XPK Start: Sat Apr 18 13:44:11 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-18 13:44:36.374378: 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) I0418 13:44:36.554501 137490195486528 max_utils.py:273] Attempting to initialize the jax distributed system... I0418 13:44:45.595274 137490195486528 distributed.py:149] Starting JAX distributed service on [::]:8482 I0418 13:44:45.597677 137490195486528 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-tihd8-slice-job-0-0.mt-11-optimizer-offload-tihd8:8482 I0418 13:44:46.518724 137490195486528 max_utils.py:284] Jax distributed system initialized! I0418 13:44:51.670285 137490195486528 max_utils.py:800] System Information: Jax Version: 0.9.2 I0418 13:44:51.670384 137490195486528 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0418 13:44:51.670423 137490195486528 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365 I0418 13:44:51.670475 137490195486528 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0418 13:44:52.360352 137490195486528 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0418 13:44:52.360638 137490195486528 checkpointing.py:677] Setting up checkpoint logger... I0418 13:44:52.360692 137490195486528 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0418 13:44:52.360738 137490195486528 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0418 13:44:52.361080 137490195486528 base_pytree_checkpoint_handler.py:441] 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 0x7d0b52d1d190>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0418 13:44:55.628249 137490195486528 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0418 13:44:55.628502 137490195486528 checkpoint_manager.py:708] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cf6c843c260>}, handler_registry=None I0418 13:44:55.628739 137490195486528 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cf6c843c260>` 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`. I0418 13:44:55.628787 137490195486528 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cf6c84e33b0>` 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`. I0418 13:44:55.628822 137490195486528 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 0x7cf6c843c260>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cf6c843c260>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cf6c84e33b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cf6c84e33b0>}). I0418 13:44:55.629139 137490195486528 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0418 13:44:55.629208 137490195486528 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7cf6a47dbc40> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0418 13:44:56.400727 137490195486528 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints I0418 13:44:57.259381 137490195486528 checkpoint_manager.py:929] [process=5][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_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7cf6c84e2ba0> I0418 13:44:57.259583 137490195486528 checkpointing.py:301] Checkpoint manager created! I0418 13:44:58.177853 137490195486528 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0418 13:44:58.177968 137490195486528 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 13:44:58.555690 137490195486528 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 13:44:58.555783 137490195486528 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 13:44:58.572089 137490195486528 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 13:44:58.572148 137490195486528 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 13:44:58.595498 137490195486528 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0418 13:44:58.595566 137490195486528 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 13:44:58.611839 137490195486528 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0418 13:44:58.611896 137490195486528 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 13:44:58.628062 137490195486528 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0418 13:44:58.628121 137490195486528 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 13:44:58.652446 137490195486528 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0418 13:44:58.652521 137490195486528 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 13:44:58.673492 137490195486528 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0418 13:44:58.673557 137490195486528 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0418 13:44:58.889905 137490195486528 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0418 13:44:58.890010 137490195486528 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0418 13:45:00.456702 137490195486528 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0418 13:45:00.456856 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0418 13:45:00.456927 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0418 13:45:00.457007 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0418 13:45:00.457085 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0418 13:45:00.457144 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0418 13:45:00.457227 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0418 13:45:00.457305 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: P('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0418 13:45:00.457364 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0418 13:45:00.457416 137490195486528 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0418 13:45:00.457542 137490195486528 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0418 13:45:00.457633 137490195486528 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0418 13:45:00.946256 137490195486528 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 13:45:00.946354 137490195486528 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0418 13:45:00.961576 137490195486528 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 13:45:00.961639 137490195486528 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0418 13:45:11.707351 137490195486528 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB. I0418 13:45:11.708178 137490195486528 metric_logger.py:301] number parameters: 1.104 billion I0418 13:45:22.864076 137490195486528 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0418 13:45:23.040858 137490195486528 checkpointing.py:776] Waited 0.17676448822021484 seconds for step 0 to finish before starting checkpointing. I0418 13:45:23.043329 137490195486528 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 13:45:23.044944 137490195486528 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 0 I0418 13:45:23.046434 137490195486528 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints/0. I0418 13:45:23.809771 137490195486528 signaling_client.py:364] Using JaxDistributedSignalingClient I0418 13:45:23.810997 137490195486528 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 13:45:23.811052 137490195486528 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 I0418 13:45:24.089573 137490195486528 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279884s I0418 13:45:24.089742 137490195486528 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.391 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2861356735229492 s) (per-host) I0418 13:45:24.089795 137490195486528 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.286198s (batch_requests_ready=0.002325s, total_serialization_initiated=0.283802s, others=0.000071s) I0418 13:45:24.089881 137490195486528 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.290473s (all_items=0.000017s, per_item={'items': '0.00001717'}, temp_paths=0.290456) I0418 13:45:24.090652 137490195486528 event_tracking.py:125] [process=5] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints/0. I0418 13:45:24.091000 137357693941504 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:05:24.090962 I0418 13:45:24.093017 137490195486528 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0418 13:45:24.093289 137357427787520 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0418 13:45:24.093438 137490195486528 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776519923.043311, 'wait_for_prev_duration_secs': 5.9604644775390625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776519923.0449877, 'checkpointer_blocking_duration_secs': 1.04616117477417, 'get_old_steps_start_time': 1776519924.091173, 'get_old_steps_duration_secs': 3.075599670410156e-05, 'checkpoint_manager_blocking_start_time': 1776519923.041404, 'checkpoint_manager_blocking_duration_secs': 1.0519964694976807} I0418 13:45:24.093556 137490195486528 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0418 13:45:24.093605 137490195486528 max_utils.py:750] Memstats: After params initialized: I0418 13:45:24.093652 137490195486528 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_18(process=5,(2,4,0,0)) I0418 13:45:24.093682 137490195486528 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_19(process=5,(3,4,0,0)) I0418 13:45:24.093708 137490195486528 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_22(process=5,(2,5,0,0)) I0418 13:45:24.093733 137490195486528 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_23(process=5,(3,5,0,0)) I0418 13:45:24.403969 137490195486528 metric_logger.py:196] completed step: 0, seconds: 11.156, TFLOP/s/device: 1.218, Tokens/s/device: 183.582, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0418 13:45:24.600712 137490195486528 metric_logger.py:196] completed step: 1, seconds: 1.538, TFLOP/s/device: 8.832, Tokens/s/device: 1331.241, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0418 13:45:25.020648 137490195486528 metric_logger.py:196] completed step: 2, seconds: 0.034, TFLOP/s/device: 398.590, Tokens/s/device: 60079.793, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0418 13:45:25.183648 137490195486528 metric_logger.py:196] completed step: 3, seconds: 0.420, TFLOP/s/device: 32.364, Tokens/s/device: 4878.293, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.047 I0418 13:45:25.509792 137490195486528 metric_logger.py:196] completed step: 4, seconds: 0.169, TFLOP/s/device: 80.550, Tokens/s/device: 12141.405, total_weights: 65536, loss: 9.278, lm_loss: 9.278, perplexity: 10694.968 I0418 13:45:25.515607 137490195486528 metric_logger.py:196] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.405, Tokens/s/device: 12571.744, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.520 I0418 13:45:27.482867 2734 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0418 13:45:29.842488 137357436180224 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_5 I0418 13:45:48.926148 137490195486528 metric_logger.py:196] completed step: 6, seconds: 0.327, TFLOP/s/device: 41.597, Tokens/s/device: 6269.938, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.688 I0418 13:45:49.088962 137490195486528 metric_logger.py:196] completed step: 7, seconds: 23.248, TFLOP/s/device: 0.584, Tokens/s/device: 88.092, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4377.010 I0418 13:45:49.251711 137490195486528 metric_logger.py:196] completed step: 8, seconds: 0.168, TFLOP/s/device: 80.874, Tokens/s/device: 12190.259, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3845.943 I0418 13:45:49.414031 137490195486528 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0418 13:45:49.414710 137490195486528 checkpointing.py:776] Waited 0.0007052421569824219 seconds for step 9 to finish before starting checkpointing. I0418 13:45:49.416719 137490195486528 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 13:46:02.139306 137357693941504 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 41.203 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.33566212654114 s) (per-host) I0418 13:46:02.139427 137357693941504 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.048317s. I0418 13:46:10.760777 137357693941504 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 46.669651s. I0418 13:46:10.761041 137357427787520 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 13:46:10.761159 137357427787520 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0418 13:46:10.761210 137357427787520 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0418 13:46:10.763140 137357427787520 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0418 13:46:10.763316 137490195486528 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0418 13:46:10.763445 137490195486528 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.346728 seconds. If this number is high, consider checkpointing less frequently. I0418 13:46:10.765259 137490195486528 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 9 I0418 13:46:10.767443 137490195486528 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints/9. I0418 13:46:11.482010 137490195486528 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 13:46:11.482106 137490195486528 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 I0418 13:46:11.523898 137490195486528 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.042856s I0418 13:46:11.524057 137490195486528 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 33.329 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04628109931945801 s) (per-host) I0418 13:46:11.524106 137490195486528 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.046340s (batch_requests_ready=0.001922s, total_serialization_initiated=0.044351s, others=0.000068s) I0418 13:46:11.524177 137490195486528 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.050504s (all_items=0.000022s, per_item={'items': '0.00002170'}, temp_paths=0.050482) I0418 13:46:11.524884 137490195486528 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints/9. I0418 13:46:11.525235 137357427787520 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:06:11.525198 I0418 13:46:11.527181 137490195486528 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0418 13:46:11.527415 137335723185920 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0418 13:46:11.527558 137490195486528 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776519949.4166884, 'wait_for_prev_duration_secs': 21.346728086471558, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776519970.7653098, 'checkpointer_blocking_duration_secs': 0.760073184967041, 'get_old_steps_start_time': 1776519971.5254056, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776519949.414956, 'checkpoint_manager_blocking_duration_secs': 22.112570762634277} I0418 13:46:11.527727 137490195486528 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0418 13:46:11.527773 137490195486528 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 13:46:18.669691 137357436180224 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_5 I0418 13:46:53.467314 137357427787520 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 37.617 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.989498138427734 s) (per-host) I0418 13:46:53.467438 137357427787520 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.942091s. I0418 13:47:03.150671 137357427787520 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 51.625308s. I0418 13:47:03.150957 137335723185920 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 13:47:03.151072 137335723185920 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0418 13:47:03.151120 137335723185920 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0418 13:47:03.153653 137335723185920 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0418 13:47:03.153833 137490195486528 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0418 13:47:03.153988 137490195486528 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 13:47:03.154941 137490195486528 metric_logger.py:196] completed step: 9, seconds: 0.163, TFLOP/s/device: 83.473, Tokens/s/device: 12582.017, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.739 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sat Apr 18 13:47:12 UTC 2026 EXIT_CODE=0