Model Card for Model ID

RoLlama2 is a family of pretrained and fine-tuned generative text models for Romanian. This is the repository for the foundational 7B model. Links to other models can be found at the bottom of this page.

Model Details

Model Description

OpenLLM 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: Llama2 Community License Agreement
  • Continual pretrained from model: Llama-2-7b
  • Trained using: CulturaX

Model Sources

Intended Use

Intended Use Cases

RoLlama2 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.

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("OpenLLM-Ro/RoLlama2-7b-Base-2024-05-14")
model = AutoModelForCausalLM.from_pretrained("OpenLLM-Ro/RoLlama2-7b-Base-2024-05-14")

input_text = "Mihai Eminescu a fost "
input_ids = tokenizer(input_text, return_tensors="pt")

outputs = model.generate(**input_ids, max_new_tokens=100)
print(tokenizer.decode(outputs[0]))

Academic Benchmarks

Model
Average
ARC
MMLU
Winogrande
Hellaswag
GSM8k
TruthfulQA
Llama-2-7b
37.04
36.05
33.66
57.56
48.00
4.75
42.22
RoLlama2-7b-Base-2024-05-14
38.03
37.95
27.22
59.29
57.22
2.53
44.00

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-2-7b
93.19
54.11
98.43
87.22
14.90
26.61
24.95
39.09
RoLlama2-7b-Base-2024-05-14
83.25
61.04
98.97
87.72
10.01
13.03
27.85
39.30
XQuAD
STS
Few-shot
Finetuned
Few-shot
Finetuned
Model
(EM)
(F1)
(EM)
(F1)
(Spearman)
(Pearson)
(Spearman)
(Pearson)
Llama-2-7b
38.91
56.82
65.46
79.42
9.08
9.07
79.93
81.08
RoLlama2-7b-Base-2024-05-14
30.15
47.03
67.06
79.96
7.89
7.98
71.75
71.99

RoLlama2 Model Family

Model Link
RoLlama2-7b-Base-2024-05-14 link
RoLlama2-7b-Instruct-2024-05-14 link
RoLlama2-7b-Instruct-2024-10-09 link
RoLlama2-7b-Instruct-DPO-2024-10-09 link

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}, 
}
Downloads last month
376
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for OpenLLM-Ro/RoLlama2-7b-Base-2024-05-14

Finetuned
(619)
this model
Quantizations
2 models

Dataset used to train OpenLLM-Ro/RoLlama2-7b-Base-2024-05-14

Collection including OpenLLM-Ro/RoLlama2-7b-Base-2024-05-14

Evaluation results