DevQuasar

Just sharing a result of a homelab infrastructure experiment:

I've managed to setup a distributed inference infra at home using a DGX Spark (128GB unified gddr6) and a linux workstation with an RTX 6000 Pro (96GB gddr7) connected via 100Gbps RoCEv2. The model I've used (https://lnkd.in/gx6J7YuB) is about 140GB so could not fit either of the GPU. Full setup and tutorial soon on devquasar.com

Screen recording:
https://lnkd.in/gKM9H5GJ

3 replies

csabakecskemeti

posted an update about 1 month ago

Post

1269

FYI: Mistral.Ministral-3 dequantizer FP8->BF16

https://github.com/csabakecskemeti/ministral-3_dequantizer_fp8-bf16

(The instruct model weights are in FP8)

csabakecskemeti

posted an update about 1 month ago

Post

2078

Looking for some help to test an INT8 Deepseek 3.2:
SGLang supports Channel wise INT8 quants on CPUs with AMX instructions (Xeon 5 and above AFAIK)
https://lmsys.org/blog/2025-07-14-intel-xeon-optimization/

Currently uploading an INT8 version of Deepseek 3.2 Speciale:
DevQuasar/deepseek-ai.DeepSeek-V3.2-Speciale-Channel-INT8

I cannot test this I'm on AMD
"AssertionError: W8A8Int8LinearMethod on CPU requires that CPU has AMX support"
(I assumed it can fall back to some non optimized kernel but seems not)

If anyone with the required resources (Intel Xeon 5/6 + ~768-1TB ram) can help to test this that would be awesome.

If you have hints how to make this work on AMD Threadripper 7000 Pro series please guide me.

Thanks all!

8 replies

csabakecskemeti

posted an update 2 months ago

Post

305

Recently there are so much activity on token efficient formats, I've also build a package (inspired by toon).

Deep-TOON

My goal was to token efficiently handle json structures with complex embeddings.

So this is what I've built on the weekend. Feel free try:

https://pypi.org/project/deep-toon/0.1.0/

csabakecskemeti

posted an update 3 months ago

Post

2628

Christmas came early this year

3 replies

csabakecskemeti

posted an update 7 months ago

Post

3074

Has anyone ever backed up a model to a sequential tape drive, or I'm the world first? :D
Just played around with my retro PC that has got a tape drive—did it just because I can.

5 replies

csabakecskemeti

posted an update 8 months ago

Post

477

Deepseek R1 0528 Q2 locally.
(I believe it has overthinking it a bit :) )
https://youtu.be/Iqu5s9aFaXA?si=QWZe293iTKf_3ELU

DevQuasar/deepseek-ai.DeepSeek-R1-0528-GGUF

csabakecskemeti

posted an update 9 months ago

Post

2121

Local Llama4 Maverick Q2
https://youtu.be/4F8g_LThli0?si=MGba2SUTHt6xYw3T
Quants uploading now

Big thanks to @ngxson !

csabakecskemeti

posted an update 9 months ago

Post

1752

Why the 'how many r's in strawberry' prompt "breaks" llama4? :D

Quants DevQuasar/meta-llama.Llama-4-Scout-17B-16E-Instruct-GGUF

3 replies

csabakecskemeti

posted an update 10 months ago

Post

3425

I'm collecting llama-bench results for inference with a llama 3.1 8B q4 and q8 reference models on varoius GPUs. The results are average of 5 executions.
The system varies (different motherboard and CPU ... but that probably that has little effect on the inference performance).

https://devquasar.com/gpu-gguf-inference-comparison/
the exact models user are in the page

I'd welcome results from other GPUs is you have access do anything else you've need in the post. Hopefully this is useful information everyone.

csabakecskemeti

posted an update 10 months ago

Post

2413

Managed to get my hands on a 5090FE, it's beefy

| llama 8B Q8_0 | 7.95 GiB | 8.03 B | CUDA | 99 | pp512 | 12207.44 ± 481.67 |
| llama 8B Q8_0 | 7.95 GiB | 8.03 B | CUDA | 99 | tg128 | 143.18 ± 0.18 |

Comparison with others GPUs
http://devquasar.com/gpu-gguf-inference-comparison/

csabakecskemeti

posted an update 10 months ago

Post

1852

GTC new model announcement now from Nvidia
nvidia/Llama-3_3-Nemotron-Super-49B-v1

GGUFs:
DevQuasar/nvidia.Llama-3_3-Nemotron-Super-49B-v1-GGUF

Enjoy!

AI & ML interests

Recent Activity

Team members 1

DevQuasar's activity