Gryphe
Models by this creator
⛏️
MythoMax-L2-13b
234
The MythoMax-L2-13b is an AI model created by Gryphe, an improved and potentially perfected variant of Gryphe's previous MythoMix model. It combines Gryphe's MythoLogic-L2 and Huginn models using a highly experimental tensor type merge technique. This results in a model that excels at both roleplaying and storywriting tasks. Model inputs and outputs The MythoMax-L2-13b model primarily accepts Alpaca-style prompts for optimal performance. These prompts include an instruction for the model, and can optionally include a suggestion for the model to roleplay as a character and respond accordingly. Inputs Prompt**: A text instruction for the model to follow, such as "Write a story about llamas" or "Roleplay as a wise old wizard and respond to my greeting." Outputs Generated text**: The model's response to the given prompt, which can range from short informative responses to longer, more detailed roleplaying replies. Capabilities The MythoMax-L2-13b model is capable of both informative and creative text generation. It can provide factual information, but also excel at roleplaying and storytelling tasks. Due to its unique merging process, the model maintains a high level of coherency across its responses. What can I use it for? The MythoMax-L2-13b model is well-suited for a variety of creative writing and roleplaying applications. It could be used to generate character dialogues, short stories, or even extended narratives. The model's capabilities make it a useful tool for writers, game developers, and others looking to explore the intersection of AI and creative expression. Things to try One interesting aspect of the MythoMax-L2-13b model is its ability to seamlessly transition between informative and roleplaying responses. Try providing the model with a prompt that starts with a factual question, but then invites it to respond in character. See how the model maintains coherency as it shifts into a roleplaying mode. You could also experiment with the model's storytelling abilities by providing it with a scenario or prompt and asking it to continue the narrative, building on its previous responses. The model's unique merging process allows it to maintain a strong narrative flow.
Updated 5/28/2024
❗
Pantheon-RP-1.0-8b-Llama-3
43
The Pantheon-RP-1.0-8b-Llama-3 is a diverse roleplay model developed by Gryphe. It features a wide range of datasets, including variations of the No-Robots dataset, an extensive collection of GPT 4 and Claude Opus data, and the LimaRP dataset for "human factor". The model also includes the Pantheon Roleplay personas created using Claude 1.3 data, as well as additional datasets for Aiva's persona covering DM world building, Python coding, and RSS summarization. The model is designed for interactive roleplaying experiences, with a focus on shorter, character-driven responses. Model inputs and outputs The Pantheon-RP-1.0-8b-Llama-3 model is designed for text-to-text generation tasks, particularly interactive roleplay scenarios. It can handle a variety of prompts, from general instructions to open-ended roleplay situations. Inputs Roleplay prompts**: The model is optimized for character-driven roleplay scenarios, where the user provides a prompt or context for the model to continue the narrative. General instructions**: The model can also handle more general prompts, such as task descriptions or open-ended questions, drawing from its diverse training data. Outputs Roleplay responses**: The model generates character-driven responses that are typically one to two paragraphs in length, using an asterisk action, no quote for speech style. Instructional responses**: The model can also provide helpful responses to more general prompts, leveraging its broad knowledge base. Capabilities The Pantheon-RP-1.0-8b-Llama-3 model excels at interactive roleplay scenarios, where it can fluently embody a variety of personas and engage in dynamic, character-driven exchanges. The model's diverse training data allows it to handle a wide range of situations and topics, from fantastical adventures to everyday interactions. What can I use it for? The Pantheon-RP-1.0-8b-Llama-3 model is well-suited for projects that require interactive, character-driven storytelling or roleplay. This could include interactive fiction, tabletop role-playing game assistants, or even creative writing tools that allow users to collaborate with an AI character. The model's ability to handle general instructions also makes it useful for more open-ended tasks, such as providing helpful information or completing simple prompts. Things to try One interesting aspect of the Pantheon-RP-1.0-8b-Llama-3 model is its ability to maintain consistent character personalities throughout an exchange. Try providing the model with a detailed character prompt and see how it adapts its responses to stay true to that persona. You can also experiment with mixing in different types of prompts, such as general instructions or open-ended questions, to see how the model navigates the transitions between modes.
Updated 9/6/2024