W-61/llama-3-8b-base-sft-hh-helpful-8xh200
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
cda0f8938a1c1acd6c3889ad4b2be067738652dc
llama-3-8b-base-sft-hh-help…/train.log
ModelHub XC cda0f8938a 初始化项目，由ModelHub XC社区提供模型 
Model: W-61/llama-3-8b-base-sft-hh-helpful-8xh200
Source: Original Platform
2026-04-24 12:05:40 +08:00

1241 lines
193 KiB
Plaintext
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/transformers/utils/hub.py:105: FutureWarning: Using `TRANSFORMERS_CACHE` is deprecated and will be removed in v5 of Transformers. Use `HF_HOME` instead.
warnings.warn(
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 2, device: cuda:2, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 7, device: cuda:7, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 1, device: cuda:1, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - INFO - __main__ - Model parameters ModelArguments(base_model_revision=None, model_name_or_path='/scratch/feng.yulu/dynamic-dpo-v4/base_models/Meta-Llama-3-8B', model_revision='main', model_code_revision=None, torch_dtype='bfloat16', tokenizer_name_or_path=None, trust_remote_code=False, attn_implementation='flash_attention_2', use_peft=False, lora_r=16, lora_alpha=32, lora_dropout=0.05, lora_target_modules=None, lora_modules_to_save=None, load_in_8bit=False, load_in_4bit=False, bnb_4bit_quant_type='nf4', use_bnb_nested_quant=False, bnb_4bit_quant_storage='uint8')
2026-04-10 13:38:21 - INFO - __main__ - Data parameters DataArguments(chat_template="{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}", dataset_mixer={'Anthropic/hh-rlhf': 1.0}, text_column='text', dataset_splits=['train', 'test'], dataset_configs=['helpful-base'], dataset_dir=None, preprocessing_num_workers=12, use_persistent_hf_cache=False, hf_cache_dir=None, truncation_side=None, auto_insert_empty_system_msg=True, preprocessing_log_samples=0, preprocessing_log_dir=None)
2026-04-10 13:38:21 - INFO - __main__ - Training/evaluation parameters SFTConfig(
_n_gpu=1,
accelerator_config={'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None, 'use_configured_state': False},
adafactor=False,
adam_beta1=0.9,
adam_beta2=0.999,
adam_epsilon=1e-08,
auto_find_batch_size=False,
average_tokens_across_devices=False,
batch_eval_metrics=False,
bf16=True,
bf16_full_eval=False,
chars_per_token=<CHARS_PER_TOKEN>,
data_seed=None,
dataloader_drop_last=False,
dataloader_num_workers=0,
dataloader_persistent_workers=False,
dataloader_pin_memory=True,
dataloader_prefetch_factor=None,
dataset_batch_size=1000,
dataset_kwargs=None,
dataset_num_proc=None,
dataset_text_field=None,
ddp_backend=None,
ddp_broadcast_buffers=None,
ddp_bucket_cap_mb=None,
ddp_find_unused_parameters=None,
ddp_timeout=1800,
debug=[],
deepspeed=None,
disable_tqdm=False,
do_eval=True,
do_predict=False,
do_train=False,
eval_accumulation_steps=None,
eval_delay=0,
eval_do_concat_batches=True,
eval_on_start=False,
eval_packing=None,
eval_steps=100,
eval_strategy=IntervalStrategy.STEPS,
eval_use_gather_object=False,
fp16=False,
fp16_backend=auto,
fp16_full_eval=False,
fp16_opt_level=O1,
fsdp=[],
fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False},
fsdp_min_num_params=0,
fsdp_transformer_layer_cls_to_wrap=None,
full_determinism=False,
gradient_accumulation_steps=1,
gradient_checkpointing=True,
gradient_checkpointing_kwargs={'use_reentrant': False},
greater_is_better=None,
group_by_length=False,
half_precision_backend=auto,
hub_always_push=False,
hub_model_id=W-61/llama-3-8b-base-sft-hh-helpful-4xh200,
hub_model_revision=main,
hub_private_repo=None,
hub_strategy=HubStrategy.END,
hub_token=<HUB_TOKEN>,
ignore_data_skip=False,
include_for_metrics=[],
include_inputs_for_metrics=False,
include_num_input_tokens_seen=False,
include_tokens_per_second=False,
jit_mode_eval=False,
label_names=None,
label_smoothing_factor=0.0,
learning_rate=2e-05,
length_column_name=length,
load_best_model_at_end=False,
local_rank=0,
log_level=info,
log_level_replica=warning,
log_on_each_node=True,
logging_dir=outputs/llama-3-8b-base-sft-hh-helpful-4xh200/runs/Apr10_13-38-20_d4054,
logging_first_step=True,
logging_nan_inf_filter=True,
logging_steps=5,
logging_strategy=IntervalStrategy.STEPS,
lr_scheduler_kwargs={},
lr_scheduler_type=SchedulerType.COSINE,
max_grad_norm=1.0,
max_seq_length=512,
max_steps=-1,
metric_for_best_model=None,
model_init_kwargs=None,
mp_parameters=,
neftune_noise_alpha=None,
no_cuda=False,
num_of_sequences=1024,
num_train_epochs=1,
optim=OptimizerNames.ADAMW_TORCH,
optim_args=None,
optim_target_modules=None,
output_dir=/scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758,
overwrite_output_dir=True,
packing=False,
past_index=-1,
per_device_eval_batch_size=16,
per_device_train_batch_size=16,
prediction_loss_only=False,
push_to_hub=False,
push_to_hub_model_id=None,
push_to_hub_organization=None,
push_to_hub_token=<PUSH_TO_HUB_TOKEN>,
ray_scope=last,
remove_unused_columns=True,
report_to=['wandb'],
restore_callback_states_from_checkpoint=False,
resume_from_checkpoint=None,
run_name=llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758,
save_on_each_node=False,
save_only_model=False,
save_safetensors=True,
save_steps=200,
save_strategy=SaveStrategy.STEPS,
save_total_limit=2,
seed=42,
skip_memory_metrics=True,
tf32=None,
torch_compile=False,
torch_compile_backend=None,
torch_compile_mode=None,
torch_empty_cache_steps=None,
torchdynamo=None,
tp_size=0,
tpu_metrics_debug=False,
tpu_num_cores=None,
use_cpu=False,
use_ipex=False,
use_legacy_prediction_loop=False,
use_liger=False,
use_liger_kernel=False,
use_mps_device=False,
warmup_ratio=0.1,
warmup_steps=0,
weight_decay=0.0,
)
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 4, device: cuda:4, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 5, device: cuda:5, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 3, device: cuda:3, n_gpu: 1 distributed training: True, 16-bits training: False
2026-04-10 13:38:21 - WARNING - __main__ - Process rank: 6, device: cuda:6, n_gpu: 1 distributed training: True, 16-bits training: False
Downloading readme: 0%| | 0.00/5.77k [00:00<?, ?B/s]
Downloading readme: 39%|███▉ | 2.24k/5.77k [00:00<00:00, 19.7kB/s]
Downloading readme: 100%|██████████| 5.77k/5.77k [00:00<00:00, 49.3kB/s]
Downloading data: 0%| | 0.00/16.2M [00:00<?, ?B/s]No config specified, defaulting to the single config: hh-rlhf/default
2026-04-10 13:38:22 - INFO - datasets.builder - No config specified, defaulting to the single config: hh-rlhf/default
Using custom data configuration default-cfba128a0ab1b99f
2026-04-10 13:38:22 - INFO - datasets.builder - Using custom data configuration default-cfba128a0ab1b99f
Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/json
2026-04-10 13:38:22 - INFO - datasets.info - Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/json
Downloading data: 65%|██████▍ | 10.5M/16.2M [00:00<00:00, 14.0MB/s]
Downloading data: 100%|██████████| 16.2M/16.2M [00:01<00:00, 15.4MB/s]
Downloading data: 100%|██████████| 16.2M/16.2M [00:01<00:00, 14.6MB/s]
Downloading data: 0%| | 0.00/875k [00:00<?, ?B/s]
Downloading data: 100%|██████████| 875k/875k [00:00<00:00, 7.10MB/s]
Downloading data: 100%|██████████| 875k/875k [00:00<00:00, 6.86MB/s]
Generating train split: 0 examples [00:00, ? examples/s]
Generating train split: 13255 examples [00:00, 92408.74 examples/s]
Generating train split: 26541 examples [00:00, 97720.85 examples/s]
Generating train split: 39887 examples [00:00, 102388.55 examples/s]
Generating train split: 43835 examples [00:00, 88529.69 examples/s]
Generating test split: 0 examples [00:00, ? examples/s]
Generating test split: 2354 examples [00:00, 70834.38 examples/s]
Found cached dataset hh-rlhf (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa)
2026-04-10 13:38:24 - INFO - datasets.builder - Found cached dataset hh-rlhf (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa)
Loading Dataset info from /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa
2026-04-10 13:38:24 - INFO - datasets.info - Loading Dataset info from /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa
2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d6e6bfbe34161664.arrow
2026-04-10 13:38:26 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d6e6bfbe34161664.arrow
2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
2026-04-10 13:38:26 - WARNING - alignment.data - Dropped 237 non-canonical HH preference examples from split `train` before normalization (126 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 111 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]
Normalizing raw HH preferences (train): 0%| | 0/43598 [00:00<?, ? examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1101/43598 [00:00<00:03, 10959.87 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1128/43598 [00:00<00:03, 11219.64 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1125/43598 [00:00<00:03, 11174.40 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1164/43598 [00:00<00:03, 11586.70 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1162/43598 [00:00<00:03, 11564.67 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1154/43598 [00:00<00:03, 11484.37 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1167/43598 [00:00<00:03, 11623.90 examples/s]
Normalizing raw HH preferences (train): 3%|▎ | 1161/43598 [00:00<00:03, 11555.60 examples/s]
Normalizing raw HH preferences (train): 5%|▌ | 2341/43598 [00:00<00:03, 11801.60 examples/s]
Normalizing raw HH preferences (train): 5%|▌ | 2369/43598 [00:00<00:03, 11908.11 examples/s]
Normalizing raw HH preferences (train): 5%|▌ | 2374/43598 [00:00<00:03, 11940.63 examples/s]
Normalizing raw HH preferences (train): 6%|▌ | 2456/43598 [00:00<00:03, 12363.79 examples/s]
Normalizing raw HH preferences (train): 6%|▌ | 2460/43598 [00:00<00:03, 12391.17 examples/s]
Normalizing raw HH preferences (train): 6%|▌ | 2429/43598 [00:00<00:03, 12224.80 examples/s]
Normalizing raw HH preferences (train): 6%|▌ | 2466/43598 [00:00<00:03, 12419.94 examples/s]
Normalizing raw HH preferences (train): 6%|▌ | 2451/43598 [00:00<00:03, 12338.10 examples/s]
Normalizing raw HH preferences (train): 8%|▊ | 3704/43598 [00:00<00:03, 12380.88 examples/s]
Normalizing raw HH preferences (train): 8%|▊ | 3700/43598 [00:00<00:03, 12307.91 examples/s]
Normalizing raw HH preferences (train): 8%|▊ | 3705/43598 [00:00<00:03, 12412.59 examples/s]
Normalizing raw HH preferences (train): 9%|▊ | 3741/43598 [00:00<00:03, 12583.64 examples/s]
Normalizing raw HH preferences (train): 9%|▊ | 3748/43598 [00:00<00:03, 12610.03 examples/s]
Normalizing raw HH preferences (train): 8%|▊ | 3702/43598 [00:00<00:03, 12448.81 examples/s]
Normalizing raw HH preferences (train): 9%|▊ | 3757/43598 [00:00<00:03, 12635.97 examples/s]
Normalizing raw HH preferences (train): 9%|▊ | 3732/43598 [00:00<00:03, 12550.55 examples/s]
Normalizing raw HH preferences (train): 11%|█▏ | 4974/43598 [00:00<00:03, 12499.00 examples/s]
Normalizing raw HH preferences (train): 11%|█▏ | 4936/43598 [00:00<00:03, 12324.82 examples/s]
Normalizing raw HH preferences (train): 11%|█▏ | 4977/43598 [00:00<00:03, 12529.11 examples/s]
Normalizing raw HH preferences (train): 11%|█▏ | 4951/43598 [00:00<00:03, 12459.13 examples/s]
Normalizing raw HH preferences (train): 13%|█▎ | 5474/43598 [00:00<00:03, 12070.50 examples/s]
Normalizing raw HH preferences (train): 13%|█▎ | 5705/43598 [00:00<00:03, 12484.42 examples/s]
Normalizing raw HH preferences (train): 13%|█▎ | 5550/43598 [00:00<00:03, 12295.75 examples/s]
Normalizing raw HH preferences (train): 13%|█▎ | 5574/43598 [00:00<00:03, 12412.73 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6838/43598 [00:00<00:02, 12461.94 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6774/43598 [00:00<00:02, 12286.78 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6847/43598 [00:00<00:02, 12496.35 examples/s]
Normalizing raw HH preferences (train): 15%|█▌ | 6740/43598 [00:00<00:03, 12258.77 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6983/43598 [00:00<00:02, 12573.32 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6779/43598 [00:00<00:02, 12331.32 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6817/43598 [00:00<00:02, 12412.03 examples/s]
Normalizing raw HH preferences (train): 16%|█▌ | 6836/43598 [00:00<00:02, 12478.11 examples/s]
Normalizing raw HH preferences (train): 18%|█▊ | 8000/43598 [00:00<00:02, 12160.25 examples/s]
Normalizing raw HH preferences (train): 20%|█▉ | 8715/43598 [00:00<00:02, 12479.66 examples/s]
Normalizing raw HH preferences (train): 20%|█▉ | 8691/43598 [00:00<00:02, 12366.95 examples/s]
Normalizing raw HH preferences (train): 20%|██ | 8736/43598 [00:00<00:02, 12529.29 examples/s]
Normalizing raw HH preferences (train): 20%|██ | 8869/43598 [00:00<00:02, 12570.47 examples/s]
Normalizing raw HH preferences (train): 20%|█▉ | 8684/43598 [00:00<00:02, 12348.12 examples/s]
Normalizing raw HH preferences (train): 20%|█▉ | 8698/43598 [00:00<00:02, 12460.95 examples/s]
Normalizing raw HH preferences (train): 20%|█▉ | 8701/43598 [00:00<00:02, 12457.83 examples/s]
Normalizing raw HH preferences (train): 21%|██▏ | 9279/43598 [00:00<00:02, 12350.15 examples/s]
Normalizing raw HH preferences (train): 23%|██▎ | 9972/43598 [00:00<00:02, 12485.23 examples/s]
Normalizing raw HH preferences (train): 23%|██▎ | 10000/43598 [00:00<00:02, 12340.30 examples/s]
Normalizing raw HH preferences (train): 23%|██▎ | 10000/43598 [00:00<00:02, 12359.42 examples/s]
Normalizing raw HH preferences (train): 23%|██▎ | 9956/43598 [00:00<00:02, 12447.31 examples/s]
Normalizing raw HH preferences (train): 23%|██▎ | 9986/43598 [00:00<00:02, 12576.17 examples/s]
Normalizing raw HH preferences (train): 23%|██▎ | 9986/43598 [00:00<00:02, 12562.60 examples/s]
Normalizing raw HH preferences (train): 24%|██▍ | 10530/43598 [00:00<00:02, 12393.96 examples/s]
Normalizing raw HH preferences (train): 26%|██▌ | 11308/43598 [00:00<00:02, 12542.06 examples/s]
Normalizing raw HH preferences (train): 25%|██▍ | 10760/43598 [00:00<00:02, 12580.36 examples/s]
Normalizing raw HH preferences (train): 26%|██▌ | 11309/43598 [00:00<00:02, 12559.61 examples/s]
Normalizing raw HH preferences (train): 27%|██▋ | 11855/43598 [00:00<00:02, 12508.01 examples/s]
Normalizing raw HH preferences (train): 27%|██▋ | 11819/43598 [00:00<00:02, 12539.64 examples/s]
Normalizing raw HH preferences (train): 29%|██▉ | 12704/43598 [00:01<00:02, 12684.33 examples/s]
Normalizing raw HH preferences (train): 27%|██▋ | 11824/43598 [00:00<00:02, 12447.62 examples/s]
Normalizing raw HH preferences (train): 29%|██▉ | 12704/43598 [00:01<00:02, 12707.26 examples/s]
Normalizing raw HH preferences (train): 27%|██▋ | 11887/43598 [00:00<00:02, 12607.31 examples/s]
Normalizing raw HH preferences (train): 27%|██▋ | 11851/43598 [00:00<00:02, 12510.56 examples/s]
Normalizing raw HH preferences (train): 29%|██▉ | 12704/43598 [00:01<00:02, 12615.09 examples/s]
Normalizing raw HH preferences (train): 31%|███▏ | 13715/43598 [00:01<00:02, 12467.82 examples/s]
Normalizing raw HH preferences (train): 32%|███▏ | 13976/43598 [00:01<00:02, 12691.52 examples/s]
Normalizing raw HH preferences (train): 31%|███▏ | 13691/43598 [00:01<00:02, 12478.07 examples/s]
Normalizing raw HH preferences (train): 32%|███▏ | 13983/43598 [00:01<00:02, 12654.79 examples/s]
Normalizing raw HH preferences (train): 31%|███▏ | 13684/43598 [00:01<00:02, 12415.05 examples/s]
Normalizing raw HH preferences (train): 33%|███▎ | 14527/43598 [00:01<00:02, 12495.63 examples/s]
Normalizing raw HH preferences (train): 32%|███▏ | 13761/43598 [00:01<00:02, 12562.67 examples/s]
Normalizing raw HH preferences (train): 31%|███▏ | 13703/43598 [00:01<00:02, 12450.01 examples/s]
Normalizing raw HH preferences (train): 34%|███▍ | 14995/43598 [00:01<00:02, 12548.27 examples/s]
Normalizing raw HH preferences (train): 34%|███▍ | 14970/43598 [00:01<00:02, 12560.46 examples/s]
Normalizing raw HH preferences (train): 36%|███▋ | 15875/43598 [00:01<00:02, 12676.01 examples/s]
Normalizing raw HH preferences (train): 34%|███▍ | 14956/43598 [00:01<00:02, 12487.60 examples/s]
Normalizing raw HH preferences (train): 36%|███▋ | 15834/43598 [00:01<00:02, 12645.49 examples/s]
Normalizing raw HH preferences (train): 34%|███▍ | 14976/43598 [00:01<00:02, 12518.37 examples/s]
Normalizing raw HH preferences (train): 36%|███▋ | 15889/43598 [00:01<00:02, 12670.54 examples/s]
Normalizing raw HH preferences (train): 36%|███▌ | 15704/43598 [00:01<00:02, 12580.64 examples/s]
Normalizing raw HH preferences (train): 39%|███▊ | 16877/43598 [00:01<00:02, 12544.86 examples/s]
Normalizing raw HH preferences (train): 39%|███▊ | 16849/43598 [00:01<00:02, 12544.75 examples/s]
Normalizing raw HH preferences (train): 41%|████ | 17746/43598 [00:01<00:02, 12603.43 examples/s]
Normalizing raw HH preferences (train): 39%|███▊ | 16822/43598 [00:01<00:02, 12467.93 examples/s]
Normalizing raw HH preferences (train): 41%|████ | 17715/43598 [00:01<00:02, 12603.96 examples/s]
Normalizing raw HH preferences (train): 39%|███▊ | 16848/43598 [00:01<00:02, 12503.24 examples/s]
Normalizing raw HH preferences (train): 39%|███▉ | 16994/43598 [00:01<00:02, 12654.46 examples/s]
Normalizing raw HH preferences (train): 41%|████ | 17765/43598 [00:01<00:02, 12613.59 examples/s]
Normalizing raw HH preferences (train): 43%|████▎ | 18702/43598 [00:01<00:02, 12413.42 examples/s]
Normalizing raw HH preferences (train): 43%|████▎ | 18689/43598 [00:01<00:02, 12446.70 examples/s]
Normalizing raw HH preferences (train): 45%|████▍ | 19585/43598 [00:01<00:01, 12484.55 examples/s]
Normalizing raw HH preferences (train): 45%|████▍ | 19566/43598 [00:01<00:01, 12511.08 examples/s]
Normalizing raw HH preferences (train): 43%|████▎ | 18672/43598 [00:01<00:02, 12377.51 examples/s]
Normalizing raw HH preferences (train): 43%|████▎ | 18679/43598 [00:01<00:02, 12403.68 examples/s]
Normalizing raw HH preferences (train): 43%|████▎ | 18851/43598 [00:01<00:01, 12560.05 examples/s]
Normalizing raw HH preferences (train): 46%|████▌ | 19970/43598 [00:01<00:01, 12475.61 examples/s]
Normalizing raw HH preferences (train): 45%|████▌ | 19707/43598 [00:01<00:01, 12565.12 examples/s]
Normalizing raw HH preferences (train): 46%|████▌ | 19959/43598 [00:01<00:01, 12506.03 examples/s]
Normalizing raw HH preferences (train): 46%|████▌ | 19938/43598 [00:01<00:01, 12443.14 examples/s]
Normalizing raw HH preferences (train): 46%|████▌ | 19952/43598 [00:01<00:01, 12478.40 examples/s]
Normalizing raw HH preferences (train): 48%|████▊ | 21000/43598 [00:01<00:02, 9221.06 examples/s]
Normalizing raw HH preferences (train): 47%|████▋ | 20364/43598 [00:01<00:02, 9284.15 examples/s]
Normalizing raw HH preferences (train): 48%|████▊ | 21000/43598 [00:01<00:02, 9226.63 examples/s]
Normalizing raw HH preferences (train): 50%|████▉ | 21696/43598 [00:01<00:02, 9278.67 examples/s]
Normalizing raw HH preferences (train): 51%|█████ | 22287/43598 [00:01<00:02, 9941.96 examples/s]
Normalizing raw HH preferences (train): 50%|████▉ | 21692/43598 [00:01<00:02, 9495.20 examples/s]
Normalizing raw HH preferences (train): 48%|████▊ | 21000/43598 [00:01<00:02, 7707.51 examples/s]
Normalizing raw HH preferences (train): 50%|████▉ | 21685/43598 [00:01<00:02, 9939.41 examples/s]
Normalizing raw HH preferences (train): 51%|█████ | 22293/43598 [00:01<00:02, 9944.62 examples/s]
Normalizing raw HH preferences (train): 50%|████▉ | 21694/43598 [00:01<00:02, 9093.83 examples/s]
Normalizing raw HH preferences (train): 53%|█████▎ | 22957/43598 [00:01<00:02, 9935.83 examples/s]
Normalizing raw HH preferences (train): 54%|█████▍ | 23581/43598 [00:02<00:01, 10601.27 examples/s]
Normalizing raw HH preferences (train): 50%|████▉ | 21694/43598 [00:02<00:02, 7748.23 examples/s]
Normalizing raw HH preferences (train): 53%|█████▎ | 22962/43598 [00:01<00:02, 10136.53 examples/s]
Normalizing raw HH preferences (train): 51%|█████ | 22287/43598 [00:02<00:02, 8592.60 examples/s]
Normalizing raw HH preferences (train): 53%|█████▎ | 22965/43598 [00:01<00:01, 10560.63 examples/s]
Normalizing raw HH preferences (train): 54%|█████▍ | 23597/43598 [00:02<00:01, 10615.44 examples/s]
Normalizing raw HH preferences (train): 53%|█████▎ | 22971/43598 [00:02<00:02, 9807.52 examples/s]
Normalizing raw HH preferences (train): 55%|█████▌ | 24124/43598 [00:02<00:01, 10316.84 examples/s]
Normalizing raw HH preferences (train): 57%|█████▋ | 24854/43598 [00:02<00:01, 11108.01 examples/s]
Normalizing raw HH preferences (train): 53%|█████▎ | 22960/43598 [00:02<00:02, 8597.63 examples/s]
Normalizing raw HH preferences (train): 55%|█████▌ | 24133/43598 [00:02<00:01, 10485.12 examples/s]
Normalizing raw HH preferences (train): 54%|█████▍ | 23572/43598 [00:02<00:02, 9430.78 examples/s]
Normalizing raw HH preferences (train): 55%|█████▌ | 24144/43598 [00:02<00:01, 10848.35 examples/s]
Normalizing raw HH preferences (train): 57%|█████▋ | 24869/43598 [00:02<00:01, 11112.67 examples/s]
Normalizing raw HH preferences (train): 55%|█████▌ | 24139/43598 [00:02<00:01, 10212.41 examples/s]
Normalizing raw HH preferences (train): 58%|█████▊ | 25389/43598 [00:02<00:01, 10871.82 examples/s]
Normalizing raw HH preferences (train): 55%|█████▌ | 24130/43598 [00:02<00:02, 9212.58 examples/s]
Normalizing raw HH preferences (train): 58%|█████▊ | 25412/43598 [00:02<00:01, 11040.26 examples/s]
Normalizing raw HH preferences (train): 57%|█████▋ | 24863/43598 [00:02<00:01, 10194.00 examples/s]
Normalizing raw HH preferences (train): 58%|█████▊ | 25426/43598 [00:02<00:01, 11342.98 examples/s]
Normalizing raw HH preferences (train): 61%|██████▏ | 26715/43598 [00:02<00:01, 11547.52 examples/s]
Normalizing raw HH preferences (train): 58%|█████▊ | 25424/43598 [00:02<00:01, 10835.41 examples/s]
Normalizing raw HH preferences (train): 61%|██████ | 26691/43598 [00:02<00:01, 11331.99 examples/s]
Normalizing raw HH preferences (train): 61%|██████▏ | 26733/43598 [00:02<00:01, 11551.82 examples/s]
Normalizing raw HH preferences (train): 58%|█████▊ | 25406/43598 [00:02<00:01, 9994.61 examples/s]
Normalizing raw HH preferences (train): 61%|██████ | 26688/43598 [00:02<00:01, 11442.11 examples/s]
Normalizing raw HH preferences (train): 61%|██████ | 26697/43598 [00:02<00:01, 11701.58 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27984/43598 [00:02<00:01, 11820.71 examples/s]
Normalizing raw HH preferences (train): 61%|██████ | 26691/43598 [00:02<00:01, 11291.51 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27949/43598 [00:02<00:01, 11661.22 examples/s]
Normalizing raw HH preferences (train): 61%|██████▏ | 26737/43598 [00:02<00:01, 10925.08 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27995/43598 [00:02<00:01, 11808.61 examples/s]
Normalizing raw HH preferences (train): 61%|██████ | 26690/43598 [00:02<00:01, 10634.94 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27941/43598 [00:02<00:01, 11731.47 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27957/43598 [00:02<00:01, 11944.68 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27948/43598 [00:02<00:01, 11630.15 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 28000/43598 [00:02<00:01, 11137.06 examples/s]
Normalizing raw HH preferences (train): 69%|██████▊ | 29878/43598 [00:02<00:01, 12096.32 examples/s]
Normalizing raw HH preferences (train): 64%|██████▍ | 27950/43598 [00:02<00:01, 11132.58 examples/s]
Normalizing raw HH preferences (train): 68%|██████▊ | 29830/43598 [00:02<00:01, 11977.77 examples/s]
Normalizing raw HH preferences (train): 69%|██████▊ | 29895/43598 [00:02<00:01, 12100.48 examples/s]
Normalizing raw HH preferences (train): 68%|██████▊ | 29825/43598 [00:02<00:01, 12032.78 examples/s]
Normalizing raw HH preferences (train): 67%|██████▋ | 29309/43598 [00:02<00:01, 11616.32 examples/s]
Normalizing raw HH preferences (train): 68%|██████▊ | 29841/43598 [00:02<00:01, 12167.96 examples/s]
Normalizing raw HH preferences (train): 68%|██████▊ | 29836/43598 [00:02<00:01, 11974.21 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31776/43598 [00:02<00:00, 12274.17 examples/s]
Normalizing raw HH preferences (train): 68%|██████▊ | 29833/43598 [00:02<00:01, 11631.59 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31708/43598 [00:02<00:00, 12161.13 examples/s]
Normalizing raw HH preferences (train): 70%|███████ | 30712/43598 [00:02<00:01, 12021.59 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31794/43598 [00:02<00:00, 12280.85 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31701/43598 [00:02<00:00, 12192.12 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31727/43598 [00:02<00:00, 12306.64 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31720/43598 [00:02<00:00, 12169.90 examples/s]
Normalizing raw HH preferences (train): 76%|███████▌ | 32988/43598 [00:02<00:00, 12314.35 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 32000/43598 [00:02<00:00, 12026.16 examples/s]
Normalizing raw HH preferences (train): 77%|███████▋ | 33690/43598 [00:02<00:00, 12347.21 examples/s]
Normalizing raw HH preferences (train): 73%|███████▎ | 31721/43598 [00:02<00:00, 11949.12 examples/s]
Normalizing raw HH preferences (train): 76%|███████▌ | 32974/43598 [00:02<00:00, 12321.59 examples/s]
Normalizing raw HH preferences (train): 76%|███████▌ | 32997/43598 [00:02<00:00, 12400.38 examples/s]
Normalizing raw HH preferences (train): 77%|███████▋ | 33693/43598 [00:02<00:00, 12361.71 examples/s]
Normalizing raw HH preferences (train): 76%|███████▌ | 32998/43598 [00:02<00:00, 12318.96 examples/s]
Normalizing raw HH preferences (train): 76%|███████▋ | 33291/43598 [00:02<00:00, 12266.45 examples/s]
Normalizing raw HH preferences (train): 80%|████████ | 34967/43598 [00:02<00:00, 12443.56 examples/s]
Normalizing raw HH preferences (train): 76%|███████▌ | 33000/43598 [00:02<00:00, 11939.87 examples/s]
Normalizing raw HH preferences (train): 80%|███████▉ | 34850/43598 [00:02<00:00, 12344.85 examples/s]
Normalizing raw HH preferences (train): 80%|████████ | 34979/43598 [00:02<00:00, 12474.85 examples/s]
Normalizing raw HH preferences (train): 80%|███████▉ | 34833/43598 [00:02<00:00, 12343.93 examples/s]
Normalizing raw HH preferences (train): 80%|███████▉ | 34871/43598 [00:02<00:00, 12428.49 examples/s]
Normalizing raw HH preferences (train): 79%|███████▉ | 34585/43598 [00:03<00:00, 12453.78 examples/s]
Normalizing raw HH preferences (train): 80%|███████▉ | 34861/43598 [00:02<00:00, 12349.69 examples/s]
Normalizing raw HH preferences (train): 79%|███████▊ | 34269/43598 [00:03<00:00, 12126.84 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36826/43598 [00:03<00:00, 12423.96 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36698/43598 [00:03<00:00, 12313.94 examples/s]
Normalizing raw HH preferences (train): 82%|████████▏ | 35863/43598 [00:03<00:00, 12542.32 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36840/43598 [00:03<00:00, 12447.75 examples/s]
Normalizing raw HH preferences (train): 82%|████████▏ | 35546/43598 [00:03<00:00, 12293.37 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36688/43598 [00:03<00:00, 12276.47 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36717/43598 [00:03<00:00, 12385.74 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36698/43598 [00:03<00:00, 12301.49 examples/s]
Normalizing raw HH preferences (train): 87%|████████▋ | 37960/43598 [00:03<00:00, 12384.55 examples/s]
Normalizing raw HH preferences (train): 89%|████████▉ | 38704/43598 [00:03<00:00, 12421.08 examples/s]
Normalizing raw HH preferences (train): 84%|████████▍ | 36798/43598 [00:03<00:00, 12353.76 examples/s]
Normalizing raw HH preferences (train): 87%|████████▋ | 37949/43598 [00:03<00:00, 12355.03 examples/s]
Normalizing raw HH preferences (train): 87%|████████▋ | 37978/43598 [00:03<00:00, 12436.89 examples/s]
Normalizing raw HH preferences (train): 87%|████████▋ | 37735/43598 [00:03<00:00, 12515.09 examples/s]
Normalizing raw HH preferences (train): 89%|████████▉ | 38712/43598 [00:03<00:00, 12455.25 examples/s]
Normalizing raw HH preferences (train): 87%|████████▋ | 37955/43598 [00:03<00:00, 12365.45 examples/s]
Normalizing raw HH preferences (train): 92%|█████████▏| 39977/43598 [00:03<00:00, 12490.93 examples/s]
Normalizing raw HH preferences (train): 91%|█████████▏| 39816/43598 [00:03<00:00, 12378.21 examples/s]
Normalizing raw HH preferences (train): 89%|████████▉ | 39000/43598 [00:03<00:00, 12335.14 examples/s]
Normalizing raw HH preferences (train): 92%|█████████▏| 39987/43598 [00:03<00:00, 12523.03 examples/s]
Normalizing raw HH preferences (train): 89%|████████▉ | 38701/43598 [00:03<00:00, 12393.58 examples/s]
Normalizing raw HH preferences (train): 91%|█████████▏| 39789/43598 [00:03<00:00, 12322.39 examples/s]
Normalizing raw HH preferences (train): 91%|█████████▏| 39835/43598 [00:03<00:00, 12415.72 examples/s]
Normalizing raw HH preferences (train): 91%|█████████▏| 39807/43598 [00:03<00:00, 12355.65 examples/s]
Normalizing raw HH preferences (train): 92%|█████████▏| 40291/43598 [00:03<00:00, 12489.12 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41853/43598 [00:03<00:00, 12490.53 examples/s]
Normalizing raw HH preferences (train): 92%|█████████▏| 39965/43598 [00:03<00:00, 12457.25 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41699/43598 [00:03<00:00, 12402.03 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41868/43598 [00:03<00:00, 12527.42 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41702/43598 [00:03<00:00, 12366.89 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41713/43598 [00:03<00:00, 12445.59 examples/s]
Normalizing raw HH preferences (train): 95%|█████████▌| 41596/43598 [00:03<00:00, 12644.38 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41701/43598 [00:03<00:00, 12398.51 examples/s]
Normalizing raw HH preferences (train): 99%|█████████▊| 42964/43598 [00:03<00:00, 12459.48 examples/s]
Normalizing raw HH preferences (train): 96%|█████████▌| 41829/43598 [00:03<00:00, 12443.51 examples/s]
Normalizing raw HH preferences (train): 99%|█████████▊| 42958/43598 [00:03<00:00, 12410.99 examples/s]
Normalizing raw HH preferences (train): 99%|█████████▊| 42982/43598 [00:03<00:00, 12502.24 examples/s]
Normalizing raw HH preferences (train): 98%|█████████▊| 42871/43598 [00:03<00:00, 12672.72 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:03<00:00, 11006.83 examples/s]
Normalizing raw HH preferences (train): 99%|█████████▊| 42964/43598 [00:03<00:00, 12451.89 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:03<00:00, 9696.86 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:03<00:00, 9249.74 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10761.19 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10670.57 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10739.50 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10703.26 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10714.52 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10685.42 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10547.11 examples/s]
Normalizing raw HH preferences (train): 100%|██████████| 43598/43598 [00:04<00:00, 10475.58 examples/s]
No config specified, defaulting to the single config: hh-rlhf/default
2026-04-10 13:38:30 - INFO - datasets.builder - No config specified, defaulting to the single config: hh-rlhf/default
Using custom data configuration default-cfba128a0ab1b99f
2026-04-10 13:38:30 - INFO - datasets.builder - Using custom data configuration default-cfba128a0ab1b99f
Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/json
2026-04-10 13:38:30 - INFO - datasets.info - Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/json
Overwrite dataset info from restored data version if exists.
2026-04-10 13:38:30 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
Loading Dataset info from /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa
2026-04-10 13:38:30 - INFO - datasets.info - Loading Dataset info from /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa
Found cached dataset hh-rlhf (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa)
2026-04-10 13:38:30 - INFO - datasets.builder - Found cached dataset hh-rlhf (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa)
Loading Dataset info from /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa
2026-04-10 13:38:30 - INFO - datasets.info - Loading Dataset info from /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa
2026-04-10 13:38:30 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
2026-04-10 13:38:30 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]2026-04-10 13:38:30 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]2026-04-10 13:38:31 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]2026-04-10 13:38:31 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]2026-04-10 13:38:31 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]
Normalizing raw HH preferences (test): 49%|████▉ | 1144/2339 [00:00<00:00, 11391.80 examples/s]2026-04-10 13:38:31 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 50%|████▉ | 1168/2339 [00:00<00:00, 11635.35 examples/s]
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-fa6f4b7acba8a3e1.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-fa6f4b7acba8a3e1.arrow
Normalizing raw HH preferences (test): 51%|█████▏ | 1199/2339 [00:00<00:00, 11943.03 examples/s]2026-04-10 13:38:31 - WARNING - alignment.data - Dropped 15 non-canonical HH preference examples from split `test` before normalization (9 x HH preprocessing expects exactly one final assistant response in chosen/rejected suffixes., 6 x HH chosen/rejected transcripts must each contain a divergent assistant response.).
Normalizing raw HH preferences (test): 0%| | 0/2339 [00:00<?, ? examples/s]
Normalizing raw HH preferences (test): 50%|████▉ | 1163/2339 [00:00<00:00, 11586.53 examples/s]
Normalizing raw HH preferences (test): 49%|████▉ | 1157/2339 [00:00<00:00, 11519.39 examples/s]
Normalizing raw HH preferences (test): 51%|█████ | 1188/2339 [00:00<00:00, 11831.46 examples/s]
Normalizing raw HH preferences (test): 48%|████▊ | 1134/2339 [00:00<00:00, 11289.81 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 11436.57 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 11265.03 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 10462.08 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 10523.37 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 10435.30 examples/s]
Normalizing raw HH preferences (test): 50%|████▉ | 1165/2339 [00:00<00:00, 11601.36 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 11198.28 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 11425.53 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 9436.64 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 11051.63 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 10304.93 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 9847.18 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 10092.37 examples/s]
Loading cached shuffled indices for dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-be0876dd0add1b31.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Loading cached shuffled indices for dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-be0876dd0add1b31.arrow
Loading cached shuffled indices for dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-40e942b49dfd026a.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Loading cached shuffled indices for dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-40e942b49dfd026a.arrow
2026-04-10 13:38:31 - INFO - __main__ - Training on the following datasets and their proportions: ['train : 43598', 'test : 2339']
[INFO|tokenization_utils_base.py:2058] 2026-04-10 13:38:31,308 >> loading file tokenizer.json
[INFO|tokenization_utils_base.py:2058] 2026-04-10 13:38:31,308 >> loading file tokenizer.model
[INFO|tokenization_utils_base.py:2058] 2026-04-10 13:38:31,308 >> loading file added_tokens.json
[INFO|tokenization_utils_base.py:2058] 2026-04-10 13:38:31,308 >> loading file special_tokens_map.json
[INFO|tokenization_utils_base.py:2058] 2026-04-10 13:38:31,308 >> loading file tokenizer_config.json
[INFO|tokenization_utils_base.py:2058] 2026-04-10 13:38:31,308 >> loading file chat_template.jinja
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 11399.82 examples/s]
Normalizing raw HH preferences (test): 100%|██████████| 2339/2339 [00:00<00:00, 9899.47 examples/s]
[INFO|tokenization_utils_base.py:2323] 2026-04-10 13:38:31,605 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
2026-04-10 13:38:31 - INFO - __main__ - *** Load pretrained model ***
Process #0 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00000_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #0 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00000_of_00012.arrow
Process #1 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00001_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #1 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00001_of_00012.arrow
Process #2 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00002_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #2 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00002_of_00012.arrow
Process #3 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00003_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #3 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00003_of_00012.arrow
Process #4 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00004_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #4 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00004_of_00012.arrow
Process #5 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00005_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #5 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00005_of_00012.arrow
Process #6 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00006_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #6 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00006_of_00012.arrow
Process #7 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00007_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #7 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00007_of_00012.arrow
Process #8 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00008_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #8 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00008_of_00012.arrow
Process #9 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00009_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #9 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00009_of_00012.arrow
Process #10 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00010_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #10 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00010_of_00012.arrow
Process #11 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00011_of_00012.arrow
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Process #11 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00011_of_00012.arrow
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]Spawning 12 processes
2026-04-10 13:38:31 - INFO - datasets.arrow_dataset - Spawning 12 processes
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/43598 [00:00<?, ? examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00000_of_00012.arrow
2026-04-10 13:38:32 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00000_of_00012.arrow
Applying chat template (num_proc=12): 0%| | 58/43598 [00:00<10:22, 69.96 examples/s]
Applying chat template (num_proc=12): 0%| | 96/43598 [00:00<06:36, 109.72 examples/s]
Applying chat template (num_proc=12): 0%| | 110/43598 [00:00<05:42, 126.83 examples/s]
Applying chat template (num_proc=12): 0%| | 1/43598 [00:00<10:29:27, 1.15 examples/s]
Applying chat template (num_proc=12): 0%| | 68/43598 [00:00<09:03, 80.07 examples/s]
Applying chat template (num_proc=12): 0%| | 1/43598 [00:00<10:36:21, 1.14 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00001_of_00012.arrow
2026-04-10 13:38:32 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00001_of_00012.arrow
Applying chat template (num_proc=12): 0%| | 22/43598 [00:01<33:02, 21.98 examples/s]
Applying chat template (num_proc=12): 1%| | 271/43598 [00:01<02:11, 329.03 examples/s]
Applying chat template (num_proc=12): 0%| | 63/43598 [00:01<12:01, 60.32 examples/s]
Applying chat template (num_proc=12): 1%| | 341/43598 [00:01<01:51, 389.39 examples/s]
Applying chat template (num_proc=12): 1%| | 423/43598 [00:01<01:32, 467.71 examples/s]
Applying chat template (num_proc=12): 0%| | 111/43598 [00:01<05:56, 121.83 examples/s]
Applying chat template (num_proc=12): 0%| | 126/43598 [00:01<05:17, 136.88 examples/s]
Applying chat template (num_proc=12): 1%| | 340/43598 [00:01<02:04, 348.56 examples/s]
Applying chat template (num_proc=12): 1%|▏ | 639/43598 [00:01<01:07, 635.93 examples/s]
Applying chat template (num_proc=12): 3%|▎ | 1336/43598 [00:01<00:28, 1502.30 examples/s]
Applying chat template (num_proc=12): 2%|▏ | 936/43598 [00:01<00:43, 985.53 examples/s]
Applying chat template (num_proc=12): 3%|▎ | 1160/43598 [00:01<00:40, 1053.34 examples/s]
Applying chat template (num_proc=12): 3%|▎ | 1328/43598 [00:01<00:33, 1274.29 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00002_of_00012.arrow
2026-04-10 13:38:33 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00002_of_00012.arrow
Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00003_of_00012.arrow
2026-04-10 13:38:33 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00003_of_00012.arrow
Applying chat template (num_proc=12): 4%|▍ | 1897/43598 [00:01<00:23, 1812.25 examples/s]
Applying chat template (num_proc=12): 5%|▍ | 2132/43598 [00:01<00:20, 2056.49 examples/s]
Applying chat template (num_proc=12): 4%|▍ | 1682/43598 [00:01<00:26, 1572.64 examples/s]
Applying chat template (num_proc=12): 3%|▎ | 1404/43598 [00:01<00:34, 1206.43 examples/s]
Applying chat template (num_proc=12): 5%|▍ | 2000/43598 [00:01<00:25, 1608.13 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00004_of_00012.arrow
2026-04-10 13:38:33 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00004_of_00012.arrow
Applying chat template (num_proc=12): 7%|▋ | 3034/43598 [00:01<00:16, 2464.02 examples/s]
Applying chat template (num_proc=12): 9%|▊ | 3756/43598 [00:01<00:11, 3607.12 examples/s]
Applying chat template (num_proc=12): 7%|▋ | 3138/43598 [00:01<00:17, 2364.14 examples/s]
Applying chat template (num_proc=12): 7%|▋ | 3175/43598 [00:01<00:15, 2533.82 examples/s]
Applying chat template (num_proc=12): 16%|█▌ | 6929/43598 [00:01<00:06, 6072.45 examples/s]
Applying chat template (num_proc=12): 9%|▉ | 3974/43598 [00:01<00:13, 3024.26 examples/s]
Applying chat template (num_proc=12): 11%|█ | 4882/43598 [00:02<00:11, 3448.70 examples/s]
Applying chat template (num_proc=12): 11%|█▏ | 5011/43598 [00:02<00:10, 3649.16 examples/s]
Applying chat template (num_proc=12): 14%|█▍ | 6111/43598 [00:02<00:07, 4691.78 examples/s]
Applying chat template (num_proc=12): 12%|█▏ | 5074/43598 [00:02<00:11, 3388.25 examples/s]
Applying chat template (num_proc=12): 15%|█▌ | 6690/43598 [00:02<00:07, 4636.13 examples/s]
Applying chat template (num_proc=12): 20%|██ | 8885/43598 [00:02<00:05, 6255.74 examples/s]
Applying chat template (num_proc=12): 15%|█▌ | 6584/43598 [00:02<00:08, 4164.55 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00005_of_00012.arrow
2026-04-10 13:38:34 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00005_of_00012.arrow
Applying chat template (num_proc=12): 15%|█▍ | 6388/43598 [00:02<00:09, 3886.84 examples/s]
Applying chat template (num_proc=12): 20%|██ | 8917/43598 [00:02<00:06, 5772.69 examples/s]
Applying chat template (num_proc=12): 19%|█▉ | 8494/43598 [00:02<00:06, 5206.20 examples/s]
Applying chat template (num_proc=12): 22%|██▏ | 9454/43598 [00:02<00:05, 6219.38 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00007_of_00012.arrow
2026-04-10 13:38:34 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00007_of_00012.arrow
Applying chat template (num_proc=12): 20%|██ | 8861/43598 [00:02<00:07, 4781.99 examples/s]
Applying chat template (num_proc=12): 25%|██▍ | 10835/43598 [00:02<00:05, 5835.92 examples/s]
Applying chat template (num_proc=12): 28%|██▊ | 12210/43598 [00:02<00:05, 6209.45 examples/s]
Applying chat template (num_proc=12): 26%|██▌ | 11150/43598 [00:02<00:05, 5792.92 examples/s]
Applying chat template (num_proc=12): 30%|███ | 13215/43598 [00:02<00:03, 7807.57 examples/s]
Applying chat template (num_proc=12): 24%|██▎ | 10325/43598 [00:02<00:06, 5072.15 examples/s]
Applying chat template (num_proc=12): 27%|██▋ | 11591/43598 [00:02<00:05, 5964.63 examples/s]
Applying chat template (num_proc=12): 40%|███▉ | 17299/43598 [00:03<00:02, 11494.16 examples/s]
Applying chat template (num_proc=12): 22%|██▏ | 9566/43598 [00:03<00:07, 4478.78 examples/s]
Applying chat template (num_proc=12): 44%|████▎ | 19028/43598 [00:03<00:02, 11740.53 examples/s]
Applying chat template (num_proc=12): 32%|███▏ | 14155/43598 [00:03<00:04, 6910.36 examples/s]
Applying chat template (num_proc=12): 36%|███▋ | 15881/43598 [00:03<00:03, 7572.24 examples/s]
Applying chat template (num_proc=12): 38%|███▊ | 16683/43598 [00:03<00:03, 7826.60 examples/s]
Applying chat template (num_proc=12): 33%|███▎ | 14271/43598 [00:03<00:04, 6423.94 examples/s]
Applying chat template (num_proc=12): 47%|████▋ | 20604/43598 [00:03<00:01, 11668.41 examples/s]
Applying chat template (num_proc=12): 46%|████▋ | 20221/43598 [00:03<00:02, 10514.44 examples/s]
Applying chat template (num_proc=12): 31%|███▏ | 13644/43598 [00:03<00:05, 5478.94 examples/s]
Applying chat template (num_proc=12): 30%|███ | 13166/43598 [00:03<00:05, 5267.21 examples/s]
Applying chat template (num_proc=12): 51%|█████ | 22105/43598 [00:03<00:01, 11594.96 examples/s]
Applying chat template (num_proc=12): 31%|███▏ | 13638/43598 [00:03<00:04, 6201.69 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00008_of_00012.arrow
2026-04-10 13:38:35 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00008_of_00012.arrow
Applying chat template (num_proc=12): 50%|█████ | 21872/43598 [00:03<00:02, 10527.31 examples/s]
Applying chat template (num_proc=12): 54%|█████▍ | 23494/43598 [00:03<00:01, 11825.06 examples/s]
Applying chat template (num_proc=12): 38%|███▊ | 16627/43598 [00:03<00:03, 6787.25 examples/s]
Applying chat template (num_proc=12): 46%|████▌ | 19986/43598 [00:03<00:02, 8480.70 examples/s]
Applying chat template (num_proc=12): 54%|█████▎ | 23402/43598 [00:03<00:01, 10456.88 examples/s]
Applying chat template (num_proc=12): 41%|████ | 17813/43598 [00:03<00:03, 7374.30 examples/s]
Applying chat template (num_proc=12): 57%|█████▋ | 24887/43598 [00:03<00:01, 11217.58 examples/s]
Applying chat template (num_proc=12): 45%|████▌ | 19707/43598 [00:03<00:02, 9024.21 examples/s]
Applying chat template (num_proc=12): 54%|█████▎ | 23420/43598 [00:03<00:01, 10943.35 examples/s]
Applying chat template (num_proc=12): 52%|█████▏ | 22878/43598 [00:03<00:01, 11239.22 examples/s]
Applying chat template (num_proc=12): 37%|███▋ | 16232/43598 [00:03<00:04, 5794.39 examples/s]
Applying chat template (num_proc=12): 57%|█████▋ | 24769/43598 [00:03<00:01, 10055.79 examples/s]
Applying chat template (num_proc=12): 60%|█████▉ | 26150/43598 [00:03<00:01, 10032.17 examples/s]
Applying chat template (num_proc=12): 48%|████▊ | 21141/43598 [00:03<00:02, 8962.91 examples/s]
Applying chat template (num_proc=12): 41%|████ | 17862/43598 [00:03<00:03, 6932.09 examples/s]
Applying chat template (num_proc=12): 48%|████▊ | 20840/43598 [00:03<00:02, 9676.78 examples/s]
Applying chat template (num_proc=12): 57%|█████▋ | 25025/43598 [00:03<00:01, 11659.65 examples/s]
Applying chat template (num_proc=12): 58%|█████▊ | 25146/43598 [00:03<00:01, 10302.42 examples/s]
Applying chat template (num_proc=12): 39%|███▊ | 16816/43598 [00:03<00:04, 6321.53 examples/s]
Applying chat template (num_proc=12): 63%|██████▎ | 27257/43598 [00:03<00:01, 10248.85 examples/s]
Applying chat template (num_proc=12): 60%|█████▉ | 25942/43598 [00:03<00:01, 9630.70 examples/s]
Applying chat template (num_proc=12): 50%|████▉ | 21721/43598 [00:03<00:02, 9630.16 examples/s]
Applying chat template (num_proc=12): 51%|█████ | 22298/43598 [00:03<00:02, 8862.11 examples/s]
Applying chat template (num_proc=12): 62%|██████▏ | 26915/43598 [00:04<00:01, 12464.24 examples/s]
Applying chat template (num_proc=12): 51%|█████ | 22299/43598 [00:03<00:02, 10541.44 examples/s]
Applying chat template (num_proc=12): 66%|██████▌ | 28698/43598 [00:04<00:01, 10939.34 examples/s]
Applying chat template (num_proc=12): 62%|██████▏ | 27153/43598 [00:04<00:01, 9946.67 examples/s]
Applying chat template (num_proc=12): 52%|█████▏ | 22774/43598 [00:04<00:02, 9669.92 examples/s]
Applying chat template (num_proc=12): 54%|█████▍ | 23508/43598 [00:04<00:02, 9379.12 examples/s]
Applying chat template (num_proc=12): 61%|██████ | 26660/43598 [00:04<00:01, 9705.48 examples/s]
Applying chat template (num_proc=12): 70%|██████▉ | 30385/43598 [00:04<00:01, 12378.53 examples/s]
Applying chat template (num_proc=12): 65%|██████▍ | 28305/43598 [00:04<00:01, 10014.05 examples/s]
Applying chat template (num_proc=12): 57%|█████▋ | 24752/43598 [00:04<00:01, 9942.93 examples/s]
Applying chat template (num_proc=12): 66%|██████▌ | 28846/43598 [00:04<00:01, 12160.33 examples/s]
Applying chat template (num_proc=12): 53%|█████▎ | 23288/43598 [00:04<00:02, 9088.05 examples/s]
Applying chat template (num_proc=12): 56%|█████▌ | 24320/43598 [00:04<00:01, 10827.59 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00006_of_00012.arrow
2026-04-10 13:38:36 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00006_of_00012.arrow
Applying chat template (num_proc=12): 68%|██████▊ | 29586/43598 [00:04<00:01, 10600.97 examples/s]
Applying chat template (num_proc=12): 73%|███████▎ | 31809/43598 [00:04<00:00, 12299.03 examples/s]
Applying chat template (num_proc=12): 59%|█████▉ | 25932/43598 [00:04<00:01, 10321.00 examples/s]
Applying chat template (num_proc=12): 56%|█████▌ | 24460/43598 [00:04<00:02, 9343.77 examples/s]
Applying chat template (num_proc=12): 70%|██████▉ | 30492/43598 [00:04<00:01, 11951.60 examples/s]
Applying chat template (num_proc=12): 56%|█████▋ | 24530/43598 [00:04<00:02, 8945.02 examples/s]
Applying chat template (num_proc=12): 71%|███████ | 30974/43598 [00:04<00:01, 11394.34 examples/s]
Applying chat template (num_proc=12): 76%|███████▋ | 33296/43598 [00:04<00:00, 12722.97 examples/s]
Applying chat template (num_proc=12): 63%|██████▎ | 27458/43598 [00:04<00:01, 11437.26 examples/s]
Applying chat template (num_proc=12): 64%|██████▍ | 28000/43598 [00:04<00:01, 8115.65 examples/s]
Applying chat template (num_proc=12): 60%|██████ | 26175/43598 [00:04<00:01, 10421.90 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00011_of_00012.arrow
2026-04-10 13:38:36 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00011_of_00012.arrow
Applying chat template (num_proc=12): 59%|█████▉ | 25888/43598 [00:04<00:01, 9300.35 examples/s]
Applying chat template (num_proc=12): 60%|█████▉ | 25967/43598 [00:04<00:01, 9650.55 examples/s]
Applying chat template (num_proc=12): 73%|███████▎ | 31990/43598 [00:04<00:00, 11757.08 examples/s]
Applying chat template (num_proc=12): 80%|███████▉ | 34871/43598 [00:04<00:00, 13532.65 examples/s]
Applying chat template (num_proc=12): 64%|██████▎ | 27764/43598 [00:04<00:01, 11018.15 examples/s]
Applying chat template (num_proc=12): 67%|██████▋ | 29148/43598 [00:04<00:01, 8150.13 examples/s]
Applying chat template (num_proc=12): 62%|██████▏ | 27193/43598 [00:04<00:01, 9834.58 examples/s]
Applying chat template (num_proc=12): 63%|██████▎ | 27398/43598 [00:04<00:01, 10436.72 examples/s]
Applying chat template (num_proc=12): 66%|██████▌ | 28758/43598 [00:04<00:01, 10214.11 examples/s]
Applying chat template (num_proc=12): 74%|███████▍ | 32194/43598 [00:04<00:01, 9340.31 examples/s]
Applying chat template (num_proc=12): 84%|████████▍ | 36603/43598 [00:04<00:00, 14524.60 examples/s]
Applying chat template (num_proc=12): 77%|███████▋ | 33520/43598 [00:04<00:00, 12103.69 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00010_of_00012.arrow
2026-04-10 13:38:36 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00010_of_00012.arrow
Applying chat template (num_proc=12): 67%|██████▋ | 29330/43598 [00:04<00:01, 11728.15 examples/s]
Applying chat template (num_proc=12): 70%|██████▉ | 30321/43598 [00:04<00:01, 8658.42 examples/s]
Applying chat template (num_proc=12): 66%|██████▋ | 28983/43598 [00:04<00:01, 11312.58 examples/s]
Applying chat template (num_proc=12): 69%|██████▉ | 30083/43598 [00:04<00:01, 10917.47 examples/s]
Applying chat template (num_proc=12): 76%|███████▋ | 33316/43598 [00:04<00:01, 9765.48 examples/s]
Applying chat template (num_proc=12): 66%|██████▌ | 28761/43598 [00:04<00:01, 10387.48 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00009_of_00012.arrow
2026-04-10 13:38:36 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d3917bc8eb716f92_00009_of_00012.arrow
Applying chat template (num_proc=12): 80%|████████ | 34987/43598 [00:04<00:00, 12548.56 examples/s]
Applying chat template (num_proc=12): 88%|████████▊ | 38170/43598 [00:04<00:00, 14261.91 examples/s]
Applying chat template (num_proc=12): 71%|███████ | 31046/43598 [00:04<00:00, 12754.63 examples/s]
Applying chat template (num_proc=12): 73%|███████▎ | 32030/43598 [00:04<00:01, 10111.71 examples/s]
Applying chat template (num_proc=12): 73%|███████▎ | 31850/43598 [00:04<00:00, 12607.74 examples/s]
Applying chat template (num_proc=12): 71%|███████ | 30784/43598 [00:04<00:01, 12490.47 examples/s]
Applying chat template (num_proc=12): 80%|████████ | 34994/43598 [00:04<00:00, 11485.50 examples/s]
Applying chat template (num_proc=12): 69%|██████▉ | 30247/43598 [00:04<00:01, 11309.21 examples/s]
Applying chat template (num_proc=12): 84%|████████▍ | 36566/43598 [00:04<00:00, 12788.74 examples/s]
Applying chat template (num_proc=12): 91%|█████████ | 39760/43598 [00:04<00:00, 14243.55 examples/s]
Applying chat template (num_proc=12): 78%|███████▊ | 33944/43598 [00:04<00:00, 11965.67 examples/s]
Applying chat template (num_proc=12): 77%|███████▋ | 33764/43598 [00:04<00:00, 14287.66 examples/s]
Applying chat template (num_proc=12): 75%|███████▍ | 32504/43598 [00:04<00:00, 13556.72 examples/s]
Applying chat template (num_proc=12): 84%|████████▍ | 36705/43598 [00:04<00:00, 12941.80 examples/s]
Applying chat template (num_proc=12): 75%|███████▌ | 32844/43598 [00:04<00:00, 12320.60 examples/s]
Applying chat template (num_proc=12): 73%|███████▎ | 31669/43598 [00:04<00:00, 11969.45 examples/s]
Applying chat template (num_proc=12): 88%|████████▊ | 38259/43598 [00:04<00:00, 13780.85 examples/s]
Applying chat template (num_proc=12): 95%|█████████▍| 41341/43598 [00:04<00:00, 13324.22 examples/s]
Applying chat template (num_proc=12): 82%|████████▏ | 35916/43598 [00:04<00:00, 16239.63 examples/s]
Applying chat template (num_proc=12): 79%|███████▉ | 34651/43598 [00:04<00:00, 15478.99 examples/s]
Applying chat template (num_proc=12): 82%|████████▏ | 35780/43598 [00:04<00:00, 13158.84 examples/s]
Applying chat template (num_proc=12): 89%|████████▊ | 38631/43598 [00:04<00:00, 14482.70 examples/s]
Applying chat template (num_proc=12): 79%|███████▉ | 34638/43598 [00:04<00:00, 13509.56 examples/s]
Applying chat template (num_proc=12): 77%|███████▋ | 33417/43598 [00:04<00:00, 13320.45 examples/s]
Applying chat template (num_proc=12): 92%|█████████▏| 39966/43598 [00:05<00:00, 13115.48 examples/s]
Applying chat template (num_proc=12): 86%|████████▋ | 37694/43598 [00:05<00:00, 16495.38 examples/s]
Applying chat template (num_proc=12): 84%|████████▍ | 36751/43598 [00:04<00:00, 16895.70 examples/s]
Applying chat template (num_proc=12): 86%|████████▋ | 37683/43598 [00:05<00:00, 14563.40 examples/s]
Applying chat template (num_proc=12): 98%|█████████▊| 42841/43598 [00:05<00:00, 13275.00 examples/s]
Applying chat template (num_proc=12): 82%|████████▏ | 35653/43598 [00:05<00:00, 15667.27 examples/s]
Applying chat template (num_proc=12): 93%|█████████▎| 40340/43598 [00:05<00:00, 14080.38 examples/s]
Applying chat template (num_proc=12): 84%|████████▎ | 36495/43598 [00:05<00:00, 13657.11 examples/s]
Applying chat template (num_proc=12): 95%|█████████▍| 41360/43598 [00:05<00:00, 13319.28 examples/s]
Applying chat template (num_proc=12): 91%|█████████ | 39682/43598 [00:05<00:00, 17443.41 examples/s]
Applying chat template (num_proc=12): 91%|█████████ | 39751/43598 [00:05<00:00, 15692.25 examples/s]
Applying chat template (num_proc=12): 88%|████████▊ | 38177/43598 [00:05<00:00, 18246.25 examples/s]
Applying chat template (num_proc=12): 89%|████████▉ | 38717/43598 [00:05<00:00, 16352.15 examples/s]
Applying chat template (num_proc=12): 87%|████████▋ | 38075/43598 [00:05<00:00, 14160.09 examples/s]
Applying chat template (num_proc=12): 97%|█████████▋| 42364/43598 [00:05<00:00, 15140.93 examples/s]
Applying chat template (num_proc=12): 96%|█████████▌| 41788/43598 [00:05<00:00, 16494.29 examples/s]
Applying chat template (num_proc=12): 94%|█████████▎| 40825/43598 [00:05<00:00, 17275.08 examples/s]
Applying chat template (num_proc=12): 92%|█████████▏| 40166/43598 [00:05<00:00, 17975.81 examples/s]
Applying chat template (num_proc=12): 96%|█████████▌| 41702/43598 [00:05<00:00, 15936.46 examples/s]
Applying chat template (num_proc=12): 91%|█████████ | 39705/43598 [00:05<00:00, 14457.72 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs6b1b54126f0c5c7f000017c8'
Applying chat template (num_proc=12): 99%|█████████▊| 42965/43598 [00:05<00:00, 11330.73 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 8156.47 examples/s]
Applying chat template (num_proc=12): 95%|█████████▌| 41512/43598 [00:05<00:00, 15401.02 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs41589db42cf35be2000017c9'
Applying chat template (num_proc=12): 100%|█████████▉| 43418/43598 [00:05<00:00, 15552.63 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 15012.52 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 8050.64 examples/s]
Applying chat template (num_proc=12): 98%|█████████▊| 42662/43598 [00:05<00:00, 15738.77 examples/s]
Applying chat template (num_proc=12): 97%|█████████▋| 42145/43598 [00:05<00:00, 16212.44 examples/s]
Applying chat template (num_proc=12): 99%|█████████▉| 43148/43598 [00:05<00:00, 15370.57 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs48dd7446b3148e06000017d4'
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 7823.71 examples/s]
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsce245982b5ea6739000017d7'
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsba63fcbf067574be000017d8'
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 7759.10 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 7742.02 examples/s]
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfse90b0255ee86dfd4000017d9'
Concatenating 12 shards
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Concatenating 12 shards
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs5e49ad14b0fadbef000017db'
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 7727.41 examples/s]
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsa53626dc662ef5e4000017da'
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 7683.55 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 43598/43598 [00:05<00:00, 7672.16 examples/s]
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]Process #0 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00000_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #0 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00000_of_00012.arrow
Process #1 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00001_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #1 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00001_of_00012.arrow
Process #2 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00002_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #2 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00002_of_00012.arrow
Process #3 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00003_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #3 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00003_of_00012.arrow
Process #4 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00004_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #4 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00004_of_00012.arrow
Process #5 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00005_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #5 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00005_of_00012.arrow
Process #6 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00006_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #6 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00006_of_00012.arrow
Process #7 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00007_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #7 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00007_of_00012.arrow
Process #8 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00008_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #8 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00008_of_00012.arrow
Process #9 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00009_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #9 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00009_of_00012.arrow
Process #10 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00010_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #10 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00010_of_00012.arrow
Process #11 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00011_of_00012.arrow
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Process #11 will write at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00011_of_00012.arrow
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]Spawning 12 processes
2026-04-10 13:38:37 - INFO - datasets.arrow_dataset - Spawning 12 processes
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 0%| | 0/2339 [00:00<?, ? examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00000_of_00012.arrow
2026-04-10 13:38:38 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00000_of_00012.arrow
Applying chat template (num_proc=12): 8%|▊ | 195/2339 [00:00<00:08, 261.03 examples/s]
Applying chat template (num_proc=12): 5%|▌ | 125/2339 [00:00<00:15, 146.41 examples/s]
Applying chat template (num_proc=12): 13%|█▎ | 303/2339 [00:00<00:06, 330.98 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00001_of_00012.arrow
2026-04-10 13:38:38 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00001_of_00012.arrow
Applying chat template (num_proc=12): 5%|▌ | 121/2339 [00:00<00:17, 126.59 examples/s]
Applying chat template (num_proc=12): 0%| | 9/2339 [00:01<04:36, 8.42 examples/s]
Applying chat template (num_proc=12): 8%|▊ | 196/2339 [00:01<00:10, 201.88 examples/s]
Applying chat template (num_proc=12): 6%|▌ | 136/2339 [00:00<00:14, 153.07 examples/s]
Applying chat template (num_proc=12): 3%|▎ | 66/2339 [00:01<00:34, 65.30 examples/s]
Applying chat template (num_proc=12): 17%|█▋ | 401/2339 [00:01<00:05, 359.35 examples/s]
Applying chat template (num_proc=12): 3%|▎ | 61/2339 [00:00<00:34, 66.45 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00002_of_00012.arrow
2026-04-10 13:38:38 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00002_of_00012.arrow
Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00003_of_00012.arrow
Applying chat template (num_proc=12): 3%|▎ | 80/2339 [00:01<00:29, 75.97 examples/s]2026-04-10 13:38:38 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00003_of_00012.arrow
Applying chat template (num_proc=12): 17%|█▋ | 391/2339 [00:01<00:04, 396.62 examples/s]
Applying chat template (num_proc=12): 12%|█▏ | 291/2339 [00:01<00:07, 274.80 examples/s]
Applying chat template (num_proc=12): 8%|▊ | 195/2339 [00:01<00:12, 178.08 examples/s]
Applying chat template (num_proc=12): 16%|█▋ | 381/2339 [00:01<00:04, 397.10 examples/s]
Applying chat template (num_proc=12): 17%|█▋ | 391/2339 [00:01<00:05, 374.20 examples/s]
Applying chat template (num_proc=12): 28%|██▊ | 651/2339 [00:01<00:02, 565.18 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00004_of_00012.arrow
2026-04-10 13:38:39 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00004_of_00012.arrow
Applying chat template (num_proc=12): 12%|█▏ | 292/2339 [00:01<00:07, 283.62 examples/s]
Applying chat template (num_proc=12): 28%|██▊ | 644/2339 [00:01<00:02, 631.55 examples/s]
Applying chat template (num_proc=12): 23%|██▎ | 531/2339 [00:01<00:03, 488.41 examples/s]
Applying chat template (num_proc=12): 11%|█ | 261/2339 [00:01<00:08, 231.21 examples/s]
Applying chat template (num_proc=12): 40%|███▉ | 932/2339 [00:01<00:01, 769.38 examples/s]
Applying chat template (num_proc=12): 28%|██▊ | 644/2339 [00:01<00:03, 521.57 examples/s]
Applying chat template (num_proc=12): 30%|██▉ | 694/2339 [00:01<00:02, 592.36 examples/s]
Applying chat template (num_proc=12): 19%|█▊ | 436/2339 [00:01<00:05, 359.51 examples/s]
Applying chat template (num_proc=12): 22%|██▏ | 522/2339 [00:01<00:03, 458.77 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00005_of_00012.arrow
2026-04-10 13:38:39 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00005_of_00012.arrow
Applying chat template (num_proc=12): 33%|███▎ | 780/2339 [00:01<00:02, 648.31 examples/s]
Applying chat template (num_proc=12): 47%|████▋ | 1106/2339 [00:01<00:01, 714.44 examples/s]
Applying chat template (num_proc=12): 36%|███▌ | 832/2339 [00:01<00:02, 570.07 examples/s]
Applying chat template (num_proc=12): 42%|████▏ | 976/2339 [00:01<00:01, 690.96 examples/s]
Applying chat template (num_proc=12): 17%|█▋ | 404/2339 [00:01<00:06, 279.84 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00006_of_00012.arrow
2026-04-10 13:38:39 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00006_of_00012.arrow
Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00007_of_00012.arrow
2026-04-10 13:38:39 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00007_of_00012.arrow
Applying chat template (num_proc=12): 42%|████▏ | 975/2339 [00:01<00:02, 627.34 examples/s]
Applying chat template (num_proc=12): 52%|█████▏ | 1225/2339 [00:02<00:01, 795.32 examples/s]
Applying chat template (num_proc=12): 28%|██▊ | 645/2339 [00:01<00:04, 397.79 examples/s]
Applying chat template (num_proc=12): 50%|█████ | 1170/2339 [00:02<00:01, 769.33 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00008_of_00012.arrow
2026-04-10 13:38:39 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00008_of_00012.arrow
Applying chat template (num_proc=12): 58%|█████▊ | 1365/2339 [00:02<00:01, 715.72 examples/s]
Applying chat template (num_proc=12): 36%|███▋ | 850/2339 [00:02<00:03, 426.58 examples/s]
Applying chat template (num_proc=12): 69%|██████▉ | 1625/2339 [00:02<00:00, 1204.88 examples/s]
Applying chat template (num_proc=12): 27%|██▋ | 623/2339 [00:02<00:05, 330.00 examples/s]
Applying chat template (num_proc=12): 58%|█████▊ | 1365/2339 [00:02<00:01, 948.15 examples/s]
Applying chat template (num_proc=12): 35%|███▍ | 811/2339 [00:02<00:02, 518.56 examples/s]
Applying chat template (num_proc=12): 47%|████▋ | 1105/2339 [00:02<00:01, 751.58 examples/s]
Applying chat template (num_proc=12): 65%|██████▌ | 1525/2339 [00:02<00:00, 863.44 examples/s]
Applying chat template (num_proc=12): 42%|████▏ | 975/2339 [00:02<00:02, 582.98 examples/s]
Applying chat template (num_proc=12): 37%|███▋ | 858/2339 [00:02<00:03, 455.59 examples/s]
Applying chat template (num_proc=12): 53%|█████▎ | 1242/2339 [00:02<00:01, 714.89 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00009_of_00012.arrow
2026-04-10 13:38:40 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00009_of_00012.arrow
Applying chat template (num_proc=12): 78%|███████▊ | 1819/2339 [00:02<00:00, 871.93 examples/s]
Applying chat template (num_proc=12): 58%|█████▊ | 1365/2339 [00:02<00:01, 580.88 examples/s]
Applying chat template (num_proc=12): 66%|██████▌ | 1535/2339 [00:02<00:01, 568.22 examples/s]
Applying chat template (num_proc=12): 58%|█████▊ | 1352/2339 [00:02<00:01, 869.57 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00010_of_00012.arrow
2026-04-10 13:38:40 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00010_of_00012.arrow
Applying chat template (num_proc=12): 53%|█████▎ | 1235/2339 [00:02<00:02, 537.65 examples/s]
Applying chat template (num_proc=12): 89%|████████▉ | 2079/2339 [00:02<00:00, 1056.37 examples/s]
Applying chat template (num_proc=12): 58%|█████▊ | 1365/2339 [00:02<00:01, 684.50 examples/s]
Applying chat template (num_proc=12): 48%|████▊ | 1128/2339 [00:02<00:02, 593.72 examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00011_of_00012.arrow
2026-04-10 13:38:40 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-0f820217b8a8b27e_00011_of_00012.arrow
Applying chat template (num_proc=12): 75%|███████▌ | 1755/2339 [00:02<00:00, 1202.39 examples/s]
Applying chat template (num_proc=12): 83%|████████▎ | 1950/2339 [00:02<00:00, 969.14 examples/s]
Applying chat template (num_proc=12): 83%|████████▎ | 1950/2339 [00:02<00:00, 1037.60 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsc7f9bea9a34682f900001814'
Applying chat template (num_proc=12): 67%|██████▋ | 1560/2339 [00:02<00:00, 812.93 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:02<00:00, 790.02 examples/s]
Applying chat template (num_proc=12): 65%|██████▍ | 1519/2339 [00:03<00:01, 673.72 examples/s]Concatenating 12 shards
2026-04-10 13:38:40 - INFO - datasets.arrow_dataset - Concatenating 12 shards
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 83%|████████▎ | 1948/2339 [00:02<00:00, 1174.18 examples/s]
Applying chat template (num_proc=12): 74%|███████▎ | 1724/2339 [00:03<00:00, 768.94 examples/s]
Applying chat template (num_proc=12): 92%|█████████▏| 2145/2339 [00:03<00:00, 970.15 examples/s]
Applying chat template (num_proc=12): 90%|████████▉ | 2094/2339 [00:03<00:00, 849.83 examples/s]
Applying chat template (num_proc=12): 83%|████████▎ | 1950/2339 [00:02<00:00, 1028.68 examples/s]
Applying chat template (num_proc=12): 88%|████████▊ | 2052/2339 [00:03<00:00, 768.81 examples/s]
Applying chat template (num_proc=12): 58%|█████▊ | 1366/2339 [00:02<00:01, 595.06 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 1046.21 examples/s]
Applying chat template (num_proc=12): 92%|█████████▏| 2145/2339 [00:03<00:00, 1140.39 examples/s]
Applying chat template (num_proc=12): 83%|████████▎ | 1950/2339 [00:03<00:00, 857.46 examples/s]
Applying chat template (num_proc=12): 92%|█████████▏| 2145/2339 [00:03<00:00, 1083.43 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsf04c0444c06ed9dc00001822'
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 966.32 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 718.19 examples/s]
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 83%|████████▎ | 1950/2339 [00:03<00:00, 1060.92 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs83e9dfb0602b1a1800001826'
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 696.12 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 1201.82 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 1124.18 examples/s]
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 99%|█████████▉| 2324/2339 [00:03<00:00, 789.78 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
Traceback (most recent call last):
SystemExit: 0
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs88e529554f7d28b80000182a'
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs41bb7f3dcce740cd0000182b'
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 1100.59 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 694.22 examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 694.19 examples/s]
Applying chat template (num_proc=12): 92%|█████████▏| 2145/2339 [00:03<00:00, 1118.95 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfs52010f168a9739e20000182e'
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 633.80 examples/s]
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsf45fad7f528b56430000182f'
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 635.95 examples/s]
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 1062.70 examples/s]Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 314, in _bootstrap
self.run()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 600, in _run_server
server.serve_forever()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/managers.py", line 184, in serve_forever
sys.exit(0)
SystemExit: 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 300, in _run_finalizers
finalizer()
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 224, in __call__
res = self._callback(*self._args, **self._kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/multiprocess/util.py", line 133, in _remove_temp_dir
rmtree(tempdir)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 752, in rmtree
_rmtree_safe_fd(fd, path, onerror)
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 703, in _rmtree_safe_fd
onerror(os.unlink, fullname, sys.exc_info())
File "/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/shutil.py", line 701, in _rmtree_safe_fd
os.unlink(entry.name, dir_fd=topfd)
OSError: [Errno 16] Device or resource busy: '.nfsfc85ae0694b60e5100001831'
Applying chat template (num_proc=12): 100%|██████████| 2339/2339 [00:03<00:00, 642.64 examples/s]
Filter: 0%| | 0/43598 [00:00<?, ? examples/s]Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d924339979854155.arrow
2026-04-10 13:38:50 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-d924339979854155.arrow
Filter: 23%|██▎ | 10000/43598 [00:09<00:31, 1060.64 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:30, 1101.67 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:30, 1085.72 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:31, 1056.15 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:31, 1077.72 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:31, 1060.72 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:31, 1071.19 examples/s]
Filter: 23%|██▎ | 10000/43598 [00:09<00:31, 1075.25 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:22, 1071.34 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:21, 1106.41 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:21, 1104.77 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:21, 1091.13 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:21, 1082.12 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:21, 1076.15 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:21, 1084.44 examples/s]
Filter: 46%|████▌ | 20000/43598 [00:18<00:22, 1070.42 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1100.97 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1098.03 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1074.07 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1083.51 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1088.19 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1074.76 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1084.57 examples/s]
Filter: 69%|██████▉ | 30000/43598 [00:27<00:12, 1072.17 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:36<00:03, 1099.29 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:37<00:03, 1083.81 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:36<00:03, 1092.73 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:36<00:03, 1092.84 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:36<00:03, 1088.25 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:36<00:03, 1084.25 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:36<00:03, 1091.77 examples/s]
Filter: 92%|█████████▏| 40000/43598 [00:37<00:03, 1079.73 examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1098.72 examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1087.31 examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1098.90 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1097.25 examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1079.87 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1089.45 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1097.41 examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1090.10 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1088.43 examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1091.34 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1088.80 examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1082.77 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:39<00:00, 1096.15 examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1089.78 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1086.32 examples/s]
Filter: 100%|██████████| 43598/43598 [00:40<00:00, 1078.74 examples/s]
Filter: 0%| | 0/2339 [00:00<?, ? examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1111.72 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1081.44 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1106.62 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1077.69 examples/s]
Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-45f8b5bf4d93c54f.arrow
2026-04-10 13:39:23 - INFO - datasets.arrow_dataset - Caching processed dataset at /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/Anthropic___hh-rlhf/default-cfba128a0ab1b99f/0.0.0/09be8c5bbc57cb3887f3a9732ad6aa7ec602a1fa/cache-45f8b5bf4d93c54f.arrow
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1086.30 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1080.21 examples/s]
2026-04-10 13:39:23 - INFO - __main__ - Decontaminated 0 (0.00%) samples from the training set.
2026-04-10 13:39:23 - INFO - __main__ - Sample 41905 of the processed training set:
<|begin_of_text|><|start_header_id|>user<|end_header_id|>
What is the best way to clean my refrigerator?<|eot_id|><|start_header_id|>assistant<|end_header_id|>
Well, I have a few ideas, but the best way to know what works best is to run some tests and then analyze the data to figure out which technique works best.<|eot_id|><|start_header_id|>user<|end_header_id|>
What are your ideas?<|eot_id|><|start_header_id|>assistant<|end_header_id|>
Some of my ideas include wiping down the interior with a rag, using bleach and water to wipe down the door, emptying the contents of the fridge to scrub the shelves with a clean cloth, cleaning the interior of the fridge with vinegar and water, and using a vacuum cleaner to clean the interior.<|eot_id|>
2026-04-10 13:39:23 - INFO - __main__ - Sample 7296 of the processed training set:
<|begin_of_text|><|start_header_id|>user<|end_header_id|>
How can I get blood out of cotton?<|eot_id|><|start_header_id|>assistant<|end_header_id|>
Blood? What do you mean?<|eot_id|>
2026-04-10 13:39:23 - INFO - __main__ - Sample 1639 of the processed training set:
<|begin_of_text|><|start_header_id|>user<|end_header_id|>
I'm trying to get my friend to open up<|eot_id|><|start_header_id|>assistant<|end_header_id|>
Sounds great! What have you tried so far?<|eot_id|><|start_header_id|>user<|end_header_id|>
I've tried to give my friend a safe space to open up about their feelings but they still aren't comfortable<|eot_id|><|start_header_id|>assistant<|end_header_id|>
It’s really important that people feel safe sharing their feelings, but I totally hear that this isn’t easy for them. And in my experience, that has a lot to do with them feeling that you can accept them and what they’re saying, even if they’re feeling difficult or uncomfortable feelings. Do you think you’ve done that, and that they still haven’t opened up?<|eot_id|>
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1093.93 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1088.80 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1092.40 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1083.66 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1074.30 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1068.17 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1076.92 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1072.00 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1107.86 examples/s]
Filter: 100%|██████████| 2339/2339 [00:02<00:00, 1103.67 examples/s]
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': model_init_kwargs, dataset_text_field, max_seq_length, packing. Will not be supported from version '1.0.0'.
Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.
warnings.warn(message, FutureWarning)
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:158: UserWarning: You passed `model_init_kwargs` to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:185: UserWarning: You passed a model_id to the SFTTrainer. This will automatically create an `AutoModelForCausalLM` or a `PeftModel` (if you passed a `peft_config`) for you.
warnings.warn(
[INFO|configuration_utils.py:691] 2026-04-10 13:39:26,149 >> loading configuration file /scratch/feng.yulu/dynamic-dpo-v4/base_models/Meta-Llama-3-8B/config.json
[INFO|configuration_utils.py:765] 2026-04-10 13:39:26,151 >> Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 8192,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.51.0",
"use_cache": false,
"vocab_size": 128256
}
[INFO|modeling_utils.py:1121] 2026-04-10 13:39:26,165 >> loading weights file /scratch/feng.yulu/dynamic-dpo-v4/base_models/Meta-Llama-3-8B/model.safetensors.index.json
[WARNING|logging.py:328] 2026-04-10 13:39:26,166 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[INFO|modeling_utils.py:2167] 2026-04-10 13:39:26,166 >> Instantiating LlamaForCausalLM model under default dtype torch.bfloat16.
[WARNING|logging.py:328] 2026-04-10 13:39:26,166 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[WARNING|logging.py:328] 2026-04-10 13:39:26,166 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[WARNING|logging.py:328] 2026-04-10 13:39:26,166 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[WARNING|logging.py:328] 2026-04-10 13:39:26,169 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[WARNING|logging.py:328] 2026-04-10 13:39:26,170 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[WARNING|logging.py:328] 2026-04-10 13:39:26,170 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[WARNING|logging.py:328] 2026-04-10 13:39:26,170 >> You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
[INFO|configuration_utils.py:1142] 2026-04-10 13:39:26,171 >> Generate config GenerationConfig {
"bos_token_id": 128000,
"eos_token_id": 128001,
"use_cache": false
}
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 463.37it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 473.40it/s]
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 464.76it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 471.85it/s]
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 349.00it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 349.98it/s]
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 360.32it/s]
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 82.44it/s]
[INFO|modeling_utils.py:4926] 2026-04-10 13:39:26,261 >> All model checkpoint weights were used when initializing LlamaForCausalLM.
[INFO|modeling_utils.py:4934] 2026-04-10 13:39:26,262 >> All the weights of LlamaForCausalLM were initialized from the model checkpoint at /scratch/feng.yulu/dynamic-dpo-v4/base_models/Meta-Llama-3-8B.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
[INFO|configuration_utils.py:1095] 2026-04-10 13:39:26,264 >> loading configuration file /scratch/feng.yulu/dynamic-dpo-v4/base_models/Meta-Llama-3-8B/generation_config.json
[INFO|configuration_utils.py:1142] 2026-04-10 13:39:26,265 >> Generate config GenerationConfig {
"bos_token_id": 128000,
"do_sample": true,
"eos_token_id": 128001,
"max_length": 4096,
"temperature": 0.6,
"top_p": 0.9
}
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:195: UserWarning: You passed a `packing` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:283: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:321: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.
warnings.warn(
Using custom data configuration default-39b52f6e03e85a82
2026-04-10 13:39:26 - INFO - datasets.builder - Using custom data configuration default-39b52f6e03e85a82
Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/generator
2026-04-10 13:39:26 - INFO - datasets.info - Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/generator
Generating dataset generator (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-39b52f6e03e85a82/0.0.0)
2026-04-10 13:39:26 - INFO - datasets.builder - Generating dataset generator (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-39b52f6e03e85a82/0.0.0)
Downloading and preparing dataset generator/default to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-39b52f6e03e85a82/0.0.0...
2026-04-10 13:39:26 - INFO - datasets.builder - Downloading and preparing dataset generator/default to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-39b52f6e03e85a82/0.0.0...
Generating train split
2026-04-10 13:39:26 - INFO - datasets.builder - Generating train split
Generating train split: 0 examples [00:00, ? examples/s]
Generating train split: 1 examples [00:00, 1.74 examples/s]
Generating train split: 705 examples [00:01, 599.40 examples/s]
Generating train split: 1413 examples [00:01, 818.34 examples/s]
Generating train split: 2120 examples [00:02, 927.30 examples/s]
Generating train split: 2825 examples [00:03, 999.25 examples/s]
Generating train split: 3530 examples [00:03, 1022.46 examples/s]
Generating train split: 4236 examples [00:04, 1046.23 examples/s]
Generating train split: 4943 examples [00:05, 1075.23 examples/s]
Generating train split: 5650 examples [00:05, 1079.58 examples/s]
Generating train split: 6358 examples [00:06, 1008.34 examples/s]
Generating train split: 7061 examples [00:07, 1030.47 examples/s]
Generating train split: 7767 examples [00:07, 1054.48 examples/s]
Generating train split: 8476 examples [00:08, 1064.89 examples/s]
Generating train split: 9186 examples [00:09, 1074.60 examples/s]
Generating train split: 9895 examples [00:09, 1095.85 examples/s]
Generating train split: 10604 examples [00:10, 1094.68 examples/s]
Generating train split: 11310 examples [00:11, 1092.87 examples/s]
Generating train split: 12000 examples [00:11, 1364.84 examples/s]
Generating train split: 12723 examples [00:12, 974.19 examples/s]
Generating train split: 13430 examples [00:13, 1005.80 examples/s]
Generating train split: 14136 examples [00:13, 1032.77 examples/s]
Generating train split: 14842 examples [00:14, 1063.97 examples/s]
Generating train split: 15550 examples [00:15, 1071.50 examples/s]
Generating train split: 16260 examples [00:15, 1300.26 examples/s]
Generating train split: 16516 examples [00:15, 1064.79 examples/s]
Unable to verify splits sizes.
2026-04-10 13:39:41 - INFO - datasets.utils.info_utils - Unable to verify splits sizes.
Dataset generator downloaded and prepared to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-39b52f6e03e85a82/0.0.0. Subsequent calls will reuse this data.
2026-04-10 13:39:41 - INFO - datasets.builder - Dataset generator downloaded and prepared to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-39b52f6e03e85a82/0.0.0. Subsequent calls will reuse this data.
Using custom data configuration default-1519231937de8df3
2026-04-10 13:39:42 - INFO - datasets.builder - Using custom data configuration default-1519231937de8df3
Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/generator
2026-04-10 13:39:42 - INFO - datasets.info - Loading Dataset Infos from /home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/datasets/packaged_modules/generator
Generating dataset generator (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-1519231937de8df3/0.0.0)
2026-04-10 13:39:42 - INFO - datasets.builder - Generating dataset generator (/scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-1519231937de8df3/0.0.0)
Downloading and preparing dataset generator/default to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-1519231937de8df3/0.0.0...
2026-04-10 13:39:42 - INFO - datasets.builder - Downloading and preparing dataset generator/default to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-1519231937de8df3/0.0.0...
Generating train split
2026-04-10 13:39:42 - INFO - datasets.builder - Generating train split
Generating train split: 0 examples [00:00, ? examples/s]
Generating train split: 1 examples [00:00, 1.72 examples/s]
Generating train split: 711 examples [00:00, 1190.80 examples/s]
Generating train split: 895 examples [00:00, 1099.18 examples/s]
Unable to verify splits sizes.
2026-04-10 13:39:42 - INFO - datasets.utils.info_utils - Unable to verify splits sizes.
Dataset generator downloaded and prepared to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-1519231937de8df3/0.0.0. Subsequent calls will reuse this data.
2026-04-10 13:39:43 - INFO - datasets.builder - Dataset generator downloaded and prepared to /scratch/feng.yulu/dynamic-dpo-v4/hf/datasets/generator/default-1519231937de8df3/0.0.0. Subsequent calls will reuse this data.
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/trl/trainer/sft_trainer.py:412: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `SFTTrainer.__init__`. Use `processing_class` instead.
super().__init__(
[INFO|trainer.py:748] 2026-04-10 13:39:44,030 >> Using auto half precision backend
2026-04-10 13:39:44 - INFO - __main__ - *** Train ***
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/accelerate/accelerator.py:1557: UserWarning: Upcasted low precision parameters in LlamaForCausalLM because mixed precision turned on in FSDP. Affects: model.embed_tokens.weight, model.norm.weight, lm_head.weight.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/accelerate/accelerator.py:1557: UserWarning: Upcasted low precision parameters in LlamaDecoderLayer because mixed precision turned on in FSDP. Affects: self_attn.q_proj.weight, self_attn.k_proj.weight, self_attn.v_proj.weight, self_attn.o_proj.weight, mlp.gate_proj.weight, mlp.up_proj.weight, mlp.down_proj.weight, input_layernorm.weight, post_attention_layernorm.weight.
warnings.warn(
/home/feng.yulu/.conda/envs/dpo_venv/lib/python3.11/site-packages/accelerate/accelerator.py:1563: UserWarning: FSDP upcast of low precision parameters may affect the precision of model checkpoints.
warnings.warn(
[INFO|trainer.py:2414] 2026-04-10 13:39:48,781 >> ***** Running training *****
[INFO|trainer.py:2415] 2026-04-10 13:39:48,781 >> Num examples = 16,516
[INFO|trainer.py:2416] 2026-04-10 13:39:48,781 >> Num Epochs = 1
[INFO|trainer.py:2417] 2026-04-10 13:39:48,781 >> Instantaneous batch size per device = 16
[INFO|trainer.py:2420] 2026-04-10 13:39:48,781 >> Total train batch size (w. parallel, distributed & accumulation) = 128
[INFO|trainer.py:2421] 2026-04-10 13:39:48,781 >> Gradient Accumulation steps = 1
[INFO|trainer.py:2422] 2026-04-10 13:39:48,781 >> Total optimization steps = 130
[INFO|trainer.py:2423] 2026-04-10 13:39:48,781 >> Number of trainable parameters = 1,003,782,656
[INFO|integration_utils.py:831] 2026-04-10 13:39:48,782 >> Automatic Weights & Biases logging enabled, to disable set os.environ["WANDB_DISABLED"] = "true"
wandb: Currently logged in as: can-not-fand (can-not-fand-northeastern-university). Use `wandb login --relogin` to force relogin
wandb: wandb version 0.25.1 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.5
wandb: Run data is saved locally in /scratch/feng.yulu/dynamic-dpo-v4/wandb/wandb/run-20260410_133950-cqdvywmp
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758
wandb: ⭐️ View project at https://wandb.ai/can-not-fand-northeastern-university/huggingface
wandb: 🚀 View run at https://wandb.ai/can-not-fand-northeastern-university/huggingface/runs/cqdvywmp
0%| | 0/130 [00:00<?, ?it/s]
1%| | 1/130 [00:01<03:05, 1.44s/it]
{'loss': 3.1196, 'grad_norm': inf, 'learning_rate': 0.0, 'epoch': 0.01}
1%| | 1/130 [00:01<03:05, 1.44s/it]
2%|▏ | 2/130 [00:02<02:44, 1.29s/it]
2%|▏ | 3/130 [00:03<02:35, 1.23s/it]
3%|▎ | 4/130 [00:04<02:30, 1.20s/it]
4%|▍ | 5/130 [00:06<02:27, 1.18s/it]
{'loss': 3.1223, 'grad_norm': inf, 'learning_rate': 6.153846153846155e-06, 'epoch': 0.04}
4%|▍ | 5/130 [00:06<02:27, 1.18s/it]
5%|▍ | 6/130 [00:07<02:25, 1.17s/it]
5%|▌ | 7/130 [00:08<02:23, 1.16s/it]
6%|▌ | 8/130 [00:09<02:21, 1.16s/it]
7%|▋ | 9/130 [00:10<02:19, 1.16s/it]
8%|▊ | 10/130 [00:11<02:18, 1.15s/it]
{'loss': 3.1246, 'grad_norm': inf, 'learning_rate': 1.3846153846153847e-05, 'epoch': 0.08}
8%|▊ | 10/130 [00:11<02:18, 1.15s/it]
8%|▊ | 11/130 [00:12<02:16, 1.15s/it]
9%|▉ | 12/130 [00:14<02:15, 1.15s/it]
10%|█ | 13/130 [00:15<02:14, 1.15s/it]
11%|█ | 14/130 [00:16<02:13, 1.15s/it]
12%|█▏ | 15/130 [00:17<02:12, 1.15s/it]
{'loss': 3.1264, 'grad_norm': inf, 'learning_rate': 1.9996395276708856e-05, 'epoch': 0.12}
12%|█▏ | 15/130 [00:17<02:12, 1.15s/it]
12%|█▏ | 16/130 [00:18<02:11, 1.15s/it]
13%|█▎ | 17/130 [00:19<02:09, 1.15s/it]
14%|█▍ | 18/130 [00:21<02:08, 1.15s/it]
15%|█▍ | 19/130 [00:22<02:07, 1.15s/it]
15%|█▌ | 20/130 [00:23<02:06, 1.15s/it]
{'loss': 2.9981, 'grad_norm': 35.01002883911133, 'learning_rate': 1.9870502626379127e-05, 'epoch': 0.15}
15%|█▌ | 20/130 [00:23<02:06, 1.15s/it]
16%|█▌ | 21/130 [00:24<02:05, 1.15s/it]
17%|█▋ | 22/130 [00:25<02:04, 1.15s/it]
18%|█▊ | 23/130 [00:26<02:03, 1.15s/it]
18%|█▊ | 24/130 [00:27<02:02, 1.15s/it]
19%|█▉ | 25/130 [00:29<02:00, 1.15s/it]
{'loss': 2.531, 'grad_norm': 11.861149787902832, 'learning_rate': 1.9566964208274254e-05, 'epoch': 0.19}
19%|█▉ | 25/130 [00:29<02:00, 1.15s/it]
20%|██ | 26/130 [00:30<01:59, 1.15s/it]
21%|██ | 27/130 [00:31<01:58, 1.15s/it]
22%|██▏ | 28/130 [00:32<01:57, 1.15s/it]
22%|██▏ | 29/130 [00:33<01:56, 1.15s/it]
23%|██▎ | 30/130 [00:34<01:55, 1.15s/it]
{'loss': 2.2014, 'grad_norm': 7.360829830169678, 'learning_rate': 1.909124299802724e-05, 'epoch': 0.23}
23%|██▎ | 30/130 [00:34<01:55, 1.15s/it]
24%|██▍ | 31/130 [00:35<01:53, 1.15s/it]
25%|██▍ | 32/130 [00:37<01:52, 1.15s/it]
25%|██▌ | 33/130 [00:38<01:51, 1.15s/it]
26%|██▌ | 34/130 [00:39<01:50, 1.15s/it]
27%|██▋ | 35/130 [00:40<01:49, 1.15s/it]
{'loss': 2.0137, 'grad_norm': 6.061845779418945, 'learning_rate': 1.845190085543795e-05, 'epoch': 0.27}
27%|██▋ | 35/130 [00:40<01:49, 1.15s/it]
28%|██▊ | 36/130 [00:41<01:53, 1.20s/it]
28%|██▊ | 37/130 [00:43<01:50, 1.19s/it]
29%|██▉ | 38/130 [00:44<01:48, 1.18s/it]
30%|███ | 39/130 [00:45<01:46, 1.17s/it]
31%|███ | 40/130 [00:46<01:45, 1.17s/it]
{'loss': 1.8515, 'grad_norm': 3.25486421585083, 'learning_rate': 1.766044443118978e-05, 'epoch': 0.31}
31%|███ | 40/130 [00:46<01:45, 1.17s/it]
32%|███▏ | 41/130 [00:47<01:43, 1.17s/it]
32%|███▏ | 42/130 [00:48<01:42, 1.16s/it]
33%|███▎ | 43/130 [00:50<01:41, 1.16s/it]
34%|███▍ | 44/130 [00:51<01:39, 1.16s/it]
35%|███▍ | 45/130 [00:52<01:38, 1.16s/it]
{'loss': 1.7909, 'grad_norm': 2.8577518463134766, 'learning_rate': 1.67311180742757e-05, 'epoch': 0.35}
35%|███▍ | 45/130 [00:52<01:38, 1.16s/it]
35%|███▌ | 46/130 [00:53<01:37, 1.16s/it]
36%|███▌ | 47/130 [00:54<01:36, 1.16s/it]
37%|███▋ | 48/130 [00:55<01:35, 1.16s/it]
38%|███▊ | 49/130 [00:56<01:33, 1.16s/it]
38%|███▊ | 50/130 [00:58<01:32, 1.16s/it]
{'loss': 1.7104, 'grad_norm': 3.2024171352386475, 'learning_rate': 1.568064746731156e-05, 'epoch': 0.38}
38%|███▊ | 50/130 [00:58<01:32, 1.16s/it]
39%|███▉ | 51/130 [00:59<01:31, 1.16s/it]
40%|████ | 52/130 [01:00<01:30, 1.16s/it]
41%|████ | 53/130 [01:01<01:29, 1.16s/it]
42%|████▏ | 54/130 [01:02<01:31, 1.20s/it]
42%|████▏ | 55/130 [01:04<01:29, 1.19s/it]
{'loss': 1.6393, 'grad_norm': 2.368898630142212, 'learning_rate': 1.4527938603696376e-05, 'epoch': 0.42}
42%|████▏ | 55/130 [01:04<01:29, 1.19s/it]
43%|████▎ | 56/130 [01:05<01:27, 1.18s/it]
44%|████▍ | 57/130 [01:06<01:25, 1.17s/it]
45%|████▍ | 58/130 [01:07<01:24, 1.17s/it]
45%|████▌ | 59/130 [01:08<01:22, 1.17s/it]
46%|████▌ | 60/130 [01:09<01:21, 1.17s/it]
{'loss': 1.5979, 'grad_norm': 2.016225814819336, 'learning_rate': 1.3293737524320798e-05, 'epoch': 0.46}
46%|████▌ | 60/130 [01:09<01:21, 1.17s/it]
47%|████▋ | 61/130 [01:11<01:20, 1.16s/it]
48%|████▊ | 62/130 [01:12<01:19, 1.16s/it]
48%|████▊ | 63/130 [01:13<01:17, 1.16s/it]
49%|████▉ | 64/130 [01:14<01:16, 1.16s/it]
50%|█████ | 65/130 [01:15<01:15, 1.16s/it]
{'loss': 1.5492, 'grad_norm': 2.3714423179626465, 'learning_rate': 1.2000256937760446e-05, 'epoch': 0.5}
50%|█████ | 65/130 [01:15<01:15, 1.16s/it]
51%|█████ | 66/130 [01:16<01:14, 1.16s/it]
52%|█████▏ | 67/130 [01:17<01:13, 1.16s/it]
52%|█████▏ | 68/130 [01:19<01:14, 1.21s/it]
53%|█████▎ | 69/130 [01:20<01:12, 1.19s/it]
54%|█████▍ | 70/130 [01:21<01:10, 1.18s/it]
{'loss': 1.5275, 'grad_norm': 2.853393793106079, 'learning_rate': 1.0670776443910024e-05, 'epoch': 0.54}
54%|█████▍ | 70/130 [01:21<01:10, 1.18s/it]
55%|█████▍ | 71/130 [01:22<01:09, 1.18s/it]
55%|█████▌ | 72/130 [01:23<01:07, 1.17s/it]
56%|█████▌ | 73/130 [01:25<01:06, 1.17s/it]
57%|█████▋ | 74/130 [01:26<01:05, 1.17s/it]
58%|█████▊ | 75/130 [01:27<01:04, 1.16s/it]
{'loss': 1.4923, 'grad_norm': 1.7769665718078613, 'learning_rate': 9.329223556089976e-06, 'epoch': 0.58}
58%|█████▊ | 75/130 [01:27<01:04, 1.16s/it]
58%|█████▊ | 76/130 [01:28<01:02, 1.16s/it]
59%|█████▉ | 77/130 [01:29<01:01, 1.16s/it]
60%|██████ | 78/130 [01:30<01:00, 1.16s/it]
61%|██████ | 79/130 [01:32<00:59, 1.16s/it]
62%|██████▏ | 80/130 [01:33<00:58, 1.16s/it]
{'loss': 1.4681, 'grad_norm': 2.4375193119049072, 'learning_rate': 7.999743062239557e-06, 'epoch': 0.62}
62%|██████▏ | 80/130 [01:33<00:58, 1.16s/it]
62%|██████▏ | 81/130 [01:34<00:56, 1.16s/it]
63%|██████▎ | 82/130 [01:35<00:55, 1.16s/it]
64%|██████▍ | 83/130 [01:36<00:54, 1.16s/it]
65%|██████▍ | 84/130 [01:37<00:53, 1.16s/it]
65%|██████▌ | 85/130 [01:39<00:52, 1.16s/it]
{'loss': 1.4528, 'grad_norm': 2.583644151687622, 'learning_rate': 6.706262475679205e-06, 'epoch': 0.65}
65%|██████▌ | 85/130 [01:39<00:52, 1.16s/it]
66%|██████▌ | 86/130 [01:40<00:51, 1.16s/it]
67%|██████▋ | 87/130 [01:41<00:49, 1.16s/it]
68%|██████▊ | 88/130 [01:42<00:48, 1.16s/it]
68%|██████▊ | 89/130 [01:43<00:47, 1.16s/it]
69%|██████▉ | 90/130 [01:44<00:46, 1.16s/it]
{'loss': 1.4174, 'grad_norm': 2.249284267425537, 'learning_rate': 5.47206139630363e-06, 'epoch': 0.69}
69%|██████▉ | 90/130 [01:44<00:46, 1.16s/it]
70%|███████ | 91/130 [01:46<00:45, 1.16s/it]
71%|███████ | 92/130 [01:47<00:44, 1.16s/it]
72%|███████▏ | 93/130 [01:48<00:42, 1.16s/it]
72%|███████▏ | 94/130 [01:49<00:41, 1.16s/it]
73%|███████▎ | 95/130 [01:50<00:40, 1.16s/it]
{'loss': 1.4044, 'grad_norm': 2.0346739292144775, 'learning_rate': 4.319352532688444e-06, 'epoch': 0.73}
73%|███████▎ | 95/130 [01:50<00:40, 1.16s/it]
74%|███████▍ | 96/130 [01:51<00:40, 1.20s/it]
75%|███████▍ | 97/130 [01:53<00:39, 1.19s/it]
75%|███████▌ | 98/130 [01:54<00:37, 1.18s/it]
76%|███████▌ | 99/130 [01:55<00:36, 1.17s/it]
77%|███████▋ | 100/130 [01:56<00:35, 1.17s/it]
{'loss': 1.3924, 'grad_norm': 1.5525047779083252, 'learning_rate': 3.2688819257242963e-06, 'epoch': 0.77}
77%|███████▋ | 100/130 [01:56<00:35, 1.17s/it][INFO|trainer.py:4307] 2026-04-10 13:41:49,893 >>
***** Running Evaluation *****
[INFO|trainer.py:4309] 2026-04-10 13:41:49,893 >> Num examples = 895
[INFO|trainer.py:4312] 2026-04-10 13:41:49,893 >> Batch size = 16
0%| | 0/7 [00:00<?, ?it/s]
29%|██▊ | 2/7 [00:00<00:00, 6.67it/s]
43%|████▎ | 3/7 [00:00<00:00, 4.60it/s]
57%|█████▋ | 4/7 [00:00<00:00, 3.94it/s]
71%|███████▏ | 5/7 [00:01<00:00, 3.70it/s]
86%|████████▌ | 6/7 [00:01<00:00, 3.55it/s]
100%|██████████| 7/7 [00:01<00:00, 3.45it/s]
{'eval_loss': 1.3881758451461792, 'eval_runtime': 2.1626, 'eval_samples_per_second': 413.847, 'eval_steps_per_second': 3.237, 'epoch': 0.77}
77%|███████▋ | 100/130 [01:58<00:35, 1.17s/it]
100%|██████████| 7/7 [00:01<00:00, 3.45it/s]

78%|███████▊ | 101/130 [02:00<00:53, 1.86s/it]
78%|███████▊ | 102/130 [02:01<00:46, 1.65s/it]
79%|███████▉ | 103/130 [02:02<00:40, 1.50s/it]
80%|████████ | 104/130 [02:03<00:36, 1.40s/it]
81%|████████ | 105/130 [02:04<00:33, 1.33s/it]
{'loss': 1.38, 'grad_norm': 1.5717474222183228, 'learning_rate': 2.339555568810221e-06, 'epoch': 0.81}
81%|████████ | 105/130 [02:04<00:33, 1.33s/it]
82%|████████▏ | 106/130 [02:05<00:30, 1.28s/it]
82%|████████▏ | 107/130 [02:06<00:28, 1.24s/it]
83%|████████▎ | 108/130 [02:08<00:26, 1.22s/it]
84%|████████▍ | 109/130 [02:09<00:25, 1.20s/it]
85%|████████▍ | 110/130 [02:10<00:23, 1.19s/it]
{'loss': 1.375, 'grad_norm': 1.4260509014129639, 'learning_rate': 1.5480991445620541e-06, 'epoch': 0.85}
85%|████████▍ | 110/130 [02:10<00:23, 1.19s/it]
85%|████████▌ | 111/130 [02:11<00:22, 1.18s/it]
86%|████████▌ | 112/130 [02:12<00:21, 1.17s/it]
87%|████████▋ | 113/130 [02:13<00:19, 1.17s/it]
88%|████████▊ | 114/130 [02:15<00:18, 1.17s/it]
88%|████████▊ | 115/130 [02:16<00:17, 1.16s/it]
{'loss': 1.3491, 'grad_norm': 1.3539237976074219, 'learning_rate': 9.08757001972762e-07, 'epoch': 0.88}
88%|████████▊ | 115/130 [02:16<00:17, 1.16s/it]
89%|████████▉ | 116/130 [02:17<00:16, 1.16s/it]
90%|█████████ | 117/130 [02:18<00:15, 1.16s/it]
91%|█████████ | 118/130 [02:19<00:13, 1.16s/it]
92%|█████████▏| 119/130 [02:20<00:12, 1.16s/it]
92%|█████████▏| 120/130 [02:22<00:12, 1.20s/it]
{'loss': 1.3533, 'grad_norm': 1.4586122035980225, 'learning_rate': 4.3303579172574884e-07, 'epoch': 0.92}
92%|█████████▏| 120/130 [02:22<00:12, 1.20s/it]
93%|█████████▎| 121/130 [02:23<00:10, 1.19s/it]
94%|█████████▍| 122/130 [02:24<00:09, 1.18s/it]
95%|█████████▍| 123/130 [02:25<00:08, 1.17s/it]
95%|█████████▌| 124/130 [02:26<00:07, 1.17s/it]
96%|█████████▌| 125/130 [02:27<00:05, 1.17s/it]
{'loss': 1.3503, 'grad_norm': 1.3738539218902588, 'learning_rate': 1.2949737362087156e-07, 'epoch': 0.96}
96%|█████████▌| 125/130 [02:28<00:05, 1.17s/it]
97%|█████████▋| 126/130 [02:29<00:04, 1.16s/it]
98%|█████████▊| 127/130 [02:30<00:03, 1.16s/it]
98%|█████████▊| 128/130 [02:31<00:02, 1.16s/it]
99%|█████████▉| 129/130 [02:32<00:01, 1.16s/it]
100%|██████████| 130/130 [02:33<00:00, 1.16s/it]
{'loss': 1.3395, 'grad_norm': 1.3676426410675049, 'learning_rate': 3.6047232911462506e-09, 'epoch': 1.0}
100%|██████████| 130/130 [02:33<00:00, 1.16s/it][INFO|trainer.py:3984] 2026-04-10 13:42:42,665 >> Saving model checkpoint to /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/checkpoint-130
[INFO|configuration_utils.py:419] 2026-04-10 13:42:42,671 >> Configuration saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/checkpoint-130/config.json
[INFO|configuration_utils.py:911] 2026-04-10 13:42:42,675 >> Configuration saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/checkpoint-130/generation_config.json
[INFO|modeling_utils.py:3580] 2026-04-10 13:43:25,431 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 6 checkpoint shards. You can find where each parameters has been saved in the index located at /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/checkpoint-130/model.safetensors.index.json.
[INFO|tokenization_utils_base.py:2510] 2026-04-10 13:43:25,448 >> tokenizer config file saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/checkpoint-130/tokenizer_config.json
[INFO|tokenization_utils_base.py:2519] 2026-04-10 13:43:25,451 >> Special tokens file saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/checkpoint-130/special_tokens_map.json
[INFO|trainer.py:2681] 2026-04-10 13:46:32,794 >>
Training completed. Do not forget to share your model on huggingface.co/models =)
{'train_runtime': 404.0129, 'train_samples_per_second': 40.88, 'train_steps_per_second': 0.322, 'train_loss': 1.8291644793290358, 'epoch': 1.0}
100%|██████████| 130/130 [06:39<00:00, 1.16s/it]
100%|██████████| 130/130 [06:39<00:00, 3.07s/it]
***** train metrics *****
epoch = 1.0
total_flos = 44661265GF
train_loss = 1.8292
train_runtime = 0:06:44.01
train_samples = 43598
train_samples_per_second = 40.88
train_steps_per_second = 0.322
2026-04-10 13:46:32 - INFO - __main__ - *** Save model ***
[INFO|configuration_utils.py:419] 2026-04-10 13:46:51,065 >> Configuration saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/config.json
[INFO|configuration_utils.py:911] 2026-04-10 13:46:51,069 >> Configuration saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/generation_config.json
[INFO|modeling_utils.py:3580] 2026-04-10 13:47:37,204 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 7 checkpoint shards. You can find where each parameters has been saved in the index located at /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/model.safetensors.index.json.
[INFO|tokenization_utils_base.py:2510] 2026-04-10 13:47:37,213 >> tokenizer config file saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/tokenizer_config.json
[INFO|tokenization_utils_base.py:2519] 2026-04-10 13:47:37,217 >> Special tokens file saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/special_tokens_map.json
2026-04-10 13:47:37 - INFO - __main__ - Saved HF-compatible model artifacts to /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758
2026-04-10 13:47:37 - INFO - __main__ - Saved validated HF-compatible model artifacts to /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758
[INFO|modelcard.py:450] 2026-04-10 13:47:37,519 >> Dropping the following result as it does not have all the necessary fields:
{'dataset': {'name': 'Anthropic/hh-rlhf', 'type': 'Anthropic/hh-rlhf', 'config': 'default', 'split': 'train', 'args': 'default'}}
[INFO|configuration_utils.py:419] 2026-04-10 13:47:37,528 >> Configuration saved in /scratch/feng.yulu/dynamic-dpo-v4/outputs/llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758/config.json
2026-04-10 13:47:37 - INFO - __main__ - *** Evaluate ***
[INFO|trainer.py:4307] 2026-04-10 13:47:37,531 >>
***** Running Evaluation *****
[INFO|trainer.py:4309] 2026-04-10 13:47:37,531 >> Num examples = 895
[INFO|trainer.py:4312] 2026-04-10 13:47:37,531 >> Batch size = 16
0%| | 0/7 [00:00<?, ?it/s]
29%|██▊ | 2/7 [00:00<00:00, 6.84it/s]
43%|████▎ | 3/7 [00:00<00:00, 4.83it/s]
57%|█████▋ | 4/7 [00:00<00:00, 4.03it/s]
71%|███████▏ | 5/7 [00:01<00:00, 3.79it/s]
86%|████████▌ | 6/7 [00:01<00:00, 3.65it/s]
100%|██████████| 7/7 [00:01<00:00, 3.53it/s]
100%|██████████| 7/7 [00:01<00:00, 3.84it/s]
***** eval metrics *****
epoch = 1.0
eval_loss = 1.3573
eval_runtime = 0:00:02.10
eval_samples = 2339
eval_samples_per_second = 424.776
eval_steps_per_second = 3.322
2026-04-10 13:47:39 - INFO - __main__ - *** Training complete ***
wandb: - 0.014 MB of 0.014 MB uploaded
wandb: \ 0.014 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb: \ 0.037 MB of 0.037 MB uploaded
wandb: | 0.037 MB of 0.037 MB uploaded
wandb: / 0.037 MB of 0.037 MB uploaded
wandb: - 0.037 MB of 0.037 MB uploaded
wandb:
wandb: Run history:
wandb: eval/loss █▁
wandb: eval/runtime █▁
wandb: eval/samples_per_second ▁█
wandb: eval/steps_per_second ▁█
wandb: train/epoch ▁▁▁▂▂▂▃▃▃▃▄▄▄▄▅▅▅▆▆▆▆▆▇▇▇▇████
wandb: train/global_step ▁▁▁▂▂▂▃▃▃▃▄▄▄▄▅▅▅▆▆▆▆▆▇▇▇▇████
wandb: train/grad_norm █▃▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁
wandb: train/learning_rate ▁▃▆████▇▇▇▆▆▆▅▅▄▄▃▃▃▂▂▂▁▁▁▁
wandb: train/loss ████▇▆▄▄▃▃▂▂▂▂▂▂▂▁▁▁▁▁▁▁▁▁▁
wandb:
wandb: Run summary:
wandb: eval/loss 1.35726
wandb: eval/runtime 2.107
wandb: eval/samples_per_second 424.776
wandb: eval/steps_per_second 3.322
wandb: total_flos 4.795466914988032e+16
wandb: train/epoch 1.0
wandb: train/global_step 130
wandb: train/grad_norm 1.36764
wandb: train/learning_rate 0.0
wandb: train/loss 1.3395
wandb: train_loss 1.82916
wandb: train_runtime 404.0129
wandb: train_samples_per_second 40.88
wandb: train_steps_per_second 0.322
wandb:
wandb: 🚀 View run llama-3-8b-base-sft-hh-helpful-8xh200-20260410-133758 at: https://wandb.ai/can-not-fand-northeastern-university/huggingface/runs/cqdvywmp
wandb: ⭐️ View project at: https://wandb.ai/can-not-fand-northeastern-university/huggingface
wandb: Synced 6 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
wandb: Find logs at: /scratch/feng.yulu/dynamic-dpo-v4/wandb/wandb/run-20260410_133950-cqdvywmp/logs
wandb: WARNING The new W&B backend becomes opt-out in version 0.18.0; try it out with `wandb.require("core")`! See https://wandb.me/wandb-core for more information.
Reference in New Issue View Git Blame Copy Permalink