概述
Its fitted results favor scaling both together in the evaluated regime; that objective is different from minimizing total deployment cost or guaranteeing one token ratio for every dataset.
深入探讨
The 2022 study fitted loss behavior across model and data budgets and estimated training-compute-optimal parameter and token allocations. The finding that both quantities should grow in similar proportions means comparable scaling exponents, not equal numerical counts. Chinchilla used about seventy billion parameters and 1.4 trillion training tokens, giving a twenty-token-per-parameter example. It was compared with the much larger Gopher under a comparable training-compute budget. The study argued that many evaluated contemporary large models were undertrained relative to the estimated optimum. These relationships are fitted results that depend on the data and setup; a fixed ratio is not a universal design law. LLaMA's paper discusses training smaller models longer as a way to consider inference efficiency, a different objective from training compute alone.
现实世界的实施
Comparing the 70-billion-parameter Chinchilla experiment with the 280-billion-parameter Gopher baseline.
Using fitted training-budget relationships as a starting point rather than a universal ratio rule.
Distinguishing a from-scratch training optimum from a smaller model chosen for repeated inference.
Evaluating whether parameters or data provide more value under a defined compute and quality objective.
龙猫缩放法则的未来
Scaling research continues to examine data quality, repeated data, model architecture, and the cost of training versus serving. State the objective and evaluation conditions before calling an allocation optimal. A useful empirical trend is different from a guarantee that one parameter-token recipe dominates every workload.
付诸实践
为什么这很重要
- 速度与规模. 语言工作流程可以在不牺牲一致性的情况下更快地移动。
- 交通与覆盖范围. 它扩展了跨语言和沟通方式的访问。
- 更清晰的判决. 团队可以花更多时间进行判断，而自动化则可以处理重复。
留意
- 幻觉的事实可以悄悄地进入报告、支持流程或研究成果。
- 及时的敏感性可能会在类似的请求中产生不一致的结果。
- 如果访问控制薄弱，敏感文本数据可能会暴露。
后续步骤
- 在推出之前定义输出格式、语气和质量标准。
- 当准确性很重要时，请使用可信来源进行地面响应。
- 为高风险输出保留人工审查检查点。
- 跟踪故障模式并定期重新训练提示或工作流程。
不断探索
常见问题
What is Chinchilla Scaling Laws?
Chinchilla studied how to allocate a training-compute budget between parameters and training tokens. Its fitted results favor scaling both together in the evaluated regime; that objective is different from minimizing total deployment cost or guaranteeing one token ratio for every dataset.
What does Chinchilla's finding that parameters and tokens should scale in similar proportions mean?
Similar proportional growth is not a literal one-token-per-parameter rule.
What token-per-parameter ratio does the 70B-parameter, 1.4T-token Chinchilla experiment illustrate?
1.4 trillion divided by seventy billion equals twenty; this example is not a universal optimum for every setup.
尽管龙猫体型小得多，但哪种型号的性能优于哪款？
70B 参数的 Chinchilla 使用相同的计算击败了 DeepMind 自己的 280B 参数的 Gopher，因为它训练的数据要多得多。
Chinchilla 当时对 GPT-3 这样的模型意味着什么？
那个时代的许多大型模型训练不足，这意味着如果参数计数有更多数据，它们会表现得更好。
Chinchilla 分析中的计算大致是如何进行的？
训练计算 (FLOP) 大约与参数数量乘以训练令牌数量成正比。