XPK Start: Sat Apr 18 19:36:50 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 19:37:14.820258: 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 19:37:15.002583 137743107589952 max_utils.py:273] Attempting to initialize the jax distributed system... I0418 19:37:24.044033 137743107589952 distributed.py:149] Starting JAX distributed service on [::]:8482 I0418 19:37:24.046510 137743107589952 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-false-rs-od8gg-slice-job-0-0.mt-15-ocdbt-false-rs-od8gg:8482 I0418 19:37:24.964478 137743107589952 max_utils.py:284] Jax distributed system initialized! I0418 19:37:31.054237 137743107589952 max_utils.py:800] System Information: Jax Version: 0.9.2 I0418 19:37:31.054337 137743107589952 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0418 19:37:31.054377 137743107589952 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 19:37:31.054410 137743107589952 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0418 19:37:31.767921 137743107589952 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0418 19:37:31.768209 137743107589952 checkpointing.py:677] Setting up checkpoint logger... I0418 19:37:31.768264 137743107589952 checkpointing.py:233] Creating checkpoint manager with ocdbt=False and zarr3=True I0418 19:37:31.768310 137743107589952 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0418 19:37:31.768637 137743107589952 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 0x7d46383143b0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0418 19:37:33.053604 137743107589952 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0418 19:37:33.053843 137743107589952 checkpoint_manager.py:708] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d323c21ab70>}, handler_registry=None I0418 19:37:33.054093 137743107589952 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d323c21ab70>` 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 19:37:33.054143 137743107589952 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d3204788170>` 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 19:37:33.054178 137743107589952 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 0x7d323c21ab70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d323c21ab70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d3204788170>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d3204788170>}). I0418 19:37:33.054489 137743107589952 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0418 19:37:33.054558 137743107589952 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d32043a7d80> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0418 19:37:35.918026 137621472794368 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776540896520773622, 'commit_timestamp_nsecs': 1776540910386215549, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/5/_CHECKPOINT_METADATA I0418 19:37:36.169757 137621472794368 checkpoint_manager.py:1871] Missing metrics for step 5 I0418 19:37:36.620144 137743107589952 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints I0418 19:37:38.284042 137743107589952 checkpoint_manager.py:929] [process=7][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_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d450ebd5880> I0418 19:37:38.284261 137743107589952 checkpointing.py:301] Checkpoint manager created! I0418 19:37:39.218278 137743107589952 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0418 19:37:39.218391 137743107589952 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 19:37:39.595433 137743107589952 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 19:37:39.595524 137743107589952 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 19:37:39.611475 137743107589952 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 19:37:39.611531 137743107589952 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 19:37:39.634850 137743107589952 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 19:37:39.634920 137743107589952 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 19:37:39.651107 137743107589952 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 19:37:39.651170 137743107589952 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 19:37:39.667158 137743107589952 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 19:37:39.667217 137743107589952 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 19:37:39.692257 137743107589952 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0418 19:37:39.692328 137743107589952 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 19:37:39.713213 137743107589952 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0418 19:37:39.713275 137743107589952 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0418 19:37:39.931152 137743107589952 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0418 19:37:39.931258 137743107589952 checkpointing.py:581] restoring from this run's directory step 9 I0418 19:37:40.617120 137743107589952 event_tracking.py:70] [process=7] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/9. I0418 19:37:40.869405 137743107589952 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/9. I0418 19:37:40.869834 137743107589952 event_tracking.py:125] [process=7] [sync] Finished blocking load in 0.25 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/9. I0418 19:37:45.253312 1489 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0418 19:37:53.760826 137743107589952 jax_array_handlers.py:843] [process=7] /jax/orbax/read/worker/io/requested throughput: 158.431 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 9.974536895751953 s) (per-host) I0418 19:37:53.767486 137743107589952 base_pytree_checkpoint_handler.py:130] [process=7] /jax/checkpoint/read/gbytes_per_sec: 146.761 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.783496379852295 s) (per-host) I0418 19:37:54.003854 137743107589952 event_tracking.py:138] [process=7] [sync] Finished load in 13.39 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/9 I0418 19:37:54.004008 137743107589952 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints', 'checkpointer_start_time': 1776541060.6169958, 'checkpointer_duration_secs': 13.386980533599854, 'checkpoint_manager_start_time': 1776541059.9314954, 'checkpoint_manager_duration_secs': 14.07248330116272} fsdp: 32 I0418 19:37:54.418152 137743107589952 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0418 19:37:54.418279 137743107589952 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 19:37:54.418332 137743107589952 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 19:37:54.418395 137743107589952 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 19:37:54.418447 137743107589952 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 19:37:54.418484 137743107589952 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 19:37:54.418536 137743107589952 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 19:37:54.418590 137743107589952 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 19:37:54.418630 137743107589952 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 19:37:54.418666 137743107589952 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 19:37:54.418713 137743107589952 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0418 19:37:54.418758 137743107589952 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0418 19:37:54.900488 137743107589952 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 19:37:54.900581 137743107589952 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0418 19:37:54.915858 137743107589952 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 19:37:54.915916 137743107589952 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0418 19:38:05.840214 137743107589952 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 19:38:05.841034 137743107589952 metric_logger.py:301] number parameters: 1.104 billion I0418 19:38:17.141427 137743107589952 checkpointing.py:772] Waiting for step 10 to finish before checkpoint... I0418 19:38:17.314374 137743107589952 checkpointing.py:776] Waited 0.17292523384094238 seconds for step 10 to finish before starting checkpointing. I0418 19:38:17.316653 137743107589952 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 19:38:17.318281 137743107589952 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 10 I0418 19:38:17.319791 137743107589952 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/10. I0418 19:38:18.057103 137743107589952 signaling_client.py:364] Using JaxDistributedSignalingClient I0418 19:38:18.058039 137743107589952 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 19:38:18.058111 137743107589952 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 19:38:18.334566 137743107589952 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.277551s I0418 19:38:18.334743 137743107589952 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.479 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2815382480621338 s) (per-host) I0418 19:38:18.334798 137743107589952 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.281604s (batch_requests_ready=0.002257s, total_serialization_initiated=0.279272s, others=0.000075s) I0418 19:38:18.334898 137743107589952 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285977s (all_items=0.000018s, per_item={'items': '0.00001812'}, temp_paths=0.285959) I0418 19:38:18.335632 137743107589952 event_tracking.py:125] [process=7] [async] Finished blocking save in 1.02 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/10. I0418 19:38:18.335983 137617097721600 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 19:58:18.335944 I0418 19:38:18.343255 137743107589952 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize I0418 19:38:18.343490 137616004101888 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0418 19:38:18.343615 137743107589952 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776541097.3166354, 'wait_for_prev_duration_secs': 6.175041198730469e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776541097.318321, 'checkpointer_blocking_duration_secs': 1.017775535583496, 'get_old_steps_start_time': 1776541098.3361194, 'get_old_steps_duration_secs': 3.266334533691406e-05, 'checkpoint_manager_blocking_start_time': 1776541097.3149037, 'checkpoint_manager_blocking_duration_secs': 1.0286765098571777} I0418 19:38:18.343721 137743107589952 checkpointing.py:408] Started an asynchronous checkpoint save for step 10 I0418 19:38:18.343772 137743107589952 max_utils.py:750] Memstats: After params initialized: I0418 19:38:18.343819 137743107589952 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0)) I0418 19:38:18.343852 137743107589952 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0)) I0418 19:38:18.343879 137743107589952 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0)) I0418 19:38:18.343903 137743107589952 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0)) I0418 19:38:18.666202 137743107589952 metric_logger.py:196] completed step: 10, seconds: 11.300, TFLOP/s/device: 1.202, Tokens/s/device: 181.235, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869 I0418 19:38:18.834760 137743107589952 metric_logger.py:196] completed step: 11, seconds: 1.523, TFLOP/s/device: 8.920, Tokens/s/device: 1344.569, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659 I0418 19:38:18.982622 137743107589952 metric_logger.py:196] completed step: 12, seconds: 0.021, TFLOP/s/device: 635.031, Tokens/s/device: 95718.826, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068 I0418 19:38:19.129988 137743107589952 metric_logger.py:196] completed step: 13, seconds: 0.153, TFLOP/s/device: 88.520, Tokens/s/device: 13342.715, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001 I0418 19:38:19.134671 137743107589952 checkpointing.py:772] Waiting for step 15 to finish before checkpoint... I0418 19:38:19.424801 137743107589952 checkpointing.py:776] Waited 0.29009294509887695 seconds for step 15 to finish before starting checkpointing. I0418 19:38:19.427585 137743107589952 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 19:38:24.825389 137616012494592 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/10/items/array_metadatas/process_7 I0418 19:38:29.272528 137617097721600 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 140.793 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.219304323196411 s) (per-host) I0418 19:38:29.272639 137617097721600 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.936033s. I0418 19:38:36.629271 137617097721600 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 18.292646s. I0418 19:38:36.629573 137616004101888 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 19:38:36.629697 137616004101888 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0418 19:38:36.629746 137616004101888 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts... I0418 19:38:36.631334 137616004101888 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts. I0418 19:38:36.631497 137743107589952 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10. W0418 19:38:36.631628 137743107589952 checkpoint_manager.py:1452] Waiting for previous save to complete took 17.204043 seconds. If this number is high, consider checkpointing less frequently. I0418 19:38:36.633856 137743107589952 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 15 I0418 19:38:36.636498 137743107589952 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/15. I0418 19:38:38.197127 137743107589952 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 19:38:38.197257 137743107589952 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 19:38:38.225873 137743107589952 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.452553s I0418 19:38:38.226042 137743107589952 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 3.377 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.45676493644714355 s) (per-host) I0418 19:38:38.226115 137743107589952 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.456845s (batch_requests_ready=0.002551s, total_serialization_initiated=0.454206s, others=0.000088s) I0418 19:38:38.226213 137743107589952 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.461829s (all_items=0.000012s, per_item={'items': '0.00001168'}, temp_paths=0.461818) I0418 19:38:38.226876 137743107589952 event_tracking.py:125] [process=7] [async] Finished blocking save in 1.59 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/15. I0418 19:38:38.227189 137615995709184 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 19:58:38.227150 I0418 19:38:38.235786 137743107589952 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize I0418 19:38:38.236092 137616004101888 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0418 19:38:38.236251 137743107589952 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776541099.427556, 'wait_for_prev_duration_secs': 17.20404314994812, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776541116.6338964, 'checkpointer_blocking_duration_secs': 1.5934321880340576, 'get_old_steps_start_time': 1776541118.2273488, 'get_old_steps_duration_secs': 3.075599670410156e-05, 'checkpoint_manager_blocking_start_time': 1776541099.4254541, 'checkpoint_manager_blocking_duration_secs': 18.810763359069824} I0418 19:38:38.236357 137743107589952 checkpointing.py:408] Started an asynchronous checkpoint save for step 15 I0418 19:38:38.237157 137743107589952 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.940, Tokens/s/device: 13858.157, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193 I0418 19:38:38.244107 137743107589952 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.237, Tokens/s/device: 13903.032, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661 I0418 19:38:43.819754 137616012494592 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/15/items/array_metadatas/process_7 I0418 19:38:49.112173 137615995709184 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 139.259 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.342856407165527 s) (per-host) I0418 19:38:49.112299 137615995709184 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.884999s. I0418 19:38:55.201166 137615995709184 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 16.973849s. I0418 19:38:55.201430 137616004101888 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 19:38:55.201541 137616004101888 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0418 19:38:55.201590 137616004101888 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts... I0418 19:38:55.204465 137616004101888 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts. I0418 19:38:59.390670 137743107589952 metric_logger.py:196] completed step: 16, seconds: 19.108, TFLOP/s/device: 0.711, Tokens/s/device: 107.182, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872 I0418 19:38:59.538125 137743107589952 metric_logger.py:196] completed step: 17, seconds: 21.001, TFLOP/s/device: 0.647, Tokens/s/device: 97.519, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975 I0418 19:38:59.685942 137743107589952 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.276, Tokens/s/device: 13456.598, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333 I0418 19:38:59.832902 137743107589952 checkpointing.py:772] Waiting for step 19 to finish before checkpoint... I0418 19:38:59.833494 137743107589952 checkpointing.py:776] Waited 0.0006139278411865234 seconds for step 19 to finish before starting checkpointing. I0418 19:38:59.835736 137743107589952 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 19:38:59.837515 137743107589952 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 19 I0418 19:38:59.838967 137743107589952 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/19. I0418 19:39:00.548761 137743107589952 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 19:39:00.548866 137743107589952 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 19:39:00.574342 137743107589952 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026742s I0418 19:39:00.574470 137743107589952 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 50.898 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030307292938232422 s) (per-host) I0418 19:39:00.574516 137743107589952 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030363s (batch_requests_ready=0.001884s, total_serialization_initiated=0.028418s, others=0.000061s) I0418 19:39:00.574592 137743107589952 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034536s (all_items=0.000011s, per_item={'items': '0.00001073'}, temp_paths=0.034525) I0418 19:39:00.575306 137743107589952 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.74 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/19. I0418 19:39:00.575582 137612245309184 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 19:59:00.575547 I0418 19:39:00.579744 137743107589952 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize I0418 19:39:00.580034 137616004101888 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0418 19:39:00.580212 137743107589952 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776541139.835706, 'wait_for_prev_duration_secs': 7.128715515136719e-05, 'time_between_consecutive_saves_sec': 4.631187677383423, 'checkpointer_blocking_start_time': 1776541139.8375542, 'checkpointer_blocking_duration_secs': 0.7381720542907715, 'get_old_steps_start_time': 1776541140.575744, 'get_old_steps_duration_secs': 2.7179718017578125e-05, 'checkpoint_manager_blocking_start_time': 1776541139.8337252, 'checkpoint_manager_blocking_duration_secs': 0.746452808380127} I0418 19:39:00.580326 137743107589952 checkpointing.py:408] Started an asynchronous checkpoint save for step 19 I0418 19:39:00.580372 137743107589952 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 19:39:06.341861 137616012494592 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_180002/linen_xpk_main_20260418_180002_15_ocdbt_false/checkpoints/19/items/array_metadatas/process_7 I0418 19:39:10.389174 137612245309184 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 160.447 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 9.844970226287842 s) (per-host) I0418 19:39:10.389283 137612245309184 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 9.813585s. I0418 19:39:15.597910 137612245309184 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 15.022176s. I0418 19:39:15.598228 137616004101888 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 19:39:15.598347 137616004101888 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0418 19:39:15.598390 137616004101888 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts... I0418 19:39:15.599891 137616004101888 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts. I0418 19:39:15.600068 137743107589952 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19. I0418 19:39:15.600231 137743107589952 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 19:39:15.601142 137743107589952 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.278, Tokens/s/device: 13909.169, 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 19:39:26 UTC 2026 EXIT_CODE=0