Run 35B MoE models in ~2.5 GB of RAM. Weights stream from SSD; a trained router predicts loads a step ahead. Mac, iPhone, Android, Windows supported
Python
README: quality section wording, OpenCompass attribution
linyubupa8c7a7feab516
diff --git a/README.md b/README.mdindex 097df68..70f8c61 100644--- a/README.md+++ b/README.md@@ -207,13 +207,13 @@ this exact path).- Both adapters are required for the prerouter + LoRA pipeline; if afile is missing, `edge0` fails with a clear message (or pass`--no-prerouter` / `--no-lora` to run the plain base model).-## Quality (self-evaluation)+## Quality-Internal self-evaluation, included only to quantify the loss of the edge0-pipeline (int4 + trained adapters + prerouter routing) relative to the-original fp16 base models — the loss is small: **3.9 points on average for-edge0-35b, 2.8 for edge0-8b** (MMLU-Pro is even above the base). All scores-self-run, max 100:+All benchmarks were run by us with [OpenCompass](https://github.com/open-compass/opencompass)+under identical settings and parameters for both the edge0 models (int4 ++trained adapters + prerouter routing) and the original fp16 base models.+The loss of the edge0 pipeline is small: **3.9 points on average for+edge0-35b, 2.8 for edge0-8b** (MMLU-Pro is even above the base). Max 100:| Benchmark | edge0-35b (int4) | Qwen3.5-MoE 35B-A3B (fp16) | edge0-8b (int4) | Ling 3.0 tiny (fp16) ||---|---:|---:|---:|---:|diff --git a/README_zh.md b/README_zh.mdindex 14b567a..fd74cf4 100644--- a/README_zh.md+++ b/README_zh.md@@ -173,21 +173,12 @@ engine.close() # 释放 mmap / 专家缓存- prerouter + LoRA 两条适配器都必需;缺文件时 `edge0` 会给出明确报错(也可加 `--no-prerouter` / `--no-lora` 直接跑裸基模)。-## 质量自测--内部自测分数,仅用于对比 edge0 管线(int4 量化 + 训练适配器 + prerouter-路由)相对原 fp16 基座模型的**损失——损失很小:edge0-35b 平均仅落后-3.9 分、edge0-8b 落后 2.8 分**(MMLU-Pro 甚至反超基座)。均为自测,-满分 100:--| 评测项 | edge0-35b (int4) | Qwen3.5-MoE 35B-A3B(fp16) | edge0-8b (int4) | Ling 3.0 tiny(fp16) |-|---|---:|---:|---:|---:|-| AIME 2026 | 86.6 | 92.7 | 63.3 | 73.3 |-| HumanEval | 90.9 | 95.1 | 91.5 | 92.7 |-| GPQA-Diamond | 79.8 | 81.8 | 70.7 | 71.2 |-| MMLU-Pro | 81.0 | 84.6 | 70.1 | 65.8 |-| IFBench | 57.9 | 61.7 | 53.9 | 60.6 |-| **平均** | **79.2** | **83.2** | **69.9** | **72.7** |+## 质量++全部评测由我们使用 [OpenCompass](https://github.com/open-compass/opencompass)、+在完全相同的设置与参数下对 edge0 模型(int4 + 训练适配器 + prerouter 路由)+与原 fp16 基座模型测得。edge0 管线的损失很小:**edge0-35b 平均仅落后+3.9 分、edge0-8b 落后 2.8 分**(MMLU-Pro 甚至反超基座)。满分 100:## 性能实测