XPK Start: Sat Apr 18 14:41:10 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 14:41:34.329068: 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 14:41:34.510744 133876584777536 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 14:41:43.550493 133876584777536 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 14:41:43.552862 133876584777536 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-false-rs-wre3k-slice-job-0-0.mt-16-zarr3-false-rs-wre3k:8482
I0418 14:41:44.933263 133876584777536 max_utils.py:284] Jax distributed system initialized!
I0418 14:41:50.129238 133876584777536 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 14:41:50.129339 133876584777536 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 14:41:50.129379 133876584777536 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 14:41:50.129413 133876584777536 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 14:41:50.825346 133876584777536 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 14:41:50.825634 133876584777536 checkpointing.py:677] Setting up checkpoint logger...
I0418 14:41:50.825690 133876584777536 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=False
I0418 14:41:50.825748 133876584777536 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 14:41:50.826086 133876584777536 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=False, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x79c1dae0b5f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 14:41:52.513486 133876584777536 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 14:41:52.513720 133876584777536 checkpoint_manager.py:708] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x79adac14a330>}, handler_registry=None
I0418 14:41:52.513952 133876584777536 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x79adac14a330>` 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 14:41:52.514001 133876584777536 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79adac1ead20>` 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 14:41:52.514036 133876584777536 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 0x79adac14a330>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x79adac14a330>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79adac1ead20>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79adac1ead20>}).
I0418 14:41:52.514343 133876584777536 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 14:41:52.514411 133876584777536 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x79ad945dc180> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 14:41:55.374142 133754324506368 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776523143947309883, 'commit_timestamp_nsecs': 1776523192302928074, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/9/_CHECKPOINT_METADATA
I0418 14:41:55.629143 133754324506368 checkpoint_manager.py:1871] Missing metrics for step 9
I0418 14:41:56.940060 133876584777536 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints
I0418 14:41:57.385762 133876584777536 checkpoint_manager.py:929] [process=5][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=5), 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_16_zarr3_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x79c0cea39100>
I0418 14:41:57.385940 133876584777536 checkpointing.py:301] Checkpoint manager created!
I0418 14:41:58.309333 133876584777536 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 14:41:58.309479 133876584777536 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:41:58.687746 133876584777536 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 14:41:58.687832 133876584777536 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:41:58.703858 133876584777536 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 14:41:58.703914 133876584777536 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:41:58.727438 133876584777536 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 14:41:58.727516 133876584777536 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:41:58.743853 133876584777536 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 14:41:58.743913 133876584777536 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:41:58.760110 133876584777536 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 14:41:58.760169 133876584777536 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:41:58.785843 133876584777536 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 14:41:58.785912 133876584777536 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:41:58.806881 133876584777536 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 14:41:58.806951 133876584777536 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 14:41:59.024287 133876584777536 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 14:41:59.024394 133876584777536 checkpointing.py:581] restoring from this run's directory step 9
I0418 14:41:59.702665 133876584777536 event_tracking.py:70] [process=5] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/9.
I0418 14:41:59.951934 133876584777536 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/9.
I0418 14:41:59.952359 133876584777536 event_tracking.py:125] [process=5] [sync] Finished blocking load in 0.25 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/9.
I0418 14:42:02.902323 1466 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 14:42:08.259126 133876584777536 jax_array_handlers.py:843] [process=5] /jax/orbax/read/worker/io/requested throughput: 294.677 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 5.362722635269165 s) (per-host)
I0418 14:42:08.264848 133876584777536 base_pytree_checkpoint_handler.py:130] [process=5] /jax/checkpoint/read/gbytes_per_sec: 274.132 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 5.773115158081055 s) (per-host)
I0418 14:42:10.207712 133876584777536 event_tracking.py:138] [process=5] [sync] Finished load in 10.51 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/9
I0418 14:42:10.207865 133876584777536 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints', 'checkpointer_start_time': 1776523319.7025657, 'checkpointer_duration_secs': 10.50526738166809, 'checkpoint_manager_start_time': 1776523319.0246542, 'checkpoint_manager_duration_secs': 11.183181047439575}
fsdp: 32
I0418 14:42:10.619040 133876584777536 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 14:42:10.619164 133876584777536 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 14:42:10.619220 133876584777536 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 14:42:10.619300 133876584777536 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 14:42:10.619410 133876584777536 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 14:42:10.619474 133876584777536 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 14:42:10.619554 133876584777536 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 14:42:10.619611 133876584777536 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 14:42:10.619651 133876584777536 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 14:42:10.619688 133876584777536 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 14:42:10.619737 133876584777536 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 14:42:10.619786 133876584777536 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 14:42:11.100728 133876584777536 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 14:42:11.100820 133876584777536 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 14:42:11.115942 133876584777536 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 14:42:11.116000 133876584777536 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 14:42:22.084959 133876584777536 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0418 14:42:22.085814 133876584777536 metric_logger.py:301] number parameters: 1.104 billion
I0418 14:42:33.374661 133876584777536 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 14:42:33.567801 133876584777536 checkpointing.py:776] Waited 0.19311738014221191 seconds for step 10 to finish before starting checkpointing.
I0418 14:42:33.570169 133876584777536 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:42:33.571931 133876584777536 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 10
I0418 14:42:33.573531 133876584777536 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/10.
I0418 14:42:34.736017 133876584777536 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 14:42:34.737071 133876584777536 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:42:34.737125 133876584777536 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 14:42:35.013018 133876584777536 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.277076s
I0418 14:42:35.013189 133876584777536 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.492 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2808835506439209 s) (per-host)
I0418 14:42:35.013243 133876584777536 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.280948s (batch_requests_ready=0.002108s, total_serialization_initiated=0.278767s, others=0.000073s)
I0418 14:42:35.013331 133876584777536 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.284980s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.284964)
I0418 14:42:35.014107 133876584777536 event_tracking.py:125] [process=5] [async] Finished blocking save in 1.44 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/10.
I0418 14:42:35.014391 133754290935552 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 15:02:35.014360
I0418 14:42:35.020275 133876584777536 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:42:35.020519 133749950981888 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:42:35.020634 133876584777536 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776523353.570151, 'wait_for_prev_duration_secs': 6.127357482910156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776523353.571969, 'checkpointer_blocking_duration_secs': 1.4425277709960938, 'get_old_steps_start_time': 1776523355.014518, 'get_old_steps_duration_secs': 3.218650817871094e-05, 'checkpoint_manager_blocking_start_time': 1776523353.568301, 'checkpoint_manager_blocking_duration_secs': 1.4522991180419922}
I0418 14:42:35.020743 133876584777536 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 14:42:35.020794 133876584777536 max_utils.py:750]
Memstats: After params initialized:
I0418 14:42:35.020842 133876584777536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0))
I0418 14:42:35.020874 133876584777536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0))
I0418 14:42:35.020900 133876584777536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0))
I0418 14:42:35.020924 133876584777536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0))
I0418 14:42:35.330110 133876584777536 metric_logger.py:196] completed step: 10, seconds: 11.289, TFLOP/s/device: 1.204, Tokens/s/device: 181.420, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0418 14:42:35.510833 133876584777536 metric_logger.py:196] completed step: 11, seconds: 1.954, TFLOP/s/device: 6.954, Tokens/s/device: 1048.160, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0418 14:42:35.658583 133876584777536 metric_logger.py:196] completed step: 12, seconds: 0.033, TFLOP/s/device: 405.646, Tokens/s/device: 61143.454, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0418 14:42:35.806002 133876584777536 metric_logger.py:196] completed step: 13, seconds: 0.153, TFLOP/s/device: 88.661, Tokens/s/device: 13363.959, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0418 14:42:35.810631 133876584777536 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0418 14:42:36.100497 133876584777536 checkpointing.py:776] Waited 0.289808988571167 seconds for step 15 to finish before starting checkpointing.
I0418 14:42:36.103526 133876584777536 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:42:41.203713 133749963937536 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/10/items/array_metadatas/process_5
I0418 14:43:17.599278 133754290935552 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 36.849 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.86693549156189 s) (per-host)
I0418 14:43:17.599406 133754290935552 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.584385s.
I0418 14:43:24.896806 133754290935552 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.881768s.
I0418 14:43:24.897033 133749950981888 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:43:24.897100 133749950981888 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:43:24.897147 133749950981888 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:43:24.898746 133749950981888 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 14:43:24.898912 133876584777536 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0418 14:43:24.899054 133876584777536 checkpoint_manager.py:1452] Waiting for previous save to complete took 48.795529 seconds. If this number is high, consider checkpointing less frequently.
I0418 14:43:24.901031 133876584777536 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 15
I0418 14:43:24.903534 133876584777536 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/15.
I0418 14:43:25.620599 133876584777536 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:43:25.620694 133876584777536 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 14:43:25.662258 133876584777536 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.042853s
I0418 14:43:25.662426 133876584777536 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 33.363 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04623603820800781 s) (per-host)
I0418 14:43:25.662490 133876584777536 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.046308s (batch_requests_ready=0.001734s, total_serialization_initiated=0.044493s, others=0.000081s)
I0418 14:43:25.662580 133876584777536 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.051161s (all_items=0.000013s, per_item={'items': '0.00001264'}, temp_paths=0.051148)
I0418 14:43:25.663304 133876584777536 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/15.
I0418 14:43:25.663652 133749950981888 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 15:03:25.663613
I0418 14:43:25.665757 133876584777536 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:43:25.666010 133748347987712 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:43:25.666177 133876584777536 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776523356.1034954, 'wait_for_prev_duration_secs': 48.79552912712097, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776523404.90107, 'checkpointer_blocking_duration_secs': 0.7627365589141846, 'get_old_steps_start_time': 1776523405.6638308, 'get_old_steps_duration_secs': 3.24249267578125e-05, 'checkpoint_manager_blocking_start_time': 1776523356.1010842, 'checkpoint_manager_blocking_duration_secs': 49.56505823135376}
I0418 14:43:25.666294 133876584777536 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0418 14:43:25.667114 133876584777536 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.956, Tokens/s/device: 13860.595, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0418 14:43:25.673256 133876584777536 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.144, Tokens/s/device: 13888.889, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0418 14:43:32.000099 133749963937536 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/15/items/array_metadatas/process_5
I0418 14:43:46.955354 133876584777536 metric_logger.py:196] completed step: 16, seconds: 49.861, TFLOP/s/device: 0.272, Tokens/s/device: 41.074, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0418 14:43:47.102669 133876584777536 metric_logger.py:196] completed step: 17, seconds: 21.136, TFLOP/s/device: 0.643, Tokens/s/device: 96.897, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0418 14:43:47.250104 133876584777536 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.121, Tokens/s/device: 13433.296, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0418 14:43:47.396738 133876584777536 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 14:43:47.397418 133876584777536 checkpointing.py:776] Waited 0.0007014274597167969 seconds for step 19 to finish before starting checkpointing.
I0418 14:43:47.399569 133876584777536 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:44:07.557881 133749950981888 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 37.662 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.941662549972534 s) (per-host)
I0418 14:44:07.557987 133749950981888 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.894217s.
I0418 14:44:14.892489 133749950981888 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.228681s.
I0418 14:44:14.892793 133748347987712 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:44:14.892920 133748347987712 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:44:14.892970 133748347987712 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:44:14.895339 133748347987712 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0418 14:44:14.895534 133876584777536 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0418 14:44:14.895687 133876584777536 checkpoint_manager.py:1452] Waiting for previous save to complete took 27.496110 seconds. If this number is high, consider checkpointing less frequently.
I0418 14:44:14.897012 133876584777536 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 19
I0418 14:44:14.898955 133876584777536 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/19.
I0418 14:44:15.616409 133876584777536 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:44:15.616517 133876584777536 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 14:44:15.659767 133876584777536 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044433s
I0418 14:44:15.659927 133876584777536 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 32.209 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04789304733276367 s) (per-host)
I0418 14:44:15.659977 133876584777536 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.047952s (batch_requests_ready=0.001641s, total_serialization_initiated=0.046246s, others=0.000065s)
I0418 14:44:15.660072 133876584777536 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.052094s (all_items=0.000010s, per_item={'items': '0.00001049'}, temp_paths=0.052083)
I0418 14:44:15.660789 133876584777536 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/19.
I0418 14:44:15.661123 133748347987712 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 15:04:15.661086
I0418 14:44:15.663174 133876584777536 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:44:15.663415 133748339595008 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:44:15.663581 133876584777536 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776523427.3995392, 'wait_for_prev_duration_secs': 27.496110439300537, 'time_between_consecutive_saves_sec': 22.50075340270996, 'checkpointer_blocking_start_time': 1776523454.897052, 'checkpointer_blocking_duration_secs': 0.7642199993133545, 'get_old_steps_start_time': 1776523455.6612895, 'get_old_steps_duration_secs': 2.7179718017578125e-05, 'checkpoint_manager_blocking_start_time': 1776523427.3976593, 'checkpoint_manager_blocking_duration_secs': 28.26589059829712}
I0418 14:44:15.663695 133876584777536 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 14:44:15.663737 133876584777536 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:44:21.675761 133749963937536 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_16_zarr3_false/checkpoints/19/items/array_metadatas/process_5
I0418 14:44:57.814540 133748347987712 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 37.429 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.202465534210205 s) (per-host)
I0418 14:44:57.814662 133748347987712 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.153427s.
I0418 14:45:05.876521 133748347987712 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 50.215271s.
I0418 14:45:05.876786 133748339595008 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:45:05.876906 133748339595008 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:45:05.876959 133748339595008 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:45:05.879446 133748339595008 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 14:45:05.879634 133876584777536 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0418 14:45:05.879791 133876584777536 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:45:05.880766 133876584777536 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.169, Tokens/s/device: 13892.752, total_weights: 65536, loss: 5.391, lm_loss: 5.391, perplexity: 219.349
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 14:45:15 UTC 2026
EXIT_CODE=0