---
license: other
license_name: model-license
library_name: transformers
pipeline_tag: text-generation
tags:
  - minimax
  - mixture-of-experts
  - moe
  - pruning
  - expert-pruning
  - fp8
base_model: morriszjm/MiniMax-M2.5-tiny
---

# MiniMax-M2.5-tiny-24e

Training-free expert-pruned variant of [`morriszjm/MiniMax-M2.5-tiny`](https://huggingface.co/morriszjm/MiniMax-M2.5-tiny),
produced by the [`minimax_expert_pruning`](https://github.com/-) pipeline.

## What changed

- `num_local_experts`: **32 → 24** (pruning rate: **25.0 %**)
- All non-MoE tensors (attention, layernorms, embeddings, lm_head, MTP heads if any)
  are **bit-identical** to the source model.
- `gate.weight` and `e_score_correction_bias` per MoE layer are row-sliced to the
  kept experts; per-expert tensors of dropped experts are absent; kept experts
  are renumbered contiguously to `0..23`.
- `top_k = num_experts_per_tok` is unchanged (8).

## Method (one-paragraph)

We run a small calibration set (64 prompts spanning Nokia AI4Code,
general English Q&A, multilingual, and reasoning) through the unpruned source
model and hook every MoE layer's router. Per layer, we accumulate each
expert's **selected probability mass** — the post-sigmoid routing weight that
the expert receives, summed over all calibration tokens that selected it
in their top-8. We keep the top-K by this score per layer (uniform K)
and atomically slice the on-disk per-expert tensors. No gradients, no
fine-tuning.

## Layer-level statistics

- Layers covered: **8**
- Tokens per layer (calibration): **1,851**
- Calibration prompts by bucket: `{"ai4code": 1008, "general_en": 416, "reasoning": 257, "multilingual": 170}`
- Median per-layer "kept-min vs drop-max" routing-mass gap: **+0.7197**
  (positive = clean separation between the kept and dropped experts;
   close to zero or negative = experts of similar utility, expect more
   quality risk)

## Intended use

Production-style serving of the source model's domain (Nokia / Merlin AI4Code
plus general English) at reduced HBM footprint. Expect graceful quality
degradation versus the unpruned source on tasks well-covered by the
calibration mix; quality on out-of-distribution domains may drop further.

## Limitations

- Training-free: no fine-tune recovery, no distillation, no merge.
- Uniform K per layer: late layers may tolerate more pruning than early ones,
  unexploited here.
- Calibration mix is small (64 prompts). Domain coverage is biased
  toward the included buckets.

## Files

`config.json`, `model-NNNNN-of-NNNNN.safetensors` (FP8), `model.safetensors.index.json`,
tokenizer, custom `modeling_minimax_m2.py` + `configuration_minimax_m2.py`, and
`expert_prune_plan.json` (full record of which experts were kept per layer).

## Loading

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

tok = AutoTokenizer.from_pretrained("morriszjm/MiniMax-M2.5-tiny-24e", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    "morriszjm/MiniMax-M2.5-tiny-24e", trust_remote_code=True,
    torch_dtype=torch.bfloat16, device_map="auto",
)
```

For vLLM serving, pass `--trust-remote-code` and (on multi-GPU) match
`--data-parallel-size` to the EP topology you compiled the K against.