Update README.md
Browse files
README.md
CHANGED
|
@@ -1,38 +1,17 @@
|
|
| 1 |
---
|
| 2 |
-
base_model:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
tags:
|
| 4 |
-
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
---
|
| 8 |
-
# out_model
|
| 9 |
-
|
| 10 |
-
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
| 11 |
-
|
| 12 |
-
## Merge Details
|
| 13 |
-
### Merge Method
|
| 14 |
-
|
| 15 |
-
This model was merged using the SLERP merge method.
|
| 16 |
-
|
| 17 |
-
### Models Merged
|
| 18 |
-
|
| 19 |
-
The following models were included in the merge:
|
| 20 |
-
* model_end
|
| 21 |
-
* model4
|
| 22 |
-
|
| 23 |
-
### Configuration
|
| 24 |
-
|
| 25 |
-
The following YAML configuration was used to produce this model:
|
| 26 |
-
|
| 27 |
-
```yaml
|
| 28 |
-
models:
|
| 29 |
-
- model: model4
|
| 30 |
-
- model: model_end
|
| 31 |
-
merge_method: slerp
|
| 32 |
-
base_model: model4
|
| 33 |
-
parameters:
|
| 34 |
-
t:
|
| 35 |
-
- value: 0.4 # fallback for rest of tensors
|
| 36 |
-
dtype: bfloat16
|
| 37 |
-
|
| 38 |
-
```
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model:
|
| 3 |
+
- euclaise/Memphis-CoT-3B
|
| 4 |
+
license: cc-by-4.0
|
| 5 |
+
datasets:
|
| 6 |
+
- euclaise/TinyCoT
|
| 7 |
+
- euclaise/mathoverflow-accepted
|
| 8 |
+
- euclaise/reddit-instruct
|
| 9 |
+
- euclaise/WritingPrompts_curated
|
| 10 |
+
- sablo/oasst2_curated
|
| 11 |
+
- euclaise/mathqa_programs
|
| 12 |
+
- BEE-spoke-data/coedit-reworded-deduped
|
| 13 |
+
- pszemraj/booksum-short
|
| 14 |
+
library_name: transformers
|
| 15 |
tags:
|
| 16 |
+
- supertrainer2000
|
| 17 |
+
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|