XPK Start: Sun Apr 19 18:10:38 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-19 18:11:02.608743: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0419 18:11:02.787613 136337834829632 max_utils.py:273] Attempting to initialize the jax distributed system... I0419 18:11:11.828593 136337834829632 distributed.py:149] Starting JAX distributed service on [::]:8482 I0419 18:11:11.830913 136337834829632 distributed.py:172] Connecting to JAX distributed service on mt-02-synthetic-6gfwy-slice-job-0-0.mt-02-synthetic-6gfwy:8482 I0419 18:11:13.523016 136337834829632 max_utils.py:284] Jax distributed system initialized! I0419 18:11:19.744524 136337834829632 max_utils.py:800] System Information: Jax Version: 0.9.2 I0419 18:11:19.744626 136337834829632 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0419 18:11:19.744665 136337834829632 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365 I0419 18:11:19.744696 136337834829632 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 18:11:20.445570 136337834829632 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 18:11:20.445844 136337834829632 checkpointing.py:677] Setting up checkpoint logger... I0419 18:11:20.445900 136337834829632 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 18:11:20.445944 136337834829632 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0419 18:11:20.446276 136337834829632 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7bff387ae6c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 18:11:23.408188 136337834829632 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0419 18:11:23.408430 136337834829632 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 0x7bead85be8a0>}, handler_registry=None I0419 18:11:23.408667 136337834829632 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bead85be8a0>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0419 18:11:23.408715 136337834829632 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bead85bcb30>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0419 18:11:23.408749 136337834829632 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 0x7bead85be8a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bead85be8a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bead85bcb30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bead85bcb30>}). I0419 18:11:23.409073 136337834829632 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0419 18:11:23.409149 136337834829632 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7bead81d3d80> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0419 18:11:24.535063 136337834829632 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints I0419 18:11:24.537507 136337834829632 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=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7bead85bd160> I0419 18:11:24.537616 136337834829632 checkpointing.py:301] Checkpoint manager created! I0419 18:11:25.464351 136337834829632 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 18:11:25.464481 136337834829632 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 18:11:25.845324 136337834829632 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 18:11:25.845428 136337834829632 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 18:11:25.861695 136337834829632 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 18:11:25.861751 136337834829632 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 18:11:25.885149 136337834829632 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:11:25.885215 136337834829632 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:11:25.901420 136337834829632 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:11:25.901478 136337834829632 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:11:25.917639 136337834829632 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:11:25.917698 136337834829632 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:11:25.942086 136337834829632 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 18:11:25.942154 136337834829632 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 18:11:25.963654 136337834829632 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 18:11:25.963716 136337834829632 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 18:11:26.183911 136337834829632 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0419 18:11:26.184021 136337834829632 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0419 18:11:27.606010 136337834829632 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0419 18:11:27.606138 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 18:11:27.606192 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 18:11:27.606250 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0419 18:11:27.606305 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0419 18:11:27.606343 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0419 18:11:27.606403 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 18:11:27.606505 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: P('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0419 18:11:27.606557 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0419 18:11:27.606604 136337834829632 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 18:11:27.606657 136337834829632 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0419 18:11:27.606709 136337834829632 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0419 18:11:28.093503 136337834829632 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 18:11:28.093601 136337834829632 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0419 18:11:28.108912 136337834829632 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 18:11:28.108974 136337834829632 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0419 18:11:39.132679 136337834829632 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0419 18:11:39.133528 136337834829632 metric_logger.py:301] number parameters: 1.104 billion I0419 18:11:50.412328 136337834829632 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0419 18:11:50.565636 136337834829632 checkpointing.py:776] Waited 0.1532881259918213 seconds for step 0 to finish before starting checkpointing. I0419 18:11:50.568304 136337834829632 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 18:11:50.569905 136337834829632 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0 I0419 18:11:50.571420 136337834829632 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints/0. I0419 18:11:51.280312 136337834829632 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 18:11:51.281395 136337834829632 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0419 18:11:51.281467 136337834829632 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0419 18:11:51.561807 136337834829632 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.281567s I0419 18:11:51.561984 136337834829632 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.375 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2869729995727539 s) (per-host) I0419 18:11:51.562037 136337834829632 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.287036s (batch_requests_ready=0.002217s, total_serialization_initiated=0.284750s, others=0.000070s) I0419 18:11:51.562126 136337834829632 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.291122s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.291103) I0419 18:11:51.562972 136337834829632 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.99 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints/0. I0419 18:11:51.563333 136209963665152 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 18:31:51.563296 I0419 18:11:51.571395 136337834829632 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 18:11:51.571681 136208906708736 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0419 18:11:51.571811 136337834829632 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776622310.568278, 'wait_for_prev_duration_secs': 9.703636169433594e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776622310.5699625, 'checkpointer_blocking_duration_secs': 0.9934723377227783, 'get_old_steps_start_time': 1776622311.5634592, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776622310.5662146, 'checkpoint_manager_blocking_duration_secs': 1.0055577754974365} I0419 18:11:51.571913 136337834829632 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0419 18:11:51.571962 136337834829632 max_utils.py:750] Memstats: After params initialized: I0419 18:11:51.572023 136337834829632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0419 18:11:51.572057 136337834829632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0419 18:11:51.572085 136337834829632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0419 18:11:51.572110 136337834829632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0419 18:11:51.883791 136337834829632 metric_logger.py:196] completed step: 0, seconds: 11.279, TFLOP/s/device: 1.205, Tokens/s/device: 181.581, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0419 18:11:52.064648 136337834829632 metric_logger.py:196] completed step: 1, seconds: 1.470, TFLOP/s/device: 9.244, Tokens/s/device: 1393.287, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0419 18:11:52.493886 136337834829632 metric_logger.py:196] completed step: 2, seconds: 0.034, TFLOP/s/device: 402.248, Tokens/s/device: 60631.180, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0419 18:11:52.641229 136337834829632 metric_logger.py:196] completed step: 3, seconds: 0.430, TFLOP/s/device: 31.615, Tokens/s/device: 4765.317, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203 I0419 18:11:52.936371 136337834829632 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 89.025, Tokens/s/device: 13418.861, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614 I0419 18:11:52.942545 136337834829632 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.217, Tokens/s/device: 13899.918, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717 I0419 18:11:54.740492 2776 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 18:11:57.234428 136208915101440 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints/0/items/array_metadatas/process_6 I0419 18:12:16.106925 136337834829632 metric_logger.py:196] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.944, Tokens/s/device: 6925.166, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759 I0419 18:12:16.254164 136337834829632 metric_logger.py:196] completed step: 7, seconds: 23.017, TFLOP/s/device: 0.590, Tokens/s/device: 88.976, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905 I0419 18:12:16.401564 136337834829632 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 89.088, Tokens/s/device: 13428.363, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406 I0419 18:12:16.548269 136337834829632 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0419 18:12:16.548956 136337834829632 checkpointing.py:776] Waited 0.0007050037384033203 seconds for step 9 to finish before starting checkpointing. I0419 18:12:16.551061 136337834829632 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 18:12:28.648435 136209963665152 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 42.265 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.37337899208069 s) (per-host) I0419 18:12:28.648561 136209963665152 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.084603s. I0419 18:12:37.903805 136209963665152 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 46.339831s. I0419 18:12:37.904091 136208906708736 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 18:12:37.904233 136208906708736 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0419 18:12:37.904296 136208906708736 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 18:12:37.906045 136208906708736 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 18:12:37.906236 136337834829632 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. W0419 18:12:37.906397 136337834829632 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.355335 seconds. If this number is high, consider checkpointing less frequently. I0419 18:12:37.908423 136337834829632 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9 I0419 18:12:37.910475 136337834829632 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints/9. I0419 18:12:38.620580 136337834829632 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0419 18:12:38.620671 136337834829632 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0419 18:12:38.662622 136337834829632 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043395s I0419 18:12:38.662771 136337834829632 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 32.954 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04681038856506348 s) (per-host) I0419 18:12:38.662821 136337834829632 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.046869s (batch_requests_ready=0.001795s, total_serialization_initiated=0.045008s, others=0.000066s) I0419 18:12:38.662907 136337834829632 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.050931s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.050914) I0419 18:12:38.663565 136337834829632 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints/9. I0419 18:12:38.663909 136208906708736 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 18:32:38.663872 I0419 18:12:38.665903 136337834829632 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 18:12:38.666142 136205639649024 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0419 18:12:38.666244 136337834829632 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776622336.551031, 'wait_for_prev_duration_secs': 21.355334758758545, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776622357.9084616, 'checkpointer_blocking_duration_secs': 0.7555959224700928, 'get_old_steps_start_time': 1776622358.6640825, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776622336.549184, 'checkpoint_manager_blocking_duration_secs': 22.11702823638916} I0419 18:12:38.666426 136337834829632 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0419 18:12:38.666473 136337834829632 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 18:12:44.789904 136208915101440 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_02_synthetic/checkpoints/9/items/array_metadatas/process_6 I0419 18:13:20.959665 136208906708736 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.304 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.343666076660156 s) (per-host) I0419 18:13:20.959786 136208906708736 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.295761s. I0419 18:13:28.121254 136208906708736 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 49.457211s. I0419 18:13:28.121536 136205639649024 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 18:13:28.121610 136205639649024 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0419 18:13:28.121665 136205639649024 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 18:13:28.123120 136205639649024 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 18:13:28.123306 136337834829632 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. I0419 18:13:28.123476 136337834829632 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 18:13:28.124479 136337834829632 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.288, Tokens/s/device: 13910.681, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 18:13:38 UTC 2026 EXIT_CODE=0