Add link to Neuron-optimized version
#8
by
badaoui
HF Staff
- opened
README.md
CHANGED
|
@@ -133,4 +133,16 @@ SmolLM2 models primarily understand and generate content in English. They can pr
|
|
| 133 |
primaryClass={cs.CL},
|
| 134 |
url={https://arxiv.org/abs/2502.02737},
|
| 135 |
}
|
| 136 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 133 |
primaryClass={cs.CL},
|
| 134 |
url={https://arxiv.org/abs/2502.02737},
|
| 135 |
}
|
| 136 |
+
```
|
| 137 |
+
|
| 138 |
+
---
|
| 139 |
+
## 🚀 AWS Neuron Optimized Version Available
|
| 140 |
+
|
| 141 |
+
A Neuron-optimized version of this model is available for improved performance on AWS Inferentia/Trainium instances:
|
| 142 |
+
|
| 143 |
+
**[badaoui/HuggingFaceTB-SmolLM2-1.7B-neuron](https://huggingface.co/badaoui/HuggingFaceTB-SmolLM2-1.7B-neuron)**
|
| 144 |
+
|
| 145 |
+
The Neuron-optimized version provides:
|
| 146 |
+
- Pre-compiled artifacts for faster loading
|
| 147 |
+
- Optimized performance on AWS Neuron devices
|
| 148 |
+
- Same model capabilities with improved inference speed
|