Unverified80% confidenceFactTime unknown
LLaMA是Meta开源的大语言模型系列，以较小的参数量实现了接近GPT-3.5的性能
1
Sources
80%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Hugging Face Transformers：16万Star背后的技术架构与生态解析
githubhuggingface
Related Entities
Related Claims
UnverifiedLLaMA是Meta开源的大语言模型系列，以较小的参数量实现了接近GPT-3.5的性能100% similarUnverifiedLLaMA是Meta开源的大语言模型系列，以较小的参数量实现了接近GPT-3.5的性能100% similarUnverifiedLlama is a series of open-source large language models from Meta; its latest version has approached GPT-4-level performance on multiple benchmarks and is completely free.81% similarUnverifiedLM Studio and GPT4All are competitors of Ollama in the local large model tooling space76% similarUnverifiedLlama 3提供了8B和70B两个版本，在英文通用任务上的表现已经接近GPT-4级别的闭源模型74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/31774API
curl https://kongchang.com/api/v1/knowledge/claims/31774MCP
get_claim(id=31774)