RichardErkhov/acrastt_-_OmegLLaMA-3B-4bits

Quantization made by Richard Erkhov.

Github

Discord

Request more models

OmegLLaMA-3B - bnb 4bits

• Model creator: https://huggingface.co/acrastt/
• Original model: https://huggingface.co/acrastt/OmegLLaMA-3B/

Original model description:

language: - en license: apache-2.0 library_name: transformers tags: - not-for-all-audiences datasets: - anon8231489123/Omegle_logs_dataset pipeline_tag: text-generation model-index: - name: OmegLLaMA-3B results: - task: type: text-generation name: Text Generation dataset: name: AI2 Reasoning Challenge (25-Shot) type: ai2_arc config: ARC-Challenge split: test args: num_few_shot: 25 metrics: - type: acc_norm value: 40.36 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=acrastt/OmegLLaMA-3B name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: HellaSwag (10-Shot) type: hellaswag split: validation args: num_few_shot: 10 metrics: - type: acc_norm value: 66.13 name: normalized accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=acrastt/OmegLLaMA-3B name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: MMLU (5-Shot) type: cais/mmlu config: all split: test args: num_few_shot: 5 metrics: - type: acc value: 28.0 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=acrastt/OmegLLaMA-3B name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: TruthfulQA (0-shot) type: truthful_qa config: multiple_choice split: validation args: num_few_shot: 0 metrics: - type: mc2 value: 33.31 source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=acrastt/OmegLLaMA-3B name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: Winogrande (5-shot) type: winogrande config: winogrande_xl split: validation args: num_few_shot: 5 metrics: - type: acc value: 61.64 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=acrastt/OmegLLaMA-3B name: Open LLM Leaderboard - task: type: text-generation name: Text Generation dataset: name: GSM8k (5-shot) type: gsm8k config: main split: test args: num_few_shot: 5 metrics: - type: acc value: 0.23 name: accuracy source: url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=acrastt/OmegLLaMA-3B name: Open LLM Leaderboard

This is Xander Boyce's OmegLLaMA LoRA merged with OpenLLama 3B.

Prompt format:

Interests: {interests}
Conversation:
You: {prompt}
Stranger: 

For multiple interests, seperate them with space. Repeat You and Stranger for multi-turn conversations, which means Interests and Conversation are technically part of the system prompt.

GGUF quantizations available here.

This model is very good at NSFW ERP and sexting(For a 3B model). I recommend using this with Faraday.dev if you want ERP or sexting.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	38.28
ARC (25-shot)	40.36
HellaSwag (10-shot)	66.13
MMLU (5-shot)	28.0
TruthfulQA (0-shot)	33.31
Winogrande (5-shot)	61.64
GSM8K (5-shot)	0.23

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	38.28
AI2 Reasoning Challenge (25-Shot)	40.36
HellaSwag (10-Shot)	66.13
MMLU (5-Shot)	28.00
TruthfulQA (0-shot)	33.31
Winogrande (5-shot)	61.64
GSM8k (5-shot)	0.23