mirror of
https://github.com/ROCm/ROCm.git
synced 2026-01-09 22:58:17 -05:00
4f53183696
* Update previous versions * Add data file * fix filename and anchors * add templates * update .wordlist.txt * Update template and data add missing step fix fmt * update template * fix data * add jax 0.6.0 * update history * update quantized training note
73 lines
2.5 KiB
YAML
73 lines
2.5 KiB
YAML
dockers:
|
|
- pull_tag: rocm/jax-training:maxtext-v25.7
|
|
docker_hub_url: https://hub.docker.com/layers/rocm/jax-training/maxtext-v25.7/images/sha256-45f4c727d4019a63fc47313d3a5f5a5105569539294ddfd2d742218212ae9025
|
|
components:
|
|
ROCm: 6.4.1
|
|
JAX: 0.5.0
|
|
Python: 3.10.12
|
|
Transformer Engine: 2.1.0+90d703dd
|
|
hipBLASLt: 1.x.x
|
|
- pull_tag: rocm/jax-training:maxtext-v25.7-jax060
|
|
docker_hub_url: https://hub.docker.com/layers/rocm/jax-training/maxtext-v25.7/images/sha256-45f4c727d4019a63fc47313d3a5f5a5105569539294ddfd2d742218212ae9025
|
|
components:
|
|
ROCm: 6.4.1
|
|
JAX: 0.6.0
|
|
Python: 3.10.12
|
|
Transformer Engine: 2.1.0+90d703dd
|
|
hipBLASLt: 1.1.0-499ece1c21
|
|
model_groups:
|
|
- group: Meta Llama
|
|
tag: llama
|
|
models:
|
|
- 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"]
|
|
- 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 8B
|
|
mad_tag: jax_maxtext_train_llama-3-8b
|
|
multinode_training_script: llama3_8b_multinode.sh
|
|
doc_options: ["multi-node"]
|
|
- model: Llama 3 70B
|
|
mad_tag: jax_maxtext_train_llama-3-70b
|
|
multinode_training_script: llama3_70b_multinode.sh
|
|
doc_options: ["multi-node"]
|
|
- 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"]
|
|
- 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"]
|