Instructions to use SLM-Archive/NanoDex-Test-500K-200M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SLM-Archive/NanoDex-Test-500K-200M with Transformers:

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

pipe = pipeline("text-generation", model="SLM-Archive/NanoDex-Test-500K-200M")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SLM-Archive/NanoDex-Test-500K-200M")
model = AutoModelForCausalLM.from_pretrained("SLM-Archive/NanoDex-Test-500K-200M", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SLM-Archive/NanoDex-Test-500K-200M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SLM-Archive/NanoDex-Test-500K-200M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SLM-Archive/NanoDex-Test-500K-200M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/SLM-Archive/NanoDex-Test-500K-200M

SGLang

How to use SLM-Archive/NanoDex-Test-500K-200M 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 "SLM-Archive/NanoDex-Test-500K-200M" \
    --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": "SLM-Archive/NanoDex-Test-500K-200M",
		"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 "SLM-Archive/NanoDex-Test-500K-200M" \
        --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": "SLM-Archive/NanoDex-Test-500K-200M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use SLM-Archive/NanoDex-Test-500K-200M with Docker Model Runner:
```
docker model run hf.co/SLM-Archive/NanoDex-Test-500K-200M
```

NanoDex-Test-500K-200M / README.md

DedeProGames

Update README.md

948f629 verified about 1 month ago

preview code

Raw History Blame Contribute Delete

2.69 kB

---
license: odc-by
datasets:
  - HuggingFaceFW/fineweb-edu
language:
  - en
library_name: transformers
pipeline_tag: text-generation
tags:
  - nanodex
  - tiny-lm
  - pretrained-from-scratch
---
<div align="center">
  <h3 align="center">Model archived by:</h3>

  <table align="center">
    <tr>
      <td align="center" valign="middle" width="140">
        <a href="https://huggingface.co/DedeProGames" target="_blank">
          <img
            src="https://cdn-avatars.huggingface.co/v1/production/uploads/685ea8ff7b4139b6845ce395/Im--QSnbrnAhHPPhpX8L0.png"
            width="80"
            height="80"
            alt="DedeProGames"
            style="display:block; margin:0 auto; border-radius:50%; object-fit:cover;"
          />
          <br />
          <b>DedeProGames</b>
        </a>
      </td>
    </tr>
  </table>
</div>
---
# NanoDex-Test-500K-200M

A **492,192-parameter** decoder-only language model pre-trained
**from scratch** on [fineweb-edu](https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu),
using the [NanoDex Trainer](https://huggingface.co/spaces/hugging-science/nanodex-trainer) Space.

## Architecture

A standard `LlamaForCausalLM` decoder-only transformer — SiLU MLP, RMSNorm,
rotary position embeddings, grouped-query attention, tied embeddings, no biases —
scaled down in width and depth to fit the parameter budget.

| | |
|---|---|
| Parameters | 492,192 |
| Hidden size | 96 |
| Layers | 3 |
| Attention heads | 6 (KV: 2) |
| FFN size | 256 |
| Context length | 512 |
| Vocab | 2,048 (custom BPE trained on fineweb-edu) |

## Training

| | |
|---|---|
| Tokens seen | 199,884,800 |
| Steps | 1,525 |
| Tokens / step | 131,072 |
| Optimizer | AdamW(0.9, 0.95) wd=0.1 clip=1.0 |
| LR schedule | warmup 2% + cosine to 10% (peak 4e-03) |
| Final loss | 3.8176 (ppl 45.5) |
| Wall time | 5.2 min |
| Trained by | [@DedeProGames](https://huggingface.co/DedeProGames) |

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("DedeProGames/NanoDex-Test-500K-200M")
model = AutoModelForCausalLM.from_pretrained("DedeProGames/NanoDex-Test-500K-200M")
ids = tok("The mitochondria is", return_tensors="pt").input_ids
print(tok.decode(model.generate(ids, max_new_tokens=60, do_sample=True,
                                temperature=0.8, top_k=50)[0]))
```

## Caveats

This is a **nano-scale research artifact**. At this parameter count and token
budget the model learns word shapes, common collocations and a little syntax —
it is not a useful assistant and its output is not factual. It exists to make
"pre-train a transformer from scratch" something you can actually watch happen.

	---
	license: odc-by
	datasets:
	- HuggingFaceFW/fineweb-edu
	language:
	- en
	library_name: transformers
	pipeline_tag: text-generation
	tags:
	- nanodex
	- tiny-lm
	- pretrained-from-scratch
	---
	<div align="center">
	<h3 align="center">Model archived by:</h3>

	<table align="center">
	<tr>
	<td align="center" valign="middle" width="140">
	<a href="https://huggingface.co/DedeProGames" target="_blank">
	<img
	src="https://cdn-avatars.huggingface.co/v1/production/uploads/685ea8ff7b4139b6845ce395/Im--QSnbrnAhHPPhpX8L0.png"
	width="80"
	height="80"
	alt="DedeProGames"
	style="display:block; margin:0 auto; border-radius:50%; object-fit:cover;"
	/>
	<br />
	<b>DedeProGames</b>
	</a>
	</td>
	</tr>
	</table>
	</div>

	---
	# NanoDex-Test-500K-200M

	A 492,192-parameter decoder-only language model pre-trained
	from scratch on [fineweb-edu](https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu),
	using the [NanoDex Trainer](https://huggingface.co/spaces/hugging-science/nanodex-trainer) Space.

	## Architecture

	A standard `LlamaForCausalLM` decoder-only transformer — SiLU MLP, RMSNorm,
	rotary position embeddings, grouped-query attention, tied embeddings, no biases —
	scaled down in width and depth to fit the parameter budget.

	\| \| \|
	\|---\|---\|
	\| Parameters \| 492,192 \|
	\| Hidden size \| 96 \|
	\| Layers \| 3 \|
	\| Attention heads \| 6 (KV: 2) \|
	\| FFN size \| 256 \|
	\| Context length \| 512 \|
	\| Vocab \| 2,048 (custom BPE trained on fineweb-edu) \|

	## Training

	\| \| \|
	\|---\|---\|
	\| Tokens seen \| 199,884,800 \|
	\| Steps \| 1,525 \|
	\| Tokens / step \| 131,072 \|
	\| Optimizer \| AdamW(0.9, 0.95) wd=0.1 clip=1.0 \|
	\| LR schedule \| warmup 2% + cosine to 10% (peak 4e-03) \|
	\| Final loss \| 3.8176 (ppl 45.5) \|
	\| Wall time \| 5.2 min \|
	\| Trained by \| [@DedeProGames](https://huggingface.co/DedeProGames) \|

	## Usage

	```python
	from transformers import AutoModelForCausalLM, AutoTokenizer

	tok = AutoTokenizer.from_pretrained("DedeProGames/NanoDex-Test-500K-200M")
	model = AutoModelForCausalLM.from_pretrained("DedeProGames/NanoDex-Test-500K-200M")

	ids = tok("The mitochondria is", return_tensors="pt").input_ids
	print(tok.decode(model.generate(ids, max_new_tokens=60, do_sample=True,
	temperature=0.8, top_k=50)[0]))
	```

	## Caveats

	This is a nano-scale research artifact. At this parameter count and token
	budget the model learns word shapes, common collocations and a little syntax —
	it is not a useful assistant and its output is not factual. It exists to make
	"pre-train a transformer from scratch" something you can actually watch happen.