XPK Start: Sat Apr 25 16:45:45 UTC 2026 2026-04-25 16:45:49.817906: 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 00:00:1777135549.830728 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1777135549.834401 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1777135549.845781 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777135549.845798 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777135549.845800 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777135549.845802 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-25 16:46:09.006814: 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) I0425 16:46:09.533733 137889738295104 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-25 16:46:18,574:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0425 16:46:18.574292 137889738295104 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-25 16:46:18,576:jax._src.distributed:157: Connecting to JAX distributed service on mt-06-grad-accum-ay96a-slice-job-0-0.mt-06-grad-accum-ay96a:8482 I0425 16:46:18.576830 137889738295104 distributed.py:157] Connecting to JAX distributed service on mt-06-grad-accum-ay96a-slice-job-0-0.mt-06-grad-accum-ay96a:8482 I0425 16:46:19.614596 137889738295104 max_utils.py:284] Jax distributed system initialized! I0425 16:46:26.433846 137889738295104 max_utils.py:800] System Information: Jax Version: 0.8.1 I0425 16:46:26.433952 137889738295104 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0425 16:46:26.433994 137889738295104 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 I0425 16:46:26.434027 137889738295104 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0425 16:46:27.053582 137889738295104 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 16:46:27.168680 137889738295104 checkpointing.py:688] Setting up checkpoint logger... I0425 16:46:27.168801 137889738295104 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0425 16:46:27.168846 137889738295104 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0425 16:46:27.169058 137889738295104 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 0x7d6864775310>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 16:46:30.026440 137889738295104 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0425 16:46:30.026686 137889738295104 checkpoint_manager.py:709] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d543c5fc680>}, handler_registry=None I0425 16:46:30.026921 137889738295104 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d543c5fc680>` 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`. I0425 16:46:30.026970 137889738295104 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d5604806960>` 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`. I0425 16:46:30.027006 137889738295104 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 0x7d543c5fc680>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d543c5fc680>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d5604806960>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d5604806960>}). I0425 16:46:30.027349 137889738295104 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0425 16:46:30.027426 137889738295104 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d5604724f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0425 16:46:31.091557 137889738295104 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints I0425 16:46:31.101421 137889738295104 checkpoint_manager.py:929] [process=5][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/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d56067ace30> I0425 16:46:31.101544 137889738295104 checkpointing.py:302] Checkpoint manager created! I0425 16:46:32.853816 137889738295104 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0425 16:46:32.853932 137889738295104 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0425 16:46:34.294157 137889738295104 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0425 16:46:34.294279 137889738295104 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0425 16:46:34.294332 137889738295104 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0425 16:46:34.294377 137889738295104 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0425 16:46:34.294415 137889738295104 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0425 16:46:34.294450 137889738295104 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0425 16:46:34.294487 137889738295104 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 16:46:34.294523 137889738295104 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0425 16:46:34.294557 137889738295104 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 16:46:34.294592 137889738295104 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 16:46:34.294627 137889738295104 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0425 16:46:34.294674 137889738295104 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0425 16:46:34.674220 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048]............................................... Unknown. I0425 16:46:34.674317 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048]............................................... (None,). I0425 16:46:34.688775 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,16,7168]....................................... Unknown. I0425 16:46:34.688833 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,16,7168]....................................... ('fsdp', None, None). I0425 16:46:34.717223 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[7168,16,2048]....................................... Unknown. I0425 16:46:34.717286 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[7168,16,2048]....................................... (None, None, 'fsdp'). I0425 16:46:34.731330 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,16]............................................ Unknown. I0425 16:46:34.731384 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,16]............................................ (None, None). I0425 16:46:34.759451 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,16,16,128]..................................... Unknown. I0425 16:46:34.759521 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,16,16,128]..................................... ('fsdp', None, None, None). I0425 16:46:34.773773 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[16,16,128,2048]..................................... Unknown. I0425 16:46:34.773828 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[16,16,128,2048]..................................... (None, None, None, 'fsdp'). I0425 16:46:34.816188 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,32000]......................................... Unknown. I0425 16:46:34.816257 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,32000]......................................... ('fsdp', None). I0425 16:46:34.830305 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[32000,2048]......................................... Unknown. I0425 16:46:34.830367 137889738295104 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[32000,2048]......................................... (None, 'fsdp'). I0425 16:46:35.047851 137889738295104 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 16:46:35.047940 137889738295104 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:46:35.051932 137889738295104 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 16:46:35.051988 137889738295104 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:46:35.068100 137889738295104 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 16:46:35.068160 137889738295104 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:46:35.083710 137889738295104 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 16:46:35.083770 137889738295104 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 16:46:35.107324 137889738295104 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 16:46:35.107394 137889738295104 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:46:35.123328 137889738295104 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 16:46:35.123386 137889738295104 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:46:35.139112 137889738295104 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 16:46:35.139170 137889738295104 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:46:35.164003 137889738295104 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0425 16:46:35.164072 137889738295104 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 16:46:35.183465 137889738295104 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0425 16:46:35.183527 137889738295104 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0425 16:46:46.790335 137889738295104 max_utils.py:791] Total memory size: 1.7 GB, Output size: 0.4 GB, Temp size: 1.3 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0425 16:46:46.792730 137889738295104 metric_logger.py:289] number parameters: 1.104 billion I0425 16:46:48.669862 137889738295104 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0425 16:46:58.921121 137889738295104 checkpointing.py:798] Waited 10.25124478340149 seconds for step 0 to finish before starting checkpointing. I0425 16:46:58.923385 137889738295104 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 16:46:58.925177 137889738295104 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0425 16:46:58.928009 137889738295104 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints/0. I0425 16:46:59.498522 137889738295104 signaling_client.py:364] Using JaxDistributedSignalingClient I0425 16:46:59.499746 137889738295104 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 16:46:59.499806 137889738295104 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 I0425 16:46:59.737135 137889738295104 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.239740s I0425 16:46:59.737302 137889738295104 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.120 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.4944319725036621 s) (per-host) I0425 16:46:59.737352 137889738295104 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.494491s (batch_requests_ready=0.247877s, total_serialization_initiated=0.246548s, others=0.000066s) I0425 16:46:59.737444 137889738295104 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.498800s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.498780) I0425 16:46:59.738432 137767803655936 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0425 16:46:59.738604 137889738295104 async_checkpointer.py:561] Finished blocking save. Time taken: 0.813355s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints/0. I0425 16:47:00.195970 137889738295104 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0425 16:47:00.196349 137764087523072 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0425 16:47:00.196517 137889738295104 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777135618.923367, 'wait_for_prev_duration_secs': 6.341934204101562e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777135618.925216, 'checkpointer_blocking_duration_secs': 0.8135108947753906, 'get_old_steps_start_time': 1777135619.7387476, 'get_old_steps_duration_secs': 2.6226043701171875e-05, 'checkpoint_manager_blocking_start_time': 1777135618.9215577, 'checkpoint_manager_blocking_duration_secs': 1.2749192714691162} I0425 16:47:00.196626 137889738295104 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0425 16:47:00.196708 137889738295104 max_utils.py:750] Memstats: After params initialized: I0425 16:47:00.196766 137889738295104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0425 16:47:00.196800 137889738295104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0425 16:47:00.196828 137889738295104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0425 16:47:00.196851 137889738295104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0425 16:47:00.454581 137889738295104 metric_logger.py:185] completed step: 0, seconds: 1.876, TFLOP/s/device: 28.969, Tokens/s/device: 4366.545, total_weights: 262144, loss: 10.876 I0425 16:47:01.072558 137889738295104 metric_logger.py:185] completed step: 1, seconds: 11.783, TFLOP/s/device: 4.612, Tokens/s/device: 695.232, total_weights: 262144, loss: 10.876 I0425 16:47:01.651085 137889738295104 metric_logger.py:185] completed step: 2, seconds: 0.041, TFLOP/s/device: 1313.242, Tokens/s/device: 197946.116, total_weights: 262144, loss: 10.561 I0425 16:47:02.229445 137889738295104 metric_logger.py:185] completed step: 3, seconds: 0.584, TFLOP/s/device: 93.101, Tokens/s/device: 14033.188, total_weights: 262144, loss: 10.270 I0425 16:47:02.964976 2392 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0425 16:47:03.386802 137889738295104 metric_logger.py:185] completed step: 4, seconds: 0.579, TFLOP/s/device: 93.864, Tokens/s/device: 14148.165, total_weights: 262144, loss: 10.021 I0425 16:47:03.393809 137889738295104 metric_logger.py:185] completed step: 5, seconds: 0.578, TFLOP/s/device: 94.032, Tokens/s/device: 14173.550, total_weights: 262144, loss: 9.819 I0425 16:47:04.870854 137767812048640 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints/0/items/array_metadatas/process_5 I0425 16:47:37.811672 137767803655936 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 40.955 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.568753242492676 s) (per-host) I0425 16:47:37.811789 137767803655936 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.073245s. I0425 16:47:40.967795 137889738295104 metric_logger.py:185] completed step: 6, seconds: 1.158, TFLOP/s/device: 46.949, Tokens/s/device: 7076.594, total_weights: 262144, loss: 9.666 I0425 16:47:41.546223 137889738295104 metric_logger.py:185] completed step: 7, seconds: 36.997, TFLOP/s/device: 1.469, Tokens/s/device: 221.425, total_weights: 262144, loss: 9.560 I0425 16:47:42.124595 137889738295104 metric_logger.py:185] completed step: 8, seconds: 0.584, TFLOP/s/device: 93.001, Tokens/s/device: 14018.156, total_weights: 262144, loss: 9.495 I0425 16:47:42.704094 137889738295104 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0425 16:47:42.704714 137889738295104 checkpointing.py:798] Waited 0.0006320476531982422 seconds for step 9 to finish before starting checkpointing. I0425 16:47:42.706531 137889738295104 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 16:48:19.346405 137767803655936 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 79.607844s. I0425 16:48:19.346688 137764087523072 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 16:48:19.346807 137764087523072 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0425 16:48:19.346859 137764087523072 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0425 16:48:19.352010 137764087523072 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0425 16:48:19.352180 137889738295104 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0425 16:48:19.352312 137889738295104 checkpoint_manager.py:1458] Waiting for previous save to complete took 36.645785 seconds. If this number is high, consider checkpointing less frequently. I0425 16:48:19.354592 137889738295104 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0425 16:48:19.357885 137889738295104 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints/9. I0425 16:48:20.187252 137889738295104 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 16:48:20.187359 137889738295104 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 I0425 16:48:20.236849 137889738295104 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.051855s I0425 16:48:20.237016 137889738295104 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.136 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3003566265106201 s) (per-host) I0425 16:48:20.237065 137889738295104 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.300414s (batch_requests_ready=0.243991s, total_serialization_initiated=0.056358s, others=0.000065s) I0425 16:48:20.237157 137889738295104 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.304560s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.304543) I0425 16:48:20.237982 137763517122304 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0425 16:48:20.238090 137889738295104 async_checkpointer.py:561] Finished blocking save. Time taken: 0.883425s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints/9. I0425 16:48:20.434726 137889738295104 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0425 16:48:20.435111 137764087523072 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0425 16:48:20.435286 137889738295104 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777135662.7065032, 'wait_for_prev_duration_secs': 36.64578461647034, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777135699.354631, 'checkpointer_blocking_duration_secs': 0.883570671081543, 'get_old_steps_start_time': 1777135700.2382195, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1777135662.7049391, 'checkpoint_manager_blocking_duration_secs': 37.730309009552} I0425 16:48:20.435396 137889738295104 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0425 16:48:20.435440 137889738295104 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 16:48:25.037922 137767812048640 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_06_grad_accum/checkpoints/9/items/array_metadatas/process_5 I0425 16:49:02.376014 137763517122304 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.220 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.439316511154175 s) (per-host) I0425 16:49:02.376128 137763517122304 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.138086s. I0425 16:49:08.350701 137763517122304 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 48.112643s. I0425 16:49:08.350993 137764087523072 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 16:49:08.351122 137764087523072 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0425 16:49:08.351172 137764087523072 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0425 16:49:08.352631 137764087523072 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0425 16:49:08.352816 137889738295104 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0425 16:49:08.352977 137889738295104 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 16:49:08.353682 137889738295104 metric_logger.py:185] completed step: 9, seconds: 0.578, TFLOP/s/device: 94.014, Tokens/s/device: 14170.755, total_weights: 262144, loss: 9.457 Per train step: Total TFLOPs: 54.35 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sat Apr 25 16:49:20 UTC 2026 EXIT_CODE=0