TUBii on Nostr: Running #deepseek r1 llama 8B model right out from the vram. Quick as hell and giving ...
Running #deepseek r1 llama 8B model right out from the vram. Quick as hell and giving much better results than any other model of this size.
Not the best for lexical knowledge but as a coding helper it is flawless. #ai #aicoding #llm
Published at
2025-02-06 08:03:23Event JSON
{
"id": "6e4c91d19191a7cc2a3f062ccad0c317a864af35a628de728524a4466b6f7c86",
"pubkey": "09ed4307acbb03cdef4d733300ec3b59dd12582f57eae919ca9e2c7998347f63",
"created_at": 1738829003,
"kind": 1,
"tags": [
[
"t",
"deepseek"
],
[
"t",
"ai"
],
[
"t",
"aicoding"
],
[
"t",
"llm"
]
],
"content": "Running #deepseek r1 llama 8B model right out from the vram. Quick as hell and giving much better results than any other model of this size. \nNot the best for lexical knowledge but as a coding helper it is flawless. #ai #aicoding #llm",
"sig": "5a6373e4ca74c46ace355277de529a9067dc6f33d84a423b8e37daaf09fb0566f4656c7fe3f67e7f68466cabbf797b97ff578876cd4f7c80b6aa52e64bdb4a36"
}
