melvincarvalho on Nostr: Maverick and Goose! Llama 4 Maverick: - 17B active parameters, 128 experts, 400B ...
Maverick and Goose!
Llama 4 Maverick:
- 17B active parameters, 128 experts, 400B total.
- 1M token context window.
- Not single-GPU; runs on one H100 DGX host or can be distributed for greater efficiency.
- Outperforms GPT-4o and Gemini 2.0 Flash on coding, reasoning, and multilingual tests at a competitive cost.
- Maintains strong image understanding and grounded reasoning ability.
Published at
2025-04-05 19:59:40 GMTEvent JSON
{
"id": "71998177260e437c69c23c00f2adb6d1bf5c5350015f5d8580f7154d31073a14",
"pubkey": "de7ecd1e2976a6adb2ffa5f4db81a7d812c8bb6698aa00dcf1e76adb55efd645",
"created_at": 1743883180,
"kind": 1,
"tags": [
[
"client",
"Ditto",
"31990:15b68d319a088a9b0c6853d2232aff0d69c8c58f0dccceabfb9a82bd4fd19c58:ditto",
"wss://ditto.pub/relay"
]
],
"content": "Maverick and Goose!\n\nLlama 4 Maverick:\n - 17B active parameters, 128 experts, 400B total.\n - 1M token context window.\n - Not single-GPU; runs on one H100 DGX host or can be distributed for greater efficiency.\n - Outperforms GPT-4o and Gemini 2.0 Flash on coding, reasoning, and multilingual tests at a competitive cost.\n - Maintains strong image understanding and grounded reasoning ability.",
"sig": "02fbafada3e28ff4f66399a80c1f388756c7a9f60b76d8e20be225af96973196ba20011bcd28cd02c9916408f420760752494fb249c0fcc9812bde22bf8f838b"
}
