Hugging Face's logo

TRL documentation

Command Line Interfaces (CLIs)

Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

Command Line Interfaces (CLIs)

You can use TRL to fine-tune your Language Model with Supervised Fine-Tuning (SFT) or Direct Policy Optimization (DPO) or even chat with your model using the TRL CLIs.

Currently supported CLIs are:

  • trl sft: fine-tune a LLM on a text/instruction dataset
  • trl dpo: fine-tune a LLM with DPO on a preference dataset
  • trl chat: quickly spin up a LLM fine-tuned for chatting

Fine-tuning with the CLI

Before getting started, pick up a Language Model from Hugging Face Hub. Supported models can be found with the filter “text-generation” within models. Also make sure to pick up a relevant dataset for your task.

Before using the sft or dpo commands make sure to run:

accelerate config

and pick up the right configuration for your training setup (single / multi-GPU, DeepSpeed, etc.). Make sure to complete all steps of accelerate config before running any CLI command.

We also recommend you passing a YAML config file to configure your training protocol. Below is a simple example of a YAML file that you can use for training your models with trl sft command.

model_name_or_path:
  HuggingFaceM4/tiny-random-LlamaForCausalLM
dataset_name:
  imdb
dataset_text_field:
  text
report_to:
  none
learning_rate:
  0.0001
lr_scheduler_type:
  cosine

Save that config in a .yaml and get directly started ! Note you can overwrite the arguments from the config file by explicitly passing them to the CLI, e.g.:

trl sft --config example_config.yaml --output_dir test-trl-cli --lr_scheduler_type cosine_with_restarts

Will force-use cosine_with_restarts for lr_scheduler_type.

Supported Arguments

We do support all arguments from transformers.TrainingArguments, for loading your model, we support all arguments from ~trl.ModelConfig:

class trl.ModelConfig

< >

( model_name_or_path: Optional = None model_revision: str = 'main' torch_dtype: Optional = None trust_remote_code: bool = False attn_implementation: Optional = None use_peft: bool = False lora_r: Optional = 16 lora_alpha: Optional = 32 lora_dropout: Optional = 0.05 lora_target_modules: Optional = None lora_modules_to_save: Optional = None lora_task_type: str = 'CAUSAL_LM' load_in_8bit: bool = False load_in_4bit: bool = False bnb_4bit_quant_type: Optional = 'nf4' use_bnb_nested_quant: bool = False )

Arguments which define the model and tokenizer to load.

You can pass any of these arguments either to the CLI or the YAML file.

Supervised Fine-tuning (SFT)

Follow the basic instructions above and run trl sft --output_dir <output_dir> <*args>:

trl sft --model_name_or_path facebook/opt-125m --dataset_name imdb --output_dir opt-sft-imdb

The SFT CLI is based on the examples/scripts/sft.py script.

Direct Policy Optimization (DPO)

To use the DPO CLI, you need to have a dataset in the TRL format such as

These datasets always have at least three columns prompt, chosen, rejected:

  • prompt is a list of strings.
  • chosen is the chosen response in chat format
  • rejected is the rejected response chat format

To do a quick start, you can run the following command:

trl dpo --model_name_or_path facebook/opt-125m --output_dir trl-hh-rlhf --dataset_name trl-internal-testing/hh-rlhf-trl-style

The DPO CLI is based on the examples/scripts/dpo.py script.

Custom preference dataset

Format the dataset into TRL format (you can adapt the examples/datasets/anthropic_hh.py):

python examples/datasets/anthropic_hh.py --push_to_hub --hf_entity your-hf-org

Chat interface

The chat CLI lets you quickly load the model and talk to it. Simply run the following:

trl chat --model_name_or_path  Qwen/Qwen1.5-0.5B-Chat

To use the chat CLI with the developer installation, you must run make dev

Note that the chat interface relies on the tokenizer’s chat template to format the inputs for the model. Make sure your tokenizer has a chat template defined.

Besides talking to the model there are a few commands you can use:

  • clear: clears the current conversation and start a new one
  • example {NAME}: load example named {NAME} from the config and use it as the user input
  • set {SETTING_NAME}={SETTING_VALUE};: change the system prompt or generation settings (multiple settings are separated by a ’;’).
  • reset: same as clear but also resets the generation configs to defaults if they have been changed by set
  • save {SAVE_NAME} (optional): save the current chat and settings to file by default to ./chat_history/{MODEL_NAME}/chat_{DATETIME}.yaml or {SAVE_NAME} if provided
  • exit: closes the interface

The default examples are defined in examples/scripts/config/default_chat_config.yaml but you can pass your own with --config CONFIG_FILE where you can also specify the default generation parameters.

←Installation PPO Training FAQ→