XPK Start: Fri Apr 17 01:22:31 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-17 01:22:55.505051: 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) I0417 01:22:55.681729 132358857959232 max_utils.py:273] Attempting to initialize the jax distributed system... I0417 01:23:04.721145 132358857959232 distributed.py:149] Starting JAX distributed service on [::]:8482 I0417 01:23:04.723531 132358857959232 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-false-dh067-slice-job-0-0.mt-10-shardy-false-dh067:8482 I0417 01:23:05.721420 132358857959232 max_utils.py:284] Jax distributed system initialized! I0417 01:23:10.734651 132358857959232 max_utils.py:800] System Information: Jax Version: 0.9.2 I0417 01:23:10.734754 132358857959232 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0417 01:23:10.734793 132358857959232 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 I0417 01:23:10.734844 132358857959232 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0417 01:23:11.779367 132358857959232 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0417 01:23:11.892302 132358857959232 checkpointing.py:688] Setting up checkpoint logger... I0417 01:23:11.892418 132358857959232 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0417 01:23:11.892463 132358857959232 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0417 01:23:11.892662 132358857959232 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 0x786092314440>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0417 01:23:14.866514 132358857959232 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0417 01:23:14.866749 132358857959232 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x785f6ebdb380>}, handler_registry=None I0417 01:23:14.867007 132358857959232 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x785f6ebdb380>` 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`. I0417 01:23:14.867057 132358857959232 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x785f6eb933e0>` 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`. I0417 01:23:14.867093 132358857959232 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 0x785f6ebdb380>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x785f6ebdb380>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x785f6eb933e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x785f6eb933e0>}). I0417 01:23:14.867415 132358857959232 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0417 01:23:14.867486 132358857959232 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x784cbc6ee340> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0417 01:23:16.264146 132358857959232 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints I0417 01:23:16.730211 132358857959232 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x785f6ebe9fd0> I0417 01:23:16.730382 132358857959232 checkpointing.py:302] Checkpoint manager created! I0417 01:23:17.709929 132358857959232 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0417 01:23:17.710048 132358857959232 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0417 01:23:19.092524 132358857959232 maxtext_utils.py:1805] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0417 01:23:19.092644 132358857959232 maxtext_utils.py:1805] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0417 01:23:19.092699 132358857959232 maxtext_utils.py:1805] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0417 01:23:19.092742 132358857959232 maxtext_utils.py:1805] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0417 01:23:19.092778 132358857959232 maxtext_utils.py:1805] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0417 01:23:19.092812 132358857959232 maxtext_utils.py:1805] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0417 01:23:19.092864 132358857959232 maxtext_utils.py:1805] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0417 01:23:19.092900 132358857959232 maxtext_utils.py:1805] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0417 01:23:19.092935 132358857959232 maxtext_utils.py:1805] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0417 01:23:19.092968 132358857959232 maxtext_utils.py:1805] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0417 01:23:19.093000 132358857959232 maxtext_utils.py:1805] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0417 01:23:19.093031 132358857959232 maxtext_utils.py:1805] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0417 01:23:19.491587 132358857959232 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0417 01:23:19.491681 132358857959232 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 01:23:19.497027 132358857959232 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0417 01:23:19.497081 132358857959232 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 01:23:19.513474 132358857959232 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 01:23:19.513532 132358857959232 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 01:23:19.529268 132358857959232 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 01:23:19.529325 132358857959232 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 01:23:19.553372 132358857959232 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 01:23:19.553438 132358857959232 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 01:23:19.569198 132358857959232 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 01:23:19.569257 132358857959232 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 01:23:19.585023 132358857959232 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 01:23:19.585081 132358857959232 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 01:23:19.615117 132358857959232 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0417 01:23:19.615188 132358857959232 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 01:23:19.634935 132358857959232 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0417 01:23:19.634996 132358857959232 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0417 01:23:30.902723 132358857959232 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. I0417 01:23:30.905059 132358857959232 metric_logger.py:289] number parameters: 1.104 billion I0417 01:23:42.221634 132358857959232 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0417 01:23:42.372624 132358857959232 checkpointing.py:798] Waited 0.15098142623901367 seconds for step 0 to finish before starting checkpointing. I0417 01:23:42.374942 132358857959232 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 01:23:42.376646 132358857959232 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0 I0417 01:23:42.378021 132358857959232 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints/0. I0417 01:23:43.121928 132358857959232 signaling_client.py:364] Using JaxDistributedSignalingClient I0417 01:23:43.122947 132358857959232 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0417 01:23:43.123004 132358857959232 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 I0417 01:23:43.401292 132358857959232 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.280127s I0417 01:23:43.401470 132358857959232 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.333 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2892582416534424 s) (per-host) I0417 01:23:43.401527 132358857959232 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289325s (batch_requests_ready=0.003047s, total_serialization_initiated=0.286204s, others=0.000074s) I0417 01:23:43.401622 132358857959232 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293458s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.293439) I0417 01:23:43.402414 132358857959232 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.03 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints/0. I0417 01:23:43.402685 132232510527232 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 01:43:43.402662 I0417 01:23:43.407104 132358857959232 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0417 01:23:43.407347 132226608002816 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0417 01:23:43.407498 132358857959232 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776389022.3749218, 'wait_for_prev_duration_secs': 6.628036499023438e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776389022.3766844, 'checkpointer_blocking_duration_secs': 1.0261025428771973, 'get_old_steps_start_time': 1776389023.4028087, 'get_old_steps_duration_secs': 5.555152893066406e-05, 'checkpoint_manager_blocking_start_time': 1776389022.3730404, 'checkpoint_manager_blocking_duration_secs': 1.0344204902648926} I0417 01:23:43.407603 132358857959232 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0417 01:23:43.407672 132358857959232 max_utils.py:750] Memstats: After params initialized: I0417 01:23:43.407720 132358857959232 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0417 01:23:43.407754 132358857959232 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0417 01:23:43.407791 132358857959232 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0417 01:23:43.407829 132358857959232 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0417 01:23:43.726237 132358857959232 metric_logger.py:185] completed step: 0, seconds: 11.316, TFLOP/s/device: 1.201, Tokens/s/device: 180.991, total_weights: 65536, loss: 10.874 I0417 01:23:43.899159 132358857959232 metric_logger.py:185] completed step: 1, seconds: 1.503, TFLOP/s/device: 9.040, Tokens/s/device: 1362.635, total_weights: 65536, loss: 10.874 I0417 01:23:44.331787 132358857959232 metric_logger.py:185] completed step: 2, seconds: 0.027, TFLOP/s/device: 506.926, Tokens/s/device: 76409.357, total_weights: 65536, loss: 10.262 I0417 01:23:44.479217 132358857959232 metric_logger.py:185] completed step: 3, seconds: 0.422, TFLOP/s/device: 32.177, Tokens/s/device: 4850.127, total_weights: 65536, loss: 9.732 I0417 01:23:44.775058 132358857959232 metric_logger.py:185] completed step: 4, seconds: 0.164, TFLOP/s/device: 82.664, Tokens/s/device: 12459.998, total_weights: 65536, loss: 9.274 I0417 01:23:44.782634 132358857959232 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.434, Tokens/s/device: 13932.636, total_weights: 65536, loss: 8.888 I0417 01:23:46.412024 2762 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0417 01:23:49.175533 132228161750784 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints/0/items/array_metadatas/process_6 I0417 01:24:06.509202 132358857959232 metric_logger.py:185] completed step: 6, seconds: 0.297, TFLOP/s/device: 45.803, Tokens/s/device: 6903.852, total_weights: 65536, loss: 8.589 I0417 01:24:06.656549 132358857959232 metric_logger.py:185] completed step: 7, seconds: 21.580, TFLOP/s/device: 0.630, Tokens/s/device: 94.901, total_weights: 65536, loss: 8.380 I0417 01:24:06.803905 132358857959232 metric_logger.py:185] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.804, Tokens/s/device: 13385.446, total_weights: 65536, loss: 8.251 I0417 01:24:06.952258 132358857959232 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0417 01:24:06.952942 132358857959232 checkpointing.py:798] Waited 0.0006966590881347656 seconds for step 9 to finish before starting checkpointing. I0417 01:24:06.955025 132358857959232 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 01:24:19.629608 132232510527232 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 43.256 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.51735234260559 s) (per-host) I0417 01:24:19.629727 132232510527232 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.226969s. I0417 01:24:27.742999 132232510527232 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 44.340225s. I0417 01:24:27.743305 132226608002816 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 01:24:27.743426 132226608002816 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0417 01:24:27.743479 132226608002816 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0417 01:24:27.745615 132226608002816 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0417 01:24:27.745791 132358857959232 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0417 01:24:27.745953 132358857959232 checkpoint_manager.py:1452] Waiting for previous save to complete took 20.790928 seconds. If this number is high, consider checkpointing less frequently. I0417 01:24:27.747656 132358857959232 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9 I0417 01:24:27.749750 132358857959232 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints/9. I0417 01:24:28.461698 132358857959232 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0417 01:24:28.461788 132358857959232 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 I0417 01:24:28.497191 132358857959232 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037094s I0417 01:24:28.497338 132358857959232 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 35.043 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04401969909667969 s) (per-host) I0417 01:24:28.497389 132358857959232 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044078s (batch_requests_ready=0.002625s, total_serialization_initiated=0.041389s, others=0.000064s) I0417 01:24:28.497473 132358857959232 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.048216s (all_items=0.000014s, per_item={'items': '0.00001383'}, temp_paths=0.048203) I0417 01:24:28.498164 132358857959232 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints/9. I0417 01:24:28.498521 132226608002816 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 01:44:28.498484 I0417 01:24:28.500462 132358857959232 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0417 01:24:28.500753 132228161750784 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0417 01:24:28.500899 132358857959232 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776389046.954994, 'wait_for_prev_duration_secs': 20.79092812538147, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776389067.7476952, 'checkpointer_blocking_duration_secs': 0.7509748935699463, 'get_old_steps_start_time': 1776389068.4986916, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1776389046.953173, 'checkpoint_manager_blocking_duration_secs': 21.547691583633423} I0417 01:24:28.501051 132358857959232 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0417 01:24:28.501097 132358857959232 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 01:24:36.181304 132210208229120 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_003418/nnx_xpk_feat_nnx_post_train_fixes_20260417_003418_10_shardy_false/checkpoints/9/items/array_metadatas/process_6 I0417 01:25:10.997360 132226608002816 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.129 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.54400110244751 s) (per-host) I0417 01:25:10.997491 132226608002816 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.498856s. I0417 01:25:20.987576 132226608002816 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 52.488925s. I0417 01:25:20.987905 132228161750784 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 01:25:20.988033 132228161750784 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0417 01:25:20.988080 132228161750784 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0417 01:25:20.990408 132228161750784 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0417 01:25:20.990583 132358857959232 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0417 01:25:20.990751 132358857959232 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 01:25:20.991375 132358857959232 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 91.904, Tokens/s/device: 13852.720, total_weights: 65536, loss: 8.175 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Fri Apr 17 01:25:32 UTC 2026 EXIT_CODE=0