61 lines
2.5 KiB
Markdown
61 lines
2.5 KiB
Markdown
|
---
|
||
|
|
library_name: transformers
|
||
|
|
license: apache-2.0
|
||
|
|
base_model:
|
||
|
|
- mistralai/Mistral-Nemo-Base-2407
|
||
|
|
---
|
||
|
|
|
||
|
|
|
||
|
|
|
||
|
|
# Zinakha-12b 🧙♂️
|
||
|
|
|
||
|
|
|
||
|
|
Zinakha 12b tries to become the perfect companion for any chat which involves multiple roles. The ability to understand context is pretty awesome and excels in creativity and storytelling.
|
||
|
|
It is built on Nemo 12b and trained on different datasets as well as some layer merges to ehance its capabilities.
|
||
|
|
|
||
|
|
## Model Details 📊
|
||
|
|
|
||
|
|
- **Developed by:** Aixon Lab
|
||
|
|
- **Model type:** Causal Language Model
|
||
|
|
- **Language(s):** English (primarily), may support other languages
|
||
|
|
- **License:** Apache 2.0
|
||
|
|
- **Repository:** https://huggingface.co/aixonlab/Zinakha-12b
|
||
|
|
|
||
|
|
## Quantization
|
||
|
|
- **GGUF:** https://huggingface.co/mradermacher/Zinakha-12b-GGUF
|
||
|
|
- **iMatrix GGUF:** https://huggingface.co/mradermacher/Zinakha-12b-i1-GGUF
|
||
|
|
|
||
|
|
## Model Architecture 🏗️
|
||
|
|
|
||
|
|
- **Base model:** mistralai/Mistral-Nemo-Base-2407
|
||
|
|
- **Parameter count:** ~12 billion
|
||
|
|
- **Architecture specifics:** Transformer-based language model
|
||
|
|
|
||
|
|
## Intended Use 🎯
|
||
|
|
As an advanced language model for various natural language processing tasks, including but not limited to text generation (excels in chat), question-answering, and analysis.
|
||
|
|
|
||
|
|
## Ethical Considerations 🤔
|
||
|
|
As a model based on multiple sources, Zinakha-12b may inherit biases and limitations from its constituent models. Users should be aware of potential biases in generated content and use the model responsibly.
|
||
|
|
|
||
|
|
## Performance and Evaluation
|
||
|
|
Performance metrics and evaluation results for Zinakha-12b are yet to be determined. Users are encouraged to contribute their findings and benchmarks.
|
||
|
|
|
||
|
|
## Limitations and Biases
|
||
|
|
The model may exhibit biases present in its training data and constituent models. It's crucial to critically evaluate the model's outputs and use them in conjunction with human judgment.
|
||
|
|
|
||
|
|
## Additional Information
|
||
|
|
For more details on the base model and constituent models, please refer to their respective model cards and documentation.
|
||
|
|
|
||
|
|
## How to Use
|
||
|
|
```python
|
||
|
|
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||
|
|
|
||
|
|
model = AutoModelForCausalLM.from_pretrained("aixonlab/Zinakha-12b")
|
||
|
|
tokenizer = AutoTokenizer.from_pretrained("aixonlab/Zinakha-12b")
|
||
|
|
|
||
|
|
prompt = "Once upon a time"
|
||
|
|
input_ids = tokenizer(prompt, return_tensors="pt").input_ids
|
||
|
|
|
||
|
|
generated_ids = model.generate(input_ids, max_length=100)
|
||
|
|
generated_text = tokenizer.decode(generated_ids, skip_special_tokens=True)
|
||
|
|
print(generated_text)
|