2025-04-25 00:40:58 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 5 distributed training: True, 16-bits training: False
2025-04-25 00:40:58 - INFO - __main__ - Model parameters ModelArguments(base_model_revision=None, model_name_or_path='google/paligemma2-3b-pt-224', model_revision='main', model_code_revision=None, torch_dtype='bfloat16', tokenizer_name_or_path=None, processor_name_or_path=None, trust_remote_code=False, attn_implementation='eager', use_peft=True, lora_r=16, lora_alpha=16, lora_dropout=0.05, lora_target_modules=['q_proj', 'k_proj', 'v_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj'], 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')
2025-04-25 00:40:58 - INFO - __main__ - Data parameters DataArguments(chat_template=None, dataset_mixer={'mateoguaman/tartandrive_every1_100pct_sub5': 1.0, 'mateoguaman/scand_every1_50pct_sub5': 1.0, 'mateoguaman/coda_every1_25pct_sub5': 1.0, 'mateoguaman/spot_every1_sub5': 1.0}, eval_dataset_mixer={'mateoguaman/tartandrive_every1_100pct_sub5': 1.0, 'mateoguaman/scand_every1_50pct_sub5': 1.0, 'mateoguaman/coda_every1_25pct_sub5': 1.0, 'mateoguaman/spot_every1_sub5': 1.0}, text_column='text', dataset_splits=['train', 'test'], train_splits=['train', 'train', 'train', 'train'], validation_splits=['validation', 'validation', 'validation', 'validation'], processing_params={}, dataset_configs=None, preprocessing_num_workers=12, truncation_side=None, auto_insert_empty_system_msg=True, auto_set_chat_template=False, cache_dataset_only=True, just_combine_data=False, output_dataset_name='mateoguaman/vlmn_tartandrive100_scand50_coda25_spot100_sub5', output_dataset_description="VLN Navigation dataset with 100% of tartandrive data, 50% of scand data, 25% of coda data, and 100% of in-domain spot data. Whenever daatsets aren't 100%, they are ranked by curvature and output of length 5.", hf_entity=None)
2025-04-25 00:40:58 - INFO - __main__ - Training/evaluation parameters SFTConfig(
_n_gpu=5,
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=None,
dataset_kwargs=None,
dataset_num_proc=None,
dataset_text_field=text,
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,
dispatch_batches=None,
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=0.25,
eval_strategy=IntervalStrategy.STEPS,
eval_use_gather_object=False,
evaluation_strategy=None,
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=False,
gradient_checkpointing_kwargs=None,
greater_is_better=False,
group_by_length=False,
half_precision_backend=auto,
hub_always_push=False,
hub_model_id=paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5,
hub_model_revision=main,
hub_private_repo=None,
hub_strategy=HubStrategy.EVERY_SAVE,
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=0.0001,
length_column_name=length,
load_best_model_at_end=True,
local_rank=0,
log_level=info,
log_level_replica=warning,
log_on_each_node=True,
logging_dir=data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5/runs/Apr25_00-40-51_g3090,
logging_first_step=True,
logging_nan_inf_filter=True,
logging_steps=20,
logging_strategy=IntervalStrategy.STEPS,
lr_scheduler_kwargs={},
lr_scheduler_type=SchedulerType.COSINE,
max_grad_norm=1.0,
max_seq_length=2048,
max_steps=-1,
metric_for_best_model=loss,
model_init_kwargs=None,
mp_parameters=,
neftune_noise_alpha=None,
no_cuda=False,
num_of_sequences=None,
num_train_epochs=1,
optim=OptimizerNames.ADAMW_TORCH,
optim_args=None,
optim_target_modules=None,
output_dir=data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5,
overwrite_output_dir=True,
packing=True,
past_index=-1,
per_device_eval_batch_size=8,
per_device_train_batch_size=8,
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=data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5,
save_on_each_node=False,
save_only_model=False,
save_safetensors=True,
save_steps=0.25,
save_strategy=SaveStrategy.STEPS,
save_total_limit=None,
seed=42,
skip_memory_metrics=True,
split_batches=None,
tf32=None,
torch_compile=False,
torch_compile_backend=None,
torch_compile_mode=None,
torch_empty_cache_steps=None,
torchdynamo=None,
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=1e-05,
)
2025-04-25 00:41:05 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 00:41:05 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 00:41:05 - INFO - datasets.builder - Found cached dataset tartandrive_every1_100pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6)
2025-04-25 00:41:05 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 00:41:12 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-43db4b7381c5a51f.arrow
2025-04-25 00:41:12 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-4ee463c5e044885b.arrow
2025-04-25 00:41:15 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 00:41:15 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 00:41:15 - INFO - datasets.builder - Found cached dataset scand_every1_50pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea)
2025-04-25 00:41:15 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 00:41:44 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-04a411d0fbb1c538.arrow
2025-04-25 00:41:44 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-23308ad257d6fb04.arrow
2025-04-25 00:41:46 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 00:41:46 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 00:41:46 - INFO - datasets.builder - Found cached dataset coda_every1_25pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3)
2025-04-25 00:41:46 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 00:41:50 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-7254b303fba2f136.arrow
2025-04-25 00:41:50 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-c7ced277dacbb68c.arrow
2025-04-25 00:41:52 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 00:41:52 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 00:41:52 - INFO - datasets.builder - Found cached dataset spot_every1_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8)
2025-04-25 00:41:52 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 00:41:53 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-a99fa30e1ff675e7.arrow
2025-04-25 00:41:53 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-59690bd57af08437.arrow
2025-04-25 07:36:47 - INFO - datasets.arrow_dataset - Number of files to upload is larger than 50. Splitting the push into multiple commits.
2025-04-25 07:36:48 - INFO - datasets.arrow_dataset - Commit #1 completed (still 3 to go).
2025-04-25 07:36:49 - INFO - datasets.arrow_dataset - Commit #2 completed (still 2 to go).
2025-04-25 07:36:50 - INFO - datasets.arrow_dataset - Commit #3 completed (still 1 to go).
2025-04-25 07:36:51 - INFO - datasets.arrow_dataset - Commit #4 completed.
2025-04-25 07:36:54 - INFO - datasets.builder - No config specified, defaulting to the single config: dataset-mix-cached/default
2025-04-25 07:36:54 - INFO - datasets.info - Loading Dataset Infos from /gscratch/weirdlab/mateogc/miniconda3/envs/vlmn/lib/python3.11/site-packages/datasets/packaged_modules/parquet
2025-04-25 07:36:54 - INFO - datasets.builder - Generating dataset dataset-mix-cached (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/30c748399cd97840162506ed9fe8ffd7595c9ae1)
2025-04-25 07:36:54 - INFO - datasets.builder - Downloading and preparing dataset dataset-mix-cached/default to /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/30c748399cd97840162506ed9fe8ffd7595c9ae1...
2025-04-25 08:09:18 - INFO - datasets.download.download_manager - Downloading took 32.0 min
2025-04-25 08:09:18 - INFO - datasets.download.download_manager - Checksum Computation took 0.0 min
2025-04-25 08:09:48 - INFO - datasets.builder - Generating train split
2025-04-25 08:12:45 - INFO - datasets.utils.info_utils - Unable to verify splits sizes.
2025-04-25 08:12:45 - INFO - datasets.builder - Dataset dataset-mix-cached downloaded and prepared to /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/30c748399cd97840162506ed9fe8ffd7595c9ae1. Subsequent calls will reuse this data.
2025-04-25 08:14:02 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 08:14:02 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 08:14:02 - INFO - datasets.builder - Found cached dataset tartandrive_every1_100pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6)
2025-04-25 08:14:02 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 08:14:02 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-bb30a5067ff7c026.arrow
2025-04-25 08:14:02 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-a98ab9d069bce5ba.arrow
2025-04-25 08:14:04 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 08:14:04 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 08:14:04 - INFO - datasets.builder - Found cached dataset scand_every1_50pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea)
2025-04-25 08:14:04 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 08:14:04 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-2db1ac0f5a365a5f.arrow
2025-04-25 08:14:04 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-c6ff109c78f60201.arrow
2025-04-25 08:14:05 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 08:14:05 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 08:14:05 - INFO - datasets.builder - Found cached dataset coda_every1_25pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3)
2025-04-25 08:14:05 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 08:14:05 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-2d8d652dec47f509.arrow
2025-04-25 08:14:05 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-22ba4c412b8e9062.arrow
2025-04-25 08:14:07 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 08:14:07 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 08:14:08 - INFO - datasets.builder - Found cached dataset spot_every1_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8)
2025-04-25 08:14:08 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 08:14:08 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-a2d6cb7e6ded3626.arrow
2025-04-25 08:14:08 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-s3fwsz9e/cache-8a4806cf28ca3c53.arrow
2025-04-25 08:46:08 - INFO - datasets.arrow_dataset - Number of files to upload is larger than 50. Splitting the push into multiple commits.
2025-04-25 08:46:09 - INFO - datasets.arrow_dataset - Commit #1 completed (still 1 to go).
2025-04-25 08:46:09 - INFO - datasets.arrow_dataset - Commit #2 completed.
2025-04-25 08:46:13 - INFO - datasets.builder - No config specified, defaulting to the single config: dataset-mix-cached/default
2025-04-25 08:46:13 - INFO - datasets.info - Loading Dataset Infos from /gscratch/weirdlab/mateogc/miniconda3/envs/vlmn/lib/python3.11/site-packages/datasets/packaged_modules/parquet
2025-04-25 08:46:13 - INFO - datasets.builder - Generating dataset dataset-mix-cached (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/51b4cf1a929fcd3852276055bb81e53708b51caa)
2025-04-25 08:46:13 - INFO - datasets.builder - Downloading and preparing dataset dataset-mix-cached/default to /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/51b4cf1a929fcd3852276055bb81e53708b51caa...
2025-04-25 08:54:06 - INFO - datasets.download.download_manager - Downloading took 7.0 min
2025-04-25 08:54:06 - INFO - datasets.download.download_manager - Checksum Computation took 0.0 min
2025-04-25 08:54:12 - INFO - datasets.builder - Generating train split
2025-04-25 08:54:53 - INFO - datasets.utils.info_utils - Unable to verify splits sizes.
2025-04-25 08:54:53 - INFO - datasets.builder - Dataset dataset-mix-cached downloaded and prepared to /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/51b4cf1a929fcd3852276055bb81e53708b51caa. Subsequent calls will reuse this data.
2025-04-25 08:55:13 - INFO - __main__ - Consolidating dataset into standalone dataset at: mateoguaman/vlmn_tartandrive100_scand50_coda25_spot100_sub5
2025-04-25 08:55:14 - INFO - datasets.dataset_dict - Pushing split train to the Hub.
2025-04-25 10:25:54 - INFO - datasets.dataset_dict - Pushing split validation to the Hub.
2025-04-25 10:45:49 - INFO - datasets.dataset_dict - Number of files to upload is larger than 50. Splitting the push into multiple commits.
2025-04-25 10:45:51 - INFO - datasets.dataset_dict - Commit #1 completed (still 4 to go).
2025-04-25 10:45:53 - INFO - datasets.dataset_dict - Commit #2 completed (still 3 to go).
2025-04-25 10:45:54 - INFO - datasets.dataset_dict - Commit #3 completed (still 2 to go).
2025-04-25 10:45:56 - INFO - datasets.dataset_dict - Commit #4 completed (still 1 to go).
2025-04-25 10:45:58 - INFO - datasets.dataset_dict - Commit #5 completed.
2025-04-25 10:46:01 - INFO - datasets.builder - No config specified, defaulting to the single config: vlmn_tartandrive100_scand50_coda25_spot100_sub5/default
2025-04-25 10:46:01 - INFO - datasets.info - Loading Dataset Infos from /gscratch/weirdlab/mateogc/miniconda3/envs/vlmn/lib/python3.11/site-packages/datasets/packaged_modules/parquet
2025-04-25 10:46:01 - INFO - datasets.builder - Generating dataset vlmn_tartandrive100_scand50_coda25_spot100_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___vlmn_tartandrive100_scand50_coda25_spot100_sub5/default/0.0.0/6a8c723298f7298236874a7b73690dcb3dcb2cd0)
2025-04-25 10:46:01 - INFO - datasets.builder - Downloading and preparing dataset vlmn_tartandrive100_scand50_coda25_spot100_sub5/default to /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___vlmn_tartandrive100_scand50_coda25_spot100_sub5/default/0.0.0/6a8c723298f7298236874a7b73690dcb3dcb2cd0...
2025-04-25 11:26:36 - INFO - datasets.download.download_manager - Downloading took 40.0 min
2025-04-25 11:26:36 - INFO - datasets.download.download_manager - Checksum Computation took 0.0 min
2025-04-25 11:27:01 - INFO - datasets.builder - Generating train split
2025-04-25 11:29:16 - INFO - datasets.builder - Generating validation split
2025-04-25 11:29:51 - INFO - datasets.utils.info_utils - Unable to verify splits sizes.
2025-04-25 11:29:51 - INFO - datasets.builder - Dataset vlmn_tartandrive100_scand50_coda25_spot100_sub5 downloaded and prepared to /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___vlmn_tartandrive100_scand50_coda25_spot100_sub5/default/0.0.0/6a8c723298f7298236874a7b73690dcb3dcb2cd0. Subsequent calls will reuse this data.
2025-04-25 11:30:44 - INFO - __main__ - *** Dataset cached! Exiting now. ***
2025-04-25 12:11:19 - WARNING - __main__ - Process rank: 6, device: cuda:6, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:19 - WARNING - __main__ - Process rank: 3, device: cuda:3, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:19 - WARNING - __main__ - Process rank: 2, device: cuda:2, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:19 - WARNING - __main__ - Process rank: 5, device: cuda:5, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:19 - WARNING - __main__ - Process rank: 4, device: cuda:4, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:19 - WARNING - __main__ - Process rank: 1, device: cuda:1, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:24 - WARNING - __main__ - Process rank: 0, device: cuda:0, n_gpu: 1 distributed training: True, 16-bits training: False
2025-04-25 12:11:24 - INFO - __main__ - Model parameters ModelArguments(base_model_revision=None, model_name_or_path='google/paligemma2-3b-pt-224', model_revision='main', model_code_revision=None, torch_dtype='bfloat16', tokenizer_name_or_path=None, processor_name_or_path=None, trust_remote_code=False, attn_implementation='eager', use_peft=True, lora_r=16, lora_alpha=16, lora_dropout=0.05, lora_target_modules=['q_proj', 'k_proj', 'v_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj'], 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')
2025-04-25 12:11:24 - INFO - __main__ - Data parameters DataArguments(chat_template=None, dataset_mixer={'mateoguaman/tartandrive_every1_100pct_sub5': 1.0, 'mateoguaman/scand_every1_50pct_sub5': 1.0, 'mateoguaman/coda_every1_25pct_sub5': 1.0, 'mateoguaman/spot_every1_sub5': 1.0}, eval_dataset_mixer={'mateoguaman/tartandrive_every1_100pct_sub5': 1.0, 'mateoguaman/scand_every1_50pct_sub5': 1.0, 'mateoguaman/coda_every1_25pct_sub5': 1.0, 'mateoguaman/spot_every1_sub5': 1.0}, text_column='text', dataset_splits=['train', 'test'], train_splits=['train', 'train', 'train', 'train'], validation_splits=['validation', 'validation', 'validation', 'validation'], processing_params={}, dataset_configs=None, preprocessing_num_workers=12, truncation_side=None, auto_insert_empty_system_msg=True, auto_set_chat_template=False, cache_dataset_only=False, just_combine_data=False, output_dataset_name=None, output_dataset_description=None, hf_entity=None)
2025-04-25 12:11:24 - 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=None,
dataset_kwargs=None,
dataset_num_proc=None,
dataset_text_field=text,
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,
dispatch_batches=None,
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=0.25,
eval_strategy=IntervalStrategy.STEPS,
eval_use_gather_object=False,
evaluation_strategy=None,
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=False,
gradient_checkpointing_kwargs=None,
greater_is_better=False,
group_by_length=False,
half_precision_backend=auto,
hub_always_push=False,
hub_model_id=paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5,
hub_model_revision=main,
hub_private_repo=None,
hub_strategy=HubStrategy.EVERY_SAVE,
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=0.0001,
length_column_name=length,
load_best_model_at_end=True,
local_rank=0,
log_level=info,
log_level_replica=warning,
log_on_each_node=True,
logging_dir=data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5/runs/Apr25_12-11-18_g3090,
logging_first_step=True,
logging_nan_inf_filter=True,
logging_steps=20,
logging_strategy=IntervalStrategy.STEPS,
lr_scheduler_kwargs={},
lr_scheduler_type=SchedulerType.COSINE,
max_grad_norm=1.0,
max_seq_length=2048,
max_steps=-1,
metric_for_best_model=loss,
model_init_kwargs=None,
mp_parameters=,
neftune_noise_alpha=None,
no_cuda=False,
num_of_sequences=None,
num_train_epochs=1,
optim=OptimizerNames.ADAMW_TORCH,
optim_args=None,
optim_target_modules=None,
output_dir=data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5,
overwrite_output_dir=True,
packing=True,
past_index=-1,
per_device_eval_batch_size=8,
per_device_train_batch_size=8,
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=data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5,
save_on_each_node=False,
save_only_model=False,
save_safetensors=True,
save_steps=0.25,
save_strategy=SaveStrategy.STEPS,
save_total_limit=None,
seed=42,
skip_memory_metrics=True,
split_batches=None,
tf32=None,
torch_compile=False,
torch_compile_backend=None,
torch_compile_mode=None,
torch_empty_cache_steps=None,
torchdynamo=None,
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=1e-05,
)
2025-04-25 12:11:31 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:11:31 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 12:11:31 - INFO - datasets.builder - Found cached dataset tartandrive_every1_100pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6)
2025-04-25 12:11:31 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 12:11:31 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-1bd7669bf66c3790.arrow
2025-04-25 12:11:31 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-e84039a5b15ae04f.arrow
2025-04-25 12:11:33 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:11:33 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 12:11:33 - INFO - datasets.builder - Found cached dataset scand_every1_50pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea)
2025-04-25 12:11:33 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 12:11:42 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-aa4685d858ca5e56.arrow
2025-04-25 12:11:42 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-e0724c0c5f348260.arrow
2025-04-25 12:11:44 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:11:44 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 12:11:44 - INFO - datasets.builder - Found cached dataset coda_every1_25pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3)
2025-04-25 12:11:44 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 12:11:46 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-eea726b58f41f1b3.arrow
2025-04-25 12:11:46 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-6fb556f1e11b4e32.arrow
2025-04-25 12:11:48 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:11:48 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 12:11:48 - INFO - datasets.builder - Found cached dataset spot_every1_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8)
2025-04-25 12:11:48 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 12:11:49 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-7d563c8b2fcfedbe.arrow
2025-04-25 12:11:49 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-65ef6a060032ce44.arrow
2025-04-25 12:11:50 - INFO - datasets.builder - No config specified, defaulting to the single config: dataset-mix-cached/default
2025-04-25 12:11:50 - INFO - datasets.info - Loading Dataset Infos from /gscratch/weirdlab/mateogc/miniconda3/envs/vlmn/lib/python3.11/site-packages/datasets/packaged_modules/parquet
2025-04-25 12:11:50 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:11:50 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/30c748399cd97840162506ed9fe8ffd7595c9ae1
2025-04-25 12:11:50 - INFO - datasets.builder - Found cached dataset dataset-mix-cached (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/30c748399cd97840162506ed9fe8ffd7595c9ae1)
2025-04-25 12:11:50 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/30c748399cd97840162506ed9fe8ffd7595c9ae1
2025-04-25 12:12:01 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:12:01 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 12:12:01 - INFO - datasets.builder - Found cached dataset tartandrive_every1_100pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6)
2025-04-25 12:12:01 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___tartandrive_every1_100pct_sub5/default/0.0.0/ad39ab9c2f4ec53a3828b1459a3e33ca6ad5bda6
2025-04-25 12:12:02 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-5e6bb7b90a16ec8e.arrow
2025-04-25 12:12:02 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-c14b5d5953b7da7e.arrow
2025-04-25 12:12:03 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:12:03 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 12:12:03 - INFO - datasets.builder - Found cached dataset scand_every1_50pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea)
2025-04-25 12:12:03 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___scand_every1_50pct_sub5/default/0.0.0/de3a3ac0719a1970ac7f343c5e7d785efcc4b7ea
2025-04-25 12:12:05 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-2f560f79902d7de7.arrow
2025-04-25 12:12:05 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-ca1366b838254328.arrow
2025-04-25 12:12:07 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:12:07 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 12:12:07 - INFO - datasets.builder - Found cached dataset coda_every1_25pct_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3)
2025-04-25 12:12:07 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___coda_every1_25pct_sub5/default/0.0.0/1052e14cd42860e5cffc0a549067d0ed8b47bdf3
2025-04-25 12:12:07 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-34715c0dfff245c9.arrow
2025-04-25 12:12:07 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-8956e9a63073617c.arrow
2025-04-25 12:12:09 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:12:09 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 12:12:09 - INFO - datasets.builder - Found cached dataset spot_every1_sub5 (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8)
2025-04-25 12:12:09 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___spot_every1_sub5/default/0.0.0/3e6067c05e134722719c0c1f637e1f7451da81a8
2025-04-25 12:12:09 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-4863baa776c0a5d0.arrow
2025-04-25 12:12:09 - INFO - datasets.arrow_dataset - Caching indices mapping at /gscratch/robotics/mateogc/tmp/hf_datasets-waoiz9xn/cache-518a4275d94895ff.arrow
2025-04-25 12:12:11 - INFO - datasets.builder - No config specified, defaulting to the single config: dataset-mix-cached/default
2025-04-25 12:12:11 - INFO - datasets.info - Loading Dataset Infos from /gscratch/weirdlab/mateogc/miniconda3/envs/vlmn/lib/python3.11/site-packages/datasets/packaged_modules/parquet
2025-04-25 12:12:11 - INFO - datasets.builder - Overwrite dataset info from restored data version if exists.
2025-04-25 12:12:11 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/51b4cf1a929fcd3852276055bb81e53708b51caa
2025-04-25 12:12:11 - INFO - datasets.builder - Found cached dataset dataset-mix-cached (/gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/51b4cf1a929fcd3852276055bb81e53708b51caa)
2025-04-25 12:12:11 - INFO - datasets.info - Loading Dataset info from /gscratch/weirdlab/mateogc/huggingface/datasets/mateoguaman___dataset-mix-cached/default/0.0.0/51b4cf1a929fcd3852276055bb81e53708b51caa
2025-04-25 12:12:14 - INFO - __main__ - *** Load pretrained model ***
2025-04-25 12:12:27 - INFO - __main__ - Sample 335243 of the processed training set:

Navigate to x=<loc0430>, y=<loc0530>

<loc0481><loc0746><loc0463><loc0651><loc0448><loc0601><loc0437><loc0567><loc0430><loc0530>
2025-04-25 12:12:27 - INFO - __main__ - Sample 58369 of the processed training set:

Navigate to x=<loc1010>, y=<loc0515>

<loc0449><loc0615><loc0479><loc0532><loc0583><loc0485><loc0663><loc0478><loc1010><loc0515>
2025-04-25 12:12:27 - INFO - __main__ - Sample 13112 of the processed training set:

Navigate to x=<loc0576>, y=<loc0237>

<loc0532><loc0406><loc0568><loc0325><loc0584><loc0282><loc0581><loc0254><loc0576><loc0237>
2025-04-25 12:12:28 - INFO - __main__ - Loaded model is not a PeftModel. Getting PEFT config based on model_args.
2025-04-25 12:12:29 - WARNING - accelerate.utils.other - Detected kernel version 4.18.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.
2025-04-25 12:13:11 - INFO - __main__ - *** Train ***
2025-04-25 18:07:02 - INFO - __main__ - *** Save model ***
2025-04-25 18:07:04 - INFO - __main__ - Model saved to data/paligemma2-3b-pt-224-sft-lora-magicsoup_no_cfiphone_no_insta_sub5
2025-04-25 18:07:04 - INFO - __main__ - *** Evaluate ***
2025-04-25 18:23:31 - INFO - __main__ - *** Training complete ***