XPK Start: Sun Apr 19 03:49:20 UTC 2026 2026-04-19 03:49:24.813706: 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 00:00:1776570564.826563 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776570564.830276 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776570564.841514 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776570564.841533 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776570564.841535 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776570564.841537 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 03:49:44.007790: 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) I0419 03:49:44.533139 140573438011200 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 03:49:53,573:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 03:49:53.573222 140573438011200 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 03:49:53,575:jax._src.distributed:157: Connecting to JAX distributed service on mt-09-pdb-lt-1-n0q4t-slice-job-0-0.mt-09-pdb-lt-1-n0q4t:8482 I0419 03:49:53.575506 140573438011200 distributed.py:157] Connecting to JAX distributed service on mt-09-pdb-lt-1-n0q4t-slice-job-0-0.mt-09-pdb-lt-1-n0q4t:8482 I0419 03:49:54.899932 140573438011200 max_utils.py:284] Jax distributed system initialized! I0419 03:50:01.962739 140573438011200 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 03:50:01.962922 140573438011200 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 03:50:01.962967 140573438011200 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 I0419 03:50:01.963003 140573438011200 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 03:50:02.583944 140573438011200 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0419 03:50:02.584600 140573438011200 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0419 03:50:02.584820 140573438011200 checkpointing.py:688] Setting up checkpoint logger... I0419 03:50:02.584881 140573438011200 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 03:50:02.584947 140573438011200 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 03:50:02.585368 140573438011200 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 0x7fd944f75580>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 03:50:05.874004 140573438011200 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0419 03:50:05.874254 140573438011200 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7fc4f81b1b80>}, handler_registry=None I0419 03:50:05.874493 140573438011200 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7fc4f81b1b80>` 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`. I0419 03:50:05.874541 140573438011200 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7fc4f81b7140>` 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`. I0419 03:50:05.874576 140573438011200 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 0x7fc4f81b1b80>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7fc4f81b1b80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7fc4f81b7140>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7fc4f81b7140>}). I0419 03:50:05.875092 140573438011200 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 03:50:05.875183 140573438011200 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7fc4e0390180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 03:50:07.865875 140573438011200 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints I0419 03:50:08.739555 140573438011200 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_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7fc528102db0> I0419 03:50:08.739738 140573438011200 checkpointing.py:302] Checkpoint manager created! I0419 03:50:09.630255 140573438011200 nnx_wrappers.py:455] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 03:50:09.630369 140573438011200 nnx_wrappers.py:455] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 03:50:10.002105 140573438011200 attentions.py:1084] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 03:50:10.002208 140573438011200 attentions.py:1084] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 03:50:10.018065 140573438011200 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 03:50:10.018125 140573438011200 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 03:50:10.040882 140573438011200 attentions.py:1150] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 03:50:10.040951 140573438011200 attentions.py:1150] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 03:50:10.056940 140573438011200 attentions.py:1151] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 03:50:10.057003 140573438011200 attentions.py:1151] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 03:50:10.072932 140573438011200 attentions.py:1152] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 03:50:10.072993 140573438011200 attentions.py:1152] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 03:50:10.097934 140573438011200 attentions.py:1193] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 03:50:10.098006 140573438011200 attentions.py:1193] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 03:50:10.117667 140573438011200 linears.py:541] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 03:50:10.117735 140573438011200 linears.py:541] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor'). I0419 03:50:10.308854 140573438011200 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0419 03:50:10.308964 140573438011200 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 8 tensor: 4 I0419 03:50:11.757956 140573438011200 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: ('tensor',) I0419 03:50:11.758090 140573438011200 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, 'tensor') I0419 03:50:11.758153 140573438011200 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, 'tensor') I0419 03:50:11.758214 140573438011200 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: ('tensor', None, 'fsdp') I0419 03:50:11.758266 140573438011200 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: ('tensor', None) I0419 03:50:11.758308 140573438011200 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: ('tensor', None) I0419 03:50:11.758365 140573438011200 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, 'tensor', None) I0419 03:50:11.758420 140573438011200 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: ('tensor', None, None, 'fsdp') I0419 03:50:11.758462 140573438011200 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, 'tensor', None) I0419 03:50:11.758500 140573438011200 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, 'tensor', None) I0419 03:50:11.758551 140573438011200 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', 'tensor') I0419 03:50:11.758601 140573438011200 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: ('tensor', 'fsdp') I0419 03:50:12.262659 140573438011200 train.py:158] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0419 03:50:12.262757 140573438011200 train.py:158] train/xent Physical: float32[8,2048]............................................. ('fsdp', None). I0419 03:50:12.278044 140573438011200 train.py:165] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0419 03:50:12.278105 140573438011200 train.py:165] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None). I0419 03:50:22.098127 140573438011200 max_utils.py:791] Total memory size: 0.9 GB, Output size: 0.4 GB, Temp size: 0.5 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0419 03:50:22.109892 140573438011200 metric_logger.py:289] number parameters: 1.104 billion I0419 03:50:23.910644 140573438011200 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0419 03:50:32.137058 140573438011200 checkpointing.py:798] Waited 8.226396322250366 seconds for step 0 to finish before starting checkpointing. I0419 03:50:32.139539 140573438011200 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 03:50:32.141447 140573438011200 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0419 03:50:32.144194 140573438011200 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0. I0419 03:50:33.265573 140573438011200 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 03:50:33.534587 140447208945408 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0 I0419 03:50:33.996272 140573438011200 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 03:50:33.996366 140573438011200 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 I0419 03:50:34.219713 140573438011200 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.224462s I0419 03:50:34.220520 140573438011200 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.629 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.946845531463623 s) (per-host) I0419 03:50:34.220591 140573438011200 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.946938s (batch_requests_ready=0.717929s, total_serialization_initiated=0.228292s, others=0.000718s) I0419 03:50:34.220668 140573438011200 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.950859s (all_items=0.000020s, per_item={'items': '0.00001979'}, temp_paths=0.950839) I0419 03:50:34.221590 140446131017472 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 03:50:34.221749 140573438011200 async_checkpointer.py:561] Finished blocking save. Time taken: 2.080232s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0. I0419 03:50:34.236034 140573438011200 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 03:50:34.236312 140447217338112 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 03:50:34.236470 140573438011200 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776570632.139522, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776570632.141485, 'checkpointer_blocking_duration_secs': 2.0803825855255127, 'get_old_steps_start_time': 1776570634.2218902, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776570632.1375241, 'checkpoint_manager_blocking_duration_secs': 2.0989022254943848} I0419 03:50:34.236584 140573438011200 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0419 03:50:34.236636 140573438011200 max_utils.py:750] Memstats: After params initialized: I0419 03:50:34.236687 140573438011200 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0419 03:50:34.236720 140573438011200 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0419 03:50:34.236746 140573438011200 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0419 03:50:34.236770 140573438011200 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0419 03:50:34.526360 140573438011200 metric_logger.py:185] completed step: 0, seconds: 1.546, TFLOP/s/device: 2.197, Tokens/s/device: 331.133, total_weights: 16384, loss: 10.862 I0419 03:50:34.529265 140573438011200 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/tensorboard/' I0419 03:50:34.925927 140573438011200 metric_logger.py:185] completed step: 1, seconds: 10.614, TFLOP/s/device: 0.320, Tokens/s/device: 48.238, total_weights: 16384, loss: 10.862 I0419 03:50:34.983257 140573438011200 metric_logger.py:185] completed step: 2, seconds: 0.400, TFLOP/s/device: 8.486, Tokens/s/device: 1279.054, total_weights: 16384, loss: 9.763 I0419 03:50:35.042118 140573438011200 metric_logger.py:185] completed step: 3, seconds: 0.008, TFLOP/s/device: 406.071, Tokens/s/device: 61207.412, total_weights: 16384, loss: 8.823 I0419 03:50:48.469114 140573438011200 metric_logger.py:185] completed step: 4, seconds: 0.058, TFLOP/s/device: 58.490, Tokens/s/device: 8816.186, total_weights: 16384, loss: 7.947 I0419 03:50:48.477533 140573438011200 metric_logger.py:185] completed step: 5, seconds: 0.059, TFLOP/s/device: 57.515, Tokens/s/device: 8669.297, total_weights: 16384, loss: 7.212 I0419 03:50:48.534769 140573438011200 metric_logger.py:185] completed step: 6, seconds: 13.426, TFLOP/s/device: 0.253, Tokens/s/device: 38.135, total_weights: 16384, loss: 6.652 I0419 03:50:48.593388 140573438011200 metric_logger.py:185] completed step: 7, seconds: 0.006, TFLOP/s/device: 542.097, Tokens/s/device: 81710.820, total_weights: 16384, loss: 6.269 I0419 03:50:48.652204 140573438011200 metric_logger.py:185] completed step: 8, seconds: 0.059, TFLOP/s/device: 58.038, Tokens/s/device: 8748.099, total_weights: 16384, loss: 6.031 I0419 03:50:48.710404 140573438011200 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0419 03:50:48.711050 140573438011200 checkpointing.py:798] Waited 0.0006620883941650391 seconds for step 9 to finish before starting checkpointing. I0419 03:50:48.712866 140573438011200 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 03:50:48.970695 140446654056192 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items I0419 03:50:49.553725 140445051836160 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776570648714669428, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776570648714669428, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/_CHECKPOINT_METADATA I0419 03:50:50.063604 2496 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 03:50:52.519842 140450967041792 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_0 I0419 03:51:27.644783 140450958649088 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.584638s (commit=35.775887s, array_metadata_write=1.808751s) I0419 03:51:27.646886 140446131017472 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 29.051 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 54.37320899963379 s) (per-host) I0419 03:51:27.646951 140446131017472 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 53.425251s. I0419 03:51:29.198401 140446131017472 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776570648714669428, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/_CHECKPOINT_METADATA I0419 03:51:29.951319 140445051836160 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776570648714669428, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/_CHECKPOINT_METADATA I0419 03:51:30.565938 140446131017472 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 03:51:30.716793 140446131017472 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000236s. I0419 03:51:31.472001 140446131017472 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.146503s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items I0419 03:51:31.473609 140446131017472 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items I0419 03:51:32.290120 140446131017472 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0 I0419 03:51:33.938040 140446131017472 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0`. I0419 03:51:33.938839 140446131017472 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 61.797311s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0 I0419 03:51:33.940941 140446131017472 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 59.719240s. I0419 03:51:33.941131 140447217338112 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 03:51:33.941248 140447217338112 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 03:51:33.941309 140447217338112 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 03:51:33.943319 140447217338112 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 03:51:33.943460 140573438011200 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 03:51:33.943530 140573438011200 checkpoint_manager.py:1458] Waiting for previous save to complete took 45.230670 seconds. If this number is high, consider checkpointing less frequently. I0419 03:51:33.945607 140573438011200 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0419 03:51:33.949171 140573438011200 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9. I0419 03:51:34.895178 140446131017472 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9 I0419 03:51:34.902099 140573438011200 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 03:51:34.902208 140573438011200 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 I0419 03:51:34.931919 140573438011200 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.030847s I0419 03:51:34.932718 140573438011200 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.484 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2812936305999756 s) (per-host) I0419 03:51:34.932794 140573438011200 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.281387s (batch_requests_ready=0.247586s, total_serialization_initiated=0.033059s, others=0.000742s) I0419 03:51:34.932886 140573438011200 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285426s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.285409) I0419 03:51:34.933704 140446151988992 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 03:51:34.933873 140573438011200 async_checkpointer.py:561] Finished blocking save. Time taken: 0.988195s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9. I0419 03:51:35.390013 140573438011200 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 03:51:35.390362 140447217338112 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 03:51:35.390532 140573438011200 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776570648.7128356, 'wait_for_prev_duration_secs': 45.23066997528076, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776570693.9456453, 'checkpointer_blocking_duration_secs': 0.9883368015289307, 'get_old_steps_start_time': 1776570694.934006, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776570648.7112663, 'checkpoint_manager_blocking_duration_secs': 46.6792311668396} I0419 03:51:35.390639 140573438011200 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0419 03:51:35.390684 140573438011200 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 03:51:38.366523 140440229623552 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items I0419 03:51:43.202426 140446654056192 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_0 I0419 03:52:21.113608 140446160381696 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 40.376383s (commit=39.421680s, array_metadata_write=0.954703s) I0419 03:52:21.114630 140446151988992 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 33.997 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.46317958831787 s) (per-host) I0419 03:52:21.114730 140446151988992 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.180904s. I0419 03:52:23.661202 140446151988992 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 03:52:24.694853 140446151988992 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000242s. I0419 03:52:24.756368 140446151988992 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.920295s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items I0419 03:52:24.757937 140446151988992 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items I0419 03:52:25.149371 140446151988992 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9 I0419 03:52:26.784402 140446151988992 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9`. I0419 03:52:26.785171 140446151988992 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 52.839500s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9 I0419 03:52:26.786618 140446151988992 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 51.852790s. I0419 03:52:26.786740 140447217338112 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 03:52:26.786823 140447217338112 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 03:52:26.786904 140447217338112 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 03:52:26.788696 140447217338112 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 03:52:26.788833 140573438011200 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 03:52:26.788900 140573438011200 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 03:52:26.789537 140573438011200 metric_logger.py:185] completed step: 9, seconds: 0.059, TFLOP/s/device: 57.736, Tokens/s/device: 8702.599, total_weights: 16384, loss: 5.892 Per train step: Total TFLOPs: 3.40 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 03:52:36 UTC 2026 EXIT_CODE=0