sophosympatheia/Nova-Tempus-70B-v0.3

Hugging Face
TEXT GENERATIONPricing:Input $2.88 / Output $2.88Concurrent Unit Cost:4Model Size:70BQuant:FP8Context Size:32kTool Calling:SupportedLicense:llama3.3Architecture:Transformer0.0K Featherless Exclusive Warm

Nova-Tempus-70B-v0.3 by sophosympatheia is a 70 billion parameter merged language model, combining Nova-Tempus-70B-v0.1 with DeepSeek-R1-Distill-Llama-70B using the SCE method. This model is specifically designed and optimized for creative writing, roleplaying, and storytelling, excelling at generating long, formally structured prose. It is uncensored and supports a 32768 token context length.

Loading preview...

Nova-Tempus-70B-v0.3 Overview

Nova-Tempus-70B-v0.3 is a 70 billion parameter language model developed by sophosympatheia. It is a merge of sophosympatheia/Nova-Tempus-70B-v0.1 and deepseek-ai/DeepSeek-R1-Distill-Llama-70B, utilizing the SCE (Selective Channel Expansion) merge method with meta-llama/Llama-3.3-70B-Instruct as its base. This version resolves tokenizer issues present in earlier iterations.

Key Capabilities

  • Creative Writing & Roleplaying: Specifically designed and optimized for generating immersive roleplaying experiences and detailed storytelling.
  • Formal Prose Generation: Excels at producing long outputs with a formal prose structure, akin to novel writing.
  • Uncensored Content: The model is uncensored, allowing for a wide range of narrative themes, including adult content, violence, and graphic descriptions, as guided by the user.
  • Context Length: Supports a context length of 32768 tokens.

When to Use This Model

  • Roleplaying Scenarios: Ideal for collaborative storytelling and roleplaying where detailed character interactions and narrative depth are desired.
  • Creative Writing Projects: Suitable for generating long-form creative content, including fiction, where a formal and consistent writing style is preferred.
  • Unrestricted Content Generation: For use cases requiring uncensored outputs and the exploration of mature or sensitive themes in a narrative context.

Users are advised to consult the Llama 3.3 Community License Agreement due to the uncertain licensing terms of this merged model.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p