Here’s I tried to finetune gpt2 tempdeltavalue/temp_l/blob/main/finetune_seq2seq.ipynb { "cells": [ { "cell_type": "code", "execution_count": 1, "id": "225b0a06-4dd3-4c84-89ed-86de4adcd24b", "metadata": {}, "outputs": [ { "name": "stderr", "output_type…

GPT2 returns sequence of <|endoftext|> after finetuning

tempdeltavalue January 27, 2024, 12:20am 2

upd:

It’s reproducible even without finetuning

Topic		Replies	Views
How does GPT decide to stop generating sentences without EOS token? 🤗Transformers	13	24519	August 19, 2024
Issue with finetuning a seq-to-seq model 🤗Transformers	30	3957	August 11, 2022
Infinity output from gpt2 model? Beginners	2	153	June 22, 2024
GPT2 finetuned with eos token will never yield eos token during generation Beginners	6	3374	April 12, 2024
[Help appreciated] GPT2 Finetuning results in Only Padding output 🤗Transformers	2	1613	June 5, 2023