---
license: apache-2.0
---

EXL2 quants of [Granite-20B-Code-Instruct](https://huggingface.co/ibm-granite/granite-20b-code-instruct)

pad_token_id has been changed in config.json to allow masking/batched inference to work.

[2.50 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/2.5bpw)     
[3.00 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/3.0bpw)     
[3.50 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/3.5bpw)     
[4.00 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/4.0bpw)     
[4.50 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/4.5bpw)     
[5.00 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/5.0bpw)     
[6.00 bits per weight](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/tree/6.0bpw)     

[measurement.json](https://huggingface.co/turboderp/granite-20b-code-instruct-exl2/blob/main/measurement.json)