MaxView

XPK Start: Fri Apr 17 05:34:57 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 05:35:21.721715: 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 05:35:22.258277 132252995491648 max_utils.py:273] Attempting to initialize the jax distributed system...
I0417 05:35:31.298139 132252995491648 distributed.py:149] Starting JAX distributed service on [::]:8482
I0417 05:35:31.300477 132252995491648 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-4elvm-slice-job-0-0.mt-11-optimizer-offload-4elvm:8482
I0417 05:35:32.221123 132252995491648 max_utils.py:284] Jax distributed system initialized!
I0417 05:35:38.305999 132252995491648 max_utils.py:800] System Information: Jax Version: 0.9.2
I0417 05:35:38.306098 132252995491648 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0417 05:35:38.306138 132252995491648 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 05:35:38.306172 132252995491648 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0417 05:35:38.994964 132252995491648 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0417 05:35:39.107938 132252995491648 checkpointing.py:688] Setting up checkpoint logger...
I0417 05:35:39.108051 132252995491648 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0417 05:35:39.108096 132252995491648 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0417 05:35:39.108285 132252995491648 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 0x7847d5050110>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0417 05:35:41.905543 132252995491648 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0417 05:35:41.905776 132252995491648 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 0x78338874bec0>}, handler_registry=None
I0417 05:35:41.906037 132252995491648 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78338874bec0>` 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 05:35:41.906087 132252995491648 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78339846d220>` 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 05:35:41.906122 132252995491648 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 0x78338874bec0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78338874bec0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78339846d220>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78339846d220>}).
I0417 05:35:41.906437 132252995491648 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0417 05:35:41.906507 132252995491648 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7833984ddee0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0417 05:35:43.051608 132252995491648 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints
I0417 05:35:43.920795 132252995491648 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_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78339848e4e0>
I0417 05:35:43.920994 132252995491648 checkpointing.py:302] Checkpoint manager created!
I0417 05:35:44.882191 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to host
I0417 05:35:44.882303 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0417 05:35:44.882342 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882376 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882406 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882434 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0417 05:35:44.882460 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0417 05:35:44.882486 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882509 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882533 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882562 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882586 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882611 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0417 05:35:44.882633 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0417 05:35:44.882657 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882680 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882703 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882726 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0417 05:35:44.882750 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0417 05:35:44.882773 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882795 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882817 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882854 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882878 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0417 05:35:44.882901 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0417 05:35:44.882923 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to host
I0417 05:35:44.882946 132252995491648 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to host
I0417 05:35:44.884066 132252995491648 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0417 05:35:44.884122 132252995491648 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0417 05:35:46.338394 132252995491648 maxtext_utils.py:1805]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0417 05:35:46.338510 132252995491648 maxtext_utils.py:1805]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0417 05:35:46.338568 132252995491648 maxtext_utils.py:1805]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0417 05:35:46.338610 132252995491648 maxtext_utils.py:1805]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0417 05:35:46.338654 132252995491648 maxtext_utils.py:1805]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0417 05:35:46.338690 132252995491648 maxtext_utils.py:1805]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0417 05:35:46.338728 132252995491648 maxtext_utils.py:1805]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0417 05:35:46.338764 132252995491648 maxtext_utils.py:1805]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0417 05:35:46.338798 132252995491648 maxtext_utils.py:1805]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0417 05:35:46.338844 132252995491648 maxtext_utils.py:1805]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)

I0417 05:35:46.338878 132252995491648 maxtext_utils.py:1805]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0417 05:35:46.338910 132252995491648 maxtext_utils.py:1805]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0417 05:35:46.744837 132252995491648 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 05:35:46.744928 132252995491648 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 05:35:46.750194 132252995491648 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 05:35:46.750250 132252995491648 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 05:35:46.766443 132252995491648 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 05:35:46.766502 132252995491648 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 05:35:46.782036 132252995491648 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 05:35:46.782093 132252995491648 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 05:35:46.805763 132252995491648 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 05:35:46.805840 132252995491648 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 05:35:46.821555 132252995491648 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 05:35:46.821616 132252995491648 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 05:35:46.837297 132252995491648 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 05:35:46.837356 132252995491648 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 05:35:46.866812 132252995491648 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0417 05:35:46.866898 132252995491648 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 05:35:46.886534 132252995491648 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0417 05:35:46.886603 132252995491648 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0417 05:35:47.032919 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0417 05:35:47.033029 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:47.033070 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033109 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033139 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033167 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:47.033194 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:47.033218 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033244 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033269 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033296 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033321 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033347 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0417 05:35:47.033371 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:47.033396 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033419 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033443 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033466 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:47.033490 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:47.033516 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033546 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033570 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033594 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033617 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:47.033642 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0417 05:35:47.033664 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0417 05:35:47.033687 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0417 05:35:57.985313 132252995491648 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0417 05:35:57.987524 132252995491648 metric_logger.py:289] number parameters: 1.104 billion
I0417 05:35:58.526565 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0417 05:35:58.526680 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:58.526721 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.526762 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.526793 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.526834 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:58.526873 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:58.526901 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.526928 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.526955 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.526981 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527005 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527031 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0417 05:35:58.527061 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:58.527087 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527113 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527138 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527164 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:58.527187 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0417 05:35:58.527210 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527233 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527256 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527282 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527306 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0417 05:35:58.527330 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0417 05:35:58.527353 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0417 05:35:58.527375 132252995491648 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0417 05:36:08.898175 132252995491648 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0417 05:36:09.067771 132252995491648 checkpointing.py:798] Waited 0.16958379745483398 seconds for step 0 to finish before starting checkpointing.
I0417 05:36:09.070534 132252995491648 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 05:36:09.072456 132252995491648 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0417 05:36:09.073954 132252995491648 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints/0.
I0417 05:36:09.818037 132252995491648 signaling_client.py:364] Using JaxDistributedSignalingClient
I0417 05:36:09.818960 132252995491648 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0417 05:36:09.819018 132252995491648 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 05:36:10.097695 132252995491648 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.280441s
I0417 05:36:10.097883 132252995491648 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.313 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29034852981567383 s) (per-host)
I0417 05:36:10.097936 132252995491648 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.290411s (batch_requests_ready=0.003008s, total_serialization_initiated=0.287332s, others=0.000071s)
I0417 05:36:10.098028 132252995491648 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.294648s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.294629)
I0417 05:36:10.098747 132252995491648 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_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints/0.
I0417 05:36:10.100415 132118680164096 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 05:56:10.100379
I0417 05:36:10.102402 132252995491648 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 05:36:10.102711 132117978732288 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 05:36:10.102878 132252995491648 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776404169.0705073, 'wait_for_prev_duration_secs': 9.989738464355469e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776404169.0725138, 'checkpointer_blocking_duration_secs': 1.028052568435669, 'get_old_steps_start_time': 1776404170.1005902, 'get_old_steps_duration_secs': 3.2901763916015625e-05, 'checkpoint_manager_blocking_start_time': 1776404169.068195, 'checkpoint_manager_blocking_duration_secs': 1.0346426963806152}
I0417 05:36:10.102991 132252995491648 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0417 05:36:10.103062 132252995491648 max_utils.py:750] 
Memstats: After params initialized:
I0417 05:36:10.103113 132252995491648 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_24(process=6,(0,6,0,0))
I0417 05:36:10.103146 132252995491648 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_25(process=6,(1,6,0,0))
I0417 05:36:10.103173 132252995491648 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_28(process=6,(0,7,0,0))
I0417 05:36:10.103199 132252995491648 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_29(process=6,(1,7,0,0))
I0417 05:36:10.415132 132252995491648 metric_logger.py:185] completed step: 0, seconds: 10.910, TFLOP/s/device: 1.245, Tokens/s/device: 187.726, total_weights: 65536, loss: 10.874
I0417 05:36:10.608039 132252995491648 metric_logger.py:185] completed step: 1, seconds: 1.515, TFLOP/s/device: 8.967, Tokens/s/device: 1351.553, total_weights: 65536, loss: 10.874
I0417 05:36:11.000963 132252995491648 metric_logger.py:185] completed step: 2, seconds: 0.031, TFLOP/s/device: 432.051, Tokens/s/device: 65123.378, total_weights: 65536, loss: 10.262
I0417 05:36:11.163726 132252995491648 metric_logger.py:185] completed step: 3, seconds: 0.383, TFLOP/s/device: 35.517, Tokens/s/device: 5353.465, total_weights: 65536, loss: 9.732
I0417 05:36:11.490102 132252995491648 metric_logger.py:185] completed step: 4, seconds: 0.180, TFLOP/s/device: 75.625, Tokens/s/device: 11398.993, total_weights: 65536, loss: 9.274
I0417 05:36:11.497467 132252995491648 metric_logger.py:185] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.607, Tokens/s/device: 12602.224, total_weights: 65536, loss: 8.888
I0417 05:36:13.598006    2756 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0417 05:36:16.316143 132117987124992 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_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_6
I0417 05:36:33.319890 132252995491648 metric_logger.py:185] completed step: 6, seconds: 0.327, TFLOP/s/device: 41.563, Tokens/s/device: 6264.798, total_weights: 65536, loss: 8.589
I0417 05:36:33.482656 132252995491648 metric_logger.py:185] completed step: 7, seconds: 21.661, TFLOP/s/device: 0.627, Tokens/s/device: 94.548, total_weights: 65536, loss: 8.380
I0417 05:36:33.645476 132252995491648 metric_logger.py:185] completed step: 8, seconds: 0.168, TFLOP/s/device: 80.757, Tokens/s/device: 12172.507, total_weights: 65536, loss: 8.251
I0417 05:36:33.809379 132252995491648 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0417 05:36:33.810059 132252995491648 checkpointing.py:798] Waited 0.0006906986236572266 seconds for step 9 to finish before starting checkpointing.
I0417 05:36:33.812066 132252995491648 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 05:36:47.545328 132118680164096 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 41.855 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.73777484893799 s) (per-host)
I0417 05:36:47.545431 132118680164096 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.444908s.
I0417 05:36:56.020269 132118680164096 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 45.919730s.
I0417 05:36:56.020559 132117978732288 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 05:36:56.020674 132117978732288 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 05:36:56.020730 132117978732288 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0417 05:36:56.022953 132117978732288 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0417 05:36:56.023157 132252995491648 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 05:36:56.023293 132252995491648 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.211230 seconds. If this number is high, consider checkpointing less frequently.
I0417 05:36:56.025071 132252995491648 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0417 05:36:56.027235 132252995491648 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints/9.
I0417 05:36:56.740297 132252995491648 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0417 05:36:56.740392 132252995491648 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 05:36:56.777336 132252995491648 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.038697s
I0417 05:36:56.777507 132252995491648 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 33.700 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04577231407165527 s) (per-host)
I0417 05:36:56.777558 132252995491648 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.045834s (batch_requests_ready=0.002877s, total_serialization_initiated=0.042888s, others=0.000068s)
I0417 05:36:56.777666 132252995491648 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.050070s (all_items=0.000014s, per_item={'items': '0.00001431'}, temp_paths=0.050056)
I0417 05:36:56.778328 132252995491648 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_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints/9.
I0417 05:36:56.778646 132098341979904 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 05:56:56.778607
I0417 05:36:56.780458 132252995491648 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 05:36:56.780687 132117978732288 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 05:36:56.780806 132252995491648 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776404193.8120348, 'wait_for_prev_duration_secs': 22.2112295627594, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776404216.02511, 'checkpointer_blocking_duration_secs': 0.7536840438842773, 'get_old_steps_start_time': 1776404216.7788162, 'get_old_steps_duration_secs': 4.124641418457031e-05, 'checkpoint_manager_blocking_start_time': 1776404193.8102922, 'checkpoint_manager_blocking_duration_secs': 22.97048044204712}
I0417 05:36:56.780920 132252995491648 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0417 05:36:56.780962 132252995491648 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 05:37:02.719737 132117987124992 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_023710/nnx_xpk_feat_nnx_post_train_fixes_20260417_023710_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_6
I0417 05:37:38.864007 132098341979904 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.490 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.13224530220032 s) (per-host)
I0417 05:37:38.864114 132098341979904 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.085353s.
I0417 05:37:46.662157 132098341979904 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 49.883380s.
I0417 05:37:46.662410 132117978732288 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 05:37:46.662466 132117978732288 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 05:37:46.662516 132117978732288 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0417 05:37:46.664226 132117978732288 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0417 05:37:46.664405 132252995491648 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 05:37:46.664572 132252995491648 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 05:37:46.665223 132252995491648 metric_logger.py:185] completed step: 9, seconds: 0.163, TFLOP/s/device: 83.272, Tokens/s/device: 12551.635, 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 05:37:58 UTC 2026
EXIT_CODE=0