78 lines
2.1 KiB
Markdown
78 lines
2.1 KiB
Markdown
|
---
|
||
|
|
tags:
|
||
|
|
- merge
|
||
|
|
- mergekit
|
||
|
|
- lazymergekit
|
||
|
|
- bardsai/jaskier-7b-dpo-v3.3
|
||
|
|
- Kquant03/NeuralTrix-7B-dpo-laser
|
||
|
|
- CultriX/NeuralTrix-v4-bf16
|
||
|
|
- CultriX/NeuralTrix-V2
|
||
|
|
base_model:
|
||
|
|
- bardsai/jaskier-7b-dpo-v3.3
|
||
|
|
- Kquant03/NeuralTrix-7B-dpo-laser
|
||
|
|
- CultriX/NeuralTrix-v4-bf16
|
||
|
|
- CultriX/NeuralTrix-V2
|
||
|
|
license: apache-2.0
|
||
|
|
---
|
||
|
|
|
||
|
|
# NeuralTrixlaser-bf16
|
||
|
|
|
||
|
|
NeuralTrixlaser-bf16 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
||
|
|
* [bardsai/jaskier-7b-dpo-v3.3](https://huggingface.co/bardsai/jaskier-7b-dpo-v3.3)
|
||
|
|
* [Kquant03/NeuralTrix-7B-dpo-laser](https://huggingface.co/Kquant03/NeuralTrix-7B-dpo-laser)
|
||
|
|
* [CultriX/NeuralTrix-v4-bf16](https://huggingface.co/CultriX/NeuralTrix-v4-bf16)
|
||
|
|
* [CultriX/NeuralTrix-V2](https://huggingface.co/CultriX/NeuralTrix-V2)
|
||
|
|
|
||
|
|
## 🧩 Configuration
|
||
|
|
|
||
|
|
```yaml
|
||
|
|
models:
|
||
|
|
- model: eren23/dpo-binarized-NeuralTrix-7B
|
||
|
|
# no parameters necessary for base model
|
||
|
|
- model: bardsai/jaskier-7b-dpo-v3.3
|
||
|
|
parameters:
|
||
|
|
density: 0.65
|
||
|
|
weight: 0.4
|
||
|
|
- model: Kquant03/NeuralTrix-7B-dpo-laser
|
||
|
|
parameters:
|
||
|
|
density: 0.6
|
||
|
|
weight: 0.35
|
||
|
|
- model: CultriX/NeuralTrix-v4-bf16
|
||
|
|
parameters:
|
||
|
|
density: 0.55
|
||
|
|
weight: 0.15
|
||
|
|
- model: CultriX/NeuralTrix-V2
|
||
|
|
parameters:
|
||
|
|
density: 0.55
|
||
|
|
weight: 0.15
|
||
|
|
merge_method: dare_ties
|
||
|
|
base_model: eren23/dpo-binarized-NeuralTrix-7B
|
||
|
|
parameters:
|
||
|
|
int8_mask: true
|
||
|
|
dtype: bfloat16
|
||
|
|
```
|
||
|
|
|
||
|
|
## 💻 Usage
|
||
|
|
|
||
|
|
```python
|
||
|
|
!pip install -qU transformers accelerate
|
||
|
|
|
||
|
|
from transformers import AutoTokenizer
|
||
|
|
import transformers
|
||
|
|
import torch
|
||
|
|
|
||
|
|
model = "CultriX/"
|
||
|
|
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"])
|
||
|
|
```
|