sivakumar·Sep 10Configuring AMD Instinct GPU SR-IOV to Virtualize GPUThis guide explains how to configure AMD GPU virtualization using SR-IOV on Cisco UCS C-Series GPU servers. The configuration creates GPU…
Jatinkumar Parmar·Aug 23Exploring Speculative Decoding in vLLM on AMD GPUsTL;DR: Speculative decoding allows vLLM to verify multiple drafted tokens in a single target-model pass. In our experiments, its effect on…
Inazhar labsbyazhar·Aug 12+4.6% Free GEMM Throughput on MI210: The Triton Gate No One RevisitedTriton ships two scheduling optimizations for AMD GPUs that quietly skip MI210. Not because the hardware can’t use them because the…
InGoPenAIbyJatinkumar Parmar·Aug 2Helion DSL for GPU kernel development and assessment on AMD GPUsAuthor: Charles Yang , jatinkumar parmar Knowledge level: Intermediate
Zecchi Zero·Jul 31Solución al error torch.cuda.is_available()Este artículo documenta cómo solucionar el error OSError: [Errno 28] No queda espacio en el dispositivo y el AttributeError: module 'torch'…
Balaji Raghuram·May 22The Strix Halo Chronicles #1: Can WSL 2 Beat Native Windows for Qwen 3.6A hands-on engineering log documenting my attempt to run massive LLMs on AMD’s latest 8060S iGPU. This series tracks the setup, hurdles…
AgentZ·Mar 5How to Fix ROCm / PyTorch Memory Faults on AMD GPUs (Segmentation Fault / Page Not Present)Let me get straight to the point: this issue drained days of my life, but I finally found the solution.
Gadget Salvation·Feb 13AMD vs NVIDIA GPU: Which One Gamers Like More?Oh, the eternal AMD vs NVIDIA debate of the PC building world. It continues to spark arguments in forums, dominating Reddit threads. And in…
duckdev·Jul 9, 2024[개발살집] Gemma 한국어 요약 모델 파인 튜닝 해보기LLM 파인튜닝 해보기. 하지만, 장렬히 실패했다.A response icon1A response icon1
![[개발살집] Gemma 한국어 요약 모델 파인 튜닝 해보기](https://miro.medium.com/v2/resize:fill:160:106/1*3SFp_ZJyfcBga4qJ2iFsJg.png)
![[개발살집] Gemma 한국어 요약 모델 파인 튜닝 해보기](https://miro.medium.com/v2/resize:fill:320:214/1*3SFp_ZJyfcBga4qJ2iFsJg.png)