---
license: cc-by-nc-4.0
datasets:
- meta-math/MetaMathQA
language:
- en
pipeline_tag: text-generation
tags:
- Math
- exl2
---
# Merged-AGI-7B

- Model creator: [Talha Rüzgar Akkuş](https://huggingface.co/Q-bert)
- Original model: [Merged AGI 7B](https://huggingface.co/Q-bert/Merged-AGI-7B)

## EXL2 Quants

- [3.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/3.0bpw)
- [4.0bpw (main)](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/main)
- [6.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/6.0bpw)
- [8.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/8.0bpw)

Zipped Quantization (if you want to download a single file)
- [3.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/3.0bpw-zip)
- [4.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/4.0bpw-zip)
- [6.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/6.0bpw-zip)
- [8.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/8.0bpw-zip)

## Calibration Dataset
[wikitext-103-v1](https://huggingface.co/datasets/wikitext/blob/9a9e482b5987f9d25b3a9b2883fc6cc9fd8071b3/wikitext-103-v1/wikitext-test.parquet)

## Memory Usage
Measured using ExLlamav2_HF and 4096 max_seq_len with [Oobabooga's Text Generation WebUI](https://github.com/oobabooga/text-generation-webui/tree/main).
| Branch | BPW | VRAM Usage | Description |
| ------ | --- | ---------- | ----------- |
[3.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/3.0bpw)|3.0|3.7 GB|For >=6GB VRAM cards
[4.0bpw (main)](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/main)|4.0|4.4 GB|For >=6GB VRAM cards
[6.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/6.0bpw)|6.0|6.1 GB|For >=8GB VRAM cards
[8.0bpw](https://huggingface.co/hgloow/Merged-AGI-7B-EXL2/tree/8.0bpw)|8.0|7.7 GB|For >=10GB VRAM cards

## Prompt template: ChatML

```
<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

```

# Original Info

Merge [Q-bert/MetaMath-Cybertron-Starling](https://huggingface.co/Q-bert/MetaMath-Cybertron-Starling) and [fblgit/juanako-7b-UNA](https://huggingface.co/fblgit/juanako-7b-UNA) using slerp merge.

You can use ChatML format.

# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [Coming soon]()

| Metric                | Value                     |
|-----------------------|---------------------------|
| Avg.                  | Coming soon               |
| ARC (25-shot)         | Coming soon               |
| HellaSwag (10-shot)   | Coming soon               |
| MMLU (5-shot)         | Coming soon               |
| TruthfulQA (0-shot)   | Coming soon               |
| Winogrande (5-shot)   | Coming soon               |
| GSM8K (5-shot)        | Coming soon               |