Commit
•
f8df579
1
Parent(s):
428a2d2
Added metrics
Browse files
README.md
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
tags:
|
| 4 |
+
- cohere
|
| 5 |
+
- conversational
|
| 6 |
+
- 10languages
|
| 7 |
+
- text-generation-inference
|
| 8 |
+
- Inference Endpoints
|
| 9 |
+
---
|
| 10 |
+
Official AQLM quantization of `CohereForAI/c4ai-command-r-v01`.
|
| 11 |
+
|
| 12 |
+
For this quantization, we used 1 codebook of 16 bits.
|
| 13 |
+
|
| 14 |
+
Results:
|
| 15 |
+
| Model | Quantization | MMLU | GSM8k | Model size, Gb |
|
| 16 |
+
|------|------|-------|------|------|
|
| 17 |
+
|CohereForAI/c4ai-command-r-v01| None |0.6755 | 0.6065 | 70.0 |
|
| 18 |
+
| | 1x16 | 0.5719 | 0.3760 | 12.7 |
|