|
--- |
|
base_model: |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
- win10/Qwen2.5-2B-Instruct |
|
tags: |
|
- merge |
|
- mergekit |
|
- lazymergekit |
|
- win10/Qwen2.5-2B-Instruct |
|
--- |
|
|
|
# Qwen2.5-4B-Instruct |
|
|
|
Qwen2.5-4B-Instruct is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing): |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
* [win10/Qwen2.5-2B-Instruct](https://huggingface.co/win10/Qwen2.5-2B-Instruct) |
|
|
|
## 🧩 Configuration |
|
|
|
```yaml |
|
dtype: bfloat16 |
|
merge_method: passthrough |
|
slices: |
|
- sources: |
|
- layer_range: [0, 2] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [1, 3] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [2, 4] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [3, 5] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [4, 6] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [5, 7] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [6, 8] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [7, 9] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [8, 10] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [9, 11] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [10, 12] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [11, 13] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [12, 14] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [13, 15] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [14, 16] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [15, 17] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [16, 18] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [17, 19] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [18, 20] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [19, 21] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [20, 22] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [21, 23] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [22, 24] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [23, 25] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [24, 26] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [25, 27] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [26, 28] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [27, 29] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [28, 30] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [29, 31] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [30, 32] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [31, 33] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [32, 34] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [33, 35] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [34, 36] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [35, 37] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [36, 38] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [37, 39] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [38, 40] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [39, 41] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [40, 42] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [41, 43] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [42, 44] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [43, 45] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [44, 46] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [45, 47] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [46, 48] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [47, 49] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [48, 50] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [49, 51] |
|
model: win10/Qwen2.5-2B-Instruct |
|
- sources: |
|
- layer_range: [50, 52] |
|
model: win10/Qwen2.5-2B-Instruct |
|
``` |
|
|
|
## 💻 Usage |
|
|
|
```python |
|
!pip install -qU transformers accelerate |
|
|
|
from transformers import AutoTokenizer |
|
import transformers |
|
import torch |
|
|
|
model = "win10/Qwen2.5-4B-Instruct" |
|
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"]) |
|
``` |