|
|
--- |
|
|
license: mit |
|
|
base_model: Ex0bit/GLM-4.7-PRISM |
|
|
model_name: Elbaz-GLM-4.7-PRISM |
|
|
tags: |
|
|
- abliteration |
|
|
- SOTA Abliteration Pipeline - PRISM |
|
|
- glm |
|
|
- glm4_moe |
|
|
- quantized |
|
|
- finetuned |
|
|
- uncensored |
|
|
- abliterated |
|
|
- mlx |
|
|
- mlx-my-repo |
|
|
language: |
|
|
- en |
|
|
- zh |
|
|
library_name: transformers |
|
|
pipeline_tag: text-generation |
|
|
--- |
|
|
|
|
|
# cs2764/GLM-4.7-PRISM-mlx-4Bit |
|
|
|
|
|
The Model [cs2764/GLM-4.7-PRISM-mlx-4Bit](https://huggingface.co/cs2764/GLM-4.7-PRISM-mlx-4Bit) was converted to MLX format from [Ex0bit/GLM-4.7-PRISM](https://huggingface.co/Ex0bit/GLM-4.7-PRISM) using mlx-lm version **0.30.4**. |
|
|
|
|
|
## Quantization Details |
|
|
|
|
|
This model was converted with the following quantization settings: |
|
|
|
|
|
- **Quantization Strategy**: 4-bit quantization |
|
|
- **Average bits per weight**: 4.502 |
|
|
|
|
|
## Use with mlx |
|
|
|
|
|
```bash |
|
|
pip install mlx-lm |
|
|
``` |
|
|
|
|
|
```python |
|
|
from mlx_lm import load, generate |
|
|
|
|
|
model, tokenizer = load("cs2764/GLM-4.7-PRISM-mlx-4Bit") |
|
|
|
|
|
prompt="hello" |
|
|
|
|
|
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None: |
|
|
messages = [{"role": "user", "content": prompt}] |
|
|
prompt = tokenizer.apply_chat_template( |
|
|
messages, tokenize=False, add_generation_prompt=True |
|
|
) |
|
|
|
|
|
response = generate(model, tokenizer, prompt=prompt, verbose=True) |
|
|
``` |
|
|
|
