Hastagaras/Halu-8B-Llama3-Blackroot
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:8kPublished:May 28, 2024License:llama3Architecture:Transformer0.0K Warm

Hastagaras/Halu-8B-Llama3-Blackroot is an experimental 8 billion parameter language model with an 8192-token context length, merged using the Model Stock method from Hastagaras/Halu-8B-Llama3-v0.3 and Blackroot's Llama-3-LongStory-LORA and Llama-3-8B-Abomination-LORA. This model is noted for its "human-like" output, potentially due to the use of human-generated data in its constituent LoRAs, and is particularly characterized by its unusual and potentially unsafe behavior in roleplay scenarios. It achieves an average score of 69.78 on the Open LLM Leaderboard, with specific strengths in HellaSwag (84.55) and Winogrande (79.48).

Loading preview...

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