|
|
---
|
|
|
license: cc-by-nc-4.0
|
|
|
language:
|
|
|
- ro
|
|
|
base_model:
|
|
|
- OpenLLM-Ro/RoGemma2-9b-Instruct-2025-04-23
|
|
|
datasets:
|
|
|
- OpenLLM-Ro/ro_dpo_helpsteer
|
|
|
- OpenLLM-Ro/ro_dpo_ultrafeedback
|
|
|
- OpenLLM-Ro/ro_dpo_magpie
|
|
|
- OpenLLM-Ro/ro_dpo_argilla_magpie
|
|
|
- OpenLLM-Ro/ro_dpo_helpsteer2
|
|
|
model-index:
|
|
|
- name: OpenLLM-Ro/RoGemma2-9b-Instruct-DPO-2025-04-23
|
|
|
results:
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: RoMT-Bench
|
|
|
type: RoMT-Bench
|
|
|
metrics:
|
|
|
- name: Score
|
|
|
type: Score
|
|
|
value: 7.26
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: RoCulturaBench
|
|
|
type: RoCulturaBench
|
|
|
metrics:
|
|
|
- name: Score
|
|
|
type: Score
|
|
|
value: 5.36
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: Romanian_Academic_Benchmarks
|
|
|
type: Romanian_Academic_Benchmarks
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 59.79
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_arc_challenge
|
|
|
type: OpenLLM-Ro/ro_arc_challenge
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 55.66
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_mmlu
|
|
|
type: OpenLLM-Ro/ro_mmlu
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 64.00
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_winogrande
|
|
|
type: OpenLLM-Ro/ro_winogrande
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 73.16
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_hellaswag
|
|
|
type: OpenLLM-Ro/ro_hellaswag
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 64.26
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_gsm8k
|
|
|
type: OpenLLM-Ro/ro_gsm8k
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 37.80
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_truthfulqa
|
|
|
type: OpenLLM-Ro/ro_truthfulqa
|
|
|
metrics:
|
|
|
- name: Average accuracy
|
|
|
type: accuracy
|
|
|
value: 63.86
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: LaRoSeDa_binary
|
|
|
type: LaRoSeDa_binary
|
|
|
metrics:
|
|
|
- name: Average macro-f1
|
|
|
type: macro-f1
|
|
|
value: 82.84
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: LaRoSeDa_multiclass
|
|
|
type: LaRoSeDa_multiclass
|
|
|
metrics:
|
|
|
- name: Average macro-f1
|
|
|
type: macro-f1
|
|
|
value: 65.95
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: WMT_EN-RO
|
|
|
type: WMT_EN-RO
|
|
|
metrics:
|
|
|
- name: Average bleu
|
|
|
type: bleu
|
|
|
value: 28.16
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: WMT_RO-EN
|
|
|
type: WMT_RO-EN
|
|
|
metrics:
|
|
|
- name: Average bleu
|
|
|
type: bleu
|
|
|
value: 19.34
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: XQuAD
|
|
|
type: XQuAD
|
|
|
metrics:
|
|
|
- name: Average exact_match
|
|
|
type: exact_match
|
|
|
value: 30.82
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: XQuAD
|
|
|
type: XQuAD
|
|
|
metrics:
|
|
|
- name: Average f1
|
|
|
type: f1
|
|
|
value: 48.53
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: STS
|
|
|
type: STS
|
|
|
metrics:
|
|
|
- name: Average spearman
|
|
|
type: spearman
|
|
|
value: 73.24
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: STS
|
|
|
type: STS
|
|
|
metrics:
|
|
|
- name: Average pearson
|
|
|
type: pearson
|
|
|
value: 73.13
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: RoMT-Bench
|
|
|
type: RoMT-Bench
|
|
|
metrics:
|
|
|
- name: First turn
|
|
|
type: Score
|
|
|
value: 7.65
|
|
|
- name: Second turn
|
|
|
type: Score
|
|
|
value: 6.86
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_arc_challenge
|
|
|
type: OpenLLM-Ro/ro_arc_challenge
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: accuracy
|
|
|
value: 52.44
|
|
|
- name: 1-shot
|
|
|
type: accuracy
|
|
|
value: 55.70
|
|
|
- name: 3-shot
|
|
|
type: accuracy
|
|
|
value: 56.47
|
|
|
- name: 5-shot
|
|
|
type: accuracy
|
|
|
value: 55.70
|
|
|
- name: 10-shot
|
|
|
type: accuracy
|
|
|
value: 57.16
|
|
|
- name: 25-shot
|
|
|
type: accuracy
|
|
|
value: 56.47
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_mmlu
|
|
|
type: OpenLLM-Ro/ro_mmlu
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: accuracy
|
|
|
value: 65.20
|
|
|
- name: 1-shot
|
|
|
type: accuracy
|
|
|
value: 63.27
|
|
|
- name: 3-shot
|
|
|
type: accuracy
|
|
|
value: 63.83
|
|
|
- name: 5-shot
|
|
|
type: accuracy
|
|
|
value: 63.69
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_winogrande
|
|
|
type: OpenLLM-Ro/ro_winogrande
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: accuracy
|
|
|
value: 74.11
|
|
|
- name: 1-shot
|
|
|
type: accuracy
|
|
|
value: 72.53
|
|
|
- name: 3-shot
|
|
|
type: accuracy
|
|
|
value: 72.93
|
|
|
- name: 5-shot
|
|
|
type: accuracy
|
|
|
value: 73.09
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_hellaswag
|
|
|
type: OpenLLM-Ro/ro_hellaswag
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: accuracy
|
|
|
value: 65.90
|
|
|
- name: 1-shot
|
|
|
type: accuracy
|
|
|
value: 66.06
|
|
|
- name: 3-shot
|
|
|
type: accuracy
|
|
|
value: 62.36
|
|
|
- name: 5-shot
|
|
|
type: accuracy
|
|
|
value: 61.87
|
|
|
- name: 10-shot
|
|
|
type: accuracy
|
|
|
value: 65.11
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: OpenLLM-Ro/ro_gsm8k
|
|
|
type: OpenLLM-Ro/ro_gsm8k
|
|
|
metrics:
|
|
|
- name: 1-shot
|
|
|
type: accuracy
|
|
|
value: 16.83
|
|
|
- name: 3-shot
|
|
|
type: accuracy
|
|
|
value: 43.21
|
|
|
- name: 5-shot
|
|
|
type: accuracy
|
|
|
value: 53.37
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: LaRoSeDa_binary
|
|
|
type: LaRoSeDa_binary
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: macro-f1
|
|
|
value: 39.18
|
|
|
- name: 1-shot
|
|
|
type: macro-f1
|
|
|
value: 96.59
|
|
|
- name: 3-shot
|
|
|
type: macro-f1
|
|
|
value: 97.63
|
|
|
- name: 5-shot
|
|
|
type: macro-f1
|
|
|
value: 97.97
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: LaRoSeDa_multiclass
|
|
|
type: LaRoSeDa_multiclass
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: macro-f1
|
|
|
value: 58.94
|
|
|
- name: 1-shot
|
|
|
type: macro-f1
|
|
|
value: 64.99
|
|
|
- name: 3-shot
|
|
|
type: macro-f1
|
|
|
value: 68.86
|
|
|
- name: 5-shot
|
|
|
type: macro-f1
|
|
|
value: 71.03
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: WMT_EN-RO
|
|
|
type: WMT_EN-RO
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: bleu
|
|
|
value: 26.89
|
|
|
- name: 1-shot
|
|
|
type: bleu
|
|
|
value: 31.18
|
|
|
- name: 3-shot
|
|
|
type: bleu
|
|
|
value: 30.65
|
|
|
- name: 5-shot
|
|
|
type: bleu
|
|
|
value: 23.91
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: WMT_RO-EN
|
|
|
type: WMT_RO-EN
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: bleu
|
|
|
value: 2.98
|
|
|
- name: 1-shot
|
|
|
type: bleu
|
|
|
value: 20.30
|
|
|
- name: 3-shot
|
|
|
type: bleu
|
|
|
value: 30.08
|
|
|
- name: 5-shot
|
|
|
type: bleu
|
|
|
value: 24.01
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: XQuAD_EM
|
|
|
type: XQuAD_EM
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: exact_match
|
|
|
value: 26.39
|
|
|
- name: 1-shot
|
|
|
type: exact_match
|
|
|
value: 23.87
|
|
|
- name: 3-shot
|
|
|
type: exact_match
|
|
|
value: 34.03
|
|
|
- name: 5-shot
|
|
|
type: exact_match
|
|
|
value: 38.99
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: XQuAD_F1
|
|
|
type: XQuAD_F1
|
|
|
metrics:
|
|
|
- name: 0-shot
|
|
|
type: f1
|
|
|
value: 43.28
|
|
|
- name: 1-shot
|
|
|
type: f1
|
|
|
value: 37.38
|
|
|
- name: 3-shot
|
|
|
type: f1
|
|
|
value: 54.08
|
|
|
- name: 5-shot
|
|
|
type: f1
|
|
|
value: 59.38
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: STS_Spearman
|
|
|
type: STS_Spearman
|
|
|
metrics:
|
|
|
- name: 1-shot
|
|
|
type: spearman
|
|
|
value: 73.46
|
|
|
- name: 3-shot
|
|
|
type: spearman
|
|
|
value: 73.55
|
|
|
- name: 5-shot
|
|
|
type: spearman
|
|
|
value: 72.70
|
|
|
- task:
|
|
|
type: text-generation
|
|
|
dataset:
|
|
|
name: STS_Pearson
|
|
|
type: STS_Pearson
|
|
|
metrics:
|
|
|
- name: 1-shot
|
|
|
type: pearson
|
|
|
value: 74.87
|
|
|
- name: 3-shot
|
|
|
type: pearson
|
|
|
value: 72.96
|
|
|
- name: 5-shot
|
|
|
type: pearson
|
|
|
value: 71.55
|
|
|
|
|
|
---
|
|
|
|
|
|
# Model Card for Model ID
|
|
|
|
|
|
<!-- Provide a quick summary of what the model is/does. -->
|
|
|
|
|
|
RoGemma2 is a family of pretrained and fine-tuned generative text models for Romanian. This is the repository for the **human aligned instruct 9B model**. Links to other models can be found at the bottom of this page.
|
|
|
|
|
|
## Model Details
|
|
|
|
|
|
### Model Description
|
|
|
|
|
|
<!-- Provide a longer summary of what this model is. -->
|
|
|
OpenLLM-Ro represents the first open-source effort to build a LLM specialized for Romanian. OpenLLM-Ro developed and publicly releases a collection of Romanian LLMs, both in the form of foundational model and instruct and chat variants.
|
|
|
|
|
|
|
|
|
- **Developed by:** OpenLLM-Ro
|
|
|
<!-- - **Funded by [optional]:** [More Information Needed] -->
|
|
|
<!-- - **Shared by [optional]:** [More Information Needed] -->
|
|
|
<!-- - **Model type:** [More Information Needed] -->
|
|
|
- **Language(s):** Romanian
|
|
|
- **License:** cc-by-nc-4.0
|
|
|
- **Finetuned from model:** [RoGemma2-9b-Instruct-2025-04-23](https://huggingface.co/OpenLLM-Ro/RoGemma2-9b-Instruct-2025-04-23)
|
|
|
- **Trained using:** [RoHelpSteer](https://huggingface.co/datasets/OpenLLM-Ro/ro_dpo_helpsteer), [RoUltraFeedback](https://huggingface.co/datasets/OpenLLM-Ro/ro_dpo_ultrafeedback), [RoMagpieDPO](https://huggingface.co/datasets/OpenLLM-Ro/ro_dpo_magpie), [RoArgillaMagpie](https://huggingface.co/datasets/OpenLLM-Ro/ro_dpo_argilla_magpie), [RoHelpSteer2](https://huggingface.co/datasets/OpenLLM-Ro/ro_dpo_helpsteer2)
|
|
|
|
|
|
|
|
|
### Model Sources
|
|
|
|
|
|
<!-- Provide the basic links for the model. -->
|
|
|
|
|
|
- **Repository:** https://github.com/OpenLLM-Ro/LLaMA-Factory
|
|
|
- **Paper:** https://arxiv.org/abs/2406.18266
|
|
|
|
|
|
## Intended Use
|
|
|
|
|
|
### Intended Use Cases
|
|
|
|
|
|
RoGemma2 is intented for research use in Romanian. Base models can be adapted for a variety of natural language tasks while instruction and chat tuned models are intended for assistant-like chat.
|
|
|
|
|
|
### Out-of-Scope Use
|
|
|
|
|
|
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
|
|
|
|
|
Use in any manner that violates the license, any applicable laws or regluations, use in languages other than Romanian.
|
|
|
|
|
|
|
|
|
|
|
|
## How to Get Started with the Model
|
|
|
|
|
|
Use the code below to get started with the model.
|
|
|
|
|
|
```python
|
|
|
from transformers import AutoTokenizer, AutoModelForCausalLM
|
|
|
|
|
|
tokenizer = AutoTokenizer.from_pretrained("OpenLLM-Ro/RoGemma2-9b-Instruct-DPO-2025-04-23")
|
|
|
model = AutoModelForCausalLM.from_pretrained("OpenLLM-Ro/RoGemma2-9b-Instruct-DPO-2025-10-23")
|
|
|
|
|
|
instruction = "Ce jocuri de societate pot juca cu prietenii mei?"
|
|
|
chat = [
|
|
|
{"role": "user", "content": instruction},
|
|
|
]
|
|
|
prompt = tokenizer.apply_chat_template(chat, tokenize=False, system_message="")
|
|
|
|
|
|
inputs = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt")
|
|
|
outputs = model.generate(input_ids=inputs, max_new_tokens=128)
|
|
|
print(tokenizer.decode(outputs[0]))
|
|
|
```
|
|
|
|
|
|
## Academic Benchmarks
|
|
|
|
|
|
<table>
|
|
|
<tbody>
|
|
|
<tr>
|
|
|
<td><strong>Model</strong></td>
|
|
|
<td><strong><center>Average</center></strong></td>
|
|
|
<td><strong><center>ARC</center></strong></td>
|
|
|
<td><strong><center>MMLU</center></strong></td>
|
|
|
<td><strong><center>Winogrande</center></strong></td>
|
|
|
<td><strong><center>Hellaswag</center></strong></td>
|
|
|
<td><strong><center>GSM8k</center></strong></td>
|
|
|
<td><strong><center>TruthfulQA</center></strong></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>gemma-2-9b-it</td><td><center>56.22</center></td><td><center>50.33</center></td><td><center><strong>64.01</strong></center></td><td><center>64.88</center></td><td><center>63.11</center></td><td><center>41.95</center></td><td><center>53.03</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2024-10-09</td><td><center>57.06</center></td><td><center><strong>56.20</strong></center></td><td><center>62.98</center></td><td><center>71.00</center></td><td><center>60.52</center></td><td><center>37.86</center></td><td><center>53.77</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2025-04-23</td><td><center>54.39</center></td><td><center>50.24</center></td><td><center>62.00</center></td><td><center>70.38</center></td><td><center>52.25</center></td><td><center>40.51</center></td><td><center>50.97</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-DPO-2024-10-09</td><td><center>59.08</center></td><td><center>54.10</center></td><td><center>63.41</center></td><td><center>70.02</center></td><td><center>59.35</center></td><td><center><strong>57.24</strong></center></td><td><center>50.39</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><em>RoGemma2-9b-Instruct-DPO-2025-04-23</em></td><td><center><em><strong>59.79</strong></em></center></td><td><center><em>55.66</em></center></td><td><center><em>64.00</em></center></td><td><center><em><strong>73.16</strong></em></center></td><td><center><em><strong>64.26</strong></em></center></td><td><center><em>37.80</em></center></td><td><center><em><strong>63.86</strong></em></center></td>
|
|
|
</tr>
|
|
|
</tbody>
|
|
|
</table>
|
|
|
|
|
|
|
|
|
## Downstream tasks
|
|
|
|
|
|
<table>
|
|
|
<tbody>
|
|
|
<tr>
|
|
|
<td></td>
|
|
|
<td colspan="4"><center><strong>LaRoSeDa</strong></center></td>
|
|
|
<td colspan="4"><center><strong>WMT</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td></td>
|
|
|
<td colspan="2"><center><strong>Few-shot</strong></center></td>
|
|
|
<td colspan="2"><center><strong>Finetuned</strong></center></td>
|
|
|
<td colspan="2"><center><strong>Few-shot</strong></center></td>
|
|
|
<td colspan="2"><center><strong>Finetuned</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><strong>Model</strong></td>
|
|
|
<td><center><strong>Binary<br>(Macro F1)</strong></center></td>
|
|
|
<td><center><strong>Multiclass<br>(Macro F1)</strong></center></td>
|
|
|
<td><center><strong>Binary<br>(Macro F1)</strong></center></td>
|
|
|
<td><center><strong>Multiclass<br>(Macro F1)</strong></center></td>
|
|
|
<td><center><strong>EN-RO<br>(Bleu)</strong></center></td>
|
|
|
<td><center><strong>RO-EN<br>(Bleu)</strong></center></td>
|
|
|
<td><center><strong>EN-RO<br>(Bleu)</strong></center></td>
|
|
|
<td><center><strong>RO-EN<br>(Bleu)</strong></center>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>gemma-2-9b-it</td><td><center>90.82</center></td><td><center>52.51</center></td><td><center><strong>98.97</strong></center></td><td><center>86.02</center></td><td><center>19.97</center></td><td><center><strong>28.94</strong></center></td><td><center>27.94</center></td><td><center><strong>41.61</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2024-10-09</td><td><center>96.19</center></td><td><center>62.49</center></td><td><center>98.93</center></td><td><center><strong>88.33</strong></center></td><td><center>25.74</center></td><td><center>23.16</center></td><td><center><strong>28.43</strong></center></td><td><center>40.94</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2025-04-23</td><td><center>84.23</center></td><td><center>60.14</center></td><td><center>-</center></td><td><center>-</center></td><td><center>17.78</center></td><td><center>18.24</center></td><td><center>-</center></td><td><center>-</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-DPO-2024-10-09</td><td><center><strong>97.74</strong></center></td><td><center><strong>67.40</strong></center></td><td><center>-</center></td><td><center>-</center></td><td><center>27.32</center></td><td><center>15.96</center></td><td><center>-</center></td><td><center>-</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><em>RoGemma2-9b-Instruct-DPO-2025-04-23</em></td><td><center><em>82.84</em></center></td><td><center><em>65.95</em></center></td><td><center><em>-</em></center></td><td><center><em>-</em></center></td><td><center><em><strong>28.16</strong></em></center></td><td><center><em>19.34</em></center></td><td><center><em>-</em></center></td><td><center><em>-</em></center></td>
|
|
|
</tr>
|
|
|
</tbody>
|
|
|
</table>
|
|
|
|
|
|
|
|
|
<table>
|
|
|
<tbody>
|
|
|
<tr>
|
|
|
<td></td>
|
|
|
<td colspan="4"><center><strong>XQuAD</strong></center></td>
|
|
|
<td colspan="4"><center><strong>STS</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td></td>
|
|
|
<td colspan="2"><center><strong>Few-shot</strong></center></td>
|
|
|
<td colspan="2"><center><strong>Finetuned</strong></center></td>
|
|
|
<td colspan="2"><center><strong>Few-shot</strong></center></td>
|
|
|
<td colspan="2"><center><strong>Finetuned</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><strong>Model</strong></td>
|
|
|
<td><center><strong>(EM)</strong></center></td>
|
|
|
<td><center><strong>(F1)</strong></center></td>
|
|
|
<td><center><strong>(EM)</strong></center></td>
|
|
|
<td><center><strong>(F1)</strong></center></td>
|
|
|
<td><center><strong>(Spearman)</strong></center></td>
|
|
|
<td><center><strong>(Pearson)</strong></center></td>
|
|
|
<td><center><strong>(Spearman)</strong></center></td>
|
|
|
<td><center><strong>(Pearson)</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>gemma-2-9b-it</td><td><center>37.56</center></td><td><center>57.48</center></td><td><center><strong>71.09</strong></center></td><td><center><strong>84.78</strong></center></td><td><center>71.39</center></td><td><center>71.73</center></td><td><center>89.07</center></td><td><center>89.29</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2024-10-09</td><td><center><strong>51.37</strong></center></td><td><center><strong>70.74</strong></center></td><td><center>50.00</center></td><td><center>64.10</center></td><td><center>77.15</center></td><td><center>77.10</center></td><td><center><strong>89.45</strong></center></td><td><center><strong>89.89</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2025-04-23</td><td><center>49.22</center></td><td><center>66.33</center></td><td><center>-</center></td><td><center>-</center></td><td><center>70.17</center></td><td><center>70.80</center></td><td><center>-</center></td><td><center>-</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-DPO-2024-10-09</td><td><center>32.42</center></td><td><center>58.68</center></td><td><center>-</center></td><td><center>-</center></td><td><center><strong>80.82</strong></center></td><td><center><strong>81.50</strong></center></td><td><center>-</center></td><td><center>-</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><em>RoGemma2-9b-Instruct-DPO-2025-04-23</em></td><td><center><em>30.82</em></center></td><td><center><em>48.53</em></center></td><td><center><em>-</em></center></td><td><center><em>-</em></center></td><td><center><em>73.24</em></center></td><td><center><em>73.13</em></center></td><td><center><em>-</em></center></td><td><center><em>-</em></center></td>
|
|
|
</tr>
|
|
|
</tbody>
|
|
|
</table>
|
|
|
|
|
|
|
|
|
## MT-Bench
|
|
|
|
|
|
<table>
|
|
|
<tbody>
|
|
|
<tr>
|
|
|
<td><strong>Model</strong></td>
|
|
|
<td><strong><center>Average</center></strong></td>
|
|
|
<td><strong><center>1st turn</center></strong></td>
|
|
|
<td><strong><center>2nd turn</center></strong></td>
|
|
|
<td><strong><center>Answers in Ro</center></strong></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>gemma-2-9b-it</td><td><center><strong>7.50</strong></center></td><td><center><strong>7.91</strong></center></td><td><center><strong>7.09</strong></center></td><td><center>159/160</center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2024-10-09</td><td><center>6.08</center></td><td><center>6.78</center></td><td><center>5.39</center></td><td><center><strong>160/160</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2025-04-23</td><td><center>6.78</center></td><td><center>7.00</center></td><td><center>6.55</center></td><td><center><strong>160/160</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-DPO-2024-10-09</td><td><center>6.77</center></td><td><center>7.24</center></td><td><center>6.30</center></td><td><center><strong>160/160</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><em>RoGemma2-9b-Instruct-DPO-2025-04-23</em></td><td><center><em>7.26</em></center></td><td><center><em>7.65</em></center></td><td><center><em>6.86</em></center></td><td><center><em><strong>160/160</strong></em></center></td>
|
|
|
</tr>
|
|
|
</tbody>
|
|
|
</table>
|
|
|
|
|
|
|
|
|
## RoCulturaBench
|
|
|
|
|
|
<table>
|
|
|
<tbody>
|
|
|
<tr>
|
|
|
<td><strong>Model</strong></td>
|
|
|
<td><strong><center>Average</center></strong></td>
|
|
|
<td><strong><center>Answers in Ro</center></strong></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>gemma-2-9b-it</td><td><center><strong>5.68</strong></center></td><td><center><strong>100/100</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2024-10-09</td><td><center>4.20</center></td><td><center><strong>100/100</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-2025-04-23</td><td><center>4.89</center></td><td><center><strong>100/100</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td>RoGemma2-9b-Instruct-DPO-2024-10-09</td><td><center>4.83</center></td><td><center><strong>100/100</strong></center></td>
|
|
|
</tr>
|
|
|
<tr>
|
|
|
<td><em>RoGemma2-9b-Instruct-DPO-2025-04-23</em></td><td><center><em>5.36</em></center></td><td><center><em><strong>100/100</strong></em></center></td>
|
|
|
</tr>
|
|
|
</tbody>
|
|
|
</table>
|
|
|
|
|
|
|
|
|
## RoGemma2 Model Family
|
|
|
|
|
|
| Model | Link |
|
|
|
|--------------------|:--------:|
|
|
|
|RoGemma2-9b-Instruct-2024-10-09| [link](https://huggingface.co/OpenLLM-Ro/RoGemma2-9b-Instruct-2024-10-09) |
|
|
|
|RoGemma2-9b-Instruct-2025-04-23| [link](https://huggingface.co/OpenLLM-Ro/RoGemma2-9b-Instruct-2024-10-09) |
|
|
|
|RoGemma2-9b-Instruct-DPO-2024-10-09| [link](https://huggingface.co/OpenLLM-Ro/RoGemma2-9b-Instruct-DPO-2024-10-09) |
|
|
|
|*RoGemma2-9b-Instruct-DPO-2025-04-23*| [link](https://huggingface.co/OpenLLM-Ro/RoGemma2-9b-Instruct-DPO-2024-10-09) |
|
|
|
|
|
|
|
|
|
|
|
|
## Citation
|
|
|
|
|
|
```
|
|
|
@misc{masala2024vorbecstiromanecsterecipetrain,
|
|
|
title={"Vorbe\c{s}ti Rom\^ane\c{s}te?" A Recipe to Train Powerful Romanian LLMs with English Instructions},
|
|
|
author={Mihai Masala and Denis C. Ilie-Ablachim and Alexandru Dima and Dragos Corlatescu and Miruna Zavelca and Ovio Olaru and Simina Terian-Dan and Andrei Terian-Dan and Marius Leordeanu and Horia Velicu and Marius Popescu and Mihai Dascalu and Traian Rebedea},
|
|
|
year={2024},
|
|
|
eprint={2406.18266},
|
|
|
archivePrefix={arXiv},
|
|
|
primaryClass={cs.CL},
|
|
|
url={https://arxiv.org/abs/2406.18266},
|
|
|
}
|
|
|
```
|
|
|
<!-- **APA:**
|
|
|
|
|
|
[More Information Needed] --> |
