Commit
·
54d821a
1
Parent(s):
a192d29
"Update README.md"
Browse files
README.md
CHANGED
|
@@ -21,10 +21,11 @@ I'm constantly enhancing these model descriptions to provide you with the most r
|
|
| 21 |
- Original model: [slimorca-stablelm-3b-4e1t](https://huggingface.co/pansophic/slimorca-stablelm-3b-4e1t)
|
| 22 |
|
| 23 |
# StableLM
|
| 24 |
-
This is a Model based on StableLM.
|
|
|
|
| 25 |
|
| 26 |
## Note:
|
| 27 |
-
Current (as of
|
| 28 |
The model will crash immediately if -ngl is larger than 34.
|
| 29 |
The model works fine however without any gpu acceleration.
|
| 30 |
|
|
|
|
| 21 |
- Original model: [slimorca-stablelm-3b-4e1t](https://huggingface.co/pansophic/slimorca-stablelm-3b-4e1t)
|
| 22 |
|
| 23 |
# StableLM
|
| 24 |
+
This is a Model based on StableLM.
|
| 25 |
+
Stablelm is a familiy of Language Models by Stability AI.
|
| 26 |
|
| 27 |
## Note:
|
| 28 |
+
Current (as of 2023-11-15) implementations of Llama.cpp only support GPU offloading up to 34 Layers with these StableLM Models.
|
| 29 |
The model will crash immediately if -ngl is larger than 34.
|
| 30 |
The model works fine however without any gpu acceleration.
|
| 31 |
|