feat: Add LLaMA-3 instruct prompt strategies for fine-tuning (#1553)

improve tool handling roles (#1587)

Feat: Add sharegpt multirole (#1137)

Add Glaive conversation format support (#1365)

Standardize system prompt format for AlpacaPrompter (#1190) [skip ci]

fix: remove excessive newlines in system prompt(s) for alpaca (#936)

various bugfixes (#856)

cleanup the old multipack dataloader (#841)

Create preprocess CLI (#785)

misc sharegpt fixes (#723)

use fastchat conversations template (#578)

split completion text to sequence_len (#616)

Prompters: ShareGPT: Allow for custom system prompts

support user defined prompters, pretokenized datasets in config, local parquet, local arrow files (#348)

Error msg for sharegpt if conv has less than 2 msg (#379)

Attention mask and position id fixes for packing (#285)

Fix(message): Improve error message for bad format (#365)

Fixed pre-commit problems, fixed small bug in logging_config to handle LOG_LEVEL env var

Adding logging enhancement

add tests and supoort for loader for sys prompt data

initial wip to get sys prompt from dataset

add typehints

add new sharegpt, refactor prompt so it can be customized later, add exception if no data is processed

Fix sharegpt prompt

Apply isort then black

Fix mypy typing

Remove fixme disable

Remove disable too many arg

Lint prompters.py

Fix: Remove base class inherit for CompletionPrompter

Update src/axolotl/prompters.py

fixes w/ example for super basic lora starter

apply black formatting

more qlora support

optionally be able to specify alpaca or chat style prompts

concise multiple choice and tldr summarize

add alpaca multiple choice instruct dataset support

fix prompters, especially the sharegpt prompter

black formatting

Rename variable to use same convention

Add CompletionPrompt type

Jeopardy bot! (#17)

fix sharegpt tokenization, refactor tokenization debugging

fix sharegpt handling from hf, don't worry about loading llama if using earlier transformers release

WIP large refactor to make finetune script a little more manageable (#3)

add support for alpaca reflect training (#2)

black formatting

make it work with pythia in the cloud

WIP for axolotl trainer