XPK Start: Mon Apr 20 22:33:57 UTC 2026 2026-04-20 22:34:01.249835: 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 00:00:1776724441.262527 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776724441.266214 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776724441.278049 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776724441.278080 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776724441.278083 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776724441.278085 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-20 22:34:36.052957: 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) I0420 22:34:36.572578 133693649635136 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-20 22:34:45,617:jax._src.distributed:157: Connecting to JAX distributed service on mt-01-base-whlsw-slice-job-0-0.mt-01-base-whlsw:8482 I0420 22:34:45.617171 133693649635136 distributed.py:157] Connecting to JAX distributed service on mt-01-base-whlsw-slice-job-0-0.mt-01-base-whlsw:8482 I0420 22:34:55.479961 133693649635136 max_utils.py:284] Jax distributed system initialized! I0420 22:35:03.280356 133693649635136 max_utils.py:800] System Information: Jax Version: 0.8.1 I0420 22:35:03.280469 133693649635136 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0420 22:35:03.280511 133693649635136 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 I0420 22:35:03.280543 133693649635136 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 22:35:03.898241 133693649635136 maxtext_utils.py:1517] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 22:35:03.898524 133693649635136 checkpointing.py:677] Setting up checkpoint logger... I0420 22:35:03.898579 133693649635136 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0420 22:35:03.898625 133693649635136 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0420 22:35:03.899054 133693649635136 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 0x799780970f20>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 22:35:07.548583 133693649635136 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0420 22:35:07.548947 133693649635136 checkpoint_manager.py:709] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x798470293e00>}, handler_registry=None I0420 22:35:07.549189 133693649635136 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x798470293e00>` 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`. I0420 22:35:07.549239 133693649635136 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79847028eb40>` 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`. I0420 22:35:07.549275 133693649635136 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 0x798470293e00>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x798470293e00>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79847028eb40>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79847028eb40>}). I0420 22:35:07.549800 133693649635136 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0420 22:35:07.549878 133693649635136 async_checkpointer.py:177] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x798458352840> timeout: 600 secs and primary_host=0 for async checkpoint writes I0420 22:35:09.072267 133693649635136 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints I0420 22:35:09.921834 133693649635136 checkpoint_manager.py:929] [process=7][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_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x798470290e90> I0420 22:35:09.922008 133693649635136 checkpointing.py:301] Checkpoint manager created! I0420 22:35:10.816046 133693649635136 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 22:35:10.816153 133693649635136 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:35:11.183932 133693649635136 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 22:35:11.184025 133693649635136 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:35:11.199521 133693649635136 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 22:35:11.199579 133693649635136 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:35:11.222819 133693649635136 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 22:35:11.222889 133693649635136 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:35:11.238524 133693649635136 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 22:35:11.238588 133693649635136 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:35:11.254082 133693649635136 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 22:35:11.254142 133693649635136 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:35:11.278025 133693649635136 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 22:35:11.278092 133693649635136 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:35:11.297306 133693649635136 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 22:35:11.297371 133693649635136 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 22:35:11.487027 133693649635136 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0420 22:35:11.487136 133693649635136 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0420 22:35:12.903721 133693649635136 maxtext_utils.py:1620] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0420 22:35:12.903878 133693649635136 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0420 22:35:12.903934 133693649635136 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0420 22:35:12.904003 133693649635136 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0420 22:35:12.904058 133693649635136 maxtext_utils.py:1620] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0420 22:35:12.904097 133693649635136 maxtext_utils.py:1620] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0420 22:35:12.904147 133693649635136 maxtext_utils.py:1620] 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) I0420 22:35:12.904198 133693649635136 maxtext_utils.py:1620] 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') I0420 22:35:12.904236 133693649635136 maxtext_utils.py:1620] 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) I0420 22:35:12.904270 133693649635136 maxtext_utils.py:1620] 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) I0420 22:35:12.904316 133693649635136 maxtext_utils.py:1620] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0420 22:35:12.904361 133693649635136 maxtext_utils.py:1620] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0420 22:35:13.397938 133693649635136 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0420 22:35:13.398051 133693649635136 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0420 22:35:13.412701 133693649635136 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0420 22:35:13.412780 133693649635136 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0420 22:35:23.889244 133693649635136 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. I0420 22:35:23.890068 133693649635136 metric_logger.py:289] number parameters: 1.104 billion I0420 22:35:34.642882 133693649635136 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0420 22:35:35.254597 133693649635136 checkpointing.py:776] Waited 0.6116962432861328 seconds for step 0 to finish before starting checkpointing. I0420 22:35:35.257219 133693649635136 checkpoint_manager.py:2013] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 22:35:35.258899 133693649635136 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 0 I0420 22:35:35.261821 133693649635136 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints/0. I0420 22:35:35.810438 133693649635136 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 22:35:35.811551 133693649635136 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0420 22:35:35.811610 133693649635136 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 I0420 22:35:36.056270 133693649635136 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.245910s I0420 22:35:36.056438 133693649635136 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.080 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5008304119110107 s) (per-host) I0420 22:35:36.056493 133693649635136 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.500895s (batch_requests_ready=0.251433s, total_serialization_initiated=0.249389s, others=0.000072s) I0420 22:35:36.056589 133693649635136 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.505006s (all_items=0.000020s, per_item={'items': '0.00002027'}, temp_paths=0.504986) I0420 22:35:36.057520 133575932360448 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started. I0420 22:35:36.057691 133693649635136 async_checkpointer.py:561] Finished blocking save. Time taken: 0.798724s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints/0. I0420 22:35:36.505769 133693649635136 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0420 22:35:36.506172 133557703927552 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0420 22:35:36.506344 133693649635136 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776724535.2572021, 'wait_for_prev_duration_secs': 6.175041198730469e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776724535.2589383, 'checkpointer_blocking_duration_secs': 0.7988684177398682, 'get_old_steps_start_time': 1776724536.0578356, 'get_old_steps_duration_secs': 2.2411346435546875e-05, 'checkpoint_manager_blocking_start_time': 1776724535.2552543, 'checkpoint_manager_blocking_duration_secs': 1.251047134399414} I0420 22:35:36.506453 133693649635136 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0420 22:35:36.506503 133693649635136 max_utils.py:750] Memstats: After params initialized: I0420 22:35:36.506561 133693649635136 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0)) I0420 22:35:36.506594 133693649635136 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0)) I0420 22:35:36.506622 133693649635136 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0)) I0420 22:35:36.506647 133693649635136 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0)) I0420 22:35:36.765977 133693649635136 metric_logger.py:185] completed step: 0, seconds: 10.753, TFLOP/s/device: 1.264, Tokens/s/device: 190.464, total_weights: 65536, loss: 10.874 I0420 22:35:36.949577 133693649635136 metric_logger.py:185] completed step: 1, seconds: 2.122, TFLOP/s/device: 6.404, Tokens/s/device: 965.212, total_weights: 65536, loss: 10.874 I0420 22:35:37.331932 133693649635136 metric_logger.py:185] completed step: 2, seconds: 0.036, TFLOP/s/device: 375.366, Tokens/s/device: 56579.275, total_weights: 65536, loss: 10.263 I0420 22:35:37.479282 133693649635136 metric_logger.py:185] completed step: 3, seconds: 0.383, TFLOP/s/device: 35.483, Tokens/s/device: 5348.348, total_weights: 65536, loss: 9.734 I0420 22:35:37.774557 133693649635136 metric_logger.py:185] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.890, Tokens/s/device: 13398.494, total_weights: 65536, loss: 9.278 I0420 22:35:37.780227 133693649635136 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.374, Tokens/s/device: 13923.638, total_weights: 65536, loss: 8.892 I0420 22:35:57.756424 133693649635136 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.953, Tokens/s/device: 6926.594, total_weights: 65536, loss: 8.592 I0420 22:35:57.903921 133693649635136 metric_logger.py:185] completed step: 7, seconds: 19.829, TFLOP/s/device: 0.685, Tokens/s/device: 103.281, total_weights: 65536, loss: 8.384 I0420 22:35:58.051193 133693649635136 metric_logger.py:185] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.202, Tokens/s/device: 13445.555, total_weights: 65536, loss: 8.255 I0420 22:35:58.198223 133693649635136 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0420 22:35:58.198858 133693649635136 checkpointing.py:776] Waited 0.0006539821624755859 seconds for step 9 to finish before starting checkpointing. I0420 22:35:58.200864 133693649635136 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 22:35:59.658769 2359 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 22:36:01.943558 133575940753152 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints/0/items/array_metadatas/process_7 I0420 22:36:33.525269 133575932360448 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 27.249 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 57.96961855888367 s) (per-host) I0420 22:36:33.525393 133575932360448 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 57.467761s. I0420 22:36:40.801023 133575932360448 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 64.743375s. I0420 22:36:40.801336 133557703927552 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 22:36:40.801460 133557703927552 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0420 22:36:40.801511 133557703927552 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0420 22:36:40.802802 133557703927552 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0420 22:36:40.802989 133693649635136 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0420 22:36:40.803133 133693649635136 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.602275 seconds. If this number is high, consider checkpointing less frequently. I0420 22:36:40.806349 133693649635136 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 9 I0420 22:36:40.809793 133693649635136 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints/9. I0420 22:36:41.804063 133693649635136 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0420 22:36:41.804163 133693649635136 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 I0420 22:36:41.848585 133693649635136 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.046073s I0420 22:36:41.848766 133693649635136 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.315 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29021167755126953 s) (per-host) I0420 22:36:41.848824 133693649635136 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.290280s (batch_requests_ready=0.242324s, total_serialization_initiated=0.047870s, others=0.000086s) I0420 22:36:41.848918 133693649635136 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.294540s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.294524) I0420 22:36:41.849822 133560607237888 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started. I0420 22:36:41.849989 133693649635136 async_checkpointer.py:561] Finished blocking save. Time taken: 1.043564s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints/9. I0420 22:36:42.299571 133693649635136 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0420 22:36:42.299960 133557703927552 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0420 22:36:42.300142 133693649635136 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776724558.200836, 'wait_for_prev_duration_secs': 42.602275133132935, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776724600.8063903, 'checkpointer_blocking_duration_secs': 1.043710708618164, 'get_old_steps_start_time': 1776724601.850122, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776724558.1990793, 'checkpoint_manager_blocking_duration_secs': 44.10102677345276} I0420 22:36:42.300255 133693649635136 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0420 22:36:42.300301 133693649635136 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 22:36:47.988145 133575940753152 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260420_223221_01_base/checkpoints/9/items/array_metadatas/process_7 I0420 22:37:24.878964 133560607237888 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 36.463 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.32037663459778 s) (per-host) I0420 22:37:24.879072 133560607237888 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.029135s. I0420 22:37:33.937241 133560607237888 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 52.087291s. I0420 22:37:33.937520 133557703927552 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 22:37:33.937639 133557703927552 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0420 22:37:33.937689 133557703927552 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0420 22:37:33.939038 133557703927552 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0420 22:37:33.939234 133693649635136 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0420 22:37:33.939381 133693649635136 checkpoint_manager.py:2013] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 22:37:33.940068 133693649635136 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.079, Tokens/s/device: 13879.194, total_weights: 65536, loss: 8.179 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Mon Apr 20 22:37:47 UTC 2026 EXIT_CODE=0