XPK Start: Fri Apr 24 21:13:04 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. 2026-04-24 21:13:29.792299: 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) I0424 21:13:30.005127 132553424877376 max_utils.py:273] Attempting to initialize the jax distributed system... I0424 21:13:39.046865 132553424877376 distributed.py:149] Starting JAX distributed service on [::]:8482 I0424 21:13:39.049197 132553424877376 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-false-sv-lokcq-slice-job-0-0.mt-15-ocdbt-false-sv-lokcq:8482 I0424 21:13:39.913409 132553424877376 max_utils.py:284] Jax distributed system initialized! I0424 21:13:45.787986 132553424877376 max_utils.py:800] System Information: Jax Version: 0.9.2 I0424 21:13:45.788089 132553424877376 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0424 21:13:45.788148 132553424877376 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 I0424 21:13:45.788183 132553424877376 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0424 21:13:46.480945 132553424877376 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0424 21:13:46.481238 132553424877376 checkpointing.py:677] Setting up checkpoint logger... I0424 21:13:46.481293 132553424877376 checkpointing.py:233] Creating checkpoint manager with ocdbt=False and zarr3=True I0424 21:13:46.481338 132553424877376 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0424 21:13:46.481679 132553424877376 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 0x788de971c770>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0424 21:13:49.346133 132553424877376 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0424 21:13:49.346375 132553424877376 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 0x787a305378f0>}, handler_registry=None I0424 21:13:49.346687 132553424877376 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x787a305378f0>` 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`. I0424 21:13:49.346741 132553424877376 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x787a30539cd0>` 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`. I0424 21:13:49.346779 132553424877376 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 0x787a305378f0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x787a305378f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x787a30539cd0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x787a30539cd0>}). I0424 21:13:49.347108 132553424877376 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0424 21:13:49.347180 132553424877376 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x787a3038b2e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0424 21:13:50.095814 132553424877376 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints I0424 21:13:50.120355 132553424877376 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_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x787a30536b70> I0424 21:13:50.120487 132553424877376 checkpointing.py:301] Checkpoint manager created! I0424 21:13:51.060904 132553424877376 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 21:13:51.061024 132553424877376 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 21:13:51.449376 132553424877376 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 21:13:51.449466 132553424877376 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 21:13:51.466115 132553424877376 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 21:13:51.466173 132553424877376 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 21:13:51.490043 132553424877376 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 21:13:51.490126 132553424877376 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 21:13:51.507059 132553424877376 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 21:13:51.507134 132553424877376 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 21:13:51.523843 132553424877376 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 21:13:51.523907 132553424877376 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 21:13:51.549254 132553424877376 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0424 21:13:51.549321 132553424877376 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 21:13:51.570592 132553424877376 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0424 21:13:51.570662 132553424877376 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0424 21:13:51.786198 132553424877376 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0424 21:13:51.786310 132553424877376 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0424 21:13:53.228380 132553424877376 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0424 21:13:53.228520 132553424877376 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) I0424 21:13:53.228578 132553424877376 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) I0424 21:13:53.228637 132553424877376 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') I0424 21:13:53.228689 132553424877376 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) I0424 21:13:53.228728 132553424877376 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) I0424 21:13:53.228781 132553424877376 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) I0424 21:13:53.228833 132553424877376 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') I0424 21:13:53.228874 132553424877376 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) I0424 21:13:53.228910 132553424877376 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) I0424 21:13:53.228958 132553424877376 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0424 21:13:53.229006 132553424877376 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0424 21:13:53.730425 132553424877376 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0424 21:13:53.730518 132553424877376 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0424 21:13:53.746190 132553424877376 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0424 21:13:53.746250 132553424877376 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0424 21:14:04.795652 132553424877376 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. I0424 21:14:04.796509 132553424877376 metric_logger.py:301] number parameters: 1.104 billion I0424 21:14:16.530893 132553424877376 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0424 21:14:16.685462 132553424877376 checkpointing.py:776] Waited 0.15454983711242676 seconds for step 0 to finish before starting checkpointing. I0424 21:14:16.687893 132553424877376 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 21:14:16.689612 132553424877376 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0 I0424 21:14:16.691055 132553424877376 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/0. I0424 21:14:17.022104 132553424877376 signaling_client.py:364] Using JaxDistributedSignalingClient I0424 21:14:17.022938 132553424877376 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 21:14:17.022994 132553424877376 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 I0424 21:14:17.299391 132553424877376 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.277373s I0424 21:14:17.299571 132553424877376 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.443 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2834310531616211 s) (per-host) I0424 21:14:17.299626 132553424877376 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283496s (batch_requests_ready=0.002706s, total_serialization_initiated=0.280717s, others=0.000073s) I0424 21:14:17.299720 132553424877376 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287583s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.287565) I0424 21:14:17.300523 132553424877376 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_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/0. I0424 21:14:17.300868 132426271303424 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 21:34:17.300826 I0424 21:14:17.310790 132553424877376 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0424 21:14:17.311043 132427897927424 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0424 21:14:17.311160 132553424877376 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777065256.6878748, 'wait_for_prev_duration_secs': 6.198883056640625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777065256.6896505, 'checkpointer_blocking_duration_secs': 0.6113786697387695, 'get_old_steps_start_time': 1777065257.3010545, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1777065256.6859996, 'checkpoint_manager_blocking_duration_secs': 0.6251218318939209} I0424 21:14:17.311316 132553424877376 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0424 21:14:17.311369 132553424877376 max_utils.py:750] Memstats: After params initialized: I0424 21:14:17.311419 132553424877376 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0424 21:14:17.311452 132553424877376 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0424 21:14:17.311480 132553424877376 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0424 21:14:17.311506 132553424877376 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0424 21:14:17.623503 132553424877376 metric_logger.py:196] completed step: 0, seconds: 11.734, TFLOP/s/device: 1.158, Tokens/s/device: 174.532, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0424 21:14:17.793535 132553424877376 metric_logger.py:196] completed step: 1, seconds: 1.091, TFLOP/s/device: 12.453, Tokens/s/device: 1877.019, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0424 21:14:18.207546 132553424877376 metric_logger.py:196] completed step: 2, seconds: 0.023, TFLOP/s/device: 596.424, Tokens/s/device: 89899.478, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719 I0424 21:14:18.355258 132553424877376 metric_logger.py:196] completed step: 3, seconds: 0.414, TFLOP/s/device: 32.794, Tokens/s/device: 4943.099, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506 I0424 21:14:18.359716 132553424877376 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0424 21:14:18.649357 132553424877376 checkpointing.py:776] Waited 0.28960442543029785 seconds for step 5 to finish before starting checkpointing. I0424 21:14:18.652228 132553424877376 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 21:14:20.392150 2758 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0424 21:14:22.417171 132426279696128 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_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/0/items/array_metadatas/process_6 I0424 21:14:27.535156 132426271303424 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 150.167 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.518958330154419 s) (per-host) I0424 21:14:27.535283 132426271303424 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.234298s. I0424 21:14:33.161449 132426271303424 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 15.860447s. I0424 21:14:33.161741 132427897927424 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 21:14:33.161856 132427897927424 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0424 21:14:33.161906 132427897927424 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0424 21:14:33.163649 132427897927424 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0424 21:14:33.163819 132553424877376 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. W0424 21:14:33.163954 132553424877376 checkpoint_manager.py:1452] Waiting for previous save to complete took 14.511722 seconds. If this number is high, consider checkpointing less frequently. I0424 21:14:33.166018 132553424877376 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 5 I0424 21:14:33.168561 132553424877376 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/5. I0424 21:14:33.834149 132553424877376 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 21:14:33.834244 132553424877376 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 I0424 21:14:33.870250 132553424877376 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037156s I0424 21:14:33.870417 132553424877376 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 37.779 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04083108901977539 s) (per-host) I0424 21:14:33.870464 132553424877376 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.040888s (batch_requests_ready=0.001989s, total_serialization_initiated=0.038836s, others=0.000062s) I0424 21:14:33.870557 132553424877376 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.046027s (all_items=0.000021s, per_item={'items': '0.00002074'}, temp_paths=0.046006) I0424 21:14:33.871254 132553424877376 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.71 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/5. I0424 21:14:33.871609 132427389331200 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 21:34:33.871569 I0424 21:14:33.878148 132553424877376 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0424 21:14:33.878391 132427897927424 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0424 21:14:33.878553 132553424877376 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777065258.652199, 'wait_for_prev_duration_secs': 14.51172161102295, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777065273.166057, 'checkpointer_blocking_duration_secs': 0.7057068347930908, 'get_old_steps_start_time': 1777065273.8717856, 'get_old_steps_duration_secs': 3.266334533691406e-05, 'checkpoint_manager_blocking_start_time': 1777065258.6500146, 'checkpoint_manager_blocking_duration_secs': 15.228504419326782} I0424 21:14:33.878669 132553424877376 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0424 21:14:33.879459 132553424877376 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.831, Tokens/s/device: 13389.559, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355 I0424 21:14:33.885441 132553424877376 metric_logger.py:196] completed step: 5, seconds: 0.148, TFLOP/s/device: 92.071, Tokens/s/device: 13877.877, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467 I0424 21:14:38.859502 132426279696128 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_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/5/items/array_metadatas/process_6 I0424 21:14:44.994615 132427389331200 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 141.478 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.164989471435547 s) (per-host) I0424 21:14:44.994744 132427389331200 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 11.123016s. I0424 21:14:50.458961 132427389331200 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 16.587218s. I0424 21:14:50.459243 132427897927424 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 21:14:50.459299 132427897927424 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0424 21:14:50.459341 132427897927424 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0424 21:14:50.461215 132427897927424 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0424 21:14:56.842908 132553424877376 metric_logger.py:196] completed step: 6, seconds: 15.524, TFLOP/s/device: 0.875, Tokens/s/device: 131.922, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740 I0424 21:14:56.990370 132553424877376 metric_logger.py:196] completed step: 7, seconds: 22.811, TFLOP/s/device: 0.596, Tokens/s/device: 89.779, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151 I0424 21:14:57.137764 132553424877376 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.140, Tokens/s/device: 13436.204, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098 I0424 21:14:57.284517 132553424877376 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0424 21:14:57.285226 132553424877376 checkpointing.py:776] Waited 0.0007319450378417969 seconds for step 9 to finish before starting checkpointing. I0424 21:14:57.287254 132553424877376 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 21:14:57.289330 132553424877376 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9 I0424 21:14:57.290940 132553424877376 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/9. I0424 21:14:57.599819 132553424877376 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 21:14:57.599937 132553424877376 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 I0424 21:14:57.634206 132553424877376 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035218s I0424 21:14:57.634377 132553424877376 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 39.627 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03892779350280762 s) (per-host) I0424 21:14:57.634427 132553424877376 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038988s (batch_requests_ready=0.001848s, total_serialization_initiated=0.037075s, others=0.000065s) I0424 21:14:57.634521 132553424877376 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043518s (all_items=0.000011s, per_item={'items': '0.00001097'}, temp_paths=0.043507) I0424 21:14:57.635154 132553424877376 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.35 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/9. I0424 21:14:57.635421 132426814748416 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 21:34:57.635387 I0424 21:14:57.637186 132553424877376 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0424 21:14:57.637438 132400101136128 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0424 21:14:57.637597 132553424877376 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777065297.2872214, 'wait_for_prev_duration_secs': 7.152557373046875e-05, 'time_between_consecutive_saves_sec': 6.825963258743286, 'checkpointer_blocking_start_time': 1777065297.2893682, 'checkpointer_blocking_duration_secs': 0.3461954593658447, 'get_old_steps_start_time': 1777065297.6355777, 'get_old_steps_duration_secs': 2.3365020751953125e-05, 'checkpoint_manager_blocking_start_time': 1777065297.285466, 'checkpoint_manager_blocking_duration_secs': 0.3521006107330322} I0424 21:14:57.637695 132553424877376 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0424 21:14:57.637742 132553424877376 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 21:15:02.825607 132399849486080 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_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_15_ocdbt_false/checkpoints/9/items/array_metadatas/process_6 I0424 21:15:08.319964 132426814748416 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 147.289 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.724478006362915 s) (per-host) I0424 21:15:08.320088 132426814748416 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.684555s. I0424 21:15:14.243527 132426814748416 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 16.607978s. I0424 21:15:14.243812 132400101136128 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 21:15:14.243934 132400101136128 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0424 21:15:14.243975 132400101136128 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0424 21:15:14.245323 132400101136128 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0424 21:15:14.245543 132553424877376 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. I0424 21:15:14.245691 132553424877376 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 21:15:14.246635 132553424877376 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.347, Tokens/s/device: 13919.568, 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: Fri Apr 24 21:15:25 UTC 2026 EXIT_CODE=0