XPK Start: Sun Apr 19 18:45:36 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-19 18:46:11.804050: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0419 18:46:12.334956 138208051050304 max_utils.py:273] Attempting to initialize the jax distributed system... I0419 18:46:21.375836 138208051050304 distributed.py:149] Starting JAX distributed service on [::]:8482 I0419 18:46:21.378349 138208051050304 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-true-sgwg7-slice-job-0-0.mt-10-shardy-true-sgwg7:8482 I0419 18:46:29.473897 138208051050304 max_utils.py:284] Jax distributed system initialized! I0419 18:46:35.761136 138208051050304 max_utils.py:800] System Information: Jax Version: 0.9.2 I0419 18:46:35.761245 138208051050304 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0419 18:46:35.761284 138208051050304 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 I0419 18:46:35.761318 138208051050304 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 18:46:36.454037 138208051050304 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 18:46:36.454315 138208051050304 checkpointing.py:677] Setting up checkpoint logger... I0419 18:46:36.454381 138208051050304 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 18:46:36.454428 138208051050304 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0419 18:46:36.454744 138208051050304 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 0x7db282d18140>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 18:46:39.615736 138208051050304 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0419 18:46:39.615978 138208051050304 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d9e380ebb90>}, handler_registry=None I0419 18:46:39.616246 138208051050304 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d9e380ebb90>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0419 18:46:39.616298 138208051050304 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d9e380ef920>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0419 18:46:39.616344 138208051050304 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 0x7d9e380ebb90>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d9e380ebb90>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d9e380ef920>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d9e380ef920>}). I0419 18:46:39.616694 138208051050304 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0419 18:46:39.616777 138208051050304 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d9e084eb880> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0419 18:46:40.600214 138208051050304 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints I0419 18:46:41.064049 138208051050304 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d9e380edbb0> I0419 18:46:41.064213 138208051050304 checkpointing.py:301] Checkpoint manager created! I0419 18:46:41.962221 138208051050304 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 18:46:41.962332 138208051050304 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 18:46:42.340074 138208051050304 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 18:46:42.340165 138208051050304 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 18:46:42.356409 138208051050304 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 18:46:42.356468 138208051050304 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 18:46:42.379922 138208051050304 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:46:42.379987 138208051050304 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:46:42.396188 138208051050304 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:46:42.396245 138208051050304 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:46:42.412636 138208051050304 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:46:42.412693 138208051050304 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:46:42.437844 138208051050304 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 18:46:42.437916 138208051050304 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:46:42.465378 138208051050304 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 18:46:42.465474 138208051050304 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 18:46:42.681917 138208051050304 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0419 18:46:42.682021 138208051050304 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0419 18:46:44.105644 138208051050304 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0419 18:46:44.105795 138208051050304 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) I0419 18:46:44.105867 138208051050304 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) I0419 18:46:44.105943 138208051050304 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') I0419 18:46:44.106024 138208051050304 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) I0419 18:46:44.106090 138208051050304 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) I0419 18:46:44.106167 138208051050304 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) I0419 18:46:44.106248 138208051050304 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') I0419 18:46:44.106304 138208051050304 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) I0419 18:46:44.106361 138208051050304 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) I0419 18:46:44.106532 138208051050304 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0419 18:46:44.106612 138208051050304 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0419 18:46:44.588236 138208051050304 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 18:46:44.588326 138208051050304 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0419 18:46:44.603491 138208051050304 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 18:46:44.603550 138208051050304 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0419 18:46:55.450826 138208051050304 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. I0419 18:46:55.458792 138208051050304 metric_logger.py:301] number parameters: 1.104 billion I0419 18:47:06.758355 138208051050304 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0419 18:47:06.912579 138208051050304 checkpointing.py:776] Waited 0.1542055606842041 seconds for step 0 to finish before starting checkpointing. I0419 18:47:06.915327 138208051050304 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 18:47:06.917164 138208051050304 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0 I0419 18:47:06.918535 138208051050304 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0. I0419 18:47:07.632148 138208051050304 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 18:47:07.646444 138208051050304 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0419 18:47:07.646509 138208051050304 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0419 18:47:07.900162 138079100712704 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0 I0419 18:47:07.915785 138208051050304 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.270161s I0419 18:47:07.916644 138208051050304 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.576 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27663540840148926 s) (per-host) I0419 18:47:07.916719 138208051050304 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.276729s (batch_requests_ready=0.002214s, total_serialization_initiated=0.273709s, others=0.000805s) I0419 18:47:07.916799 138208051050304 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.280661s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.280643) I0419 18:47:07.917583 138208051050304 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.00 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0. I0419 18:47:07.917912 138079583954688 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 19:07:07.917876 I0419 18:47:07.937323 138208051050304 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 18:47:07.937670 138077973124864 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 18:47:07.937828 138208051050304 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776624426.915308, 'wait_for_prev_duration_secs': 7.653236389160156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776624426.9172018, 'checkpointer_blocking_duration_secs': 1.000854730606079, 'get_old_steps_start_time': 1776624427.9180813, 'get_old_steps_duration_secs': 2.6941299438476562e-05, 'checkpoint_manager_blocking_start_time': 1776624426.9130723, 'checkpoint_manager_blocking_duration_secs': 1.024716854095459} I0419 18:47:07.937931 138208051050304 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0419 18:47:07.937982 138208051050304 max_utils.py:750] Memstats: After params initialized: I0419 18:47:07.938029 138208051050304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0419 18:47:07.938061 138208051050304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0419 18:47:07.938087 138208051050304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0419 18:47:07.938109 138208051050304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0419 18:47:08.274029 138208051050304 metric_logger.py:196] completed step: 0, seconds: 11.005, TFLOP/s/device: 1.235, Tokens/s/device: 186.096, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0419 18:47:08.277469 138208051050304 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/tensorboard/' I0419 18:47:08.685871 138208051050304 metric_logger.py:196] completed step: 1, seconds: 1.514, TFLOP/s/device: 8.975, Tokens/s/device: 1352.828, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0419 18:47:08.831951 138208051050304 metric_logger.py:196] completed step: 2, seconds: 0.413, TFLOP/s/device: 32.936, Tokens/s/device: 4964.427, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0419 18:47:08.979132 138208051050304 metric_logger.py:196] completed step: 3, seconds: 0.009, TFLOP/s/device: 1467.610, Tokens/s/device: 221214.085, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203 I0419 18:47:11.165350 138076916172544 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776624430756867160, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776624430756867160, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/_CHECKPOINT_METADATA I0419 18:47:11.452840 138079071352576 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/items I0419 18:47:13.030648 2844 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 18:47:15.217403 138078023325440 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/items/array_metadatas/process_0 I0419 18:47:31.889558 138208051050304 metric_logger.py:196] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.527, Tokens/s/device: 13946.678, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614 I0419 18:47:31.898001 138208051050304 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.374, Tokens/s/device: 13923.543, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717 I0419 18:47:32.044133 138208051050304 metric_logger.py:196] completed step: 6, seconds: 22.909, TFLOP/s/device: 0.593, Tokens/s/device: 89.397, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759 I0419 18:47:32.191542 138208051050304 metric_logger.py:196] completed step: 7, seconds: 0.007, TFLOP/s/device: 2070.893, Tokens/s/device: 312147.538, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905 I0419 18:47:32.339073 138208051050304 metric_logger.py:196] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.363, Tokens/s/device: 13921.934, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406 I0419 18:47:32.485748 138208051050304 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0419 18:47:32.486496 138208051050304 checkpointing.py:776] Waited 0.0007677078247070312 seconds for step 9 to finish before starting checkpointing. I0419 18:47:32.488517 138208051050304 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 18:47:50.903251 138078014932736 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.871639s (commit=35.937102s, array_metadata_write=1.934537s) I0419 18:47:50.904415 138079583954688 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 36.510 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.26435923576355 s) (per-host) I0419 18:47:50.904549 138079583954688 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.986526s. I0419 18:47:51.574604 138079583954688 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776624430756867160, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/_CHECKPOINT_METADATA I0419 18:47:52.759002 138076916172544 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776624430756867160, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/_CHECKPOINT_METADATA I0419 18:47:53.086045 138079583954688 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 18:47:53.114648 138079583954688 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000277s. I0419 18:47:54.027521 138079583954688 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.332347s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/items I0419 18:47:54.029176 138079583954688 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0/items I0419 18:47:54.979246 138079583954688 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0 I0419 18:47:56.414289 138079583954688 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0`. I0419 18:47:56.415186 138079583954688 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 49.50 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/0 I0419 18:47:56.416407 138079583954688 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 48.498383s. I0419 18:47:56.416616 138077973124864 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 18:47:56.416732 138077973124864 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 18:47:56.416803 138077973124864 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 18:47:56.418896 138077973124864 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 18:47:56.419017 138208051050304 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 18:47:56.419086 138208051050304 checkpoint_manager.py:1452] Waiting for previous save to complete took 23.930576 seconds. If this number is high, consider checkpointing less frequently. I0419 18:47:56.421168 138208051050304 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9 I0419 18:47:56.423209 138208051050304 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9. I0419 18:47:57.139356 138208051050304 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0419 18:47:57.139462 138208051050304 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0419 18:47:57.166493 138208051050304 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027802s I0419 18:47:57.167217 138208051050304 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 47.472 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.032494306564331055 s) (per-host) I0419 18:47:57.167281 138208051050304 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.032577s (batch_requests_ready=0.001817s, total_serialization_initiated=0.030087s, others=0.000673s) I0419 18:47:57.167351 138208051050304 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.036431s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.036414) I0419 18:47:57.168022 138208051050304 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9. I0419 18:47:57.168308 138077973124864 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 19:07:57.168276 I0419 18:47:57.183033 138208051050304 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 18:47:57.183309 138078526088960 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 18:47:57.183481 138208051050304 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776624452.488484, 'wait_for_prev_duration_secs': 23.93057632446289, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776624476.4212108, 'checkpointer_blocking_duration_secs': 0.7472436428070068, 'get_old_steps_start_time': 1776624477.168478, 'get_old_steps_duration_secs': 3.075599670410156e-05, 'checkpoint_manager_blocking_start_time': 1776624452.4867394, 'checkpoint_manager_blocking_duration_secs': 24.69670271873474} I0419 18:47:57.183616 138208051050304 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0419 18:47:57.183665 138208051050304 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 18:47:57.383852 138079583954688 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9 I0419 18:47:59.584295 138078559659776 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9/items I0419 18:48:04.657053 138078551267072 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9/items/array_metadatas/process_0 I0419 18:48:44.104170 138078542874368 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.943521s (commit=40.875808s, array_metadata_write=2.067713s) I0419 18:48:44.105245 138077973124864 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 33.630 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.97049617767334 s) (per-host) I0419 18:48:44.105383 138077973124864 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.936939s. I0419 18:48:46.844965 138077973124864 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 18:48:46.933991 138077973124864 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000203s. I0419 18:48:47.946257 138077973124864 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.080621s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9/items I0419 18:48:47.951905 138077973124864 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9/items I0419 18:48:48.491814 138077973124864 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9 I0419 18:48:50.327277 138077973124864 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9`. I0419 18:48:50.327967 138077973124864 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.91 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_10_shardy_true/checkpoints/9 I0419 18:48:50.329308 138077973124864 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 53.160877s. I0419 18:48:50.329514 138078526088960 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 18:48:50.329634 138078526088960 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 18:48:50.329701 138078526088960 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 18:48:50.332329 138078526088960 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 18:48:50.332502 138208051050304 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 18:48:50.332647 138208051050304 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 18:48:50.333719 138208051050304 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.172, Tokens/s/device: 13893.223, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 18:49:01 UTC 2026 EXIT_CODE=0