Top New Best Ask Show Jobs

gdiamos

871 karma · joined December 18, 2014

submissions comments

Outrageously Small NNs: Emergent Reasoning at 6,616 Tok/s on One Intel AMX Core

gregdiamos.com·1 pts·gdiamos·

Outrageously Small Neural Networks: 6,616 tok/s on One Intel AMX Core [pdf]

huggingface.co·2 pts·gdiamos·

CUDA-like programming of Cerebras WSE

github.com·2 pts·gdiamos·

I aint gonna work on Maggie's Datacenter no more [video]

youtube.com·3 pts·gdiamos·

Demo: Fold your coding sessions into LLM weights

app.scalarlmforge.com·2 pts·gdiamos·

LLM-Deflate: Extracting LLMs into Datasets

scalarlm.com·77 pts·gdiamos·

AMD MI300X Memcpy Peer Deep Dive

scalarlm.com·1 pts·gdiamos·

Craylm: Open-source unified LLM training and inference for R1

blog.cray-lm.com·9 pts·gdiamos·

Notes on Building LLMs in 2024

4 pts·gdiamos·

LLM Inference at the Memory Wall

lamini.ai·1 pts·gdiamos·

LLMs hallucinate in critical enterprise scenarios

lamini.ai·6 pts·gdiamos·

Training LLMs in the Wild Wild West – AMD multi-node

lamini.ai·2 pts·gdiamos·

Open source ICD11 Mistral trained on AMD GPUs [video]

youtube.com·2 pts·gdiamos·

Good old fashion ML is dying

20 pts·gdiamos·

AI startup Lamini bets future on AMD's Instinct GPUs

theregister.com·22 pts·gdiamos·

How to use a LLM to classify text

github.com·2 pts·gdiamos·

Laminify: Instantly classify data with Lamini and Llama 2

github.com·2 pts·gdiamos·

Cramming an LLM platform into one container

lamini-ai.github.io·3 pts·gdiamos·

1.109B times faster serving of finetuned LLMs

lamini.ai·2 pts·gdiamos·

Specializing LLMs on Databricks

databricks.com·3 pts·gdiamos·

MLPerf Releases Inference Results

mlperf.org·2 pts·gdiamos·