Update README.md
Browse filesUpdate benchmark results
README.md
CHANGED
@@ -28,18 +28,18 @@ This model is trained with [notus](https://github.com/argilla-io/notus) code bas
|
|
28 |
|
29 |
### Training Datasets
|
30 |
|
31 |
-
-
|
32 |
|
33 |
The dataset is machine translated version of Ultrafeedback. Some samples are missing because of API request failure.
|
34 |
Will redeem the dataset and train again.
|
35 |
|
36 |
|
37 |
-
### Benchmarks
|
38 |
|
39 |
-
| Model | Average | jcommonsenseqa | jnli
|
40 |
-
|
41 |
-
| japanese-stablelm-instruct-gamma-7b |
|
42 |
-
| this model |
|
43 |
|
44 |
These benchmark performances are evaluated by [JP Language Model Evaluation Harness](https://github.com/Stability-AI/lm-evaluation-harness/tree/jp-stable).
|
45 |
|
|
|
28 |
|
29 |
### Training Datasets
|
30 |
|
31 |
+
- Machine Translated [Ultrafeedback dataset](https://huggingface.co/datasets/argilla/ultrafeedback-binarized-preferences-cleaned)
|
32 |
|
33 |
The dataset is machine translated version of Ultrafeedback. Some samples are missing because of API request failure.
|
34 |
Will redeem the dataset and train again.
|
35 |
|
36 |
|
37 |
+
### Benchmarks
|
38 |
|
39 |
+
| Model | Average | jcommonsenseqa | jnli | marc_ja | jsquad | jaqket_v2 | xlsum_ja | xwinograd_ja | mgsm |
|
40 |
+
|-------------------------------------|-----------|----------------|-----------|-----------|-----------|-----------|-----------|--------------|-----------|
|
41 |
+
| japanese-stablelm-instruct-gamma-7b | 59.86 | 83.47 | 18.65 | **95.79** | **76.29** | **82.13** | 21.47 | 81.44 | 19.60 |
|
42 |
+
| this model | **63.28** | **87.04** | **43.84** | 95.65 | 75.30 | 80.24 | **22.25** | **81.54** | **20.40** |
|
43 |
|
44 |
These benchmark performances are evaluated by [JP Language Model Evaluation Harness](https://github.com/Stability-AI/lm-evaluation-harness/tree/jp-stable).
|
45 |
|