6778856804
* feat: add num_proc and load from cache for rl mapping
* fix: refactor sft and rl trainer to set same base args
* feat: add report_to to set run name
* fix: consolidate handling of fp16, bf16, tf32 kwarg
* chore: consolidate eval_strat, loraplus, lr sched, max_length
* fix: deprecate old types
* fix: adding missing Any
* fix: max_steps incorrectly set
* fix: remove unnecessary datacollator kwarg insert and pop
* fix: update default max_steps
* fix: add missing weight_decay handling
* fix: ignore max_length for grpo
* feat: update CI on trainer_builder
* fix: comments
* improve handling of warmup/logging steps
* use transformers default for logging steps, not None
* fix: remove redundant override
* fix: lint
* feat: allow custom optim for rl methods
* fix: duplicate optim setting
* fix(test): set sequence_parallel_degree default in base cfg
* feat: add handling for seed and SP/ring-attn config
* chore: add back return typing from rebase
* fix(test): use RLType directly to skip needing to validate
* feat: split training builder into sub modules
* fix: remove deprecated clause
* chore: add missing config to doc
* fix: update quarto autodoc
* fix: import path for trainer builder and submodules
* fix: remove redundant configs from rebase mistake
* chore: simplify dynamo check
* fix: optimizer_cls_and_kwargs to be passed into trainer_kwargs
* fix: add missing rex from rebase
* fix: move pop optimizer_cls_and_kwargs
* fix: pop optimizer cls in rl too
* fix: leftover bug from rebase
* fix: update handling of trainer_cls in RL
* fix: address pr feedback
* feat: call hook_pre_create_trainer for rl
* chore: lint
* fix: return notimplemented for ppo
* feat: moved torch compile to base and refactor collator setting
* chore: remove unused importlib.util import
* fix: optimizer cls not being popped
* feat: move epoch setting to base
* fix: catch unhandled custom optimizer
* fix: remove duplicate lora plus setting
* chore: refactor if condition
* chore: refactor set_base_training_args into smaller modules
* fix: address TrainerBuilderBase class variables to instance var
* fix: add handling for beta3 and episilon2
* fix: change to pass dict via arg instead of updating dict
* chore: simplify if condition
* fix: force access to lr & weight decay in case not provided to early error
* fix: remove log sweep
* chore: refactor if condition
* fix: address renamed cfg
* fix: improve handling of cosine hyp
* fix: remove unused params
* chore: refactor
* chore: clarify doc safetensors
* fix: update import path to be unified following comments
* fix: duplicate kwargs passed
* feat: return separate trainer_kwargs
* chore: refactor
* chore: refactor based on comments
* chore: refactor based on comments
* fix: move gpustats callback to base
* chore: create trainer_cls_args first based on comments
* fix: ipo label smoothing passed incorrectly
* feat: add optimizer parity for RL methods with test
* feat: add parity for optimizer in RM/PRM and add test
* fix: remove redundant function override for orpo/cpo batch metrics
* fix: improve handling of dpo_label_smoothing and merge issue
* fix: test fixture returning wrong field
* fix: address avoid direct modify fixture
* chore: minor refactor
* Revert "chore: refactor"
This reverts commit 99c8859eb0.
* feat: rename trainer_builder to builders
---------
Co-authored-by: Wing Lian <wing@axolotl.ai>
310 lines
8.8 KiB
YAML
310 lines
8.8 KiB
YAML
project:
|
|
type: website
|
|
|
|
quartodoc:
|
|
dir: docs/api
|
|
package: axolotl
|
|
title: API Reference
|
|
parser: google
|
|
|
|
sections:
|
|
- title: Core
|
|
desc: Core functionality for training
|
|
contents:
|
|
- train
|
|
- evaluate
|
|
- datasets
|
|
- convert
|
|
- prompt_tokenizers
|
|
- logging_config
|
|
- core.builders.base
|
|
- core.builders.causal
|
|
- core.builders.rl
|
|
- core.training_args
|
|
- core.chat.messages
|
|
- core.chat.format.chatml
|
|
- core.chat.format.llama3x
|
|
- core.chat.format.shared
|
|
- core.datasets.chat
|
|
- core.datasets.transforms.chat_builder
|
|
- title: CLI
|
|
desc: Command-line interface
|
|
contents:
|
|
- cli.main
|
|
- cli.train
|
|
- cli.evaluate
|
|
- cli.args
|
|
- cli.checks
|
|
- cli.config
|
|
- cli.inference
|
|
- cli.merge_lora
|
|
- cli.merge_sharded_fsdp_weights
|
|
- cli.preprocess
|
|
- cli.sweeps
|
|
- cli.utils
|
|
- cli.vllm_serve
|
|
- cli.cloud.base
|
|
- cli.cloud.modal_
|
|
- cli.quantize
|
|
- title: Trainers
|
|
desc: Training implementations
|
|
contents:
|
|
- core.trainers.base
|
|
- core.trainers.trl
|
|
- core.trainers.mamba
|
|
- core.trainers.relora
|
|
- core.trainers.dpo.trainer
|
|
- core.trainers.grpo.trainer
|
|
- core.trainers.grpo.sampler
|
|
- core.trainers.utils
|
|
- title: Model Loading
|
|
desc: Functionality for loading and patching models, tokenizers, etc.
|
|
contents:
|
|
- loaders.model
|
|
- loaders.tokenizer
|
|
- loaders.processor
|
|
- loaders.adapter
|
|
- loaders.patch_manager
|
|
- loaders.constants
|
|
- title: Mixins
|
|
desc: Mixin classes for augmenting trainers
|
|
contents:
|
|
- core.trainers.mixins.optimizer
|
|
- core.trainers.mixins.rng_state_loader
|
|
- core.trainers.mixins.scheduler
|
|
- title: Context Managers
|
|
desc: Context managers for altering trainer behaviors
|
|
contents:
|
|
- utils.ctx_managers.sequence_parallel
|
|
- title: Prompt Strategies
|
|
desc: Prompt formatting strategies
|
|
contents:
|
|
- prompt_strategies.base
|
|
- prompt_strategies.chat_template
|
|
- prompt_strategies.alpaca_chat
|
|
- prompt_strategies.alpaca_instruct
|
|
- prompt_strategies.alpaca_w_system
|
|
- prompt_strategies.user_defined
|
|
- prompt_strategies.llama2_chat
|
|
- prompt_strategies.completion
|
|
- prompt_strategies.input_output
|
|
- prompt_strategies.stepwise_supervised
|
|
- prompt_strategies.metharme
|
|
- prompt_strategies.orcamini
|
|
- prompt_strategies.pygmalion
|
|
- prompt_strategies.messages.chat
|
|
- prompt_strategies.dpo.chat_template
|
|
- prompt_strategies.dpo.llama3
|
|
- prompt_strategies.dpo.chatml
|
|
- prompt_strategies.dpo.zephyr
|
|
- prompt_strategies.dpo.user_defined
|
|
- prompt_strategies.dpo.passthrough
|
|
- prompt_strategies.kto.llama3
|
|
- prompt_strategies.kto.chatml
|
|
- prompt_strategies.kto.user_defined
|
|
- prompt_strategies.orpo.chat_template
|
|
- prompt_strategies.bradley_terry.llama3
|
|
- title: Kernels
|
|
desc: Low-level performance optimizations
|
|
contents:
|
|
- kernels.lora
|
|
- kernels.geglu
|
|
- kernels.swiglu
|
|
- kernels.quantize
|
|
- kernels.utils
|
|
- title: Monkey Patches
|
|
desc: Runtime patches for model optimizations
|
|
contents:
|
|
- monkeypatch.llama_attn_hijack_flash
|
|
- monkeypatch.llama_attn_hijack_xformers
|
|
- monkeypatch.mistral_attn_hijack_flash
|
|
- monkeypatch.multipack
|
|
- monkeypatch.relora
|
|
- monkeypatch.llama_expand_mask
|
|
- monkeypatch.lora_kernels
|
|
- monkeypatch.utils
|
|
- monkeypatch.btlm_attn_hijack_flash
|
|
- monkeypatch.llama_patch_multipack
|
|
- monkeypatch.stablelm_attn_hijack_flash
|
|
- monkeypatch.trainer_fsdp_optim
|
|
- monkeypatch.transformers_fa_utils
|
|
- monkeypatch.unsloth_
|
|
- monkeypatch.attention.mllama
|
|
- monkeypatch.data.batch_dataset_fetcher
|
|
- monkeypatch.mixtral
|
|
- monkeypatch.gradient_checkpointing.offload_cpu
|
|
- monkeypatch.gradient_checkpointing.offload_disk
|
|
- title: Utils
|
|
desc: Utility functions
|
|
contents:
|
|
- utils.tokenization
|
|
- utils.chat_templates
|
|
- utils.lora
|
|
- utils.model_shard_quant
|
|
- utils.bench
|
|
- utils.freeze
|
|
- utils.trainer
|
|
- utils.schedulers
|
|
- utils.distributed
|
|
- utils.dict
|
|
- utils.optimizers.adopt
|
|
- utils.data.pretraining
|
|
- utils.data.sft
|
|
- utils.quantization
|
|
- title: Schemas
|
|
desc: Pydantic data models for Axolotl config
|
|
contents:
|
|
- utils.schemas.config
|
|
- utils.schemas.model
|
|
- utils.schemas.training
|
|
- utils.schemas.datasets
|
|
- utils.schemas.peft
|
|
- utils.schemas.trl
|
|
- utils.schemas.multimodal
|
|
- utils.schemas.integrations
|
|
- utils.schemas.enums
|
|
- utils.schemas.utils
|
|
- title: Integrations
|
|
desc: Third-party integrations and extensions
|
|
contents:
|
|
- integrations.base
|
|
- integrations.cut_cross_entropy.args
|
|
- integrations.grokfast.optimizer
|
|
- integrations.kd.trainer
|
|
- integrations.liger.args
|
|
- integrations.lm_eval.args
|
|
- integrations.spectrum.args
|
|
- title: Common
|
|
desc: Common utilities and shared functionality
|
|
contents:
|
|
- common.architectures
|
|
- common.const
|
|
- common.datasets
|
|
- title: Models
|
|
desc: Custom model implementations
|
|
contents:
|
|
- models.mamba.modeling_mamba
|
|
- title: Data Processing
|
|
desc: Data processing utilities
|
|
contents:
|
|
- utils.collators.core
|
|
- utils.collators.batching
|
|
- utils.collators.mamba
|
|
- utils.collators.mm_chat
|
|
- utils.samplers.multipack
|
|
- title: Callbacks
|
|
desc: Training callbacks
|
|
contents:
|
|
- utils.callbacks.perplexity
|
|
- utils.callbacks.profiler
|
|
- utils.callbacks.lisa
|
|
- utils.callbacks.mlflow_
|
|
- utils.callbacks.comet_
|
|
- utils.callbacks.qat
|
|
website:
|
|
title: "Axolotl"
|
|
description: "We make fine-tuning accessible, scalable, and fun"
|
|
favicon: favicon.jpg
|
|
|
|
google-analytics: "G-9KYCVJBNMQ"
|
|
|
|
navbar:
|
|
logo: image/axolotl_logo_digital_white.svg
|
|
title: false
|
|
background: dark
|
|
pinned: false
|
|
collapse: false
|
|
tools:
|
|
- icon: twitter
|
|
href: https://twitter.com/axolotl_ai
|
|
- icon: github
|
|
href: https://github.com/axolotl-ai-cloud/axolotl/
|
|
- icon: discord
|
|
href: https://discord.gg/7m9sfhzaf3
|
|
|
|
sidebar:
|
|
pinned: true
|
|
collapse-level: 2
|
|
style: docked
|
|
contents:
|
|
- text: Home
|
|
href: index.qmd
|
|
|
|
- section: "Getting Started"
|
|
contents:
|
|
- docs/getting-started.qmd
|
|
- docs/installation.qmd
|
|
- docs/inference.qmd
|
|
- docs/cli.qmd
|
|
- docs/config.qmd
|
|
- text: "API Reference"
|
|
href: docs/api
|
|
|
|
- section: "Dataset Formats"
|
|
contents: docs/dataset-formats/*
|
|
|
|
- section: "Deployments"
|
|
contents:
|
|
- docs/docker.qmd
|
|
- docs/multi-gpu.qmd
|
|
- docs/multi-node.qmd
|
|
- docs/ray-integration.qmd
|
|
- docs/amd_hpc.qmd
|
|
- docs/mac.qmd
|
|
|
|
- section: "How To Guides"
|
|
contents:
|
|
- docs/multimodal.qmd
|
|
- docs/rlhf.qmd
|
|
- docs/reward_modelling.qmd
|
|
- docs/lr_groups.qmd
|
|
- docs/lora_optims.qmd
|
|
- docs/dataset_loading.qmd
|
|
- docs/qat.qmd
|
|
- docs/quantize.qmd
|
|
|
|
- section: "Core Concepts"
|
|
contents:
|
|
- docs/batch_vs_grad.qmd
|
|
- docs/dataset_preprocessing.qmd
|
|
- docs/multipack.qmd
|
|
|
|
- section: "Advanced Features"
|
|
contents:
|
|
- docs/fsdp_qlora.qmd
|
|
- docs/unsloth.qmd
|
|
- docs/torchao.qmd
|
|
- docs/custom_integrations.qmd
|
|
- docs/sequence_parallelism.qmd
|
|
|
|
- section: "Troubleshooting"
|
|
contents:
|
|
- docs/faq.qmd
|
|
- docs/debugging.qmd
|
|
- docs/nccl.qmd
|
|
|
|
format:
|
|
html:
|
|
theme: darkly
|
|
css: styles.css
|
|
toc: true
|
|
# Enable better handling of line breaks in markdown
|
|
preserve-tabs: true
|
|
html-math-method: mathjax
|
|
# Improved markdown processing options
|
|
md-extensions:
|
|
- markdown_it
|
|
- def_list
|
|
- attr_list
|
|
- fenced_divs
|
|
- tables
|
|
- html_admonition
|
|
- lineblocks
|
|
- fancy_lists
|
|
# Control whitespace handling
|
|
whitespace: preserve
|
|
# Process newlines in paragraphs
|
|
wrap: preserve
|
|
# Better line break handling
|
|
preserve-linebreaks: true
|