Update README.md
Browse files
README.md
CHANGED
|
@@ -4,4 +4,11 @@ license: cc-by-nc-4.0
|
|
| 4 |
|
| 5 |
# Command R+ GGUF
|
| 6 |
|
|
|
|
| 7 |
This repository contains experimental GGUF weights that are currently compatible only with the following fork: [https://github.com/Noeda/llama.cpp/tree/53f71f0026cbed4588b2ad16c51db630d2745794](https://github.com/Noeda/llama.cpp/tree/53f71f0026cbed4588b2ad16c51db630d2745794). I will update them once support for Command R+ is merged into the llama.cpp repository
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
|
| 5 |
# Command R+ GGUF
|
| 6 |
|
| 7 |
+
## Description
|
| 8 |
This repository contains experimental GGUF weights that are currently compatible only with the following fork: [https://github.com/Noeda/llama.cpp/tree/53f71f0026cbed4588b2ad16c51db630d2745794](https://github.com/Noeda/llama.cpp/tree/53f71f0026cbed4588b2ad16c51db630d2745794). I will update them once support for Command R+ is merged into the llama.cpp repository
|
| 9 |
+
|
| 10 |
+
## Concatenating Weights
|
| 11 |
+
For every variant (except Q2_K), you must concatenate the weights, as they exceed the 50 GB single file size limit on HuggingFace. You can accomplish this using the `cat` command on Linux (example for the Q3 variant):
|
| 12 |
+
```bash
|
| 13 |
+
cat command-r-plus-Q3_K_L-0000* > command-r-plus-Q3_K_L.gguf
|
| 14 |
+
```
|