XPK Start: Fri Apr 17 03:26:30 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-17 03:26:54.244614: 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) I0417 03:26:54.770900 139826229937984 max_utils.py:273] Attempting to initialize the jax distributed system... I0417 03:27:03.810579 139826229937984 distributed.py:149] Starting JAX distributed service on [::]:8482 I0417 03:27:03.812912 139826229937984 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-false-cmlqq-slice-job-0-0.mt-10-shardy-false-cmlqq:8482 I0417 03:27:04.742444 139826229937984 max_utils.py:284] Jax distributed system initialized! I0417 03:27:10.835656 139826229937984 max_utils.py:800] System Information: Jax Version: 0.9.2 I0417 03:27:10.835756 139826229937984 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0417 03:27:10.835794 139826229937984 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365 I0417 03:27:10.835830 139826229937984 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0417 03:27:11.504470 139826229937984 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0417 03:27:11.505062 139826229937984 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0417 03:27:11.505237 139826229937984 checkpointing.py:688] Setting up checkpoint logger... I0417 03:27:11.505287 139826229937984 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0417 03:27:11.505331 139826229937984 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0417 03:27:11.505654 139826229937984 base_pytree_checkpoint_handler.py:441] 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 0x7f2b1d048b60>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0417 03:27:14.539578 139826229937984 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0417 03:27:14.539804 139826229937984 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f2a113227b0>}, handler_registry=None I0417 03:27:14.540054 139826229937984 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f2a113227b0>` 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`. I0417 03:27:14.540106 139826229937984 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f16f0197d10>` 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`. I0417 03:27:14.540142 139826229937984 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 0x7f2a113227b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f2a113227b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f16f0197d10>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f16f0197d10>}). I0417 03:27:14.540473 139826229937984 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0417 03:27:14.540540 139826229937984 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f16d027c680> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0417 03:27:16.022349 139826229937984 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints I0417 03:27:16.886935 139826229937984 checkpoint_manager.py:929] [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_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f16f0192870> I0417 03:27:16.887107 139826229937984 checkpointing.py:302] Checkpoint manager created! I0417 03:27:17.794050 139826229937984 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0417 03:27:17.794151 139826229937984 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 03:27:18.174777 139826229937984 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 03:27:18.174900 139826229937984 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 03:27:18.191079 139826229937984 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 03:27:18.191135 139826229937984 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 03:27:18.214644 139826229937984 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 03:27:18.214709 139826229937984 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 03:27:18.231025 139826229937984 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 03:27:18.231086 139826229937984 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 03:27:18.247203 139826229937984 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 03:27:18.247260 139826229937984 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 03:27:18.272341 139826229937984 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0417 03:27:18.272406 139826229937984 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 03:27:18.292979 139826229937984 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0417 03:27:18.293041 139826229937984 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0417 03:27:18.506574 139826229937984 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0417 03:27:18.506680 139826229937984 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0417 03:27:19.889113 139826229937984 maxtext_utils.py:1796] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0417 03:27:19.889252 139826229937984 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0417 03:27:19.889307 139826229937984 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0417 03:27:19.889366 139826229937984 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0417 03:27:19.889417 139826229937984 maxtext_utils.py:1796] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0417 03:27:19.889456 139826229937984 maxtext_utils.py:1796] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0417 03:27:19.889508 139826229937984 maxtext_utils.py:1796] 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) I0417 03:27:19.889560 139826229937984 maxtext_utils.py:1796] 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') I0417 03:27:19.889602 139826229937984 maxtext_utils.py:1796] 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) I0417 03:27:19.889639 139826229937984 maxtext_utils.py:1796] 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) I0417 03:27:19.889685 139826229937984 maxtext_utils.py:1796] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0417 03:27:19.889732 139826229937984 maxtext_utils.py:1796] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0417 03:27:20.383424 139826229937984 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0417 03:27:20.383521 139826229937984 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0417 03:27:20.404511 139826229937984 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0417 03:27:20.404613 139826229937984 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0417 03:27:31.240714 139826229937984 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. I0417 03:27:31.248314 139826229937984 metric_logger.py:289] number parameters: 1.104 billion I0417 03:27:43.152150 139826229937984 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0417 03:27:43.305146 139826229937984 checkpointing.py:798] Waited 0.15297222137451172 seconds for step 0 to finish before starting checkpointing. I0417 03:27:43.307780 139826229937984 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 03:27:43.309453 139826229937984 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0 I0417 03:27:43.311182 139826229937984 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0. I0417 03:27:43.643499 139826229937984 signaling_client.py:364] Using JaxDistributedSignalingClient I0417 03:27:43.657696 139826229937984 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 03:27:43.657779 139826229937984 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 I0417 03:27:43.895747 139696124217088 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0 I0417 03:27:43.932044 139826229937984 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.275130s I0417 03:27:43.932931 139826229937984 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.483 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2813529968261719 s) (per-host) I0417 03:27:43.933008 139826229937984 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.281452s (batch_requests_ready=0.002064s, total_serialization_initiated=0.278579s, others=0.000808s) I0417 03:27:43.933090 139826229937984 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285447s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.285430) I0417 03:27:43.933953 139826229937984 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.62 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0. I0417 03:27:43.934242 139696132609792 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 03:47:43.934207 I0417 03:27:43.940288 139826229937984 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0417 03:27:43.940559 139694012626688 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 03:27:43.940682 139826229937984 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776396463.3077514, 'wait_for_prev_duration_secs': 0.00012993812561035156, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776396463.3095357, 'checkpointer_blocking_duration_secs': 0.6248538494110107, 'get_old_steps_start_time': 1776396463.9344146, 'get_old_steps_duration_secs': 2.7179718017578125e-05, 'checkpoint_manager_blocking_start_time': 1776396463.3057032, 'checkpoint_manager_blocking_duration_secs': 0.6349406242370605} I0417 03:27:43.940783 139826229937984 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0417 03:27:43.940829 139826229937984 max_utils.py:750] Memstats: After params initialized: I0417 03:27:43.940890 139826229937984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0417 03:27:43.940928 139826229937984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0417 03:27:43.940955 139826229937984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0417 03:27:43.940979 139826229937984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0417 03:27:44.286956 139826229937984 metric_logger.py:185] completed step: 0, seconds: 11.595, TFLOP/s/device: 1.172, Tokens/s/device: 176.629, total_weights: 65536, loss: 10.874 I0417 03:27:44.289525 139826229937984 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/tensorboard/' I0417 03:27:44.692136 139826229937984 metric_logger.py:185] completed step: 1, seconds: 1.133, TFLOP/s/device: 11.990, Tokens/s/device: 1807.201, total_weights: 65536, loss: 10.874 I0417 03:27:44.838067 139826229937984 metric_logger.py:185] completed step: 2, seconds: 0.406, TFLOP/s/device: 33.474, Tokens/s/device: 5045.603, total_weights: 65536, loss: 10.262 I0417 03:27:44.985387 139826229937984 metric_logger.py:185] completed step: 3, seconds: 0.009, TFLOP/s/device: 1597.358, Tokens/s/device: 240771.220, total_weights: 65536, loss: 9.734 I0417 03:27:46.176658 139696107431680 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/items I0417 03:27:46.367279 139693960681216 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776396465924886111, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776396465924886111, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA I0417 03:27:47.338126 2853 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0417 03:27:49.488878 139695057004288 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/items/array_metadatas/process_0 I0417 03:28:07.143525 139826229937984 metric_logger.py:185] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.589, Tokens/s/device: 13956.087, total_weights: 65536, loss: 9.277 I0417 03:28:07.152495 139826229937984 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.275, Tokens/s/device: 13908.697, total_weights: 65536, loss: 8.892 I0417 03:28:07.298442 139826229937984 metric_logger.py:185] completed step: 6, seconds: 22.158, TFLOP/s/device: 0.613, Tokens/s/device: 92.428, total_weights: 65536, loss: 8.592 I0417 03:28:07.445894 139826229937984 metric_logger.py:185] completed step: 7, seconds: 0.007, TFLOP/s/device: 2053.367, Tokens/s/device: 309505.818, total_weights: 65536, loss: 8.384 I0417 03:28:07.593253 139826229937984 metric_logger.py:185] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.215, Tokens/s/device: 13899.635, total_weights: 65536, loss: 8.255 I0417 03:28:07.740180 139826229937984 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0417 03:28:07.740860 139826229937984 checkpointing.py:798] Waited 0.0006804466247558594 seconds for step 9 to finish before starting checkpointing. I0417 03:28:07.743010 139826229937984 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 03:28:25.196893 139695036012288 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.859099s (commit=36.260412s, array_metadata_write=1.598687s) I0417 03:28:25.198003 139696132609792 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 38.020 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.54638361930847 s) (per-host) I0417 03:28:25.198130 139696132609792 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.263779s. I0417 03:28:25.947665 139696132609792 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776396465924886111, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA I0417 03:28:26.759974 139693960681216 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776396465924886111, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA I0417 03:28:27.072890 139696132609792 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000271s. I0417 03:28:27.457358 139696132609792 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 03:28:28.393436 139696132609792 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.317787s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/items I0417 03:28:28.395216 139696132609792 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0/items I0417 03:28:28.939111 139696132609792 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0 I0417 03:28:30.785191 139696132609792 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0`. I0417 03:28:30.785917 139696132609792 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 47.48 seconds @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/0 I0417 03:28:30.787224 139696132609792 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 46.852871s. I0417 03:28:30.787450 139694012626688 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 03:28:30.787572 139694012626688 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 03:28:30.787637 139694012626688 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0417 03:28:30.790428 139694012626688 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0417 03:28:30.790564 139826229937984 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0417 03:28:30.790653 139826229937984 checkpoint_manager.py:1452] Waiting for previous save to complete took 23.047646 seconds. If this number is high, consider checkpointing less frequently. I0417 03:28:30.792546 139826229937984 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9 I0417 03:28:30.794659 139826229937984 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9. I0417 03:28:31.497408 139826229937984 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 03:28:31.497500 139826229937984 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 I0417 03:28:31.523312 139826229937984 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026614s I0417 03:28:31.524214 139826229937984 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 49.507 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.031158447265625 s) (per-host) I0417 03:28:31.524292 139826229937984 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031255s (batch_requests_ready=0.001828s, total_serialization_initiated=0.028602s, others=0.000826s) I0417 03:28:31.524378 139826229937984 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035129s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.035114) I0417 03:28:31.525076 139826229937984 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.73 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9. I0417 03:28:31.525358 139694012626688 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 03:48:31.525331 I0417 03:28:31.533622 139826229937984 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0417 03:28:31.533930 139695086368512 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 03:28:31.534107 139826229937984 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776396487.7429788, 'wait_for_prev_duration_secs': 23.047646284103394, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776396510.7925856, 'checkpointer_blocking_duration_secs': 0.7328884601593018, 'get_old_steps_start_time': 1776396511.525498, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776396487.7410965, 'checkpoint_manager_blocking_duration_secs': 23.792969226837158} I0417 03:28:31.534216 139826229937984 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0417 03:28:31.534260 139826229937984 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 03:28:32.159798 139696132609792 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9 I0417 03:28:34.316231 139697217066752 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9/items I0417 03:28:37.644252 139697194645248 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9/items/array_metadatas/process_0 I0417 03:29:16.977653 139696107431680 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.523384s (commit=40.342228s, array_metadata_write=1.181156s) I0417 03:29:16.978660 139694012626688 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.727 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.485578536987305 s) (per-host) I0417 03:29:16.978781 139694012626688 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.453328s. I0417 03:29:19.164949 139694012626688 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 03:29:19.381695 139694012626688 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000227s. I0417 03:29:20.258247 139694012626688 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.917434s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9/items I0417 03:29:20.259887 139694012626688 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9/items I0417 03:29:21.700655 139694012626688 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9 I0417 03:29:23.557984 139694012626688 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9`. I0417 03:29:23.558806 139694012626688 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 52.77 seconds @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_10_shardy_false/checkpoints/9 I0417 03:29:23.560235 139694012626688 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.034780s. I0417 03:29:23.560443 139695086368512 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 03:29:23.560560 139695086368512 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 03:29:23.560624 139695086368512 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0417 03:29:23.563446 139695086368512 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0417 03:29:23.563577 139826229937984 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0417 03:29:23.563673 139826229937984 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 03:29:23.564374 139826229937984 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.389, Tokens/s/device: 13925.815, total_weights: 65536, loss: 8.179 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Fri Apr 17 03:29:36 UTC 2026 EXIT_CODE=0