Update README.md
Browse files
README.md
CHANGED
|
@@ -2,4 +2,18 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model:
|
| 4 |
- mistralai/Devstral-Small-2507
|
| 5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model:
|
| 4 |
- mistralai/Devstral-Small-2507
|
| 5 |
+
language:
|
| 6 |
+
- fr
|
| 7 |
+
- en
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
This model is a RTN/W8A8-FP8 quantization of the original MistralAI/Devstral-small-2507 24B model.
|
| 11 |
+
We reduced the amount of shards from 10 to 6 making it practically usable on a single GPU node.
|
| 12 |
+
|
| 13 |
+
We tested it with the following consumer grade hardware.
|
| 14 |
+
|
| 15 |
+
Test machine:
|
| 16 |
+
CPU: AMD Ryzen 7 7700X 8-Core
|
| 17 |
+
GPU: AMD Radeon RX9070XT 16Gb
|
| 18 |
+
RAM: 64Gb 2x32Gb DDR5-5200
|
| 19 |
+
HDD: 2x 1Tb M.2 NVMe PCIe 5.0
|