Instructions to use ForgeWorks/ForgePlex-M2-9M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ForgeWorks/ForgePlex-M2-9M with Transformers:

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

pipe = pipeline("text-generation", model="ForgeWorks/ForgePlex-M2-9M", trust_remote_code=True)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ForgeWorks/ForgePlex-M2-9M", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ForgeWorks/ForgePlex-M2-9M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ForgeWorks/ForgePlex-M2-9M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ForgeWorks/ForgePlex-M2-9M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ForgeWorks/ForgePlex-M2-9M

SGLang

How to use ForgeWorks/ForgePlex-M2-9M 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 "ForgeWorks/ForgePlex-M2-9M" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ForgeWorks/ForgePlex-M2-9M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "ForgeWorks/ForgePlex-M2-9M" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ForgeWorks/ForgePlex-M2-9M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ForgeWorks/ForgePlex-M2-9M with Docker Model Runner:
```
docker model run hf.co/ForgeWorks/ForgePlex-M2-9M
```

ForgePlex-M2-9M

File size: 2,740 Bytes

30cac1d
 
93e48a8
 
 
 
 
 
 
 
 
 
 
 
 
 
30cac1d
3cc1d37
f796c50
93e48a8
 
 
78a6b39
31bf5f2
3cc1d37
 
31bf5f2
 
 
 
 
 
93e48a8
 
 
bb56738
063d24b
 
 
83d6b98
93e48a8
 
e6fd953
 
f48cc99
 
e6fd953
 
 
 
 
 
 
93e48a8
 
564c1ba
93e48a8

---
license: apache-2.0
language:
- en
library_name: transformers
tags:
- language-model
- forgeplex
- forgeworks
- rope
- swiglu
- gqa
- attn-output-gate
- refresh-gate
- custom_code
pipeline_tag: text-generation
---

We would like to thank [Axiomic Labs](https://huggingface.co/AxiomicLabs) for allowing us to use their TrainWork framework to train this model.

# ForgePlex-M2-9M

**ForgePlex-M2-9M** is a ~9.95M-parameter decoder-only language model from **ForgeWorks**. Trained on 30B tokens. Our second attempt at creating a <10m parameter model. We're proud of this product, while M1 was a promising start, M2 shows what we can do.

![ForgePlex-M2-9M architecture](M2-Complete.jpg)

## Q&A

**What was the motivation behind M2?**

"That is an excellent question. To be completely honest. George Mallory was asked why he wanted to climb Everest. He said, “Because it’s there.” I just see M2 as a mountain to climb"

| Metric | Value |
|---|---|
| Unique parameters | **9,949,698** |
| Intelligence Index | **9.14** |
| HellaSwag | 28.02% |
| ARC (combined) | 30.00% |
| PIQA | 57.18% |
| ArithMark-3 | 34.60% |
| License | Apache-2.0 |

## Training

Trained on an all new dataset

| Data | Percentage |
|---|---|
| Finephrase | 50% |
| DCLM - Baseline | 30% |
| Proprietary Axiomic Labs dataset **releasing soon** | 10% |
| Finemath | 10% |

## Architecture

GQA + NeoX-style RoPE + RMSNorm + SwiGLU, with Qwen3.5-style **attention output gates** and **Axiomic Labs TX4 style refresh gates** on inject layers `[5, 10]` (kernel 9). **XSA is off.** Weights keep training key layout (no Llama remapping).

| Component | Details |
|---|---|
| Position encoding | **RoPE** (theta=5,000, NeoX even/odd) |
| Normalization | **RMSNorm** (eps=1e-6) |
| Feed-forward | **SwiGLU** (intermediate 707) |
| Attention | **GQA** — 8Q / 2KV, head_dim=32 + attn output gate |
| Refresh | Layers **5, 10**, kernel **9** |
| Bias | **None** |
| Embedding | **Weight tying** |
| Depth × width | **11 layers × 256 hidden** |
| Context | **1024 tokens** |
| Vocab | **4,096** custom BPE |

## Usage

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

model_path = r"C:\slm\ForgePlex-M2-9M"
tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_path,
    trust_remote_code=True,
    torch_dtype=torch.float32,
    device_map="auto",
)

prompt = "Once upon a time"
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
with torch.inference_mode():
    out = model.generate(**inputs, max_new_tokens=80, do_sample=False)
print(tokenizer.decode(out[0], skip_special_tokens=True))
```

Or run `python usage.py` from this folder.