Upload README.md with huggingface_hub

README.md

@@ -30,7 +30,7 @@ pipeline_tag: text-generation
 </div>
 
 
-**Edge0-35b-a3b** — a 35B MoE LLM that runs at viable speed on portable devices in under **2.9 GiB of active memory** (1/8 of its 23 GB weight footprint), 
+**Edge0-35b-a3b** — a 35B MoE LLM that runs at viable speed on a phone in under **3 GiB of active memory** (1/8 of its 23 GB weight footprint), 
 via the [edge0](https://github.com/Edge0-AI/edge0) streaming inference framework.
 
 > **Preview status:** this is an early preview release of the edge0
@@ -41,10 +41,10 @@ via the [edge0](https://github.com/Edge0-AI/edge0) streaming inference framework
 
 - **Runs in phone-class memory**: the full 4-bit checkpoint stays on
   storage and experts are streamed on demand, so only the active
-  weights are in RAM — under **2.9 GiB**, with no sharding and no
+  weights are in RAM — under **3 GiB**, with no sharding and no
   upfront download of the weights into memory.
-- **Fast enough for interactive use**: 14.9–17.7 tok/s decode on a
-  Mac mini M4 Pro; long prompts fill in at 113–140 tok/s.
+- **Fast enough for interactive use**: 15 tok/s decode on a phone;
+  long prompts fill in at 140 tok/s.
 - **Quality kept after quantization**: Recover-LoRA distillation keeps
   the int4 model within **3.9 points** of its fp16 base.
 - **Works out of the box**: base, LoRA and prerouter adapters ship
@@ -126,7 +126,7 @@ SSD on demand and are not resident.
 - The MLX backend currently targets Apple Silicon; other backends are
   on the edge0 roadmap.
 - Long contexts grow the KV cache; use shorter contexts to keep peak
-  memory at 2.9 GiB.
+  memory at 3 GiB.
 
 ## Quick start