---
license: apache-2.0
base_model: openbmb/MiniCPM-o-4_5
tags:
  - minicpm
  - multimodal
  - sculpt
  - structural-pruning
  - dystrio
---

# MiniCPM-o 4.5 — Sculpt Throughput (keep_frac=0.82)

18% compression — moderate quality tradeoff

Structurally pruned from [openbmb/MiniCPM-o-4_5](https://huggingface.co/openbmb/MiniCPM-o-4_5) using [Dystrio Sculpt](https://github.com/clusteroptimizerengine/BumbleB). Only the Qwen3-8B LLM backbone is pruned — vision (SigLip2), audio (Whisper), and TTS (CosyVoice2) modules are untouched.

## Quality (Downstream Probe — 250 questions)

| Metric | Baseline | This Model | Retention |
|--------|----------|------------|-----------|
| **Weighted Accuracy** | 0.6756 | 0.5709 | 84.5% |
| MMLU | 0.6700 | 0.5400 | 80.6% |
| HellaSwag | 0.7625 | 0.6750 | 88.5% |
| ARC-Challenge | 0.6000 | 0.5286 | 88.1% |

## Compression Details

- **keep_frac**: 0.82 (18% of MLP intermediate neurons removed)
- **Method**: Structural pruning with live teacher distillation (alpha=0.5)
- **Repair**: Full repair pass with workload-matched training data
- **Architecture**: All multimodal modules preserved; only LLM MLP layers compressed

## Intended Use

Drop-in replacement for MiniCPM-o 4.5 with reduced memory footprint. Suitable for:
- LoRA fine-tuning on memory-constrained GPUs
- File description and indexing workloads
- Multimodal inference with lower VRAM requirements

## How to Use

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "dystrio/MiniCPM-o-4_5-Sculpt-Throughput",
    torch_dtype=torch.bfloat16,
    trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained(
    "dystrio/MiniCPM-o-4_5-Sculpt-Throughput",
    trust_remote_code=True,
)
```