README.md

---
license: apache-2.0
tags:
- finetuned
- quantized
- 4-bit
- gptq
- transformers
- safetensors
- llama
- text-generation
- dataset:ai2_arc
- dataset:unalignment/spicy-3.1
- dataset:codeparrot/apps
- dataset:facebook/belebele
- dataset:boolq
- dataset:jondurbin/cinematika-v0.1
- dataset:drop
- dataset:lmsys/lmsys-chat-1m
- dataset:TIGER-Lab/MathInstruct
- dataset:cais/mmlu
- dataset:Muennighoff/natural-instructions
- dataset:openbookqa
- dataset:piqa
- dataset:Vezora/Tested-22k-Python-Alpaca
- dataset:cakiki/rosetta-code
- dataset:Open-Orca/SlimOrca
- dataset:spider
- dataset:squad_v2
- dataset:migtissera/Synthia-v1.3
- dataset:datasets/winogrande
- dataset:nvidia/HelpSteer
- dataset:Intel/orca_dpo_pairs
- dataset:unalignment/toxic-dpo-v0.1
- dataset:jondurbin/truthy-dpo-v0.1
- dataset:allenai/ultrafeedback_binarized_cleaned
- dataset:Squish42/bluemoon-fandom-1-1-rp-cleaned
- dataset:LDJnr/Capybara
- dataset:JULIELab/EmoBank
- dataset:kingbri/PIPPA-shareGPT
- license:other
- autotrain_compatible
- endpoints_compatible
- text-generation-inference
- region:us
- has_space
model_name: UNA-34Beagles-32K-bf16-v1-GPTQ
base_model: one-man-army/UNA-34Beagles-32K-bf16-v1
inference: false
model_creator: one-man-army
pipeline_tag: text-generation
quantized_by: MaziyarPanahi
---
# Description
[MaziyarPanahi/UNA-34Beagles-32K-bf16-v1-GPTQ](https://huggingface.co/MaziyarPanahi/UNA-34Beagles-32K-bf16-v1-GPTQ) is a quantized (GPTQ) version of [one-man-army/UNA-34Beagles-32K-bf16-v1](https://huggingface.co/one-man-army/UNA-34Beagles-32K-bf16-v1)

## How to use
### Install the necessary packages

```
pip install --upgrade accelerate auto-gptq transformers
```

### Example Python code


```python
from transformers import AutoTokenizer, pipeline
from auto_gptq import AutoGPTQForCausalLM, BaseQuantizeConfig
import torch

model_id = "MaziyarPanahi/UNA-34Beagles-32K-bf16-v1-GPTQ"

quantize_config = BaseQuantizeConfig(
        bits=4,
        group_size=128,
        desc_act=False
    )

model = AutoGPTQForCausalLM.from_quantized(
        model_id,
        use_safetensors=True,
        device="cuda:0",
        quantize_config=quantize_config)

tokenizer = AutoTokenizer.from_pretrained(model_id)

pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    max_new_tokens=512,
    temperature=0.7,
    top_p=0.95,
    repetition_penalty=1.1
)

outputs = pipe("What is a large language model?")
print(outputs[0]["generated_text"])
```