XPK Start: Sat Apr 25 16:14:33 UTC 2026 2026-04-25 16:14:37.244390: 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 00:00:1777133677.257695 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1777133677.261464 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1777133677.273368 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777133677.273387 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777133677.273389 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777133677.273391 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-25 16:14:56.437319: 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) I0425 16:14:56.960567 135176477742912 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-25 16:15:06,001:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0425 16:15:06.001176 135176477742912 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-25 16:15:06,003:jax._src.distributed:157: Connecting to JAX distributed service on mt-16-zarr3-true-sv-dwj5y-slice-job-0-0.mt-16-zarr3-true-sv-dwj5y:8482 I0425 16:15:06.003569 135176477742912 distributed.py:157] Connecting to JAX distributed service on mt-16-zarr3-true-sv-dwj5y-slice-job-0-0.mt-16-zarr3-true-sv-dwj5y:8482 I0425 16:15:07.100935 135176477742912 max_utils.py:284] Jax distributed system initialized! I0425 16:15:13.957050 135176477742912 max_utils.py:800] System Information: Jax Version: 0.8.1 I0425 16:15:13.957223 135176477742912 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0425 16:15:13.957267 135176477742912 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 I0425 16:15:13.957302 135176477742912 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0425 16:15:14.579245 135176477742912 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 16:15:14.579834 135176477742912 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 16:15:14.580016 135176477742912 checkpointing.py:688] Setting up checkpoint logger... I0425 16:15:14.580067 135176477742912 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0425 16:15:14.580132 135176477742912 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0425 16:15:14.580495 135176477742912 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 0x7af0b1179370>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 16:15:17.503207 135176477742912 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0425 16:15:17.503559 135176477742912 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7adc600f4ec0>}, handler_registry=None I0425 16:15:17.503791 135176477742912 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7adc600f4ec0>` 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`. I0425 16:15:17.503841 135176477742912 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7adc600f9e80>` 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`. I0425 16:15:17.503878 135176477742912 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 0x7adc600f4ec0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7adc600f4ec0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7adc600f9e80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7adc600f9e80>}). I0425 16:15:17.504392 135176477742912 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0425 16:15:17.504472 135176477742912 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7adc48764180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0425 16:15:18.279302 135176477742912 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints I0425 16:15:18.313817 135176477742912 checkpoint_manager.py:929] [process=6][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=5), 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_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7add4bae07a0> I0425 16:15:18.313959 135176477742912 checkpointing.py:302] Checkpoint manager created! I0425 16:15:19.212185 135176477742912 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 16:15:19.212295 135176477742912 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:15:19.580022 135176477742912 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 16:15:19.580141 135176477742912 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:15:19.596018 135176477742912 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 16:15:19.596078 135176477742912 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:15:19.618943 135176477742912 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 16:15:19.619014 135176477742912 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:15:19.635071 135176477742912 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 16:15:19.635145 135176477742912 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:15:19.650979 135176477742912 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 16:15:19.651039 135176477742912 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:15:19.675895 135176477742912 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0425 16:15:19.675965 135176477742912 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:15:19.695637 135176477742912 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0425 16:15:19.695702 135176477742912 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0425 16:15:19.885250 135176477742912 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0425 16:15:19.885360 135176477742912 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0425 16:15:21.325731 135176477742912 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0425 16:15:21.325902 135176477742912 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) I0425 16:15:21.325991 135176477742912 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) I0425 16:15:21.326084 135176477742912 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') I0425 16:15:21.326177 135176477742912 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) I0425 16:15:21.326239 135176477742912 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) I0425 16:15:21.326311 135176477742912 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) I0425 16:15:21.326370 135176477742912 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') I0425 16:15:21.326413 135176477742912 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) I0425 16:15:21.326454 135176477742912 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) I0425 16:15:21.326505 135176477742912 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0425 16:15:21.326576 135176477742912 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0425 16:15:21.828388 135176477742912 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0425 16:15:21.828481 135176477742912 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0425 16:15:21.843567 135176477742912 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0425 16:15:21.843627 135176477742912 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0425 16:15:32.730078 135176477742912 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. I0425 16:15:32.730942 135176477742912 metric_logger.py:289] number parameters: 1.104 billion I0425 16:15:43.975556 135176477742912 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0425 16:15:44.130979 135176477742912 checkpointing.py:798] Waited 0.15541529655456543 seconds for step 0 to finish before starting checkpointing. I0425 16:15:44.133589 135176477742912 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 16:15:44.135344 135176477742912 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0 I0425 16:15:44.138054 135176477742912 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/0. I0425 16:15:44.663480 135176477742912 signaling_client.py:364] Using JaxDistributedSignalingClient I0425 16:15:44.664790 135176477742912 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0425 16:15:44.664846 135176477742912 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 I0425 16:15:44.913297 135176477742912 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249896s I0425 16:15:44.913473 135176477742912 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.080 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5007584095001221 s) (per-host) I0425 16:15:44.913524 135176477742912 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.500818s (batch_requests_ready=0.247276s, total_serialization_initiated=0.253474s, others=0.000068s) I0425 16:15:44.913619 135176477742912 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.504826s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.504807) I0425 16:15:44.914555 135051387844352 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0425 16:15:44.914724 135176477742912 async_checkpointer.py:561] Finished blocking save. Time taken: 0.779305s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/0. I0425 16:15:45.288781 135176477742912 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0425 16:15:45.289208 135051413022464 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0425 16:15:45.289378 135176477742912 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777133744.1335702, 'wait_for_prev_duration_secs': 6.532669067382812e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777133744.1353853, 'checkpointer_blocking_duration_secs': 0.7794468402862549, 'get_old_steps_start_time': 1777133744.9148514, 'get_old_steps_duration_secs': 2.8371810913085938e-05, 'checkpoint_manager_blocking_start_time': 1777133744.131589, 'checkpoint_manager_blocking_duration_secs': 1.157745122909546} I0425 16:15:45.289494 135176477742912 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0425 16:15:45.289548 135176477742912 max_utils.py:750] Memstats: After params initialized: I0425 16:15:45.289608 135176477742912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0425 16:15:45.289641 135176477742912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0425 16:15:45.289669 135176477742912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0425 16:15:45.289693 135176477742912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0425 16:15:45.548686 135176477742912 metric_logger.py:185] completed step: 0, seconds: 11.244, TFLOP/s/device: 1.208, Tokens/s/device: 182.134, total_weights: 65536, loss: 10.874 I0425 16:15:45.718415 135176477742912 metric_logger.py:185] completed step: 1, seconds: 1.572, TFLOP/s/device: 8.644, Tokens/s/device: 1302.971, total_weights: 65536, loss: 10.874 I0425 16:15:46.132236 135176477742912 metric_logger.py:185] completed step: 2, seconds: 0.023, TFLOP/s/device: 601.680, Tokens/s/device: 90691.701, total_weights: 65536, loss: 10.560 I0425 16:15:46.279892 135176477742912 metric_logger.py:185] completed step: 3, seconds: 0.415, TFLOP/s/device: 32.771, Tokens/s/device: 4939.641, total_weights: 65536, loss: 9.980 I0425 16:15:46.284591 135176477742912 checkpointing.py:794] Waiting for step 5 to finish before checkpoint... I0425 16:15:46.574420 135176477742912 checkpointing.py:798] Waited 0.2897953987121582 seconds for step 5 to finish before starting checkpointing. I0425 16:15:46.577262 135176477742912 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 16:15:47.712127 2580 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0425 16:15:49.618675 135051396237056 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/0/items/array_metadatas/process_6 I0425 16:16:21.202451 135051387844352 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 42.936 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.78969478607178 s) (per-host) I0425 16:16:21.202577 135051387844352 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.287916s. I0425 16:16:30.689600 135051387844352 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 45.774923s. I0425 16:16:30.689904 135051413022464 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 16:16:30.690029 135051413022464 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0425 16:16:30.690079 135051413022464 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0425 16:16:30.691443 135051413022464 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0425 16:16:30.691643 135176477742912 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0425 16:16:30.691780 135176477742912 checkpoint_manager.py:1458] Waiting for previous save to complete took 44.114522 seconds. If this number is high, consider checkpointing less frequently. I0425 16:16:30.693661 135176477742912 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 5 I0425 16:16:30.697653 135176477742912 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/5. I0425 16:16:31.229578 135176477742912 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0425 16:16:31.229675 135176477742912 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 I0425 16:16:31.274513 135176477742912 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.046097s I0425 16:16:31.274686 135176477742912 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.136 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30033087730407715 s) (per-host) I0425 16:16:31.274739 135176477742912 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.300393s (batch_requests_ready=0.252401s, total_serialization_initiated=0.047924s, others=0.000068s) I0425 16:16:31.274833 135176477742912 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.305036s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.305021) I0425 16:16:31.275788 135051404629760 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0425 16:16:31.275944 135176477742912 async_checkpointer.py:561] Finished blocking save. Time taken: 0.582213s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/5. I0425 16:16:31.287046 135176477742912 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0425 16:16:31.287343 135051413022464 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0425 16:16:31.287502 135176477742912 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777133746.577233, 'wait_for_prev_duration_secs': 44.114521980285645, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777133790.693699, 'checkpointer_blocking_duration_secs': 0.5823540687561035, 'get_old_steps_start_time': 1777133791.2760727, 'get_old_steps_duration_secs': 3.7670135498046875e-05, 'checkpoint_manager_blocking_start_time': 1777133746.5750127, 'checkpoint_manager_blocking_duration_secs': 44.712453842163086} I0425 16:16:31.287612 135176477742912 checkpointing.py:409] Started an asynchronous checkpoint save for step 5 I0425 16:16:31.288160 135176477742912 metric_logger.py:185] completed step: 4, seconds: 0.152, TFLOP/s/device: 89.165, Tokens/s/device: 13439.908, total_weights: 65536, loss: 9.460 I0425 16:16:31.293947 135176477742912 metric_logger.py:185] completed step: 5, seconds: 0.148, TFLOP/s/device: 92.071, Tokens/s/device: 13877.971, total_weights: 65536, loss: 8.959 I0425 16:16:50.336367 135176477742912 metric_logger.py:185] completed step: 6, seconds: 45.008, TFLOP/s/device: 0.302, Tokens/s/device: 45.503, total_weights: 65536, loss: 8.469 I0425 16:16:50.483908 135176477742912 metric_logger.py:185] completed step: 7, seconds: 18.897, TFLOP/s/device: 0.719, Tokens/s/device: 108.380, total_weights: 65536, loss: 8.003 I0425 16:16:50.631197 135176477742912 metric_logger.py:185] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.400, Tokens/s/device: 13475.280, total_weights: 65536, loss: 7.572 I0425 16:16:50.778197 135176477742912 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0425 16:16:50.778867 135176477742912 checkpointing.py:798] Waited 0.0006887912750244141 seconds for step 9 to finish before starting checkpointing. I0425 16:16:50.781962 135176477742912 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 16:16:55.646931 135047612962560 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/5/items/array_metadatas/process_6 I0425 16:17:32.781144 135051404629760 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 25.557 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 61.80673122406006 s) (per-host) I0425 16:17:32.781250 135051404629760 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 61.505355s. I0425 16:17:40.024594 135051404629760 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 68.748683s. I0425 16:17:40.024871 135051413022464 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 16:17:40.025003 135051413022464 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0425 16:17:40.025053 135051413022464 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0425 16:17:40.026501 135051413022464 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0425 16:17:40.026688 135176477742912 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0425 16:17:40.026822 135176477742912 checkpoint_manager.py:1458] Waiting for previous save to complete took 49.244870 seconds. If this number is high, consider checkpointing less frequently. I0425 16:17:40.028467 135176477742912 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9 I0425 16:17:40.031943 135176477742912 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/9. I0425 16:17:40.587984 135176477742912 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0425 16:17:40.588079 135176477742912 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 I0425 16:17:40.634057 135176477742912 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047239s I0425 16:17:40.634233 135176477742912 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.150 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29951906204223633 s) (per-host) I0425 16:17:40.634288 135176477742912 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.299583s (batch_requests_ready=0.250265s, total_serialization_initiated=0.049249s, others=0.000069s) I0425 16:17:40.634377 135176477742912 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.303620s (all_items=0.000011s, per_item={'items': '0.00001144'}, temp_paths=0.303608) I0425 16:17:40.635196 135051387844352 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0425 16:17:40.635356 135176477742912 async_checkpointer.py:561] Finished blocking save. Time taken: 0.606820s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/9. I0425 16:17:40.900247 135176477742912 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0425 16:17:40.900592 135051413022464 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0425 16:17:40.900761 135176477742912 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777133810.7819345, 'wait_for_prev_duration_secs': 49.24486994743347, 'time_between_consecutive_saves_sec': 20.090453147888184, 'checkpointer_blocking_start_time': 1777133860.028506, 'checkpointer_blocking_duration_secs': 0.6069602966308594, 'get_old_steps_start_time': 1777133860.6354778, 'get_old_steps_duration_secs': 1.9550323486328125e-05, 'checkpoint_manager_blocking_start_time': 1777133810.7790914, 'checkpoint_manager_blocking_duration_secs': 50.12163543701172} I0425 16:17:40.900867 135176477742912 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0425 16:17:40.900910 135176477742912 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 16:17:45.335434 135051396237056 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/9/items/array_metadatas/process_6 I0425 16:18:22.365806 135051387844352 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.582 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.031049966812134 s) (per-host) I0425 16:18:22.365923 135051387844352 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.730613s. I0425 16:18:31.858868 135051387844352 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 51.223543s. I0425 16:18:31.859090 135051413022464 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 16:18:31.859181 135051413022464 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0425 16:18:31.859224 135051413022464 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0425 16:18:31.862255 135051413022464 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0425 16:18:31.862413 135176477742912 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0425 16:18:31.862557 135176477742912 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 16:18:31.863299 135176477742912 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.121, Tokens/s/device: 13885.499, total_weights: 65536, loss: 7.181 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sat Apr 25 16:18:44 UTC 2026 EXIT_CODE=0