XPK Start: Fri Apr 24 08:26:20 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 08:26:45.290315: 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 08:26:45.498596 134101681002304 max_utils.py:273] Attempting to initialize the jax distributed system... I0424 08:26:54.540722 134101681002304 distributed.py:149] Starting JAX distributed service on [::]:8482 I0424 08:26:54.545242 134101681002304 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-false-sv-b7buj-slice-job-0-0.mt-16-zarr3-false-sv-b7buj:8482 I0424 08:26:55.513823 134101681002304 max_utils.py:284] Jax distributed system initialized! I0424 08:27:01.878430 134101681002304 max_utils.py:800] System Information: Jax Version: 0.9.2 I0424 08:27:01.878541 134101681002304 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0424 08:27:01.878580 134101681002304 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 08:27:01.878615 134101681002304 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0424 08:27:02.577913 134101681002304 maxtext_utils.py:1604] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1) I0424 08:27:02.578215 134101681002304 checkpointing.py:677] Setting up checkpoint logger... I0424 08:27:02.578274 134101681002304 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=False I0424 08:27:02.578326 134101681002304 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0424 08:27:02.578684 134101681002304 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=False, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x79f664918650>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0424 08:27:05.522728 134101681002304 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0424 08:27:05.522965 134101681002304 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 0x79e1745aa960>}, handler_registry=None I0424 08:27:05.523216 134101681002304 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x79e1745aa960>` 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 08:27:05.523265 134101681002304 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79e1745ad6d0>` 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 08:27:05.523301 134101681002304 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 0x79e1745aa960>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x79e1745aa960>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79e1745ad6d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x79e1745ad6d0>}). I0424 08:27:05.523630 134101681002304 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0424 08:27:05.523701 134101681002304 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x79e1547e5800> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0424 08:27:06.508043 134101681002304 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints I0424 08:27:06.522021 134101681002304 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_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x79f53735b500> I0424 08:27:06.522162 134101681002304 checkpointing.py:301] Checkpoint manager created! I0424 08:27:07.464248 134101681002304 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 08:27:07.464370 134101681002304 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 08:27:07.865709 134101681002304 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch_attn', 'activation_length_attn', 'activation_embed_attn'). I0424 08:27:07.865803 134101681002304 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 08:27:07.882537 134101681002304 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch_attn', 'activation_length_attn', 'activation_embed_attn'). I0424 08:27:07.882602 134101681002304 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 08:27:07.906894 134101681002304 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 08:27:07.906965 134101681002304 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 08:27:07.923756 134101681002304 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 08:27:07.923818 134101681002304 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 08:27:07.940649 134101681002304 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 08:27:07.940710 134101681002304 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 08:27:07.965670 134101681002304 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch_attn', 'activation_length_attn', 'activation_heads', 'activation_kv'). I0424 08:27:07.965739 134101681002304 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 08:27:07.987213 134101681002304 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0424 08:27:07.987279 134101681002304 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0424 08:27:08.203077 134101681002304 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0424 08:27:08.203197 134101681002304 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0424 08:27:09.636245 134101681002304 maxtext_utils.py:1707] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0424 08:27:09.636378 134101681002304 maxtext_utils.py:1707] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0424 08:27:09.636432 134101681002304 maxtext_utils.py:1707] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0424 08:27:09.636489 134101681002304 maxtext_utils.py:1707] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0424 08:27:09.636546 134101681002304 maxtext_utils.py:1707] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0424 08:27:09.636586 134101681002304 maxtext_utils.py:1707] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0424 08:27:09.636638 134101681002304 maxtext_utils.py:1707] 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 08:27:09.636689 134101681002304 maxtext_utils.py:1707] 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 08:27:09.636728 134101681002304 maxtext_utils.py:1707] 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 08:27:09.636764 134101681002304 maxtext_utils.py:1707] 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 08:27:09.636811 134101681002304 maxtext_utils.py:1707] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0424 08:27:09.636859 134101681002304 maxtext_utils.py:1707] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0424 08:27:10.129678 134101681002304 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0424 08:27:10.129772 134101681002304 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0424 08:27:10.145466 134101681002304 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0424 08:27:10.145529 134101681002304 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0424 08:27:21.025619 134101681002304 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 08:27:21.026454 134101681002304 metric_logger.py:301] number parameters: 1.104 billion I0424 08:27:32.427545 134101681002304 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0424 08:27:32.820245 134101681002304 checkpointing.py:776] Waited 0.39267659187316895 seconds for step 0 to finish before starting checkpointing. I0424 08:27:32.822646 134101681002304 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 08:27:32.824358 134101681002304 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0 I0424 08:27:32.825798 134101681002304 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/0. I0424 08:27:33.166866 134101681002304 signaling_client.py:364] Using JaxDistributedSignalingClient I0424 08:27:33.167777 134101681002304 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 08:27:33.167831 134101681002304 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 08:27:33.450154 134101681002304 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.283429s I0424 08:27:33.450330 134101681002304 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.337 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28903865814208984 s) (per-host) I0424 08:27:33.450383 134101681002304 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289101s (batch_requests_ready=0.002366s, total_serialization_initiated=0.286666s, others=0.000069s) I0424 08:27:33.450474 134101681002304 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293277s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.293259) I0424 08:27:33.451241 134101681002304 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.63 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/0. I0424 08:27:33.451588 133973401376512 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 08:47:33.451543 I0424 08:27:33.460803 134101681002304 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0424 08:27:33.461136 133972348626688 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0424 08:27:33.461291 134101681002304 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777019252.8226283, 'wait_for_prev_duration_secs': 6.008148193359375e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777019252.8243968, 'checkpointer_blocking_duration_secs': 0.6273412704467773, 'get_old_steps_start_time': 1777019253.4517639, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1777019252.820786, 'checkpoint_manager_blocking_duration_secs': 0.640465259552002} I0424 08:27:33.461405 134101681002304 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0424 08:27:33.461453 134101681002304 max_utils.py:750] Memstats: After params initialized: I0424 08:27:33.461506 134101681002304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0424 08:27:33.461539 134101681002304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0424 08:27:33.461565 134101681002304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0424 08:27:33.461591 134101681002304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0424 08:27:33.771408 134101681002304 metric_logger.py:196] completed step: 0, seconds: 11.401, TFLOP/s/device: 1.192, Tokens/s/device: 179.634, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0424 08:27:33.947651 134101681002304 metric_logger.py:196] completed step: 1, seconds: 1.342, TFLOP/s/device: 10.122, Tokens/s/device: 1525.661, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0424 08:27:34.363768 134101681002304 metric_logger.py:196] completed step: 2, seconds: 0.029, TFLOP/s/device: 466.223, Tokens/s/device: 70274.165, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719 I0424 08:27:34.511024 134101681002304 metric_logger.py:196] completed step: 3, seconds: 0.416, TFLOP/s/device: 32.650, Tokens/s/device: 4921.385, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506 I0424 08:27:34.515823 134101681002304 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0424 08:27:34.804993 134101681002304 checkpointing.py:776] Waited 0.28913378715515137 seconds for step 5 to finish before starting checkpointing. I0424 08:27:34.808058 134101681002304 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 08:27:36.371966 2770 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0424 08:27:38.390138 133972860344064 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/0/items/array_metadatas/process_6 I0424 08:28:15.036293 133973401376512 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.722 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.87494397163391 s) (per-host) I0424 08:28:15.036419 133973401376512 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.584720s. I0424 08:28:23.152091 133973401376512 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 49.700375s. I0424 08:28:23.152346 133972348626688 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 08:28:23.152412 133972348626688 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0424 08:28:23.152461 133972348626688 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0424 08:28:23.154074 133972348626688 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0424 08:28:23.154282 134101681002304 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 08:28:23.154420 134101681002304 checkpoint_manager.py:1452] Waiting for previous save to complete took 48.346364 seconds. If this number is high, consider checkpointing less frequently. I0424 08:28:23.156296 134101681002304 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 5 I0424 08:28:23.158675 134101681002304 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/5. I0424 08:28:23.877481 134101681002304 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 08:28:23.877573 134101681002304 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 08:28:23.917898 134101681002304 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.041581s I0424 08:28:23.918070 134101681002304 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 33.679 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04580259323120117 s) (per-host) I0424 08:28:23.918133 134101681002304 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.045874s (batch_requests_ready=0.002581s, total_serialization_initiated=0.043214s, others=0.000079s) I0424 08:28:23.918224 134101681002304 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.051369s (all_items=0.000014s, per_item={'items': '0.00001431'}, temp_paths=0.051355) I0424 08:28:23.918942 134101681002304 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/5. I0424 08:28:23.919243 133972348626688 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 08:48:23.919209 I0424 08:28:23.926988 134101681002304 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0424 08:28:23.927287 133972860344064 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0424 08:28:23.927427 134101681002304 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777019254.8080277, 'wait_for_prev_duration_secs': 48.34636449813843, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777019303.1563342, 'checkpointer_blocking_duration_secs': 0.7630705833435059, 'get_old_steps_start_time': 1777019303.9194305, 'get_old_steps_duration_secs': 2.7894973754882812e-05, 'checkpoint_manager_blocking_start_time': 1777019254.8057158, 'checkpoint_manager_blocking_duration_secs': 49.1216766834259} I0424 08:28:23.927587 134101681002304 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0424 08:28:23.928499 134101681002304 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.780, Tokens/s/device: 13381.860, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355 I0424 08:28:23.934406 134101681002304 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.133, Tokens/s/device: 13887.288, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467 I0424 08:28:29.029891 133948189435648 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/5/items/array_metadatas/process_6 I0424 08:28:46.374685 134101681002304 metric_logger.py:196] completed step: 6, seconds: 49.417, TFLOP/s/device: 0.275, Tokens/s/device: 41.443, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740 I0424 08:28:46.522140 134101681002304 metric_logger.py:196] completed step: 7, seconds: 22.294, TFLOP/s/device: 0.609, Tokens/s/device: 91.862, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151 I0424 08:28:46.669186 134101681002304 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.452, Tokens/s/device: 13483.176, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098 I0424 08:28:46.815730 134101681002304 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0424 08:28:46.816418 134101681002304 checkpointing.py:776] Waited 0.0007090568542480469 seconds for step 9 to finish before starting checkpointing. I0424 08:28:46.818457 134101681002304 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 08:29:06.361988 133972348626688 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.176 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.489681005477905 s) (per-host) I0424 08:29:06.362123 133972348626688 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.442764s. I0424 08:29:13.339376 133972348626688 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 49.420001s. I0424 08:29:13.339677 133972860344064 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 08:29:13.339796 133972860344064 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0424 08:29:13.339849 133972860344064 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0424 08:29:13.341603 133972860344064 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0424 08:29:13.341806 134101681002304 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0424 08:29:13.341933 134101681002304 checkpoint_manager.py:1452] Waiting for previous save to complete took 26.523483 seconds. If this number is high, consider checkpointing less frequently. I0424 08:29:13.343358 134101681002304 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9 I0424 08:29:13.345461 134101681002304 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/9. I0424 08:29:13.646200 134101681002304 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 08:29:13.646291 134101681002304 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 08:29:13.687172 134101681002304 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.041742s I0424 08:29:13.687343 134101681002304 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 34.048 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.045305490493774414 s) (per-host) I0424 08:29:13.687393 134101681002304 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.045366s (batch_requests_ready=0.001772s, total_serialization_initiated=0.043524s, others=0.000069s) I0424 08:29:13.687482 134101681002304 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.049420s (all_items=0.000010s, per_item={'items': '0.00000954'}, temp_paths=0.049411) I0424 08:29:13.688196 134101681002304 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/9. I0424 08:29:13.688486 133972860344064 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 08:49:13.688455 I0424 08:29:13.690382 134101681002304 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0424 08:29:13.690646 133970210666240 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0424 08:29:13.690774 134101681002304 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777019326.818426, 'wait_for_prev_duration_secs': 26.52348303794861, 'time_between_consecutive_saves_sec': 23.664292812347412, 'checkpointer_blocking_start_time': 1777019353.3433979, 'checkpointer_blocking_duration_secs': 0.3452339172363281, 'get_old_steps_start_time': 1777019353.6886501, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1777019326.8166533, 'checkpoint_manager_blocking_duration_secs': 26.874091625213623} I0424 08:29:13.690920 134101681002304 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0424 08:29:13.690960 134101681002304 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 08:29:19.265356 133948189435648 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260424_070227/linen_xpk_main_20260424_070227_16_zarr3_false/checkpoints/9/items/array_metadatas/process_6 I0424 08:29:55.176678 133972860344064 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 38.031 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.534600019454956 s) (per-host) I0424 08:29:55.176787 133972860344064 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.488187s. I0424 08:30:04.683521 133972860344064 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 50.994905s. I0424 08:30:04.683818 133970210666240 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 08:30:04.683933 133970210666240 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0424 08:30:04.683974 133970210666240 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0424 08:30:04.685411 133970210666240 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0424 08:30:04.685598 134101681002304 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 08:30:04.685750 134101681002304 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 08:30:04.686750 134101681002304 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.146, Tokens/s/device: 13889.266, 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 08:30:16 UTC 2026 EXIT_CODE=0