Q: Hello Rafael,
What should all the Appsumo buyers do now that you are officially ending development?
Will you support our refund requests from appsumo?
Appreciate all you did, sorry it didn't work our for you or us. You should definitely open-source it. Other sumo deals like Gyana, that stopped development, planned to do the same.
Anyway. Best of luck with the new project you are working on.
DISCORD UPDATE rafael — 08/10/2024 12:23 PM Hello @everyone!
Last months was super crazy for us, we've been focusing on solving a different problem and building a different product due to the limited traction we've seen with Metatext as is today.
We understand that many people here find significant value in our current product, and we want to assure you that we're committed to its maintenance but with limited resources.
We've addressed the recent bugs that some of you encountered, and these fixes are now live, particularly for the training of classification models.
We will continue to support and maintain the current service but, as i sayd, with limited resources, that why we have plans to open-source it in the future, while keeping the managed services running.
Would you be interested in joining us to maintain Metatext as an open-source project?
Q: I see there's an API for interacting from our own app.
However I'm not sure there's any mean to upload a PDF and extract some specific data. Can you clarify if it's only about sending raw text or if one can upload a PDF document for data extraction?
Q: Concerned what happens if you decide to close the business in the future.
Is there a way to export the models to another platform? It is a bit daunting thinking any databases my business will rely on may disappear if the platform is shutdown.
Q: Hello,
I would like to know the foundational language model used for training.
Additionally,
I'm interested in whether I can use the trained model locally.
I use the N8N tool to generate and create automations based on artificial intelligence insights.
The language model trained with your tool would allow for greater precision.
My use case involves a company that frequently responds to requests for proposals. I could train the model using all the past proposals I have responded to, each paired with the response I provided.
This would help train the model. I would then use the trained model to suggest responses to new proposals based on its prior training.
To summarize, I want to know the base model used for training.
Is it Llama, Mistral, or something else?
Also, once trained, can I use it locally in my self-hosted N8N workflow on my server? Thank you.
Q: Hello Rafael, What should all the Appsumo buyers do now that you are officially ending development?
Will you support our refund requests from appsumo?
Appreciate all you did, sorry it didn't work our for you or us. You should definitely open-source it. Other sumo deals like Gyana, that stopped development, planned to do the same.
Anyway. Best of luck with the new project you are working on.
DISCORD UPDATE
rafael — 08/10/2024 12:23 PM
Hello @everyone!
Last months was super crazy for us, we've been focusing on solving a different problem and building a different product due to the limited traction we've seen with Metatext as is today.
We understand that many people here find significant value in our current product, and we want to assure you that we're committed to its maintenance but with limited resources.
We've addressed the recent bugs that some of you encountered, and these fixes are now live, particularly for the training of classification models.
We will continue to support and maintain the current service but, as i sayd, with limited resources, that why we have plans to open-source it in the future, while keeping the managed services running.
Would you be interested in joining us to maintain Metatext as an open-source project?
Q: Your changelog hasn't been updated in 8 months.
https://docs.metatext.ai/v1/product/changelog shows that last update was in October 2023. I'm writing this mid-June '24. Is this product abandoned?
Q: I see there's an API for interacting from our own app.
However I'm not sure there's any mean to upload a PDF and extract some specific data.
Can you clarify if it's only about sending raw text or if one can upload a PDF document for data extraction?
Q: Concerned what happens if you decide to close the business in the future.
Is there a way to export the models to another platform? It is a bit daunting thinking any databases my business will rely on may disappear if the platform is shutdown.
Q: Hello, I would like to know the foundational language model used for training.
Additionally,
I'm interested in whether I can use the trained model locally.
I use the N8N tool to generate and create automations based on artificial intelligence insights.
The language model trained with your tool would allow for greater precision.
My use case involves a company that frequently responds to requests for proposals. I could train the model using all the past proposals I have responded to, each paired with the response I provided.
This would help train the model. I would then use the trained model to suggest responses to new proposals based on its prior training.
To summarize, I want to know the base model used for training.
Is it Llama, Mistral, or something else?
Also, once trained, can I use it locally in my self-hosted N8N workflow on my server? Thank you.