Downloads · 30 days
358
12% of all-time downloads
RichardErkhov/swap-uniba_-_LLaMAntino-2-13b-hf-ITA-gguf
swap-uniba_-_LLaMAntino-2-13b-hf-ITA-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Downloads · 30 days
358
12% of all-time downloads
All-time downloads
3K
Public
Repo size
131 GB
Likes
0
Public
Click a slice to open those files.
.gguf147 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
LLaMAntino-2-13b-hf-ITA - GGUF
| Name | Quant method | Size |
|---|---|---|
| LLaMAntino-2-13b-hf-ITA.Q2_K.gguf | Q2_K | 4.52GB |
| LLaMAntino-2-13b-hf-ITA.Q3_K_S.gguf | Q3_K_S | 5.27GB |
| LLaMAntino-2-13b-hf-ITA.Q3_K.gguf | Q3_K | 5.9GB |
| LLaMAntino-2-13b-hf-ITA.Q3_K_M.gguf | Q3_K_M | 5.9GB |
| LLaMAntino-2-13b-hf-ITA.Q3_K_L.gguf | Q3_K_L | 6.45GB |
| LLaMAntino-2-13b-hf-ITA.IQ4_XS.gguf | IQ4_XS | 6.54GB |
| LLaMAntino-2-13b-hf-ITA.Q4_0.gguf | Q4_0 | 6.86GB |
| LLaMAntino-2-13b-hf-ITA.IQ4_NL.gguf | IQ4_NL | 6.9GB |
| LLaMAntino-2-13b-hf-ITA.Q4_K_S.gguf | Q4_K_S | 6.91GB |
| LLaMAntino-2-13b-hf-ITA.Q4_K.gguf | Q4_K | 7.33GB |
| LLaMAntino-2-13b-hf-ITA.Q4_K_M.gguf | Q4_K_M | 4.64GB |
| LLaMAntino-2-13b-hf-ITA.Q4_1.gguf | Q4_1 | 7.61GB |
| LLaMAntino-2-13b-hf-ITA.Q5_0.gguf | Q5_0 | 8.36GB |
| LLaMAntino-2-13b-hf-ITA.Q5_K_S.gguf | Q5_K_S | 4.22GB |
| LLaMAntino-2-13b-hf-ITA.Q5_K.gguf | Q5_K | 8.6GB |
| LLaMAntino-2-13b-hf-ITA.Q5_K_M.gguf | Q5_K_M | 8.6GB |
| LLaMAntino-2-13b-hf-ITA.Q5_1.gguf | Q5_1 | 9.1GB |
| LLaMAntino-2-13b-hf-ITA.Q6_K.gguf | Q6_K | 9.95GB |
| LLaMAntino-2-13b-hf-ITA.Q8_0.gguf | Q8_0 | 12.88GB |
license: llama2 language:
Last Update: 22/01/2024<br>
LLaMAntino-2-13b is a Large Language Model (LLM) that is an italian-adapted LLaMA 2. This model aims to provide Italian NLP researchers with a base model for natural language generation tasks.
The model was trained using QLora and using as training data clean_mc4_it medium. If you are interested in more details regarding the training procedure, you can find the code we used at the following link:
NOTICE: the code has not been released yet, we apologize for the delay, it will be available asap!
Below you can find an example of model usage:
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "swap-uniba/LLaMAntino-2-13b-hf-ITA"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
prompt = "Scrivi qui un possibile prompt"
input_ids = tokenizer(prompt, return_tensors="pt").input_ids
outputs = model.generate(input_ids=input_ids)
print(tokenizer.batch_decode(outputs.detach().cpu().numpy()[:, input_ids.shape[1]:], skip_special_tokens=True)[0])
If you are facing issues when loading the model, you can try to load it quantized:
model = AutoModelForCausalLM.from_pretrained(model_id, load_in_8bit=True)
Note: The model loading strategy above requires the bitsandbytes and accelerate libraries
If you use this model in your research, please cite the following:
@misc{basile2023llamantino,
title={LLaMAntino: LLaMA 2 Models for Effective Text Generation in Italian Language},
author={Pierpaolo Basile and Elio Musacchio and Marco Polignano and Lucia Siciliani and Giuseppe Fiameni and Giovanni Semeraro},
year={2023},
eprint={2312.09993},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
Notice: Llama 2 is licensed under the LLAMA 2 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved. License