Instructions to use morriszjm/MiniMax-M2.5-tiny-24e with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use morriszjm/MiniMax-M2.5-tiny-24e with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="morriszjm/MiniMax-M2.5-tiny-24e", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = 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)
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use morriszjm/MiniMax-M2.5-tiny-24e with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "morriszjm/MiniMax-M2.5-tiny-24e"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "morriszjm/MiniMax-M2.5-tiny-24e",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/morriszjm/MiniMax-M2.5-tiny-24e

SGLang

How to use morriszjm/MiniMax-M2.5-tiny-24e with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "morriszjm/MiniMax-M2.5-tiny-24e" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "morriszjm/MiniMax-M2.5-tiny-24e",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "morriszjm/MiniMax-M2.5-tiny-24e" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "morriszjm/MiniMax-M2.5-tiny-24e",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use morriszjm/MiniMax-M2.5-tiny-24e with Docker Model Runner:
```
docker model run hf.co/morriszjm/MiniMax-M2.5-tiny-24e
```

MiniMax-M2.5-tiny-24e / README.md

morriszjm

training-free expert prune K=24/32 (PR=25%) via routing-mass calibration

51d7883 verified 11 days ago

preview code

raw

history blame contribute delete

3.26 kB

	---
	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.