XPK Start: Tue Apr 21 15:52:05 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. 2026-04-21 15:52:30.123328: 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) I0421 15:52:30.334112 137180641048384 max_utils.py:273] Attempting to initialize the jax distributed system... I0421 15:52:39.377353 137180641048384 distributed.py:149] Starting JAX distributed service on [::]:8482 I0421 15:52:39.379932 137180641048384 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-false-sv-4oczz-slice-job-0-0.mt-15-ocdbt-false-sv-4oczz:8482 I0421 15:52:40.349617 137180641048384 max_utils.py:284] Jax distributed system initialized! I0421 15:52:46.513726 137180641048384 max_utils.py:800] System Information: Jax Version: 0.9.2 I0421 15:52:46.513855 137180641048384 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0421 15:52:46.513895 137180641048384 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 I0421 15:52:46.513932 137180641048384 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0421 15:52:47.215804 137180641048384 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0421 15:52:47.216083 137180641048384 checkpointing.py:677] Setting up checkpoint logger... I0421 15:52:47.216143 137180641048384 checkpointing.py:233] Creating checkpoint manager with ocdbt=False and zarr3=True I0421 15:52:47.216187 137180641048384 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0421 15:52:47.216571 137180641048384 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 0x7cc34c7f5940>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0421 15:52:50.106685 137180641048384 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0421 15:52:50.107115 137180641048384 checkpoint_manager.py:708] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7caf045fa6c0>}, handler_registry=None I0421 15:52:50.107382 137180641048384 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7caf045fa6c0>` 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`. I0421 15:52:50.107446 137180641048384 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7caf045fc1d0>` 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`. I0421 15:52:50.107503 137180641048384 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 0x7caf045fa6c0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7caf045fa6c0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7caf045fc1d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7caf045fc1d0>}). I0421 15:52:50.107863 137180641048384 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0421 15:52:50.107952 137180641048384 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7caf04375800> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0421 15:52:52.113269 137180641048384 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints I0421 15:52:52.541602 137180641048384 checkpoint_manager.py:929] [process=7][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=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_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7cc2177befc0> I0421 15:52:52.541807 137180641048384 checkpointing.py:301] Checkpoint manager created! I0421 15:52:53.475223 137180641048384 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0421 15:52:53.475342 137180641048384 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:52:53.854893 137180641048384 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 15:52:53.854985 137180641048384 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:52:53.871481 137180641048384 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 15:52:53.871541 137180641048384 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:52:53.895311 137180641048384 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0421 15:52:53.895386 137180641048384 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:52:53.911866 137180641048384 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0421 15:52:53.911925 137180641048384 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:52:53.928555 137180641048384 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0421 15:52:53.928615 137180641048384 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:52:53.958206 137180641048384 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0421 15:52:53.958292 137180641048384 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:52:53.980280 137180641048384 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0421 15:52:53.980356 137180641048384 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0421 15:52:54.196896 137180641048384 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0421 15:52:54.197011 137180641048384 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0421 15:52:55.626282 137180641048384 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0421 15:52:55.626410 137180641048384 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) I0421 15:52:55.626464 137180641048384 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) I0421 15:52:55.626522 137180641048384 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') I0421 15:52:55.626574 137180641048384 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) I0421 15:52:55.626613 137180641048384 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) I0421 15:52:55.626666 137180641048384 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) I0421 15:52:55.626721 137180641048384 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') I0421 15:52:55.626778 137180641048384 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) I0421 15:52:55.626817 137180641048384 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) I0421 15:52:55.626872 137180641048384 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0421 15:52:55.626923 137180641048384 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0421 15:52:56.115104 137180641048384 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 15:52:56.115199 137180641048384 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0421 15:52:56.130703 137180641048384 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 15:52:56.130775 137180641048384 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0421 15:53:07.031003 137180641048384 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. I0421 15:53:07.031845 137180641048384 metric_logger.py:301] number parameters: 1.104 billion I0421 15:53:18.916033 137180641048384 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0421 15:53:19.069008 137180641048384 checkpointing.py:776] Waited 0.1529543399810791 seconds for step 0 to finish before starting checkpointing. I0421 15:53:19.071719 137180641048384 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:53:19.073448 137180641048384 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 0 I0421 15:53:19.074976 137180641048384 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/0. I0421 15:53:19.818143 137180641048384 signaling_client.py:364] Using JaxDistributedSignalingClient I0421 15:53:19.819168 137180641048384 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:53:19.819224 137180641048384 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 I0421 15:53:20.102403 137180641048384 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.284336s I0421 15:53:20.102580 137180641048384 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.320 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28997325897216797 s) (per-host) I0421 15:53:20.102634 137180641048384 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.290039s (batch_requests_ready=0.002372s, total_serialization_initiated=0.287593s, others=0.000073s) I0421 15:53:20.102730 137180641048384 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.294738s (all_items=0.000018s, per_item={'items': '0.00001812'}, temp_paths=0.294720) I0421 15:53:20.103586 137180641048384 event_tracking.py:125] [process=7] [async] Finished blocking save in 1.03 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/0. I0421 15:53:20.103945 137053090424576 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:13:20.103902 I0421 15:53:20.105985 137180641048384 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:53:20.106254 137051122177792 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:53:20.106383 137180641048384 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776786799.0716994, 'wait_for_prev_duration_secs': 7.557868957519531e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776786799.0734863, 'checkpointer_blocking_duration_secs': 1.0306291580200195, 'get_old_steps_start_time': 1776786800.1041422, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776786799.0695317, 'checkpoint_manager_blocking_duration_secs': 1.0368154048919678} I0421 15:53:20.106540 137180641048384 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0421 15:53:20.106592 137180641048384 max_utils.py:750] Memstats: After params initialized: I0421 15:53:20.106642 137180641048384 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0)) I0421 15:53:20.106675 137180641048384 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0)) I0421 15:53:20.106704 137180641048384 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0)) I0421 15:53:20.106741 137180641048384 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0)) I0421 15:53:20.422021 137180641048384 metric_logger.py:196] completed step: 0, seconds: 11.884, TFLOP/s/device: 1.143, Tokens/s/device: 172.331, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 15:53:20.609238 137180641048384 metric_logger.py:196] completed step: 1, seconds: 1.504, TFLOP/s/device: 9.032, Tokens/s/device: 1361.371, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 15:53:21.015865 137180641048384 metric_logger.py:196] completed step: 2, seconds: 0.040, TFLOP/s/device: 340.606, Tokens/s/device: 51339.901, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719 I0421 15:53:21.163086 137180641048384 metric_logger.py:196] completed step: 3, seconds: 0.407, TFLOP/s/device: 33.377, Tokens/s/device: 5031.014, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506 I0421 15:53:21.167626 137180641048384 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0421 15:53:21.457279 137180641048384 checkpointing.py:776] Waited 0.28962063789367676 seconds for step 5 to finish before starting checkpointing. I0421 15:53:21.460254 137180641048384 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 15:53:23.421784 2790 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0421 15:53:25.925482 137051636606720 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/0/items/array_metadatas/process_7 I0421 15:53:30.957210 137053090424576 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 141.737 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.144579410552979 s) (per-host) I0421 15:53:30.957319 137053090424576 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.853251s. I0421 15:53:37.426341 137053090424576 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 17.322256s. I0421 15:53:37.426610 137051122177792 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:53:37.426752 137051122177792 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:53:37.426808 137051122177792 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:53:37.428294 137051122177792 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0421 15:53:37.428462 137180641048384 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0421 15:53:37.428602 137180641048384 checkpoint_manager.py:1452] Waiting for previous save to complete took 15.968345 seconds. If this number is high, consider checkpointing less frequently. I0421 15:53:37.430609 137180641048384 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 5 I0421 15:53:37.433148 137180641048384 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/5. I0421 15:53:38.139798 137180641048384 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:53:38.139894 137180641048384 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 I0421 15:53:38.178816 137180641048384 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.040281s I0421 15:53:38.178986 137180641048384 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 35.175 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.043854475021362305 s) (per-host) I0421 15:53:38.179037 137180641048384 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.043916s (batch_requests_ready=0.001913s, total_serialization_initiated=0.041936s, others=0.000067s) I0421 15:53:38.179140 137180641048384 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.048945s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.048928) I0421 15:53:38.179834 137180641048384 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/5. I0421 15:53:38.180194 137051122177792 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:13:38.180156 I0421 15:53:38.182256 137180641048384 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:53:38.182586 137052196693760 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:53:38.182722 137180641048384 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776786801.4602253, 'wait_for_prev_duration_secs': 15.968345165252686, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776786817.4306486, 'checkpointer_blocking_duration_secs': 0.7496964931488037, 'get_old_steps_start_time': 1776786818.1803663, 'get_old_steps_duration_secs': 3.3855438232421875e-05, 'checkpoint_manager_blocking_start_time': 1776786801.4579265, 'checkpoint_manager_blocking_duration_secs': 16.724761724472046} I0421 15:53:38.182897 137180641048384 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0421 15:53:38.183763 137180641048384 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.819, Tokens/s/device: 13387.721, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355 I0421 15:53:38.189930 137180641048384 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.314, Tokens/s/device: 13914.556, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467 I0421 15:53:44.147919 137051636606720 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/5/items/array_metadatas/process_7 I0421 15:53:48.562915 137051122177792 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 151.480 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.42774510383606 s) (per-host) I0421 15:53:48.563041 137051122177792 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.382730s. I0421 15:53:55.382182 137051122177792 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 17.201855s. I0421 15:53:55.382452 137052196693760 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:53:55.382569 137052196693760 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:53:55.382617 137052196693760 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:53:55.384115 137052196693760 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0421 15:54:00.650329 137180641048384 metric_logger.py:196] completed step: 6, seconds: 17.021, TFLOP/s/device: 0.798, Tokens/s/device: 120.324, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740 I0421 15:54:00.797818 137180641048384 metric_logger.py:196] completed step: 7, seconds: 22.315, TFLOP/s/device: 0.609, Tokens/s/device: 91.779, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151 I0421 15:54:00.945263 137180641048384 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.441, Tokens/s/device: 13481.578, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098 I0421 15:54:01.092031 137180641048384 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0421 15:54:01.092700 137180641048384 checkpointing.py:776] Waited 0.0006914138793945312 seconds for step 9 to finish before starting checkpointing. I0421 15:54:01.094759 137180641048384 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:54:01.096428 137180641048384 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 9 I0421 15:54:01.097884 137180641048384 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/9. I0421 15:54:01.802950 137180641048384 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:54:01.803043 137180641048384 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 I0421 15:54:01.841293 137180641048384 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.039250s I0421 15:54:01.841464 137180641048384 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 35.994 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.042856454849243164 s) (per-host) I0421 15:54:01.841514 137180641048384 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.042918s (batch_requests_ready=0.001781s, total_serialization_initiated=0.041071s, others=0.000067s) I0421 15:54:01.841607 137180641048384 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.047192s (all_items=0.000012s, per_item={'items': '0.00001168'}, temp_paths=0.047180) I0421 15:54:01.842344 137180641048384 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/9. I0421 15:54:01.842639 137052196693760 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:14:01.842605 I0421 15:54:01.844630 137180641048384 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:54:01.844938 137030042343168 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:54:01.845077 137180641048384 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776786841.094711, 'wait_for_prev_duration_secs': 8.893013000488281e-05, 'time_between_consecutive_saves_sec': 5.7105560302734375, 'checkpointer_blocking_start_time': 1776786841.0964668, 'checkpointer_blocking_duration_secs': 0.7462868690490723, 'get_old_steps_start_time': 1776786841.842772, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776786841.0929542, 'checkpoint_manager_blocking_duration_secs': 0.7520880699157715} I0421 15:54:01.845248 137180641048384 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0421 15:54:01.845296 137180641048384 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 15:54:06.880774 137030503745280 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_15_ocdbt_false/checkpoints/9/items/array_metadatas/process_7 I0421 15:54:12.484055 137052196693760 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 147.827 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.685407400131226 s) (per-host) I0421 15:54:12.484180 137052196693760 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.640906s. I0421 15:54:17.328591 137052196693760 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 15.485300s. I0421 15:54:17.328902 137030042343168 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:54:17.329033 137030042343168 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:54:17.329078 137030042343168 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:54:17.331393 137030042343168 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0421 15:54:17.331573 137180641048384 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0421 15:54:17.331725 137180641048384 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:54:17.332653 137180641048384 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.089, Tokens/s/device: 13880.605, 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: Tue Apr 21 15:54:25 UTC 2026 EXIT_CODE=0