Week 2 Lab Errors Cannot complete the lab

Course Q&A Generative AI with Large Language Models
1

Even after executin step8 I am seeing following error.


nstruct_model = AutoModelForSeq2SeqLM.from_pretrained("./flan-dialogue-summary-checkpoint", torch_dtype=torch.bfloat16)

instruct_model = AutoModelForSeq2SeqLM.from_pretrained(“./flan-dialogue-summary-checkpoint”, torch_dtype=torch.bfloat16)

--------------------------------------------------------------------------- NameError Traceback (most recent call last) in ----> 1 instruct_model = AutoModelForSeq2SeqLM.from_pretrained(“./flan-dialogue-summary-checkpoint”, torch_dtype=torch.bfloat16) NameError: name ‘AutoModelForSeq2SeqLM’ is not defined

1 Like
2

Dear @Yogesh_Kolte,

Have you run all the steps from 1 to 7 before executing step8.
Have you get any error before step8?

Thanks

3

:slight_smile: yes of course…I tried sunning steps multiple times. I get error on step 8. I saw timeout error on step 7 once but it was executed without error of following attempts.

4

Dear @Yogesh_Kolte,

If possible please send me a personal message with your notebook, so that I can have a look, what’s going wrong.

5

image
image1277×197 12.9 KB

Got exactly the same error!
Run the previous cells again to double check! But to no avail.

1 Like
6

You have a number [1] showing to the left of the step. It probably means that this is the first step you ran. It is necessary to execute the steps in order from top to bottom. It should show something like [7] there if it was run in the correct order.

Before, you probably have a step with something like:

from datasets import load_dataset
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer, GenerationConfig, TrainingArguments, Trainer
import torch
import time
import evaluate
import pandas as pd
import numpy as np

This code is necessary to define AutoModelForSeq2SeqLM.

1 Like
7

I had selected wrong aws instance. I was able to resolve the error when I selected instance with 8 core and 32gb memory.

8

image
image1108×349 47.6 KB

Now that I selected the right size, the kernel doesn’t start.
It happened again today after I tried again after a gap of couple of days. Will it affect my working of Lab 3? They say the two are interconnected!

1 Like
9

image
So it is the second instance with the same specifications that must be chosen. No problems and works like magic!

1 Like