TinyStories-Korean-800K

A tiny autoregressive language model trained from scratch.

• Architecture: Llama
• Vocab size: 4096
• Hidden size: 64
• Layers: 5
• Heads: 8 (MHA)
• Context length: up to 512 tokens

Note

This model was trained for experimental purposes and the generated output may not make any sense at all.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('northwind33/TinyStories-Korean-800K')
tokenizer = AutoTokenizer.from_pretrained('northwind33/TinyStories-Korean-800K')

input_text = ''
input_ids = tokenizer(input_text, return_tensors='pt').input_ids

output = model.generate(input_ids, max_length=512, do_sample=True, temperature=0.5)