Update README.md
Browse files
README.md
CHANGED
|
@@ -72,4 +72,20 @@ Spring is a cascading dare-ties merge of the following models:
|
|
| 72 |
]
|
| 73 |
```
|
| 74 |
|
| 75 |
-
I'm finding my iq4_xs to be working well. Llama 3 instruct format works well, but minimal format is highly creative.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
]
|
| 73 |
```
|
| 74 |
|
| 75 |
+
I'm finding my iq4_xs to be working well. Llama 3 instruct format works well, but minimal format is highly creative.
|
| 76 |
+
|
| 77 |
+
## Scores
|
| 78 |
+
|
| 79 |
+
Not greater than the sum of it's parts, based on the scores; but it is really smart for an emotive RP model.
|
| 80 |
+
|
| 81 |
+
Metric | Score
|
| 82 |
+
---|---
|
| 83 |
+
Average | 65.89
|
| 84 |
+
ARC | 63.05
|
| 85 |
+
HellaSwag | 82.49
|
| 86 |
+
MMLU | 64.45
|
| 87 |
+
TruthfulQA | 51.63
|
| 88 |
+
Winogrande | 76.24
|
| 89 |
+
GSM8K | 51.63
|
| 90 |
+
|
| 91 |
+
[Details](https://huggingface.co/datasets/open-llm-leaderboard/details_maldv__spring-chicken-8x8b)
|