Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,72 @@
|
|
1 |
-
---
|
2 |
-
license: gemma
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: gemma
|
3 |
+
language:
|
4 |
+
- tr
|
5 |
+
base_model:
|
6 |
+
- google/gemma-2-9b-it
|
7 |
+
pipeline_tag: text-generation
|
8 |
+
|
9 |
+
model-index:
|
10 |
+
- name: Gemma-2-9b-it-TR-DPO-V1
|
11 |
+
results:
|
12 |
+
- task:
|
13 |
+
type: multiple-choice
|
14 |
+
dataset:
|
15 |
+
type: multiple-choice
|
16 |
+
name: MMLU_TR_V0.2
|
17 |
+
metrics:
|
18 |
+
- name: 5-shot
|
19 |
+
type: 5-shot
|
20 |
+
value: 0.5169
|
21 |
+
verified: false
|
22 |
+
- task:
|
23 |
+
type: multiple-choice
|
24 |
+
dataset:
|
25 |
+
type: multiple-choice
|
26 |
+
name: Truthful_QA_V0.2
|
27 |
+
metrics:
|
28 |
+
- name: 0-shot
|
29 |
+
type: 0-shot
|
30 |
+
value: 0.5472
|
31 |
+
verified: false
|
32 |
+
- task:
|
33 |
+
type: multiple-choice
|
34 |
+
dataset:
|
35 |
+
type: multiple-choice
|
36 |
+
name: ARC_TR_V0.2
|
37 |
+
metrics:
|
38 |
+
- name: 25-shot
|
39 |
+
type: 25-shot
|
40 |
+
value: 0.5282
|
41 |
+
verified: false
|
42 |
+
- task:
|
43 |
+
type: multiple-choice
|
44 |
+
dataset:
|
45 |
+
type: multiple-choice
|
46 |
+
name: HellaSwag_TR_V0.2
|
47 |
+
metrics:
|
48 |
+
- name: 10-shot
|
49 |
+
type: 10-shot
|
50 |
+
value: 0.5116
|
51 |
+
verified: false
|
52 |
+
- task:
|
53 |
+
type: multiple-choice
|
54 |
+
dataset:
|
55 |
+
type: multiple-choice
|
56 |
+
name: GSM8K_TR_V0.2
|
57 |
+
metrics:
|
58 |
+
- name: 5-shot
|
59 |
+
type: 5-shot
|
60 |
+
value: 0.6507
|
61 |
+
verified: false
|
62 |
+
- task:
|
63 |
+
type: multiple-choice
|
64 |
+
dataset:
|
65 |
+
type: multiple-choice
|
66 |
+
name: Winogrande_TR_V0.2
|
67 |
+
metrics:
|
68 |
+
- name: 5-shot
|
69 |
+
type: 5-shot
|
70 |
+
value: 0.5529
|
71 |
+
verified: false
|
72 |
+
---
|