forked from flagos-ai/FlagScale
-
Notifications
You must be signed in to change notification settings - Fork 2
Expand file tree
/
Copy pathmetax.yml
More file actions
57 lines (48 loc) · 1.72 KB
/
Copy pathmetax.yml
File metadata and controls
57 lines (48 loc) · 1.72 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
# MetaX C550 Hardware Configuration
# This file defines CI/CD settings for MetaX C550 GPU (MACA) testing
# Test configurations are defined in tests/test_utils/config/platforms/metax.yaml
hardware_name: metax
display_name: 'MetaX C550 Tests'
# Docker images for MetaX C550 (MACA-based)
ci_image: harbor.baai.ac.cn/flagscale/megatron-lm-with-te:202603231839
ci_train_image: harbor.baai.ac.cn/flagscale/megatron-lm-with-te:202603231839
ci_inference_image: harbor.baai.ac.cn/flagscale/megatron-lm-with-te:202603231839
# Directory to store image tar files (shared between build and test workflows)
tar_dir: /mnt/op/cicd/image_tar
# Runner labels for MetaX C550 hardware
runner_labels: ['flagscale-metax-c550-gpu2-8c-256g']
# Container volumes (hardware-specific paths)
container_volumes:
- /mnt/op/cicd/baai_datasets:/home/gitlab-runner/data
- /mnt/op/cicd/baai_tokenizers:/home/gitlab-runner/tokenizers
# Container options for MetaX C550 (MACA runtime)
container_options: >-
--hostname=flagscale-ci
--ipc=host
--group-add video
--shm-size=100g
--ulimit memlock=-1
--security-opt seccomp=unconfined
--security-opt apparmor=unconfined
--device=/dev/dri
--device=/dev/mxcd
--device=/dev/infiniband
--user root
--ulimit nofile=65535:65535
--cap-add SYS_ADMIN
--cap-add SYS_PTRACE
--cap-add IPC_LOCK
--cap-add NET_ADMIN
--security-opt seccomp=unconfined
# =============================================================================
# Package Manager Configuration
# =============================================================================
pkg_mgr: 'conda'
# Environment path (conda installation path)
env_path: '/opt/conda'
# Conda environment names
env_names:
train: 'base'
hetero_train: ''
inference: ''
rl: ''