--- license: cc-by-nc-4.0 language: - ro base_model: - OpenLLM-Ro/RoLlama3-8b-Instruct-2024-10-09 datasets: - OpenLLM-Ro/ro_dpo_helpsteer model-index: - name: OpenLLM-Ro/RoLlama3-8b-Instruct-DPO-2024-10-09 results: - task: type: text-generation dataset: name: RoMT-Bench type: RoMT-Bench metrics: - name: Score type: Score value: 5.87 - task: type: text-generation dataset: name: RoCulturaBench type: RoCulturaBench metrics: - name: Score type: Score value: 4.40 - task: type: text-generation dataset: name: Romanian_Academic_Benchmarks type: Romanian_Academic_Benchmarks metrics: - name: Average accuracy type: accuracy value: 49.96 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_arc_challenge type: OpenLLM-Ro/ro_arc_challenge metrics: - name: Average accuracy type: accuracy value: 46.29 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_mmlu type: OpenLLM-Ro/ro_mmlu metrics: - name: Average accuracy type: accuracy value: 53.29 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_winogrande type: OpenLLM-Ro/ro_winogrande metrics: - name: Average accuracy type: accuracy value: 65.57 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_hellaswag type: OpenLLM-Ro/ro_hellaswag metrics: - name: Average accuracy type: accuracy value: 58.15 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_gsm8k type: OpenLLM-Ro/ro_gsm8k metrics: - name: Average accuracy type: accuracy value: 34.77 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_truthfulqa type: OpenLLM-Ro/ro_truthfulqa metrics: - name: Average accuracy type: accuracy value: 41.70 - task: type: text-generation dataset: name: LaRoSeDa_binary type: LaRoSeDa_binary metrics: - name: Average macro-f1 type: macro-f1 value: 97.48 - task: type: text-generation dataset: name: LaRoSeDa_multiclass type: LaRoSeDa_multiclass metrics: - name: Average macro-f1 type: macro-f1 value: 54.00 - task: type: text-generation dataset: name: LaRoSeDa_binary_finetuned type: LaRoSeDa_binary_finetuned metrics: - name: Average macro-f1 type: macro-f1 value: 0.00 - task: type: text-generation dataset: name: LaRoSeDa_multiclass_finetuned type: LaRoSeDa_multiclass_finetuned metrics: - name: Average macro-f1 type: macro-f1 value: 0.00 - task: type: text-generation dataset: name: WMT_EN-RO type: WMT_EN-RO metrics: - name: Average bleu type: bleu value: 22.09 - task: type: text-generation dataset: name: WMT_RO-EN type: WMT_RO-EN metrics: - name: Average bleu type: bleu value: 23.00 - task: type: text-generation dataset: name: WMT_EN-RO_finetuned type: WMT_EN-RO_finetuned metrics: - name: Average bleu type: bleu value: 0.00 - task: type: text-generation dataset: name: WMT_RO-EN_finetuned type: WMT_RO-EN_finetuned metrics: - name: Average bleu type: bleu value: 0.00 - task: type: text-generation dataset: name: XQuAD type: XQuAD metrics: - name: Average exact_match type: exact_match value: 26.05 - task: type: text-generation dataset: name: XQuAD type: XQuAD metrics: - name: Average f1 type: f1 value: 42.77 - task: type: text-generation dataset: name: XQuAD_finetuned type: XQuAD_finetuned metrics: - name: Average exact_match type: exact_match value: 0.00 - task: type: text-generation dataset: name: XQuAD_finetuned type: XQuAD_finetuned metrics: - name: Average f1 type: f1 value: 0.00 - task: type: text-generation dataset: name: STS type: STS metrics: - name: Average spearman type: spearman value: 79.64 - task: type: text-generation dataset: name: STS type: STS metrics: - name: Average pearson type: pearson value: 79.52 - task: type: text-generation dataset: name: STS_finetuned type: STS_finetuned metrics: - name: Average spearman type: spearman value: 0.00 - task: type: text-generation dataset: name: STS_finetuned type: STS_finetuned metrics: - name: Average pearson type: pearson value: 0.00 - task: type: text-generation dataset: name: RoMT-Bench type: RoMT-Bench metrics: - name: First turn type: Score value: 6.22 - name: Second turn type: Score value: 5.49 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_arc_challenge type: OpenLLM-Ro/ro_arc_challenge metrics: - name: 0-shot type: accuracy value: 44.56 - name: 1-shot type: accuracy value: 45.42 - name: 3-shot type: accuracy value: 46.10 - name: 5-shot type: accuracy value: 46.27 - name: 10-shot type: accuracy value: 46.96 - name: 25-shot type: accuracy value: 48.41 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_mmlu type: OpenLLM-Ro/ro_mmlu metrics: - name: 0-shot type: accuracy value: 52.33 - name: 1-shot type: accuracy value: 52.86 - name: 3-shot type: accuracy value: 54.06 - name: 5-shot type: accuracy value: 53.90 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_winogrande type: OpenLLM-Ro/ro_winogrande metrics: - name: 0-shot type: accuracy value: 64.33 - name: 1-shot type: accuracy value: 64.72 - name: 3-shot type: accuracy value: 66.30 - name: 5-shot type: accuracy value: 66.93 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_hellaswag type: OpenLLM-Ro/ro_hellaswag metrics: - name: 0-shot type: accuracy value: 57.45 - name: 1-shot type: accuracy value: 57.65 - name: 3-shot type: accuracy value: 58.18 - name: 5-shot type: accuracy value: 58.64 - name: 10-shot type: accuracy value: 58.85 - task: type: text-generation dataset: name: OpenLLM-Ro/ro_gsm8k type: OpenLLM-Ro/ro_gsm8k metrics: - name: 1-shot type: accuracy value: 32.52 - name: 3-shot type: accuracy value: 33.97 - name: 5-shot type: accuracy value: 37.83 - task: type: text-generation dataset: name: LaRoSeDa_binary type: LaRoSeDa_binary metrics: - name: 0-shot type: macro-f1 value: 97.67 - name: 1-shot type: macro-f1 value: 97.07 - name: 3-shot type: macro-f1 value: 97.40 - name: 5-shot type: macro-f1 value: 97.80 - task: type: text-generation dataset: name: LaRoSeDa_multiclass type: LaRoSeDa_multiclass metrics: - name: 0-shot type: macro-f1 value: 58.49 - name: 1-shot type: macro-f1 value: 55.93 - name: 3-shot type: macro-f1 value: 47.70 - name: 5-shot type: macro-f1 value: 53.89 - task: type: text-generation dataset: name: WMT_EN-RO type: WMT_EN-RO metrics: - name: 0-shot type: bleu value: 8.63 - name: 1-shot type: bleu value: 25.89 - name: 3-shot type: bleu value: 26.79 - name: 5-shot type: bleu value: 27.05 - task: type: text-generation dataset: name: WMT_RO-EN type: WMT_RO-EN metrics: - name: 0-shot type: bleu value: 3.56 - name: 1-shot type: bleu value: 20.66 - name: 3-shot type: bleu value: 33.56 - name: 5-shot type: bleu value: 34.22 - task: type: text-generation dataset: name: XQuAD_EM type: XQuAD_EM metrics: - name: 0-shot type: exact_match value: 11.26 - name: 1-shot type: exact_match value: 34.29 - name: 3-shot type: exact_match value: 29.24 - name: 5-shot type: exact_match value: 29.41 - task: type: text-generation dataset: name: XQuAD_F1 type: XQuAD_F1 metrics: - name: 0-shot type: f1 value: 22.98 - name: 1-shot type: f1 value: 54.48 - name: 3-shot type: f1 value: 46.18 - name: 5-shot type: f1 value: 47.43 - task: type: text-generation dataset: name: STS_Spearman type: STS_Spearman metrics: - name: 1-shot type: spearman value: 79.99 - name: 3-shot type: spearman value: 78.42 - name: 5-shot type: spearman value: 80.51 - task: type: text-generation dataset: name: STS_Pearson type: STS_Pearson metrics: - name: 1-shot type: pearson value: 80.59 - name: 3-shot type: pearson value: 78.11 - name: 5-shot type: pearson value: 79.87 --- # Model Card for Model ID *Built with Meta Llama 3* This model points/is identical to [RoLlama3-8b-Instruct-DPO-2024-10-09](https://huggingface.co/OpenLLM-Ro/RoLlama3-8b-Instruct-DPO-2024-10-09). RoLlama3 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 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 - **Language(s):** Romanian - **License:** cc-by-nc-4.0 - **Finetuned from model:** [RoLlama3-8b-Instruct-2024-10-09](https://huggingface.co/OpenLLM-Ro/RoLlama3-8b-Instruct-2024-10-09) - **Trained using:** [RoHelpSteer](https://huggingface.co/datasets/OpenLLM-Ro/ro_dpo_helpsteer) ### Model Sources - **Repository:** https://github.com/OpenLLM-Ro/LLaMA-Factory - **Paper:** https://arxiv.org/abs/2406.18266 ## Intended Use ### Intended Use Cases RoLlama3 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 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/RoLlama3-8b-Instruct-DPO-2024-10-09") model = AutoModelForCausalLM.from_pretrained("OpenLLM-Ro/RoLlama3-8b-Instruct-DPO-2024-10-09") 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
Model
Average
ARC
MMLU
Winogrande
Hellaswag
GSM8k
TruthfulQA
Llama-3-8B-Instruct
50.62
43.69
52.04
59.33
53.19
43.87
51.59
RoLlama3-8b-Instruct-2024-06-28
50.56
44.70
52.19
67.23
57.69
30.23
51.34
RoLlama3-8b-Instruct-2024-10-09
52.21
47.94
53.50
66.06
59.72
40.16
45.90
RoLlama3-8b-Instruct-DPO-2024-10-09
49.96
46.29
53.29
65.57
58.15
34.77
41.70
## Downstream tasks
LaRoSeDa
WMT
Few-shot
Finetuned
Few-shot
Finetuned
Model
Binary
(Macro F1)
Multiclass
(Macro F1)
Binary
(Macro F1)
Multiclass
(Macro F1)
EN-RO
(Bleu)
RO-EN
(Bleu)
EN-RO
(Bleu)
RO-EN
(Bleu)
Llama-3-8B-Instruct
95.88
56.21
98.53
86.19
18.88
30.98
28.02
40.28
RoLlama3-8b-Instruct-2024-06-28
97.52
67.41
94.15
87.13
24.01
27.36
26.53
40.36
RoLlama3-8b-Instruct-2024-10-09
95.58
61.20
96.46
87.26
22.92
24.28
27.31
40.52
RoLlama3-8b-Instruct-DPO-2024-10-09
97.48
54.00
-
-
22.09
23.00
-
-
XQuAD
STS
Few-shot
Finetuned
Few-shot
Finetuned
Model
(EM)
(F1)
(EM)
(F1)
(Spearman)
(Pearson)
(Spearman)
(Pearson)
Llama-3-8B-Instruct
39.47
58.67
67.65
82.77
73.04
72.36
83.49
84.06
RoLlama3-8b-Instruct-2024-06-28
39.43
59.50
44.45
59.76
77.20
77.87
85.80
86.05
RoLlama3-8b-Instruct-2024-10-09
18.89
31.79
50.84
65.18
77.60
76.86
86.70
87.09
RoLlama3-8b-Instruct-DPO-2024-10-09
26.05
42.77
-
-
79.64
79.52
-
-
## MT-Bench
Model
Average
1st turn
2nd turn
Answers in Ro
Llama-3-8B-Instruct
5.96
6.16
5.76
158/160
RoLlama3-8b-Instruct-2024-06-28
5.15
6.03
4.28
160/160
RoLlama3-8b-Instruct-2024-10-09
5.38
6.09
4.67
160/160
RoLlama3-8b-Instruct-DPO-2024-10-09
5.87
6.22
5.49
160/160
## RoCulturaBench
Model
Average
Answers in Ro
Llama-3-8B-Instruct
4.62
100/100
RoLlama3-8b-Instruct-2024-06-28
3.71
100/100
RoLlama3-8b-Instruct-2024-10-09
3.81
100/100
RoLlama3-8b-Instruct-DPO-2024-10-09
4.40
100/100
## RoLlama3 Model Family | Model | Link | |--------------------|:--------:| |RoLlama3-8b-Instruct-2024-06-28| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3-8b-Instruct-2024-06-28) | |RoLlama3-8b-Instruct-2024-10-09| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3-8b-Instruct-2024-10-09) | |*RoLlama3-8b-Instruct-DPO-2024-10-09*| [link](https://huggingface.co/OpenLLM-Ro/RoLlama3-8b-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}, } ```