XPK Start: Sun Apr 19 05:26:49 UTC 2026 2026-04-19 05:26:53.559606: 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 00:00:1776576413.572218 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776576413.575991 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776576413.587307 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776576413.587325 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776576413.587327 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776576413.587329 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 05:27:12.647024: 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 05:27:13.175526 135825742202688 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 05:27:22,216:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 05:27:22.216435 135825742202688 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 05:27:22,218:jax._src.distributed:157: Connecting to JAX distributed service on mt-09-pdb-lt-1-zi4se-slice-job-0-0.mt-09-pdb-lt-1-zi4se:8482 I0419 05:27:22.218728 135825742202688 distributed.py:157] Connecting to JAX distributed service on mt-09-pdb-lt-1-zi4se-slice-job-0-0.mt-09-pdb-lt-1-zi4se:8482 I0419 05:27:24.275489 135825742202688 max_utils.py:284] Jax distributed system initialized! I0419 05:27:31.225117 135825742202688 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 05:27:31.225227 135825742202688 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 05:27:31.225268 135825742202688 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 05:27:31.225302 135825742202688 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 05:27:31.831274 135825742202688 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0419 05:27:31.945137 135825742202688 checkpointing.py:688] Setting up checkpoint logger... I0419 05:27:31.945258 135825742202688 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 05:27:31.945303 135825742202688 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 05:27:31.945569 135825742202688 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 0x7b87eb3791c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 05:27:36.169359 135825742202688 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0419 05:27:36.169605 135825742202688 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 0x7b74a763faa0>}, handler_registry=None I0419 05:27:36.169908 135825742202688 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b74a763faa0>` 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 05:27:36.169975 135825742202688 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b73703d5d90>` 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 05:27:36.170015 135825742202688 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 0x7b74a763faa0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b74a763faa0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b73703d5d90>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b73703d5d90>}). I0419 05:27:36.170358 135825742202688 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 05:27:36.170441 135825742202688 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b7370344f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 05:27:36.895887 135825742202688 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints I0419 05:27:37.774567 135825742202688 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/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b73703c6fc0> I0419 05:27:37.774748 135825742202688 checkpointing.py:302] Checkpoint manager created! I0419 05:27:38.716440 135825742202688 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0419 05:27:38.716560 135825742202688 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 8 tensor: 4 I0419 05:27:40.163362 135825742202688 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: ('tensor',) I0419 05:27:40.163495 135825742202688 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, 'tensor') I0419 05:27:40.163550 135825742202688 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, 'tensor') I0419 05:27:40.163594 135825742202688 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: ('tensor', None, 'fsdp') I0419 05:27:40.163634 135825742202688 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: ('tensor', None) I0419 05:27:40.163671 135825742202688 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: ('tensor', None) I0419 05:27:40.163709 135825742202688 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, 'tensor', None) I0419 05:27:40.163748 135825742202688 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: ('tensor', None, None, 'fsdp') I0419 05:27:40.163784 135825742202688 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, 'tensor', None) I0419 05:27:40.163845 135825742202688 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, 'tensor', None) I0419 05:27:40.163886 135825742202688 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', 'tensor') I0419 05:27:40.163920 135825742202688 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: ('tensor', 'fsdp') I0419 05:27:40.558709 135825742202688 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 05:27:40.558872 135825742202688 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 05:27:40.564247 135825742202688 nnx_decoders.py:616] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 05:27:40.564339 135825742202688 nnx_decoders.py:616] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 05:27:40.712166 135825742202688 attentions.py:1084] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 05:27:40.712298 135825742202688 attentions.py:1084] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 05:27:40.858110 135825742202688 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 05:27:40.858243 135825742202688 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 05:27:41.013467 135825742202688 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 05:27:41.013603 135825742202688 attentions.py:1150] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 05:27:41.160667 135825742202688 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 05:27:41.160802 135825742202688 attentions.py:1151] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 05:27:41.306018 135825742202688 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 05:27:41.306153 135825742202688 attentions.py:1152] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 05:27:41.462212 135825742202688 attentions.py:1193] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 05:27:41.462349 135825742202688 attentions.py:1193] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 05:27:41.484623 135825742202688 linears.py:541] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 05:27:41.484721 135825742202688 linears.py:541] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor'). I0419 05:27:51.792266 135825742202688 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 05:27:51.799750 135825742202688 metric_logger.py:289] number parameters: 1.104 billion I0419 05:27:53.407247 135825742202688 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0419 05:28:01.165111 135825742202688 checkpointing.py:798] Waited 7.757848739624023 seconds for step 0 to finish before starting checkpointing. I0419 05:28:01.167398 135825742202688 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 05:28:01.169078 135825742202688 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0419 05:28:01.172057 135825742202688 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0. I0419 05:28:01.872434 135825742202688 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 05:28:02.125744 135699693041408 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0 I0419 05:28:02.148709 135825742202688 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0419 05:28:02.148874 135825742202688 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 05:28:02.372749 135825742202688 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.226106s I0419 05:28:02.373658 135825742202688 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.126 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.4934515953063965 s) (per-host) I0419 05:28:02.373736 135825742202688 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.493548s (batch_requests_ready=0.259228s, total_serialization_initiated=0.233507s, others=0.000813s) I0419 05:28:02.373836 135825742202688 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.497470s (all_items=0.000023s, per_item={'items': '0.00002337'}, temp_paths=0.497447) I0419 05:28:02.374771 135698699028224 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 05:28:02.374969 135825742202688 async_checkpointer.py:561] Finished blocking save. Time taken: 1.205818s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0. I0419 05:28:02.830989 135825742202688 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 05:28:02.831365 135700313773824 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 05:28:02.831532 135825742202688 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_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': 1776576481.1673796, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776576481.1691184, 'checkpointer_blocking_duration_secs': 1.2059602737426758, 'get_old_steps_start_time': 1776576482.3751004, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1776576481.1655216, 'checkpoint_manager_blocking_duration_secs': 1.6659698486328125} I0419 05:28:02.831642 135825742202688 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0419 05:28:02.831713 135825742202688 max_utils.py:750] Memstats: After params initialized: I0419 05:28:02.831769 135825742202688 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0419 05:28:02.831802 135825742202688 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0419 05:28:02.831842 135825742202688 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0419 05:28:02.831866 135825742202688 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0419 05:28:03.130048 135825742202688 metric_logger.py:185] completed step: 0, seconds: 1.352, TFLOP/s/device: 2.513, Tokens/s/device: 378.773, total_weights: 16384, loss: 10.864 I0419 05:28:03.133534 135825742202688 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/tensorboard/' I0419 05:28:03.556546 135825742202688 metric_logger.py:185] completed step: 1, seconds: 9.719, TFLOP/s/device: 0.349, Tokens/s/device: 52.678, total_weights: 16384, loss: 10.864 I0419 05:28:03.609607 135825742202688 metric_logger.py:185] completed step: 2, seconds: 0.428, TFLOP/s/device: 7.943, Tokens/s/device: 1197.297, total_weights: 16384, loss: 9.760 I0419 05:28:03.668516 135825742202688 metric_logger.py:185] completed step: 3, seconds: 0.011, TFLOP/s/device: 303.095, Tokens/s/device: 45685.732, total_weights: 16384, loss: 8.820 I0419 05:28:16.881638 135825742202688 metric_logger.py:185] completed step: 4, seconds: 0.054, TFLOP/s/device: 62.641, Tokens/s/device: 9441.965, total_weights: 16384, loss: 7.947 I0419 05:28:16.891615 135825742202688 metric_logger.py:185] completed step: 5, seconds: 0.058, TFLOP/s/device: 58.075, Tokens/s/device: 8753.633, total_weights: 16384, loss: 7.209 I0419 05:28:16.947019 135825742202688 metric_logger.py:185] completed step: 6, seconds: 13.212, TFLOP/s/device: 0.257, Tokens/s/device: 38.752, total_weights: 16384, loss: 6.651 I0419 05:28:17.005709 135825742202688 metric_logger.py:185] completed step: 7, seconds: 0.008, TFLOP/s/device: 448.894, Tokens/s/device: 67662.218, total_weights: 16384, loss: 6.267 I0419 05:28:17.064272 135825742202688 metric_logger.py:185] completed step: 8, seconds: 0.057, TFLOP/s/device: 59.823, Tokens/s/device: 9017.101, total_weights: 16384, loss: 6.030 I0419 05:28:17.123873 135825742202688 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0419 05:28:17.124524 135825742202688 checkpointing.py:798] Waited 0.0006570816040039062 seconds for step 9 to finish before starting checkpointing. I0419 05:28:17.126399 135825742202688 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 05:28:17.392645 135699676256000 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items I0419 05:28:17.548523 135697099638528 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776576497145267201, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776576497145267201, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/_CHECKPOINT_METADATA I0419 05:28:18.950320 2480 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 05:28:21.465827 135703006521088 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_0 I0419 05:28:55.732460 135702998128384 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 36.803894s (commit=35.412013s, array_metadata_write=1.391881s) I0419 05:28:55.733592 135698699028224 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 29.331 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 53.853376626968384 s) (per-host) I0419 05:28:55.733658 135698699028224 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 53.358743s. I0419 05:28:56.415064 135698699028224 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776576497145267201, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/_CHECKPOINT_METADATA I0419 05:28:57.141559 135697099638528 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776576497145267201, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/_CHECKPOINT_METADATA I0419 05:28:57.581550 135698699028224 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000422s. I0419 05:28:57.770907 135698699028224 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 05:28:58.685235 135698699028224 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.151467s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items I0419 05:28:58.687987 135698699028224 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0/items I0419 05:28:59.095420 135698699028224 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0 I0419 05:29:00.388610 135698699028224 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0`. I0419 05:29:00.389436 135698699028224 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 59.220292s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/0 I0419 05:29:00.390979 135698699028224 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 58.016062s. I0419 05:29:00.391167 135700313773824 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 05:29:00.391285 135700313773824 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 05:29:00.391354 135700313773824 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 05:29:00.394214 135700313773824 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 05:29:00.394378 135825742202688 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 05:29:00.394508 135825742202688 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.268118 seconds. If this number is high, consider checkpointing less frequently. I0419 05:29:00.396574 135825742202688 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0419 05:29:00.400089 135825742202688 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9. I0419 05:29:01.365529 135698699028224 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9 I0419 05:29:01.810477 135825742202688 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0419 05:29:01.810577 135825742202688 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 05:29:01.842251 135825742202688 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.034002s I0419 05:29:01.843076 135825742202688 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.152 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.7167477607727051 s) (per-host) I0419 05:29:01.843151 135825742202688 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.716839s (batch_requests_ready=0.676311s, total_serialization_initiated=0.039785s, others=0.000742s) I0419 05:29:01.843243 135825742202688 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.721306s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.721290) I0419 05:29:01.844126 135698174682880 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 05:29:01.844273 135825742202688 async_checkpointer.py:561] Finished blocking save. Time taken: 1.447637s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9. I0419 05:29:01.873072 135825742202688 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 05:29:01.873358 135700313773824 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 05:29:01.873525 135825742202688 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_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': 1776576497.126364, 'wait_for_prev_duration_secs': 43.268118143081665, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776576540.3966134, 'checkpointer_blocking_duration_secs': 1.4477319717407227, 'get_old_steps_start_time': 1776576541.8443685, 'get_old_steps_duration_secs': 3.361701965332031e-05, 'checkpoint_manager_blocking_start_time': 1776576497.1247814, 'checkpoint_manager_blocking_duration_secs': 44.74870800971985} I0419 05:29:01.873636 135825742202688 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0419 05:29:01.873682 135825742202688 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 05:29:03.127484 135698199860992 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items I0419 05:29:06.749220 135698209912576 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_0 I0419 05:29:46.384397 135698191468288 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.277392s (commit=39.783416s, array_metadata_write=1.493976s) I0419 05:29:46.385570 135698174682880 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.901 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.259215116500854 s) (per-host) I0419 05:29:46.385680 135698174682880 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.541446s. I0419 05:29:48.512127 135698174682880 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 05:29:48.644179 135698174682880 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000388s. I0419 05:29:49.423413 135698174682880 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.170104s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items I0419 05:29:49.425039 135698174682880 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9/items I0419 05:29:49.832413 135698174682880 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9 I0419 05:29:51.967236 135698174682880 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9`. I0419 05:29:51.968020 135698174682880 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 51.571381s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_09_pdb_lt_1/checkpoints/9 I0419 05:29:51.969466 135698174682880 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.125232s. I0419 05:29:51.969646 135700313773824 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 05:29:51.969769 135700313773824 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 05:29:51.969850 135700313773824 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 05:29:51.972194 135700313773824 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 05:29:51.972368 135825742202688 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 05:29:51.972527 135825742202688 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 05:29:51.973277 135825742202688 metric_logger.py:185] completed step: 9, seconds: 0.058, TFLOP/s/device: 58.295, Tokens/s/device: 8786.833, total_weights: 16384, loss: 5.890 Per train step: Total TFLOPs: 3.40 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 05:30:00 UTC 2026 EXIT_CODE=0