初始化项目，由ModelHub XC社区提供模型
Model: OpenLLM-Ro/RoLlama3.1-8b-Instruct-DPO Source: Original Platform
This commit is contained in:
ModelHub XC
701
README.md
Normal file
701
README.md
Normal file
@@ -0,0 +1,701 @@
|
||||
---
|
||||
license: cc-by-nc-4.0
|
||||
language:
|
||||
- ro
|
||||
base_model:
|
||||
- OpenLLM-Ro/RoLlama3.1-8b-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/RoLlama3.1-8b-Instruct-DPO-2025-04-23
|
||||
results:
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: RoMT-Bench
|
||||
type: RoMT-Bench
|
||||
metrics:
|
||||
- name: Score
|
||||
type: Score
|
||||
value: 7.00
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: RoCulturaBench
|
||||
type: RoCulturaBench
|
||||
metrics:
|
||||
- name: Score
|
||||
type: Score
|
||||
value: 4.73
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: Romanian_Academic_Benchmarks
|
||||
type: Romanian_Academic_Benchmarks
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 53.76
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_arc_challenge
|
||||
type: OpenLLM-Ro/ro_arc_challenge
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 51.09
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_mmlu
|
||||
type: OpenLLM-Ro/ro_mmlu
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 56.22
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_winogrande
|
||||
type: OpenLLM-Ro/ro_winogrande
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 66.77
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_hellaswag
|
||||
type: OpenLLM-Ro/ro_hellaswag
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 59.38
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_gsm8k
|
||||
type: OpenLLM-Ro/ro_gsm8k
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 31.54
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_truthfulqa
|
||||
type: OpenLLM-Ro/ro_truthfulqa
|
||||
metrics:
|
||||
- name: Average accuracy
|
||||
type: accuracy
|
||||
value: 57.56
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: LaRoSeDa_binary
|
||||
type: LaRoSeDa_binary
|
||||
metrics:
|
||||
- name: Average macro-f1
|
||||
type: macro-f1
|
||||
value: 96.87
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: LaRoSeDa_multiclass
|
||||
type: LaRoSeDa_multiclass
|
||||
metrics:
|
||||
- name: Average macro-f1
|
||||
type: macro-f1
|
||||
value: 60.75
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: WMT_EN-RO
|
||||
type: WMT_EN-RO
|
||||
metrics:
|
||||
- name: Average bleu
|
||||
type: bleu
|
||||
value: 20.30
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: WMT_RO-EN
|
||||
type: WMT_RO-EN
|
||||
metrics:
|
||||
- name: Average bleu
|
||||
type: bleu
|
||||
value: 18.57
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: XQuAD
|
||||
type: XQuAD
|
||||
metrics:
|
||||
- name: Average exact_match
|
||||
type: exact_match
|
||||
value: 9.22
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: XQuAD
|
||||
type: XQuAD
|
||||
metrics:
|
||||
- name: Average f1
|
||||
type: f1
|
||||
value: 22.75
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: STS
|
||||
type: STS
|
||||
metrics:
|
||||
- name: Average spearman
|
||||
type: spearman
|
||||
value: 30.82
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: STS
|
||||
type: STS
|
||||
metrics:
|
||||
- name: Average pearson
|
||||
type: pearson
|
||||
value: 20.25
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: RoMT-Bench
|
||||
type: RoMT-Bench
|
||||
metrics:
|
||||
- name: First turn
|
||||
type: Score
|
||||
value: 7.30
|
||||
- name: Second turn
|
||||
type: Score
|
||||
value: 6.70
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_arc_challenge
|
||||
type: OpenLLM-Ro/ro_arc_challenge
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: accuracy
|
||||
value: 51.59
|
||||
- name: 1-shot
|
||||
type: accuracy
|
||||
value: 52.10
|
||||
- name: 3-shot
|
||||
type: accuracy
|
||||
value: 50.99
|
||||
- name: 5-shot
|
||||
type: accuracy
|
||||
value: 50.81
|
||||
- name: 10-shot
|
||||
type: accuracy
|
||||
value: 49.70
|
||||
- name: 25-shot
|
||||
type: accuracy
|
||||
value: 51.33
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_mmlu
|
||||
type: OpenLLM-Ro/ro_mmlu
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: accuracy
|
||||
value: 56.88
|
||||
- name: 1-shot
|
||||
type: accuracy
|
||||
value: 55.61
|
||||
- name: 3-shot
|
||||
type: accuracy
|
||||
value: 56.06
|
||||
- name: 5-shot
|
||||
type: accuracy
|
||||
value: 56.31
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_winogrande
|
||||
type: OpenLLM-Ro/ro_winogrande
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: accuracy
|
||||
value: 65.67
|
||||
- name: 1-shot
|
||||
type: accuracy
|
||||
value: 66.30
|
||||
- name: 3-shot
|
||||
type: accuracy
|
||||
value: 67.40
|
||||
- name: 5-shot
|
||||
type: accuracy
|
||||
value: 67.72
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_hellaswag
|
||||
type: OpenLLM-Ro/ro_hellaswag
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: accuracy
|
||||
value: 60.53
|
||||
- name: 1-shot
|
||||
type: accuracy
|
||||
value: 60.37
|
||||
- name: 3-shot
|
||||
type: accuracy
|
||||
value: 58.20
|
||||
- name: 5-shot
|
||||
type: accuracy
|
||||
value: 58.18
|
||||
- name: 10-shot
|
||||
type: accuracy
|
||||
value: 59.61
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: OpenLLM-Ro/ro_gsm8k
|
||||
type: OpenLLM-Ro/ro_gsm8k
|
||||
metrics:
|
||||
- name: 1-shot
|
||||
type: accuracy
|
||||
value: 25.09
|
||||
- name: 3-shot
|
||||
type: accuracy
|
||||
value: 30.02
|
||||
- name: 5-shot
|
||||
type: accuracy
|
||||
value: 39.50
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: LaRoSeDa_binary
|
||||
type: LaRoSeDa_binary
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: macro-f1
|
||||
value: 95.39
|
||||
- name: 1-shot
|
||||
type: macro-f1
|
||||
value: 95.90
|
||||
- name: 3-shot
|
||||
type: macro-f1
|
||||
value: 98.00
|
||||
- name: 5-shot
|
||||
type: macro-f1
|
||||
value: 98.17
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: LaRoSeDa_multiclass
|
||||
type: LaRoSeDa_multiclass
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: macro-f1
|
||||
value: 60.30
|
||||
- name: 1-shot
|
||||
type: macro-f1
|
||||
value: 64.73
|
||||
- name: 3-shot
|
||||
type: macro-f1
|
||||
value: 58.69
|
||||
- name: 5-shot
|
||||
type: macro-f1
|
||||
value: 59.30
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: WMT_EN-RO
|
||||
type: WMT_EN-RO
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: bleu
|
||||
value: 5.46
|
||||
- name: 1-shot
|
||||
type: bleu
|
||||
value: 26.08
|
||||
- name: 3-shot
|
||||
type: bleu
|
||||
value: 25.90
|
||||
- name: 5-shot
|
||||
type: bleu
|
||||
value: 23.76
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: WMT_RO-EN
|
||||
type: WMT_RO-EN
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: bleu
|
||||
value: 2.74
|
||||
- name: 1-shot
|
||||
type: bleu
|
||||
value: 20.95
|
||||
- name: 3-shot
|
||||
type: bleu
|
||||
value: 31.53
|
||||
- name: 5-shot
|
||||
type: bleu
|
||||
value: 19.05
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: XQuAD_EM
|
||||
type: XQuAD_EM
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: exact_match
|
||||
value: 12.27
|
||||
- name: 1-shot
|
||||
type: exact_match
|
||||
value: 17.98
|
||||
- name: 3-shot
|
||||
type: exact_match
|
||||
value: 5.04
|
||||
- name: 5-shot
|
||||
type: exact_match
|
||||
value: 1.60
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: XQuAD_F1
|
||||
type: XQuAD_F1
|
||||
metrics:
|
||||
- name: 0-shot
|
||||
type: f1
|
||||
value: 26.24
|
||||
- name: 1-shot
|
||||
type: f1
|
||||
value: 32.54
|
||||
- name: 3-shot
|
||||
type: f1
|
||||
value: 18.00
|
||||
- name: 5-shot
|
||||
type: f1
|
||||
value: 14.22
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: STS_Spearman
|
||||
type: STS_Spearman
|
||||
metrics:
|
||||
- name: 1-shot
|
||||
type: spearman
|
||||
value: 76.70
|
||||
- name: 3-shot
|
||||
type: spearman
|
||||
value: 2.82
|
||||
- name: 5-shot
|
||||
type: spearman
|
||||
value: 12.95
|
||||
- task:
|
||||
type: text-generation
|
||||
dataset:
|
||||
name: STS_Pearson
|
||||
type: STS_Pearson
|
||||
metrics:
|
||||
- name: 1-shot
|
||||
type: pearson
|
||||
value: 77.30
|
||||
- name: 3-shot
|
||||
type: pearson
|
||||
value: -14.56
|
||||
- name: 5-shot
|
||||
type: pearson
|
||||
value: -1.99
|
||||
|
||||
---
|
||||
|
||||
# Model Card for Model ID
|
||||
|
||||
*Built with Meta Llama 3.1*
|
||||
|
||||
This model points/is identical to [RoLlama3.1-8b-Instruct-DPO-2025-04-23](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-DPO-2025-04-23).
|
||||
|
||||
<!-- Provide a quick summary of what the model is/does. -->
|
||||
|
||||
RoLlama3.1 is a family of pretrained and fine-tuned generative text models for Romanian. This is the repository for the **human aligned instruct 8B 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:** [RoLlama3.1-8b-Instruct-2025-04-23](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-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
|
||||
|
||||
RoLlama3.1 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/RoLlama3.1-8b-Instruct-DPO")
|
||||
model = AutoModelForCausalLM.from_pretrained("OpenLLM-Ro/RoLlama3.1-8b-Instruct-DPO")
|
||||
|
||||
instruction = "Ce jocuri de societate pot juca cu prietenii mei?"
|
||||
chat = [
|
||||
{"role": "system", "content": "Ești un asistent folositor, respectuos și onest. Încearcă să ajuți cât mai mult prin informațiile oferite, excluzând răspunsuri toxice, rasiste, sexiste, periculoase și ilegale."},
|
||||
{"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>Llama-3.1-8B-Instruct</td><td><center>49.87</center></td><td><center>42.86</center></td><td><center>53.73</center></td><td><center>59.71</center></td><td><center>56.82</center></td><td><center>35.56</center></td><td><center>50.54</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2024-10-09</td><td><center>53.03</center></td><td><center>47.69</center></td><td><center>54.57</center></td><td><center>65.84</center></td><td><center>59.94</center></td><td><center><strong>44.30</strong></center></td><td><center>45.82</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2025-04-23</td><td><center>53.36</center></td><td><center>48.97</center></td><td><center>55.17</center></td><td><center>66.52</center></td><td><center><strong>60.73</strong></center></td><td><center>42.03</center></td><td><center>46.71</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-DPO-2024-10-09</td><td><center>52.74</center></td><td><center>44.84</center></td><td><center>55.06</center></td><td><center>65.87</center></td><td><center>58.67</center></td><td><center>44.17</center></td><td><center>47.82</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><em>RoLlama3.1-8b-Instruct-DPO-2025-04-23</em></td><td><center><em><strong>53.76</strong></em></center></td><td><center><em><strong>51.09</strong></em></center></td><td><center><em><strong>56.22</strong></em></center></td><td><center><em><strong>66.77</strong></em></center></td><td><center><em>59.38</em></center></td><td><center><em>31.54</em></center></td><td><center><em><strong>57.56</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>Llama-3.1-8B-Instruct</td><td><center>95.74</center></td><td><center>59.49</center></td><td><center><strong>98.57</strong></center></td><td><center>82.41</center></td><td><center>19.01</center></td><td><center><strong>27.77</strong></center></td><td><center><strong>29.02</strong></center></td><td><center>39.80</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2024-10-09</td><td><center>94.56</center></td><td><center>60.10</center></td><td><center>95.12</center></td><td><center><strong>87.53</strong></center></td><td><center>21.88</center></td><td><center>23.99</center></td><td><center>28.27</center></td><td><center><strong>40.44</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2025-04-23</td><td><center>95.32</center></td><td><center><strong>60.84</strong></center></td><td><center>-</center></td><td><center>-</center></td><td><center><strong>23.18</strong></center></td><td><center>25.11</center></td><td><center>-</center></td><td><center>-</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-DPO-2024-10-09</td><td><center>96.10</center></td><td><center>55.37</center></td><td><center>-</center></td><td><center>-</center></td><td><center>21.29</center></td><td><center>21.86</center></td><td><center>-</center></td><td><center>-</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><em>RoLlama3.1-8b-Instruct-DPO-2025-04-23</em></td><td><center><em><strong>96.87</strong></em></center></td><td><center><em>60.75</em></center></td><td><center><em>-</em></center></td><td><center><em>-</em></center></td><td><center><em>20.30</em></center></td><td><center><em>18.57</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>Llama-3.1-8B-Instruct</td><td><center><strong>44.96</strong></center></td><td><center><strong>64.45</strong></center></td><td><center><strong>69.50</strong></center></td><td><center><strong>84.31</strong></center></td><td><center>72.11</center></td><td><center>71.64</center></td><td><center>84.59</center></td><td><center>84.96</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2024-10-09</td><td><center>13.59</center></td><td><center>23.56</center></td><td><center>49.41</center></td><td><center>62.93</center></td><td><center>75.89</center></td><td><center>76.00</center></td><td><center><strong>86.86</strong></center></td><td><center><strong>87.05</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2025-04-23</td><td><center>10.74</center></td><td><center>19.75</center></td><td><center>-</center></td><td><center>-</center></td><td><center>73.53</center></td><td><center>74.93</center></td><td><center>-</center></td><td><center>-</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-DPO-2024-10-09</td><td><center>21.58</center></td><td><center>36.54</center></td><td><center>-</center></td><td><center>-</center></td><td><center><strong>78.01</strong></center></td><td><center><strong>77.98</strong></center></td><td><center>-</center></td><td><center>-</center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><em>RoLlama3.1-8b-Instruct-DPO-2025-04-23</em></td><td><center><em>9.22</em></center></td><td><center><em>22.75</em></center></td><td><center><em>-</em></center></td><td><center><em>-</em></center></td><td><center><em>30.82</em></center></td><td><center><em>20.25</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>Llama-3.1-8B-Instruct</td><td><center>5.69</center></td><td><center>5.85</center></td><td><center>5.53</center></td><td><center><strong>160/160</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2024-10-09</td><td><center>5.42</center></td><td><center>5.95</center></td><td><center>4.89</center></td><td><center><strong>160/160</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2025-04-23</td><td><center>6.43</center></td><td><center>6.78</center></td><td><center>6.09</center></td><td><center><strong>160/160</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-DPO-2024-10-09</td><td><center>6.21</center></td><td><center>6.74</center></td><td><center>5.69</center></td><td><center><strong>160/160</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><em>RoLlama3.1-8b-Instruct-DPO-2025-04-23</em></td><td><center><em><strong>7.00</strong></em></center></td><td><center><em><strong>7.30</strong></em></center></td><td><center><em><strong>6.70</strong></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>Llama-3.1-8B-Instruct</td><td><center>3.54</center></td><td><center><strong>100/100</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2024-10-09</td><td><center>3.55</center></td><td><center><strong>100/100</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-2025-04-23</td><td><center>4.28</center></td><td><center><strong>100/100</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>RoLlama3.1-8b-Instruct-DPO-2024-10-09</td><td><center>4.42</center></td><td><center><strong>100/100</strong></center></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><em>RoLlama3.1-8b-Instruct-DPO-2025-04-23</em></td><td><center><em><strong>4.73</strong></em></center></td><td><center><em><strong>100/100</strong></em></center></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
|
||||
|
||||
|
||||
## RoLlama3.1 Model Family
|
||||
|
||||
| Model | Link |
|
||||
|--------------------|:--------:|
|
||||
|RoLlama3.1-8b-Instruct-2024-10-09| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-2024-10-09) |
|
||||
|RoLlama3.1-8b-Instruct-2025-04-23| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-2025-04-23) |
|
||||
|RoLlama3.1-8b-Instruct-DPO-2024-10-09| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-DPO-2024-10-09) |
|
||||
|*RoLlama3.1-8b-Instruct-DPO-2025-04-23*| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-DPO-2025-04-23) |
|
||||
|
||||
|
||||
## Citation
|
||||
|
||||
```
|
||||
@inproceedings{masala-etal-2024-vorbesti,
|
||||
title = "``Vorbe\c{s}ti Rom{\^a}ne\c{s}te?'' A Recipe to Train Powerful {R}omanian {LLM}s with {E}nglish Instructions",
|
||||
author = "Masala, Mihai and Ilie-Ablachim, Denis and Dima, Alexandru and Corlatescu, Dragos Georgian and Zavelca, Miruna-Andreea and Olaru, Ovio and Terian, Simina-Maria and Terian, Andrei and Leordeanu, Marius and Velicu, Horia and Popescu, Marius and Dascalu, Mihai and Rebedea, Traian",
|
||||
editor = "Al-Onaizan, Yaser and Bansal, Mohit and Chen, Yun-Nung",
|
||||
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
|
||||
month = nov,
|
||||
year = "2024",
|
||||
address = "Miami, Florida, USA",
|
||||
publisher = "Association for Computational Linguistics",
|
||||
url = "https://aclanthology.org/2024.findings-emnlp.681/",
|
||||
doi = "10.18653/v1/2024.findings-emnlp.681",
|
||||
pages = "11632--11647"
|
||||
}
|
||||
```
|
||||
<!-- **APA:**
|
||||
|
||||
[More Information Needed] -->
|
||||
Reference in New Issue
Block a user