XPK Start: Fri Apr 24 14:00:02 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. `rope_parameters`'s factor field must be a float >= 1, got 40 `rope_parameters`'s beta_fast field must be a float, got 32 `rope_parameters`'s beta_slow field must be a float, got 1 DeepseekV32Config got `key=rope_scaling` in kwargs but hasn't set it as attribute. For RoPE standardization you need to set `self.rope_parameters` in model's config. 2026-04-24 14:00:27.807305: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0424 14:00:28.017516 132729113945920 max_utils.py:273] Attempting to initialize the jax distributed system... I0424 14:00:37.057011 132729113945920 distributed.py:149] Starting JAX distributed service on [::]:8482 I0424 14:00:37.059472 132729113945920 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-false-0sx8w-slice-job-0-0.mt-10-shardy-false-0sx8w:8482 I0424 14:00:37.920384 132729113945920 max_utils.py:284] Jax distributed system initialized! I0424 14:00:44.058767 132729113945920 max_utils.py:800] System Information: Jax Version: 0.9.2 I0424 14:00:44.058872 132729113945920 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0424 14:00:44.058912 132729113945920 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 I0424 14:00:44.058954 132729113945920 train_utils.py:391] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0424 14:00:44.753818 132729113945920 maxtext_utils.py:1771] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0424 14:00:44.867124 132729113945920 checkpointing.py:688] Setting up checkpoint logger... I0424 14:00:44.867241 132729113945920 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0424 14:00:44.867288 132729113945920 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0424 14:00:44.867486 132729113945920 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 0x78b6fa7c26c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0424 14:00:47.836518 132729113945920 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0424 14:00:47.836777 132729113945920 checkpoint_manager.py:708] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78a2804394f0>}, handler_registry=None I0424 14:00:47.837021 132729113945920 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78a2804394f0>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0424 14:00:47.837072 132729113945920 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78a2b02b1280>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0424 14:00:47.837108 132729113945920 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 0x78a2804394f0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78a2804394f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78a2b02b1280>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78a2b02b1280>}). I0424 14:00:47.837435 132729113945920 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0424 14:00:47.837506 132729113945920 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x78a28042b600> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0424 14:00:48.657476 132729113945920 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints I0424 14:00:48.850151 132729113945920 checkpoint_manager.py:929] [process=5][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78a2b01c3020> I0424 14:00:48.850330 132729113945920 checkpointing.py:302] Checkpoint manager created! I0424 14:00:49.821667 132729113945920 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0424 14:00:49.821785 132729113945920 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0424 14:00:51.158713 132729113945920 maxtext_utils.py:1889] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0424 14:00:51.158830 132729113945920 maxtext_utils.py:1889] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0424 14:00:51.158882 132729113945920 maxtext_utils.py:1889] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0424 14:00:51.158927 132729113945920 maxtext_utils.py:1889] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0424 14:00:51.158966 132729113945920 maxtext_utils.py:1889] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0424 14:00:51.159001 132729113945920 maxtext_utils.py:1889] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0424 14:00:51.159038 132729113945920 maxtext_utils.py:1889] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0424 14:00:51.159075 132729113945920 maxtext_utils.py:1889] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0424 14:00:51.159109 132729113945920 maxtext_utils.py:1889] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0424 14:00:51.159144 132729113945920 maxtext_utils.py:1889] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0424 14:00:51.159178 132729113945920 maxtext_utils.py:1889] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0424 14:00:51.159209 132729113945920 maxtext_utils.py:1889] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0424 14:00:51.562027 132729113945920 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 14:00:51.562120 132729113945920 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 14:00:51.567458 132729113945920 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 14:00:51.567516 132729113945920 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 14:00:51.584070 132729113945920 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 14:00:51.584131 132729113945920 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 14:00:51.599950 132729113945920 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 14:00:51.600011 132729113945920 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0424 14:00:51.623768 132729113945920 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 14:00:51.623838 132729113945920 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 14:00:51.639762 132729113945920 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 14:00:51.639823 132729113945920 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 14:00:51.655681 132729113945920 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 14:00:51.655739 132729113945920 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 14:00:51.686335 132729113945920 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0424 14:00:51.686407 132729113945920 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0424 14:00:51.706171 132729113945920 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0424 14:00:51.706237 132729113945920 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0424 14:01:03.018199 132729113945920 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0424 14:01:03.020584 132729113945920 metric_logger.py:301] number parameters: 1.104 billion I0424 14:01:14.160492 132729113945920 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0424 14:01:14.312257 132729113945920 checkpointing.py:798] Waited 0.1517488956451416 seconds for step 0 to finish before starting checkpointing. I0424 14:01:14.314476 132729113945920 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 14:01:14.316055 132729113945920 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 0 I0424 14:01:14.317512 132729113945920 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints/0. I0424 14:01:14.624452 132729113945920 signaling_client.py:364] Using JaxDistributedSignalingClient I0424 14:01:14.625549 132729113945920 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0424 14:01:14.625626 132729113945920 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0424 14:01:14.903590 132729113945920 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279938s I0424 14:01:14.903778 132729113945920 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.316 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2901613712310791 s) (per-host) I0424 14:01:14.903834 132729113945920 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.290228s (batch_requests_ready=0.003392s, total_serialization_initiated=0.286761s, others=0.000075s) I0424 14:01:14.903937 132729113945920 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.294472s (all_items=0.000022s, per_item={'items': '0.00002241'}, temp_paths=0.294450) I0424 14:01:14.904861 132729113945920 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.59 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints/0. I0424 14:01:14.905160 132600823006976 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 14:21:14.905137 I0424 14:01:14.908434 132729113945920 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0424 14:01:14.908727 132599221561088 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0424 14:01:14.908882 132729113945920 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777039274.314459, 'wait_for_prev_duration_secs': 6.008148193359375e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777039274.3160942, 'checkpointer_blocking_duration_secs': 0.5891721248626709, 'get_old_steps_start_time': 1777039274.905294, 'get_old_steps_duration_secs': 2.8371810913085938e-05, 'checkpoint_manager_blocking_start_time': 1777039274.312638, 'checkpoint_manager_blocking_duration_secs': 0.5962059497833252} I0424 14:01:14.908990 132729113945920 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0424 14:01:14.909053 132729113945920 max_utils.py:750] Memstats: After params initialized: I0424 14:01:14.909100 132729113945920 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0424 14:01:14.909132 132729113945920 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0424 14:01:14.909159 132729113945920 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0424 14:01:14.909184 132729113945920 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0424 14:01:15.224204 132729113945920 metric_logger.py:196] completed step: 0, seconds: 11.139, TFLOP/s/device: 1.220, Tokens/s/device: 183.861, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469 I0424 14:01:15.405210 132729113945920 metric_logger.py:196] completed step: 1, seconds: 1.062, TFLOP/s/device: 12.797, Tokens/s/device: 1928.880, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469 I0424 14:01:15.852176 132729113945920 metric_logger.py:196] completed step: 2, seconds: 0.035, TFLOP/s/device: 390.244, Tokens/s/device: 58821.840, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28632.133 I0424 14:01:15.999839 132729113945920 metric_logger.py:196] completed step: 3, seconds: 0.437, TFLOP/s/device: 31.097, Tokens/s/device: 4687.207, total_weights: 65536, loss: 9.732, lm_loss: 9.732, perplexity: 16855.789 I0424 14:01:16.295833 132729113945920 metric_logger.py:196] completed step: 4, seconds: 0.164, TFLOP/s/device: 82.707, Tokens/s/device: 12466.445, total_weights: 65536, loss: 9.274, lm_loss: 9.274, perplexity: 10661.594 I0424 14:01:16.303300 132729113945920 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.244, Tokens/s/device: 13904.070, total_weights: 65536, loss: 8.888, lm_loss: 8.888, perplexity: 7245.978 I0424 14:01:18.013399 2860 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0424 14:01:19.884754 132599748318976 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints/0/items/array_metadatas/process_5 I0424 14:01:38.368976 132729113945920 metric_logger.py:196] completed step: 6, seconds: 0.297, TFLOP/s/device: 45.811, Tokens/s/device: 6905.062, total_weights: 65536, loss: 8.589, lm_loss: 8.589, perplexity: 5371.156 I0424 14:01:38.516389 132729113945920 metric_logger.py:196] completed step: 7, seconds: 21.919, TFLOP/s/device: 0.620, Tokens/s/device: 93.433, total_weights: 65536, loss: 8.380, lm_loss: 8.380, perplexity: 4360.465 I0424 14:01:38.663918 132729113945920 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.862, Tokens/s/device: 13394.200, total_weights: 65536, loss: 8.251, lm_loss: 8.251, perplexity: 3831.256 I0424 14:01:38.811877 132729113945920 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0424 14:01:38.812525 132729113945920 checkpointing.py:798] Waited 0.0006611347198486328 seconds for step 9 to finish before starting checkpointing. I0424 14:01:38.814691 132729113945920 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 14:01:51.494203 132600823006976 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 42.830 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.88053560256958 s) (per-host) I0424 14:01:51.494324 132600823006976 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.589087s. I0424 14:02:00.675436 132600823006976 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 45.770182s. I0424 14:02:00.675755 132599221561088 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 14:02:00.675876 132599221561088 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0424 14:02:00.675925 132599221561088 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0424 14:02:00.679324 132599221561088 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0424 14:02:00.679513 132729113945920 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0424 14:02:00.679664 132729113945920 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.864978 seconds. If this number is high, consider checkpointing less frequently. I0424 14:02:00.681532 132729113945920 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 9 I0424 14:02:00.683433 132729113945920 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints/9. I0424 14:02:00.984249 132729113945920 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0424 14:02:00.984358 132729113945920 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False I0424 14:02:01.022897 132729113945920 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.040271s I0424 14:02:01.023032 132729113945920 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 32.501 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04746270179748535 s) (per-host) I0424 14:02:01.023082 132729113945920 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.047521s (batch_requests_ready=0.002884s, total_serialization_initiated=0.044573s, others=0.000065s) I0424 14:02:01.023189 132729113945920 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.051783s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.051767) I0424 14:02:01.023924 132729113945920 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints/9. I0424 14:02:01.024192 132599221561088 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 14:22:01.024169 I0424 14:02:01.026020 132729113945920 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0424 14:02:01.026288 132599748318976 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0424 14:02:01.026421 132729113945920 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777039298.8146434, 'wait_for_prev_duration_secs': 21.864977836608887, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777039320.6815708, 'checkpointer_blocking_duration_secs': 0.34271788597106934, 'get_old_steps_start_time': 1777039321.0243137, 'get_old_steps_duration_secs': 2.9087066650390625e-05, 'checkpoint_manager_blocking_start_time': 1777039298.8127644, 'checkpoint_manager_blocking_duration_secs': 22.213623046875} I0424 14:02:01.026582 132729113945920 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0424 14:02:01.026630 132729113945920 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 14:02:06.310303 132575406307072 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260424_120657/nnx_xpk_feat_nnx_post_train_fixes_20260424_120657_10_shardy_false/checkpoints/9/items/array_metadatas/process_5 I0424 14:02:42.490592 132599221561088 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 38.049 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.51496720314026 s) (per-host) I0424 14:02:42.490735 132599221561088 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.466468s. I0424 14:02:50.410156 132599221561088 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.385874s. I0424 14:02:50.410434 132599748318976 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 14:02:50.410554 132599748318976 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0424 14:02:50.410602 132599748318976 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0424 14:02:50.413394 132599748318976 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0424 14:02:50.413578 132729113945920 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0424 14:02:50.413769 132729113945920 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 14:02:50.414686 132729113945920 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 91.884, Tokens/s/device: 13849.816, total_weights: 65536, loss: 8.175, lm_loss: 8.175, perplexity: 3551.438 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Fri Apr 24 14:03:02 UTC 2026 EXIT_CODE=0