Spaces:
Configuration error
Configuration error
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
<p align="center">
|
| 2 |
<h1 align="center">Atlas Inference Engine</h1>
|
| 3 |
<p align="center">
|
|
@@ -26,9 +32,9 @@
|
|
| 26 |
|
| 27 |
---
|
| 28 |
|
| 29 |
-
## ⚡ What is Atlas?
|
| 30 |
|
| 31 |
-
Atlas is a high-performance, pure Rust & CUDA LLM inference engine purpose-built for prosumer workstations (NVIDIA DGX Spark / GB10 SM121 and AMD Strix Halo). No Python, no PyTorch, no bloated dependency trees—just one compact binary with hand-tuned micro-kernels.
|
| 32 |
|
| 33 |
- **Sub-90s First Token**: Boots in seconds with cached weights; zero JIT compile or Python startup lag.
|
| 34 |
- **Default Flagship Qwen 3.8 27B**: Dense hybrid GDN + Attention running at 23.59 tok/s single-stream with MTP speculative decoding on a single GB10.
|
|
@@ -196,4 +202,4 @@ curl http://localhost:8888/v1/chat/completions \
|
|
| 196 |
- **Community Edition**: Licensed under **AGPLv3**. Free and open for personal use, research, and non-commercial local deployments.
|
| 197 |
- **Enterprise Edition**: Commercial licensing for proprietary applications, SaaS hosting without AGPLv3 copyleft obligations, dedicated support, and custom hardware/kernel porting. Contact `debaterishaqui@gmail.com`.
|
| 198 |
|
| 199 |
-
<sub><b>Continuity notice.</b> Atlas is continuing. This repository, the <a href="https://github.com/Atlas-Inf">Atlas-Inf</a> GitHub organization, and <a href="https://atlasinference.dev">atlasinference.dev</a> are the replacement official Atlas channels. The existing website and GitHub repository remain disputed Atlas assets that have not been relinquished.</sub>
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: agpl-3.0
|
| 3 |
+
thumbnail: >-
|
| 4 |
+
https://cdn-uploads.huggingface.co/production/uploads/655a1f6bf6103195fd8f5633/9OTkwltwOc0QP4_7S03Rf.png
|
| 5 |
+
short_description: Atlas Inference is a pure Rust & ROCm/CUDA LLM inference eng
|
| 6 |
+
---
|
| 7 |
<p align="center">
|
| 8 |
<h1 align="center">Atlas Inference Engine</h1>
|
| 9 |
<p align="center">
|
|
|
|
| 32 |
|
| 33 |
---
|
| 34 |
|
| 35 |
+
## ⚡ What is Atlas Inference?
|
| 36 |
|
| 37 |
+
Atlas Inference is a high-performance, pure Rust & CUDA LLM inference engine purpose-built for prosumer workstations (NVIDIA DGX Spark / GB10 SM121 and AMD Strix Halo). No Python, no PyTorch, no bloated dependency trees—just one compact binary with hand-tuned micro-kernels.
|
| 38 |
|
| 39 |
- **Sub-90s First Token**: Boots in seconds with cached weights; zero JIT compile or Python startup lag.
|
| 40 |
- **Default Flagship Qwen 3.8 27B**: Dense hybrid GDN + Attention running at 23.59 tok/s single-stream with MTP speculative decoding on a single GB10.
|
|
|
|
| 202 |
- **Community Edition**: Licensed under **AGPLv3**. Free and open for personal use, research, and non-commercial local deployments.
|
| 203 |
- **Enterprise Edition**: Commercial licensing for proprietary applications, SaaS hosting without AGPLv3 copyleft obligations, dedicated support, and custom hardware/kernel porting. Contact `debaterishaqui@gmail.com`.
|
| 204 |
|
| 205 |
+
<sub><b>Continuity notice.</b> Atlas is continuing. This repository, the <a href="https://github.com/Atlas-Inf">Atlas-Inf</a> GitHub organization, and <a href="https://atlasinference.dev">atlasinference.dev</a> are the replacement official Atlas channels. The existing website and GitHub repository remain disputed Atlas assets that have not been relinquished.</sub>
|