Text Generation
Safetensors
diffusion_gemma
diffusion-gemma
ream
expert-pruning
unsloth
conversational
Instructions to use Akicou/diffusiongemma-18B-A4B-it-OP with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Local Apps Settings
- Unsloth Studio
How to use Akicou/diffusiongemma-18B-A4B-it-OP with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for Akicou/diffusiongemma-18B-A4B-it-OP to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for Akicou/diffusiongemma-18B-A4B-it-OP to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for Akicou/diffusiongemma-18B-A4B-it-OP to start chatting
Load model with FastModel
pip install unsloth from unsloth import FastModel model, tokenizer = FastModel.from_pretrained( model_name="Akicou/diffusiongemma-18B-A4B-it-OP", max_seq_length=2048, )
DiffusionGemma 26B A4B - 90 Experts
Pruned from unsloth/diffusiongemma-26B-A4B-it using expert removal.
128 experts → 90 experts (70.3% retention, 38 experts removed)
Method
Direct safetensors-level offline pruning with deterministic seed 42. Custom pruning code at Akicou/ream.
Size
| Version | Format | Size |
|---|---|---|
| Full (128e) | BF16 safetensors | 49 GB |
| Pruned (90e) | BF16 safetensors | 36 GB |
Quantization
Convert to GGUF and quantize for inference:
- Downloads last month
- -