Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,59 @@
|
|
1 |
---
|
2 |
license: mit
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: mit
|
3 |
+
tags:
|
4 |
+
- merge
|
5 |
---
|
6 |
+
---
|
7 |
+
base_model: [nisten/BigCodeLlama-169b]
|
8 |
+
tags:
|
9 |
+
- mergekit
|
10 |
+
- merge
|
11 |
+
|
12 |
+
---
|
13 |
+
# Quantizations of BigCodeLLama LFG 🚀
|
14 |
+
|
15 |
+
## Experimental CodeLlaMA frankenstein to see how it benchmarks
|
16 |
+
|
17 |
+
### Models Merged
|
18 |
+
|
19 |
+
The following models were included in the merge:
|
20 |
+
* ../CodeLlama-70b-hf
|
21 |
+
* ../CodeLlama-70b-Instruct-hf
|
22 |
+
* ../CodeLlama-70b-Python-hf
|
23 |
+
|
24 |
+
### Configuration
|
25 |
+
|
26 |
+
The following YAML configuration was used to produce this model:
|
27 |
+
|
28 |
+
```yaml
|
29 |
+
dtype: bfloat16
|
30 |
+
merge_method: passthrough
|
31 |
+
slices:
|
32 |
+
- sources:
|
33 |
+
- layer_range: [0, 69]
|
34 |
+
model:
|
35 |
+
model:
|
36 |
+
path: ../CodeLlama-70b-hf
|
37 |
+
- sources:
|
38 |
+
- layer_range: [66, 76]
|
39 |
+
model:
|
40 |
+
model:
|
41 |
+
path: ../CodeLlama-70b-Instruct-hf
|
42 |
+
- sources:
|
43 |
+
- layer_range: [42, 66]
|
44 |
+
model:
|
45 |
+
model:
|
46 |
+
path: ../CodeLlama-70b-hf
|
47 |
+
- sources:
|
48 |
+
- layer_range: [13, 37]
|
49 |
+
model:
|
50 |
+
model:
|
51 |
+
path: ../CodeLlama-70b-Python-hf
|
52 |
+
- sources:
|
53 |
+
- layer_range: [10, 80]
|
54 |
+
model:
|
55 |
+
model:
|
56 |
+
path: ../CodeLlama-70b-Instruct-hf
|
57 |
+
```
|
58 |
+
|
59 |
+
### Stay tuned for GGUFs quants
|