61 lines
1.5 KiB
Markdown
61 lines
1.5 KiB
Markdown
---
|
|
tags:
|
|
- merge
|
|
- mergekit
|
|
- lazymergekit
|
|
- prometheus-eval/prometheus-7b-v2.0
|
|
- 4bit/Mistral-7B-Instruct-v0.1
|
|
base_model:
|
|
- prometheus-eval/prometheus-7b-v2.0
|
|
- 4bit/Mistral-7B-Instruct-v0.1
|
|
license: apache-2.0
|
|
---
|
|
|
|
# Merge-Mistral-Prometheus-7B
|
|
|
|
Merge-Mistral-Prometheus-7B is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
|
* [prometheus-eval/prometheus-7b-v2.0](https://huggingface.co/prometheus-eval/prometheus-7b-v2.0)
|
|
* [4bit/Mistral-7B-Instruct-v0.1](https://huggingface.co/4bit/Mistral-7B-Instruct-v0.1)
|
|
|
|
## 🧩 Configuration
|
|
|
|
```yaml
|
|
models:
|
|
- model: prometheus-eval/prometheus-7b-v2.0
|
|
parameters:
|
|
weight: 1.0
|
|
- model: 4bit/Mistral-7B-Instruct-v0.1
|
|
parameters:
|
|
weight: 1.0
|
|
merge_method: linear
|
|
dtype: bfloat16
|
|
```
|
|
|
|
## 💻 Usage
|
|
|
|
```python
|
|
!pip install -qU transformers accelerate
|
|
|
|
from transformers import AutoTokenizer
|
|
import transformers
|
|
import torch
|
|
|
|
model = "vicgalle/test-merge-2"
|
|
messages = [{"role": "user", "content": "What is a large language model?"}]
|
|
|
|
tokenizer = AutoTokenizer.from_pretrained(model)
|
|
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
|
pipeline = transformers.pipeline(
|
|
"text-generation",
|
|
model=model,
|
|
torch_dtype=torch.float16,
|
|
device_map="auto",
|
|
)
|
|
|
|
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
|
|
print(outputs[0]["generated_text"])
|
|
```
|
|
|
|
## Paper citation
|
|
|
|
Paper: https://arxiv.org/abs/2406.07188 |