abiboe/swarmlm

Confidential sharded inference on commodity hardware about 1 hour ago

help wanted research

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

Warm-set residency policy for idle nodes about 1 hour ago

help wanted

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

OpenTimestamps anchor backend about 1 hour ago

help wanted

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

Signed announces (Ed25519) to replace address-based trust about 1 hour ago

help wanted

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

Add more model presets about 1 hour ago

good first issue

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

RTT-aware peer selection about 1 hour ago

good first issue

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

Measure the heterogeneous-hardware noise floor (sigma_hw) about 1 hour ago

good first issue research

abiboe/swarmlm

0

Run big LLMs across many small computers. Pool the machines you already own — no data-center GPU, no vendor lock-in.

Python

#decentralized#distributed-inference#inference#llama-cpp#llm#local-llm#p2p#python#self-hosted

Open Issues Need Help