XPK Start: Sun Apr 19 03:25:29 UTC 2026 2026-04-19 03:25:33.280378: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered WARNING: All log messages before absl::InitializeLog() is called are written to STDERR E0000 00:00:1776569133.292946 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776569133.296646 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776569133.308137 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776569133.308158 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776569133.308161 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776569133.308163 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 03:25:53.068975: 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 03:25:53.594715 138108591912768 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 03:26:02,636:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 03:26:02.636728 138108591912768 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 03:26:02,639:jax._src.distributed:157: Connecting to JAX distributed service on mt-04-int8-ndq1y-slice-job-0-0.mt-04-int8-ndq1y:8482 I0419 03:26:02.639091 138108591912768 distributed.py:157] Connecting to JAX distributed service on mt-04-int8-ndq1y-slice-job-0-0.mt-04-int8-ndq1y:8482 I0419 03:26:04.303702 138108591912768 max_utils.py:284] Jax distributed system initialized! I0419 03:26:11.221864 138108591912768 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 03:26:11.221971 138108591912768 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 03:26:11.222013 138108591912768 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709 I0419 03:26:11.222049 138108591912768 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 03:26:11.840358 138108591912768 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 03:26:11.840938 138108591912768 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 03:26:11.841231 138108591912768 checkpointing.py:688] Setting up checkpoint logger... I0419 03:26:11.841299 138108591912768 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 03:26:11.841350 138108591912768 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 03:26:11.841690 138108591912768 base_pytree_checkpoint_handler.py:415] 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 0x7d9b6fb79490>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 03:26:15.906601 138108591912768 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0419 03:26:15.906947 138108591912768 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d86d42c5e50>}, handler_registry=None I0419 03:26:15.907184 138108591912768 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d86d42c5e50>` 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 03:26:15.907244 138108591912768 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d86d42cb110>` 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 03:26:15.907282 138108591912768 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 0x7d86d42c5e50>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d86d42c5e50>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d86d42cb110>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d86d42cb110>}). I0419 03:26:15.907794 138108591912768 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 03:26:15.907871 138108591912768 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d86d40d84a0> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 03:26:17.650412 138108591912768 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints I0419 03:26:18.541985 138108591912768 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_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d86d42c6990> I0419 03:26:18.542166 138108591912768 checkpointing.py:302] Checkpoint manager created! I0419 03:26:19.747536 138108591912768 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 03:26:19.747653 138108591912768 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 03:26:20.123540 138108591912768 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 03:26:20.123636 138108591912768 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 03:26:20.139749 138108591912768 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 03:26:20.139811 138108591912768 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 03:26:20.218075 138108591912768 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 03:26:20.218170 138108591912768 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 03:26:20.234274 138108591912768 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 03:26:20.234335 138108591912768 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 03:26:20.250064 138108591912768 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 03:26:20.250133 138108591912768 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 03:26:20.274126 138108591912768 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 03:26:20.274196 138108591912768 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 03:26:20.355003 138108591912768 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 03:26:20.355096 138108591912768 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 03:26:20.961196 138108591912768 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0419 03:26:20.961328 138108591912768 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0419 03:26:23.640463 138108591912768 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0419 03:26:23.640609 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 03:26:23.640667 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 03:26:23.640727 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0419 03:26:23.640783 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 03:26:23.640824 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 03:26:23.640880 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 03:26:23.640937 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: PartitionSpec('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0419 03:26:23.640978 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0419 03:26:23.641016 138108591912768 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 03:26:23.641065 138108591912768 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0419 03:26:23.641114 138108591912768 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0419 03:26:25.355378 138108591912768 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 03:26:25.355474 138108591912768 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0419 03:26:25.370402 138108591912768 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 03:26:25.370461 138108591912768 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0419 03:26:39.271105 138108591912768 max_utils.py:791] Total memory size: 1.8 GB, Output size: 0.4 GB, Temp size: 1.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0419 03:26:39.271993 138108591912768 metric_logger.py:289] number parameters: 1.104 billion I0419 03:26:54.819052 138108591912768 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0419 03:26:55.060438 138108591912768 checkpointing.py:798] Waited 0.24136638641357422 seconds for step 0 to finish before starting checkpointing. I0419 03:26:55.062905 138108591912768 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 03:26:55.064600 138108591912768 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0 I0419 03:26:55.067379 138108591912768 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints/0. I0419 03:26:56.570340 138108591912768 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 03:26:56.571389 138108591912768 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 03:26:56.571450 138108591912768 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 03:26:56.814554 138108591912768 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.244312s I0419 03:26:56.814729 138108591912768 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.604 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.9619736671447754 s) (per-host) I0419 03:26:56.814783 138108591912768 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.962037s (batch_requests_ready=0.714005s, total_serialization_initiated=0.247960s, others=0.000072s) I0419 03:26:56.814876 138108591912768 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.966460s (all_items=0.000021s, per_item={'items': '0.00002122'}, temp_paths=0.966439) I0419 03:26:56.815882 137985907463936 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0419 03:26:56.816060 138108591912768 async_checkpointer.py:561] Finished blocking save. Time taken: 1.751388s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints/0. I0419 03:26:56.823990 138108591912768 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 03:26:56.824265 137976285751040 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0419 03:26:56.824419 138108591912768 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776569215.062887, 'wait_for_prev_duration_secs': 6.628036499023438e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776569215.0646384, 'checkpointer_blocking_duration_secs': 1.7515275478363037, 'get_old_steps_start_time': 1776569216.8161864, 'get_old_steps_duration_secs': 2.384185791015625e-05, 'checkpoint_manager_blocking_start_time': 1776569215.061035, 'checkpoint_manager_blocking_duration_secs': 1.7633471488952637} I0419 03:26:56.824522 138108591912768 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0419 03:26:56.824571 138108591912768 max_utils.py:750] Memstats: After params initialized: I0419 03:26:56.824619 138108591912768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0419 03:26:56.824653 138108591912768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0419 03:26:56.824679 138108591912768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0419 03:26:56.824703 138108591912768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0419 03:26:57.090170 138108591912768 metric_logger.py:185] completed step: 0, seconds: 15.547, TFLOP/s/device: 0.874, Tokens/s/device: 131.730, total_weights: 65536, loss: 10.874 I0419 03:26:57.352385 138108591912768 metric_logger.py:185] completed step: 1, seconds: 2.270, TFLOP/s/device: 5.986, Tokens/s/device: 902.311, total_weights: 65536, loss: 10.874 I0419 03:26:58.181607 138108591912768 metric_logger.py:185] completed step: 2, seconds: 0.034, TFLOP/s/device: 404.246, Tokens/s/device: 60932.433, total_weights: 65536, loss: 10.262 I0419 03:26:58.411334 138108591912768 metric_logger.py:185] completed step: 3, seconds: 0.830, TFLOP/s/device: 16.377, Tokens/s/device: 2468.493, total_weights: 65536, loss: 9.731 I0419 03:26:58.870737 138108591912768 metric_logger.py:185] completed step: 4, seconds: 0.234, TFLOP/s/device: 57.971, Tokens/s/device: 8738.059, total_weights: 65536, loss: 9.272 I0419 03:26:58.876689 138108591912768 metric_logger.py:185] completed step: 5, seconds: 0.230, TFLOP/s/device: 59.186, Tokens/s/device: 8921.182, total_weights: 65536, loss: 8.887 I0419 03:27:23.073832 138108591912768 metric_logger.py:185] completed step: 6, seconds: 0.460, TFLOP/s/device: 29.541, Tokens/s/device: 4452.784, total_weights: 65536, loss: 8.588 I0419 03:27:23.304813 138108591912768 metric_logger.py:185] completed step: 7, seconds: 23.969, TFLOP/s/device: 0.567, Tokens/s/device: 85.443, total_weights: 65536, loss: 8.380 I0419 03:27:23.534358 138108591912768 metric_logger.py:185] completed step: 8, seconds: 0.234, TFLOP/s/device: 58.166, Tokens/s/device: 8767.348, total_weights: 65536, loss: 8.251 I0419 03:27:23.764627 138108591912768 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0419 03:27:23.765324 138108591912768 checkpointing.py:798] Waited 0.0007183551788330078 seconds for step 9 to finish before starting checkpointing. I0419 03:27:23.767302 138108591912768 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 03:27:25.204823 2792 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 03:27:28.509132 137985915856640 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints/0/items/array_metadatas/process_6 I0419 03:27:55.459734 137985907463936 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 26.500 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 59.60693883895874 s) (per-host) I0419 03:27:55.459838 137985907463936 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 58.643839s. I0419 03:28:05.187423 137985907463936 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 68.371406s. I0419 03:28:05.187700 137976285751040 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 03:28:05.187821 137976285751040 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0419 03:28:05.187885 137976285751040 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 03:28:05.190384 137976285751040 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 03:28:05.190582 138108591912768 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 03:28:05.190721 138108591912768 checkpoint_manager.py:1458] Waiting for previous save to complete took 41.423425 seconds. If this number is high, consider checkpointing less frequently. I0419 03:28:05.192676 138108591912768 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9 I0419 03:28:05.196170 138108591912768 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints/9. I0419 03:28:06.590702 138108591912768 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 03:28:06.590802 138108591912768 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 03:28:06.636201 138108591912768 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.046762s I0419 03:28:06.636387 138108591912768 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.099 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.7347745895385742 s) (per-host) I0419 03:28:06.636440 138108591912768 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.734837s (batch_requests_ready=0.686097s, total_serialization_initiated=0.048671s, others=0.000068s) I0419 03:28:06.636533 138108591912768 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.739084s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.739069) I0419 03:28:06.637401 137979515369216 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0419 03:28:06.637531 138108591912768 async_checkpointer.py:561] Finished blocking save. Time taken: 1.444788s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints/9. I0419 03:28:07.061962 138108591912768 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 03:28:07.062366 137976285751040 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0419 03:28:07.062543 138108591912768 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776569243.7672722, 'wait_for_prev_duration_secs': 41.4234254360199, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776569285.1927137, 'checkpointer_blocking_duration_secs': 1.4449882507324219, 'get_old_steps_start_time': 1776569286.6377213, 'get_old_steps_duration_secs': 2.4557113647460938e-05, 'checkpoint_manager_blocking_start_time': 1776569243.7655494, 'checkpoint_manager_blocking_duration_secs': 43.29695463180542} I0419 03:28:07.062661 138108591912768 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0419 03:28:07.062709 138108591912768 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 03:28:12.101981 137985915856640 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_04_int8/checkpoints/9/items/array_metadatas/process_6 I0419 03:28:48.252052 137979515369216 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.298 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.35040616989136 s) (per-host) I0419 03:28:48.252159 137979515369216 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.614549s. I0419 03:28:58.669813 137979515369216 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 52.032189s. I0419 03:28:58.670098 137976285751040 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 03:28:58.670232 137976285751040 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0419 03:28:58.670321 137976285751040 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 03:28:58.673553 137976285751040 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 03:28:58.673733 138108591912768 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 03:28:58.673883 138108591912768 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 03:28:58.674515 138108591912768 metric_logger.py:185] completed step: 9, seconds: 0.231, TFLOP/s/device: 58.890, Tokens/s/device: 8876.483, total_weights: 65536, loss: 8.175 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 03:29:09 UTC 2026 EXIT_CODE=0