Update README.md
Browse files
README.md
CHANGED
@@ -38,24 +38,24 @@ The evaluation methdology of the [Open LLM Leaderboard](https://huggingface.co/s
|
|
38 |
|
39 |
### German benchmarks
|
40 |
|
41 |
-
| **German tasks:** | **MMLU-DE** | **Hellaswag-DE** | **ARC-DE** |
|
42 |
-
|
43 |
-
| **Models / Few-shots:** | _(5 shots)_ | _(10 shots)_ | _(24 shots)_ |
|
44 |
-
| _7B parameters_ | | | |
|
45 |
-
| llama-2-7b | 0.400 | 0.513 | 0.381 |
|
46 |
-
| leo-hessianai-7b | 0.400 | 0.609 | 0.429 |
|
47 |
-
| bloom-6b4-clp-german | 0.274 | 0.550 | 0.351 |
|
48 |
-
| mistral-7b | **0.524** | 0.588 | 0.473 |
|
49 |
-
| leo-mistral-hessianai-7b | 0.481 | 0.663 | 0.485 |
|
50 |
-
| leo-mistral-hessianai-7b-chat | 0.458 | 0.617 | 0.465 |
|
51 |
-
| DPOpenHermes-7B-v2 | TBA | 0.603 | 0.515 |
|
52 |
-
| hermeo-7b (this model) | 0.511 | **0.668** | **0.528** |
|
53 |
-
| _13B parameters_ | | | |
|
54 |
-
| llama-2-13b | 0.469 | 0.581 | 0.468 |
|
55 |
-
| leo-hessianai-13b | **0.486** | **0.658** | **0.509** |
|
56 |
-
| _70B parameters_ | | | |
|
57 |
-
| llama-2-70b | 0.597 | 0.674 | 0.561 |
|
58 |
-
| leo-hessianai-70b | **0.653** | **0.721** | **0.600** |
|
59 |
|
60 |
### English benchmarks
|
61 |
|
|
|
38 |
|
39 |
### German benchmarks
|
40 |
|
41 |
+
| **German tasks:** | **MMLU-DE** | **Hellaswag-DE** | **ARC-DE** |**Average** |
|
42 |
+
|-------------------------------|-------------|---------------|--------------|--------------|
|
43 |
+
| **Models / Few-shots:** | _(5 shots)_ | _(10 shots)_ | _(24 shots)_ | |
|
44 |
+
| _7B parameters_ | | | | |
|
45 |
+
| llama-2-7b | 0.400 | 0.513 | 0.381 | 0.431 |
|
46 |
+
| leo-hessianai-7b | 0.400 | 0.609 | 0.429 | 0.479 |
|
47 |
+
| bloom-6b4-clp-german | 0.274 | 0.550 | 0.351 | 0.392 |
|
48 |
+
| mistral-7b | **0.524** | 0.588 | 0.473 | 0.528 |
|
49 |
+
| leo-mistral-hessianai-7b | 0.481 | 0.663 | 0.485 | 0.543 |
|
50 |
+
| leo-mistral-hessianai-7b-chat | 0.458 | 0.617 | 0.465 | 0.513 |
|
51 |
+
| DPOpenHermes-7B-v2 | TBA | 0.603 | 0.515 | TBA |
|
52 |
+
| hermeo-7b (this model) | 0.511 | **0.668** | **0.528** | **0.569** |
|
53 |
+
| _13B parameters_ | | | | |
|
54 |
+
| llama-2-13b | 0.469 | 0.581 | 0.468 | 0.506 |
|
55 |
+
| leo-hessianai-13b | **0.486** | **0.658** | **0.509** | **0.551** |
|
56 |
+
| _70B parameters_ | | | | |
|
57 |
+
| llama-2-70b | 0.597 | 0.674 | 0.561 | 0.611 |
|
58 |
+
| leo-hessianai-70b | **0.653** | **0.721** | **0.600** | **0.658** |
|
59 |
|
60 |
### English benchmarks
|
61 |
|