khimaros

24 ms·

61.
by khimaros 7mo ago
this isn't quite right. the blobs are produced by GrapheneOS and are reproducible once the source code embargo lifts.
62.
MiniMax-M2.5: How to Run Guide(unsloth.ai)
3 points by khimaros 8mo ago|0 comments
63.
by khimaros 8mo ago
maxed out Framework Desktop
64.
by khimaros 8mo ago
allocation is irrelevant. as an owner of one of these you can absolutely use the full 128GB (minus OS overhead) for inference workloads
65.
by khimaros 8mo ago
i meant m2.1, but you are probably talking about kimi, not minimax
66.
by khimaros 8mo ago
i have been working on something similar, trying to build the leanest agent loop that can be self modifying. ended up building it as a plugin within OpenCode with the cow pulled out into python hooks that the agent can modify at runtime (wi
67.
by khimaros 8mo ago
minimax-m.2 is close
68.
by khimaros 8mo ago
https://github.com/Chainlift/liftkit
69.
by khimaros 8mo ago
this seems to be for custom design services. IANAL but the libraries and design language seems to be open source and free to use.
70.
by khimaros 8mo ago
what will we keep our ready power tucked in our vests under now?
71.
by khimaros 8mo ago
does this support oauth tokens for making use of Claude or Gemini subscriptions?
72.
by khimaros 8mo ago
been waiting for this for a while https://github.com/transmission/transmission/issues/47 https://github.com/arvidn/libtorrent/issues/7283
73.
by khimaros 8mo ago
reminds me a bit if ZeroNet, which still has a maintained fork somewhere out there https://github.com/zeronet-conservancy/zeronet-conservancy/
74.
by khimaros 8mo ago
unsloth quants are up https://huggingface.co/unsloth/Trinity-Large-Preview-GGUF
75.
by khimaros 8mo ago
llama.cpp
76.
by khimaros 8mo ago
this is not open source
77.
by khimaros 8mo ago
"SERA-32B is the first model in Ai2's Open Coding Agents series. It is a state-of-the-art open-source coding agent that achieves 49.5% on SWE-bench Verified, matching the performance of much larger models like Devstral-Small-2 (24
78.
by khimaros 8mo ago
it's great to see this kind of progress in reproducible weights, but color me confused. this claims to be better and smaller than Devstral-Small-2-24B, while clocking in at 32B (larger) and scoring more poorly?
79.
by khimaros 8mo ago
Minimax m2.1 rivals GLM 4.7 and fits in 128GB with 100k context at 3bit quantization.
80.
by khimaros 9mo ago
i made an epub to audiobook generator using this with optional LLM integration for dramatized output: https://github.com/khimaros/autiobook -- also submitted here: https://news.ycombinator.com/item?id=4
81.
Convert ePub files to audiobooks using qwen3-TTS(github.com)
3 points by khimaros 9mo ago|1 comments
82.
by khimaros 9mo ago
there are two generation modes: 1) a basic single voice narrator and 2) a "dramatization" mode which generates a cast list from the book contents and converts the entire text into a script with distinct voices and instructions for
83.
by khimaros 9mo ago
this is like an open weights version of DeepMind's Genie
84.
by khimaros 9mo ago
Terminal Bench 2.0
85.
by khimaros 9mo ago
has been merged
86.
by khimaros 9mo ago
there is a pretty compatible rust implementation as well which claims to target microcontrollers https://github.com/BeechatNetworkSystemsLtd/Reticulum-rs
87.
by khimaros 9mo ago
https://github.com/leejet/stable-diffusion.cpp
88.
UML based testing of nftables firewalls(github.com)
2 points by khimaros 9mo ago|1 comments
89.
by khimaros 9mo ago
i built this to scratch an itch i had. i wanted a way to assert the behavior of a complex nftables ruleset and verify verdict and forwarding behavior in a high fidelity environment. there is a large compliance testing suite in the testdata&
90.
by khimaros 9mo ago
seems like a promising alternative to obsidian, but missing [[wikilinks]] and back references

More ›