2026-04-16 13:20:11.450892: 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) I0416 13:20:11.568583 136490028534912 max_utils.py:238] Skipping jax distributed system due to skip_jax_distributed_system=True flag. I0416 13:21:07.226608 136490028534912 max_utils.py:800] System Information: Jax Version: 0.9.2 I0416 13:21:07.226716 136490028534912 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0416 13:21:07.226751 136490028534912 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 I0416 13:21:07.226779 136490028534912 train_utils.py:334] WARNING: 'dataset_path' might be pointing your local file system I0416 13:21:07.226804 136490028534912 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0416 13:21:07.226889 136490028534912 train.py:683] [DECOUPLED NO-OP] skipping cloud diagnostics wrapper. W0416 13:21:07.318091 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0416 13:21:07.706432 136490028534912 maxtext_utils.py:1517] Num_devices: 8, shape (1, 1, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0416 13:21:07.706714 136490028534912 checkpointing.py:677] Setting up checkpoint logger... I0416 13:21:07.706756 136490028534912 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0416 13:21:07.706798 136490028534912 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None I0416 13:21:07.707305 136490028534912 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 0x7c2278ff87d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0416 13:21:10.074849 136490028534912 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0416 13:21:10.075020 136490028534912 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 0x7c1bfcd39bb0>}, handler_registry=None I0416 13:21:10.075278 136490028534912 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c1bfcd39bb0>` 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`. I0416 13:21:10.075319 136490028534912 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c1bfd5d3920>` 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`. I0416 13:21:10.075348 136490028534912 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 0x7c1bfcd39bb0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c1bfcd39bb0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c1bfd5d3920>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c1bfd5d3920>}). I0416 13:21:10.075642 136490028534912 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28 I0416 13:21:10.075695 136490028534912 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>.<lambda> at 0x7c1bfcd31260> timeout: 600 secs and primary_host=0 for async checkpoint writes I0416 13:21:10.194630 136490028534912 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints I0416 13:21:10.194843 136490028534912 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=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), root_directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c1bfcd38da0> I0416 13:21:10.194920 136490028534912 checkpointing.py:301] Checkpoint manager created! W0416 13:21:10.213817 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0416 13:21:10.392606 136490028534912 nnx_wrappers.py:437] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0416 13:21:10.392705 136490028534912 nnx_wrappers.py:437] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None). I0416 13:21:10.485459 136490028534912 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0416 13:21:10.485533 136490028534912 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None). I0416 13:21:10.499555 136490028534912 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0416 13:21:10.499610 136490028534912 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None). I0416 13:21:10.520060 136490028534912 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0416 13:21:10.520116 136490028534912 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0416 13:21:10.534176 136490028534912 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0416 13:21:10.534223 136490028534912 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0416 13:21:10.548261 136490028534912 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0416 13:21:10.548310 136490028534912 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0416 13:21:10.569635 136490028534912 attentions.py:1197] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0416 13:21:10.569692 136490028534912 attentions.py:1197] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None). I0416 13:21:10.587721 136490028534912 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0416 13:21:10.587771 136490028534912 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, None). I0416 13:21:10.770667 136490028534912 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0416 13:21:10.770779 136490028534912 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. W0416 13:21:10.820011 3865786 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 I0416 13:21:11.203372 136490028534912 maxtext_utils.py:1620] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0416 13:21:11.203535 136490028534912 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0416 13:21:11.203575 136490028534912 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0416 13:21:11.203618 136490028534912 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0416 13:21:11.203656 136490028534912 maxtext_utils.py:1620] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0416 13:21:11.203683 136490028534912 maxtext_utils.py:1620] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0416 13:21:11.203721 136490028534912 maxtext_utils.py:1620] 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) I0416 13:21:11.203759 136490028534912 maxtext_utils.py:1620] 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') I0416 13:21:11.203787 136490028534912 maxtext_utils.py:1620] 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) I0416 13:21:11.203813 136490028534912 maxtext_utils.py:1620] 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) I0416 13:21:11.203856 136490028534912 maxtext_utils.py:1620] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0416 13:21:11.203891 136490028534912 maxtext_utils.py:1620] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0416 13:21:11.620248 136490028534912 train.py:155] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0416 13:21:11.620328 136490028534912 train.py:155] train/xent Physical: float32[8,2048]............................................. ('fsdp', None). I0416 13:21:11.633711 136490028534912 train.py:162] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0416 13:21:11.633759 136490028534912 train.py:162] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None). W0416 13:21:12.270015 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0416 13:21:12.386724 136490028534912 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. I0416 13:21:12.387201 136490028534912 max_utils.py:194] tensorboardX not available; using no-op SummaryWriter. I0416 13:21:12.387343 136490028534912 metric_logger.py:289] number parameters: 1.104 billion W0416 13:21:13.211897 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0416 13:21:13.307523 136490028534912 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0416 13:21:13.424787 136490028534912 checkpointing.py:776] Waited 0.11725449562072754 seconds for step 0 to finish before starting checkpointing. I0416 13:21:13.425227 136490028534912 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0416 13:21:13.425374 136490028534912 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 0 I0416 13:21:13.425723 136490028534912 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0. I0416 13:21:13.511824 136490028534912 signaling_client.py:373] Using ThreadSafeKeyValueSignalingClient I0416 13:21:13.604227 136382590748224 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0 I0416 13:21:13.609343 136490028534912 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array. I0416 13:21:13.609427 136490028534912 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 W0416 13:21:13.823544 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.834137 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.840991 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.847713 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.856699 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.870993 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.877625 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:13.882736 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0416 13:21:14.262889 136382580262464 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/items I0416 13:21:14.421967 136382548805184 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776345674181516227, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776345674181516227, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0416 13:21:14.748732 3867898 google_auth_provider.cc:149] Using credentials at ~/.config/gcloud/application_default_credentials.json I0416 13:21:14.748778 3867898 google_auth_provider.cc:156] Using OAuth2 AuthProvider I0416 13:21:15.100820 136490028534912 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 1.491897s I0416 13:21:15.105053 136490028534912 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 7.751 GiB/s (total gbytes: 12.3 GiB) (time elapsed: a second) (per-host) I0416 13:21:15.105111 136490028534912 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 1.592338s (batch_requests_ready=0.092467s, total_serialization_initiated=1.495743s, others=0.004128s) I0416 13:21:15.105187 136490028534912 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 1.592879s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=1.592860) I0416 13:21:15.106432 136382601233984 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0416 13:21:15.106511 136490028534912 async_checkpointer.py:561] Finished blocking save. Time taken: 1.681097s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0. I0416 13:21:15.106684 136490028534912 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0416 13:21:15.106838 136382485890624 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0416 13:21:15.106925 136490028534912 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776345673.425215, 'wait_for_prev_duration_secs': 4.1484832763671875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776345673.425396, 'checkpointer_blocking_duration_secs': 1.6812145709991455, 'get_old_steps_start_time': 1776345675.1066272, 'get_old_steps_duration_secs': 2.09808349609375e-05, 'checkpoint_manager_blocking_start_time': 1776345673.4251351, 'checkpoint_manager_blocking_duration_secs': 1.681764841079712} I0416 13:21:15.107033 136490028534912 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0416 13:21:15.107089 136490028534912 max_utils.py:750] Memstats: After params initialized: I0416 13:21:15.107132 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_0(process=0,(0,0,0,0)) I0416 13:21:15.107156 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_1(process=0,(1,0,0,0)) I0416 13:21:15.107175 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_2(process=0,(0,1,0,0)) I0416 13:21:15.107193 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_3(process=0,(1,1,0,0)) I0416 13:21:15.107211 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_4(process=0,(0,2,0,0)) I0416 13:21:15.107227 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_5(process=0,(1,2,0,0)) I0416 13:21:15.107242 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_6(process=0,(0,3,0,0)) I0416 13:21:15.107259 136490028534912 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_7(process=0,(1,3,0,0)) W0416 13:21:15.113186 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:15.119415 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. W0416 13:21:15.123902 3865786 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions. I0416 13:21:15.497435 136490028534912 metric_logger.py:185] completed step: 0, seconds: 0.920, TFLOP/s/device: 14.773, Tokens/s/device: 2226.770, total_weights: 16384, loss: 10.874 I0416 13:21:15.498467 136490028534912 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/tensorboard/' I0416 13:21:15.613368 136490028534912 metric_logger.py:185] completed step: 1, seconds: 2.188, TFLOP/s/device: 6.210, Tokens/s/device: 936.020, total_weights: 16384, loss: 10.874 I0416 13:21:15.732118 136490028534912 metric_logger.py:185] completed step: 2, seconds: 0.007, TFLOP/s/device: 2038.885, Tokens/s/device: 307322.929, total_weights: 16384, loss: 9.775 I0416 13:21:15.850559 136490028534912 metric_logger.py:185] completed step: 3, seconds: 0.117, TFLOP/s/device: 116.435, Tokens/s/device: 17550.324, total_weights: 16384, loss: 8.838 I0416 13:21:15.905882 136382559290944 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/items/array_metadatas/process_0 I0416 13:21:35.331115 136490028534912 metric_logger.py:185] completed step: 4, seconds: 0.119, TFLOP/s/device: 114.542, Tokens/s/device: 17265.071, total_weights: 16384, loss: 7.965 I0416 13:21:35.338801 136490028534912 metric_logger.py:185] completed step: 5, seconds: 0.119, TFLOP/s/device: 114.532, Tokens/s/device: 17263.470, total_weights: 16384, loss: 7.233 I0416 13:21:35.455435 136490028534912 metric_logger.py:185] completed step: 6, seconds: 19.482, TFLOP/s/device: 0.697, Tokens/s/device: 105.124, total_weights: 16384, loss: 6.676 I0416 13:21:35.573585 136490028534912 metric_logger.py:185] completed step: 7, seconds: 0.004, TFLOP/s/device: 3024.066, Tokens/s/device: 455820.165, total_weights: 16384, loss: 6.293 I0416 13:21:35.692185 136490028534912 metric_logger.py:185] completed step: 8, seconds: 0.117, TFLOP/s/device: 115.673, Tokens/s/device: 17435.426, total_weights: 16384, loss: 6.055 I0416 13:21:35.810429 136490028534912 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0416 13:21:35.811232 136490028534912 checkpointing.py:776] Waited 0.0008275508880615234 seconds for step 9 to finish before starting checkpointing. I0416 13:21:35.811494 136490028534912 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0416 13:21:40.507004 136382506862144 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 25.401262s (commit=24.957105s, array_metadata_write=0.444157s) I0416 13:21:40.508171 136382601233984 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 468.109 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 26 seconds) (per-host) I0416 13:21:40.508227 136382601233984 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 25.401674s. I0416 13:21:40.747349 136382601233984 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776345674181516227, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0416 13:21:40.924656 136382601233984 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0. I0416 13:21:41.124352 136382548805184 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776345674181516227, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/_CHECKPOINT_METADATA I0416 13:21:41.382868 136382601233984 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389). I0416 13:21:41.383456 136382601233984 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.598157s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/items I0416 13:21:41.384100 136382601233984 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0/items I0416 13:21:41.624567 136382601233984 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0 I0416 13:21:42.300615 136382601233984 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0`. I0416 13:21:42.301246 136382601233984 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 28.875839s. directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/0 I0416 13:21:42.301319 136382601233984 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 27.194767s. I0416 13:21:42.301999 136382485890624 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0416 13:21:42.302116 136382485890624 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0416 13:21:42.302171 136382485890624 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0416 13:21:42.302217 136382485890624 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0416 13:21:42.302462 136490028534912 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. W0416 13:21:42.302591 136490028534912 checkpoint_manager.py:1441] Waiting for previous save to complete took 6.491101 seconds. If this number is high, consider checkpointing less frequently. I0416 13:21:42.303403 136490028534912 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 9 I0416 13:21:42.303687 136490028534912 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9. I0416 13:21:42.473476 136382485890624 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9 I0416 13:21:42.478020 136490028534912 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array. I0416 13:21:42.478127 136490028534912 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 I0416 13:21:43.214889 136382475404864 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9/items I0416 13:21:52.070982 136490028534912 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 9.593357s I0416 13:21:52.075127 136490028534912 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.274 GiB/s (total gbytes: 12.3 GiB) (time elapsed: 9 seconds) (per-host) I0416 13:21:52.075188 136490028534912 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 9.687510s (batch_requests_ready=0.086987s, total_serialization_initiated=9.596500s, others=0.004023s) I0416 13:21:52.075257 136490028534912 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 9.688268s (all_items=0.000021s, per_item={'items': '0.00002146'}, temp_paths=9.688247) I0416 13:21:52.076604 136382506862144 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0416 13:21:52.076686 136490028534912 async_checkpointer.py:561] Finished blocking save. Time taken: 9.773238s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9. I0416 13:21:52.076877 136490028534912 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0416 13:21:52.077067 136382601233984 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0416 13:21:52.077179 136490028534912 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776345695.8114567, 'wait_for_prev_duration_secs': 6.491101264953613, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776345702.3034272, 'checkpointer_blocking_duration_secs': 9.773362874984741, 'get_old_steps_start_time': 1776345712.0768065, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776345695.8114154, 'checkpoint_manager_blocking_duration_secs': 16.265739679336548} I0416 13:21:52.077300 136490028534912 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0416 13:21:52.077333 136490028534912 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0416 13:21:52.852734 136382714480192 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9/items/array_metadatas/process_0 I0416 13:22:26.291769 136382475404864 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 34.215964s (commit=33.788675s, array_metadata_write=0.427289s) I0416 13:22:26.292729 136382506862144 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 287.821 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 43 seconds) (per-host) I0416 13:22:26.292784 136382506862144 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 34.216055s. I0416 13:22:26.705035 136382506862144 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0. I0416 13:22:27.163176 136382506862144 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389). I0416 13:22:27.163787 136382506862144 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.597432s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9/items I0416 13:22:27.164445 136382506862144 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9/items I0416 13:22:27.393591 136382506862144 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9 I0416 13:22:28.054452 136382506862144 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9`. I0416 13:22:28.055079 136382506862144 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 45.751637s. directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_13_scan_layers_true/checkpoints/9 I0416 13:22:28.055161 136382506862144 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 35.978433s. I0416 13:22:28.055341 136382601233984 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0416 13:22:28.055448 136382601233984 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0416 13:22:28.055499 136382601233984 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0416 13:22:28.055541 136382601233984 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0416 13:22:28.055826 136490028534912 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. I0416 13:22:28.055969 136490028534912 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0416 13:22:28.056567 136490028534912 metric_logger.py:185] completed step: 9, seconds: 0.118, TFLOP/s/device: 114.828, Tokens/s/device: 17308.115, total_weights: 16384, loss: 5.915 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops