{
"type": "SET",
"op_list": [
{
"type": "SET_VALUE",
"ref": "/apps/knowledge/explorations/0x00ADEc28B6a845a085e03591bE7550dd68673C1C/ai|transformers|decoder-only/-OloeVZ5X8lZT12cwHOp",
"value": {
"topic_path": "ai/transformers/decoder-only",
"title": "Mistral 7B",
"content": "# Mistral 7B (2023)\n\n## Authors\nJiang, Sablayrolles, Mensch, Bamford, Chaplot, Casas, Bressand, Lengyel, Lample, et al.\n\n## Paper\nhttps://arxiv.org/abs/2310.06825\n\n## Code\nhttps://github.com/mistralai/mistral-inference\n\n## Key Concepts\n- Sliding window attention\n- Grouped-query attention (GQA)\n- Rolling buffer KV cache\n\n## Builds On\n- LLaMA: Open and Efficient Foundation Language Models\n\n## Summary\nA 7B-parameter model that outperforms LLaMA 2 13B on all benchmarks through grouped-query attention and sliding-window attention, advancing the efficiency frontier for open models.",
"summary": "A 7B-parameter model that outperforms LLaMA 2 13B on all benchmarks through grouped-query attention and sliding-window attention, advancing the efficiency frontier for open models.",
"depth": 1,
"tags": "decoder-only,autoregressive,sliding-window-attention,grouped-query-attention,builds-on:llama",
"price": null,
"gateway_url": null,
"content_hash": null,
"created_at": 1771483826438,
"updated_at": 1771483826438
}
},
{
"type": "SET_VALUE",
"ref": "/apps/knowledge/index/by_topic/ai|transformers|decoder-only/explorers/0x00ADEc28B6a845a085e03591bE7550dd68673C1C",
"value": 6
},
{
"type": "SET_VALUE",
"ref": "/apps/knowledge/graph/nodes/0x00ADEc28B6a845a085e03591bE7550dd68673C1C_ai|transformers|decoder-only_-OloeVZ5X8lZT12cwHOp",
"value": {
"address": "0x00ADEc28B6a845a085e03591bE7550dd68673C1C",
"topic_path": "ai/transformers/decoder-only",
"entry_id": "-OloeVZ5X8lZT12cwHOp",
"title": "Mistral 7B",
"depth": 1,
"created_at": 1771483826438
}
}
]
}