Skip to content

Instantly share code, notes, and snippets.

View shendsaliaga's full-sized avatar

shendsaliaga

1 follower · 8 following
View GitHub Profile
@shendsaliaga
shendsaliaga / Output using 65B on a M1 MacBook Pro 14.2 with 64GB
Created March 17, 2023 03:41 — forked from zitterbewegung/Output using 65B on a M1 MacBook Pro 14.2 with 64GB
llama.cpp 65B run
(venv) # Exit:0 2023-03-12 16:59:27 [r2q2@Reformer#[:~/opt/llama.cpp]
$(: !605 ) ./main -m ./models/65B/ggml-model-q4_0.bin -t 8 -n 128
main: seed = 1678658429
llama_model_load: loading model from './models/65B/ggml-model-q4_0.bin' - please wait ...
llama_model_load: n_vocab = 32000
llama_model_load: n_ctx = 512
llama_model_load: n_embd = 8192
llama_model_load: n_mult = 256
llama_model_load: n_head = 64
llama_model_load: n_layer = 80
@shendsaliaga
shendsaliaga / steps.md
Created April 19, 2022 02:31 — forked from grigorescu/steps.md
Running DPDK Zeek as a non-privileged user

We're tripling down on acronyms, by using the VFIO driver, with IOMMU and SR-IOV.

Boot parameters

intel_iommu=on iommu=pt default_hugepagesz=1G hugepagesz=1G hugepages=16

Enable hugepages

echo 4096 | sudo tee /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages
mkdir -p /tmp/mnt/huge
mount -t hugetlbfs nodev /tmp/mnt/huge