XPK Start: Fri Apr 24 17:00:01 UTC 2026 2026-04-24 17:00:05.166546: 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 00:00:1777050005.179299 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1777050005.182961 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1777050005.194099 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777050005.194118 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777050005.194121 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777050005.194123 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-24 17:00:24.305462: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0424 17:00:24.834488 135320449079104 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-24 17:00:33,876:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0424 17:00:33.876533 135320449079104 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-24 17:00:33,878:jax._src.distributed:157: Connecting to JAX distributed service on mt-10-shardy-false-mhqop-slice-job-0-0.mt-10-shardy-false-mhqop:8482 I0424 17:00:33.878831 135320449079104 distributed.py:157] Connecting to JAX distributed service on mt-10-shardy-false-mhqop-slice-job-0-0.mt-10-shardy-false-mhqop:8482 I0424 17:00:35.149523 135320449079104 max_utils.py:284] Jax distributed system initialized! I0424 17:00:40.875017 135320449079104 max_utils.py:800] System Information: Jax Version: 0.8.1 I0424 17:00:40.875121 135320449079104 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0424 17:00:40.875161 135320449079104 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 I0424 17:00:40.875195 135320449079104 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0424 17:00:41.497364 135320449079104 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0424 17:00:41.612858 135320449079104 checkpointing.py:688] Setting up checkpoint logger... I0424 17:00:41.612977 135320449079104 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0424 17:00:41.613025 135320449079104 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0424 17:00:41.613226 135320449079104 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 0x7b12427751f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0424 17:00:45.073257 135320449079104 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0424 17:00:45.073499 135320449079104 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 0x7b000095b710>}, handler_registry=None I0424 17:00:45.073746 135320449079104 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b000095b710>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0424 17:00:45.073799 135320449079104 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7afdd435b920>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0424 17:00:45.073837 135320449079104 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 0x7b000095b710>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b000095b710>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7afdd435b920>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7afdd435b920>}). I0424 17:00:45.074181 135320449079104 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0424 17:00:45.074260 135320449079104 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7afdd446cf40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0424 17:00:45.770113 135320449079104 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints I0424 17:00:45.792754 135320449079104 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_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7afdd435bc80> I0424 17:00:45.792880 135320449079104 checkpointing.py:302] Checkpoint manager created! I0424 17:00:46.721205 135320449079104 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0424 17:00:46.721321 135320449079104 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0424 17:00:48.141268 135320449079104 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0424 17:00:48.141386 135320449079104 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0424 17:00:48.141443 135320449079104 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0424 17:00:48.141489 135320449079104 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0424 17:00:48.141529 135320449079104 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0424 17:00:48.141566 135320449079104 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0424 17:00:48.141604 135320449079104 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0424 17:00:48.141643 135320449079104 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0424 17:00:48.141694 135320449079104 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0424 17:00:48.141730 135320449079104 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0424 17:00:48.141762 135320449079104 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0424 17:00:48.141795 135320449079104 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0424 17:00:48.540523 135320449079104 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 17:00:48.540694 135320449079104 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 17:00:48.546078 135320449079104 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 17:00:48.546174 135320449079104 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 17:00:48.701107 135320449079104 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 17:00:48.701244 135320449079104 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 17:00:48.854478 135320449079104 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 17:00:48.854609 135320449079104 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 17:00:49.014712 135320449079104 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 17:00:49.014867 135320449079104 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 17:00:49.165117 135320449079104 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 17:00:49.165259 135320449079104 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 17:00:49.313560 135320449079104 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 17:00:49.313707 135320449079104 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 17:00:49.474036 135320449079104 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0424 17:00:49.474172 135320449079104 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 17:00:49.502563 135320449079104 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0424 17:00:49.502712 135320449079104 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0424 17:01:00.489512 135320449079104 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0424 17:01:00.491905 135320449079104 metric_logger.py:289] number parameters: 1.104 billion I0424 17:01:11.579591 135320449079104 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0424 17:01:11.731031 135320449079104 checkpointing.py:798] Waited 0.1514263153076172 seconds for step 0 to finish before starting checkpointing. I0424 17:01:11.733380 135320449079104 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 17:01:11.735146 135320449079104 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0424 17:01:11.737973 135320449079104 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints/0. I0424 17:01:12.331892 135320449079104 signaling_client.py:364] Using JaxDistributedSignalingClient I0424 17:01:12.334418 135320449079104 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0424 17:01:12.334478 135320449079104 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0424 17:01:12.580070 135320449079104 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249293s I0424 17:01:12.580248 135320449079104 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.022 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5105063915252686 s) (per-host) I0424 17:01:12.580303 135320449079104 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.510571s (batch_requests_ready=0.254161s, total_serialization_initiated=0.256338s, others=0.000073s) I0424 17:01:12.580403 135320449079104 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.514798s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.514780) I0424 17:01:12.581433 135193348269824 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0424 17:01:12.581562 135320449079104 async_checkpointer.py:561] Finished blocking save. Time taken: 0.846346s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints/0. I0424 17:01:12.859714 135320449079104 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0424 17:01:12.860089 135194927437568 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0424 17:01:12.860260 135320449079104 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777050071.7333627, 'wait_for_prev_duration_secs': 6.29425048828125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777050071.7351842, 'checkpointer_blocking_duration_secs': 0.8465654850006104, 'get_old_steps_start_time': 1777050072.5817692, 'get_old_steps_duration_secs': 2.384185791015625e-05, 'checkpoint_manager_blocking_start_time': 1777050071.7314277, 'checkpoint_manager_blocking_duration_secs': 1.1287903785705566} I0424 17:01:12.860374 135320449079104 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0424 17:01:12.860450 135320449079104 max_utils.py:750] Memstats: After params initialized: I0424 17:01:12.860505 135320449079104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0424 17:01:12.860540 135320449079104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0424 17:01:12.860568 135320449079104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0424 17:01:12.860594 135320449079104 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0424 17:01:13.128756 135320449079104 metric_logger.py:185] completed step: 0, seconds: 11.087, TFLOP/s/device: 1.226, Tokens/s/device: 184.727, total_weights: 65536, loss: 10.874 I0424 17:01:13.293573 135320449079104 metric_logger.py:185] completed step: 1, seconds: 1.547, TFLOP/s/device: 8.781, Tokens/s/device: 1323.594, total_weights: 65536, loss: 10.874 I0424 17:01:13.747346 135320449079104 metric_logger.py:185] completed step: 2, seconds: 0.019, TFLOP/s/device: 709.844, Tokens/s/device: 106995.455, total_weights: 65536, loss: 10.262 I0424 17:01:13.894831 135320449079104 metric_logger.py:185] completed step: 3, seconds: 0.444, TFLOP/s/device: 30.588, Tokens/s/device: 4610.505, total_weights: 65536, loss: 9.732 I0424 17:01:14.190322 135320449079104 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.123, Tokens/s/device: 12529.136, total_weights: 65536, loss: 9.274 I0424 17:01:14.197769 135320449079104 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.418, Tokens/s/device: 13930.172, total_weights: 65536, loss: 8.888 I0424 17:01:33.596474 135320449079104 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.892, Tokens/s/device: 6917.306, total_weights: 65536, loss: 8.589 I0424 17:01:33.744047 135320449079104 metric_logger.py:185] completed step: 7, seconds: 19.252, TFLOP/s/device: 0.706, Tokens/s/device: 106.376, total_weights: 65536, loss: 8.380 I0424 17:01:33.891719 135320449079104 metric_logger.py:185] completed step: 8, seconds: 0.154, TFLOP/s/device: 88.037, Tokens/s/device: 13269.921, total_weights: 65536, loss: 8.251 I0424 17:01:34.040112 135320449079104 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0424 17:01:34.040718 135320449079104 checkpointing.py:798] Waited 0.0006165504455566406 seconds for step 9 to finish before starting checkpointing. I0424 17:01:34.043163 135320449079104 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 17:01:34.455900 2644 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0424 17:01:36.889522 135194935830272 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_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints/0/items/array_metadatas/process_5 I0424 17:02:07.712513 135193348269824 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 28.388 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 55.64273023605347 s) (per-host) I0424 17:02:07.712630 135193348269824 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 55.130991s. I0424 17:02:16.819009 135193348269824 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 64.237354s. I0424 17:02:16.819330 135194927437568 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 17:02:16.819447 135194927437568 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0424 17:02:16.819497 135194927437568 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0424 17:02:16.820875 135194927437568 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0424 17:02:16.821058 135320449079104 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. W0424 17:02:16.821197 135320449079104 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.778038 seconds. If this number is high, consider checkpointing less frequently. I0424 17:02:16.823179 135320449079104 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0424 17:02:16.826596 135320449079104 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints/9. I0424 17:02:17.420009 135320449079104 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0424 17:02:17.420161 135320449079104 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0424 17:02:17.467258 135320449079104 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.049195s I0424 17:02:17.467429 135320449079104 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.381 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3520979881286621 s) (per-host) I0424 17:02:17.467482 135320449079104 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.352160s (batch_requests_ready=0.298195s, total_serialization_initiated=0.053895s, others=0.000070s) I0424 17:02:17.467581 135320449079104 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.356088s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.356072) I0424 17:02:17.468477 135193876748032 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0424 17:02:17.468632 135320449079104 async_checkpointer.py:561] Finished blocking save. Time taken: 0.645383s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints/9. I0424 17:02:17.470494 135320449079104 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0424 17:02:17.470740 135194927437568 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0424 17:02:17.470869 135320449079104 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777050094.043134, 'wait_for_prev_duration_secs': 42.778037786483765, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777050136.8232195, 'checkpointer_blocking_duration_secs': 0.6455316543579102, 'get_old_steps_start_time': 1777050137.4687698, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1777050094.0409381, 'checkpoint_manager_blocking_duration_secs': 43.42989730834961} I0424 17:02:17.470985 135320449079104 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0424 17:02:17.471030 135320449079104 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 17:02:22.440503 135194935830272 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_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_10_shardy_false/checkpoints/9/items/array_metadatas/process_5 I0424 17:02:58.712554 135193876748032 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.974 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.597200870513916 s) (per-host) I0424 17:02:58.712671 135193876748032 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.244084s. I0424 17:03:06.466635 135193876748032 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 48.998033s. I0424 17:03:06.466863 135194927437568 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 17:03:06.466929 135194927437568 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0424 17:03:06.466974 135194927437568 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0424 17:03:06.468632 135194927437568 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0424 17:03:06.468808 135320449079104 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. I0424 17:03:06.468977 135320449079104 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 17:03:06.469639 135320449079104 metric_logger.py:185] completed step: 9, seconds: 0.146, TFLOP/s/device: 92.836, Tokens/s/device: 13993.277, 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: Fri Apr 24 17:03:16 UTC 2026 EXIT_CODE=0