marcobambini

6 ms·

1.
Run DeepSeek v4.1 Flash on your Mac(github.com)
2 points by marcobambini 23h ago|1 comments
2.
Show HN: Warp – Run DeepSeek v4.1 Flash with 5 GB of RAM at 3.77 tok/s(github.com)
13 points by marcobambini 2d ago|4 comments
3.
Show HN: SQLite-Vector – Vector search for SQLite, now Apache 2.0(github.com)
2 points by marcobambini 6d ago|0 comments
4.
SQLite-Vector is now licensed under Apache 2.0(github.com)
9 points by marcobambini 7d ago|0 comments
5.
GLM-5.3-Flash on Apple Silicon(github.com)
2 points by marcobambini 20d ago|0 comments
6.
GLM-5.3-Flash at 3.3 tok/s
2 points by marcobambini 21d ago|0 comments
7.
Show HN: Warp – Run the 313B GLM-5.3-Flash on a MacBook with 8GB RAM
3 points by marcobambini 21d ago|0 comments
8.
by marcobambini 1mo ago
Developing a programming language from scratch was a project that literally drained every ounce of energy I had for years. Every moment of every day, of every month, was devoted to thinking about how to make Gravity faster, safer, more effi
9.
I once wrote a programming language by hand and now I feel like a dinosaur(marcobambini.substack.com)
3 points by marcobambini 1mo ago|1 comments
10.
by marcobambini 2mo ago
Thanks guys for all the comments, I am going to rewrite the README (without using an LLM)
11.
by marcobambini 2mo ago
We are backed by the SQLite author, and we have the right to use the SQLite name. Please do your homework before writing such comments. We used an Elastic license for some projects to protect our work from being used in SASS without a prior
12.
by marcobambini 2mo ago
I wrote tons of software, even a programming language by hand https://github.com/marcobambini/gravity . I'm using my skills to orchestrate LLMs and agents, and I can write better code much faster. As developers, we
13.
by marcobambini 2mo ago
It is not yet available, the only way is to download the official Kimi K3 model and then convert it: # 1. preflight: reachable? how big? does it fit? tools/fetch_weights.sh --dest /Volumes/staging/k3 --dry-run # 2. downl
14.
by marcobambini 2mo ago
We tried to use Metal, but for that specific project it was slower than just using NEON ARM optimizations. It is all documented in the docs.
15.
Run Kimi K3 using 29 GB of RAM at 0.50 tok/s(github.com)
341 points by marcobambini 2mo ago|166 comments
16.
Running Kimi K3 on a local computer(github.com)
2 points by marcobambini 2mo ago|0 comments
17.
by marcobambini 2mo ago
Right, we trade speed for the ability to run a 2.7T-parameter model while preserving accuracy. It is a first version, and we plan to improve the inference performance.
18.
A new inference engine to run Kimi K3 2.78T parameter with 29GB of RAM(marcobambini.substack.com)
4 points by marcobambini 2mo ago|2 comments
19.
Show HN: Run Full Kimi K3 with 29 GB of RAM(github.com)
9 points by marcobambini 2mo ago|2 comments
20.
by marcobambini 2mo ago
M5 Pro
21.
Show HN: A new engine to run Kimi K3 on a laptop
7 points by marcobambini 2mo ago|3 comments
22.
Run Kimi K3 on a local computer(github.com)
8 points by marcobambini 2mo ago|0 comments
23.
Rethinking Databases for Humans and AI Agents(marcobambini.substack.com)
3 points by marcobambini 2mo ago|0 comments
24.
Measuring and Reducing CPU Usage in SQLite(sqlite.org)
2 points by marcobambini 2mo ago|0 comments
25.
SQLite Databases for Humans and Agents(marcobambini.substack.com)
3 points by marcobambini 3mo ago|0 comments
26.
Rethinking Databases for Humans and AI Agents(marcobambini.substack.com)
2 points by marcobambini 3mo ago|0 comments
27.
The Cloud vs. Edge Debate Is Over(marcobambini.substack.com)
4 points by marcobambini 4mo ago|1 comments
28.
Block-Level CRDT: The Missing Piece for Collaborative AI Agent Memory(marcobambini.substack.com)
2 points by marcobambini 4mo ago|0 comments
29.
The Cloud vs. Edge Debate Is Over(marcobambini.substack.com)
2 points by marcobambini 4mo ago|1 comments
30.
SQLite-Vector now with Google TurboQuant for a 38x speedup(github.com)
3 points by marcobambini 4mo ago|0 comments

More ›