MaxView

XPK Start: Sat Apr 25 13:20:32 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. 
2026-04-25 13:20:57.751571: 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)
I0425 13:20:57.963004 136347792000832 max_utils.py:273] Attempting to initialize the jax distributed system...
I0425 13:21:07.005442 136347792000832 distributed.py:149] Starting JAX distributed service on [::]:8482
I0425 13:21:07.007679 136347792000832 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-true-sv-kdegq-slice-job-0-0.mt-16-zarr3-true-sv-kdegq:8482
I0425 13:21:07.819073 136347792000832 max_utils.py:284] Jax distributed system initialized!
I0425 13:21:14.026887 136347792000832 max_utils.py:800] System Information: Jax Version: 0.9.2
I0425 13:21:14.026992 136347792000832 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0425 13:21:14.027032 136347792000832 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
I0425 13:21:14.027066 136347792000832 train_utils.py:391] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0425 13:21:14.724248 136347792000832 maxtext_utils.py:1771] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0425 13:21:14.724822 136347792000832 maxtext_utils.py:1771] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0425 13:21:14.725002 136347792000832 checkpointing.py:688] Setting up checkpoint logger...
I0425 13:21:14.725048 136347792000832 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0425 13:21:14.725091 136347792000832 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0425 13:21:14.725491 136347792000832 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 0x7c0151518200>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0425 13:21:17.611956 136347792000832 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0425 13:21:17.612215 136347792000832 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 0x7becf46efb90>}, handler_registry=None
I0425 13:21:17.612470 136347792000832 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7becf46efb90>` 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`.
I0425 13:21:17.612520 136347792000832 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7becf46f35f0>` 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`.
I0425 13:21:17.612555 136347792000832 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 0x7becf46efb90>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7becf46efb90>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7becf46f35f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7becf46f35f0>}).
I0425 13:21:17.612875 136347792000832 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0425 13:21:17.612947 136347792000832 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7becf44c1d00> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0425 13:21:18.823661 136347792000832 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints
I0425 13:21:18.825918 136347792000832 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=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_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7becf46f0f50>
I0425 13:21:18.826031 136347792000832 checkpointing.py:302] Checkpoint manager created!
I0425 13:21:19.763830 136347792000832 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0425 13:21:19.763942 136347792000832 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 13:21:20.141882 136347792000832 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch_attn', 'activation_length_attn', 'activation_embed_attn').
I0425 13:21:20.141980 136347792000832 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 13:21:20.158663 136347792000832 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch_attn', 'activation_length_attn', 'activation_embed_attn').
I0425 13:21:20.158726 136347792000832 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 13:21:20.182860 136347792000832 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 13:21:20.182927 136347792000832 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 13:21:20.199575 136347792000832 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 13:21:20.199634 136347792000832 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 13:21:20.216202 136347792000832 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 13:21:20.216263 136347792000832 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 13:21:20.241320 136347792000832 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch_attn', 'activation_length_attn', 'activation_heads', 'activation_kv').
I0425 13:21:20.241392 136347792000832 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 13:21:20.262403 136347792000832 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0425 13:21:20.262467 136347792000832 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0425 13:21:20.482017 136347792000832 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0425 13:21:20.482142 136347792000832 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0425 13:21:21.926675 136347792000832 maxtext_utils.py:1880]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0425 13:21:21.926804 136347792000832 maxtext_utils.py:1880]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0425 13:21:21.926859 136347792000832 maxtext_utils.py:1880]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0425 13:21:21.926917 136347792000832 maxtext_utils.py:1880]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0425 13:21:21.926971 136347792000832 maxtext_utils.py:1880]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0425 13:21:21.927010 136347792000832 maxtext_utils.py:1880]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0425 13:21:21.927062 136347792000832 maxtext_utils.py:1880]  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)
I0425 13:21:21.927128 136347792000832 maxtext_utils.py:1880]  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')
I0425 13:21:21.927171 136347792000832 maxtext_utils.py:1880]  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)
I0425 13:21:21.927206 136347792000832 maxtext_utils.py:1880]  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)
I0425 13:21:21.927259 136347792000832 maxtext_utils.py:1880]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0425 13:21:21.927313 136347792000832 maxtext_utils.py:1880]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')

    Physical:  (None, 'fsdp')
I0425 13:21:22.427742 136347792000832 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 13:21:22.427850 136347792000832 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 13:21:22.443834 136347792000832 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 13:21:22.443896 136347792000832 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 13:21:33.472818 136347792000832 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.
I0425 13:21:33.473581 136347792000832 metric_logger.py:301] number parameters: 1.104 billion
I0425 13:21:44.812003 136347792000832 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0425 13:21:44.981860 136347792000832 checkpointing.py:798] Waited 0.16983795166015625 seconds for step 0 to finish before starting checkpointing.
I0425 13:21:44.984214 136347792000832 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 13:21:44.986086 136347792000832 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0425 13:21:44.987602 136347792000832 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/0.
I0425 13:21:45.836812 136347792000832 signaling_client.py:364] Using JaxDistributedSignalingClient
I0425 13:21:45.837767 136347792000832 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 13:21:45.837823 136347792000832 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
I0425 13:21:46.115078 136347792000832 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.278340s
I0425 13:21:46.115275 136347792000832 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.431 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2840125560760498 s) (per-host)
I0425 13:21:46.115331 136347792000832 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.284079s (batch_requests_ready=0.002480s, total_serialization_initiated=0.281524s, others=0.000076s)
I0425 13:21:46.115426 136347792000832 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.289385s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.289366)
I0425 13:21:46.116241 136347792000832 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.13 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/0.
I0425 13:21:46.116529 136218377451264 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 13:41:46.116496
I0425 13:21:46.549867 136347792000832 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 13:21:46.550267 136217260930816 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 13:21:46.550433 136347792000832 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777123304.9841952, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777123304.9861412, 'checkpointer_blocking_duration_secs': 1.1305410861968994, 'get_old_steps_start_time': 1777123306.1167076, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1777123304.9823706, 'checkpoint_manager_blocking_duration_secs': 1.5680220127105713}
I0425 13:21:46.550544 136347792000832 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0425 13:21:46.550595 136347792000832 max_utils.py:750] 
Memstats: After params initialized:
I0425 13:21:46.550650 136347792000832 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0425 13:21:46.550683 136347792000832 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0425 13:21:46.550711 136347792000832 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0425 13:21:46.550734 136347792000832 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0425 13:21:46.870257 136347792000832 metric_logger.py:196] completed step: 0, seconds: 11.338, TFLOP/s/device: 1.198, Tokens/s/device: 180.627, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0425 13:21:47.041577 136347792000832 metric_logger.py:196] completed step: 1, seconds: 2.057, TFLOP/s/device: 6.606, Tokens/s/device: 995.728, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0425 13:21:47.476634 136347792000832 metric_logger.py:196] completed step: 2, seconds: 0.024, TFLOP/s/device: 573.442, Tokens/s/device: 86435.384, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0425 13:21:47.623950 136347792000832 metric_logger.py:196] completed step: 3, seconds: 0.436, TFLOP/s/device: 31.184, Tokens/s/device: 4700.439, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0425 13:21:47.628347 136347792000832 checkpointing.py:794] Waiting for step 5 to finish before checkpoint...
I0425 13:21:47.917948 136347792000832 checkpointing.py:798] Waited 0.2895667552947998 seconds for step 5 to finish before starting checkpointing.
I0425 13:21:47.921064 136347792000832 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 13:21:48.974773    2845 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0425 13:21:51.010688 136217298097920 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/0/items/array_metadatas/process_6
I0425 13:22:22.659371 136218377451264 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 42.891 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.82804584503174 s) (per-host)
I0425 13:22:22.659503 136218377451264 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.542862s.
I0425 13:22:29.237270 136218377451264 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 43.120612s.
I0425 13:22:29.237567 136217260930816 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 13:22:29.237688 136217260930816 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 13:22:29.237740 136217260930816 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0425 13:22:29.239498 136217260930816 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0425 13:22:29.239673 136347792000832 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.
W0425 13:22:29.239811 136347792000832 checkpoint_manager.py:1452] Waiting for previous save to complete took 41.318745 seconds. If this number is high, consider checkpointing less frequently.
I0425 13:22:29.241782 136347792000832 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 5
I0425 13:22:29.244311 136347792000832 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/5.
I0425 13:22:29.523587 136347792000832 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 13:22:29.523680 136347792000832 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
I0425 13:22:29.557901 136347792000832 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035402s
I0425 13:22:29.558069 136347792000832 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 39.782 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03877592086791992 s) (per-host)
I0425 13:22:29.558133 136347792000832 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038848s (batch_requests_ready=0.001738s, total_serialization_initiated=0.037031s, others=0.000080s)
I0425 13:22:29.558221 136347792000832 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043869s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.043852)
I0425 13:22:29.558947 136347792000832 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.32 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/5.
I0425 13:22:29.559260 136217260930816 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 13:42:29.559227
I0425 13:22:29.569617 136347792000832 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 13:22:29.569907 136218344503040 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 13:22:29.570067 136347792000832 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777123307.9210348, 'wait_for_prev_duration_secs': 41.318745374679565, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777123349.2418208, 'checkpointer_blocking_duration_secs': 0.317584753036499, 'get_old_steps_start_time': 1777123349.5594292, 'get_old_steps_duration_secs': 3.337860107421875e-05, 'checkpoint_manager_blocking_start_time': 1777123307.9185922, 'checkpoint_manager_blocking_duration_secs': 41.65144109725952}
I0425 13:22:29.570197 136347792000832 checkpointing.py:409] Started an asynchronous checkpoint save for step 5
I0425 13:22:29.570987 136347792000832 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.898, Tokens/s/device: 13399.721, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0425 13:22:29.577042 136347792000832 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.301, Tokens/s/device: 13912.571, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0425 13:22:34.610820 136217298097920 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/5/items/array_metadatas/process_6
I0425 13:22:52.111527 136347792000832 metric_logger.py:196] completed step: 6, seconds: 41.947, TFLOP/s/device: 0.324, Tokens/s/device: 48.823, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0425 13:22:52.258757 136347792000832 metric_logger.py:196] completed step: 7, seconds: 22.388, TFLOP/s/device: 0.607, Tokens/s/device: 91.476, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0425 13:22:52.408524 136347792000832 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.280, Tokens/s/device: 13457.305, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0425 13:22:52.554984 136347792000832 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0425 13:22:52.555671 136347792000832 checkpointing.py:798] Waited 0.0007100105285644531 seconds for step 9 to finish before starting checkpointing.
I0425 13:22:52.557723 136347792000832 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 13:23:11.344142 136217260930816 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.767 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.82478713989258 s) (per-host)
I0425 13:23:11.344272 136217260930816 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.784898s.
I0425 13:23:17.219652 136217260930816 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 47.660262s.
I0425 13:23:17.219935 136218344503040 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 13:23:17.220084 136218344503040 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 13:23:17.220177 136218344503040 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0425 13:23:17.221511 136218344503040 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0425 13:23:17.221702 136347792000832 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0425 13:23:17.221826 136347792000832 checkpoint_manager.py:1452] Waiting for previous save to complete took 24.664109 seconds. If this number is high, consider checkpointing less frequently.
I0425 13:23:17.223269 136347792000832 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0425 13:23:17.225403 136347792000832 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/9.
I0425 13:23:17.525272 136347792000832 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 13:23:17.525363 136347792000832 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
I0425 13:23:17.558961 136347792000832 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.034695s
I0425 13:23:17.559144 136347792000832 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 40.432 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.038152217864990234 s) (per-host)
I0425 13:23:17.559200 136347792000832 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038218s (batch_requests_ready=0.001637s, total_serialization_initiated=0.036510s, others=0.000072s)
I0425 13:23:17.559292 136347792000832 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.042109s (all_items=0.000011s, per_item={'items': '0.00001073'}, temp_paths=0.042098)
I0425 13:23:17.560011 136347792000832 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/9.
I0425 13:23:17.560326 136218344503040 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 13:43:17.560291
I0425 13:23:17.571272 136347792000832 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 13:23:17.571527 136217298097920 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 13:23:17.571687 136347792000832 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777123372.5576944, 'wait_for_prev_duration_secs': 24.664109230041504, 'time_between_consecutive_saves_sec': 23.318158864974976, 'checkpointer_blocking_start_time': 1777123397.22331, 'checkpointer_blocking_duration_secs': 0.3371593952178955, 'get_old_steps_start_time': 1777123397.5604923, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1777123372.5558937, 'checkpoint_manager_blocking_duration_secs': 25.01576256752014}
I0425 13:23:17.571799 136347792000832 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0425 13:23:17.571839 136347792000832 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 13:23:22.614381 136216213182208 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260425_121355/linen_xpk_feat_nnx_post_train_fixes_20260425_121355_16_zarr3_true/checkpoints/9/items/array_metadatas/process_6
I0425 13:23:59.026747 136218344503040 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 38.057 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.5057156085968 s) (per-host)
I0425 13:23:59.026872 136218344503040 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.466437s.
I0425 13:24:05.302921 136218344503040 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 47.742470s.
I0425 13:24:05.303224 136217298097920 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 13:24:05.303344 136217298097920 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 13:24:05.303386 136217298097920 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0425 13:24:05.305025 136217298097920 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0425 13:24:05.305225 136347792000832 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.
I0425 13:24:05.305372 136347792000832 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 13:24:05.306407 136347792000832 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.313, Tokens/s/device: 13914.367, 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: Sat Apr 25 13:24:17 UTC 2026
EXIT_CODE=0