XPK Start: Thu Apr 23 16:38:31 UTC 2026 2026-04-23 16:38:35.806025: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered WARNING: All log messages before absl::InitializeLog() is called are written to STDERR E0000 00:00:1776962315.819207 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776962315.822922 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776962315.834526 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776962315.834547 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776962315.834549 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776962315.834551 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 16:38:55.012584: 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) I0423 16:38:55.540458 137734109841216 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 16:39:04,581:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 16:39:04.581169 137734109841216 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 16:39:04,583:jax._src.distributed:157: Connecting to JAX distributed service on mt-13-scan-layers-true-i3eac-slice-job-0-0.mt-13-scan-layers-true-i3eac:8482 I0423 16:39:04.583453 137734109841216 distributed.py:157] Connecting to JAX distributed service on mt-13-scan-layers-true-i3eac-slice-job-0-0.mt-13-scan-layers-true-i3eac:8482 I0423 16:39:05.662211 137734109841216 max_utils.py:284] Jax distributed system initialized! I0423 16:39:12.558301 137734109841216 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 16:39:12.558407 137734109841216 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 16:39:12.558448 137734109841216 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709 I0423 16:39:12.558484 137734109841216 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 16:39:13.179980 137734109841216 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 16:39:13.180572 137734109841216 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 16:39:13.180752 137734109841216 checkpointing.py:688] Setting up checkpoint logger... I0423 16:39:13.180803 137734109841216 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 16:39:13.180847 137734109841216 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 16:39:13.181231 137734109841216 base_pytree_checkpoint_handler.py:415] 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 0x7d4436d51370>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 16:39:16.188758 137734109841216 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 16:39:16.189128 137734109841216 checkpoint_manager.py:709] [process=4][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d303b9a5700>}, handler_registry=None I0423 16:39:16.189425 137734109841216 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d303b9a5700>` 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`. I0423 16:39:16.189478 137734109841216 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d303b9aa4b0>` 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`. I0423 16:39:16.189515 137734109841216 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 0x7d303b9a5700>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d303b9a5700>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d303b9aa4b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d303b9aa4b0>}). I0423 16:39:16.190028 137734109841216 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 16:39:16.190109 137734109841216 async_checkpointer.py:177] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d2fb8364180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 16:39:17.200240 137734109841216 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints I0423 16:39:17.649807 137734109841216 checkpoint_manager.py:929] [process=4][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_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d42f4026c30> I0423 16:39:17.650017 137734109841216 checkpointing.py:302] Checkpoint manager created! I0423 16:39:18.562179 137734109841216 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 16:39:18.562283 137734109841216 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:39:18.928307 137734109841216 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 16:39:18.928404 137734109841216 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:39:18.944454 137734109841216 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 16:39:18.944518 137734109841216 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:39:18.967604 137734109841216 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 16:39:18.967673 137734109841216 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:39:18.983626 137734109841216 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 16:39:18.983685 137734109841216 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:39:18.999433 137734109841216 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 16:39:18.999490 137734109841216 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:39:19.024644 137734109841216 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 16:39:19.024717 137734109841216 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:39:19.044363 137734109841216 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 16:39:19.044430 137734109841216 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 16:39:19.234015 137734109841216 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 16:39:19.234122 137734109841216 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 16:39:20.663331 137734109841216 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0423 16:39:20.663461 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0423 16:39:20.663516 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0423 16:39:20.663576 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0423 16:39:20.663628 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0423 16:39:20.663667 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0423 16:39:20.663719 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0423 16:39:20.663771 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: PartitionSpec('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0423 16:39:20.663810 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0423 16:39:20.663847 137734109841216 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0423 16:39:20.663911 137734109841216 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0423 16:39:20.663963 137734109841216 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0423 16:39:21.172316 137734109841216 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 16:39:21.172409 137734109841216 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0423 16:39:21.187429 137734109841216 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 16:39:21.187492 137734109841216 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0423 16:39:31.906728 137734109841216 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. I0423 16:39:31.907547 137734109841216 metric_logger.py:289] number parameters: 1.104 billion I0423 16:39:43.127864 137734109841216 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 16:39:43.306123 137734109841216 checkpointing.py:798] Waited 0.17824411392211914 seconds for step 0 to finish before starting checkpointing. I0423 16:39:43.308954 137734109841216 checkpoint_manager.py:2013] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 16:39:43.310614 137734109841216 checkpoint_manager.py:1518] [process=4] Saving checkpoint at step 0 I0423 16:39:43.313798 137734109841216 async_checkpointer.py:452] [process=4] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/0. I0423 16:39:43.913602 137734109841216 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 16:39:43.914599 137734109841216 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 16:39:43.914655 137734109841216 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 I0423 16:39:44.161154 137734109841216 base_pytree_checkpoint_handler.py:153] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.247638s I0423 16:39:44.161335 137734109841216 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.875 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5365128517150879 s) (per-host) I0423 16:39:44.161391 137734109841216 base_pytree_checkpoint_handler.py:737] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.536579s (batch_requests_ready=0.285317s, total_serialization_initiated=0.251188s, others=0.000075s) I0423 16:39:44.161480 137734109841216 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.540723s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.540704) I0423 16:39:44.162379 137606322976512 async_checkpointer.py:79] [process=4][thread=async_save] Background save thread started. I0423 16:39:44.162493 137734109841216 async_checkpointer.py:561] Finished blocking save. Time taken: 0.851810s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/0. I0423 16:39:44.575821 137734109841216 checkpoint_manager.py:1566] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 16:39:44.576249 137607901751040 async_checkpointer.py:265] [process=4][thread=save_finalize] Waiting for background save thread=async_save. I0423 16:39:44.576416 137734109841216 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776962383.3089359, 'wait_for_prev_duration_secs': 6.29425048828125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776962383.3106527, 'checkpointer_blocking_duration_secs': 0.851952314376831, 'get_old_steps_start_time': 1776962384.1626267, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776962383.306686, 'checkpoint_manager_blocking_duration_secs': 1.2696881294250488} I0423 16:39:44.576537 137734109841216 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 16:39:44.576610 137734109841216 max_utils.py:750] Memstats: After params initialized: I0423 16:39:44.576684 137734109841216 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0)) I0423 16:39:44.576742 137734109841216 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0)) I0423 16:39:44.576778 137734109841216 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0)) I0423 16:39:44.576821 137734109841216 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0)) I0423 16:39:44.845824 137734109841216 metric_logger.py:185] completed step: 0, seconds: 11.220, TFLOP/s/device: 1.211, Tokens/s/device: 182.528, total_weights: 65536, loss: 10.874 I0423 16:39:45.016093 137734109841216 metric_logger.py:185] completed step: 1, seconds: 1.717, TFLOP/s/device: 7.915, Tokens/s/device: 1193.048, total_weights: 65536, loss: 10.874 I0423 16:39:45.487765 137734109841216 metric_logger.py:185] completed step: 2, seconds: 0.023, TFLOP/s/device: 587.933, Tokens/s/device: 88619.645, total_weights: 65536, loss: 10.263 I0423 16:39:45.635320 137734109841216 metric_logger.py:185] completed step: 3, seconds: 0.472, TFLOP/s/device: 28.783, Tokens/s/device: 4338.533, total_weights: 65536, loss: 9.734 I0423 16:39:45.930514 137734109841216 metric_logger.py:185] completed step: 4, seconds: 0.153, TFLOP/s/device: 89.001, Tokens/s/device: 13415.257, total_weights: 65536, loss: 9.278 I0423 16:39:45.936425 137734109841216 metric_logger.py:185] completed step: 5, seconds: 0.148, TFLOP/s/device: 92.009, Tokens/s/device: 13868.668, total_weights: 65536, loss: 8.892 I0423 16:40:04.385919 137734109841216 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.955, Tokens/s/device: 6926.852, total_weights: 65536, loss: 8.592 I0423 16:40:04.533595 137734109841216 metric_logger.py:185] completed step: 7, seconds: 18.303, TFLOP/s/device: 0.742, Tokens/s/device: 111.895, total_weights: 65536, loss: 8.384 I0423 16:40:04.681012 137734109841216 metric_logger.py:185] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.296, Tokens/s/device: 13459.693, total_weights: 65536, loss: 8.255 I0423 16:40:04.828157 137734109841216 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 16:40:04.828798 137734109841216 checkpointing.py:798] Waited 0.0006606578826904297 seconds for step 9 to finish before starting checkpointing. I0423 16:40:04.830810 137734109841216 checkpoint_manager.py:2024] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 16:40:05.271154 2686 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 16:40:07.232616 137607910143744 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/0/items/array_metadatas/process_4 I0423 16:40:38.471727 137606322976512 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/gbytes_per_sec: 28.800 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 54.846866607666016 s) (per-host) I0423 16:40:38.471843 137606322976512 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 54.309397s. I0423 16:40:47.309310 137606322976512 async_checkpointer.py:144] [process=4][thread=async_save] Background save thread done. Time taken: 63.146847s. I0423 16:40:47.309538 137607901751040 async_checkpointer.py:273] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 16:40:47.309605 137607901751040 async_checkpointer.py:283] [process=4][thread=save_finalize] No errors found in background save thread=async_save. I0423 16:40:47.309654 137607901751040 checkpoint_manager.py:2133] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 16:40:47.311397 137607901751040 checkpoint_manager.py:2142] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 16:40:47.311585 137734109841216 checkpoint_manager.py:2036] [process=4][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0423 16:40:47.311716 137734109841216 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.480911 seconds. If this number is high, consider checkpointing less frequently. I0423 16:40:47.313720 137734109841216 checkpoint_manager.py:1518] [process=4] Saving checkpoint at step 9 I0423 16:40:47.317158 137734109841216 async_checkpointer.py:452] [process=4] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/9. I0423 16:40:47.875483 137734109841216 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 16:40:47.875583 137734109841216 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 I0423 16:40:47.920706 137734109841216 base_pytree_checkpoint_handler.py:153] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.046593s I0423 16:40:47.920865 137734109841216 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.102 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30232977867126465 s) (per-host) I0423 16:40:47.920934 137734109841216 base_pytree_checkpoint_handler.py:737] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.302408s (batch_requests_ready=0.253930s, total_serialization_initiated=0.048392s, others=0.000086s) I0423 16:40:47.921018 137734109841216 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.306574s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.306558) I0423 16:40:47.921913 137607878240000 async_checkpointer.py:79] [process=4][thread=async_save] Background save thread started. I0423 16:40:47.922072 137734109841216 async_checkpointer.py:561] Finished blocking save. Time taken: 0.608282s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/9. I0423 16:40:47.936394 137734109841216 checkpoint_manager.py:1566] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 16:40:47.936692 137607901751040 async_checkpointer.py:265] [process=4][thread=save_finalize] Waiting for background save thread=async_save. I0423 16:40:47.936858 137734109841216 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776962404.8307815, 'wait_for_prev_duration_secs': 42.480910539627075, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776962447.3137596, 'checkpointer_blocking_duration_secs': 0.6084189414978027, 'get_old_steps_start_time': 1776962447.9221983, 'get_old_steps_duration_secs': 2.4557113647460938e-05, 'checkpoint_manager_blocking_start_time': 1776962404.8290741, 'checkpoint_manager_blocking_duration_secs': 43.10774874687195} I0423 16:40:47.936974 137734109841216 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 16:40:47.937019 137734109841216 checkpoint_manager.py:2024] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 16:40:52.763736 137607910143744 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/9/items/array_metadatas/process_4 I0423 16:41:29.518301 137607878240000 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/gbytes_per_sec: 37.699 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.899733781814575 s) (per-host) I0423 16:41:29.518405 137607878240000 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.596378s. I0423 16:41:36.720808 137607878240000 async_checkpointer.py:144] [process=4][thread=async_save] Background save thread done. Time taken: 48.798766s. I0423 16:41:36.721144 137607901751040 async_checkpointer.py:273] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 16:41:36.721266 137607901751040 async_checkpointer.py:283] [process=4][thread=save_finalize] No errors found in background save thread=async_save. I0423 16:41:36.721312 137607901751040 checkpoint_manager.py:2133] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 16:41:36.722864 137607901751040 checkpoint_manager.py:2142] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 16:41:36.723056 137734109841216 checkpoint_manager.py:2036] [process=4][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0423 16:41:36.723212 137734109841216 checkpoint_manager.py:2013] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 16:41:36.723887 137734109841216 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.097, Tokens/s/device: 13881.828, total_weights: 65536, loss: 8.179 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Thu Apr 23 16:41:47 UTC 2026 EXIT_CODE=0