XPK Start: Tue Apr 21 15:22:29 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. 2026-04-21 15:22:54.443109: 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) I0421 15:22:54.652785 139473395197760 max_utils.py:273] Attempting to initialize the jax distributed system... I0421 15:23:03.694012 139473395197760 distributed.py:149] Starting JAX distributed service on [::]:8482 I0421 15:23:03.696383 139473395197760 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-w0ydx-slice-job-0-0.mt-11-optimizer-offload-w0ydx:8482 I0421 15:23:04.709656 139473395197760 max_utils.py:284] Jax distributed system initialized! I0421 15:23:10.897595 139473395197760 max_utils.py:800] System Information: Jax Version: 0.9.2 I0421 15:23:10.897703 139473395197760 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0421 15:23:10.897761 139473395197760 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 I0421 15:23:10.897801 139473395197760 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0421 15:23:11.613231 139473395197760 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0421 15:23:11.613514 139473395197760 checkpointing.py:677] Setting up checkpoint logger... I0421 15:23:11.613570 139473395197760 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0421 15:23:11.613613 139473395197760 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0421 15:23:11.613965 139473395197760 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 0x7ed90b3144d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0421 15:23:14.788511 139473395197760 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0421 15:23:14.788764 139473395197760 checkpoint_manager.py:708] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec5040d50a0>}, handler_registry=None I0421 15:23:14.789055 139473395197760 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec5040d50a0>` 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`. I0421 15:23:14.789114 139473395197760 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ec5040d4c80>` 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`. I0421 15:23:14.789157 139473395197760 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 0x7ec5040d50a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec5040d50a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ec5040d4c80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ec5040d4c80>}). I0421 15:23:14.789478 139473395197760 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0421 15:23:14.789553 139473395197760 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ec4ec6491c0> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0421 15:23:15.500415 139473395197760 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints I0421 15:23:15.929672 139473395197760 checkpoint_manager.py:929] [process=7][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_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ec5040d5760> I0421 15:23:15.929875 139473395197760 checkpointing.py:301] Checkpoint manager created! I0421 15:23:16.865507 139473395197760 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0421 15:23:16.865626 139473395197760 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:23:17.245837 139473395197760 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 15:23:17.245933 139473395197760 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:23:17.262440 139473395197760 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 15:23:17.262502 139473395197760 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:23:17.286576 139473395197760 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0421 15:23:17.286648 139473395197760 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:23:17.303289 139473395197760 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0421 15:23:17.303350 139473395197760 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:23:17.319891 139473395197760 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0421 15:23:17.319952 139473395197760 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:23:17.344772 139473395197760 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0421 15:23:17.344847 139473395197760 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:23:17.365968 139473395197760 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0421 15:23:17.366032 139473395197760 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0421 15:23:17.578527 139473395197760 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0421 15:23:17.578647 139473395197760 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0421 15:23:19.188079 139473395197760 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0421 15:23:19.188209 139473395197760 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, None) I0421 15:23:19.188264 139473395197760 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, None) I0421 15:23:19.188322 139473395197760 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0421 15:23:19.188373 139473395197760 maxtext_utils.py:1668] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0421 15:23:19.188412 139473395197760 maxtext_utils.py:1668] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0421 15:23:19.188466 139473395197760 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, None, None) I0421 15:23:19.188517 139473395197760 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: (None, None, None, 'fsdp') I0421 15:23:19.188557 139473395197760 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, None, None) I0421 15:23:19.188593 139473395197760 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, None, None) I0421 15:23:19.188639 139473395197760 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0421 15:23:19.188684 139473395197760 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0421 15:23:19.688417 139473395197760 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 15:23:19.688514 139473395197760 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0421 15:23:19.704244 139473395197760 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 15:23:19.704305 139473395197760 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0421 15:23:30.811846 139473395197760 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB. I0421 15:23:30.812670 139473395197760 metric_logger.py:301] number parameters: 1.104 billion I0421 15:23:42.065550 139473395197760 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0421 15:23:42.233788 139473395197760 checkpointing.py:776] Waited 0.16819405555725098 seconds for step 0 to finish before starting checkpointing. I0421 15:23:42.236330 139473395197760 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:23:42.238024 139473395197760 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 0 I0421 15:23:42.239547 139473395197760 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints/0. I0421 15:23:42.788862 139473395197760 signaling_client.py:364] Using JaxDistributedSignalingClient I0421 15:23:42.789943 139473395197760 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:23:42.790000 139473395197760 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 I0421 15:23:43.068060 139473395197760 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279268s I0421 15:23:43.068234 139473395197760 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.405 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2853989601135254 s) (per-host) I0421 15:23:43.068289 139473395197760 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.285464s (batch_requests_ready=0.002202s, total_serialization_initiated=0.283190s, others=0.000072s) I0421 15:23:43.068384 139473395197760 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.289659s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.289642) I0421 15:23:43.069203 139473395197760 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.83 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints/0. I0421 15:23:43.069486 139339035109120 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:43:43.069461 I0421 15:23:43.071483 139473395197760 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:23:43.071712 139331113981696 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:23:43.071835 139473395197760 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776785022.2363129, 'wait_for_prev_duration_secs': 6.0558319091796875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776785022.238062, 'checkpointer_blocking_duration_secs': 0.8315386772155762, 'get_old_steps_start_time': 1776785023.0696256, 'get_old_steps_duration_secs': 3.2901763916015625e-05, 'checkpoint_manager_blocking_start_time': 1776785022.234333, 'checkpoint_manager_blocking_duration_secs': 0.8374636173248291} I0421 15:23:43.071942 139473395197760 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0421 15:23:43.071990 139473395197760 max_utils.py:750] Memstats: After params initialized: I0421 15:23:43.072040 139473395197760 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_26(process=7,(2,6,0,0)) I0421 15:23:43.072071 139473395197760 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_27(process=7,(3,6,0,0)) I0421 15:23:43.072099 139473395197760 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_30(process=7,(2,7,0,0)) I0421 15:23:43.072126 139473395197760 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_31(process=7,(3,7,0,0)) I0421 15:23:43.387588 139473395197760 metric_logger.py:196] completed step: 0, seconds: 11.253, TFLOP/s/device: 1.207, Tokens/s/device: 182.000, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 15:23:43.574477 139473395197760 metric_logger.py:196] completed step: 1, seconds: 1.320, TFLOP/s/device: 10.289, Tokens/s/device: 1550.929, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 15:23:43.983147 139473395197760 metric_logger.py:196] completed step: 2, seconds: 0.024, TFLOP/s/device: 560.363, Tokens/s/device: 84464.057, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0421 15:23:44.146064 139473395197760 metric_logger.py:196] completed step: 3, seconds: 0.409, TFLOP/s/device: 33.236, Tokens/s/device: 5009.748, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.047 I0421 15:23:44.471656 139473395197760 metric_logger.py:196] completed step: 4, seconds: 0.168, TFLOP/s/device: 80.722, Tokens/s/device: 12167.228, total_weights: 65536, loss: 9.278, lm_loss: 9.278, perplexity: 10694.968 I0421 15:23:44.477324 139473395197760 metric_logger.py:196] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.354, Tokens/s/device: 12564.032, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.520 I0421 15:23:46.286770 2807 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0421 15:23:48.739454 139339018323712 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_7 I0421 15:24:06.588662 139473395197760 metric_logger.py:196] completed step: 6, seconds: 0.326, TFLOP/s/device: 41.695, Tokens/s/device: 6284.715, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.688 I0421 15:24:06.751859 139473395197760 metric_logger.py:196] completed step: 7, seconds: 21.949, TFLOP/s/device: 0.619, Tokens/s/device: 93.307, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4377.010 I0421 15:24:06.914580 139473395197760 metric_logger.py:196] completed step: 8, seconds: 0.168, TFLOP/s/device: 80.986, Tokens/s/device: 12207.043, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3845.943 I0421 15:24:07.077103 139473395197760 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0421 15:24:07.077690 139473395197760 checkpointing.py:776] Waited 0.0006058216094970703 seconds for step 9 to finish before starting checkpointing. I0421 15:24:07.079727 139473395197760 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 15:24:20.451666 139339035109120 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 41.932 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.6687912940979 s) (per-host) I0421 15:24:20.451803 139339035109120 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.382236s. I0421 15:24:34.968349 139339035109120 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 51.898764s. I0421 15:24:34.968562 139331113981696 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:24:34.968630 139331113981696 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:24:34.968683 139331113981696 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:24:34.970501 139331113981696 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0421 15:24:34.970690 139473395197760 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0421 15:24:34.970870 139473395197760 checkpoint_manager.py:1452] Waiting for previous save to complete took 27.891141 seconds. If this number is high, consider checkpointing less frequently. I0421 15:24:34.972747 139473395197760 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 9 I0421 15:24:34.974843 139473395197760 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints/9. I0421 15:24:35.294716 139473395197760 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:24:35.294827 139473395197760 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 I0421 15:24:35.337936 139473395197760 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044213s I0421 15:24:35.338083 139473395197760 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 32.471 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04750394821166992 s) (per-host) I0421 15:24:35.338132 139473395197760 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.047562s (batch_requests_ready=0.001791s, total_serialization_initiated=0.045706s, others=0.000064s) I0421 15:24:35.338207 139473395197760 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.051630s (all_items=0.000014s, per_item={'items': '0.00001431'}, temp_paths=0.051615) I0421 15:24:35.338877 139473395197760 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.37 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints/9. I0421 15:24:35.339237 139331113981696 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:44:35.339200 I0421 15:24:35.341189 139473395197760 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:24:35.341464 139319427708672 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:24:35.341586 139473395197760 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776785047.0796974, 'wait_for_prev_duration_secs': 27.891140699386597, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776785074.972791, 'checkpointer_blocking_duration_secs': 0.36658716201782227, 'get_old_steps_start_time': 1776785075.3394036, 'get_old_steps_duration_secs': 2.86102294921875e-05, 'checkpoint_manager_blocking_start_time': 1776785047.0779715, 'checkpoint_manager_blocking_duration_secs': 28.26357913017273} I0421 15:24:35.341691 139473395197760 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0421 15:24:35.341755 139473395197760 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 15:24:40.425596 139339018323712 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_7 I0421 15:25:16.932274 139331113981696 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 37.932 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.64164972305298 s) (per-host) I0421 15:25:16.932393 139331113981696 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.593044s. I0421 15:25:25.487476 139331113981696 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 50.148112s. I0421 15:25:25.487824 139319427708672 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:25:25.487961 139319427708672 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:25:25.488009 139319427708672 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:25:25.489431 139319427708672 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0421 15:25:25.489634 139473395197760 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0421 15:25:25.489825 139473395197760 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:25:25.490719 139473395197760 metric_logger.py:196] completed step: 9, seconds: 0.163, TFLOP/s/device: 83.375, Tokens/s/device: 12567.193, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.739 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Tue Apr 21 15:25:34 UTC 2026 EXIT_CODE=0