File size: 1,080 Bytes
158a640 dadbcfe 158a640 dadbcfe 158a640 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 |
---
tags:
- generated_from_trainer
model-index:
- name: zephyr-7b-beta
results: []
license: mit
datasets:
- HuggingFaceH4/ultrachat_200k
- HuggingFaceH4/ultrafeedback_binarized
language:
- en
base_model: mistralai/Mistral-7B-v0.1
quantized_by: bartowski
---
# Exllama v2 Quantizations of zephyr-7b-beta at 4.0
Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.7">turboderp's ExLlamaV2 v0.0.7</a> for quantization.
Conversion was done using wikitext-103-raw-v1-test.parquet as calibration dataset.
Original model: https://huggingface.co/HuggingFaceH4/zephyr-7b-beta
## Download instructions
With git:
```shell
git clone --single-branch --branch 4.0 https://huggingface.co/bartowski/zephyr-7b-beta-exl2
```
With huggingface hub (credit to TheBloke for instructions):
```shell
pip3 install huggingface-hub
```
To download from a different branch, add the `--revision` parameter:
```shell
mkdir zephyr-7b-beta-exl2
huggingface-cli download bartowski/zephyr-7b-beta-exl2 --revision --local-dir zephyr-7b-beta-exl2 --local-dir-use-symlinks False
```
|