File size: 1,567 Bytes
818b1fb d00cc49 818b1fb 3e2da8a 818b1fb 3e2da8a 818b1fb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 |
---
tags:
- merge
- mergekit
- davidkim205/Rhea-72b-v0.5
- abacusai/Smaug-72B-v0.1
base_model:
- davidkim205/Rhea-72b-v0.5
- abacusai/Smaug-72B-v0.1
license: apache-2.0
---
# Pyrhea-72B
Pyrhea-72B is a merge of the following models using [Mergekit](https://github.com/arcee-ai/mergekit):
* [davidkim205/Rhea-72b-v0.5](https://huggingface.co/davidkim205/Rhea-72b-v0.5)
* [abacusai/Smaug-72B-v0.1](https://huggingface.co/abacusai/Smaug-72B-v0.1)
## 🧩 Configuration
```yamlname: Pyrhea-72B
models:
- model: davidkim205/Rhea-72b-v0.5
parameters:
density: 0.5
weight: 0.6
# No parameters necessary for base model
- model: abacusai/Smaug-72B-v0.1
parameters:
density: 0.5
weight: 0.4
merge_method: dare_ties
base_model: davidkim205/Rhea-72b-v0.5
parameters:
int8_mask: true
dtype: bfloat16
```
## 💻 Usage
```python
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "saucam/Pyrhea-72B"
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"])
``` |