XPK Start: Sat Apr 18 14:22:15 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 14:22:39.344317: 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 14:22:39.884484 136428769281856 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 14:22:48.926912 136428769281856 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 14:22:48.929277 136428769281856 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-false-rs-m066k-slice-job-0-0.mt-15-ocdbt-false-rs-m066k:8482
I0418 14:22:50.063704 136428769281856 max_utils.py:284] Jax distributed system initialized!
I0418 14:22:56.153695 136428769281856 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 14:22:56.153800 136428769281856 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 14:22:56.153840 136428769281856 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 14:22:56.153876 136428769281856 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 14:22:56.849994 136428769281856 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 14:22:56.850275 136428769281856 checkpointing.py:677] Setting up checkpoint logger...
I0418 14:22:56.850327 136428769281856 checkpointing.py:233] Creating checkpoint manager with ocdbt=False and zarr3=True
I0418 14:22:56.850371 136428769281856 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 14:22:56.850721 136428769281856 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=False, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7c1414b4c170>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 14:22:58.587434 136428769281856 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 14:22:58.587693 136428769281856 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 0x7bffe808b800>}, handler_registry=None
I0418 14:22:58.587929 136428769281856 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bffe808b800>` 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 14:22:58.587984 136428769281856 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bffe808d460>` 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 14:22:58.588020 136428769281856 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 0x7bffe808b800>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bffe808b800>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bffe808d460>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bffe808d460>}).
I0418 14:22:58.588336 136428769281856 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 14:22:58.588405 136428769281856 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7bffd0597c40> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 14:23:00.544612 136307162478336 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776522001042770684, 'commit_timestamp_nsecs': 1776522015516503644, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/0/_CHECKPOINT_METADATA
I0418 14:23:01.654538 136307162478336 checkpoint_manager.py:1871] Missing metrics for step 0
I0418 14:23:02.540893 136428769281856 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints
I0418 14:23:02.543495 136428769281856 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_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7bffe808d5b0>
I0418 14:23:02.543609 136428769281856 checkpointing.py:301] Checkpoint manager created!
I0418 14:23:03.460021 136428769281856 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 14:23:03.460132 136428769281856 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:23:03.838777 136428769281856 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 14:23:03.838872 136428769281856 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:23:03.855041 136428769281856 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 14:23:03.855099 136428769281856 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:23:03.879339 136428769281856 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 14:23:03.879409 136428769281856 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:23:03.895630 136428769281856 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 14:23:03.895689 136428769281856 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:23:03.911815 136428769281856 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 14:23:03.911880 136428769281856 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:23:03.936381 136428769281856 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 14:23:03.936448 136428769281856 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:23:03.957793 136428769281856 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 14:23:03.957853 136428769281856 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 14:23:04.174057 136428769281856 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 14:23:04.174164 136428769281856 checkpointing.py:581] restoring from this run's directory step 9
I0418 14:23:04.428545 136428769281856 event_tracking.py:70] [process=5] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/9.
I0418 14:23:05.122411 136428769281856 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/9.
I0418 14:23:05.122854 136428769281856 event_tracking.py:125] [process=5] [sync] Finished blocking load in 0.69 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/9.
I0418 14:23:07.911484 1483 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 14:23:16.770020 136428769281856 jax_array_handlers.py:843] [process=5] /jax/orbax/read/worker/io/requested throughput: 157.323 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.04475736618042 s) (per-host)
I0418 14:23:16.777211 136428769281856 base_pytree_checkpoint_handler.py:130] [process=5] /jax/checkpoint/read/gbytes_per_sec: 151.684 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.433490991592407 s) (per-host)
I0418 14:23:19.027013 136428769281856 event_tracking.py:138] [process=5] [sync] Finished load in 14.60 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/9
I0418 14:23:19.027162 136428769281856 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints', 'checkpointer_start_time': 1776522184.4284377, 'checkpointer_duration_secs': 14.598692417144775, 'checkpoint_manager_start_time': 1776522184.174401, 'checkpoint_manager_duration_secs': 14.852731227874756}
fsdp: 32
I0418 14:23:19.441620 136428769281856 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 14:23:19.441745 136428769281856 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 14:23:19.441799 136428769281856 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 14:23:19.441856 136428769281856 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 14:23:19.441905 136428769281856 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 14:23:19.441943 136428769281856 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 14:23:19.441999 136428769281856 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 14:23:19.442052 136428769281856 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 14:23:19.442092 136428769281856 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 14:23:19.442126 136428769281856 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 14:23:19.442171 136428769281856 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 14:23:19.442216 136428769281856 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 14:23:19.924651 136428769281856 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 14:23:19.924753 136428769281856 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 14:23:19.940039 136428769281856 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 14:23:19.940105 136428769281856 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 14:23:30.893193 136428769281856 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 14:23:30.894054 136428769281856 metric_logger.py:301] number parameters: 1.104 billion
I0418 14:23:42.105767 136428769281856 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 14:23:42.304242 136428769281856 checkpointing.py:776] Waited 0.19845080375671387 seconds for step 10 to finish before starting checkpointing.
I0418 14:23:42.306591 136428769281856 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:23:42.308429 136428769281856 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 10
I0418 14:23:42.309954 136428769281856 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/10.
I0418 14:23:43.032313 136428769281856 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 14:23:43.033348 136428769281856 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:23:43.033404 136428769281856 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 14:23:43.310788 136428769281856 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.278550s
I0418 14:23:43.310960 136428769281856 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.462 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2824099063873291 s) (per-host)
I0418 14:23:43.311012 136428769281856 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.282473s (batch_requests_ready=0.002160s, total_serialization_initiated=0.280241s, others=0.000072s)
I0418 14:23:43.311101 136428769281856 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.288988s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.288970)
I0418 14:23:43.311860 136428769281856 event_tracking.py:125] [process=5] [async] Finished blocking save in 1.00 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/10.
I0418 14:23:43.312230 136304427767552 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:43:43.312185
I0418 14:23:43.319193 136428769281856 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:23:43.319424 136302844614400 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:23:43.319565 136428769281856 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776522222.3065732, 'wait_for_prev_duration_secs': 5.936622619628906e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776522222.3084807, 'checkpointer_blocking_duration_secs': 1.0039036273956299, 'get_old_steps_start_time': 1776522223.312408, 'get_old_steps_duration_secs': 3.266334533691406e-05, 'checkpoint_manager_blocking_start_time': 1776522222.304765, 'checkpoint_manager_blocking_duration_secs': 1.0147650241851807}
I0418 14:23:43.319666 136428769281856 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 14:23:43.319714 136428769281856 max_utils.py:750]
Memstats: After params initialized:
I0418 14:23:43.319762 136428769281856 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0))
I0418 14:23:43.319801 136428769281856 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0))
I0418 14:23:43.319826 136428769281856 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0))
I0418 14:23:43.319849 136428769281856 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0))
I0418 14:23:43.641611 136428769281856 metric_logger.py:196] completed step: 10, seconds: 11.212, TFLOP/s/device: 1.212, Tokens/s/device: 182.668, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0418 14:23:43.814622 136428769281856 metric_logger.py:196] completed step: 11, seconds: 1.534, TFLOP/s/device: 8.856, Tokens/s/device: 1334.805, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0418 14:23:43.962524 136428769281856 metric_logger.py:196] completed step: 12, seconds: 0.026, TFLOP/s/device: 532.703, Tokens/s/device: 80294.833, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0418 14:23:44.110038 136428769281856 metric_logger.py:196] completed step: 13, seconds: 0.154, TFLOP/s/device: 88.399, Tokens/s/device: 13324.485, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0418 14:23:44.114707 136428769281856 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0418 14:23:44.404389 136428769281856 checkpointing.py:776] Waited 0.2896449565887451 seconds for step 15 to finish before starting checkpointing.
I0418 14:23:44.407574 136428769281856 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:23:49.005408 136302853007104 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_15_ocdbt_false/checkpoints/10/items/array_metadatas/process_5
I0418 14:23:53.706236 136304427767552 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 147.935 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.677648067474365 s) (per-host)
I0418 14:23:53.706356 136304427767552 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.394011s.
I0418 14:23:59.715775 136304427767552 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 16.403412s.
I0418 14:23:59.716059 136302844614400 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:23:59.716173 136302844614400 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:23:59.716220 136302844614400 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:23:59.717796 136302844614400 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 14:23:59.717989 136428769281856 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 14:23:59.718120 136428769281856 checkpoint_manager.py:1452] Waiting for previous save to complete took 15.310545 seconds. If this number is high, consider checkpointing less frequently.
I0418 14:23:59.720098 136428769281856 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 15
I0418 14:23:59.722651 136428769281856 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/15.
I0418 14:24:00.087768 136428769281856 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:24:00.087862 136428769281856 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 14:24:00.115051 136428769281856 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028553s
I0418 14:24:00.115182 136428769281856 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 48.268 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.031958580017089844 s) (per-host)
I0418 14:24:00.115231 136428769281856 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.032016s (batch_requests_ready=0.001786s, total_serialization_initiated=0.030167s, others=0.000062s)
I0418 14:24:00.115309 136428769281856 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.037135s (all_items=0.000011s, per_item={'items': '0.00001144'}, temp_paths=0.037124)
I0418 14:24:00.115985 136428769281856 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.40 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/15.
I0418 14:24:00.116240 136302844614400 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:44:00.116216
I0418 14:24:00.543066 136428769281856 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:24:00.543437 136302330640128 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:24:00.543638 136428769281856 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776522224.4075463, 'wait_for_prev_duration_secs': 15.310545444488525, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776522239.7201366, 'checkpointer_blocking_duration_secs': 0.39620399475097656, 'get_old_steps_start_time': 1776522240.116364, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776522224.405055, 'checkpoint_manager_blocking_duration_secs': 16.138545751571655}
I0418 14:24:00.543756 136428769281856 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0418 14:24:00.544645 136428769281856 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.801, Tokens/s/device: 13837.277, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0418 14:24:00.551094 136428769281856 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.231, Tokens/s/device: 13902.088, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0418 14:24:05.981094 136302853007104 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_15_ocdbt_false/checkpoints/15/items/array_metadatas/process_5
I0418 14:24:11.325773 136302844614400 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 140.502 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.24249267578125 s) (per-host)
I0418 14:24:11.325909 136302844614400 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 11.209592s.
I0418 14:24:18.594438 136302844614400 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 18.478104s.
I0418 14:24:18.594753 136302330640128 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:24:18.594885 136302330640128 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:24:18.594933 136302330640128 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:24:18.596230 136302330640128 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0418 14:24:22.108819 136428769281856 metric_logger.py:196] completed step: 16, seconds: 16.435, TFLOP/s/device: 0.827, Tokens/s/device: 124.613, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0418 14:24:22.256397 136428769281856 metric_logger.py:196] completed step: 17, seconds: 21.412, TFLOP/s/device: 0.635, Tokens/s/device: 95.649, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0418 14:24:22.403692 136428769281856 metric_logger.py:196] completed step: 18, seconds: 0.153, TFLOP/s/device: 88.969, Tokens/s/device: 13410.338, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0418 14:24:22.550733 136428769281856 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 14:24:22.551422 136428769281856 checkpointing.py:776] Waited 0.0007114410400390625 seconds for step 19 to finish before starting checkpointing.
I0418 14:24:22.553682 136428769281856 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:24:22.555163 136428769281856 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 19
I0418 14:24:22.556599 136428769281856 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/19.
I0418 14:24:23.253847 136428769281856 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:24:23.253940 136428769281856 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 14:24:23.278903 136428769281856 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.025944s
I0418 14:24:23.279080 136428769281856 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 52.479 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.02939438819885254 s) (per-host)
I0418 14:24:23.279132 136428769281856 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.029455s (batch_requests_ready=0.001644s, total_serialization_initiated=0.027746s, others=0.000066s)
I0418 14:24:23.279227 136428769281856 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.033556s (all_items=0.000010s, per_item={'items': '0.00001025'}, temp_paths=0.033546)
I0418 14:24:23.279992 136428769281856 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.72 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints/19.
I0418 14:24:23.280322 136302330640128 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:44:23.280286
I0418 14:24:23.285659 136428769281856 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:24:23.285924 136297993709312 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:24:23.286081 136428769281856 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776522262.5536518, 'wait_for_prev_duration_secs': 7.295608520507812e-05, 'time_between_consecutive_saves_sec': 3.9573824405670166, 'checkpointer_blocking_start_time': 1776522262.5552006, 'checkpointer_blocking_duration_secs': 0.7253286838531494, 'get_old_steps_start_time': 1776522263.2805467, 'get_old_steps_duration_secs': 2.7179718017578125e-05, 'checkpoint_manager_blocking_start_time': 1776522262.551672, 'checkpoint_manager_blocking_duration_secs': 0.734375}
I0418 14:24:23.286196 136428769281856 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 14:24:23.286241 136428769281856 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:24:29.371971 136302853007104 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_15_ocdbt_false/checkpoints/19/items/array_metadatas/process_5
I0418 14:24:33.364906 136302330640128 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 156.161 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.115179061889648 s) (per-host)
I0418 14:24:33.365034 136302330640128 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.084602s.
I0418 14:24:38.840619 136302330640128 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 15.560169s.
I0418 14:24:38.840885 136297993709312 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:24:38.841007 136297993709312 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:24:38.841049 136297993709312 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:24:38.842549 136297993709312 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 14:24:38.842725 136428769281856 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 14:24:38.842873 136428769281856 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:24:38.843824 136428769281856 metric_logger.py:196] completed step: 19, seconds: 0.148, TFLOP/s/device: 92.052, Tokens/s/device: 13875.151, 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 14:24:51 UTC 2026
EXIT_CODE=0