30 lines
999 B
Markdown
30 lines
999 B
Markdown
---
|
|
license: apache-2.0
|
|
language:
|
|
- en
|
|
pipeline_tag: text-generation
|
|
tags:
|
|
- mistral
|
|
- instruct
|
|
- gguf
|
|
- imatrix
|
|
base_model: mistralai/Mistral-Nemo-Instruct-2407
|
|
---
|
|
|
|
# Quant Infos
|
|
|
|
|
|
- needs [#8604](https://github.com/ggerganov/llama.cpp/pull/8604) & latest master with tekken tokenizer fixes applied
|
|
- quants done with an importance matrix for improved quantization loss
|
|
- Quantized ggufs & imatrix from hf bf16, through bf16. `safetensors bf16 -> gguf bf16 -> quant` for optimal quant loss.
|
|
- Wide coverage of different gguf quant types from Q\_8\_0 down to IQ1\_S
|
|
- experimental custom quant types
|
|
- `_L` with `--output-tensor-type f16 --token-embedding-type f16` (same as bartowski's)
|
|
- Imatrix generated with [this](https://gist.github.com/bartowski1182/eb213dccb3571f863da82e99418f81e8) multi-purpose dataset by [bartowski](https://huggingface.co/bartowski).
|
|
```
|
|
./imatrix -m $model_name-bf16.gguf -f calibration_datav3.txt -o $model_name.imatrix
|
|
```
|
|
|
|
# Original Model Card:
|
|
|
|
TODO |