MicroT-test1-100K-TinyStories

Micro Transformer — Reference Baseline Vanilla Attention • RoPE • Byte-Level • Decoder-Only

📋 Overview

MicroT-test1-100K-TinyStories is a 97,872-parameter vanilla decoder-only transformer — multi-head causal self-attention with RoPE — pretrained on TinyStories as part of the MicroMixer-4 dataset-efficiency comparison study (analysis). TinyStories is a corpus of ~200K short synthetic children's stories (GPT-3.5/4-generated, constrained grammar and vocabulary).

⚠️ This repo is pretrain-only — it is NOT FMSP-fine-tuned. TinyStories contains free-flowing prose with no User:/Assistant: dialogue markers, so the FMSP answer-only cross-entropy recipe does not apply (there is no answer region to isolate). What you get is the raw pretrained backbone (seed 42, V76 recipe, 3 epochs). It continues stories; it does not answer questions or follow instructions.

It is the registered attention-based reference baseline — a deliberately boring standard 2018–2020 transformer recipe (no flash attention, no SwiGLU, no ALiBi, no QKNorm, no sliding window, no MQA/GQA, no MoE). The 100K preset reproduces the champion recipe at its budget.

🏗️ Architecture

graph TD
    A[Byte Input] --> B[Embed 256→48]
    B --> C[Transformer Block × 3]
    C --> D[RMSNorm]
    D --> E[LM Head Tied with Embed]
    E --> F[Byte Output]

    subgraph "Transformer Block (pre-norm)"
        X[Input 48] --> N1[RMSNorm]
        N1 --> AT["MHA 3 heads × d_head 16<br/>RoPE θ=10000 on q,k · causal SDPA"]
        AT --> R1[+ residual]
        R1 --> N2[RMSNorm]
        N2 --> MLP["GELU MLP 48→200→48"]
        MLP --> R2[+ residual]
    end

    style A fill:#007BFF,color:#fff
    style F fill:#00D620,color:#fff
    style AT fill:#FF6600,color:#fff

Model Configuration

Parameter	Value
Hidden Dimension (d_model)	48
Attention Heads	3 (d_head = 16 at every size)
Number of Blocks	3
FFN Hidden	200
Position Encoding	RoPE θ=10000 on q/k only (non-persistent buffers)
Attention	Causal MHA via F.scaled_dot_product_attention(is_causal=True)
Activation	GELU
Biases	None — no bias parameters anywhere
Normalization	RMSNorm (pre-norm)
Max Sequence Length	1024
Vocabulary Size	256 (byte-level)
Output Head	Tied with input embedding

Core Components

┌──────────────────────────────────────────────┐
│         Transformer Block (×3)               │
│  h = h + MHA(RMSNorm(h))    # RoPE q/k, causal│
│  h = h + MLP(RMSNorm(h))    # GELU d→ffn→d   │
│  no biases, no flash, no tricks — vanilla     │
└──────────────────────────────────────────────┘

The d_head=16 contract is hard-asserted across all six sizes so that attention-head behavior is comparable at every budget and never confounds the memorization measurements.

---

🎯 Generation Examples

Story-continuation probes (raw text — no User:/Assistant: prefix, since TinyStories is a plain LM; greedy decoding, repetition_penalty=1.2, no_repeat_ngram_size=4, max_new_tokens=200):

[Prompt] Once upon a time, there was a little girl named Lily.
[Output] She loved to play outside in the park. One day, she was playing with her toys and saw a big box…

_{coherent — on-topic story continuation in the TinyStories register}

[Prompt] One day, a boy named Tom found a big red balloon.
[Output] He was very happy and she wanted to play with his toys. He was very happy and she wanted to pla…

_{drifts — starts as story but falls into a repetitive loop}

[Prompt] The sun was shining and the birds were singing.
[Output] They were sad and sad and saw a big bird and said, "I want to play with me?" The bird said, "I …

_{drifts — starts as story but falls into a repetitive loop}

[Prompt] A little dog ran into the garden and saw
[Output] a big box. It was a big box of colorful box and started to stay in the sky. It was a big box an…

_{drifts — starts as story but falls into a repetitive loop}

---

📊 Results

Pretraining (TinyStories, V76 recipe, 3 epochs)

Metric	1 ep	2 ep	3 ep
Val PPL	2.51	2.43	2.28

AdamW lr 3e-3 · WSD (warmup 500) · wd 0.01 · bs 16 · seq 1024 · seed 42 · plain CE on non-pad bytes.

MicroT-test1 TinyStories family (pretrain-only, all sizes)

Size	Params	3ep Val PPL
1M	996,736	1.69
500K	498,528	1.79
300K	297,680	1.90
100K	97,872	2.28
50K	49,888	2.60
10K	9,808	4.31

_{Pretrain-only family — FMSP-based axes (chatter fluency, full-988 EM, q-relevance, OOD, unanswerable fabrication) are N/A: TinyStories has no User:/Assistant: markers, so the answer-only-CE recipe does not apply.}

---

📚 Training Data

1. Pretraining: TinyStories — synthetic short stories generated by GPT-3.5/4 with a constrained vocabulary and simple grammar, ~200K stories sampled, flattened to 1024-byte sequences, 3 epochs. No User:/Assistant: dialogue structure.

---

🔧 Usage

Files in this repository

• epoch_{0,1,2}.safetensors — per-epoch pretrained backbone weights (pickle-free safetensors). epoch_2.safetensors is the final (3rd-epoch) checkpoint. No FMSP adapter — this is the plain backbone.

Load and generate (local clone)

import torch
from safetensors.torch import load_file
from src.model_v88_transformer import MicroMixerV88Transformer, v88_transformer_100k
from src.tokenizer import ByteTokenizer

# Clone the code repository first:
# git clone https://github.com/llaa33219/MicroMixer-4.git && cd MicroMixer-4

cfg = v88_transformer_100k()
model = MicroMixerV88Transformer(cfg)                         # plain backbone — NO attach_adapter (pretrain-only)
model.load_state_dict(load_file("epoch_2.safetensors"), strict=True)
model.eval()

tok = ByteTokenizer()
prompt = "Once upon a time, there was a little girl named Lily."
ids = tok.encode(prompt)
if ids and ids[-1] == tok.eos_token_id:
    ids = ids[:-1]                       # ByteTokenizer appends EOS; the prompt must end open
ids = torch.tensor([ids])
with torch.no_grad():
    out = model.generate(
        ids, max_new_tokens=200,
        temperature=0.0,                 # greedy
        repetition_penalty=1.2,
        no_repeat_ngram_size=4,
        eos_token_id=tok.eos_token_id,
    )
print(prompt + tok.decode(out[0].tolist()[len(ids):]))

Note the differences from the FMSP cards: (1) no attach_adapter — the backbone is loaded as-is; (2) the prompt is raw story text, not the User: …\n\nAssistant: dialogue format.

Load from Hugging Face Hub (no clone of the weights needed)

import torch
from huggingface_hub import hf_hub_download
from safetensors.torch import load_file
from src.model_v88_transformer import MicroMixerV88Transformer, v88_transformer_100k

REPO = "llaa33219/MicroT-test1-100K-TinyStories"

cfg = v88_transformer_100k()
model = MicroMixerV88Transformer(cfg)
model.load_state_dict(
    load_file(hf_hub_download(REPO, "epoch_2.safetensors")), strict=True)
model.eval()
# ... continue a story as above

---

⚠️ Limitations

Limitation	Description
Pretrain-only — no instruction/QA ability	Not FMSP-fine-tuned; it only continues TinyStories-style prose. It cannot answer questions or follow instructions.
Micro parameters	97,872 parameters; capacity is the binding constraint
Knows only TinyStories	Distribution is synthetic children's stories; no real-world knowledge
Byte-level noise	256-vocab byte tokenizer; PPL not comparable to BPE baselines
Research use only	Architecture/pretraining research artifact, not a production model

---

🧬 Context

This is the 100K TinyStories-pretrained arm of the dataset-efficiency comparison study in the MicroMixer-4 project — the pretrain-only third corpus alongside the UltraChat and SmolTalk2 FMSP arms (TinyStories is excluded from the FMSP/eval battery because it has no User:/Assistant: markers). Sibling repos: llaa33219/{MicroMixer-4,MicroT-test1}-{1M..10K}-TinyStories, plus the UltraChat/SmolTalk2 arms …-{UltraChat,SmolTalk2} and the discord-pretrained baselines llaa33219/{MicroMixer-4,MicroT-test1}-{1M..10K}. Full analysis: DATASET_COMPARISON_ANALYSIS.md.

---

_{Part of the MicroMixer-4 research project — V88 transformer reference (MicroT-test1), 100K preset, TinyStories pretraining (pretrain-only)}