robust-armadillo-92

AutoResearch-tinystories-depth8

AutoResearch-tinystories-depth8 is a 285.2M parameter decoder-only Transformer trained from scratch on TinyStories (karpathy/tinystories-gpt4-clean).

This model is part of the AutoResearch project, which focuses on training, evaluating, and releasing efficient language models with reproducible research workflows.

---

Overview

This is a 8-layer decoder-only Transformer trained on the TinyStories (karpathy/tinystories-gpt4-clean) dataset for 0.2 hours of wall-clock training time. The model achieves a validation bits-per-byte (val_bpb) of 0.567264 (perplexity: 1.4817) on the held-out validation set.

---

References

Papers

• NanoGPT / NanoChat architecture patterns

Datasets

• Training: TinyStories (karpathy/tinystories-gpt4-clean)
• Tokenizer: climbmix-400b-shuffle

Related Projects

• karpathy/autoresearch

WANDB Run

• https://wandb.ai/dustinsdelivery-delivery-ai/autoresearch/runs/tl2jijhc

---

Highlights

• Trained from scratch
• 285.2M parameters
• Trained on 19.9M tokens (38 steps)
• 8-layer decoder-only Transformer with sliding window attention
• RoPE positional encoding, RMSNorm, ReLU² activation
• MuonAdamW optimizer (Muon for matrices, AdamW for embeddings)
• Mixture of Experts (8 routed + 1 shared, top-2 routing)
• Hugging Face Transformers compatible

---

Model Architecture

---

Training

This model was trained from scratch for 0.2 hours (620s) of wall-clock training time.

Training Configuration

Hardware

• GPU: NVIDIA GeForce RTX 4060 Ti
• VRAM: 16.0 GB
• Peak VRAM Used: 6.3 GB
• MFU: 40.76%
• Framework: PyTorch 2.9.1+cu128

---

Dataset

• Name: TinyStories (karpathy/tinystories-gpt4-clean)
• Language: English

Preprocessing

Data is packed into fixed-length sequences of 2048 tokens using the nanochat-compatible BPE tokenizer (16,384 vocabulary, 9 special tokens). No additional filtering or deduplication is applied beyond what is in the source dataset.

---

Intended Use

This model is intended for:

• Educational purposes and research
• Text generation experiments
• Studying small language model training dynamics

Not recommended for:

• Production use or safety-critical applications
• Tasks requiring factual accuracy

---

Evaluation

Results

---

Example Generations

Example 1

Prompt

Once upon a time,

Generation

Once upon a time, there was a big bell in the yard. The bell would ring every day and when the bell ring, the bell would ring and the bell would ring when it ringed.
One day, a little boy came to the yard. He saw the bell

Example 2

Prompt

A lonely dragon

Generation

A lonely dragon heard the lonely dragon's lone lonely lonely lonely lonely lone l lonely lonely lonely l lonely lonely lone l l lonely l lone

Example 3

Prompt

The opposite of boy is

Generation

The opposite of boy is sad. He cried and c tears. He lost his favorite toy. He wished he had cared about it. He wished he had a new toy. He wished he had a new toy. But he did not have a new

Example 4

Prompt

The opposite of queen is

Generation

The opposite of queen is kind and shared with everyone. Everyone in the village everyone had kindness. Everyone felt sad and knew they knew the story, they knew about their kindness. And the story has been kindness to the kind queen and the people of the kindness

Example 5

Prompt

My name is

Generation

My name is Pr King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King King

Usage

import torch
import pickle
import json
from train import GPT, GPTConfig, Tokenizer

# Load config
with open('config.json', 'r') as f:
    config_dict = json.load(f)
config = GPTConfig(**{k: v for k, v in config_dict.items() if k in GPTConfig.__dataclass_fields__})

# Load model
model = GPT(config)
state_dict = torch.load('model.pt', map_location='cpu')['state_dict']
model.load_state_dict(state_dict)
model.eval()

# Load tokenizer
with open('tokenizer.pkl', 'rb') as f:
    tokenizer = pickle.load(f)

# Generate
prompt = 'Once upon a time, '
input_ids = tokenizer.encode(prompt)
x = torch.tensor([input_ids], dtype=torch.long)
with torch.no_grad():
    for _ in range(50):
        logits = model(x)
        probs = torch.softmax(logits[:, -1, :] / 0.8, dim=-1)
        next_token = torch.multinomial(probs, num_samples=1)
        input_ids.append(next_token.item())
        x = torch.tensor([input_ids], dtype=torch.long)
print(tokenizer.decode(input_ids))

---

Repository Structure

model.pt                  # Model weights
config.json               # Model architecture config
dataset.txt               # Dataset name used for training
token_bytes.pt            # Token byte mappings
tokenizer.pkl             # Trained BPE tokenizer
tokenizer_config.json     # Tokenizer configuration
training_metrics.json     # Training metrics
README.md                 # This file

---

Limitations

• Small model size limits language understanding and coherence
• Trained on a single dataset (TinyStories) — limited domain
• Fixed time budget training — not fully trained to convergence
• No RLHF or safety alignment

---

Ethical Considerations

• This is a research artifact, not a production model
• The training data consists of synthetic stories (GPT-4 generated)
• No harmful content filtering was applied
• Intended for research and educational use only

---

Citation

@misc{autoresearch_tinystories_depth8,
  title={AutoResearch-tinystories-depth8},
  author={Dustin Loring},
  year={2026},
  howpublished={\url{https://huggingface.co/quik-models/robust-armadillo-92}}
}}

---

Version History

---

Acknowledgements

Built with the AutoResearch training framework.

Thanks to:

• Hugging Face
• PyTorch
• The creators of the TinyStories dataset
• The open-source AI research community

---

License

This model is released under the MIT License unless otherwise specified.