yavin5 on Nostr: It actually depends on what you're trying to do with it. Llama 3.1 is newer than ...
It actually depends on what you're trying to do with it. Llama 3.1 is newer than Mixtral.. but Llama 3.1 has wokeness built into it. Mixtral is less woke, and highly capable, but generally will perform lower on logic and reasoning than the newer Llama 3.1. Claude Haiku is nice but too small, and not open source. GPT models perform well but are both somewhat woke plus closed source.
Published at
2024-09-02 02:08:57Event JSON
{
"id": "9b01262be4aabcd601f42a295657c63fa97eebd711bcb4ffec9c8b9182147113",
"pubkey": "e93524a304c1a00e0cd4c274d2f53b81fe1b90aa769c6c53c416180316ddf2da",
"created_at": 1725242937,
"kind": 1,
"tags": [
[
"e",
"13c660c56ec11b1b4138c22bf4cc822c2bc96f4644c9b78ddd56aae3ce4c1d2a",
"",
"root"
],
[
"p",
"3e6e0735b8a2e96f8cf663f64d04bb9cea931afcc57f5427c35bb6859e95c8a2"
]
],
"content": "It actually depends on what you're trying to do with it. Llama 3.1 is newer than Mixtral.. but Llama 3.1 has wokeness built into it. Mixtral is less woke, and highly capable, but generally will perform lower on logic and reasoning than the newer Llama 3.1. Claude Haiku is nice but too small, and not open source. GPT models perform well but are both somewhat woke plus closed source.",
"sig": "29f42bde8758577e8820decc736301408b8a75787f63d3ed74aa7e062ad0564be73bb6958411bd96e18a3809b38ce9d564bd06b816ea271c69d7e61e9ba7d496"
}
