llama3_1b_chinchilla_8132025

Model Details

This model is a 1B llama3 model pretrained from scratch with torchtitan on fineweb-edu with AdamW optimizer. 20x chinchilla rule for 20B tokens seen.

How to use

import torch
from transformers import pipeline


pipe = pipeline(
    "text-generation",
    model="kz919/llama3_1b_chinchilla_8132025",
)

print(pipe("The key to life is"))

Downstream Eval

ARC, Hellaswag, Lambda_OpenAI, OpenbookQA, PIQA

lm_eval --model hf --model_args pretrained=kz919/llama3_1b_chinchilla_8142025,dtype="bfloat16",add_bos_token=True --tasks lambada_openai,hellaswag,piqa,arc_easy,arc_challenge,openbookqa --device cuda:7 --batch_size 8

MMLU