TeeZee/Xwin-LM-70B-V0.1_Limarpv3
TeeZee/Xwin-LM-70B-V0.1_Limarpv3 is a 69 billion parameter merged language model combining Xwin-LM-70B-V0.1 with the LimaRP Lora v3. This model retains the qualities of the base Xwin-LM-70B-V0.1 while enhancing its ability to produce both SFW and NSFW content and seamlessly switch contexts. It is designed for versatile text generation tasks, including creative writing and roleplay scenarios.
Loading preview...
Model Overview
TeeZee/Xwin-LM-70B-V0.1_Limarpv3 is a 69 billion parameter language model created by merging the Xwin-LM/Xwin-LM-70B-V0.1 base model with the Doctor-Shotgun/limarpv3-llama2-70b-qlora Lora. This merge aims to combine the robust capabilities of the original Xwin-LM-70B-V0.1 with the specific enhancements provided by the LimaRP Lora, particularly in content generation flexibility.
Key Capabilities
- Content Generation: Capable of producing both SFW (Safe For Work) and NSFW (Not Safe For Work) content.
- Context Switching: Demonstrates seamless transitions between different conversational or narrative contexts.
- Base Model Qualities: Retains the inherent strengths and performance characteristics of the Xwin-LM-70B-V0.1 model.
Performance Metrics
Evaluated on the Open LLM Leaderboard, the model achieved an average score of 69.16. Specific benchmark results include:
- AI2 Reasoning Challenge (25-Shot): 70.82
- HellaSwag (10-Shot): 86.97
- MMLU (5-Shot): 69.28
- TruthfulQA (0-shot): 57.15
- Winogrande (5-shot): 81.77
- GSM8k (5-shot): 48.98
Good For
- Applications requiring flexible content generation, including both general-purpose and explicit material.
- Scenarios where dynamic context switching is beneficial, such as interactive storytelling or role-playing.
- Developers seeking a 70B parameter model with enhanced adaptability for diverse text outputs.