declare-lab / flan-alpaca

This repository contains code for extending the Stanford Alpaca synthetic instruction tuning to existing instruction-tuned models such as Flan-T5.

declare-lab/flan-alpaca Issues

Performance of the model on gsm8k/SVAMP/MultiArith.
Updated a year ago1
Any plan to support trl-peft load_in_8bit for training.py ?
Updated a year ago1
Loss value is NaN
Updated a year ago3
Unable to train on 4-5 gtx 1070s
Updated a year ago10
unable to use new flan-alpaca-gpt4-xl in pipeline
Closed a year ago2
OMP: Error #100: Fatal system error detected.
Updated a year ago1
I dont see any progress logs
Closed a year ago1
Trouble training
Closed a year ago2
RuntimeError: Trying to resize storage that is not resizable
Closed a year ago1
What are the available parameters for generator?
Closed a year ago1
Quantized
Closed a year ago1
Usage example in readme doesn't work
Closed a year ago2
Performance on MMLU
Updated a year ago6
Commercial Use?
Closed a year ago2
wget https://raw.githubusercontent.com/tloen/alpaca-lora/main/alpaca_data_cleaned.json -O data/alpaca_clean.json
Closed a year ago3
Flan Data
Closed a year ago2
In ShareGPT, why the conversation from human is accumulated?
Updated a year ago5
LoRA + FSDP -- issue
Closed a year ago2
issue in training declare-lab/flan-alpaca-base model
Closed a year ago2
is there any plan for flan-ul2?
Updated a year ago6
use gpt4all dataset
Closed a year ago4