XPK Start: Thu Apr 23 19:15:14 UTC 2026 2026-04-23 19:15:18.151506: 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 00:00:1776971718.164377 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776971718.168112 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776971718.179668 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776971718.179689 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776971718.179692 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776971718.179694 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 19:15:37.915638: 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) I0423 19:15:38.443674 139244345337664 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 19:15:47,485:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 19:15:47.485767 139244345337664 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 19:15:47,488:jax._src.distributed:157: Connecting to JAX distributed service on mt-04-int8-dnoup-slice-job-0-0.mt-04-int8-dnoup:8482 I0423 19:15:47.488181 139244345337664 distributed.py:157] Connecting to JAX distributed service on mt-04-int8-dnoup-slice-job-0-0.mt-04-int8-dnoup:8482 I0423 19:15:48.529618 139244345337664 max_utils.py:284] Jax distributed system initialized! I0423 19:15:55.292390 139244345337664 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 19:15:55.292503 139244345337664 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 19:15:55.292545 139244345337664 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 I0423 19:15:55.292578 139244345337664 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 19:15:55.947137 139244345337664 maxtext_utils.py:1517] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 19:15:55.947560 139244345337664 checkpointing.py:677] Setting up checkpoint logger... I0423 19:15:55.947618 139244345337664 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 19:15:55.947662 139244345337664 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 19:15:55.948009 139244345337664 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 0x7ea3d7d71880>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 19:15:58.820516 139244345337664 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0423 19:15:58.820760 139244345337664 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 0x7e9087601d60>}, handler_registry=None I0423 19:15:58.820997 139244345337664 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e9087601d60>` 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`. I0423 19:15:58.821045 139244345337664 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9086fc3410>` 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`. I0423 19:15:58.821081 139244345337664 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 0x7e9087601d60>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e9087601d60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9086fc3410>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9086fc3410>}). I0423 19:15:58.821618 139244345337664 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 19:15:58.821696 139244345337664 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e8f681fa840> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 19:16:00.034173 139244345337664 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints I0423 19:16:00.468547 139244345337664 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_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e9087ae7d70> I0423 19:16:00.468727 139244345337664 checkpointing.py:301] Checkpoint manager created! I0423 19:16:01.656512 139244345337664 nnx_wrappers.py:442] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 19:16:01.656628 139244345337664 nnx_wrappers.py:442] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 19:16:02.025416 139244345337664 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 19:16:02.025513 139244345337664 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 19:16:02.047296 139244345337664 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 19:16:02.047404 139244345337664 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 19:16:02.127021 139244345337664 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 19:16:02.127129 139244345337664 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 19:16:02.143048 139244345337664 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 19:16:02.143122 139244345337664 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 19:16:02.158967 139244345337664 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 19:16:02.159031 139244345337664 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 19:16:02.187566 139244345337664 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 19:16:02.187668 139244345337664 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 19:16:02.269420 139244345337664 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 19:16:02.269511 139244345337664 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 19:16:02.871454 139244345337664 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0423 19:16:02.871576 139244345337664 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 19:16:05.555644 139244345337664 maxtext_utils.py:1620] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0423 19:16:05.555778 139244345337664 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0423 19:16:05.555834 139244345337664 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0423 19:16:05.555893 139244345337664 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0423 19:16:05.555946 139244345337664 maxtext_utils.py:1620] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0423 19:16:05.555985 139244345337664 maxtext_utils.py:1620] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0423 19:16:05.556039 139244345337664 maxtext_utils.py:1620] 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) I0423 19:16:05.556106 139244345337664 maxtext_utils.py:1620] 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') I0423 19:16:05.556153 139244345337664 maxtext_utils.py:1620] 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) I0423 19:16:05.556192 139244345337664 maxtext_utils.py:1620] 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) I0423 19:16:05.556243 139244345337664 maxtext_utils.py:1620] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0423 19:16:05.556298 139244345337664 maxtext_utils.py:1620] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0423 19:16:07.273229 139244345337664 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 19:16:07.273335 139244345337664 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0423 19:16:07.293218 139244345337664 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 19:16:07.293313 139244345337664 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0423 19:16:21.409548 139244345337664 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. I0423 19:16:21.410484 139244345337664 metric_logger.py:289] number parameters: 1.104 billion I0423 19:16:36.563765 139244345337664 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0423 19:16:36.803613 139244345337664 checkpointing.py:776] Waited 0.23982787132263184 seconds for step 0 to finish before starting checkpointing. I0423 19:16:36.806157 139244345337664 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 19:16:36.808343 139244345337664 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0 I0423 19:16:36.811469 139244345337664 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints/0. I0423 19:16:38.271861 139244345337664 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 19:16:38.273076 139244345337664 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 19:16:38.273144 139244345337664 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 I0423 19:16:38.523083 139244345337664 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.251302s I0423 19:16:38.523267 139244345337664 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.650 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.9348530769348145 s) (per-host) I0423 19:16:38.523320 139244345337664 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.934916s (batch_requests_ready=0.679911s, total_serialization_initiated=0.254934s, others=0.000071s) I0423 19:16:38.523411 139244345337664 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.939450s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.939431) I0423 19:16:38.524443 139121859204864 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0423 19:16:38.524622 139244345337664 async_checkpointer.py:561] Finished blocking save. Time taken: 1.716202s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints/0. I0423 19:16:38.526512 139244345337664 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 19:16:38.526750 139111671613184 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 19:16:38.526876 139244345337664 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776971796.8061385, 'wait_for_prev_duration_secs': 6.532669067382812e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776971796.8083847, 'checkpointer_blocking_duration_secs': 1.7163503170013428, 'get_old_steps_start_time': 1776971798.5247564, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776971796.8041782, 'checkpoint_manager_blocking_duration_secs': 1.722661018371582} I0423 19:16:38.527034 139244345337664 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0423 19:16:38.527086 139244345337664 max_utils.py:750] Memstats: After params initialized: I0423 19:16:38.527155 139244345337664 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0423 19:16:38.527191 139244345337664 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0423 19:16:38.527220 139244345337664 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0423 19:16:38.527245 139244345337664 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0423 19:16:38.799757 139244345337664 metric_logger.py:185] completed step: 0, seconds: 15.153, TFLOP/s/device: 0.897, Tokens/s/device: 135.153, total_weights: 65536, loss: 10.874 I0423 19:16:39.051534 139244345337664 metric_logger.py:185] completed step: 1, seconds: 2.235, TFLOP/s/device: 6.080, Tokens/s/device: 916.480, total_weights: 65536, loss: 10.874 I0423 19:16:39.466678 139244345337664 metric_logger.py:185] completed step: 2, seconds: 0.023, TFLOP/s/device: 586.866, Tokens/s/device: 88458.880, total_weights: 65536, loss: 10.262 I0423 19:16:39.697280 139244345337664 metric_logger.py:185] completed step: 3, seconds: 0.416, TFLOP/s/device: 32.697, Tokens/s/device: 4928.444, total_weights: 65536, loss: 9.731 I0423 19:16:40.156689 139244345337664 metric_logger.py:185] completed step: 4, seconds: 0.234, TFLOP/s/device: 58.023, Tokens/s/device: 8745.858, total_weights: 65536, loss: 9.272 I0423 19:16:40.162501 139244345337664 metric_logger.py:185] completed step: 5, seconds: 0.231, TFLOP/s/device: 58.922, Tokens/s/device: 8881.295, total_weights: 65536, loss: 8.887 I0423 19:17:02.610497 2805 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 19:17:02.865443 139244345337664 metric_logger.py:185] completed step: 6, seconds: 0.460, TFLOP/s/device: 29.547, Tokens/s/device: 4453.694, total_weights: 65536, loss: 8.588 I0423 19:17:03.096224 139244345337664 metric_logger.py:185] completed step: 7, seconds: 22.470, TFLOP/s/device: 0.605, Tokens/s/device: 91.144, total_weights: 65536, loss: 8.380 I0423 19:17:03.325792 139244345337664 metric_logger.py:185] completed step: 8, seconds: 0.240, TFLOP/s/device: 56.499, Tokens/s/device: 8516.194, total_weights: 65536, loss: 8.251 I0423 19:17:03.554993 139244345337664 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0423 19:17:03.555683 139244345337664 checkpointing.py:776] Waited 0.0007114410400390625 seconds for step 9 to finish before starting checkpointing. I0423 19:17:03.557669 139244345337664 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 19:17:05.050979 139121867597568 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints/0/items/array_metadatas/process_6 I0423 19:17:32.868080 139121859204864 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 28.575 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 55.279622316360474 s) (per-host) I0423 19:17:32.868215 139121859204864 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 54.343658s. I0423 19:17:41.430538 139121859204864 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 62.905964s. I0423 19:17:41.430860 139111671613184 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 19:17:41.430989 139111671613184 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 19:17:41.431039 139111671613184 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 19:17:41.432405 139111671613184 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 19:17:41.432617 139244345337664 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. W0423 19:17:41.432747 139244345337664 checkpoint_manager.py:1458] Waiting for previous save to complete took 37.875083 seconds. If this number is high, consider checkpointing less frequently. I0423 19:17:41.434532 139244345337664 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9 I0423 19:17:41.437843 139244345337664 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints/9. I0423 19:17:42.508446 139244345337664 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 19:17:42.508550 139244345337664 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 I0423 19:17:42.551789 139244345337664 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044378s I0423 19:17:42.551955 139244345337664 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.840 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.8385241031646729 s) (per-host) I0423 19:17:42.552004 139244345337664 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.838582s (batch_requests_ready=0.792121s, total_serialization_initiated=0.046396s, others=0.000066s) I0423 19:17:42.552117 139244345337664 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.842632s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.842616) I0423 19:17:42.552997 139114930239232 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0423 19:17:42.553141 139244345337664 async_checkpointer.py:561] Finished blocking save. Time taken: 1.118531s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints/9. I0423 19:17:42.560823 139244345337664 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 19:17:42.561060 139111671613184 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 19:17:42.561201 139244345337664 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776971823.5576398, 'wait_for_prev_duration_secs': 37.875083208084106, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776971861.4345794, 'checkpointer_blocking_duration_secs': 1.1187384128570557, 'get_old_steps_start_time': 1776971862.553338, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776971823.5559087, 'checkpoint_manager_blocking_duration_secs': 39.00525760650635} I0423 19:17:42.561314 139244345337664 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0423 19:17:42.561360 139244345337664 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 19:17:47.547968 139121867597568 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_04_int8/checkpoints/9/items/array_metadatas/process_6 I0423 19:18:23.990304 139114930239232 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.363 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.27684807777405 s) (per-host) I0423 19:18:23.990419 139114930239232 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.437199s. I0423 19:18:32.061242 139114930239232 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 49.508006s. I0423 19:18:32.061589 139111671613184 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 19:18:32.061709 139111671613184 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 19:18:32.061758 139111671613184 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 19:18:32.063312 139111671613184 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 19:18:32.063511 139244345337664 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. I0423 19:18:32.063660 139244345337664 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 19:18:32.064385 139244345337664 metric_logger.py:185] completed step: 9, seconds: 0.230, TFLOP/s/device: 59.015, Tokens/s/device: 8895.336, 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: Thu Apr 23 19:18:44 UTC 2026 EXIT_CODE=0