XPK Start: Sat Apr 18 13:08:41 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-18 13:09:05.625800: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)
I0418 13:09:06.155379 138286100473664 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 13:09:15.197744 138286100473664 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 13:09:15.200301 138286100473664 distributed.py:172] Connecting to JAX distributed service on mt-03-dropout-pfnui-slice-job-0-0.mt-03-dropout-pfnui:8482
I0418 13:09:17.567279 138286100473664 max_utils.py:284] Jax distributed system initialized!
I0418 13:09:22.791757 138286100473664 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 13:09:22.791856 138286100473664 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 13:09:22.791894 138286100473664 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0418 13:09:22.791928 138286100473664 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 13:09:23.493480 138286100473664 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 13:09:23.493774 138286100473664 checkpointing.py:677] Setting up checkpoint logger...
I0418 13:09:23.493832 138286100473664 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 13:09:23.493878 138286100473664 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 13:09:23.494201 138286100473664 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7dc4b4dbe6c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 13:09:26.533048 138286100473664 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 13:09:26.533341 138286100473664 checkpoint_manager.py:708] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dc3783dcd70>}, handler_registry=None
I0418 13:09:26.533579 138286100473664 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dc3783dcd70>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`.
I0418 13:09:26.533640 138286100473664 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7db10041ac60>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`.
I0418 13:09:26.533677 138286100473664 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dc3783dcd70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dc3783dcd70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7db10041ac60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7db10041ac60>}).
I0418 13:09:26.534004 138286100473664 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 13:09:26.534087 138286100473664 async_checkpointer.py:192] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7db0d87afce0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 13:09:30.380715 138286100473664 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints
I0418 13:09:30.383111 138286100473664 checkpoint_manager.py:929] [process=3][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7db10041ac00>
I0418 13:09:30.383223 138286100473664 checkpointing.py:301] Checkpoint manager created!
I0418 13:09:30.727637 138286100473664 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 13:09:30.727751 138286100473664 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0418 13:09:31.104781 138286100473664 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:09:31.104874 138286100473664 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0418 13:09:31.121697 138286100473664 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:09:31.121763 138286100473664 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0418 13:09:31.145923 138286100473664 attentions.py:1154] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:09:31.145987 138286100473664 attentions.py:1154] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0418 13:09:31.162917 138286100473664 attentions.py:1155] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:09:31.162976 138286100473664 attentions.py:1155] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0418 13:09:31.179822 138286100473664 attentions.py:1156] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 13:09:31.179880 138286100473664 attentions.py:1156] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0418 13:09:31.206131 138286100473664 attentions.py:1197] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 13:09:31.206202 138286100473664 attentions.py:1197] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0418 13:09:31.232805 138286100473664 linears.py:525] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 13:09:31.232868 138286100473664 linears.py:525] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0418 13:09:31.463197 138286100473664 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 13:09:31.463305 138286100473664 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0418 13:09:32.897771 138286100473664 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 13:09:32.897896 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 13:09:32.897949 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 13:09:32.898023 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0418 13:09:32.898082 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 13:09:32.898122 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 13:09:32.898173 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 13:09:32.898224 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0418 13:09:32.898270 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0418 13:09:32.898307 138286100473664 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 13:09:32.898354 138286100473664 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 13:09:32.898404 138286100473664 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 13:09:33.428801 138286100473664 train.py:155] train/xent Logical: float32[32,128]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:09:33.428894 138286100473664 train.py:155] train/xent Physical: float32[32,128]............................................. ('fsdp', None).
I0418 13:09:33.444557 138286100473664 train.py:162] train/z_loss Logical: float32[32,128]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:09:33.444633 138286100473664 train.py:162] train/z_loss Physical: float32[32,128]............................................. ('fsdp', None).
I0418 13:09:37.006290 138286100473664 max_utils.py:791] Total memory size: 0.8 GB, Output size: 0.4 GB, Temp size: 0.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0418 13:09:37.007181 138286100473664 metric_logger.py:301] number parameters: 1.104 billion
I0418 13:09:41.188527 138286100473664 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0418 13:09:41.508722 138286100473664 checkpointing.py:776] Waited 0.320171594619751 seconds for step 0 to finish before starting checkpointing.
I0418 13:09:41.511160 138286100473664 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:09:41.513130 138286100473664 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 0
I0418 13:09:41.514601 138286100473664 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints/0.
I0418 13:09:42.250519 138286100473664 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 13:09:42.251531 138286100473664 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 13:09:42.251587 138286100473664 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 13:09:42.536417 138286100473664 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.285967s
I0418 13:09:42.536600 138286100473664 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 5.296 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2912907600402832 s) (per-host)
I0418 13:09:42.536669 138286100473664 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.291369s (batch_requests_ready=0.002169s, total_serialization_initiated=0.289114s, others=0.000087s)
I0418 13:09:42.536761 138286100473664 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.295625s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.295608)
I0418 13:09:42.537513 138286100473664 event_tracking.py:125] [process=3] [async] Finished blocking save in 1.02 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints/0.
I0418 13:09:42.537871 138161265207040 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:29:42.537834
I0418 13:09:42.540603 138286100473664 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:09:42.540920 138160201783040 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:09:42.541090 138286100473664 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776517781.5111418, 'wait_for_prev_duration_secs': 6.246566772460938e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776517781.5131712, 'checkpointer_blocking_duration_secs': 1.0248656272888184, 'get_old_steps_start_time': 1776517782.538066, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776517781.509318, 'checkpoint_manager_blocking_duration_secs': 1.0317330360412598}
I0418 13:09:42.541202 138286100473664 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0418 13:09:42.541252 138286100473664 max_utils.py:750]
Memstats: After params initialized:
I0418 13:09:42.541301 138286100473664 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_10(process=3,(2,2,0,0))
I0418 13:09:42.541333 138286100473664 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_11(process=3,(3,2,0,0))
I0418 13:09:42.541361 138286100473664 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_14(process=3,(2,3,0,0))
I0418 13:09:42.541387 138286100473664 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_15(process=3,(3,3,0,0))
I0418 13:09:42.856224 138286100473664 metric_logger.py:196] completed step: 0, seconds: 4.181, TFLOP/s/device: 0.192, Tokens/s/device: 30.613, total_weights: 4096, loss: 10.889, lm_loss: 10.889, perplexity: 53560.699
I0418 13:09:42.955830 138286100473664 metric_logger.py:196] completed step: 1, seconds: 1.666, TFLOP/s/device: 0.481, Tokens/s/device: 76.829, total_weights: 4096, loss: 10.902, lm_loss: 10.902, perplexity: 54298.523
I0418 13:09:43.362672 138286100473664 metric_logger.py:196] completed step: 2, seconds: 0.030, TFLOP/s/device: 27.099, Tokens/s/device: 4331.055, total_weights: 4096, loss: 9.956, lm_loss: 9.956, perplexity: 21082.051
I0418 13:09:43.435035 138286100473664 metric_logger.py:196] completed step: 3, seconds: 0.407, TFLOP/s/device: 1.967, Tokens/s/device: 314.319, total_weights: 4096, loss: 9.139, lm_loss: 9.139, perplexity: 9314.691
I0418 13:09:43.577292 138286100473664 metric_logger.py:196] completed step: 4, seconds: 0.076, TFLOP/s/device: 10.533, Tokens/s/device: 1683.369, total_weights: 4096, loss: 8.459, lm_loss: 8.459, perplexity: 4715.493
I0418 13:09:43.583598 138286100473664 metric_logger.py:196] completed step: 5, seconds: 0.072, TFLOP/s/device: 11.114, Tokens/s/device: 1776.248, total_weights: 4096, loss: 7.924, lm_loss: 7.924, perplexity: 2762.373
I0418 13:09:47.069780 2724 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 13:09:50.834223 138160210175744 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints/0/items/array_metadatas/process_3
I0418 13:09:54.211390 138286100473664 metric_logger.py:196] completed step: 6, seconds: 0.143, TFLOP/s/device: 5.596, Tokens/s/device: 894.379, total_weights: 4096, loss: 7.553, lm_loss: 7.553, perplexity: 1906.617
I0418 13:09:54.284013 138286100473664 metric_logger.py:196] completed step: 7, seconds: 10.559, TFLOP/s/device: 0.076, Tokens/s/device: 12.123, total_weights: 4096, loss: 7.267, lm_loss: 7.267, perplexity: 1431.653
I0418 13:09:54.354851 138286100473664 metric_logger.py:196] completed step: 8, seconds: 0.075, TFLOP/s/device: 10.696, Tokens/s/device: 1709.470, total_weights: 4096, loss: 7.111, lm_loss: 7.111, perplexity: 1225.983
I0418 13:09:54.424958 138286100473664 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0418 13:09:54.425592 138286100473664 checkpointing.py:776] Waited 0.0006542205810546875 seconds for step 9 to finish before starting checkpointing.
I0418 13:09:54.427672 138286100473664 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:10:16.133441 138161265207040 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 46.612 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 33.888092279434204 s) (per-host)
I0418 13:10:16.133566 138161265207040 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 33.595574s.
I0418 13:10:27.845384 138161265207040 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 45.307375s.
I0418 13:10:27.845700 138160201783040 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:10:27.845822 138160201783040 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:10:27.845872 138160201783040 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:10:27.847560 138160201783040 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0418 13:10:27.847766 138286100473664 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0418 13:10:27.847893 138286100473664 checkpoint_manager.py:1452] Waiting for previous save to complete took 33.420221 seconds. If this number is high, consider checkpointing less frequently.
I0418 13:10:27.849759 138286100473664 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 9
I0418 13:10:27.851785 138286100473664 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints/9.
I0418 13:10:28.574667 138286100473664 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 13:10:28.574764 138286100473664 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 13:10:28.611145 138286100473664 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037576s
I0418 13:10:28.611316 138286100473664 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 37.603 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04102301597595215 s) (per-host)
I0418 13:10:28.611367 138286100473664 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.041083s (batch_requests_ready=0.001782s, total_serialization_initiated=0.039234s, others=0.000067s)
I0418 13:10:28.611460 138286100473664 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.045196s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.045180)
I0418 13:10:28.612212 138286100473664 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints/9.
I0418 13:10:28.612575 138160201783040 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:30:28.612538
I0418 13:10:28.614585 138286100473664 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:10:28.614832 138160210175744 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:10:28.614987 138286100473664 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776517794.4276423, 'wait_for_prev_duration_secs': 33.42022132873535, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776517827.8497984, 'checkpointer_blocking_duration_secs': 0.7629468441009521, 'get_old_steps_start_time': 1776517828.6127703, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776517794.4258301, 'checkpoint_manager_blocking_duration_secs': 34.18912482261658}
I0418 13:10:28.615095 138286100473664 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0418 13:10:28.615140 138286100473664 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:10:36.250261 138134289303296 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_03_dropout/checkpoints/9/items/array_metadatas/process_3
I0418 13:11:09.907140 138160201783040 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 38.213 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.336814165115356 s) (per-host)
I0418 13:11:09.907252 138160201783040 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.294537s.
I0418 13:11:19.493574 138160201783040 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 50.880844s.
I0418 13:11:19.493898 138160210175744 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:11:19.494023 138160210175744 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:11:19.494072 138160210175744 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:11:19.496608 138160210175744 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0418 13:11:19.496805 138286100473664 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0418 13:11:19.496955 138286100473664 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:11:19.497905 138286100473664 metric_logger.py:196] completed step: 9, seconds: 0.073, TFLOP/s/device: 11.030, Tokens/s/device: 1762.794, total_weights: 4096, loss: 6.981, lm_loss: 6.981, perplexity: 1075.867
Per train step:
Total TFLOPs: 0.80
split as 99.60% learnable weight flops and 0.40% attention flops
XPK End: Sat Apr 18 13:11:31 UTC 2026
EXIT_CODE=0