XPK Start: Tue Apr 21 15:00:20 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 15:00:45.461803: 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 15:00:45.672158 138990160594752 max_utils.py:273] Attempting to initialize the jax distributed system... I0421 15:00:54.714177 138990160594752 distributed.py:149] Starting JAX distributed service on [::]:8482 I0421 15:00:54.716423 138990160594752 distributed.py:172] Connecting to JAX distributed service on mt-07-eval-odhg7-slice-job-0-0.mt-07-eval-odhg7:8482 I0421 15:00:55.825120 138990160594752 max_utils.py:284] Jax distributed system initialized! I0421 15:01:02.061098 138990160594752 max_utils.py:800] System Information: Jax Version: 0.9.2 I0421 15:01:02.061207 138990160594752 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0421 15:01:02.061247 138990160594752 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 15:01:02.061283 138990160594752 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0421 15:01:02.753000 138990160594752 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0421 15:01:02.753282 138990160594752 checkpointing.py:677] Setting up checkpoint logger... I0421 15:01:02.753335 138990160594752 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0421 15:01:02.753378 138990160594752 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0421 15:01:02.753708 138990160594752 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 0x7e6873744530>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0421 15:01:05.729366 138990160594752 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0421 15:01:05.729605 138990160594752 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 0x7e5478621760>}, handler_registry=None I0421 15:01:05.729863 138990160594752 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e5478621760>` 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 15:01:05.729912 138990160594752 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e5478622240>` 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 15:01:05.729949 138990160594752 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 0x7e5478621760>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e5478621760>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e5478622240>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e5478622240>}). I0421 15:01:05.730277 138990160594752 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0421 15:01:05.730346 138990160594752 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e54603dd620> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0421 15:01:06.821123 138990160594752 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints I0421 15:01:07.267343 138990160594752 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_07_eval/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e54786205f0> I0421 15:01:07.267516 138990160594752 checkpointing.py:301] Checkpoint manager created! I0421 15:01:08.194894 138990160594752 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0421 15:01:08.195020 138990160594752 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:01:08.574128 138990160594752 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 15:01:08.574223 138990160594752 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:01:08.590828 138990160594752 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 15:01:08.590887 138990160594752 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 15:01:08.614853 138990160594752 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 15:01:08.614919 138990160594752 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:01:08.631463 138990160594752 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 15:01:08.631524 138990160594752 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:01:08.648041 138990160594752 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 15:01:08.648101 138990160594752 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:01:08.672877 138990160594752 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0421 15:01:08.672951 138990160594752 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 15:01:08.694035 138990160594752 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0421 15:01:08.694096 138990160594752 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0421 15:01:08.902563 138990160594752 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0421 15:01:08.902675 138990160594752 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0421 15:01:10.326942 138990160594752 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0421 15:01:10.327077 138990160594752 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 15:01:10.327130 138990160594752 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 15:01:10.327189 138990160594752 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 15:01:10.327242 138990160594752 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 15:01:10.327279 138990160594752 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 15:01:10.327329 138990160594752 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 15:01:10.327379 138990160594752 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 15:01:10.327417 138990160594752 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 15:01:10.327451 138990160594752 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 15:01:10.327496 138990160594752 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0421 15:01:10.327541 138990160594752 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0421 15:01:10.822089 138990160594752 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 15:01:10.822185 138990160594752 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0421 15:01:10.837729 138990160594752 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 15:01:10.837808 138990160594752 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0421 15:01:21.842700 138990160594752 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. I0421 15:01:21.843559 138990160594752 metric_logger.py:301] number parameters: 1.104 billion I0421 15:01:33.123090 138990160594752 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0421 15:01:33.303894 138990160594752 checkpointing.py:776] Waited 0.18079304695129395 seconds for step 0 to finish before starting checkpointing. I0421 15:01:33.306252 138990160594752 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:01:33.308096 138990160594752 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 0 I0421 15:01:33.309661 138990160594752 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints/0. I0421 15:01:33.651312 138990160594752 signaling_client.py:364] Using JaxDistributedSignalingClient I0421 15:01:33.652213 138990160594752 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:01:33.652267 138990160594752 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 15:01:33.922652 138990160594752 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.271414s I0421 15:01:33.922864 138990160594752 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.568 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2770669460296631 s) (per-host) I0421 15:01:33.922922 138990160594752 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.277136s (batch_requests_ready=0.002456s, total_serialization_initiated=0.274603s, others=0.000077s) I0421 15:01:33.923018 138990160594752 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.281618s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.281600) I0421 15:01:33.923879 138990160594752 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.62 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints/0. I0421 15:01:33.924155 138862443022080 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:21:33.924123 I0421 15:01:33.967133 138990160594752 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:01:33.967435 138859245467392 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:01:33.967577 138990160594752 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776783693.3062344, 'wait_for_prev_duration_secs': 6.175041198730469e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776783693.3081348, 'checkpointer_blocking_duration_secs': 0.6161749362945557, 'get_old_steps_start_time': 1776783693.924334, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1776783693.3042955, 'checkpoint_manager_blocking_duration_secs': 0.6632411479949951} I0421 15:01:33.967684 138990160594752 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0421 15:01:33.967753 138990160594752 max_utils.py:750] Memstats: After params initialized: I0421 15:01:33.967806 138990160594752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0)) I0421 15:01:33.967836 138990160594752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0)) I0421 15:01:33.967863 138990160594752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0)) I0421 15:01:33.967888 138990160594752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0)) I0421 15:01:34.277886 138990160594752 metric_logger.py:196] completed step: 0, seconds: 11.279, TFLOP/s/device: 1.205, Tokens/s/device: 181.570, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 15:01:34.460865 138990160594752 metric_logger.py:196] completed step: 1, seconds: 1.153, TFLOP/s/device: 11.782, Tokens/s/device: 1775.866, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 15:01:34.885748 138990160594752 metric_logger.py:196] completed step: 2, seconds: 0.036, TFLOP/s/device: 379.476, Tokens/s/device: 57198.715, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129 I0421 15:01:35.033076 138990160594752 metric_logger.py:196] completed step: 3, seconds: 0.425, TFLOP/s/device: 31.960, Tokens/s/device: 4817.384, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203 I0421 15:01:35.327842 138990160594752 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.941, Tokens/s/device: 13406.212, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614 I0421 15:01:35.333507 138990160594752 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.145, Tokens/s/device: 13889.077, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717 I0421 15:01:37.396724 2840 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0421 15:01:39.602346 138859725473536 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_07_eval/checkpoints/0/items/array_metadatas/process_7 I0421 15:01:57.106157 138990160594752 metric_logger.py:196] completed step: 6, seconds: 0.295, TFLOP/s/device: 46.054, Tokens/s/device: 6941.832, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759 I0421 15:01:57.253682 138990160594752 metric_logger.py:196] completed step: 7, seconds: 21.626, TFLOP/s/device: 0.628, Tokens/s/device: 94.700, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905 I0421 15:01:57.400928 138990160594752 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 89.092, Tokens/s/device: 13428.980, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406 I0421 15:01:57.547651 138990160594752 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0421 15:01:57.548348 138990160594752 checkpointing.py:776] Waited 0.0007169246673583984 seconds for step 9 to finish before starting checkpointing. I0421 15:01:57.550414 138990160594752 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 15:02:11.006345 138862443022080 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 42.280 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.360512018203735 s) (per-host) I0421 15:02:11.006455 138862443022080 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.082184s. I0421 15:02:18.724689 138862443022080 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 44.800402s. I0421 15:02:18.724989 138859245467392 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:02:18.725132 138859245467392 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:02:18.725183 138859245467392 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:02:18.726746 138859245467392 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0421 15:02:18.726908 138990160594752 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 15:02:18.727051 138990160594752 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.176636 seconds. If this number is high, consider checkpointing less frequently. I0421 15:02:18.728779 138990160594752 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 9 I0421 15:02:18.730843 138990160594752 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints/9. I0421 15:02:19.319672 138990160594752 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 15:02:19.319781 138990160594752 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 15:02:19.352367 138990160594752 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.033540s I0421 15:02:19.352526 138990160594752 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 41.670 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.037018537521362305 s) (per-host) I0421 15:02:19.352575 138990160594752 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.037076s (batch_requests_ready=0.001784s, total_serialization_initiated=0.035228s, others=0.000064s) I0421 15:02:19.352660 138990160594752 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.041233s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.041217) I0421 15:02:19.353419 138990160594752 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.62 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints/9. I0421 15:02:19.353783 138859245467392 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:22:19.353745 I0421 15:02:19.357552 138990160594752 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0421 15:02:19.357875 138839976830720 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 15:02:19.358040 138990160594752 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776783717.5503852, 'wait_for_prev_duration_secs': 21.1766357421875, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776783738.7288163, 'checkpointer_blocking_duration_secs': 0.6251187324523926, 'get_old_steps_start_time': 1776783739.3539581, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776783717.5486302, 'checkpoint_manager_blocking_duration_secs': 21.80937361717224} I0421 15:02:19.358155 138990160594752 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0421 15:02:19.358198 138990160594752 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 15:02:24.551011 138859725473536 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_07_eval/checkpoints/9/items/array_metadatas/process_7 I0421 15:03:02.062855 138859245467392 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 36.952 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.74730849266052 s) (per-host) I0421 15:03:02.062985 138859245467392 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.709080s. I0421 15:03:08.061604 138859245467392 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 48.707683s. I0421 15:03:08.061940 138839976830720 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 15:03:08.062069 138839976830720 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 15:03:08.062117 138839976830720 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0421 15:03:08.063498 138839976830720 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0421 15:03:08.063686 138990160594752 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 15:03:08.063862 138990160594752 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 15:03:08.064764 138990160594752 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.326, Tokens/s/device: 13916.447, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Tue Apr 21 15:03:20 UTC 2026 EXIT_CODE=0