It is more of a general questions. I am not able to find a public package for torch, i.e. habana_frameworks.torch. Would be nice if wrls was distributed such that one could create the same setup as in docker container.
Would that be possible?
Inside the container the following packages are used (bold ones are available via pypi):
habana_gpu_migration 1.17.1.40
habana-media-loader 1.17.1.40
habana-pyhlml 1.17.1.40
habana_quantization_toolkit 1.17.1.40
habana-torch-dataloader 1.17.1.40
habana-torch-plugin 1.17.1.40 → habana_torch_plugin-1.17.1.40-cp310-cp310-linux_x86_64.whl
lightning-habana 1.6.0
There is also neural-compressor · PyPI which can be used with habana devices, but I am not clear yet how/if that overlaps.
What is the Details of the Environment
- Docker or not docker
Hopefully "not docker. "
- Build from source or binary distribution
Well, I’m hoping for binary distributions.
- OS version: uname -a
root@h001:~# uname -a
Linux h001 5.15.0-122-generic #128-Ubuntu SMP Thu Aug 29 13:45:52 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
- Software versions: (dpkg -l | grep habanalabs)
root@h001:~# dpkg -l | grep habanalabs
ii habanalabs-container-runtime 1.17.0-495 amd64 HABANA container runtime
ii habanalabs-dkms 1.17.0-495 all habanalabs driver in DKMS format.
ii habanalabs-firmware 1.17.0-495 amd64 Firmware package for Habana Labs processing accelerators
ii habanalabs-firmware-tools 1.17.0-495 amd64 Habanalabs firmware tools package
ii habanalabs-graph 1.17.0-495 amd64 habanalabs graph compiler
ii habanalabs-qual 1.17.0-495 amd64 This package contains Habanalabs qualification package. It designed to assist server vendors to qualify their Goya based server on the production line.
ii habanalabs-qual-workloads 1.17.0-495 all Habanalabs qual workloads data files
ii habanalabs-rdma-core 1.17.0-495 all Habana Labs rdma-core components.
ii habanalabs-thunk 1.17.0-495 all habanalabs thunk
- Python versions used: python –version
I have all versions of python installed. Default is 3.10.
root@h001:~# which python3 python3.7 python3.8 python3.9 python3.10 python3.11 python3.12
/usr/bin/python3
/usr/bin/python3.7
/usr/bin/python3.8
/usr/bin/python3.9
/usr/bin/python3.10
/usr/bin/python3.11
/usr/bin/python3.12
root@h001:~# python3 -V
Python 3.10.12
- Please attach the dmesg dump, dmesg.log: dmesg > dmesg.log
NA
If Bare Metal, please share the current Habana release version and Firmware version by running this command: sudo hl-smi -q
root@h001:~# sudo hl-smi
±----------------------------------------------------------------------------+
| HL-SMI Version: hl-1.17.0-fw-51.3.0 |
| Driver Version: 1.17.0-28a11ca |
|-------------------------------±---------------------±---------------------+
| AIP Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | AIP-Util Compute M. |
|===============================+======================+======================|
| 0 HL-205 N/A | 0000:b3:00.0 N/A | 0 |
| N/A 33C N/A 104W / 350W | 512MiB / 32768MiB | 15% N/A |
|-------------------------------±---------------------±---------------------+
| 1 HL-205 N/A | 0000:b4:00.0 N/A | 0 |
| N/A 30C N/A 102W / 350W | 512MiB / 32768MiB | 14% N/A |
|-------------------------------±---------------------±---------------------+
| 2 HL-205 N/A | 0000:cd:00.0 N/A | 0 |
| N/A 36C N/A 103W / 350W | 512MiB / 32768MiB | 15% N/A |
|-------------------------------±---------------------±---------------------+
| 3 HL-205 N/A | 0000:cc:00.0 N/A | 0 |
| N/A 30C N/A 103W / 350W | 512MiB / 32768MiB | 15% N/A |
|-------------------------------±---------------------±---------------------+
| 4 HL-205 N/A | 0000:19:00.0 N/A | 0 |
| N/A 33C N/A 96W / 350W | 512MiB / 32768MiB | 12% N/A |
|-------------------------------±---------------------±---------------------+
| 5 HL-205 N/A | 0000:1a:00.0 N/A | 0 |
| N/A 34C N/A 105W / 350W | 512MiB / 32768MiB | 15% N/A |
|-------------------------------±---------------------±---------------------+
| 6 HL-205 N/A | 0000:33:00.0 N/A | 0 |
| N/A 32C N/A 100W / 350W | 512MiB / 32768MiB | 14% N/A |
|-------------------------------±---------------------±---------------------+
| 7 HL-205 N/A | 0000:34:00.0 N/A | 0 |
| N/A 35C N/A 103W / 350W | 512MiB / 32768MiB | 15% N/A |
|-------------------------------±---------------------±---------------------+
–Tore