Unverified50% confidenceFactExact time
LLaMA-2 70B 总参数约 700 亿均匀分布于 80 层，每层约 8.75 亿参数，FP16 精度下单层约占 1.75GB 显存
1
Sources
50%
Confidence
Long-term
Relevance
7/19/2026
First Seen
Sources
Related Claims
UnverifiedLLaMA-2 70B 模型共有 80 个 Transformer 层，每层约 8.75 亿参数，FP16 精度下单层约占 1.75GB 显存83% similarUnverifiedRunning an open-source model at the 70-billion-parameter scale requires at least approximately 140GB of VRAM in FP16 precision, i.e., 8 H100s in parallel.77% similarPartially VerifiedTaking the 65B-parameter LLaMA model as an example, traditional full fine-tuning requires more than 780GB of VRAM, whereas QLoRA requires only approximately 48GB.77% similarVerifiedTaking a 70B parameter model with FP16 precision and 32 attention head layers as an example, the KV Cache for a single 200K-length request requires approximately 80–120 GB of VRAM, far exceeding the 80 GB VRAM limit of a single A100 GPU.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/563762API
curl https://kongchang.com/api/v1/knowledge/claims/563762MCP
get_claim(id=563762)