama 2 7B FP8, run:
|
||||
|
||||
@@ -428,10 +416,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama2_7B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama2_7B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama2_7B-FP8-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -443,10 +429,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama2_7B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama2_7B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama2_7B-FP8-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
To run pre-training for Llama 2 7B BF16, run:
|
||||
|
||||
@@ -457,10 +441,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama2_7B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama2_7B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama2_7B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -472,16 +454,14 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama2_7B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama2_7B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama2_7B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_llama-2-70b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 2 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run pre-training for Llama 2 70B BF16, run:
|
||||
|
||||
@@ -492,10 +472,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama2_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/llama2_70B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/llama2_70B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -507,16 +485,14 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama2_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/llama2_70B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/llama2_70B-BF16-pretrain.yaml \
|
||||
bash ./examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_deepseek-v3-proxy
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to DeepSeek-V3.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run training on a single node for DeepSeek-V3 (MoE with expert parallel) BF16 with 3-layer proxy,
|
||||
use the following command:
|
||||
@@ -528,15 +504,13 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v3_proxy.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/deepseek_v3-BF16-pretrain.yaml \
|
||||
--num_layers 3 \
|
||||
--moe_layer_freq 1 \
|
||||
--train_iters 50 \
|
||||
--micro_batch_size 8 \
|
||||
--global_batch_size 64
|
||||
EXP=examples/megatron/configs/MI355X/deepseek_v3-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh \
|
||||
--num_layers 3 \
|
||||
--moe_layer_freq 1 \
|
||||
--train_iters 50 \
|
||||
--micro_batch_size 8 \
|
||||
--global_batch_size 64
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -548,21 +522,17 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v3_proxy.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/deepseek_v3-BF16-pretrain.yaml \
|
||||
--num_layers 3 \
|
||||
--moe_layer_freq 1 \
|
||||
--micro_batch_size 3 \
|
||||
--global_batch_size 192 \
|
||||
--train_iters 50
|
||||
EXP=examples/megatron/configs/MI300X/deepseek_v3-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh \
|
||||
--num_layers 3 \
|
||||
--moe_layer_freq 1 \
|
||||
--train_iters 50
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_deepseek-v2-lite-16b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to DeepSeek-V2-Lite.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run training on a single node for DeepSeek-V2-Lite (MoE with expert parallel) BF16,
|
||||
use the following command:
|
||||
@@ -574,10 +544,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v2_lite.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs//MI355X/deepseek_v2_lite-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/deepseek_v2_lite-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -589,16 +557,14 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v2_lite.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/deepseek_v2_lite-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/deepseek_v2_lite-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_mixtral-8x7b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Mixtral 8x7B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run training on a single node for Mixtral 8x7B (MoE with expert parallel),
|
||||
use the following command:
|
||||
@@ -610,10 +576,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_mixtral_8x7B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/mixtral_8x7B_v0.1-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/mixtral_8x7B_v0.1-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -625,16 +589,15 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_mixtral_8x7B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/mixtral_8x7B_v0.1-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/mixtral_8x7B_v0.1-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh \
|
||||
--train_iters 50
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_mixtral-8x22b-proxy
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Mixtral 8x22B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run training on a single node for Mixtral 8x22B BF16 (MoE with expert parallel) 4-layer proxy,
|
||||
use the following command:
|
||||
@@ -646,10 +609,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_mixtral_8x22B_proxy.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/mixtral_8x22B_v0.1-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/mixtral_8x22B_v0.1-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -661,21 +622,19 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_mixtral_8x22B_proxy.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/mixtral_8x22B_v0.1-BF16-pretrain.yaml \
|
||||
--num_layers 4 \
|
||||
--pipeline_model_parallel_size 1 \
|
||||
--micro_batch_size 1 \
|
||||
--global_batch_size 16 \
|
||||
--train_iters 50
|
||||
EXP=examples/megatron/configs/MI300X/mixtral_8x22B_v0.1-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh \
|
||||
--train_iters 50 \
|
||||
--num_layers 4 \
|
||||
--pipeline_model_parallel_size 1 \
|
||||
--micro_batch_size 1 \
|
||||
--global_batch_size 16
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_qwen2.5-7b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Qwen 2.5 7B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run training on a single node for Qwen 2.5 7B BF16, use the following
|
||||
command:
|
||||
@@ -687,10 +646,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_qwen2.5_7B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/qwen2.5_7B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/qwen2.5_7B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -702,10 +659,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_qwen2.5_7B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/qwen2.5_7B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/qwen2.5_7B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
For FP8, use the following command.
|
||||
|
||||
@@ -716,10 +671,8 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_qwen2.5_7B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/qwen2.5_7B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/qwen2.5_7B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -731,16 +684,14 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_qwen2.5_7B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/qwen2.5_7B-FP8-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/qwen2.5_7B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_qwen2.5-72b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Qwen 2.5 72B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To run the training on a single node for Qwen 2.5 72B BF16, use the following command.
|
||||
|
||||
@@ -751,10 +702,11 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_qwen2.5_72B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI355X/qwen2.5_72B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI355X/qwen2.5_72B-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh \
|
||||
--train_iters 50 \
|
||||
--micro_batch_size 16 \
|
||||
--global_batch_size 256
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI325X and MI300X
|
||||
@@ -766,12 +718,10 @@ To run training on a single node, navigate to ``/workspace/Primus`` and use the
|
||||
export PRIMUS_TURBO_ATTN_V3_ATOMIC_FP32=1
|
||||
export NVTE_CK_IS_V3_ATOMIC_FP32=1
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_qwen2.5_72B.log \
|
||||
-- train pretrain \
|
||||
--config examples/megatron/configs/MI300X/qwen2.5_72B-BF16-pretrain.yaml
|
||||
EXP=examples/megatron/configs/MI300X/qwen2.5_72B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. _amd-primus-megatron-multi-node-examples-v26.01:
|
||||
.. _amd-primus-megatron-multi-node-examples-v25.11:
|
||||
|
||||
Multi-node training examples
|
||||
----------------------------
|
||||
@@ -780,7 +730,7 @@ Refer to :doc:`/how-to/rocm-for-ai/system-setup/multi-node-setup` to configure y
|
||||
training.
|
||||
|
||||
To run training on multiple nodes, you can use the
|
||||
`run_slurm_pretrain.sh <https://github.com/AMD-AGI/Primus/blob/9c529cd4a934a68a880ede036c3e97b792e38167/examples/run_slurm_pretrain.sh>`__
|
||||
`run_slurm_pretrain.sh <https://github.com/AMD-AGI/Primus/blob/main/examples/run_slurm_pretrain.sh>`__
|
||||
to launch the multi-node workload. Use the following steps to setup your environment:
|
||||
|
||||
.. datatemplate:yaml:: /data/how-to/rocm-for-ai/training/primus-megatron-benchmark-models.yaml
|
||||
@@ -813,13 +763,13 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
* If ``NCCL_IB_HCA`` and ``NCCL_SOCKET_IFNAME`` are not set, Primus will try to auto-detect. However, since NICs can vary accross different cluster, it is encouraged to explicitly export your NCCL parameters for the cluster.
|
||||
* To find your network interface, you can use ``ip a``.
|
||||
* To find RDMA interfaces, you can use ``ibv_devices`` to get the list of all the RDMA/IB devices.
|
||||
* Remember to set ``DOCKER_IMAGE`` and ``HF_TOKEN`` (see :ref:`amd-primus-megatron-lm-tokenizer-v26.01`) as appropriate.
|
||||
* Remember to set ``DOCKER_IMAGE`` and ``HF_TOKEN`` (see :ref:`amd-primus-megatron-lm-tokenizer-v25.11`) as appropriate.
|
||||
|
||||
.. container:: model-doc primus_pyt_megatron_lm_train_llama-3.1-8b
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 3.1 8B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Llama 3.1 8B FP8 on 8 nodes, run:
|
||||
|
||||
@@ -836,7 +786,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 2 7B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Llama 2 7B FP8 on 8 nodes, run:
|
||||
|
||||
@@ -853,7 +803,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 3.1 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Llama 3.1 70B FP8 on 8 nodes, run:
|
||||
|
||||
@@ -883,7 +833,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 2 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Llama 2 70B FP8 on 8 nodes, run:
|
||||
|
||||
@@ -913,7 +863,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 3.3 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Llama 3.3 70B FP8 on 8 nodes, run:
|
||||
|
||||
@@ -943,7 +893,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 2 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Mixtral 8x7B BF16 on 8 nodes, run:
|
||||
|
||||
@@ -961,7 +911,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
|
||||
Once setup is complete, run the appropriate training command.
|
||||
The following run commands are tailored to Llama 2 70B.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-megatron-lm-model-support-v25.11` to switch to another available model.
|
||||
|
||||
To train Qwen2.5 72B FP8 on 8 nodes, run:
|
||||
|
||||
@@ -976,7 +926,7 @@ to launch the multi-node workload. Use the following steps to setup your environ
|
||||
--global_batch_size 512 \
|
||||
--recompute_num_layers 80 \
|
||||
|
||||
.. _amd-primus-megatron-lm-benchmark-test-vars-v26.01:
|
||||
.. _amd-primus-megatron-lm-benchmark-test-vars-v25.11:
|
||||
|
||||
Key options
|
||||
-----------
|
||||
|
||||
@@ -45,7 +45,7 @@ with Primus Turbo optimizations.
|
||||
- {{ component_version }}
|
||||
{% endfor %}
|
||||
|
||||
.. _amd-primus-pytorch-model-support-v26.01:
|
||||
.. _amd-primus-pytorch-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
@@ -91,7 +91,7 @@ vary by model -- select one to get started.
|
||||
For additional workloads, including Llama 3.3, Llama 3.2, Llama 2, GPT OSS, Qwen, and Flux models,
|
||||
see the documentation :doc:`pytorch-training` (without Primus)
|
||||
|
||||
.. _amd-primus-pytorch-performance-measurements-v26.01:
|
||||
.. _amd-primus-pytorch-performance-measurements-v25.11:
|
||||
|
||||
System validation
|
||||
=================
|
||||
@@ -146,7 +146,7 @@ tweak some configurations (such as batch sizes).
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following run command is tailored to {{ model.model }}.
|
||||
See :ref:`amd-primus-pytorch-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-pytorch-model-support-v25.11` to switch to another available model.
|
||||
|
||||
1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
|
||||
directory and install the required packages on the host machine.
|
||||
@@ -184,7 +184,7 @@ tweak some configurations (such as batch sizes).
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following run commands are tailored to {{ model.model }}.
|
||||
See :ref:`amd-primus-pytorch-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-primus-pytorch-model-support-v25.11` to switch to another available model.
|
||||
|
||||
.. rubric:: Download the Docker image and required packages
|
||||
|
||||
@@ -220,8 +220,8 @@ tweak some configurations (such as batch sizes).
|
||||
docker start training_env
|
||||
docker exec -it training_env bash
|
||||
|
||||
The Docker container hosts verified commit ``9c529cd4`` of the `Primus
|
||||
<https://github.com/AMD-AGI/Primus/tree/9c529cd4a934a68a880ede036c3e97b792e38167/>`__ repository.
|
||||
The Docker container hosts verified commit ``c4c083de`` of the `Primus
|
||||
<https://github.com/AMD-AGI/Primus/tree/c4c083de64ba3e8f19ccc9629411267108931f9e/>`__ repository.
|
||||
|
||||
.. rubric:: Prepare training datasets and dependencies
|
||||
|
||||
@@ -257,31 +257,24 @@ tweak some configurations (such as batch sizes).
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI355X/llama3.1_8B-BF16-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
--training.local_batch_size 6
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 6
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
To train Llama 3.1 8B with FP8 precision, use the following command.
|
||||
|
||||
@@ -292,31 +285,24 @@ tweak some configurations (such as batch sizes).
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI355X/llama3.1_8B-FP8-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
--training.local_batch_size 7
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 7
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_8B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_8B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_train_llama-3.1-70b
|
||||
|
||||
@@ -329,31 +315,24 @@ tweak some configurations (such as batch sizes).
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI355X/llama3.1_70B-BF16-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
--training.local_batch_size 6
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 6
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-BF16-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
To train Llama 3.1 70B with FP8 precision, use the following command.
|
||||
|
||||
@@ -364,31 +343,24 @@ tweak some configurations (such as batch sizes).
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI355X/llama3.1_70B-FP8-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI355X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
--training.local_batch_size 5
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 5
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_llama3.1_70B_fp8.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI300X/llama3.1_70B-FP8-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. container:: model-doc primus_pyt_train_deepseek-v3-16b
|
||||
|
||||
@@ -401,31 +373,24 @@ tweak some configurations (such as batch sizes).
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v3_16b.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI355X/deepseek_v3_16b-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI355X/deepseek_v3_16b-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
|
||||
.. tab-item:: MI325X
|
||||
:sync: MI325X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v3_16b.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/deepseek_v3_16b-pretrain.yaml \
|
||||
--training.local_batch_size 10
|
||||
EXP=examples/torchtitan/configs/MI300X/deepseek_v3_16b-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh --training.local_batch_size 10
|
||||
|
||||
.. tab-item:: MI300X
|
||||
:sync: MI300X
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
bash runner/primus-cli direct \
|
||||
--log_file /tmp/primus_deepseek_v3_16b.log \
|
||||
-- train pretrain \
|
||||
--config examples/torchtitan/configs/MI300X/deepseek_v3_16b-pretrain.yaml
|
||||
EXP=examples/torchtitan/configs/MI300X/deepseek_v3_16b-pretrain.yaml \
|
||||
bash examples/run_pretrain.sh
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
|
||||
|
||||
@@ -43,7 +43,7 @@ training workloads:
|
||||
- {{ component_version }}
|
||||
{% endfor %}
|
||||
|
||||
.. _amd-pytorch-training-model-support-v26.01:
|
||||
.. _amd-pytorch-training-model-support-v25.11:
|
||||
|
||||
Supported models
|
||||
================
|
||||
@@ -85,7 +85,7 @@ one to get started.
|
||||
</div>
|
||||
</div>
|
||||
|
||||
.. _amd-pytorch-training-supported-training-modes-v26.01:
|
||||
.. _amd-pytorch-training-supported-training-modes-v25.11:
|
||||
|
||||
The following table lists supported training modes per model.
|
||||
|
||||
@@ -120,7 +120,7 @@ The following table lists supported training modes per model.
|
||||
unlisted fine-tuning methods by using an existing file in the
|
||||
``/workspace/torchtune/recipes/configs`` directory as a template.
|
||||
|
||||
.. _amd-pytorch-training-performance-measurements-v26.01:
|
||||
.. _amd-pytorch-training-performance-measurements-v25.11:
|
||||
|
||||
Performance measurements
|
||||
========================
|
||||
@@ -176,7 +176,7 @@ Run training
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following run command is tailored to {{ model.model }}.
|
||||
See :ref:`amd-pytorch-training-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-pytorch-training-model-support-v25.11` to switch to another available model.
|
||||
|
||||
1. Clone the ROCm Model Automation and Dashboarding (`<https://github.com/ROCm/MAD>`__) repository to a local
|
||||
directory and install the required packages on the host machine.
|
||||
@@ -214,7 +214,7 @@ Run training
|
||||
.. container:: model-doc {{ model.mad_tag }}
|
||||
|
||||
The following commands are tailored to {{ model.model }}.
|
||||
See :ref:`amd-pytorch-training-model-support-v26.01` to switch to another available model.
|
||||
See :ref:`amd-pytorch-training-model-support-v25.11` to switch to another available model.
|
||||
|
||||
{% endfor %}
|
||||
{% endfor %}
|
||||
@@ -409,10 +409,6 @@ Run training
|
||||
|
||||
{% if model.mad_tag == "pyt_train_dlrm" %}
|
||||
|
||||
.. note::
|
||||
|
||||
DLRM is supported on MI300X, MI325X, MI350X, and MI355X GPUs.
|
||||
|
||||
1. Go to the DLRM directory.
|
||||
|
||||
.. code-block:: shell
|
||||
@@ -536,7 +532,7 @@ Run training
|
||||
|
||||
To start the fine-tuning benchmark, use the following command with the
|
||||
appropriate options. See the following list of options and their descriptions.
|
||||
See :ref:`supported training modes <amd-pytorch-training-supported-training-modes-v26.01>`.
|
||||
See :ref:`supported training modes <amd-pytorch-training-supported-training-modes-v25.11>`.
|
||||
|
||||
.. code-block:: shell
|
||||
|
||||
@@ -601,7 +597,7 @@ Run training
|
||||
|
||||
For examples of benchmarking commands, see `<https://github.com/ROCm/MAD/tree/develop/benchmark/pytorch_train#benchmarking-examples>`__.
|
||||
|
||||
.. _amd-pytorch-training-multinode-examples-v26.01:
|
||||
.. _amd-pytorch-training-multinode-examples-v25.11:
|
||||
|
||||
Multi-node training
|
||||
-------------------
|
||||
|
||||
@@ -18,6 +18,7 @@
|
||||
(artificial-intelligence-apis)=
|
||||
|
||||
* {doc}`Composable Kernel <composable_kernel:index>`
|
||||
* {doc}`hipDNN <hipdnn:index>`
|
||||
* {doc}`MIGraphX <amdmigraphx:index>`
|
||||
* {doc}`MIOpen <miopen:index>`
|
||||
* {doc}`MIVisionX <mivisionx:index>`
|
||||
|
||||
@@ -37,7 +37,7 @@ click==8.3.1
|
||||
# sphinx-external-toc
|
||||
comm==0.2.3
|
||||
# via ipykernel
|
||||
cryptography==46.0.5
|
||||
cryptography==46.0.3
|
||||
# via pyjwt
|
||||
debugpy==1.8.19
|
||||
# via ipykernel
|
||||
|
||||
@@ -36,6 +36,7 @@ Machine Learning & Computer Vision
|
||||
:header: "Component", "Description"
|
||||
|
||||
":doc:`Composable Kernel <composable_kernel:index>`", "Provides a programming model for writing performance critical kernels for machine learning workloads across multiple architectures"
|
||||
":doc:`hipDNN <hipdnn:index>`", "A graph-based deep learning library that enables multi-operation fusion for improved performance on AMD GPUs. "
|
||||
":doc:`MIGraphX <amdmigraphx:index>`", "Graph inference engine that accelerates machine learning model inference"
|
||||
":doc:`MIOpen <miopen:index>`", "An open source deep-learning library"
|
||||
":doc:`MIVisionX <mivisionx:index>`", "Set of comprehensive computer vision and machine learning libraries, utilities, and applications"
|
||||
|
||||
Reference in New Issue
Block a user