Skip to content

Instantly share code, notes, and snippets.

View sids07's full-sized avatar

Siddhartha Shrestha sids07

View GitHub Profile
@sids07
sids07 / train.py
Created October 16, 2023 14:32
training on multi-gpu
from transformers import (
AutoModelForCausalLM,
AutoTokenizer,
DataCollatorForLanguageModeling,
PreTrainedTokenizer,
Trainer,
TrainingArguments,
set_seed,
get_linear_schedule_with_warmup
)