中文 ｜ English

Hy3 preview provides processes related to model training. This section details how to process training data for model training purposes.

The training data should be formatted as a list of messages. By default, the system prompt for both training and inference is empty, but you may customize it as needed.

Below is a training data example for a translation task:

# Translation task example
{"messages": [{"role": "user", "content": "将以下中文翻译为英文，只输出翻译结果，不要额外解释：\n\n实验结果证明了假设的正确性。"}, {"role": "assistant", "content": "The experimental results demonstrate the correctness of the hypothesis."}]}

You can quickly get started by following the instructions in the Quick Start Guide.

The following are the minimum hardware requirements for each model at max_seq_length = 8192:

If you only use single-machine training, you can skip this section.

The following instructions use two machines as an example, with their IPs denoted as ${ip1} and ${ip2}. All steps should be performed inside the Docker container.

First, configure passwordless SSH for each container on every machine:

ssh-keygen			# Generate id_rsa and id_rsa.pub for passwordless login
ssh-keygen -t rsa -A    # Generate /etc/ssh/ssh_host_rsa_key and ssh_host_ecdsa_key for SSH listening
/usr/sbin/sshd -p 36005 -o ListenAddress=0.0.0.0        # Start SSH listening
echo "Port 36005" > ~/.ssh/config   # Set SSH connection port to 36005
passwd root    # Set the root password to avoid monitoring platform alerts

Note: 36005 is an example port. You may use any available port, but ensure it is open and not occupied by other processes.

Next, in each machine's container, execute:

cat ~/.ssh/id_rsa.pub

Copy the output SSH public key and paste it into the ~/.ssh/authorized_keys file, one key per line. This must be done on every machine. In the end, the ~/.ssh/authorized_keys file on each machine should be identical and contain the public keys of all machines.

Please note that for multi-node training, the code executed on each node must be identical. It is recommended to mount a shared network drive. If this is not possible, you must manually copy the dataset, scripts, and code to the same directory on each machine.

This project provides two training methods. You can choose based on your needs:

• DeepSpeed Native Training (based on HuggingFace Transformers Trainer): Located in the train/deepspeed_support directory
• LLaMA-Factory Training: Located in the train/llama_factory_support directory

Reference: HuggingFace Transformers Trainer

In the train/deepspeed_support directory, the training scripts for each model are as follows:

In the train/deepspeed_support directory, install dependencies and execute the corresponding script:

pip install -r requirements.txt
# Example: Dense 1.8B full fine-tuning
bash train_dense.sh 1.8B

To launch training across multiple machines, please first complete the configuration in Configure Passwordless SSH Login Between Machines, and ensure all machines are within the same cluster.

Confirm that dependencies are installed (if not, run pip install -r requirements.txt), then set the IP_LIST environment variable in the corresponding training script:

export HOST_GPU_NUM=8
# IP list, comma separated. e.g. "192.168.1.1,192.168.1.2" or single node "192.168.1.1"
IP_LIST=${IP_LIST:-"127.0.0.1"}

Note: If the IP_LIST environment variable is not set, replace IP_LIST with the IP list! The format is:

For a single IP:
IP_LIST=${ip_1}

For multiple IPs:
IP_LIST=${ip_1},${ip_2}

Replace ${ip_1} and ${ip_2} with the actual IP addresses.

Then, on the machine with ${ip1}, execute the corresponding training script in the train/deepspeed_support/ directory. On first launch, you may see the following output:

The authenticity of host '[ip]:36005 ([ip]:36005)' can't be established.
ECDSA key fingerprint is xxxxxx.
ECDSA key fingerprint is MD5:xxxxxx.
Are you sure you want to continue connecting (yes/no)?

Type yes to continue.

The key parameters in the script are as follows:

• --deepspeed: Path to the DeepSpeed configuration file. Three default DeepSpeed configuration files are provided in the train/deepspeed_support folder: ds_zero2_no_offload.json, ds_zero3_no_offload.json, and ds_zero3_offload.json, with decreasing memory requirements in that order.
• --model_name_or_path: Path to the Hy3 preview HF pre-trained model weights to load.
• --tokenizer_name_or_path: Path to the tokenizer folder.
• --train_data_file: Path to the training file, which should be a jsonl file.
• --output_dir: Output directory where logs, tensorboard files, and model weights will be stored.
• --per_device_train_batch_size: Batch size per GPU.
• --gradient_accumulation_steps: Number of gradient accumulation steps. The global batch size is per_device_train_batch_size * gradient_accumulation_steps * dp_size.
• --max_steps: Total number of training steps.
• --save_steps: Number of steps between saving checkpoints.
• --use_lora: Whether to use LoRA training. Also accepts --lora_rank, --lora_alpha, and --lora_dropout parameters. By default, LoRA is applied to "q_proj", "k_proj", "v_proj", and "o_proj". To change this, modify the code. Note: When using LoRA training, only the LoRA weights are saved, not the base model weights. To merge LoRA weights, see the "LoRA Weight Merging" section below.
• --make_moe_param_leaf_module: When using ZeRO-3 with MoE training, treat the MoE module as a leaf module, i.e., its parameters are not partitioned by ZeRO-3. This option is expected to significantly increase memory usage.
• --gradient_checkpointing: Enable gradient checkpointing.
• --train_attention_params_only: Whether to train only attention parameters.
• --learning_rate: Maximum learning rate during training.
• --min_lr: Minimum learning rate during training.
• --use_flash_attn: Enable flash-attention for accelerated training.

Notes:

• To resume training from a previously saved checkpoint rather than loading pre-trained weights, specify --resume_from_checkpoint with the path to the checkpoint. Do not specify --model_name_or_path; this will load only the weights without the training state.
• When resuming from a checkpoint, there may be minor differences in loss due to the randomness of some non-deterministic algorithms. This is normal. See: HuggingFace Transformers Trainer Randomness
• When --model_name_or_path is specified, all model-related parameters will be ignored.
• Samples within a batch are padded to the length of the longest sample in the batch, but the maximum length of each sample is max_seq_length. Any excess will be truncated.
• If you see a warning about bias weights not being loaded, you can ignore it. Hunyuan-Large does not use bias.

Reference: DeepSpeed Configuration

You can try modifying the DeepSpeed configuration by removing the auto attribute from the following parameters and reducing their values:

• stage3_param_persistence_threshold
• stage3_prefetch_bucket_size
• stage3_max_reuse_distance

LoRA weights saved during training cannot be merged into the ZeRO-3 model at runtime, as ZeRO-3 partitions model weights across data parallel ranks. To merge LoRA weights into the base model, you can do so offline to obtain a merged weight file. Run merge_lora_weight.sh to merge the LoRA and base model weights. The parameters are:

• --base_model_path: Directory of the base model weights
• --adapter_model_path: Directory of the LoRA weights
• --output_path: Directory to save the merged weights
• --save_dtype: Data type for saving the merged weights; options are: fp16, bf16, fp32

If you are familiar with LLaMA-Factory, you may use it for fine-tuning. All scripts, code, and configuration files are archived in the train/llama_factory_support directory. Unless otherwise specified, all files mentioned below are located in this directory.

You can install LLaMA-Factory by downloading the source code from https://github.com/hiyouga/LLaMA-Factory/tree/main and following the instructions on the website.

The configuration files and launch scripts for each model are as follows:

Tip: The Dense model launch script train_lf_dense.sh uses hy_dense_1_8b_full_sft.yaml by default. You can specify other configuration files via the YAML_FILE environment variable.

Key parameters in the configuration files are as follows:

Model:

• model_name_or_path: Path to the Hy-MT HF format pre-trained model weights
• trust_remote_code: Whether to trust remote code; Hy-MT requires this to be set to true

Training Method:

• stage: Training stage, currently sft (supervised fine-tuning)
• finetuning_type: Fine-tuning type, either full (full fine-tuning) or lora (LoRA fine-tuning)
• deepspeed: DeepSpeed configuration file path; ds_zero3_offload.json is recommended for full fine-tuning, ds_zero2_offload_lora.json for LoRA fine-tuning

LoRA Parameters (only effective during LoRA fine-tuning):

• lora_rank: LoRA rank, default 64
• lora_alpha: LoRA alpha coefficient, default 128
• lora_dropout: LoRA dropout ratio, default 0.05
• lora_target: Target modules for LoRA, default q_proj,k_proj,v_proj,o_proj

Dataset:

• dataset_dir: Dataset directory path
• dataset: Dataset name, must be registered in dataset_info.json under dataset_dir
• template: Chat template; Hy-MT2-1.8B uses hy_dense_1_8b, Hy-MT2-7B uses hy_dense_7b, Hy-MT2-30B-A3B uses hy_v3
• cutoff_len: Maximum sequence length; sequences exceeding this will be truncated. For full fine-tuning, can be set to 262144 (262K); for LoRA fine-tuning, 8192 is recommended to save memory
• max_samples: Maximum number of samples per dataset
• overwrite_cache: Whether to overwrite cached preprocessed datasets

Output:

• output_dir: Output directory where logs, TensorBoard files, and weights will be stored
• logging_steps: Number of steps between logging
• save_steps: Number of steps between saving checkpoints
• plot_loss: Whether to plot the training loss curve
• overwrite_output_dir: Whether to overwrite the existing output directory
• save_only_model: Whether to save only model weights (excluding optimizer states, etc.)
• report_to: Logging tool, options: none, wandb, tensorboard, swanlab, mlflow

Training Hyperparameters:

• per_device_train_batch_size: Batch size per GPU
• gradient_accumulation_steps: Gradient accumulation steps; per_device_train_batch_size * gradient_accumulation_steps * dp_size equals the global batch size
• learning_rate: Maximum learning rate; 1.0e-5 recommended for full fine-tuning, 2.0e-4 for LoRA fine-tuning
• num_train_epochs: Number of training epochs
• lr_scheduler_type: Learning rate scheduler type; cosine_with_min_lr is recommended
• lr_scheduler_kwargs.min_lr_rate: Ratio of minimum to maximum learning rate; e.g., 0.1 means the minimum learning rate is 10% of the maximum
• warmup_ratio: Proportion of total training steps used for warmup
• bf16: Whether to use BFloat16 mixed precision training
• gradient_checkpointing: Whether to enable gradient checkpointing to save memory
• ddp_timeout: Distributed training timeout (milliseconds)
• flash_attn: Attention implementation; fa2 (FlashAttention-2) is recommended, sdpa is also available; using fa2 requires the flash-attn package
• resume_from_checkpoint: Resume training from a specified checkpoint path; set to null to start from scratch

For multi-machine training, please first complete the configuration in Configure Passwordless SSH Login Between Machines (single-machine training can skip this step).

Modify the following configuration at the beginning of the corresponding launch script:

export HOST_GPU_NUM=8
# IP list, comma separated. e.g. "192.168.1.1,192.168.1.2" or single node "192.168.1.1"
export IP_LIST=${IP_LIST:-"127.0.0.1"}

Note: If the IP_LIST environment variable is not set, replace IP_LIST with the IP list! The format is:

For a single IP:
IP_LIST=${ip_1}

For multiple IPs:
IP_LIST=${ip_1},${ip_2}

Replace ${ip_1} and ${ip_2} with the actual IP addresses.

Then, on each machine, run the corresponding launch script in the train/llama_factory_support/ directory. For example:

# Dense 1.8B full fine-tuning
bash train_lf_dense.sh

# Dense 7B LoRA fine-tuning
YAML_FILE=hy_dense_7b_lora_sft.yaml bash train_lf_dense.sh

# MoE 30B-A3B full fine-tuning
bash train_lf.sh