XPK Start: Tue Apr 21 07:32:36 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 07:33:01.247571: 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 07:33:01.459232 137014495983424 max_utils.py:273] Attempting to initialize the jax distributed system... I0421 07:33:10.498498 137014495983424 distributed.py:149] Starting JAX distributed service on [::]:8482 I0421 07:33:10.500809 137014495983424 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-true-sv-eiki1-slice-job-0-0.mt-16-zarr3-true-sv-eiki1:8482 I0421 07:33:11.525515 137014495983424 max_utils.py:284] Jax distributed system initialized! I0421 07:33:17.591296 137014495983424 max_utils.py:800] System Information: Jax Version: 0.9.2 I0421 07:33:17.591402 137014495983424 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0421 07:33:17.591444 137014495983424 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 07:33:17.591480 137014495983424 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0421 07:33:18.291302 137014495983424 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0421 07:33:18.291585 137014495983424 checkpointing.py:677] Setting up checkpoint logger... I0421 07:33:18.291643 137014495983424 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0421 07:33:18.291685 137014495983424 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0421 07:33:18.292042 137014495983424 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 0x7c9c74950a70>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0421 07:33:21.203626 137014495983424 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0421 07:33:21.203884 137014495983424 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 0x7c884c325130>}, handler_registry=None I0421 07:33:21.204126 137014495983424 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c884c325130>` 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 07:33:21.204175 137014495983424 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c884c1c5340>` 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 07:33:21.204212 137014495983424 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 0x7c884c325130>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c884c325130>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c884c1c5340>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c884c1c5340>}). I0421 07:33:21.204523 137014495983424 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0421 07:33:21.204600 137014495983424 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c87646f1580> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0421 07:33:22.263036 137014495983424 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints I0421 07:33:22.484802 137014495983424 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=5), 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_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c884c1c1730> I0421 07:33:22.484979 137014495983424 checkpointing.py:301] Checkpoint manager created! I0421 07:33:23.404316 137014495983424 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0421 07:33:23.404434 137014495983424 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 07:33:23.787494 137014495983424 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 07:33:23.787587 137014495983424 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 07:33:23.804034 137014495983424 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0421 07:33:23.804095 137014495983424 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0421 07:33:23.827956 137014495983424 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 07:33:23.828032 137014495983424 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 07:33:23.844673 137014495983424 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 07:33:23.844748 137014495983424 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 07:33:23.861159 137014495983424 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 07:33:23.861219 137014495983424 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 07:33:23.885931 137014495983424 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0421 07:33:23.886004 137014495983424 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0421 07:33:23.907111 137014495983424 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0421 07:33:23.907181 137014495983424 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0421 07:33:24.115807 137014495983424 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0421 07:33:24.115920 137014495983424 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0421 07:33:25.563974 137014495983424 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0421 07:33:25.564153 137014495983424 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 07:33:25.564233 137014495983424 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 07:33:25.564319 137014495983424 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 07:33:25.564405 137014495983424 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 07:33:25.564471 137014495983424 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 07:33:25.564554 137014495983424 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 07:33:25.564649 137014495983424 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 07:33:25.564715 137014495983424 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 07:33:25.564797 137014495983424 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 07:33:25.564869 137014495983424 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0421 07:33:25.564941 137014495983424 maxtext_utils.py:1668] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0421 07:33:26.055221 137014495983424 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 07:33:26.055310 137014495983424 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0421 07:33:26.070774 137014495983424 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0421 07:33:26.070831 137014495983424 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0421 07:33:37.011988 137014495983424 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 07:33:37.012791 137014495983424 metric_logger.py:301] number parameters: 1.104 billion I0421 07:33:48.234817 137014495983424 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0421 07:33:48.549196 137014495983424 checkpointing.py:776] Waited 0.3143596649169922 seconds for step 0 to finish before starting checkpointing. I0421 07:33:48.551928 137014495983424 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 07:33:48.553538 137014495983424 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 0 I0421 07:33:48.554939 137014495983424 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/0. I0421 07:33:49.261194 137014495983424 signaling_client.py:364] Using JaxDistributedSignalingClient I0421 07:33:49.262149 137014495983424 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 07:33:49.262207 137014495983424 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 07:33:49.544945 137014495983424 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.283824s I0421 07:33:49.545134 137014495983424 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.329 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2894749641418457 s) (per-host) I0421 07:33:49.545197 137014495983424 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289545s (batch_requests_ready=0.002503s, total_serialization_initiated=0.286963s, others=0.000080s) I0421 07:33:49.545314 137014495983424 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293811s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.293793) I0421 07:33:49.546097 137014495983424 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.99 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/0. I0421 07:33:49.546441 136886291359488 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:53:49.546402 I0421 07:33:49.548456 137014495983424 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0421 07:33:49.548811 136885238232832 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 07:33:49.548985 137014495983424 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776756828.5519097, 'wait_for_prev_duration_secs': 6.0558319091796875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776756828.5535767, 'checkpointer_blocking_duration_secs': 0.9930181503295898, 'get_old_steps_start_time': 1776756829.546619, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776756828.5497808, 'checkpoint_manager_blocking_duration_secs': 0.9991559982299805} I0421 07:33:49.549120 137014495983424 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0421 07:33:49.549180 137014495983424 max_utils.py:750] Memstats: After params initialized: I0421 07:33:49.549239 137014495983424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0)) I0421 07:33:49.549277 137014495983424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0)) I0421 07:33:49.549308 137014495983424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0)) I0421 07:33:49.549340 137014495983424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0)) I0421 07:33:49.864785 137014495983424 metric_logger.py:196] completed step: 0, seconds: 11.222, TFLOP/s/device: 1.211, Tokens/s/device: 182.500, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 07:33:50.036625 137014495983424 metric_logger.py:196] completed step: 1, seconds: 1.628, TFLOP/s/device: 8.343, Tokens/s/device: 1257.616, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875 I0421 07:33:50.446652 137014495983424 metric_logger.py:196] completed step: 2, seconds: 0.025, TFLOP/s/device: 551.068, Tokens/s/device: 83062.946, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719 I0421 07:33:50.593947 137014495983424 metric_logger.py:196] completed step: 3, seconds: 0.410, TFLOP/s/device: 33.146, Tokens/s/device: 4996.194, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506 I0421 07:33:50.598502 137014495983424 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0421 07:33:50.888104 137014495983424 checkpointing.py:776] Waited 0.2895667552947998 seconds for step 5 to finish before starting checkpointing. I0421 07:33:50.891058 137014495983424 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 07:33:53.028256 2807 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0421 07:33:55.042397 136890017502976 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_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/0/items/array_metadatas/process_7 I0421 07:34:27.207986 136886291359488 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 41.621 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.95228028297424 s) (per-host) I0421 07:34:27.208107 136886291359488 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.661551s. I0421 07:34:35.739022 136886291359488 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 46.192449s. I0421 07:34:35.739343 136885238232832 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 07:34:35.739463 136885238232832 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 07:34:35.739511 136885238232832 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0421 07:34:35.741026 136885238232832 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0421 07:34:35.741232 137014495983424 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 07:34:35.741373 137014495983424 checkpoint_manager.py:1452] Waiting for previous save to complete took 44.850312 seconds. If this number is high, consider checkpointing less frequently. I0421 07:34:35.743621 137014495983424 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 5 I0421 07:34:35.746227 137014495983424 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/5. I0421 07:34:36.467681 137014495983424 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 07:34:36.467786 137014495983424 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 07:34:36.501794 137014495983424 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035324s I0421 07:34:36.501968 137014495983424 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 39.848 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0387115478515625 s) (per-host) I0421 07:34:36.502030 137014495983424 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038779s (batch_requests_ready=0.001767s, total_serialization_initiated=0.036937s, others=0.000074s) I0421 07:34:36.502124 137014495983424 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043957s (all_items=0.000015s, per_item={'items': '0.00001454'}, temp_paths=0.043943) I0421 07:34:36.502840 137014495983424 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/5. I0421 07:34:36.503181 136885238232832 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:54:36.503143 I0421 07:34:36.510962 137014495983424 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0421 07:34:36.511211 136881963198208 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 07:34:36.511341 137014495983424 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776756830.8910294, 'wait_for_prev_duration_secs': 44.85031175613403, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776756875.7436676, 'checkpointer_blocking_duration_secs': 0.7596631050109863, 'get_old_steps_start_time': 1776756876.5033526, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776756830.8886974, 'checkpoint_manager_blocking_duration_secs': 45.62261176109314} I0421 07:34:36.511454 137014495983424 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0421 07:34:36.512275 137014495983424 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.737, Tokens/s/device: 13375.480, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355 I0421 07:34:36.518098 137014495983424 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.164, Tokens/s/device: 13891.904, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467 I0421 07:34:41.997248 136890017502976 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_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/5/items/array_metadatas/process_7 I0421 07:34:58.876376 137014495983424 metric_logger.py:196] completed step: 6, seconds: 45.918, TFLOP/s/device: 0.296, Tokens/s/device: 44.601, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740 I0421 07:34:59.023709 137014495983424 metric_logger.py:196] completed step: 7, seconds: 22.212, TFLOP/s/device: 0.612, Tokens/s/device: 92.202, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151 I0421 07:34:59.170906 137014495983424 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.316, Tokens/s/device: 13462.613, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098 I0421 07:34:59.317547 137014495983424 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0421 07:34:59.318236 137014495983424 checkpointing.py:776] Waited 0.0007100105285644531 seconds for step 9 to finish before starting checkpointing. I0421 07:34:59.320316 137014495983424 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 07:35:18.525902 136885238232832 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 37.553 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.06260585784912 s) (per-host) I0421 07:35:18.526026 136885238232832 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.022730s. I0421 07:35:27.515807 136885238232832 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 51.012496s. I0421 07:35:27.516105 136881963198208 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 07:35:27.516230 136881963198208 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 07:35:27.516278 136881963198208 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0421 07:35:27.520265 136881963198208 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0421 07:35:27.520432 137014495983424 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0421 07:35:27.520565 137014495983424 checkpoint_manager.py:1452] Waiting for previous save to complete took 28.200253 seconds. If this number is high, consider checkpointing less frequently. I0421 07:35:27.522389 137014495983424 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 9 I0421 07:35:27.524572 137014495983424 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/9. I0421 07:35:27.835317 137014495983424 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0421 07:35:27.835411 137014495983424 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 07:35:27.867793 137014495983424 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.033351s I0421 07:35:27.867962 137014495983424 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 41.915 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03680253028869629 s) (per-host) I0421 07:35:27.868015 137014495983424 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.036865s (batch_requests_ready=0.001673s, total_serialization_initiated=0.035125s, others=0.000067s) I0421 07:35:27.868129 137014495983424 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.040997s (all_items=0.000010s, per_item={'items': '0.00001049'}, temp_paths=0.040987) I0421 07:35:27.868855 137014495983424 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.35 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/9. I0421 07:35:27.869130 136881963198208 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:55:27.869101 I0421 07:35:27.872961 137014495983424 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0421 07:35:27.873222 136884164491008 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0421 07:35:27.873379 137014495983424 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776756899.3202884, 'wait_for_prev_duration_secs': 28.200253009796143, 'time_between_consecutive_saves_sec': 23.579216241836548, 'checkpointer_blocking_start_time': 1776756927.5224304, 'checkpointer_blocking_duration_secs': 0.3468003273010254, 'get_old_steps_start_time': 1776756927.8692498, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776756899.318464, 'checkpoint_manager_blocking_duration_secs': 28.554882526397705} I0421 07:35:27.873496 137014495983424 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0421 07:35:27.873538 137014495983424 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0421 07:35:33.393674 136890017502976 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_061359/linen_xpk_main_20260421_061359_16_zarr3_true/checkpoints/9/items/array_metadatas/process_7 I0421 07:36:10.223214 136881963198208 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 37.262 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.392016887664795 s) (per-host) I0421 07:36:10.223333 136881963198208 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.354128s. I0421 07:36:17.521691 136881963198208 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 49.652469s. I0421 07:36:17.521999 136884164491008 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0421 07:36:17.522121 136884164491008 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0421 07:36:17.522163 136884164491008 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0421 07:36:17.523462 136884164491008 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0421 07:36:17.523677 137014495983424 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 07:36:17.523849 137014495983424 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0421 07:36:17.524806 137014495983424 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.176, Tokens/s/device: 13893.788, total_weights: 65536, loss: 7.181, lm_loss: 7.181, perplexity: 1314.495 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Tue Apr 21 07:36:26 UTC 2026 EXIT_CODE=0