covi

6 ms·

1.
RL Is Everything, Everywhere, All at Once(skypilot.ai)
2 points by covi 17h ago|1 comments
2.
Pulley Is Shutting Down(pulley.com)
4 points by covi 2d ago|1 comments
3.
VRAM Ghost Busting: Who You Gonna Close()?(hcompany.ai)
1 points by covi 3mo ago|0 comments
4.
SkyPilot Sandboxes: Run Agent Code on Your Own Kubernetes, at Scale(blog.skypilot.co)
6 points by covi 3mo ago|0 comments
5.
RL Doesn't Work on Slurm(blog.skypilot.co)
5 points by covi 4mo ago|0 comments
6.
by covi 6mo ago
This feels like the chimpanzee with a power drill. An agent is honestly just brute-force search, but guided.
7.
by covi 7mo ago
The post says Slurm supports gang scheduling, k8s doesn't (out of the box).
8.
SkyPilot: One system to use and manage all AI compute (K8s, 20 clouds, Slurm)(github.com)
15 points by covi 8mo ago|2 comments
9.
The Evolution of AI Job Orchestration(blog.skypilot.co)
2 points by covi 1y ago|0 comments
10.
by covi 1y ago
Take a look at SkyPilot. Good for running these batch workloads. You can use spot instances to save costs.
11.
by covi 1y ago
To massively increase the reliability to get GPUs, you can use something like SkyPilot ( https://github.com/skypilot-org/skypilot ) to fall back across regions, clouds, or GPU choices. E.g., $ sky launch --gpus H100 will
12.
by covi 1y ago
Related: https://skyplane.org/en/latest/ (mentioned in OP) From what I know this idea underpins a few FAANG-level companies' data transfer systems. OP's value = a simple implementation of the idea that&#
13.
Large-Scale AI Batch Inference: 9x Faster Embedding Generation(blog.skypilot.co)
1 points by covi 1y ago|0 comments
14.
Abusing SQLite to Handle Concurrency(blog.skypilot.co)
2 points by covi 2y ago|0 comments
15.
by covi 2y ago
Congrats on the API launch (from SkyPilot)!
16.
Using DeepSeek R1 for RAG: Do's and Don'ts(blog.skypilot.co)
13 points by covi 2y ago|0 comments
17.
Building Large-Scale Image Search Using VectorDB and OpenAI Clip(blog.skypilot.co)
3 points by covi 2y ago|2 comments
18.
Fungible cloud compute and storage: AI training on any cloud(tigrisdata.com)
2 points by covi 2y ago|0 comments
19.
Getting $1M cloud credits for AI startups – and using them wisely(blog.skypilot.co)
1 points by covi 2y ago|0 comments
20.
You Ran the Operational Database on What? Testing Spot Instances(cockroachlabs.com)
3 points by covi 2y ago|1 comments
21.
Can Multi-Modal LLMs "See" Images? A Deep Dive with ASCII Art(blog.skypilot.co)
3 points by covi 2y ago|0 comments
22.
Embarrassingly parallel batch jobs on multiple clouds or regions(skypilot.readthedocs.io)
1 points by covi 2y ago|0 comments
23.
by covi 2y ago
If you want to use your own GPUs or cloud accounts but with a great dev experience, see SkyPilot.
24.
Llama 3.1 finetuning on your own data and infra, without vendor lock-in(blog.skypilot.co)
2 points by covi 2y ago|0 comments
25.
Guide: Finetune Llama 3.1 on your infra(github.com)
5 points by covi 2y ago|0 comments
26.
AI on Kubernetes Without the Pain(blog.skypilot.co)
2 points by covi 2y ago|0 comments
27.
Dagster and SkyPilot: Orchestrating LLM training cost-effectively(dagster.io)
3 points by covi 2y ago|0 comments
28.
Inflection CEO Left, Became CEO of Microsoft AI(twitter.com)
5 points by covi 2y ago|0 comments
29.
SkyServe: A cost-efficient, multi-region/cloud library for serving GenAI models(blog.skypilot.co)
4 points by covi 3y ago|0 comments
30.
SGLang: Fast and Expressive LLM Inference with RadixAttention for 5x Throughput(github.com)
2 points by covi 3y ago|0 comments

More ›