KaraKaraWarehouse/Llama-MiraiFanfare-2-3.3-70B

Hugging Face
TEXT GENERATIONPricing:Input $2.6 / Cached $0.52 / Output $3Concurrent Unit Cost:4Model Size:70BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Dec 26, 2024Architecture:Transformer0.0K Featherless Exclusive Warm

KaraKaraWitch/Llama-MiraiFanfare-2-3.3-70B is a 70 billion parameter language model based on the Llama architecture, created by merging EVA-LLaMA-3.33-70B-v0.1 and Mirai-3.0-70B using the TIES method. This model leverages the strengths of its constituent models to offer enhanced performance. With a 32,768 token context length, it is suitable for applications requiring extensive contextual understanding.

Loading preview...

Model Overview

KaraKaraWitch/Llama-MiraiFanfare-2-3.3-70B is a 70 billion parameter language model developed by KaraKaraWitch. It is a merge of pre-trained language models, specifically combining elements from EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1 and Blackroot/Mirai-3.0-70B. The merge was performed using the TIES (Trimming and Expanding) method, with EVA-LLaMA-3.33-70B-v0.1 serving as the base model.

Key Characteristics

  • Architecture: Llama-based, leveraging a 70 billion parameter count.
  • Merge Method: Utilizes the TIES merging technique for combining model weights.
  • Constituent Models: Incorporates Blackroot/Mirai-3.0-70B and EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.1.
  • Context Length: Supports a substantial context window of 32,768 tokens, enabling processing of longer inputs.

Potential Use Cases

This model is designed for general language understanding and generation tasks, benefiting from the combined capabilities of its merged components. Its large parameter count and extended context length make it suitable for:

  • Complex text analysis and summarization.
  • Advanced conversational AI and chatbots.
  • Applications requiring deep contextual understanding over long documents.

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