XPK Start: Thu Apr 23 22:21:21 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. `rope_parameters`'s factor field must be a float >= 1, got 40 `rope_parameters`'s beta_fast field must be a float, got 32 `rope_parameters`'s beta_slow field must be a float, got 1 DeepseekV32Config got `key=rope_scaling` in kwargs but hasn't set it as attribute. For RoPE standardization you need to set `self.rope_parameters` in model's config. 2026-04-23 22:21:46.755797: 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) I0423 22:21:46.969737 133003083020096 max_utils.py:273] Attempting to initialize the jax distributed system... I0423 22:21:56.011080 133003083020096 distributed.py:149] Starting JAX distributed service on [::]:8482 I0423 22:21:56.013456 133003083020096 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-false-sv-xqx5f-slice-job-0-0.mt-15-ocdbt-false-sv-xqx5f:8482 I0423 22:21:56.934331 133003083020096 max_utils.py:284] Jax distributed system initialized! I0423 22:22:03.311622 133003083020096 max_utils.py:800] System Information: Jax Version: 0.9.2 I0423 22:22:03.311729 133003083020096 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0423 22:22:03.311771 133003083020096 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 I0423 22:22:03.311805 133003083020096 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 22:22:04.005204 133003083020096 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 22:22:04.005485 133003083020096 checkpointing.py:677] Setting up checkpoint logger... I0423 22:22:04.005538 133003083020096 checkpointing.py:233] Creating checkpoint manager with ocdbt=False and zarr3=True I0423 22:22:04.005583 133003083020096 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0423 22:22:04.005923 133003083020096 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=False, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x78f679d501d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 22:22:06.956947 133003083020096 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0423 22:22:06.957206 133003083020096 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78e1ec4939b0>}, handler_registry=None I0423 22:22:06.957615 133003083020096 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78e1ec4939b0>` 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`. I0423 22:22:06.957763 133003083020096 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78e20ff06630>` 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`. I0423 22:22:06.957804 133003083020096 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 0x78e1ec4939b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78e1ec4939b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78e20ff06630>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78e20ff06630>}). I0423 22:22:06.958158 133003083020096 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0423 22:22:06.958244 133003083020096 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x78e10c7e7560> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0423 22:22:08.336292 133003083020096 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints I0423 22:22:08.338541 133003083020096 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_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78e1ec494260> I0423 22:22:08.338653 133003083020096 checkpointing.py:301] Checkpoint manager created! I0423 22:22:09.270198 133003083020096 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 22:22:09.270314 133003083020096 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 22:22:09.648384 133003083020096 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 22:22:09.648480 133003083020096 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 22:22:09.665252 133003083020096 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 22:22:09.665315 133003083020096 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 22:22:09.689160 133003083020096 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 22:22:09.689241 133003083020096 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 22:22:09.705992 133003083020096 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 22:22:09.706062 133003083020096 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 22:22:09.722900 133003083020096 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 22:22:09.722963 133003083020096 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 22:22:09.748058 133003083020096 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 22:22:09.748145 133003083020096 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 22:22:09.769568 133003083020096 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 22:22:09.769640 133003083020096 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 22:22:09.982324 133003083020096 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0423 22:22:09.982431 133003083020096 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 22:22:11.414066 133003083020096 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0423 22:22:11.414209 133003083020096 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0423 22:22:11.414266 133003083020096 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0423 22:22:11.414324 133003083020096 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0423 22:22:11.414376 133003083020096 maxtext_utils.py:1668] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0423 22:22:11.414413 133003083020096 maxtext_utils.py:1668] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0423 22:22:11.414465 133003083020096 maxtext_utils.py:1668] 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) I0423 22:22:11.414523 133003083020096 maxtext_utils.py:1668] 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') I0423 22:22:11.414563 133003083020096 maxtext_utils.py:1668] 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) I0423 22:22:11.414598 133003083020096 maxtext_utils.py:1668] 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) I0423 22:22:11.414644 133003083020096 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0423 22:22:11.414690 133003083020096 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0423 22:22:11.913740 133003083020096 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 22:22:11.913835 133003083020096 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0423 22:22:11.929623 133003083020096 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 22:22:11.929681 133003083020096 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0423 22:22:22.953184 133003083020096 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. I0423 22:22:22.953992 133003083020096 metric_logger.py:301] number parameters: 1.104 billion I0423 22:22:34.635245 133003083020096 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0423 22:22:34.788888 133003083020096 checkpointing.py:776] Waited 0.1536235809326172 seconds for step 0 to finish before starting checkpointing. I0423 22:22:34.791239 133003083020096 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 22:22:34.792969 133003083020096 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0 I0423 22:22:34.794356 133003083020096 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/0. I0423 22:22:35.127423 133003083020096 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 22:22:35.128273 133003083020096 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0423 22:22:35.128331 133003083020096 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 I0423 22:22:35.404611 133003083020096 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.277265s I0423 22:22:35.404785 133003083020096 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.452 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2829458713531494 s) (per-host) I0423 22:22:35.404837 133003083020096 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283008s (batch_requests_ready=0.002366s, total_serialization_initiated=0.280572s, others=0.000070s) I0423 22:22:35.404932 133003083020096 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287351s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.287332) I0423 22:22:35.405788 133003083020096 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.61 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/0. I0423 22:22:35.406140 132870047790848 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 22:42:35.406086 I0423 22:22:35.417752 133003083020096 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 22:22:35.418052 132872732145408 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 22:22:35.418170 133003083020096 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776982954.7912207, 'wait_for_prev_duration_secs': 6.4849853515625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776982954.7930079, 'checkpointer_blocking_duration_secs': 0.6132955551147461, 'get_old_steps_start_time': 1776982955.406329, 'get_old_steps_duration_secs': 3.3855438232421875e-05, 'checkpoint_manager_blocking_start_time': 1776982954.7894225, 'checkpoint_manager_blocking_duration_secs': 0.6287102699279785} I0423 22:22:35.418335 133003083020096 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0423 22:22:35.418388 133003083020096 max_utils.py:750] Memstats: After params initialized: I0423 22:22:35.418441 133003083020096 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0423 22:22:35.418473 133003083020096 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0423 22:22:35.418501 133003083020096 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0423 22:22:35.418528 133003083020096 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0423 22:22:35.733625 133003083020096 metric_logger.py:196] completed step: 0, seconds: 11.681, TFLOP/s/device: 1.163, Tokens/s/device: 175.325, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0423 22:22:35.903145 133003083020096 metric_logger.py:196] completed step: 1, seconds: 1.097, TFLOP/s/device: 12.388, Tokens/s/device: 1867.272, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0423 22:22:36.297814 133003083020096 metric_logger.py:196] completed step: 2, seconds: 0.022, TFLOP/s/device: 608.225, Tokens/s/device: 91678.231, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719 I0423 22:22:36.445128 133003083020096 metric_logger.py:196] completed step: 3, seconds: 0.395, TFLOP/s/device: 34.404, Tokens/s/device: 5185.729, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506 I0423 22:22:36.449578 133003083020096 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0423 22:22:36.738902 133003083020096 checkpointing.py:776] Waited 0.2892873287200928 seconds for step 5 to finish before starting checkpointing. I0423 22:22:36.741997 133003083020096 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 22:22:38.367327 2813 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 22:22:40.418053 132871069271808 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/0/items/array_metadatas/process_6 I0423 22:22:45.638217 132870047790848 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 150.204 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.516340255737305 s) (per-host) I0423 22:22:45.638348 132870047790848 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.232088s. I0423 22:22:58.457869 132870047790848 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 23.051593s. I0423 22:22:58.458169 132872732145408 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 22:22:58.458305 132872732145408 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 22:22:58.458382 132872732145408 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 22:22:58.460803 132872732145408 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 22:22:58.460969 133003083020096 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0423 22:22:58.461119 133003083020096 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.719106 seconds. If this number is high, consider checkpointing less frequently. I0423 22:22:58.463315 133003083020096 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 5 I0423 22:22:58.465652 133003083020096 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/5. I0423 22:22:58.755912 133003083020096 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0423 22:22:58.756003 133003083020096 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 I0423 22:22:58.792330 133003083020096 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037380s I0423 22:22:58.792497 133003083020096 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 37.529 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04110360145568848 s) (per-host) I0423 22:22:58.792547 133003083020096 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.041163s (batch_requests_ready=0.002050s, total_serialization_initiated=0.039046s, others=0.000066s) I0423 22:22:58.792641 133003083020096 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.046209s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.046193) I0423 22:22:58.793429 133003083020096 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.33 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/5. I0423 22:22:58.793786 132868386006784 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 22:42:58.793754 I0423 22:22:58.807065 133003083020096 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0423 22:22:58.807340 132872732145408 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 22:22:58.807468 133003083020096 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776982956.7419662, 'wait_for_prev_duration_secs': 21.719106435775757, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776982978.4633548, 'checkpointer_blocking_duration_secs': 0.33057427406311035, 'get_old_steps_start_time': 1776982978.793951, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776982956.739528, 'checkpoint_manager_blocking_duration_secs': 22.06790518760681} I0423 22:22:58.807587 133003083020096 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0423 22:22:58.808445 133003083020096 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.847, Tokens/s/device: 13392.011, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355 I0423 22:22:58.814486 133003083020096 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.278, Tokens/s/device: 13909.169, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467 I0423 22:23:04.199307 132849671841536 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/5/items/array_metadatas/process_6 I0423 22:23:09.722829 132868386006784 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 143.974 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.971396446228027 s) (per-host) I0423 22:23:09.722962 132868386006784 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.929059s. I0423 22:23:15.586431 132868386006784 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 16.792512s. I0423 22:23:15.586721 132872732145408 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 22:23:15.586778 132872732145408 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 22:23:15.586824 132872732145408 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0423 22:23:15.588588 132872732145408 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0423 22:23:21.485469 133003083020096 metric_logger.py:196] completed step: 6, seconds: 22.363, TFLOP/s/device: 0.608, Tokens/s/device: 91.578, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740 I0423 22:23:21.632788 133003083020096 metric_logger.py:196] completed step: 7, seconds: 22.525, TFLOP/s/device: 0.603, Tokens/s/device: 90.922, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151 I0423 22:23:21.780070 133003083020096 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.196, Tokens/s/device: 13444.584, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098 I0423 22:23:21.926809 133003083020096 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0423 22:23:21.927541 133003083020096 checkpointing.py:776] Waited 0.0007524490356445312 seconds for step 9 to finish before starting checkpointing. I0423 22:23:21.929526 133003083020096 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 22:23:21.931391 133003083020096 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9 I0423 22:23:21.932677 133003083020096 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/9. I0423 22:23:22.230365 133003083020096 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0423 22:23:22.230455 133003083020096 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 I0423 22:23:22.266376 133003083020096 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.036759s I0423 22:23:22.266545 133003083020096 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 38.180 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.040402889251708984 s) (per-host) I0423 22:23:22.266596 133003083020096 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.040463s (batch_requests_ready=0.001774s, total_serialization_initiated=0.038623s, others=0.000066s) I0423 22:23:22.266693 133003083020096 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.044511s (all_items=0.000011s, per_item={'items': '0.00001144'}, temp_paths=0.044500) I0423 22:23:22.267451 133003083020096 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/9. I0423 22:23:22.267787 132849671841536 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 22:43:22.267758 I0423 22:23:22.269821 133003083020096 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 22:23:22.270080 132872732145408 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 22:23:22.270254 133003083020096 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776983001.929495, 'wait_for_prev_duration_secs': 7.128715515136719e-05, 'time_between_consecutive_saves_sec': 6.340863943099976, 'checkpointer_blocking_start_time': 1776983001.931433, 'checkpointer_blocking_duration_secs': 0.33648085594177246, 'get_old_steps_start_time': 1776983002.267932, 'get_old_steps_duration_secs': 2.5510787963867188e-05, 'checkpoint_manager_blocking_start_time': 1776983001.9277759, 'checkpoint_manager_blocking_duration_secs': 0.34244585037231445} I0423 22:23:22.270374 133003083020096 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0423 22:23:22.270422 133003083020096 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 22:23:28.141836 132851299239680 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260423_211617/linen_xpk_test_pipeline_scan_nnx_20260423_211617_15_ocdbt_false/checkpoints/9/items/array_metadatas/process_6 I0423 22:23:33.738676 132849671841536 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 137.207 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.5124990940094 s) (per-host) I0423 22:23:33.738794 132849671841536 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 11.470907s. I0423 22:23:39.301489 132849671841536 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 17.033579s. I0423 22:23:39.301791 132872732145408 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 22:23:39.301912 132872732145408 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 22:23:39.301955 132872732145408 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 22:23:39.303317 132872732145408 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 22:23:39.303490 133003083020096 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0423 22:23:39.303644 133003083020096 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 22:23:39.304656 133003083020096 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.216, Tokens/s/device: 13899.729, total_weights: 65536, loss: 7.181, lm_loss: 7.181, perplexity: 1314.495 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Thu Apr 23 22:23:51 UTC 2026 EXIT_CODE=0