Add configs and stuff (#2)
Esse commit está contido em:
lewtun
+5
-5
@@ -3,7 +3,7 @@
|
||||
# make sure to test the local checkout in scripts and not the pre-installed one (don't use quotes!)
|
||||
export PYTHONPATH = src
|
||||
|
||||
check_dirs := src tests scripts
|
||||
check_dirs := src scripts
|
||||
|
||||
style:
|
||||
black --line-length 119 --target-version py310 $(check_dirs) setup.py
|
||||
@@ -18,16 +18,16 @@ quality:
|
||||
# Release stuff
|
||||
|
||||
pre-release:
|
||||
python src/alignment/release.py
|
||||
python src/open_r1/release.py
|
||||
|
||||
pre-patch:
|
||||
python src/alignment/release.py --patch
|
||||
python src/open_r1/release.py --patch
|
||||
|
||||
post-release:
|
||||
python src/alignment/release.py --post_release
|
||||
python src/open_r1/release.py --post_release
|
||||
|
||||
post-patch:
|
||||
python src/alignment/release.py --post_release --patch
|
||||
python src/open_r1/release.py --post_release --patch
|
||||
|
||||
wheels:
|
||||
python setup.py bdist_wheel && python setup.py sdist
|
||||
|
||||
+43
-1
@@ -1 +1,43 @@
|
||||
# open_r1
|
||||
# Open R1
|
||||
|
||||
## Installation instructions
|
||||
|
||||
To run the code in this project, first, create a Python virtual environment using e.g. Conda:
|
||||
|
||||
```shell
|
||||
conda create -n openr1 python=3.11 && conda activate openr1
|
||||
```
|
||||
|
||||
Next, install vLLM:
|
||||
|
||||
```shell
|
||||
pip install vllm==0.6.6.post1
|
||||
|
||||
# For HF (cluster only has CUDA 12.1)
|
||||
pip install vllm==0.6.6.post1 --extra-index-url https://download.pytorch.org/whl/cu121
|
||||
```
|
||||
|
||||
This will also install PyTorch `v2.5.1` and it is **very important** to use this version since the vLLM binaries are compiled for it. You can then install the remaining dependencies for your specific use case via `pip install -e .[LIST OF MODES]`. For most contributors, we recommend:
|
||||
|
||||
```shell
|
||||
pip install -e ".[dev]"
|
||||
```
|
||||
|
||||
Next, log into your Hugging Face and Weights and Biases accounts as follows:
|
||||
|
||||
```shell
|
||||
huggingface-cli login
|
||||
wandb login
|
||||
```
|
||||
|
||||
Finally, check your system has Git LFS installed so that you can load and push models/datasets to the Hugging Face Hub:
|
||||
|
||||
```shell
|
||||
git-lfs --version
|
||||
```
|
||||
|
||||
If it isn't installed, run:
|
||||
|
||||
```shell
|
||||
sudo apt-get install git-lfs
|
||||
```
|
||||
@@ -0,0 +1,45 @@
|
||||
# Model arguments
|
||||
model_name_or_path: Qwen/Qwen2.5-1.5B-Instruct
|
||||
model_revision: main
|
||||
torch_dtype: bfloat16
|
||||
attn_implementation: flash_attention_2
|
||||
|
||||
# Data training arguments
|
||||
dataset_mixer:
|
||||
HuggingFaceH4/Bespoke-Stratos-17k: 1.0
|
||||
dataset_splits:
|
||||
- train
|
||||
- test
|
||||
preprocessing_num_workers: 12
|
||||
|
||||
# SFT trainer config
|
||||
bf16: true
|
||||
do_eval: true
|
||||
eval_strategy: epoch
|
||||
gradient_accumulation_steps: 1
|
||||
gradient_checkpointing: true
|
||||
gradient_checkpointing_kwargs:
|
||||
use_reentrant: False
|
||||
hub_model_id: HuggingFaceH4/Qwen2.5-1.5B-R1-v00.00
|
||||
hub_strategy: every_save
|
||||
learning_rate: 2.0e-05
|
||||
log_level: info
|
||||
logging_steps: 5
|
||||
logging_strategy: steps
|
||||
lr_scheduler_type: cosine
|
||||
max_seq_length: 2048
|
||||
max_steps: -1
|
||||
num_train_epochs: 1
|
||||
output_dir: data/Qwen2.5-1.5B-Distill-R1-v00.00
|
||||
overwrite_output_dir: true
|
||||
per_device_eval_batch_size: 8
|
||||
per_device_train_batch_size: 16
|
||||
push_to_hub: true
|
||||
remove_unused_columns: true
|
||||
report_to:
|
||||
- wandb
|
||||
save_strategy: "steps"
|
||||
save_steps: 100
|
||||
save_total_limit: 1
|
||||
seed: 42
|
||||
warmup_ratio: 0.1
|
||||
@@ -0,0 +1,22 @@
|
||||
compute_environment: LOCAL_MACHINE
|
||||
debug: false
|
||||
deepspeed_config:
|
||||
deepspeed_multinode_launcher: standard
|
||||
offload_optimizer_device: none
|
||||
offload_param_device: none
|
||||
zero3_init_flag: true
|
||||
zero3_save_16bit_model: true
|
||||
zero_stage: 3
|
||||
distributed_type: DEEPSPEED
|
||||
downcast_bf16: 'no'
|
||||
machine_rank: 0
|
||||
main_training_function: main
|
||||
mixed_precision: bf16
|
||||
num_machines: 1
|
||||
num_processes: 8
|
||||
rdzv_backend: static
|
||||
same_network: true
|
||||
tpu_env: []
|
||||
tpu_use_cluster: false
|
||||
tpu_use_sudo: false
|
||||
use_cpu: false
|
||||
@@ -0,0 +1,26 @@
|
||||
compute_environment: LOCAL_MACHINE
|
||||
debug: false
|
||||
distributed_type: FSDP
|
||||
downcast_bf16: 'no'
|
||||
enable_cpu_affinity: false
|
||||
fsdp_config:
|
||||
fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
|
||||
fsdp_backward_prefetch: BACKWARD_PRE
|
||||
fsdp_cpu_ram_efficient_loading: true
|
||||
fsdp_forward_prefetch: true
|
||||
fsdp_offload_params: false
|
||||
fsdp_sharding_strategy: FULL_SHARD
|
||||
fsdp_state_dict_type: SHARDED_STATE_DICT
|
||||
fsdp_sync_module_states: true
|
||||
fsdp_use_orig_params: true
|
||||
machine_rank: 0
|
||||
main_training_function: main
|
||||
mixed_precision: bf16
|
||||
num_machines: 1
|
||||
num_processes: 8
|
||||
rdzv_backend: static
|
||||
same_network: true
|
||||
tpu_env: []
|
||||
tpu_use_cluster: false
|
||||
tpu_use_sudo: false
|
||||
use_cpu: false
|
||||
@@ -0,0 +1,25 @@
|
||||
compute_environment: LOCAL_MACHINE
|
||||
debug: false
|
||||
distributed_type: FSDP
|
||||
downcast_bf16: 'no'
|
||||
fsdp_config:
|
||||
fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
|
||||
fsdp_backward_prefetch: BACKWARD_PRE
|
||||
fsdp_cpu_ram_efficient_loading: true
|
||||
fsdp_forward_prefetch: false
|
||||
fsdp_offload_params: true
|
||||
fsdp_sharding_strategy: FULL_SHARD
|
||||
fsdp_state_dict_type: SHARDED_STATE_DICT
|
||||
fsdp_sync_module_states: true
|
||||
fsdp_use_orig_params: false
|
||||
machine_rank: 0
|
||||
main_training_function: main
|
||||
mixed_precision: 'no'
|
||||
num_machines: 1
|
||||
num_processes: 2
|
||||
rdzv_backend: static
|
||||
same_network: true
|
||||
tpu_env: []
|
||||
tpu_use_cluster: false
|
||||
tpu_use_sudo: false
|
||||
use_cpu: false
|
||||
@@ -0,0 +1,16 @@
|
||||
compute_environment: LOCAL_MACHINE
|
||||
debug: false
|
||||
distributed_type: MULTI_GPU
|
||||
downcast_bf16: 'no'
|
||||
gpu_ids: all
|
||||
machine_rank: 0
|
||||
main_training_function: main
|
||||
mixed_precision: bf16
|
||||
num_machines: 1
|
||||
num_processes: 8
|
||||
rdzv_backend: static
|
||||
same_network: true
|
||||
tpu_env: []
|
||||
tpu_use_cluster: false
|
||||
tpu_use_sudo: false
|
||||
use_cpu: false
|
||||
@@ -0,0 +1,86 @@
|
||||
#!/bin/bash
|
||||
#SBATCH --ntasks-per-node=1
|
||||
#SBATCH --exclusive
|
||||
#SBATCH --gres=gpu:8
|
||||
#SBATCH --partition=hopper-prod # Adjust this for your cluster
|
||||
#SBATCH --output=/fsx/h4/logs/%x-%j.out # Adjust this for your cluster
|
||||
#SBATCH --err=/fsx/h4/logs/%x-%j.err # Adjust this for your cluster
|
||||
|
||||
set -x -e
|
||||
|
||||
source ~/.bashrc
|
||||
conda activate openr1
|
||||
echo "START TIME: $(date)"
|
||||
|
||||
MODEL=$1
|
||||
TASK=$2
|
||||
PRECISION=$3
|
||||
ACCELERATOR=$4
|
||||
OPTIONAL_ARGS=$5
|
||||
|
||||
# Training setup
|
||||
NUM_NODES=$SLURM_NNODES
|
||||
GPUS_PER_NODE=8
|
||||
WORLD_SIZE=$(($NUM_NODES*$GPUS_PER_NODE))
|
||||
# Due to conflicts between Accelerate's DeepSpeed configs and Transformers' TrainingArguments, we need to parse the gradient accumulation steps from the config file to ensure they match
|
||||
CONFIG_FILE=recipes/$MODEL/$TASK/config_$PRECISION.yaml
|
||||
GRAD_ACC_STEPS=$(grep 'gradient_accumulation_steps' $CONFIG_FILE | awk '{print $2}')
|
||||
|
||||
# Split the string into individual arguments
|
||||
IFS=' ' read -ra ARGS <<< "$OPTIONAL_ARGS"
|
||||
|
||||
# Loop through the arguments and find the one with "--gradient_accumulation_steps"
|
||||
for arg in "${ARGS[@]}"; do
|
||||
if [[ "$arg" == "--gradient_accumulation_steps="* ]]; then
|
||||
# Extract the value after the equals sign
|
||||
GRAD_ACC_STEPS="${arg#*=}"
|
||||
break # Exit the loop once we find the desired argument
|
||||
fi
|
||||
done
|
||||
|
||||
echo "Gradient accumulation steps: $GRAD_ACC_STEPS"
|
||||
# so processes know who to talk to
|
||||
MASTER_ADDR=$(scontrol show hostnames $SLURM_JOB_NODELIST | head -n 1)
|
||||
MASTER_PORT=6000
|
||||
|
||||
export CMD=" \
|
||||
scripts/run_$TASK.py $CONFIG_FILE $OPTIONAL_ARGS
|
||||
"
|
||||
|
||||
export LAUNCHER="HF_HUB_ENABLE_HF_TRANSFER=1 ACCELERATE_LOG_LEVEL=info TRANSFORMERS_VERBOSITY=info accelerate launch \
|
||||
--config_file recipes/accelerate_configs/$ACCELERATOR.yaml \
|
||||
--gradient_accumulation_steps $GRAD_ACC_STEPS \
|
||||
--num_machines $NUM_NODES \
|
||||
--num_processes $WORLD_SIZE \
|
||||
--main_process_ip $MASTER_ADDR \
|
||||
--main_process_port $MASTER_PORT \
|
||||
--machine_rank \$SLURM_PROCID \
|
||||
--rdzv_conf "rdzv_backend=c10d,rdzv_endpoint=$MASTER_ADDR:$MASTER_PORT" \
|
||||
--max_restarts 1 \
|
||||
--role \$(hostname -s): \
|
||||
--tee 3 \
|
||||
"
|
||||
|
||||
# force crashing on nccl issues like hanging broadcast
|
||||
export NCCL_ASYNC_ERROR_HANDLING=1
|
||||
# export NCCL_DEBUG=INFO
|
||||
# export NCCL_DEBUG_SUBSYS=COLL
|
||||
# export NCCL_SOCKET_NTHREADS=1
|
||||
# export NCCL_NSOCKS_PERTHREAD=1
|
||||
# export CUDA_LAUNCH_BLOCKING=1
|
||||
|
||||
# Specific configuration optimized for the Hugging Face Compute Cluster
|
||||
# Be ye warned this may not work on other clusters!
|
||||
module load cuda/12.1
|
||||
|
||||
# srun error handling:
|
||||
# --wait=60: wait 60 sec after the first task terminates before terminating all remaining tasks
|
||||
# --kill-on-bad-exit=1: terminate a step if any task exits with a non-zero exit code
|
||||
SRUN_ARGS=" \
|
||||
--wait=60 \
|
||||
--kill-on-bad-exit=1 \
|
||||
"
|
||||
|
||||
clear; srun $SRUN_ARGS --jobid $SLURM_JOB_ID bash -c "$LAUNCHER --role \$SLURMD_NODENAME: $CMD" 2>&1
|
||||
|
||||
echo "END TIME: $(date)"
|
||||
@@ -1,141 +1 @@
|
||||
# Scripts to Train and Evaluate Chat Models
|
||||
|
||||
## Fine-tuning
|
||||
|
||||
In the handbook, we provide three main ways to align LLMs for chat:
|
||||
|
||||
- Full fine-tuning on a multi-GPU machine with DeepSpeed ZeRO-3 (tested on an 8 x A100 (80GB) node).
|
||||
- LoRA or QLoRA fine-tuning on a single consumer 24GB GPU (tested on an RTX 4090).
|
||||
- LoRA fine-tuning on a multi-GPU machine with DeepSpeed ZeRO-3 (tested on a 2 x A100s (80GB)).
|
||||
- QLoRA fine-tuning on multi-GPU machine with FSDP (tested on a 2 x A6000s (48GB)).
|
||||
|
||||
In practice, we find comparable performance for both full and QLoRA fine-tuning, with the latter having the advantage of producing small adapter weights that are fast to upload and download from the Hugging Face Hub. Here are the general commands to fine-tune your models:
|
||||
|
||||
```shell
|
||||
# Full training with ZeRO-3 on 8 GPUs
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/deepspeed_zero3.yaml scripts/run_{task}.py recipes/{model_name}/{task}/config_full.yaml
|
||||
|
||||
# QLoRA 4-bit training on a single GPU
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/multi_gpu.yaml --num_processes=1 scripts/run_{task}.py recipes/{model_name}/{task}/config_qlora.yaml
|
||||
|
||||
# LoRA training on a single GPU
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/multi_gpu.yaml --num_processes=1 scripts/run_{task}.py recipes/{model_name}/{task}/config_qlora.yaml --load_in_4bit=false
|
||||
|
||||
# LoRA training with ZeRO-3 on two or more GPUs
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/deepspeed_zero3.yaml --num_processes={num_gpus} scripts/run_{task}.py recipes/{model_name}/{task}/config_qlora.yaml --load_in_4bit=false
|
||||
|
||||
# QLoRA training with FSDP on two or more GPUs
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/fsdp+qlora.yaml --num_processes={num_gpus} scripts/run_{task}.py recipes/{model_name}/{task}/config_qlora.yaml --torch_dtype=bfloat16 --bnb_4bit_quant_storage=bfloat16
|
||||
```
|
||||
|
||||
Here `{task}` refers to the type of training you wish to run. Currently, the following tasks are supported:
|
||||
* continued pretraining `cpt` (note that `cpt` is only present in the `gpt-nl` example recipe)
|
||||
* supervised finetuning `sft`
|
||||
* direct preference optimisation `dpo`
|
||||
* odds ratio preference optimisation `orpo`
|
||||
|
||||
`{model_name}` refers to the choice of a recipe in the `recipes` directory. For example, to replicate Zephyr-7B-β you can run:
|
||||
|
||||
```shell
|
||||
# Step 1 - train SFT policy
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/deepspeed_zero3.yaml scripts/run_sft.py recipes/zephyr-7b-beta/sft/config_full.yaml
|
||||
|
||||
# Step 2 - align with DPO
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/deepspeed_zero3.yaml scripts/run_dpo.py recipes/zephyr-7b-beta/dpo/config_full.yaml
|
||||
```
|
||||
|
||||
**💡 Tip:** If you scale up/down the number of GPUs, we recommend also scaling up the per-device batch size or number of gradient accumulation steps to keep the global batch size constant (and thus replicate our results).
|
||||
|
||||
By default, these scripts will push each model to your Hugging Face Hub username, i.e. `{username}/{model_name}-{task}`. You can override the parameters in each YAML config by appending them to the command as follows:
|
||||
|
||||
```shell
|
||||
# Change batch size, number of epochs etc
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/deepspeed_zero3.yaml scripts/run_{task}.py recipes/{model_name}/{task}/config_full.yaml --per_device_train_batch_size=42 --num_train_epochs=5
|
||||
```
|
||||
|
||||
## Logging with Weights and Biases
|
||||
By default, all training metrics are logged with TensorBoard. If you have a [Weights and Biases](https://wandb.ai/site) account and are logged in, you can view the training metrics by appending `--report_to=wandb`, e.g.
|
||||
|
||||
```shell
|
||||
ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/deepspeed_zero3.yaml scripts/run_{task}.py recipes/{model_name}/{task}/config_full.yaml --report_to=wandb
|
||||
```
|
||||
|
||||
## Launching jobs on a Slurm cluster
|
||||
|
||||
If you have access to a Slurm cluster, we provide a `recipes/launch.slurm` script that will automatically queue training jobs for you. Here's how you can use it:
|
||||
|
||||
```shell
|
||||
sbatch --job-name=handbook_{task} --nodes=1 recipes/launch.slurm {model_name} {task} {precision} {accelerator}
|
||||
```
|
||||
|
||||
Here `{model_name}` and `{task}` are defined as above, while `{precision}` refers to the type of training (`full` vs `qlora`) and `{accelerator}` refers to the choice of 🤗 Accelerate config in `recipes/accelerate_configs`. If you wish to override the default config parameters, you can provide them by appending a space-separated string like `'--arg1=value1 --arg2=value2'. Here's a concrete example to run SFT on 1 node of 8 GPUs:
|
||||
|
||||
```shell
|
||||
# Launch on Slurm and override default hyperparameters
|
||||
sbatch --job-name=handbook_sft --nodes=1 recipes/launch.slurm zephyr-7b-beta sft full deepspeed_zero3 '--per_device_train_batch_size=42 --num_train_epochs=5'
|
||||
```
|
||||
|
||||
You can scale the number of nodes by increasing the `--nodes` flag.
|
||||
|
||||
**⚠️ Note:** the configuration in `recipes/launch.slurm` is optimised for the Hugging Face Compute Cluster and may require tweaking to be adapted to your own compute nodes.
|
||||
|
||||
## Fine-tuning on your datasets
|
||||
|
||||
Under the hood, each training script uses the `get_datasets()` function which allows one to easily combine multiple datasets with varying proportions. For instance, this is how one can specify multiple datasets and which splits to combine in one of the YAML configs:
|
||||
|
||||
```yaml
|
||||
datasets_mixer:
|
||||
dataset_1: 0.5 # Use 50% of the training examples
|
||||
dataset_2: 0.66 # Use 66% of the training examples
|
||||
dataset_3: 0.10 # Use 10% of the training examples
|
||||
dataset_splits:
|
||||
- train_xxx # The training splits to mix
|
||||
- test_xxx # The test splits to mix
|
||||
```
|
||||
|
||||
If you want to fine-tune on your datasets, the main thing to keep in mind is how the chat templates are applied to the dataset blend. Since each task (SFT, DPO, ORPO, etc), requires a different format, we assume the datasets have the following columns:
|
||||
|
||||
**SFT**
|
||||
|
||||
* `messages`: A list of `dicts` in the form `{"role": "{role}", "content": {content}}`.
|
||||
* See [ultrachat_200k](https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k) for an example.
|
||||
|
||||
**DPO and ORPO**
|
||||
|
||||
* `chosen`: A list of `dicts` in the form `{"role": "{role}", "content": {content}}` corresponding to the preferred dialogue.
|
||||
* `rejected`: A list of `dicts` in the form `{"role": "{role}", "content": {content}}` corresponding to the dispreferred dialogue.
|
||||
* See [ultrafeedback_binarized](https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized) for an example.
|
||||
|
||||
We also find it useful to include dedicated splits per task in our datasets, so e.g. we have:
|
||||
|
||||
* `{train,test}_sft`: Splits for SFT training.
|
||||
* `{train,test}_gen`: Splits for generation ranking like rejection sampling or PPO.
|
||||
* `{train,test}_prefs`: Splits for preference modelling, like reward modelling or DPO.
|
||||
|
||||
If you format your dataset in the same way, our training scripts should work out of the box!
|
||||
|
||||
## Evaluating chat models
|
||||
|
||||
We recommend benchmarking chat models on:
|
||||
|
||||
* [MT-Bench](https://huggingface.co/spaces/lmsys/mt-bench): a multi-turn benchmark spanning 80 dialogues and 10 domains.
|
||||
* [AlpacaEval](https://github.com/tatsu-lab/alpaca_eval): a single-turn benchmark that evaluates the helpfulness of chat and instruct models against `text-davinci-003`.
|
||||
|
||||
For both benchmarks, we have added support for the [Zephyr chat template](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full/blob/ac6e600eefcce74f5e8bae1035d4f66019e93190/tokenizer_config.json#L30) (which is the default produced by our scripts), so you can evaluate models produced by our scripts as follows:
|
||||
|
||||
**MT-Bench**
|
||||
|
||||
* Follow the installation instructions [here](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge)
|
||||
* Make sure the word `zephyr` exists in the `--model-path` argument when generating the model responses [here](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge#step-1-generate-model-answers-to-mt-bench-questions). This will ensure the correct chat template is loaded. For example, the following model name is valid: `--model-path {hub_username}/my-baby-zephyr`
|
||||
* Generate the model responses and GPT-4 rankings.
|
||||
|
||||
**AlpacaEval**
|
||||
|
||||
* Follow the installation instructions [here](https://github.com/tatsu-lab/alpaca_eval#quick-start)
|
||||
* Copy-paste the [config](https://github.com/tatsu-lab/alpaca_eval/blob/main/src/alpaca_eval/models_configs/zephyr-7b-beta/configs.yaml) for `zephyr-7b-beta` and place it in the `model_configs` directory under `{your_zephyr_model}`.
|
||||
* Next, update the [config name](https://github.com/tatsu-lab/alpaca_eval/blob/2daa6e11b194653043ca74f735728dc068e04aae/src/alpaca_eval/models_configs/zephyr-7b-beta/configs.yaml#L1) and [Hub model ID](https://github.com/tatsu-lab/alpaca_eval/blob/2daa6e11b194653043ca74f735728dc068e04aae/src/alpaca_eval/models_configs/zephyr-7b-beta/configs.yaml#L5) to match your model name.
|
||||
* Follow the steps to evaluate your model [here](https://github.com/tatsu-lab/alpaca_eval/tree/main#evaluating-a-model).
|
||||
|
||||
Note that MT-Bench and AlpacaEval rely on LLMs like GPT-4 to judge the quality of the model responses, and thus the ranking exhibits various biases including a preference for models distilled from GPTs. For that reason, we also recommend submitting your best models for human evaluation in:
|
||||
|
||||
* [Chatbot Arena](https://chat.lmsys.org): a live, human evaluation of chat models in head-to-head comparisons.
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
#!/usr/bin/env python
|
||||
# coding=utf-8
|
||||
# Copyright 2023 The HuggingFace Inc. team. All rights reserved.
|
||||
# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
@@ -204,7 +204,7 @@ def main():
|
||||
"finetuned_from": model_args.model_name_or_path,
|
||||
"dataset": list(data_args.dataset_mixer.keys()),
|
||||
"dataset_tags": list(data_args.dataset_mixer.keys()),
|
||||
"tags": ["alignment-handbook"],
|
||||
"tags": ["open-r1"],
|
||||
}
|
||||
if trainer.accelerator.is_main_process:
|
||||
trainer.create_model_card(**kwargs)
|
||||
|
||||
+1
-1
@@ -3,7 +3,7 @@ default_section = FIRSTPARTY
|
||||
ensure_newline_before_comments = True
|
||||
force_grid_wrap = 0
|
||||
include_trailing_comma = True
|
||||
known_first_party = alignment
|
||||
known_first_party = open_r1
|
||||
known_third_party =
|
||||
transformers
|
||||
datasets
|
||||
|
||||
+15
-32
@@ -41,34 +41,27 @@ if stale_egg_info.exists():
|
||||
# IMPORTANT: all dependencies should be listed here with their version requirements, if any.
|
||||
# * If a dependency is fast-moving (e.g. transformers), pin to the exact version
|
||||
_deps = [
|
||||
"accelerate>=0.29.2",
|
||||
"accelerate>=1.2.1",
|
||||
"bitsandbytes>=0.43.0",
|
||||
"black>=24.4.2",
|
||||
"datasets>=2.18.0",
|
||||
"deepspeed>=0.14.4",
|
||||
"einops>=0.6.1",
|
||||
"evaluate==0.4.0",
|
||||
"datasets>=3.2.0",
|
||||
"deepspeed==0.15.4",
|
||||
"einops>=0.8.0",
|
||||
"flake8>=6.0.0",
|
||||
"hf-doc-builder>=0.4.0",
|
||||
"hf_transfer>=0.1.4",
|
||||
"huggingface-hub>=0.19.2,<1.0",
|
||||
"huggingface-hub[cli]>=0.19.2,<1.0",
|
||||
"isort>=5.12.0",
|
||||
"ninja>=1.11.1",
|
||||
"numpy>=1.24.2",
|
||||
"lighteval @ git+https://github.com/huggingface/lighteval.git@4f381b352c0e467b5870a97d41cb66b487a2c503#egg=lighteval[math]",
|
||||
"packaging>=23.0",
|
||||
"parameterized>=0.9.0",
|
||||
"peft>=0.9.0",
|
||||
"protobuf<=3.20.2", # Needed to avoid conflicts with `transformers`
|
||||
"pytest",
|
||||
"safetensors>=0.3.3",
|
||||
"sentencepiece>=0.1.99",
|
||||
"scipy",
|
||||
"tensorboard",
|
||||
"torch>=2.1.2",
|
||||
"transformers>=4.39.3",
|
||||
"trl @ git+https://github.com/huggingface/trl.git",
|
||||
"jinja2>=3.0.0",
|
||||
"tqdm>=4.64.1",
|
||||
"torch>=2.5.1",
|
||||
"transformers==4.48.1",
|
||||
"trl @ git+https://github.com/huggingface/trl.git@6f99f42f724123409422f2fad42bf56fa91f366f", # Bump when this is merged: https://github.com/huggingface/trl/pull/2650
|
||||
"vllm==0.6.6.post1",
|
||||
"wandb>=0.19.1",
|
||||
]
|
||||
|
||||
# this is a lookup table with items like:
|
||||
@@ -88,45 +81,35 @@ extras = {}
|
||||
extras["tests"] = deps_list("pytest", "parameterized")
|
||||
extras["torch"] = deps_list("torch")
|
||||
extras["quality"] = deps_list("black", "isort", "flake8")
|
||||
extras["docs"] = deps_list("hf-doc-builder")
|
||||
extras["dev"] = extras["docs"] + extras["quality"] + extras["tests"]
|
||||
extras["dev"] = extras["quality"] + extras["tests"]
|
||||
|
||||
# core dependencies shared across the whole project - keep this to a bare minimum :)
|
||||
install_requires = [
|
||||
deps["accelerate"],
|
||||
deps["bitsandbytes"],
|
||||
deps["einops"],
|
||||
deps["evaluate"],
|
||||
deps["datasets"],
|
||||
deps["deepspeed"],
|
||||
deps["hf_transfer"],
|
||||
deps["huggingface-hub"],
|
||||
deps["jinja2"],
|
||||
deps["ninja"],
|
||||
deps["numpy"],
|
||||
deps["packaging"], # utilities from PyPA to e.g., compare versions
|
||||
deps["peft"],
|
||||
deps["protobuf"],
|
||||
deps["safetensors"],
|
||||
deps["sentencepiece"],
|
||||
deps["scipy"],
|
||||
deps["tensorboard"],
|
||||
deps["tqdm"], # progress bars in model download and training scripts
|
||||
deps["transformers"],
|
||||
deps["trl"],
|
||||
]
|
||||
|
||||
setup(
|
||||
name="open_r1",
|
||||
name="open-r1",
|
||||
version="0.1.0.dev0", # expected format is one of x.y.z.dev0, or x.y.z.rc1 or x.y.z (no to dashes, yes to dots)
|
||||
author="The Hugging Face team (past and future)",
|
||||
author_email="lewis@huggingface.co",
|
||||
description="Open R1",
|
||||
long_description=open("README.md", "r", encoding="utf-8").read(),
|
||||
long_description_content_type="text/markdown",
|
||||
keywords="nlp deep learning rlhf llm",
|
||||
keywords="llm inference-time compute reasoning",
|
||||
license="Apache",
|
||||
url="https://github.com/huggingface/open_r1",
|
||||
url="https://github.com/huggingface/open-r1",
|
||||
package_dir={"": "src"},
|
||||
packages=find_packages("src"),
|
||||
zip_safe=False,
|
||||
|
||||
@@ -38,7 +38,7 @@ REPLACE_PATTERNS = {
|
||||
README_FILE = "README.md"
|
||||
|
||||
REPLACE_FILES = {
|
||||
"init": "src/alignment/__init__.py",
|
||||
"init": "src/open_r1/__init__.py",
|
||||
"setup": "setup.py",
|
||||
"citation": "CITATION.cff",
|
||||
"readme": README_FILE,
|
||||
|
||||
Referência em uma Nova Issue
Bloquear um usuário