XPK Start: Fri Apr 24 20:40:15 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 20:40:39.842866: 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 20:40:40.052618 137890040932160 max_utils.py:273] Attempting to initialize the jax distributed system... I0424 20:40:49.093982 137890040932160 distributed.py:149] Starting JAX distributed service on [::]:8482 I0424 20:40:49.096253 137890040932160 distributed.py:172] Connecting to JAX distributed service on mt-09-pdb-lt-1-ru7va-slice-job-0-0.mt-09-pdb-lt-1-ru7va:8482 I0424 20:40:50.391107 137890040932160 max_utils.py:284] Jax distributed system initialized! I0424 20:40:56.571909 137890040932160 max_utils.py:800] System Information: Jax Version: 0.9.2 I0424 20:40:56.572018 137890040932160 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0424 20:40:56.572058 137890040932160 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 20:40:56.572095 137890040932160 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0424 20:40:57.268785 137890040932160 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0424 20:40:57.269067 137890040932160 checkpointing.py:677] Setting up checkpoint logger... I0424 20:40:57.269121 137890040932160 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0424 20:40:57.269167 137890040932160 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0424 20:40:57.269507 137890040932160 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 0x7d6868717230>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0424 20:41:00.181264 137890040932160 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0424 20:41:00.181498 137890040932160 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 0x7d545012f8c0>}, handler_registry=None I0424 20:41:00.181749 137890040932160 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d545012f8c0>` 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 20:41:00.181798 137890040932160 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d5450133320>` 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 20:41:00.181835 137890040932160 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 0x7d545012f8c0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d545012f8c0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d5450133320>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d5450133320>}). I0424 20:41:00.182171 137890040932160 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0424 20:41:00.182241 137890040932160 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d532c40b560> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0424 20:41:02.126060 137890040932160 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints I0424 20:41:02.128223 137890040932160 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/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d545012f800> I0424 20:41:02.128340 137890040932160 checkpointing.py:301] Checkpoint manager created! I0424 20:41:03.064878 137890040932160 nnx_wrappers.py:453] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 20:41:03.064998 137890040932160 nnx_wrappers.py:453] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0424 20:41:03.447872 137890040932160 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 20:41:03.447969 137890040932160 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0424 20:41:03.464353 137890040932160 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 20:41:03.464413 137890040932160 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0424 20:41:03.488007 137890040932160 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 20:41:03.488077 137890040932160 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 20:41:03.504627 137890040932160 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 20:41:03.504708 137890040932160 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 20:41:03.521067 137890040932160 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 20:41:03.521127 137890040932160 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 20:41:03.545994 137890040932160 attentions.py:1198] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0424 20:41:03.546066 137890040932160 attentions.py:1198] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 20:41:03.567178 137890040932160 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0424 20:41:03.567245 137890040932160 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor'). I0424 20:41:03.775948 137890040932160 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0424 20:41:03.776056 137890040932160 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 8 tensor: 4 I0424 20:41:05.197016 137890040932160 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: ('tensor',) I0424 20:41:05.197141 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, 'tensor') I0424 20:41:05.197194 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, 'tensor') I0424 20:41:05.197259 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: ('tensor', None, 'fsdp') I0424 20:41:05.197312 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: ('tensor', None) I0424 20:41:05.197350 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: ('tensor', None) I0424 20:41:05.197401 137890040932160 maxtext_utils.py:1668] 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, 'tensor', None) I0424 20:41:05.197451 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: P('heads', 'layers', 'kv', 'embed') Physical: ('tensor', None, None, 'fsdp') I0424 20:41:05.197490 137890040932160 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: P('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, 'tensor', None) I0424 20:41:05.197528 137890040932160 maxtext_utils.py:1668] 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, 'tensor', None) I0424 20:41:05.197577 137890040932160 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', 'tensor') I0424 20:41:05.197623 137890040932160 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: ('tensor', 'fsdp') I0424 20:41:05.686857 137890040932160 train.py:155] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0424 20:41:05.686951 137890040932160 train.py:155] train/xent Physical: float32[8,2048]............................................. ('fsdp', None). I0424 20:41:05.702433 137890040932160 train.py:162] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0424 20:41:05.702492 137890040932160 train.py:162] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None). I0424 20:41:20.760597 137890040932160 max_utils.py:791] Total memory size: 0.9 GB, Output size: 0.4 GB, Temp size: 0.5 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0424 20:41:20.761436 137890040932160 metric_logger.py:301] number parameters: 1.104 billion I0424 20:41:22.354377 137890040932160 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0424 20:41:36.966952 137890040932160 checkpointing.py:776] Waited 14.612546682357788 seconds for step 0 to finish before starting checkpointing. I0424 20:41:36.969470 137890040932160 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 20:41:36.971445 137890040932160 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 0 I0424 20:41:36.972932 137890040932160 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints/0. I0424 20:41:37.268690 137890040932160 signaling_client.py:364] Using JaxDistributedSignalingClient I0424 20:41:37.269698 137890040932160 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 20:41:37.269754 137890040932160 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 20:41:37.533592 137890040932160 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.264984s I0424 20:41:37.533779 137890040932160 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.700 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2706329822540283 s) (per-host) I0424 20:41:37.533832 137890040932160 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.270697s (batch_requests_ready=0.002354s, total_serialization_initiated=0.268270s, others=0.000072s) I0424 20:41:37.533928 137890040932160 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.274871s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.274852) I0424 20:41:37.534754 137890040932160 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.56 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints/0. I0424 20:41:37.535074 137762506864384 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 21:01:37.535041 I0424 20:41:37.547836 137890040932160 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0424 20:41:37.548140 137762409801472 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0424 20:41:37.548296 137890040932160 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777063296.9694505, 'wait_for_prev_duration_secs': 6.127357482910156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777063296.9714825, 'checkpointer_blocking_duration_secs': 0.563753604888916, 'get_old_steps_start_time': 1777063297.5352633, 'get_old_steps_duration_secs': 3.337860107421875e-05, 'checkpoint_manager_blocking_start_time': 1777063296.9674873, 'checkpoint_manager_blocking_duration_secs': 0.5807669162750244} I0424 20:41:37.548400 137890040932160 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0424 20:41:37.548449 137890040932160 max_utils.py:750] Memstats: After params initialized: I0424 20:41:37.548500 137890040932160 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0424 20:41:37.548533 137890040932160 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0424 20:41:37.548560 137890040932160 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0424 20:41:37.548588 137890040932160 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0424 20:41:37.862323 137890040932160 metric_logger.py:196] completed step: 0, seconds: 1.593, TFLOP/s/device: 2.133, Tokens/s/device: 321.443, total_weights: 16384, loss: 10.862, lm_loss: 10.862, perplexity: 52153.723 I0424 20:41:37.953025 137890040932160 metric_logger.py:196] completed step: 1, seconds: 15.506, TFLOP/s/device: 0.219, Tokens/s/device: 33.019, total_weights: 16384, loss: 10.862, lm_loss: 10.862, perplexity: 52153.723 I0424 20:41:38.371324 137890040932160 metric_logger.py:196] completed step: 2, seconds: 0.032, TFLOP/s/device: 104.719, Tokens/s/device: 15784.444, total_weights: 16384, loss: 9.763, lm_loss: 9.763, perplexity: 17373.129 I0424 20:41:38.429667 137890040932160 metric_logger.py:196] completed step: 3, seconds: 0.419, TFLOP/s/device: 8.113, Tokens/s/device: 1222.917, total_weights: 16384, loss: 8.823, lm_loss: 8.823, perplexity: 6791.817 I0424 20:41:38.546938 137890040932160 metric_logger.py:196] completed step: 4, seconds: 0.064, TFLOP/s/device: 53.360, Tokens/s/device: 8042.980, total_weights: 16384, loss: 7.947, lm_loss: 7.947, perplexity: 2828.417 I0424 20:41:38.552568 137890040932160 metric_logger.py:196] completed step: 5, seconds: 0.058, TFLOP/s/device: 58.334, Tokens/s/device: 8792.719, total_weights: 16384, loss: 7.212, lm_loss: 7.212, perplexity: 1355.339 I0424 20:41:40.798433 2578 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0424 20:41:43.363335 137762418194176 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_5 I0424 20:41:52.555682 137890040932160 metric_logger.py:196] completed step: 6, seconds: 0.118, TFLOP/s/device: 28.826, Tokens/s/device: 4344.911, total_weights: 16384, loss: 6.652, lm_loss: 6.652, perplexity: 774.659 I0424 20:41:52.614473 137890040932160 metric_logger.py:196] completed step: 7, seconds: 13.945, TFLOP/s/device: 0.244, Tokens/s/device: 36.715, total_weights: 16384, loss: 6.269, lm_loss: 6.269, perplexity: 527.696 I0424 20:41:52.672831 137890040932160 metric_logger.py:196] completed step: 8, seconds: 0.063, TFLOP/s/device: 53.734, Tokens/s/device: 8099.344, total_weights: 16384, loss: 6.031, lm_loss: 6.031, perplexity: 416.165 I0424 20:41:52.730459 137890040932160 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0424 20:41:52.731023 137890040932160 checkpointing.py:776] Waited 0.0005793571472167969 seconds for step 9 to finish before starting checkpointing. I0424 20:41:52.734296 137890040932160 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 20:42:13.579210 137762506864384 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 43.496 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.31602120399475 s) (per-host) I0424 20:42:13.579335 137762506864384 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.044141s. I0424 20:42:22.029600 137762506864384 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 44.494389s. I0424 20:42:22.029898 137762409801472 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 20:42:22.030017 137762409801472 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0424 20:42:22.030067 137762409801472 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0424 20:42:22.032432 137762409801472 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0424 20:42:22.032602 137890040932160 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 20:42:22.032747 137890040932160 checkpoint_manager.py:1452] Waiting for previous save to complete took 29.298457 seconds. If this number is high, consider checkpointing less frequently. I0424 20:42:22.034637 137890040932160 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 9 I0424 20:42:22.036605 137890040932160 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints/9. I0424 20:42:22.352884 137890040932160 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0424 20:42:22.352975 137890040932160 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 20:42:22.379266 137890040932160 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027485s I0424 20:42:22.379445 137890040932160 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 49.565 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03112196922302246 s) (per-host) I0424 20:42:22.379498 137890040932160 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031186s (batch_requests_ready=0.001959s, total_serialization_initiated=0.029155s, others=0.000072s) I0424 20:42:22.379597 137890040932160 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035333s (all_items=0.000015s, per_item={'items': '0.00001454'}, temp_paths=0.035318) I0424 20:42:22.380319 137890040932160 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.35 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints/9. I0424 20:42:22.380633 137762409801472 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 21:02:22.380597 I0424 20:42:22.383497 137890040932160 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0424 20:42:22.383781 137761428801280 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0424 20:42:22.383945 137890040932160 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777063312.7342594, 'wait_for_prev_duration_secs': 29.29845690727234, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777063342.0346901, 'checkpointer_blocking_duration_secs': 0.3460962772369385, 'get_old_steps_start_time': 1777063342.3808105, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1777063312.7312446, 'checkpoint_manager_blocking_duration_secs': 29.65266513824463} I0424 20:42:22.384063 137890040932160 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0424 20:42:22.384107 137890040932160 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 20:42:27.729034 137755573090048 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_5 I0424 20:43:03.834054 137762409801472 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 38.076 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.48571252822876 s) (per-host) I0424 20:43:03.834162 137762409801472 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.453406s. I0424 20:43:11.651332 137762409801472 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.270559s. I0424 20:43:11.651616 137761428801280 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 20:43:11.651757 137761428801280 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0424 20:43:11.651811 137761428801280 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0424 20:43:11.653561 137761428801280 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0424 20:43:11.653737 137890040932160 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 20:43:11.653882 137890040932160 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 20:43:11.654835 137890040932160 metric_logger.py:196] completed step: 9, seconds: 0.059, TFLOP/s/device: 57.936, Tokens/s/device: 8732.731, total_weights: 16384, loss: 5.892, lm_loss: 5.892, perplexity: 361.955 Per train step: Total TFLOPs: 3.40 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Fri Apr 24 20:43:20 UTC 2026 EXIT_CODE=0