you know what a spinlock is, right?
here's a spinlock:
| /** Lambdas in C. Compile with GCC! | |
| * ███ ▄████████ ███ ▄████████ ███ █▄ ▄██████▄ | |
| *▀█████████▄ ███ ███ ▀█████████▄ ███ ███ ███ ███ ███ ███ | |
| * ▀███▀▀██ ███ █▀ ▀███▀▀██ ███ █▀ ███ ███ ███ ███ | |
| * ███ ▀ ▄███▄▄▄ ███ ▀ ███ ███ ███ ███ ███ | |
| * ███ ▀▀███▀▀▀ ███ ▀███████████ ███ ███ ███ ███ | |
| * ███ ███ █▄ ███ ███ ███ ███ ███ ███ | |
| * ███ ███ ███ ███ ▄█ ███ ███ ███ ███ ███ | |
| * ▄████▀ ██████████ ▄████▀ ▄████████▀ ████████▀ ▀██████▀ | |
| * |
| # GitHub Action that will run prettier on the whole repo and commit the changes to the PR. | |
| name: Prettier | |
| on: | |
| pull_request: | |
| branches: [main] | |
| jobs: | |
| prettier: | |
| runs-on: ubuntu-latest |
| List of domains that are registered with squarespace and thus could be vulnerable: | |
| celer.network | |
| pendle.finance | |
| karak.network | |
| compound.finance | |
| hyperliquid.xyz | |
| dydx.exchange | |
| thorchain.com | |
| threshold.network |
| import numpy as np | |
| import torch | |
| import jax | |
| from tqdm import tqdm | |
| from model import LanguageModelConfig, TransformerConfig, QuantizedWeight8bit as QW8Bit | |
| from runners import InferenceRunner, ModelRunner, sample_from_model | |
| CKPT_PATH = "./checkpoints" |
Good question! I am collecting human data on how quantization affects outputs. See here for more information: ggml-org/llama.cpp#5962
In the meantime, use the largest that fully fits in your GPU. If you can comfortably fit Q4_K_S, try using a model with more parameters.
See the wiki upstream: https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix
| // A minimal example to reproduce a TSAN crash on calling vkCreateDevice on Linux/NVIDIA: | |
| // [...] | |
| // ThreadSanitizer:DEADLYSIGNAL | |
| // ==73756==ERROR: ThreadSanitizer: SEGV on unknown address 0x000000001048 (pc 0x560233ed6b07 bp 0x000000000010 sp 0x7f3ac37fe580 T73757) | |
| // ==73756==The signal is caused by a READ memory access. | |
| // [...] | |
| // | |
| // github issue: https://github.com/google/sanitizers/issues/1678 | |
| // nvidia issue: https://developer.nvidia.com/bugs/4470250 | |
| // tested on: Arch Linux 2024-01-19 with NVIDIA driver 545.29.06, clang 16.0.6 |
| <body onload=z=c.getContext`2d`,setInterval(`c.width=W=150,Y<W&&P<Y&Y<P+E|9<p?z.fillText(S++${Y=`,9,9|z.fillRect(p`}*0,Y-=--M${Y+Y},P+E,9,W),P))):p=M=Y=S=6,p=p-6||(P=S%E,W)`,E=49) onclick=M=9><canvas id=c> |
The only people entitled to say how open source 'ought' to work are people who run projects, and the scope of their entitlement extends only to their own projects.
Just because someone open sources something does not imply they owe the world a change in their status, focus and effort, e.g. from inventor to community manager.
As a user of something open source you are not thereby entitled to anything at all. You are not entitled to contribute. You are not entitled to features. You are not entitled to the attention of others. You are not entitled to having value attached to your complaints. You are not entitled to this explanation.
If you have expectations (of others) that aren't being met, those expectations are your own responsibility. You are responsible for your own needs. If you want things, make them.
Peter Naur's classic 1985 essay "Programming as Theory Building" argues that a program is not its source code. A program is a shared mental construct (he uses the word theory) that lives in the minds of the people who work on it. If you lose the people, you lose the program. The code is merely a written representation of the program, and it's lossy, so you can't reconstruct