 and we move on to the next presentation right away from Lucia Wernic Xavier, who is a private member of the Read Co-op. Thank you. I am Lucia Wernic Xavier, I'm a private member of Transcribos and I present like this, I can't stay behind this, I feel too, I like to move around. I made a very simple presentation because it's an idea I had a couple of months ago and it's to work in a general, this large copy models, huge ambitions, building Portuguese models for two important document collections. I'm working for myself in a completely volunteer project for transcribing the Inquisition Documentation we started in 2016, 2019, the year before the virus and we are still working on it but it's all volunteer. Even me as a coordinator is also a volunteer and the second collection is the Notary Books of Colonial Brazil, specifically Colonial Salvador, the by year that the capital was then the capital city of Colonial Brazil. Why do we need a general model in Portuguese for simple, the Inquisitorial Documentation has professional scriptors and private scriptors as well but it's not everything. It's a huge number of professional scriptors but still it's not so large that a general model. The by year Notary Books only professional scriptors. So a general model combining these two collections will help, we hope it will help attend more people with different collections and we know that in the Portuguese archive, Torredo Tombo, there are several collections that are only private scriptors. That's why we thought about this general model combining these two documentation. So what's the ground truth we have so far? So I have printed material like this one I used in a model for printed for the second half of the 17th century. Then I have this beauty here. It's a by year Notary Book damaged and I have that beauty over there that's an Inquisition document and everything in between it's in the model. So that's not even the worst damage. There is much worse but in this ground truth we have that one. So from this and everything you can imagine in between we have even some pages that are printed and handwritten. And what did I get for this presentation? The models, one pilaya, it got a CR of 8.5 which I consider excellent results and the validation set came out 3.8. I was very happy with that. And the city lab plus, it's a slightly different. It came out both 12, 13%. I don't know why this difference. I asked the people, I asked the staff on transcripts so I will post also in select because I personally don't understand why it came out that way. So that's it from my part. Thank you. Obrigada, bedant. And that's my email address. It's just I would like to point out this is I'm doing for fun because I am not specialized in Portuguese, paleography, even though I am Brazilian, my specialty is Dutch paleography. So I count not so much for this project, especially for the paleographer challenges and standards from transcriptions. They are not mine unfortunately, but we all have our specialties. Thank you. Thank you.