XPK Start: Sat Apr 18 01:08:42 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 01:09:05.919374: 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 01:09:06.453621 135124737042240 max_utils.py:273] Attempting to initialize the jax distributed system... I0418 01:09:15.493295 135124737042240 distributed.py:149] Starting JAX distributed service on [::]:8482 I0418 01:09:15.495674 135124737042240 distributed.py:172] Connecting to JAX distributed service on mt-13-scan-layers-true-f9vxh-slice-job-0-0.mt-13-scan-layers-true-f9vxh:8482 I0418 01:09:16.845709 135124737042240 max_utils.py:284] Jax distributed system initialized! I0418 01:09:23.073585 135124737042240 max_utils.py:800] System Information: Jax Version: 0.9.2 I0418 01:09:23.073686 135124737042240 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0418 01:09:23.073726 135124737042240 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 01:09:23.073760 135124737042240 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0418 01:09:23.759469 135124737042240 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0418 01:09:23.759744 135124737042240 checkpointing.py:677] Setting up checkpoint logger... I0418 01:09:23.759798 135124737042240 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0418 01:09:23.759864 135124737042240 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0418 01:09:23.760196 135124737042240 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 0x7ae47624c2f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0418 01:09:28.246623 135124737042240 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0418 01:09:28.246872 135124737042240 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 0x7ad00824df70>}, handler_registry=None I0418 01:09:28.247107 135124737042240 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ad00824df70>` 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 01:09:28.247155 135124737042240 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7acf2c7627b0>` 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 01:09:28.247190 135124737042240 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 0x7ad00824df70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ad00824df70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7acf2c7627b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7acf2c7627b0>}). I0418 01:09:28.247502 135124737042240 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0418 01:09:28.247570 135124737042240 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7acf0c247c40> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0418 01:09:28.934057 135124737042240 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints I0418 01:09:30.251691 135124737042240 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_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7acf2c757200> I0418 01:09:30.251877 135124737042240 checkpointing.py:301] Checkpoint manager created! I0418 01:09:31.167190 135124737042240 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0418 01:09:31.167300 135124737042240 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 01:09:31.543312 135124737042240 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 01:09:31.543402 135124737042240 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 01:09:31.559675 135124737042240 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 01:09:31.559735 135124737042240 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 01:09:31.583247 135124737042240 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 01:09:31.583315 135124737042240 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:09:31.599543 135124737042240 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 01:09:31.599600 135124737042240 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:09:31.615783 135124737042240 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 01:09:31.615853 135124737042240 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:09:31.641309 135124737042240 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0418 01:09:31.641383 135124737042240 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:09:31.662465 135124737042240 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0418 01:09:31.662533 135124737042240 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0418 01:09:31.881340 135124737042240 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0418 01:09:31.881445 135124737042240 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0418 01:09:33.306988 135124737042240 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0418 01:09:33.307117 135124737042240 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 01:09:33.307171 135124737042240 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 01:09:33.307228 135124737042240 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 01:09:33.307282 135124737042240 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 01:09:33.307321 135124737042240 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 01:09:33.307373 135124737042240 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 01:09:33.307424 135124737042240 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 01:09:33.307464 135124737042240 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 01:09:33.307501 135124737042240 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 01:09:33.307547 135124737042240 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0418 01:09:33.307593 135124737042240 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0418 01:09:33.798228 135124737042240 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 01:09:33.798321 135124737042240 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0418 01:09:33.813626 135124737042240 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 01:09:33.813687 135124737042240 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0418 01:09:44.902360 135124737042240 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 01:09:44.910444 135124737042240 metric_logger.py:301] number parameters: 1.104 billion I0418 01:09:56.364108 135124737042240 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0418 01:09:56.516869 135124737042240 checkpointing.py:776] Waited 0.15273833274841309 seconds for step 0 to finish before starting checkpointing. I0418 01:09:56.519475 135124737042240 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 01:09:56.521143 135124737042240 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0 I0418 01:09:56.522543 135124737042240 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0. I0418 01:09:57.694489 135124737042240 signaling_client.py:364] Using JaxDistributedSignalingClient I0418 01:09:57.708725 135124737042240 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 01:09:57.708791 135124737042240 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 01:09:57.978890 135124737042240 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.270960s I0418 01:09:57.979815 135124737042240 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.560 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27742481231689453 s) (per-host) I0418 01:09:57.979927 135124737042240 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.277549s (batch_requests_ready=0.002189s, total_serialization_initiated=0.274501s, others=0.000858s) I0418 01:09:57.980020 135124737042240 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.281415s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.281397) I0418 01:09:57.980802 135124737042240 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.46 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0. I0418 01:09:57.981240 134996026046208 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:29:57.981202 I0418 01:09:57.997740 135124737042240 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0418 01:09:57.998142 134992298505984 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0418 01:09:57.998280 135124737042240 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776474596.5194566, 'wait_for_prev_duration_secs': 6.079673767089844e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776474596.521181, 'checkpointer_blocking_duration_secs': 1.4602019786834717, 'get_old_steps_start_time': 1776474597.9814086, 'get_old_steps_duration_secs': 2.7418136596679688e-05, 'checkpoint_manager_blocking_start_time': 1776474596.517373, 'checkpoint_manager_blocking_duration_secs': 1.480867862701416} I0418 01:09:57.998388 135124737042240 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0418 01:09:57.998440 135124737042240 max_utils.py:750] Memstats: After params initialized: I0418 01:09:57.998489 135124737042240 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0418 01:09:57.998523 135124737042240 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0418 01:09:57.998554 135124737042240 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0418 01:09:57.998581 135124737042240 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0418 01:09:58.004008 134996017653504 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0 I0418 01:09:58.335944 135124737042240 metric_logger.py:196] completed step: 0, seconds: 11.153, TFLOP/s/device: 1.218, Tokens/s/device: 183.632, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0418 01:09:58.339781 135124737042240 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/tensorboard/' I0418 01:09:58.742448 135124737042240 metric_logger.py:196] completed step: 1, seconds: 1.970, TFLOP/s/device: 6.897, Tokens/s/device: 1039.557, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0418 01:09:58.888466 135124737042240 metric_logger.py:196] completed step: 2, seconds: 0.407, TFLOP/s/device: 33.374, Tokens/s/device: 5030.445, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0418 01:09:59.035928 135124737042240 metric_logger.py:196] completed step: 3, seconds: 0.009, TFLOP/s/device: 1484.608, Tokens/s/device: 223776.224, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203 I0418 01:10:01.055535 134996000868096 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/items I0418 01:10:01.211453 134992249616128 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776474600808776694, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776474600808776694, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0418 01:10:03.055073 2843 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0418 01:10:04.825189 134994429740800 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_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/items/array_metadatas/process_0 I0418 01:10:22.798026 135124737042240 metric_logger.py:196] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.473, Tokens/s/device: 13938.515, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614 I0418 01:10:22.806493 135124737042240 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.122, Tokens/s/device: 13885.593, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717 I0418 01:10:22.952600 135124737042240 metric_logger.py:196] completed step: 6, seconds: 23.761, TFLOP/s/device: 0.572, Tokens/s/device: 86.193, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759 I0418 01:10:23.100031 135124737042240 metric_logger.py:196] completed step: 7, seconds: 0.006, TFLOP/s/device: 2114.070, Tokens/s/device: 318655.671, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905 I0418 01:10:23.247239 135124737042240 metric_logger.py:196] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.407, Tokens/s/device: 13928.656, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406 I0418 01:10:23.393965 135124737042240 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0418 01:10:23.394586 135124737042240 checkpointing.py:776] Waited 0.0006430149078369141 seconds for step 9 to finish before starting checkpointing. I0418 01:10:23.396732 135124737042240 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 01:10:40.464893 134994387932928 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.411433s (commit=35.351935s, array_metadata_write=2.059499s) I0418 01:10:40.466022 134996026046208 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 36.938 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.76360249519348 s) (per-host) I0418 01:10:40.466160 134996026046208 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.484811s. I0418 01:10:41.198061 134996026046208 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776474600808776694, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0418 01:10:42.233472 134992249616128 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776474600808776694, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0418 01:10:42.296236 134996026046208 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000249s. I0418 01:10:42.583713 134996026046208 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0418 01:10:43.532116 134996026046208 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.198651s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/items I0418 01:10:43.533759 134996026046208 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0/items I0418 01:10:44.053135 134996026046208 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0 I0418 01:10:45.857764 134996026046208 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0`. I0418 01:10:45.858674 134996026046208 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 49.34 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/0 I0418 01:10:45.860129 134996026046208 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 47.878777s. I0418 01:10:45.860272 134992298505984 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 01:10:45.860352 134992298505984 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0418 01:10:45.860415 134992298505984 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0418 01:10:45.863544 134992298505984 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0418 01:10:45.863709 135124737042240 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. W0418 01:10:45.863853 135124737042240 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.467108 seconds. If this number is high, consider checkpointing less frequently. I0418 01:10:45.865583 135124737042240 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9 I0418 01:10:45.867622 135124737042240 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9. I0418 01:10:46.587457 135124737042240 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 01:10:46.587550 135124737042240 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 01:10:46.613473 135124737042240 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026804s I0418 01:10:46.614251 135124737042240 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 48.853 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.031575918197631836 s) (per-host) I0418 01:10:46.614315 135124737042240 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031656s (batch_requests_ready=0.001814s, total_serialization_initiated=0.029112s, others=0.000729s) I0418 01:10:46.614385 135124737042240 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035500s (all_items=0.000018s, per_item={'items': '0.00001812'}, temp_paths=0.035482) I0418 01:10:46.615083 135124737042240 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9. I0418 01:10:46.615350 134992298505984 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:30:46.615327 I0418 01:10:46.629413 135124737042240 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0418 01:10:46.629672 134994937280256 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0418 01:10:46.629817 135124737042240 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776474623.3967009, 'wait_for_prev_duration_secs': 22.467108249664307, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776474645.8656213, 'checkpointer_blocking_duration_secs': 0.749845027923584, 'get_old_steps_start_time': 1776474646.6154933, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776474623.3948102, 'checkpoint_manager_blocking_duration_secs': 23.2349636554718} I0418 01:10:46.629971 135124737042240 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0418 01:10:46.630030 135124737042240 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 01:10:46.823445 134996026046208 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9 I0418 01:10:49.845191 134994982860544 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9/items I0418 01:10:52.717955 134994954065664 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_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9/items/array_metadatas/process_0 I0418 01:11:34.240467 134994945672960 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 43.272146s (commit=41.225932s, array_metadata_write=2.046215s) I0418 01:11:34.241611 134992298505984 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 33.144 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.658888816833496 s) (per-host) I0418 01:11:34.241766 134992298505984 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.626342s. I0418 01:11:35.928096 134992298505984 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0418 01:11:36.136136 134992298505984 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000252s. I0418 01:11:37.035242 134992298505984 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.938853s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9/items I0418 01:11:37.038445 134992298505984 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9/items I0418 01:11:37.558233 134992298505984 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9 I0418 01:11:39.420180 134992298505984 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9`. I0418 01:11:39.420970 134992298505984 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.56 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_13_scan_layers_true/checkpoints/9 I0418 01:11:39.422415 134992298505984 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.806990s. I0418 01:11:39.422601 134994937280256 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 01:11:39.422723 134994937280256 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0418 01:11:39.422785 134994937280256 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0418 01:11:39.424621 134994937280256 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0418 01:11:39.424752 135124737042240 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. I0418 01:11:39.424865 135124737042240 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 01:11:39.425797 135124737042240 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.132, Tokens/s/device: 13887.194, 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: Sat Apr 18 01:11:48 UTC 2026 EXIT_CODE=0