PulseAugur
中
实时 18:48:20
English(EN) PLaMo 3 Translate 0.8B Gets Down to About 423MB of Weights

PLaMo 3 Translate 0.8B 模型通过混合量化缩小至 423MB

PLaMo 3 Translate 0.8B 模型的新混合量化版本已发布,将其权重大小减少到约 423MB。该版本使用 mlx-lm 0.32.2 转换,为特定层分配 4 位量化,其余层使用 3 位量化,平均每权重 3.804 位。虽然与统一 4 位版本相比,该模型的翻译质量指标略有下降,但在 Mac mini M2 Pro 上仍保持约 352 tokens/s 的解码速度和约 565MB 的峰值内存使用量。 AI

影响 使资源受限的设备能够运行更强大的翻译模型。

排序理由 发布现有 LLM 的更小、量化版本,并附带性能指标。[lever_c_demoted from research: ic=1 ai=1.0]

在 dev.to — LLM tag 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

PLaMo 3 Translate 0.8B 模型通过混合量化缩小至 423MB

本文如何被排名

Signal score
25 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
发布现有 LLM 的更小、量化版本,并附带性能指标。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, infra
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. dev.to — LLM tag TIER_1 English(EN) · Noptium ·

    PLaMo 3 Translate 0.8B 的权重约为 423MB

    <p>An MLX version of PLaMo 3 Translate 0.8B, the smallest model in PFN's translation-focused LLM family, now has weights of only about 423MB. It is a mixed-quantization model called "mixed_3_4", published on Hugging Face by an individual converter.</p> <p>Unless noted otherwise, …