XPK Start: Fri Apr 17 14:15:05 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-17 14:15:29.304407: 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) I0417 14:15:29.485735 132645232813888 max_utils.py:273] Attempting to initialize the jax distributed system... I0417 14:15:38.525413 132645232813888 distributed.py:149] Starting JAX distributed service on [::]:8482 I0417 14:15:38.527742 132645232813888 distributed.py:172] Connecting to JAX distributed service on mt-13-scan-layers-true-v7c7q-slice-job-0-0.mt-13-scan-layers-true-v7c7q:8482 I0417 14:15:39.637755 132645232813888 max_utils.py:284] Jax distributed system initialized! I0417 14:15:45.805742 132645232813888 max_utils.py:800] System Information: Jax Version: 0.9.2 I0417 14:15:45.805865 132645232813888 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0417 14:15:45.805911 132645232813888 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 I0417 14:15:45.805946 132645232813888 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0417 14:15:46.485635 132645232813888 maxtext_utils.py:1548] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0417 14:15:46.485932 132645232813888 checkpointing.py:677] Setting up checkpoint logger... I0417 14:15:46.485986 132645232813888 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0417 14:15:46.486031 132645232813888 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0417 14:15:46.486369 132645232813888 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x78a3513ea6c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0417 14:15:49.392220 132645232813888 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0417 14:15:49.392529 132645232813888 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x788f00361fd0>}, handler_registry=None I0417 14:15:49.392769 132645232813888 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x788f00361fd0>` 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`. I0417 14:15:49.392816 132645232813888 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x788f005835f0>` 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`. I0417 14:15:49.392873 132645232813888 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 0x788f00361fd0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x788f00361fd0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x788f005835f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x788f005835f0>}). I0417 14:15:49.393198 132645232813888 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0417 14:15:49.393272 132645232813888 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x788df4694a40> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0417 14:15:50.582885 132645232813888 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints I0417 14:15:51.053732 132645232813888 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x788f00366db0> I0417 14:15:51.053947 132645232813888 checkpointing.py:301] Checkpoint manager created! I0417 14:15:51.954574 132645232813888 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0417 14:15:51.954686 132645232813888 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 14:15:52.339468 132645232813888 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 14:15:52.339558 132645232813888 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 14:15:52.355870 132645232813888 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 14:15:52.355930 132645232813888 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 14:15:52.379652 132645232813888 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 14:15:52.379722 132645232813888 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 14:15:52.396228 132645232813888 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 14:15:52.396289 132645232813888 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 14:15:52.412462 132645232813888 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 14:15:52.412518 132645232813888 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 14:15:52.437311 132645232813888 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0417 14:15:52.437382 132645232813888 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 14:15:52.459112 132645232813888 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0417 14:15:52.459180 132645232813888 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0417 14:15:52.676894 132645232813888 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0417 14:15:52.677010 132645232813888 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0417 14:15:54.113198 132645232813888 maxtext_utils.py:1651] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0417 14:15:54.113318 132645232813888 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0417 14:15:54.113369 132645232813888 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0417 14:15:54.113427 132645232813888 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0417 14:15:54.113480 132645232813888 maxtext_utils.py:1651] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0417 14:15:54.113528 132645232813888 maxtext_utils.py:1651] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0417 14:15:54.113580 132645232813888 maxtext_utils.py:1651] 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) I0417 14:15:54.113630 132645232813888 maxtext_utils.py:1651] 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') I0417 14:15:54.113668 132645232813888 maxtext_utils.py:1651] 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) I0417 14:15:54.113704 132645232813888 maxtext_utils.py:1651] 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) I0417 14:15:54.113750 132645232813888 maxtext_utils.py:1651] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0417 14:15:54.113798 132645232813888 maxtext_utils.py:1651] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0417 14:15:54.600064 132645232813888 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0417 14:15:54.600159 132645232813888 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0417 14:15:54.615318 132645232813888 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0417 14:15:54.615381 132645232813888 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0417 14:16:05.378782 132645232813888 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. I0417 14:16:05.386437 132645232813888 metric_logger.py:301] number parameters: 1.104 billion I0417 14:16:16.876953 132645232813888 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0417 14:16:17.030660 132645232813888 checkpointing.py:776] Waited 0.15368413925170898 seconds for step 0 to finish before starting checkpointing. I0417 14:16:17.033260 132645232813888 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 14:16:17.035003 132645232813888 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0 I0417 14:16:17.036465 132645232813888 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0. I0417 14:16:17.793684 132645232813888 signaling_client.py:364] Using JaxDistributedSignalingClient I0417 14:16:17.807795 132645232813888 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 14:16:17.807869 132645232813888 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 I0417 14:16:18.079581 132516308576000 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0 I0417 14:16:18.081608 132645232813888 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.274592s I0417 14:16:18.082379 132645232813888 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.490 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28096842765808105 s) (per-host) I0417 14:16:18.082450 132645232813888 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.281059s (batch_requests_ready=0.002181s, total_serialization_initiated=0.278188s, others=0.000690s) I0417 14:16:18.082530 132645232813888 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.284928s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.284911) I0417 14:16:18.083221 132645232813888 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0. I0417 14:16:18.083521 132514112202496 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 14:36:18.083481 I0417 14:16:18.095063 132645232813888 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0417 14:16:18.095307 132516316968704 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 14:16:18.095431 132645232813888 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776435377.033238, 'wait_for_prev_duration_secs': 7.009506225585938e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776435377.0350428, 'checkpointer_blocking_duration_secs': 1.0486249923706055, 'get_old_steps_start_time': 1776435378.0836923, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776435377.0312252, 'checkpoint_manager_blocking_duration_secs': 1.0641677379608154} I0417 14:16:18.095537 132645232813888 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0417 14:16:18.095583 132645232813888 max_utils.py:750] Memstats: After params initialized: I0417 14:16:18.095632 132645232813888 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0417 14:16:18.095665 132645232813888 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0417 14:16:18.095690 132645232813888 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0417 14:16:18.095714 132645232813888 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0417 14:16:18.424637 132645232813888 metric_logger.py:196] completed step: 0, seconds: 11.187, TFLOP/s/device: 1.215, Tokens/s/device: 183.071, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0417 14:16:18.427991 132645232813888 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/tensorboard/' I0417 14:16:18.858147 132645232813888 metric_logger.py:196] completed step: 1, seconds: 1.546, TFLOP/s/device: 8.790, Tokens/s/device: 1324.906, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0417 14:16:19.004363 132645232813888 metric_logger.py:196] completed step: 2, seconds: 0.434, TFLOP/s/device: 31.284, Tokens/s/device: 4715.471, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0417 14:16:19.152565 132645232813888 metric_logger.py:196] completed step: 3, seconds: 0.008, TFLOP/s/device: 1610.040, Tokens/s/device: 242682.782, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203 I0417 14:16:25.809691 132516283397888 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/items I0417 14:16:25.955307 132512516093696 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776435385544783022, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776435385544783022, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0417 14:16:27.023321 2852 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0417 14:16:29.640936 132514166118144 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/items/array_metadatas/process_0 I0417 14:16:41.465044 132645232813888 metric_logger.py:196] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.215, Tokens/s/device: 13899.635, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614 I0417 14:16:41.473702 132645232813888 metric_logger.py:196] completed step: 5, seconds: 0.149, TFLOP/s/device: 91.180, Tokens/s/device: 13743.675, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717 I0417 14:16:41.619831 132645232813888 metric_logger.py:196] completed step: 6, seconds: 22.311, TFLOP/s/device: 0.609, Tokens/s/device: 91.794, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759 I0417 14:16:41.767056 132645232813888 metric_logger.py:196] completed step: 7, seconds: 0.007, TFLOP/s/device: 2029.141, Tokens/s/device: 305854.241, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905 I0417 14:16:41.914421 132645232813888 metric_logger.py:196] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.298, Tokens/s/device: 13912.193, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406 I0417 14:16:42.061008 132645232813888 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0417 14:16:42.061676 132645232813888 checkpointing.py:776] Waited 0.0006897449493408203 seconds for step 9 to finish before starting checkpointing. I0417 14:16:42.063700 132645232813888 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 14:17:03.793776 132514120595200 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 36.769760s (commit=35.476512s, array_metadata_write=1.293247s) I0417 14:17:03.794978 132514112202496 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.344 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.993539333343506 s) (per-host) I0417 14:17:03.795111 132514112202496 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.711475s. I0417 14:17:04.991076 132514112202496 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776435385544783022, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0417 14:17:05.747971 132512516093696 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776435385544783022, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0417 14:17:06.354656 132514112202496 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 14:17:06.422584 132514112202496 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000249s. I0417 14:17:07.282753 132514112202496 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.164690s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/items I0417 14:17:07.284350 132514112202496 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0/items I0417 14:17:07.849370 132514112202496 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0 I0417 14:17:09.418440 132514112202496 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0`. I0417 14:17:09.419302 132514112202496 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 52.38 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/0 I0417 14:17:09.420751 132514112202496 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 51.337113s. I0417 14:17:09.420953 132516316968704 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 14:17:09.421068 132516316968704 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 14:17:09.421128 132516316968704 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0417 14:17:09.423087 132516316968704 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0417 14:17:09.423250 132645232813888 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0417 14:17:09.423335 132645232813888 checkpoint_manager.py:1452] Waiting for previous save to complete took 27.359641 seconds. If this number is high, consider checkpointing less frequently. I0417 14:17:09.425136 132645232813888 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9 I0417 14:17:09.427394 132645232813888 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9. I0417 14:17:10.164181 132645232813888 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 14:17:10.164271 132645232813888 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 I0417 14:17:10.190774 132645232813888 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027366s I0417 14:17:10.191615 132645232813888 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 48.376 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.031887054443359375 s) (per-host) I0417 14:17:10.191685 132645232813888 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031974s (batch_requests_ready=0.001804s, total_serialization_initiated=0.029380s, others=0.000790s) I0417 14:17:10.191752 132645232813888 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035832s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.035815) I0417 14:17:10.192487 132645232813888 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.77 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9. I0417 14:17:10.192783 132515645880064 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 14:37:10.192749 I0417 14:17:10.206065 132645232813888 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0417 14:17:10.206327 132516316968704 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 14:17:10.206470 132645232813888 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776435402.0636659, 'wait_for_prev_duration_secs': 27.359641313552856, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776435429.4251764, 'checkpointer_blocking_duration_secs': 0.7677559852600098, 'get_old_steps_start_time': 1776435430.192955, 'get_old_steps_duration_secs': 3.1948089599609375e-05, 'checkpoint_manager_blocking_start_time': 1776435402.061926, 'checkpoint_manager_blocking_duration_secs': 28.1445095539093} I0417 14:17:10.206582 132645232813888 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0417 14:17:10.206626 132645232813888 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 14:17:10.429806 132514112202496 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9 I0417 14:17:12.717411 132510365239040 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9/items I0417 14:17:16.624566 132515771705088 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9/items/array_metadatas/process_0 I0417 14:17:54.951630 132515763312384 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.053461s (commit=39.395632s, array_metadata_write=1.657829s) I0417 14:17:54.959853 132515645880064 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 35.259 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.80007290840149 s) (per-host) I0417 14:17:54.959975 132515645880064 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.767066s. I0417 14:17:56.975039 132515645880064 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000247s. I0417 14:17:56.987710 132515645880064 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 14:17:57.626727 132515645880064 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.718925s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9/items I0417 14:17:57.628234 132515645880064 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9/items I0417 14:17:58.187630 132515645880064 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9 I0417 14:18:00.136029 132515645880064 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9`. I0417 14:18:00.136880 132515645880064 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 50.71 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_13_scan_layers_true/checkpoints/9 I0417 14:18:00.138678 132515645880064 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 49.945757s. I0417 14:18:00.138990 132516316968704 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 14:18:00.139107 132516316968704 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 14:18:00.139169 132516316968704 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0417 14:18:00.141695 132516316968704 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0417 14:18:00.141814 132645232813888 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0417 14:18:00.141931 132645232813888 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 14:18:00.142941 132645232813888 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.198, Tokens/s/device: 13897.088, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Fri Apr 17 14:18:11 UTC 2026 EXIT_CODE=0