SparkAI-47m-llama-instruct/README.md

language, license, library_name, tags, pipeline_tag, datasets, base_model

language	license	library_name	tags	pipeline_tag	datasets	base_model
en	apache-2.0	transformers	llama causal-lm text-generation sft chat pytorch fineweb-edu cosmopedia	text-generation	HuggingFaceFW/fineweb-edu HuggingFaceTB/cosmopedia-v2	vedantjadhav701/SparkAI-47m-llama-10b-token

SparkAI-47M-Llama-Instruct

Instruction-tuned checkpoint of SparkAI-47M-Llama (~48M parameter decoder-only transformer), fine-tuned for chat and instruction following.

🌟 Key Highlights & Unique Features

• ⚡ Ultra-Low Memory Footprint (~95.4 MB): Fits in under 100MB of RAM, making it suitable for edge devices, mobile apps, WebGPU, and microcontrollers.
• 🏋️ Data-Saturated Pretraining (10 Billion Tokens): Pretrained on 10B tokens (210 tokens/param) of high-quality FineWeb-Edu + Cosmopedia-v2 text, providing an empirical benchmark on capacity saturation for sub-50M models.
• 🏗️ Modern LLaMA 3 Architecture: Built with Grouped Query Attention (GQA), SwiGLU activations, RoPE positional encodings, RMSNorm pre-normalization, and tied embeddings.
• 💬 Full ChatML SFT Alignment: Fine-tuned with ChatML <|im_start|> instruction formatting and template support (chat_template.jinja).

---

📐 Architecture

• Parameters: ~48M (~47.4M non-embedding / tied embeddings)
• Layers: 8
• Hidden Size: 512
• Attention Heads: 8 query heads, 2 key/value heads (Grouped Query Attention - GQA)
• MLP: SwiGLU (Intermediate size: 1408)
• Positional Encoding: RoPE (Rotary Position Embeddings)
• Normalization: RMSNorm (Pre-normalization)
• Embeddings: Tied embeddings, no bias terms
• Vocabulary Size: 49,152 (SmolLM2 tokenizer with Chat Template)
• Sequence Length: 1024

---

🏋️ Training & Fine-Tuning Details

• Base Checkpoint: vedantjadhav701/SparkAI-47m-llama-10b-token
• Pretraining Data: FineWeb-Edu (sample-100BT) + Cosmopedia-v2 (85% / 15% mix, 10.00B tokens)
• Optimizer: AdamW with Cosine LR schedule + warmup
• Hardware: NVIDIA A100 80GB PCIe
• Final Eval Perplexity: 31.46

📈 Pretraining Progression

Tokens	Perplexity
630M	43.49
3.77B	31.30
7.00B	—
10.00B	31.46

Note on Saturation: Perplexity plateaued between 3.77B and 10.00B tokens despite continued training, indicating the model has saturated its representational capacity at this size.

---

📊 Comparison in Sub-50M Parameter Landscape

Feature	Typical Sub-50M Models	SparkAI-47M-Llama / Instruct
Token Budget	~1B – 2B tokens	10.00 Billion Tokens (210 tokens/param)
Data Quality	Raw web text / C4	FineWeb-Edu (85%) + Cosmopedia-v2 (15%)
Architecture	Basic MHA / GPT-2 style	Modern LLaMA 3 (GQA, SwiGLU, RoPE, RMSNorm)
Model Size	~100MB – 200MB	~95.4 MB (model.safetensors)
SFT Alignment	Rare / None	Instruction-tuned with ChatML (chat_template.jinja)
Benchmarking	Few metrics	Empirical capacity saturation documented at 10B tokens

---

💻 Usage with Hugging Face transformers

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "vedantjadhav701/SparkAI-47m-llama-instruct"
tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(repo_id)

messages = [
    {"role": "user", "content": "What is a computer program?"}
]

prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=80, do_sample=True, temperature=0.6)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

---

🚀 Quickstart & Local Gradio UI

Installation

pip install -r requirements.txt

Run Gradio App

python app.py

Open http://127.0.0.1:7860 in your web browser.