2026-04-15 23:27:44.379146: 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) I0415 23:27:44.494260 135837803375744 max_utils.py:238] Skipping jax distributed system due to skip_jax_distributed_system=True flag. I0415 23:28:38.568098 135837803375744 max_utils.py:800] System Information: Jax Version: 0.9.2 I0415 23:28:38.568202 135837803375744 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0415 23:28:38.568235 135837803375744 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Apr 6 2026 20:48:10 (1775533690) cl/895581894 I0415 23:28:38.568258 135837803375744 train_utils.py:334] WARNING: 'dataset_path' might be pointing your local file system I0415 23:28:38.568279 135837803375744 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0415 23:28:38.568350 135837803375744 train.py:715] [DECOUPLED NO-OP] skipping cloud diagnostics wrapper. W0415 23:28:38.658473 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0415 23:28:39.057059 135837803375744 maxtext_utils.py:1520] Num_devices: 8, shape (1, 1, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0415 23:28:39.057328 135837803375744 checkpointing.py:677] Setting up checkpoint logger... I0415 23:28:39.057374 135837803375744 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0415 23:28:39.057413 135837803375744 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None I0415 23:28:39.057889 135837803375744 base_pytree_checkpoint_handler.py:411] 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 0x7b8a9d1f8860>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0415 23:28:41.452445 135837803375744 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0415 23:28:41.452626 135837803375744 checkpoint_manager.py:702] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b8420f35f40>}, handler_registry=None I0415 23:28:41.452901 135837803375744 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b8420f35f40>` 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`. I0415 23:28:41.452942 135837803375744 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b8420f364e0>` 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`. I0415 23:28:41.452972 135837803375744 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 0x7b8420f35f40>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b8420f35f40>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b8420f364e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b8420f364e0>}). I0415 23:28:41.453377 135837803375744 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28 I0415 23:28:41.453427 135837803375744 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>.<lambda> at 0x7b8420f394e0> timeout: 600 secs and primary_host=0 for async checkpoint writes I0415 23:28:41.579379 135837803375744 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints I0415 23:28:41.579600 135837803375744 checkpoint_manager.py:921] [process=0][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_hns=False, 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=5), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False), root_directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b8420f34470> I0415 23:28:41.579681 135837803375744 checkpointing.py:301] Checkpoint manager created! W0415 23:28:41.595674 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0415 23:28:41.781587 135837803375744 nnx_wrappers.py:437] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0415 23:28:41.781789 135837803375744 nnx_wrappers.py:437] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None). I0415 23:28:41.878656 135837803375744 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0415 23:28:41.878737 135837803375744 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None). I0415 23:28:41.892939 135837803375744 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0415 23:28:41.892988 135837803375744 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None). I0415 23:28:41.913618 135837803375744 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0415 23:28:41.913674 135837803375744 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0415 23:28:41.927771 135837803375744 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0415 23:28:41.927819 135837803375744 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0415 23:28:41.941969 135837803375744 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0415 23:28:41.942017 135837803375744 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0415 23:28:41.963748 135837803375744 attentions.py:1197] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0415 23:28:41.963804 135837803375744 attentions.py:1197] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0415 23:28:41.982174 135837803375744 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0415 23:28:41.982225 135837803375744 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, None). I0415 23:28:42.166027 135837803375744 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0415 23:28:42.166199 135837803375744 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. W0415 23:28:42.217450 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. [DECOUPLED NO-OP] gcs_storage: using stubs. [DECOUPLED NO-OP] mldiagnostics: using stub. [DECOUPLED NO-OP] mldiagnostics: using stub. [DECOUPLED NO-OP] mldiagnostics: using stub. [DECOUPLED NO-OP] workload_monitor: using stub. [DECOUPLED NO-OP] vertex_tensorboard: using stub. fsdp: 8 I0415 23:28:42.599455 135837803375744 maxtext_utils.py:1623] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0415 23:28:42.599564 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0415 23:28:42.599612 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0415 23:28:42.599658 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0415 23:28:42.599702 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0415 23:28:42.599733 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0415 23:28:42.599773 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0415 23:28:42.599813 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: P('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0415 23:28:42.599841 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0415 23:28:42.599868 135837803375744 maxtext_utils.py:1623] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0415 23:28:42.599901 135837803375744 maxtext_utils.py:1623] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed', 'vocab') Physical: ('fsdp', None) I0415 23:28:42.599936 135837803375744 maxtext_utils.py:1623] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed') Physical: (None, 'fsdp') I0415 23:28:43.025806 135837803375744 train.py:155] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0415 23:28:43.025890 135837803375744 train.py:155] train/xent Physical: float32[8,2048]............................................. ('fsdp', None). I0415 23:28:43.039386 135837803375744 train.py:162] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0415 23:28:43.039435 135837803375744 train.py:162] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None). W0415 23:28:43.728521 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0415 23:28:43.843491 135837803375744 max_utils.py:791] Total memory size: 3.6 GB, Output size: 1.5 GB, Temp size: 2.0 GB, Argument size: 1.5 GB, Host temp size: 0.0 GB. I0415 23:28:43.844067 135837803375744 max_utils.py:194] tensorboardX not available; using no-op SummaryWriter. I0415 23:28:43.844248 135837803375744 metric_logger.py:289] number parameters: 1.104 billion W0415 23:28:44.718888 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0415 23:28:44.813565 135837803375744 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0415 23:28:44.930672 135837803375744 checkpointing.py:776] Waited 0.1170952320098877 seconds for step 0 to finish before starting checkpointing. I0415 23:28:44.931258 135837803375744 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0415 23:28:44.931437 135837803375744 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 0 I0415 23:28:44.931848 135837803375744 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0. I0415 23:28:45.017330 135837803375744 signaling_client.py:373] Using ThreadSafeKeyValueSignalingClient I0415 23:28:45.102670 135730284201536 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0 I0415 23:28:45.111453 135837803375744 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array. I0415 23:28:45.111542 135837803375744 replica_slices.py:410] 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 W0415 23:28:45.327953 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.342914 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.353890 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.363605 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.375296 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.382100 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.386705 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:45.391139 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0415 23:28:45.797760 135730273715776 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/items I0415 23:28:45.927321 135730242258496 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776295725698703676, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776295725698703676, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/_CHECKPOINT_METADATA I0415 23:28:46.145213 135837803375744 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 1.034187s I0415 23:28:46.145640 135837803375744 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 10.947 GiB/s (total gbytes: 12.3 GiB) (time elapsed: a second) (per-host) I0415 23:28:46.145690 135837803375744 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 1.127365s (batch_requests_ready=0.089325s, total_serialization_initiated=1.037707s, others=0.000332s) I0415 23:28:46.145747 135837803375744 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 1.127942s (all_items=0.000029s, per_item={'items': '0.00002933'}, temp_paths=1.127912) I0415 23:28:46.146521 135730221286976 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0415 23:28:46.146615 135837803375744 async_checkpointer.py:561] Finished blocking save. Time taken: 1.215138s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0. I0415 23:28:46.146815 135837803375744 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0415 23:28:46.147005 135730294687296 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0415 23:28:46.147116 135837803375744 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776295724.9312437, 'wait_for_prev_duration_secs': 4.3392181396484375e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776295724.9314575, 'checkpointer_blocking_duration_secs': 1.215268850326538, 'get_old_steps_start_time': 1776295726.1467469, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776295724.931108, 'checkpoint_manager_blocking_duration_secs': 1.2159819602966309} I0415 23:28:46.147230 135837803375744 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0415 23:28:46.147270 135837803375744 max_utils.py:750] Memstats: After params initialized: I0415 23:28:46.147346 135837803375744 max_utils.py:756] Using (GB) 1.6 / 31.25 (5.120000%) on TPU_0(process=0,(0,0,0,0)) I0415 23:28:46.147372 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_1(process=0,(1,0,0,0)) I0415 23:28:46.147391 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_2(process=0,(0,1,0,0)) I0415 23:28:46.147412 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_3(process=0,(1,1,0,0)) I0415 23:28:46.147429 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_4(process=0,(0,2,0,0)) I0415 23:28:46.147446 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_5(process=0,(1,2,0,0)) I0415 23:28:46.147464 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_6(process=0,(0,3,0,0)) I0415 23:28:46.147480 135837803375744 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_7(process=0,(1,3,0,0)) W0415 23:28:46.153141 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:46.158966 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0415 23:28:46.163890 3373288 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0415 23:28:46.291539 3375534 google_auth_provider.cc:149] Using credentials at ~/.config/gcloud/application_default_credentials.json I0415 23:28:46.291587 3375534 google_auth_provider.cc:156] Using OAuth2 AuthProvider I0415 23:28:46.355161 135837803375744 metric_logger.py:185] completed step: 0, seconds: 0.969, TFLOP/s/device: 14.025, Tokens/s/device: 2114.032, total_weights: 16384, loss: 10.874 I0415 23:28:46.356041 135837803375744 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/tensorboard/' I0415 23:28:46.471520 135837803375744 metric_logger.py:185] completed step: 1, seconds: 1.540, TFLOP/s/device: 8.823, Tokens/s/device: 1329.879, total_weights: 16384, loss: 10.874 I0415 23:28:46.589694 135837803375744 metric_logger.py:185] completed step: 2, seconds: 0.005, TFLOP/s/device: 2530.192, Tokens/s/device: 381378.026, total_weights: 16384, loss: 10.276 I0415 23:28:46.708564 135837803375744 metric_logger.py:185] completed step: 3, seconds: 0.117, TFLOP/s/device: 116.221, Tokens/s/device: 17518.048, total_weights: 16384, loss: 9.282 I0415 23:28:46.713232 135837803375744 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0415 23:28:46.945097 135837803375744 checkpointing.py:776] Waited 0.23180246353149414 seconds for step 5 to finish before starting checkpointing. I0415 23:28:46.945485 135837803375744 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0415 23:28:47.023597 135730263230016 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/items/array_metadatas/process_0 I0415 23:29:08.060509 135730284201536 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 21.766810s (commit=21.341454s, array_metadata_write=0.425356s) I0415 23:29:08.061552 135730221286976 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 548.396 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 23 seconds) (per-host) I0415 23:29:08.061658 135730221286976 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 21.914997s. I0415 23:29:08.295327 135730221286976 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776295725698703676, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/_CHECKPOINT_METADATA I0415 23:29:08.492624 135730221286976 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0. I0415 23:29:08.693725 135730242258496 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776295725698703676, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/_CHECKPOINT_METADATA I0415 23:29:08.909839 135730221286976 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389). I0415 23:29:08.910018 135730221286976 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.567477s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/items I0415 23:29:08.910696 135730221286976 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0/items I0415 23:29:09.163164 135730221286976 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0 I0415 23:29:09.828283 135730221286976 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0`. I0415 23:29:09.828913 135730221286976 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 24.897441s. directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/0 I0415 23:29:09.829000 135730221286976 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 23.682341s. I0415 23:29:09.829198 135730294687296 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0415 23:29:09.829307 135730294687296 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0415 23:29:09.829379 135730294687296 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0415 23:29:09.829421 135730294687296 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0415 23:29:09.829559 135837803375744 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0415 23:29:09.829675 135837803375744 checkpoint_manager.py:1441] Waiting for previous save to complete took 22.884201 seconds. If this number is high, consider checkpointing less frequently. I0415 23:29:09.830399 135837803375744 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 5 I0415 23:29:09.830677 135837803375744 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5. I0415 23:29:10.006719 135837803375744 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array. I0415 23:29:10.006806 135837803375744 replica_slices.py:410] 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 I0415 23:29:10.014206 135730294687296 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5 I0415 23:29:10.724500 135729741039168 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5/items I0415 23:29:22.536987 135837803375744 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 12.530659s I0415 23:29:22.537548 135837803375744 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1001.978 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 12 seconds) (per-host) I0415 23:29:22.537614 135837803375744 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 12.611928s (batch_requests_ready=0.077842s, total_serialization_initiated=12.533644s, others=0.000442s) I0415 23:29:22.537691 135837803375744 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 12.612674s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=12.612659) I0415 23:29:22.542838 135730284201536 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0415 23:29:22.542937 135837803375744 async_checkpointer.py:561] Finished blocking save. Time taken: 12.712488s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5. I0415 23:29:22.543152 135837803375744 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0415 23:29:22.543307 135730221286976 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0415 23:29:22.543383 135837803375744 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776295726.9454522, 'wait_for_prev_duration_secs': 22.88420081138611, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776295749.8304257, 'checkpointer_blocking_duration_secs': 12.712638854980469, 'get_old_steps_start_time': 1776295762.5430803, 'get_old_steps_duration_secs': 2.765655517578125e-05, 'checkpoint_manager_blocking_start_time': 1776295726.9453876, 'checkpoint_manager_blocking_duration_secs': 35.59796929359436} I0415 23:29:22.543504 135837803375744 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0415 23:29:22.543932 135837803375744 metric_logger.py:185] completed step: 4, seconds: 0.119, TFLOP/s/device: 114.209, Tokens/s/device: 17214.858, total_weights: 16384, loss: 8.327 I0415 23:29:22.550513 135837803375744 metric_logger.py:185] completed step: 5, seconds: 0.118, TFLOP/s/device: 114.847, Tokens/s/device: 17311.041, total_weights: 16384, loss: 7.367 I0415 23:29:22.666265 135837803375744 metric_logger.py:185] completed step: 6, seconds: 35.836, TFLOP/s/device: 0.379, Tokens/s/device: 57.150, total_weights: 16384, loss: 6.459 I0415 23:29:22.784496 135837803375744 metric_logger.py:185] completed step: 7, seconds: 0.005, TFLOP/s/device: 2698.536, Tokens/s/device: 406752.731, total_weights: 16384, loss: 5.622 I0415 23:29:22.902735 135837803375744 metric_logger.py:185] completed step: 8, seconds: 0.117, TFLOP/s/device: 116.248, Tokens/s/device: 17522.095, total_weights: 16384, loss: 4.864 I0415 23:29:23.020788 135837803375744 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0415 23:29:23.021663 135837803375744 checkpointing.py:776] Waited 0.0008988380432128906 seconds for step 9 to finish before starting checkpointing. I0415 23:29:23.021902 135837803375744 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0415 23:29:23.318068 135728723920448 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5/items/array_metadatas/process_0 I0415 23:29:56.213206 135729269179968 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 33.671202s (commit=33.254920s, array_metadata_write=0.416282s) I0415 23:29:56.214375 135730284201536 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 273.000 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 46 seconds) (per-host) I0415 23:29:56.214484 135730284201536 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 33.671501s. I0415 23:29:56.641226 135730284201536 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0. I0415 23:29:57.094949 135730284201536 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389). I0415 23:29:57.095131 135730284201536 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.605542s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5/items I0415 23:29:57.095749 135730284201536 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5/items I0415 23:29:57.336794 135730284201536 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5 I0415 23:29:58.018400 135730284201536 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5`. I0415 23:29:58.019061 135730284201536 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 48.188607s. directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/5 I0415 23:29:58.019135 135730284201536 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 35.476154s. I0415 23:29:58.019241 135730221286976 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0415 23:29:58.019292 135730221286976 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0415 23:29:58.019351 135730221286976 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0415 23:29:58.019391 135730221286976 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0415 23:29:58.020281 135837803375744 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0415 23:29:58.020411 135837803375744 checkpoint_manager.py:1441] Waiting for previous save to complete took 34.998507 seconds. If this number is high, consider checkpointing less frequently. I0415 23:29:58.020641 135837803375744 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 9 I0415 23:29:58.020926 135837803375744 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9. I0415 23:29:58.226875 135730221286976 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9 I0415 23:29:58.241881 135837803375744 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array. I0415 23:29:58.241973 135837803375744 replica_slices.py:410] 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 I0415 23:29:58.960716 135730294687296 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9/items I0415 23:30:09.516524 135837803375744 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 11.275020s I0415 23:30:09.520925 135837803375744 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.084 GiB/s (total gbytes: 12.3 GiB) (time elapsed: 11 seconds) (per-host) I0415 23:30:09.521186 135837803375744 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 11.379656s (batch_requests_ready=0.096556s, total_serialization_initiated=11.278634s, others=0.004466s) I0415 23:30:09.521281 135837803375744 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 11.380395s (all_items=0.000013s, per_item={'items': '0.00001335'}, temp_paths=11.380382) I0415 23:30:09.522581 135730221286976 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0415 23:30:09.522663 135837803375744 async_checkpointer.py:561] Finished blocking save. Time taken: 11.501979s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9. I0415 23:30:09.522850 135837803375744 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0415 23:30:09.523113 135730170955328 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0415 23:30:09.523231 135837803375744 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776295763.021878, 'wait_for_prev_duration_secs': 34.998507261276245, 'time_between_consecutive_saves_sec': 13.192442893981934, 'checkpointer_blocking_start_time': 1776295798.0206642, 'checkpointer_blocking_duration_secs': 11.502109289169312, 'get_old_steps_start_time': 1776295809.5227864, 'get_old_steps_duration_secs': 2.2172927856445312e-05, 'checkpoint_manager_blocking_start_time': 1776295763.021846, 'checkpoint_manager_blocking_duration_secs': 46.50136089324951} I0415 23:30:09.523380 135837803375744 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0415 23:30:09.523414 135837803375744 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0415 23:30:10.271996 135729426466368 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9/items/array_metadatas/process_0 I0415 23:30:44.646016 135730263230016 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 35.124374s (commit=34.664195s, array_metadata_write=0.460179s) I0415 23:30:44.647077 135730221286976 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 271.726 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 46 seconds) (per-host) I0415 23:30:44.647131 135730221286976 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 35.124426s. I0415 23:30:45.111680 135730221286976 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0. I0415 23:30:45.521990 135730221286976 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389). I0415 23:30:45.522554 135730221286976 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.571660s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9/items I0415 23:30:45.523190 135730221286976 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9/items I0415 23:30:45.770224 135730221286976 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9 I0415 23:30:46.454919 135730221286976 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9`. I0415 23:30:46.455534 135730221286976 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 48.434857s. directory=gs://wanglance-maxtext/linen_ckpt_feat_migrate_nnx_utils_20260415_230805/linen_feat_migrate_nnx_utils_20260415_230805_08_checkpoint_async_true/checkpoints/9 I0415 23:30:46.455600 135730221286976 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 36.932896s. I0415 23:30:46.455781 135730170955328 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0415 23:30:46.455884 135730170955328 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0415 23:30:46.455936 135730170955328 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0415 23:30:46.455975 135730170955328 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0415 23:30:46.456477 135837803375744 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0415 23:30:46.457116 135837803375744 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0415 23:30:46.458344 135837803375744 metric_logger.py:185] completed step: 9, seconds: 0.118, TFLOP/s/device: 114.929, Tokens/s/device: 17323.341, total_weights: 16384, loss: 4.189 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops