Unverified50% confidenceFactExact time
Meta's LLaMA 3 was trained on more than 15 trillion tokens.
1
Sources
50%
Confidence
Long-term
Relevance
9/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe training data volume for LLaMA 3 was significantly increased from LLaMA 2's 2 trillion tokens to more than 15 trillion tokens.77% similarVerifiedMeta's Llama 3 has up to 405 billion parameters.68% similarUnverifiedMeta的LLaMA系列通过开源迅速聚集了数万开发者，形成庞大的微调模型库与工具链60% similarUnverified训练使用约500万行英文维基百科语料约3亿token，10万上下文词表，23758个名词目标60% similarUnverifiedMeta的LLaMA模型仓库积累10万GitHub星用了约一年时间58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/866927API
curl https://kongchang.com/api/v1/knowledge/claims/866927MCP
get_claim(id=866927)