README.md

---
license: mit
license_link: https://huggingface.co/microsoft/wavecoder-ultra-6.7b/blob/main/LICENSE
language:
  - en
library_name: transformers
datasets:
  - humaneval
pipeline_tag: text-generation
tags:
  - code
metrics:
  - code_eval
quantized_by: bartowski
---

## Exllama v2 Quantizations of wavecoder-ultra-1.1-6.7b

Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.19">turboderp's ExLlamaV2 v0.0.19</a> for quantization.

<b>The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)</b>

Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.

Original model: https://huggingface.co/microsoft/wavecoder-ultra-6.7b

Note: They updated the model in place, no idea if there's actually any changes, but tagging this 1.1 just in case

## Prompt format

No chat template specified so default is used. This may be incorrect, check original model card for details.

```
<｜begin▁of▁sentence｜>[INST] <<SYS>>
{system_prompt}
<</SYS>>

{prompt} [/INST]  <｜end▁of▁sentence｜>
```

## Available sizes

No GQA - VRAM requirements will be higher

| Branch                                                         | Bits | lm_head bits | Size (4k) | Size (16k) | Description |
| -------------------------------------------------------------- | ---- | ------------ | --------- | ---------- | ----------- |
| [8_0](https://huggingface.co/bartowski/wavecoder-ultra-1.1-6.7b-exl2/tree/8_0) | 8.0 | 8.0 | 9.0 GB | 15.2 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
| [6_5](https://huggingface.co/bartowski/wavecoder-ultra-1.1-6.7b-exl2/tree/6_5) | 6.5  | 8.0 | 8.2 GB | 14.4 GB | Near unquantized performance at vastly reduced size, **recommended**. |
| [5_0](https://huggingface.co/bartowski/wavecoder-ultra-1.1-6.7b-exl2/tree/5_0) | 5.0  | 6.0 | 6.8 GB | 13.0 GB | Slightly lower quality vs 6.5, but usable on 8GB cards with 4k context. |
| [4_25](https://huggingface.co/bartowski/wavecoder-ultra-1.1-6.7b-exl2/tree/4_25) | 4.25 | 6.0 | 6.1 GB | 12.3 GB | GPTQ equivalent bits per weight. |
| [3_5](https://huggingface.co/bartowski/wavecoder-ultra-1.1-6.7b-exl2/tree/3_5) | 3.5 | 6.0 | 5.5 GB | 11.7 GB | Lower quality, not recommended. |

## Download instructions

With git:

```shell
git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/wavecoder-ultra-1.1-6.7b-exl2 wavecoder-ultra-1.1-6.7b-exl2-6_5
```

With huggingface hub (credit to TheBloke for instructions):

```shell
pip3 install huggingface-hub
```

To download a specific branch, use the `--revision` parameter. For example, to download the 6.5 bpw branch:

Linux:

```shell
huggingface-cli download bartowski/wavecoder-ultra-1.1-6.7b-exl2 --revision 6_5 --local-dir wavecoder-ultra-1.1-6.7b-exl2-6_5 --local-dir-use-symlinks False
```

Windows (which apparently doesn't like _ in folders sometimes?):

```shell
huggingface-cli download bartowski/wavecoder-ultra-1.1-6.7b-exl2 --revision 6_5 --local-dir wavecoder-ultra-1.1-6.7b-exl2-6.5 --local-dir-use-symlinks False
```

Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski