retention: prune 1 old checkpoints dirs

checkpoints/step2456209_20260828/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/scaler.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c2110dfd7b7a6c034fb62244a6f8f76a22d9936d55b6d0fe133e43e65b38631c
-size 149

checkpoints/step2456209_20260828/serving_candidate.json

@@ -1,546 +0,0 @@
-{
-  "agillm43_provenance": {
-    "batch_size": 60,
-    "block_size": 2048,
-    "checkpoint_path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt",
-    "checkpoint_type": "full",
-    "created_at_iso": "2026-08-28T04:14:06Z",
-    "created_at_unix": 1787890446.1632168,
-    "created_from_scratch": false,
-    "dataset_provenance": {
-      "chat": false,
-      "dataset_field_text": "text",
-      "dataset_hotload_config_path": "/workspace/agillm43_dataset_hotload.json",
-      "dataset_hotload_config_sha256": "2762ba666c280207bdb2293b6d382ef8077d167de79609e5bcc35d4499fd1527",
-      "dataset_hotload_config_stamp": [
-        73,
-        2196426857,
-        1787483694198261655,
-        1281
-      ],
-      "dataset_hotload_config_used": true,
-      "dataset_hotload_enabled": true,
-      "dataset_hotload_generation": 4,
-      "dataset_hotload_schema": "agillm.dataset_hotload.config.v2",
-      "dataset_revision_map_sha256": "8361995a504a2365136b20e1a879ba78944a06309f13f3f013bbd0cdfc72225b",
-      "dataset_revisions": {
-        "EleutherAI/proof-pile-2": "901a9273a770e9d4138c5ddd91802f9c5c6cdc4b",
-        "HuggingFaceFW/fineweb": "9bb295ddab0e05d785b879661af7260fed5140fc",
-        "HuggingFaceFW/fineweb-edu": "87f09149ef4734204d70ed1d046ddc9ca3f2b8f9",
-        "HuggingFaceTB/cosmopedia": "0ae6ec63f91742bd2d1eaef4f02232c55d719385",
-        "HuggingFaceTB/finemath": "e92b25a616738fe95dc186b64dfb19f9c8525594",
-        "HuggingFaceTB/smollm-corpus": "3ba9d605774198c5868892d7a8deda78031a781f",
-        "Skylion007/openwebtext": "79d93d786212f7344586290adb811d4ae6a1762c",
-        "allenai/c4": "1588ec454efa1a09f29cd18ddd04fe05fc8653a2",
-        "allenai/dolma": "7f48140530a023e9ea4c5cfb141160922727d4d3",
-        "allenai/dolmino-mix-1124": "a319f19eef1e257417b11ea8c30da266ae175557",
-        "codeparrot/codeparrot-clean": "35a59fb025bc0a102f7d96eac09d145b896d487b",
-        "mlfoundations/dclm-baseline-1.0": "a3b142c183aebe5af344955ae20836eb34dcf69b",
-        "wikimedia/wikipedia": "b04c8d1ceb2f5cd4588862100d08de323dccfbaa"
-      },
-      "dolma_payload_integrity": "external_url_manifest_pinned_payload_hashes_unavailable",
-      "dolma_url_list_sha256": "bc6ef505be920c94bb9e7147445638f606fdc3f5af622cd1d887dc48e21a2d5b",
-      "dolma_url_manifest_raw_sha256": "6fe8c40d51e5bebc9303453d7f533b232d69f7ae122123621a82621b1cd779b7",
-      "has_language_mix": true,
-      "has_numeracy": false,
-      "hot_config_expected_sha256": "",
-      "hot_config_has_append": true,
-      "hot_config_has_datasets": false,
-      "hot_config_mtime": 1787722590.529335,
-      "hot_config_path": "/workspace/agillm43_production_v14/contracts/hot_config_owner_v25.json",
-      "hot_config_used": true,
-      "hot_reload_capable": true,
-      "phase": "pretrain",
-      "proof_pile_lfs_manifest_sha256": "d020dfb796b02a75564e4e11d9e03185a55b3f9c6639f87830cfbe10737ed329",
-      "proof_pile_paths_sha256": "8a30018017e3b3d9a9bcfa06a3c404ccd113297b0b1d0c48e2e2aadf113e52b0",
-      "schema": "agillm.dataset_provenance.v1",
-      "sequence_prefetch_depth": 256,
-      "source_count": 15,
-      "source_effective": "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25,HuggingFaceTB/finemath:finemath-4plus|0.7,HuggingFaceTB/finemath:infiwebmath-4plus|0.35,HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8,allenai/dolmino-mix-1124|1,mlfoundations/dclm-baseline-1.0|1,HuggingFaceFW/fineweb-edu:sample-100BT|1.6",
-      "source_requested": "HuggingFaceFW/fineweb-edu:sample-10BT|1.60,HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25",
-      "source_specs": [
-        "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50",
-        "wikimedia/wikipedia:20231101.en|1.10",
-        "allenai/c4:en|0.40",
-        "Skylion007/openwebtext|0.35",
-        "HuggingFaceTB/cosmopedia:web_samples_v2|1.30",
-        "HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30",
-        "EleutherAI/proof-pile-2:all|1.50",
-        "allenai/dolma:v1_6-sample|0.50",
-        "codeparrot/codeparrot-clean|1.25",
-        "HuggingFaceTB/finemath:finemath-4plus|0.7",
-        "HuggingFaceTB/finemath:infiwebmath-4plus|0.35",
-        "HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8",
-        "allenai/dolmino-mix-1124|1",
-        "mlfoundations/dclm-baseline-1.0|1",
-        "HuggingFaceFW/fineweb-edu:sample-100BT|1.6"
-      ],
-      "val_source_effective": "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25,HuggingFaceTB/finemath:finemath-4plus|0.7,HuggingFaceTB/finemath:infiwebmath-4plus|0.35,HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8,allenai/dolmino-mix-1124|1,mlfoundations/dclm-baseline-1.0|1,HuggingFaceFW/fineweb-edu:sample-100BT|1.6",
-      "val_source_requested": ""
-    },
-    "effective_global_step": 4892266,
-    "effective_seen_tok": 235535175680,
-    "global_origin_seen_tok": 0,
-    "global_origin_step": 2436057,
-    "gpu": {
-      "allocated_gb": 6.83,
-      "peak_allocated_gb": 18.92,
-      "reserved_gb": 22.59
-    },
-    "hostname": "ead6505f92c4",
-    "lane": "pretrain",
-    "launch": {
-      "argv": [
-        "/workspace/agillm43_production_v14/runtime/agillm43_singlefile_intelligence_v25_satvar_pretrain_final_20260826.py",
-        "train",
-        "--preset",
-        "agillm4_floor",
-        "--block",
-        "2048",
-        "--source",
-        "HuggingFaceFW/fineweb-edu:sample-10BT|1.60,HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25",
-        "--nat_mask_token_id",
-        "2",
-        "--tie_weights",
-        "--tie_kv",
-        "--moe_ffn",
-        "--moe_experts",
-        "2",
-        "--moe_top_k",
-        "1",
-        "--moe_mlp_mult",
-        "4",
-        "--moe_shared_experts",
-        "1",
-        "--moe_shared_mlp_mult",
-        "2",
-        "--moe_aux_coef",
-        "0.01",
-        "--moe_z_coef",
-        "0.001",
-        "--amp",
-        "--grad_checkpoint",
-        "--attn_backend",
-        "sdpa",
-        "--sublinear_window",
-        "128",
-        "--sublinear_stride",
-        "128",
-        "--sublinear_max_anchors",
-        "128",
-        "--sublinear_chunk",
-        "128",
-        "--sublinear_sinks",
-        "4",
-        "--sublinear_recent_anchors",
-        "64",
-        "--no-sublinear_pooled_landmarks",
-        "--anchor_stride",
-        "256",
-        "--anchor_max",
-        "2048",
-        "--anchor_position",
-        "-1",
-        "--alibi_mode",
-        "corrected",
-        "--alibi_scale",
-        "0.0",
-        "--optimizer",
-        "adamw8bit",
-        "--weight_decay",
-        "0.0",
-        "--lr_decay",
-        "cosine",
-        "--lr_decay_tokens",
-        "49000954880",
-        "--lr_min_mult",
-        "0.2",
-        "--lr_schedule_reset_on_resume",
-        "--lr_warmup_tokens",
-        "10000000",
-        "--lr_warmup_min_mult",
-        "0.1",
-        "--dblock",
-        "--dblock_blocks",
-        "14",
-        "--dblock_schedule",
-        "roundrobin",
-        "--dblock_router",
-        "shadow",
-        "--dblock_router_blend",
-        "0.0",
-        "--dblock_router_ramp_steps",
-        "64",
-        "--dblock_warmup_steps",
-        "14",
-        "--dblock_sublayer_mode",
-        "off",
-        "--dblock_checkpoint_skip_tail",
-        "0",
-        "--dblock_loop_layers",
-        "0",
-        "--dblock_loop_start",
-        "0",
-        "--dblock_loop_cond_scale",
-        "1.0",
-        "--dblock_sigma_curriculum_steps",
-        "2000",
-        "--dblock_sigma_sampling",
-        "lognormal",
-        "--dblock_sigma_stratified",
-        "--dblock_sigma_min",
-        "0.002",
-        "--dblock_sigma_max",
-        "2.0",
-        "--dblock_sigma_pmean",
-        "-1.2",
-        "--dblock_sigma_pstd",
-        "1.2",
-        "--dblock_edm_wmax",
-        "5.0",
-        "--dblock_ar_prob",
-        "0.8",
-        "--dblock_sat_prob",
-        "0.1",
-        "--dblock_nat_prob",
-        "0.1",
-        "--dblock_ar_loss_tokens",
-        "8192",
-        "--dblock_sat_loss_tokens",
-        "4096",
-        "--dblock_nat_loss_tokens",
-        "4096",
-        "--dblock_nat_embed_noise_mode",
-        "off",
-        "--dblock_nat_embed_noise_scale",
-        "1.0",
-        "--nat_loss_weight",
-        "1.0",
-        "--nat_mask_ratio",
-        "0.5",
-        "--nat_max_tokens",
-        "0",
-        "--nat_span_mask_prob",
-        "0.35",
-        "--nat_suffix_mask_prob",
-        "0.20",
-        "--nat_span_max_tokens",
-        "0",
-        "--loss_spike_skip",
-        "0.0",
-        "--sat_every",
-        "1",
-        "--nat_every",
-        "1",
-        "--dblock_fullstack_anchor_spike_skip",
-        "0.0",
-        "--dblock_fullstack_anchor_max_ce",
-        "0.0",
-        "--dblock_anchor_softcap_mult",
-        "4.0",
-        "--dblock_anchor_softcap_min_ce",
-        "25.0",
-        "--dblock_anchor_softcap_max_ce",
-        "50.0",
-        "--dblock_fullstack_ar_every",
-        "3",
-        "--dblock_fullstack_ar_offset",
-        "0",
-        "--dblock_fullstack_ar_tokens",
-        "256",
-        "--dblock_fullstack_sat_every",
-        "3",
-        "--dblock_fullstack_sat_offset",
-        "1",
-        "--dblock_fullstack_sat_tokens",
-        "128",
-        "--dblock_fullstack_nat_every",
-        "3",
-        "--dblock_fullstack_nat_offset",
-        "2",
-        "--dblock_fullstack_nat_tokens",
-        "128",
-        "--dblock_fullstack_nat_mask_id",
-        "-1",
-        "--ckpt_codec",
-        "block-sharded-zstd",
-        "--async_update_every_steps",
-        "0",
-        "--dblock_router_hidden",
-        "64",
-        "--dblock_router_heads",
-        "4",
-        "--dblock_router_layers",
-        "2",
-        "--dblock_router_lr",
-        "0.0005",
-        "--dblock_explore",
-        "0.08",
-        "--dblock_max_stale_steps",
-        "64",
-        "--dblock_max_count_skew",
-        "1.35",
-        "--dblock_stale_bonus",
-        "0.35",
-        "--dblock_undertrain_bonus",
-        "0.25",
-        "--dblock_activation_offload_min_mb",
-        "1.0",
-        "--nat_document_boundary_aware",
-        "--dblock_fullstack_anchor_deterministic_eval",
-        "--dblock_fullstack_anchor_spike_retry_limit",
-        "3",
-        "--dblock_local_spike_retry_limit",
-        "3",
-        "--dblock_fullstack_nat_no_valid_crop_limit",
-        "8",
-        "--target_tokens",
-        "400000000000",
-        "--lr_schedule_reanchor_on_resume",
-        "--save_every_sec",
-        "1800",
-        "--delta_every_steps",
-        "0",
-        "--delta_every_sec",
-        "0",
-        "--delta_max_keep",
-        "0",
-        "--max_ckpts",
-        "2",
-        "--disk_free_floor_gb",
-        "12.0",
-        "--heartbeat_every_sec",
-        "30",
-        "--dblock_fullstack_ar_weight",
-        "1.0",
-        "--dblock_fullstack_sat_weight",
-        "0.1",
-        "--dblock_fullstack_nat_weight",
-        "0.1",
-        "--dblock_local_aux_warn_ce",
-        "0.0",
-        "--dblock_local_aux_max_ce",
-        "0.0",
-        "--lr_schedule_origin_tokens",
-        "85002387456",
-        "--dblock_stop_after_commits",
-        "0",
-        "--no-oom_auto_backoff",
-        "--require_exact_resume_state",
-        "--dblock_objective_mode",
-        "committed_9_1_1",
-        "--dblock_ar_weight",
-        "0.25",
-        "--dblock_sat_weight",
-        "0.25",
-        "--dblock_nat_weight",
-        "0.25",
-        "--dblock_aux_grad_ratio",
-        "0.5",
-        "--dblock_min_ar_share",
-        "0.8",
-        "--lr_core",
-        "2e-7",
-        "--lr_head",
-        "2e-7",
-        "--batch_size",
-        "60",
-        "--oom_known_safe_batch",
-        "56",
-        "--dblock_checkpoint_stride",
-        "2",
-        "--empty_cache_every_steps",
-        "0",
-        "--dblock_log_every",
-        "25",
-        "--cuda_max_reserved_mib",
-        "0",
-        "--cuda_min_free_mib",
-        "0",
-        "--weight_accel_mode",
-        "shadow",
-        "--weight_accel_dir",
-        "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/weight_trajectory_accel",
-        "--weight_accel_every",
-        "1",
-        "--weight_accel_train_every",
-        "4",
-        "--weight_accel_record_every",
-        "14",
-        "--weight_accel_state_every",
-        "256",
-        "--resume",
-        "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt",
-        "--save_dir",
-        "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B",
-        "--data_seed",
-        "2436099"
-      ],
-      "argv_string": "/workspace/agillm43_production_v14/runtime/agillm43_singlefile_intelligence_v25_satvar_pretrain_final_20260826.py train --preset agillm4_floor --block 2048 --source HuggingFaceFW/fineweb-edu:sample-10BT|1.60,HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25 --nat_mask_token_id 2 --tie_weights --tie_kv --moe_ffn --moe_experts 2 --moe_top_k 1 --moe_mlp_mult 4 --moe_shared_experts 1 --moe_shared_mlp_mult 2 --moe_aux_coef 0.01 --moe_z_coef 0.001 --amp --grad_checkpoint --attn_backend sdpa --sublinear_window 128 --sublinear_stride 128 --sublinear_max_anchors 128 --sublinear_chunk 128 --sublinear_sinks 4 --sublinear_recent_anchors 64 --no-sublinear_pooled_landmarks --anchor_stride 256 --anchor_max 2048 --anchor_position -1 --alibi_mode corrected --alibi_scale 0.0 --optimizer adamw8bit --weight_decay 0.0 --lr_decay cosine --lr_decay_tokens 49000954880 --lr_min_mult 0.2 --lr_schedule_reset_on_resume --lr_warmup_tokens 10000000 --lr_warmup_min_mult 0.1 --dblock --dblock_blocks 14 --dblock_schedule roundrobin --dblock_router shadow --dblock_router_blend 0.0 --dblock_router_ramp_steps 64 --dblock_warmup_steps 14 --dblock_sublayer_mode off --dblock_checkpoint_skip_tail 0 --dblock_loop_layers 0 --dblock_loop_start 0 --dblock_loop_cond_scale 1.0 --dblock_sigma_curriculum_steps 2000 --dblock_sigma_sampling lognormal --dblock_sigma_stratified --dblock_sigma_min 0.002 --dblock_sigma_max 2.0 --dblock_sigma_pmean -1.2 --dblock_sigma_pstd 1.2 --dblock_edm_wmax 5.0 --dblock_ar_prob 0.8 --dblock_sat_prob 0.1 --dblock_nat_prob 0.1 --dblock_ar_loss_tokens 8192 --dblock_sat_loss_tokens 4096 --dblock_nat_loss_tokens 4096 --dblock_nat_embed_noise_mode off --dblock_nat_embed_noise_scale 1.0 --nat_loss_weight 1.0 --nat_mask_ratio 0.5 --nat_max_tokens 0 --nat_span_mask_prob 0.35 --nat_suffix_mask_prob 0.20 --nat_span_max_tokens 0 --loss_spike_skip 0.0 --sat_every 1 --nat_every 1 --dblock_fullstack_anchor_spike_skip 0.0 --dblock_fullstack_anchor_max_ce 0.0 --dblock_anchor_softcap_mult 4.0 --dblock_anchor_softcap_min_ce 25.0 --dblock_anchor_softcap_max_ce 50.0 --dblock_fullstack_ar_every 3 --dblock_fullstack_ar_offset 0 --dblock_fullstack_ar_tokens 256 --dblock_fullstack_sat_every 3 --dblock_fullstack_sat_offset 1 --dblock_fullstack_sat_tokens 128 --dblock_fullstack_nat_every 3 --dblock_fullstack_nat_offset 2 --dblock_fullstack_nat_tokens 128 --dblock_fullstack_nat_mask_id -1 --ckpt_codec block-sharded-zstd --async_update_every_steps 0 --dblock_router_hidden 64 --dblock_router_heads 4 --dblock_router_layers 2 --dblock_router_lr 0.0005 --dblock_explore 0.08 --dblock_max_stale_steps 64 --dblock_max_count_skew 1.35 --dblock_stale_bonus 0.35 --dblock_undertrain_bonus 0.25 --dblock_activation_offload_min_mb 1.0 --nat_document_boundary_aware --dblock_fullstack_anchor_deterministic_eval --dblock_fullstack_anchor_spike_retry_limit 3 --dblock_local_spike_retry_limit 3 --dblock_fullstack_nat_no_valid_crop_limit 8 --target_tokens 400000000000 --lr_schedule_reanchor_on_resume --save_every_sec 1800 --delta_every_steps 0 --delta_every_sec 0 --delta_max_keep 0 --max_ckpts 2 --disk_free_floor_gb 12.0 --heartbeat_every_sec 30 --dblock_fullstack_ar_weight 1.0 --dblock_fullstack_sat_weight 0.1 --dblock_fullstack_nat_weight 0.1 --dblock_local_aux_warn_ce 0.0 --dblock_local_aux_max_ce 0.0 --lr_schedule_origin_tokens 85002387456 --dblock_stop_after_commits 0 --no-oom_auto_backoff --require_exact_resume_state --dblock_objective_mode committed_9_1_1 --dblock_ar_weight 0.25 --dblock_sat_weight 0.25 --dblock_nat_weight 0.25 --dblock_aux_grad_ratio 0.5 --dblock_min_ar_share 0.8 --lr_core 2e-7 --lr_head 2e-7 --batch_size 60 --oom_known_safe_batch 56 --dblock_checkpoint_stride 2 --empty_cache_every_steps 0 --dblock_log_every 25 --cuda_max_reserved_mib 0 --cuda_min_free_mib 0 --weight_accel_mode shadow --weight_accel_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/weight_trajectory_accel --weight_accel_every 1 --weight_accel_train_every 4 --weight_accel_record_every 14 --weight_accel_state_every 256 --resume /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt --save_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B --data_seed 2436099",
-      "cwd": "/workspace/agillm43_production_v14/runtime",
-      "env": {
-        "AGILLM43_DECOMPRESS_CACHE": "0",
-        "AGILLM43_SEQUENCE_PREFETCH_DEPTH": "256",
-        "AGILLM43_SINGLEFILE_PROFILE": "intelligence-v25-satvar-pretrain",
-        "AGILLM_DATASETS_STREAMING_READ_MAX_RETRIES": "2",
-        "AGILLM_DATASETS_STREAMING_READ_RETRY_INTERVAL": "1",
-        "AGILLM_DATASET_AGENT_ROUTER": "0",
-        "AGILLM_DATASET_HOTLOAD_CONFIG": "/workspace/agillm43_dataset_hotload.json",
-        "AGILLM_DATASET_HOT_RELOAD_CHECK_TOKENS": "<redacted>",
-        "AGILLM_DATASET_HOT_RELOAD_SEC": "5",
-        "AGILLM_DATASET_NN_ROUTER": "1",
-        "AGILLM_DATASET_ROUTER_BLEND": "0.25",
-        "AGILLM_DATASET_ROUTER_EXPLORE": "0.03",
-        "AGILLM_DATASET_ROUTER_LR": "0.005",
-        "AGILLM_DATASET_ROUTER_MIN_SCORE": "0.15",
-        "AGILLM_DATASET_ROUTER_SHARPNESS": "2.0",
-        "AGILLM_ENABLE_CUPY": "0",
-        "AGILLM_HF_HUB_DOWNLOAD_TIMEOUT": "30",
-        "AGILLM_HF_TRANSPORT": "geth_hf_relay",
-        "AGILLM_HOT_CONFIG": "/workspace/agillm43_production_v14/contracts/hot_config_owner_v25.json",
-        "AGILLM_HOT_CONFIG_SHA256": "",
-        "AGILLM_MAX_EXAMPLE_CHARS": "32768",
-        "AGILLM_MAX_EXAMPLE_TOKENS": "<redacted>",
-        "AGILLM_PARENT_POINTER_SHA256": "f2f30774b7ab22b7393aad444fcd41a66a55fc6c3c9c30b9d494a999543f21f7",
-        "AGILLM_REMOTE_SHARDS_PER_ITER": "4",
-        "AGILLM_RUN_ID": "singlefile_intelligence_v25_20260827T152842Z_from_step000002436057",
-        "AGILLM_STREAM_NEXT_TIMEOUT_SEC": "60",
-        "AGILLM_STREAM_OPEN_TIMEOUT_SEC": "120",
-        "AGILLM_STREAM_SOCKET_TIMEOUT_SEC": "30",
-        "AGILLM_STREAM_SOURCE_FATAL_COOLDOWN_SEC": "600",
-        "AGILLM_STREAM_SOURCE_MAX_COOLDOWN_SEC": "120",
-        "AGILLM_SYNTHETIC_TOKENIZER": "<redacted>",
-        "CUDA_MODULE_LOADING": "LAZY",
-        "CUDA_VERSION": "12.1.1",
-        "CUDA_VISIBLE_DEVICES": "0",
-        "HF_HUB_DOWNLOAD_TIMEOUT": "30",
-        "HF_HUB_ETAG_TIMEOUT": "10",
-        "PYTORCH_CUDA_ALLOC_CONF": "backend:cudaMallocAsync",
-        "PYTORCH_VERSION": "2.2.1"
-      },
-      "parent_proc_cmdline": "/opt/conda/bin/python -u /workspace/agillm43_production_v14/runtime/agillm43_singlefile_intelligence_v25_satvar_pretrain_final_20260826.py supervise --save_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B --log /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/logs/singlefile_intelligence_v25_satvar_pretrain.log --sleep_sec 15 --profile normal",
-      "pid": 2629546,
-      "ppid": 2610096,
-      "proc_cmdline": "/opt/conda/bin/python -u /workspace/agillm43_production_v14/runtime/agillm43_singlefile_intelligence_v25_satvar_pretrain_final_20260826.py train --preset agillm4_floor --block 2048 --source HuggingFaceFW/fineweb-edu:sample-10BT|1.60,HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25 --nat_mask_token_id 2 --tie_weights --tie_kv --moe_ffn --moe_experts 2 --moe_top_k 1 --moe_mlp_mult 4 --moe_shared_experts 1 --moe_shared_mlp_mult 2 --moe_aux_coef 0.01 --moe_z_coef 0.001 --amp --grad_checkpoint --attn_backend sdpa --sublinear_window 128 --sublinear_stride 128 --sublinear_max_anchors 128 --sublinear_chunk 128 --sublinear_sinks 4 --sublinear_recent_anchors 64 --no-sublinear_pooled_landmarks --anchor_stride 256 --anchor_max 2048 --anchor_position -1 --alibi_mode corrected --alibi_scale 0.0 --optimizer adamw8bit --weight_decay 0.0 --lr_decay cosine --lr_decay_tokens 49000954880 --lr_min_mult 0.2 --lr_schedule_reset_on_resume --lr_warmup_tokens 10000000 --lr_warmup_min_mult 0.1 --dblock --dblock_blocks 14 --dblock_schedule roundrobin --dblock_router shadow --dblock_router_blend 0.0 --dblock_router_ramp_steps 64 --dblock_warmup_steps 14 --dblock_sublayer_mode off --dblock_checkpoint_skip_tail 0 --dblock_loop_layers 0 --dblock_loop_start 0 --dblock_loop_cond_scale 1.0 --dblock_sigma_curriculum_steps 2000 --dblock_sigma_sampling lognormal --dblock_sigma_stratified --dblock_sigma_min 0.002 --dblock_sigma_max 2.0 --dblock_sigma_pmean -1.2 --dblock_sigma_pstd 1.2 --dblock_edm_wmax 5.0 --dblock_ar_prob 0.8 --dblock_sat_prob 0.1 --dblock_nat_prob 0.1 --dblock_ar_loss_tokens 8192 --dblock_sat_loss_tokens 4096 --dblock_nat_loss_tokens 4096 --dblock_nat_embed_noise_mode off --dblock_nat_embed_noise_scale 1.0 --nat_loss_weight 1.0 --nat_mask_ratio 0.5 --nat_max_tokens 0 --nat_span_mask_prob 0.35 --nat_suffix_mask_prob 0.20 --nat_span_max_tokens 0 --loss_spike_skip 0.0 --sat_every 1 --nat_every 1 --dblock_fullstack_anchor_spike_skip 0.0 --dblock_fullstack_anchor_max_ce 0.0 --dblock_anchor_softcap_mult 4.0 --dblock_anchor_softcap_min_ce 25.0 --dblock_anchor_softcap_max_ce 50.0 --dblock_fullstack_ar_every 3 --dblock_fullstack_ar_offset 0 --dblock_fullstack_ar_tokens 256 --dblock_fullstack_sat_every 3 --dblock_fullstack_sat_offset 1 --dblock_fullstack_sat_tokens 128 --dblock_fullstack_nat_every 3 --dblock_fullstack_nat_offset 2 --dblock_fullstack_nat_tokens 128 --dblock_fullstack_nat_mask_id -1 --ckpt_codec block-sharded-zstd --async_update_every_steps 0 --dblock_router_hidden 64 --dblock_router_heads 4 --dblock_router_layers 2 --dblock_router_lr 0.0005 --dblock_explore 0.08 --dblock_max_stale_steps 64 --dblock_max_count_skew 1.35 --dblock_stale_bonus 0.35 --dblock_undertrain_bonus 0.25 --dblock_activation_offload_min_mb 1.0 --nat_document_boundary_aware --dblock_fullstack_anchor_deterministic_eval --dblock_fullstack_anchor_spike_retry_limit 3 --dblock_local_spike_retry_limit 3 --dblock_fullstack_nat_no_valid_crop_limit 8 --target_tokens 400000000000 --lr_schedule_reanchor_on_resume --save_every_sec 1800 --delta_every_steps 0 --delta_every_sec 0 --delta_max_keep 0 --max_ckpts 2 --disk_free_floor_gb 12.0 --heartbeat_every_sec 30 --dblock_fullstack_ar_weight 1.0 --dblock_fullstack_sat_weight 0.1 --dblock_fullstack_nat_weight 0.1 --dblock_local_aux_warn_ce 0.0 --dblock_local_aux_max_ce 0.0 --lr_schedule_origin_tokens 85002387456 --dblock_stop_after_commits 0 --no-oom_auto_backoff --require_exact_resume_state --dblock_objective_mode committed_9_1_1 --dblock_ar_weight 0.25 --dblock_sat_weight 0.25 --dblock_nat_weight 0.25 --dblock_aux_grad_ratio 0.5 --dblock_min_ar_share 0.8 --lr_core 2e-7 --lr_head 2e-7 --batch_size 60 --oom_known_safe_batch 56 --dblock_checkpoint_stride 2 --empty_cache_every_steps 0 --dblock_log_every 25 --cuda_max_reserved_mib 0 --cuda_min_free_mib 0 --weight_accel_mode shadow --weight_accel_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/weight_trajectory_accel --weight_accel_every 1 --weight_accel_train_every 4 --weight_accel_record_every 14 --weight_accel_state_every 256 --resume /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt --save_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B --data_seed 2436099",
-      "schema": "agillm.launch.v1"
-    },
-    "local_step": 2456209,
-    "loss": 10.085275650024414,
-    "nat_mask_contract": "agillm.nat-mask.v1",
-    "nat_mask_contract_source": "explicit-legacy-recovery",
-    "nat_mask_eos_id": 1,
-    "nat_mask_legacy_blank_id": 1,
-    "nat_mask_migrated_from_id": 1,
-    "nat_mask_schema_version": 1,
-    "nat_mask_token_id": 2,
-    "pid": 2629546,
-    "schema_version": 1,
-    "seen_tok": 235535175680,
-    "source_is_lane_checkpoint": true,
-    "source_is_master_checkpoint": false,
-    "source_is_non_master_checkpoint": false,
-    "step": 2456209,
-    "train_argv": "/workspace/agillm43_production_v14/runtime/agillm43_singlefile_intelligence_v25_satvar_pretrain_final_20260826.py train --preset agillm4_floor --block 2048 --source HuggingFaceFW/fineweb-edu:sample-10BT|1.60,HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25 --nat_mask_token_id 2 --tie_weights --tie_kv --moe_ffn --moe_experts 2 --moe_top_k 1 --moe_mlp_mult 4 --moe_shared_experts 1 --moe_shared_mlp_mult 2 --moe_aux_coef 0.01 --moe_z_coef 0.001 --amp --grad_checkpoint --attn_backend sdpa --sublinear_window 128 --sublinear_stride 128 --sublinear_max_anchors 128 --sublinear_chunk 128 --sublinear_sinks 4 --sublinear_recent_anchors 64 --no-sublinear_pooled_landmarks --anchor_stride 256 --anchor_max 2048 --anchor_position -1 --alibi_mode corrected --alibi_scale 0.0 --optimizer adamw8bit --weight_decay 0.0 --lr_decay cosine --lr_decay_tokens 49000954880 --lr_min_mult 0.2 --lr_schedule_reset_on_resume --lr_warmup_tokens 10000000 --lr_warmup_min_mult 0.1 --dblock --dblock_blocks 14 --dblock_schedule roundrobin --dblock_router shadow --dblock_router_blend 0.0 --dblock_router_ramp_steps 64 --dblock_warmup_steps 14 --dblock_sublayer_mode off --dblock_checkpoint_skip_tail 0 --dblock_loop_layers 0 --dblock_loop_start 0 --dblock_loop_cond_scale 1.0 --dblock_sigma_curriculum_steps 2000 --dblock_sigma_sampling lognormal --dblock_sigma_stratified --dblock_sigma_min 0.002 --dblock_sigma_max 2.0 --dblock_sigma_pmean -1.2 --dblock_sigma_pstd 1.2 --dblock_edm_wmax 5.0 --dblock_ar_prob 0.8 --dblock_sat_prob 0.1 --dblock_nat_prob 0.1 --dblock_ar_loss_tokens 8192 --dblock_sat_loss_tokens 4096 --dblock_nat_loss_tokens 4096 --dblock_nat_embed_noise_mode off --dblock_nat_embed_noise_scale 1.0 --nat_loss_weight 1.0 --nat_mask_ratio 0.5 --nat_max_tokens 0 --nat_span_mask_prob 0.35 --nat_suffix_mask_prob 0.20 --nat_span_max_tokens 0 --loss_spike_skip 0.0 --sat_every 1 --nat_every 1 --dblock_fullstack_anchor_spike_skip 0.0 --dblock_fullstack_anchor_max_ce 0.0 --dblock_anchor_softcap_mult 4.0 --dblock_anchor_softcap_min_ce 25.0 --dblock_anchor_softcap_max_ce 50.0 --dblock_fullstack_ar_every 3 --dblock_fullstack_ar_offset 0 --dblock_fullstack_ar_tokens 256 --dblock_fullstack_sat_every 3 --dblock_fullstack_sat_offset 1 --dblock_fullstack_sat_tokens 128 --dblock_fullstack_nat_every 3 --dblock_fullstack_nat_offset 2 --dblock_fullstack_nat_tokens 128 --dblock_fullstack_nat_mask_id -1 --ckpt_codec block-sharded-zstd --async_update_every_steps 0 --dblock_router_hidden 64 --dblock_router_heads 4 --dblock_router_layers 2 --dblock_router_lr 0.0005 --dblock_explore 0.08 --dblock_max_stale_steps 64 --dblock_max_count_skew 1.35 --dblock_stale_bonus 0.35 --dblock_undertrain_bonus 0.25 --dblock_activation_offload_min_mb 1.0 --nat_document_boundary_aware --dblock_fullstack_anchor_deterministic_eval --dblock_fullstack_anchor_spike_retry_limit 3 --dblock_local_spike_retry_limit 3 --dblock_fullstack_nat_no_valid_crop_limit 8 --target_tokens 400000000000 --lr_schedule_reanchor_on_resume --save_every_sec 1800 --delta_every_steps 0 --delta_every_sec 0 --delta_max_keep 0 --max_ckpts 2 --disk_free_floor_gb 12.0 --heartbeat_every_sec 30 --dblock_fullstack_ar_weight 1.0 --dblock_fullstack_sat_weight 0.1 --dblock_fullstack_nat_weight 0.1 --dblock_local_aux_warn_ce 0.0 --dblock_local_aux_max_ce 0.0 --lr_schedule_origin_tokens 85002387456 --dblock_stop_after_commits 0 --no-oom_auto_backoff --require_exact_resume_state --dblock_objective_mode committed_9_1_1 --dblock_ar_weight 0.25 --dblock_sat_weight 0.25 --dblock_nat_weight 0.25 --dblock_aux_grad_ratio 0.5 --dblock_min_ar_share 0.8 --lr_core 2e-7 --lr_head 2e-7 --batch_size 60 --oom_known_safe_batch 56 --dblock_checkpoint_stride 2 --empty_cache_every_steps 0 --dblock_log_every 25 --cuda_max_reserved_mib 0 --cuda_min_free_mib 0 --weight_accel_mode shadow --weight_accel_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/weight_trajectory_accel --weight_accel_every 1 --weight_accel_train_every 4 --weight_accel_record_every 14 --weight_accel_state_every 256 --resume /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt --save_dir /workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B --data_seed 2436099",
-    "train_script": "agillm43_singlefile_intelligence_v25_satvar_pretrain_final_20260826.py",
-    "train_script_sha256": "d31faa9e5afe55f7b719109270c5d1e4498029bfa1bffef06b61885e52063d07",
-    "warmstart_base_seen_tok": 0,
-    "warmstart_base_step": 2436057,
-    "warmstart_kind": "warmstarted_from_lane_checkpoint",
-    "warmstart_source_path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt",
-    "warmstart_source_step": 2436057
-  },
-  "alibi_mode": "corrected",
-  "alibi_scale": 0.0,
-  "checkpoint_summary": "warmstarted_from_lane_checkpoint; source=/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt; path=/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt",
-  "dataset_provenance": {
-    "chat": false,
-    "dataset_field_text": "text",
-    "dataset_hotload_config_path": "/workspace/agillm43_dataset_hotload.json",
-    "dataset_hotload_config_sha256": "2762ba666c280207bdb2293b6d382ef8077d167de79609e5bcc35d4499fd1527",
-    "dataset_hotload_config_stamp": [
-      73,
-      2196426857,
-      1787483694198261655,
-      1281
-    ],
-    "dataset_hotload_config_used": true,
-    "dataset_hotload_enabled": true,
-    "dataset_hotload_generation": 4,
-    "dataset_hotload_schema": "agillm.dataset_hotload.config.v2",
-    "dataset_revision_map_sha256": "8361995a504a2365136b20e1a879ba78944a06309f13f3f013bbd0cdfc72225b",
-    "dataset_revisions": {
-      "EleutherAI/proof-pile-2": "901a9273a770e9d4138c5ddd91802f9c5c6cdc4b",
-      "HuggingFaceFW/fineweb": "9bb295ddab0e05d785b879661af7260fed5140fc",
-      "HuggingFaceFW/fineweb-edu": "87f09149ef4734204d70ed1d046ddc9ca3f2b8f9",
-      "HuggingFaceTB/cosmopedia": "0ae6ec63f91742bd2d1eaef4f02232c55d719385",
-      "HuggingFaceTB/finemath": "e92b25a616738fe95dc186b64dfb19f9c8525594",
-      "HuggingFaceTB/smollm-corpus": "3ba9d605774198c5868892d7a8deda78031a781f",
-      "Skylion007/openwebtext": "79d93d786212f7344586290adb811d4ae6a1762c",
-      "allenai/c4": "1588ec454efa1a09f29cd18ddd04fe05fc8653a2",
-      "allenai/dolma": "7f48140530a023e9ea4c5cfb141160922727d4d3",
-      "allenai/dolmino-mix-1124": "a319f19eef1e257417b11ea8c30da266ae175557",
-      "codeparrot/codeparrot-clean": "35a59fb025bc0a102f7d96eac09d145b896d487b",
-      "mlfoundations/dclm-baseline-1.0": "a3b142c183aebe5af344955ae20836eb34dcf69b",
-      "wikimedia/wikipedia": "b04c8d1ceb2f5cd4588862100d08de323dccfbaa"
-    },
-    "dolma_payload_integrity": "external_url_manifest_pinned_payload_hashes_unavailable",
-    "dolma_url_list_sha256": "bc6ef505be920c94bb9e7147445638f606fdc3f5af622cd1d887dc48e21a2d5b",
-    "dolma_url_manifest_raw_sha256": "6fe8c40d51e5bebc9303453d7f533b232d69f7ae122123621a82621b1cd779b7",
-    "has_language_mix": true,
-    "has_numeracy": false,
-    "hot_config_expected_sha256": "",
-    "hot_config_has_append": true,
-    "hot_config_has_datasets": false,
-    "hot_config_mtime": 1787722590.529335,
-    "hot_config_path": "/workspace/agillm43_production_v14/contracts/hot_config_owner_v25.json",
-    "hot_config_used": true,
-    "hot_reload_capable": true,
-    "phase": "pretrain",
-    "proof_pile_lfs_manifest_sha256": "d020dfb796b02a75564e4e11d9e03185a55b3f9c6639f87830cfbe10737ed329",
-    "proof_pile_paths_sha256": "8a30018017e3b3d9a9bcfa06a3c404ccd113297b0b1d0c48e2e2aadf113e52b0",
-    "schema": "agillm.dataset_provenance.v1",
-    "sequence_prefetch_depth": 256,
-    "source_count": 15,
-    "source_effective": "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25,HuggingFaceTB/finemath:finemath-4plus|0.7,HuggingFaceTB/finemath:infiwebmath-4plus|0.35,HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8,allenai/dolmino-mix-1124|1,mlfoundations/dclm-baseline-1.0|1,HuggingFaceFW/fineweb-edu:sample-100BT|1.6",
-    "source_requested": "HuggingFaceFW/fineweb-edu:sample-10BT|1.60,HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25",
-    "source_specs": [
-      "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50",
-      "wikimedia/wikipedia:20231101.en|1.10",
-      "allenai/c4:en|0.40",
-      "Skylion007/openwebtext|0.35",
-      "HuggingFaceTB/cosmopedia:web_samples_v2|1.30",
-      "HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30",
-      "EleutherAI/proof-pile-2:all|1.50",
-      "allenai/dolma:v1_6-sample|0.50",
-      "codeparrot/codeparrot-clean|1.25",
-      "HuggingFaceTB/finemath:finemath-4plus|0.7",
-      "HuggingFaceTB/finemath:infiwebmath-4plus|0.35",
-      "HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8",
-      "allenai/dolmino-mix-1124|1",
-      "mlfoundations/dclm-baseline-1.0|1",
-      "HuggingFaceFW/fineweb-edu:sample-100BT|1.6"
-    ],
-    "val_source_effective": "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25,HuggingFaceTB/finemath:finemath-4plus|0.7,HuggingFaceTB/finemath:infiwebmath-4plus|0.35,HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8,allenai/dolmino-mix-1124|1,mlfoundations/dclm-baseline-1.0|1,HuggingFaceFW/fineweb-edu:sample-100BT|1.6",
-    "val_source_requested": ""
-  },
-  "lr_schedule_origin_tok": 85002387456,
-  "nat_mask_contract": "agillm.nat-mask.v1",
-  "nat_mask_contract_source": "explicit-legacy-recovery",
-  "nat_mask_eos_id": 1,
-  "nat_mask_legacy_blank_id": 1,
-  "nat_mask_migrated_from_id": 1,
-  "nat_mask_schema_version": 1,
-  "nat_mask_token_id": 2,
-  "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt",
-  "promotion_required": true,
-  "quality_gate_used_for_training": false,
-  "repair_schema": "",
-  "saved_at": "2026-08-28T04:14:54Z",
-  "schema": "agillm43.serving.candidate.v1",
-  "source_effective": "HuggingFaceFW/fineweb:CC-MAIN-2024-10|0.50,wikimedia/wikipedia:20231101.en|1.10,allenai/c4:en|0.40,Skylion007/openwebtext|0.35,HuggingFaceTB/cosmopedia:web_samples_v2|1.30,HuggingFaceTB/smollm-corpus:cosmopedia-v2|1.30,EleutherAI/proof-pile-2:all|1.50,allenai/dolma:v1_6-sample|0.50,codeparrot/codeparrot-clean|1.25,HuggingFaceTB/finemath:finemath-4plus|0.7,HuggingFaceTB/finemath:infiwebmath-4plus|0.35,HuggingFaceTB/smollm-corpus:fineweb-edu-dedup|0.8,allenai/dolmino-mix-1124|1,mlfoundations/dclm-baseline-1.0|1,HuggingFaceFW/fineweb-edu:sample-100BT|1.6",
-  "step": 2456209,
-  "training_pointer_schema": "agillm43.training.resume.v2",
-  "training_pointer_sha256": "5c11781a93267577ef9a36a5713c51f029dde2f08d1d41761850712842e97b5b",
-  "warmstart_kind": "warmstarted_from_lane_checkpoint",
-  "warmstart_source_path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02436057_from02435995_20260827T1524Z.pt"
-}

checkpoints/step2456209_20260828/training_latest.json

@@ -1,228 +0,0 @@
-{
-  "checkpoint_main_nbytes": 17669618,
-  "checkpoint_main_sha256": "4e841d7ef2d0c5e6a2d1ae015ee2bedb0015ae1f43095fd6c17e2c4738fc90f4",
-  "checkpoint_name": "pretrain_step02456209_from02436057_20260828T0414Z.pt",
-  "checkpoint_path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt",
-  "continuation_state_schema": "agillm43.continuation.state.v2",
-  "continuation_state_sha256": "bde4bec7afc5d1d62d60bd3eb581fdfdb0fee457d40f79b744d7f884de3c0a7b",
-  "dblock_committed_step": 725858,
-  "package_format": "block-sharded",
-  "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt",
-  "quality_gate_used_for_training": false,
-  "record_sha256": "5c11781a93267577ef9a36a5713c51f029dde2f08d1d41761850712842e97b5b",
-  "role": "training",
-  "saved_at": "2026-08-28T04:14:54Z",
-  "schema": "agillm43.training.resume.v2",
-  "seen_tok": 235535175680,
-  "shard_dir": "pretrain_step02456209_from02436057_20260828T0414Z.pt.shards",
-  "shards": [
-    {
-      "name": "core_base.pt",
-      "nbytes": 608342946,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_base.pt",
-      "sha256": "b1c79c712a03d314eccbfdfa9c8965a204698ed36fb03d262c45817bd8d86b27"
-    },
-    {
-      "name": "core_block_0000.pt",
-      "nbytes": 117688535,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0000.pt",
-      "sha256": "154b30572599f7b03e85e34a3b4d27f755e2c8771f0f16c95daa5ac04b91e7c9"
-    },
-    {
-      "name": "core_block_0001.pt",
-      "nbytes": 116952900,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0001.pt",
-      "sha256": "cd6681ec55fe3f49dd45f82c3e2a7ea00e6032939cccfaca95fa2703edd434c7"
-    },
-    {
-      "name": "core_block_0002.pt",
-      "nbytes": 116948234,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0002.pt",
-      "sha256": "04dfeb7ad6483876ba0707645735de286a6fdb95381d6ebec71cf16bc4879453"
-    },
-    {
-      "name": "core_block_0003.pt",
-      "nbytes": 116923196,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0003.pt",
-      "sha256": "1ffb29cefc532d4e2468bf03ee6da7f7af7bb8136c48074a190972db9b518d2d"
-    },
-    {
-      "name": "core_block_0004.pt",
-      "nbytes": 117020275,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0004.pt",
-      "sha256": "d40ec06fb130d272458edf35204eb5a6ab8f631c65dc855ae2ea1896ed7dec58"
-    },
-    {
-      "name": "core_block_0005.pt",
-      "nbytes": 116980150,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0005.pt",
-      "sha256": "318ae307dd24b526062857d634bc0514242fbc5363dde62792b1cfca5aaf7b8e"
-    },
-    {
-      "name": "core_block_0006.pt",
-      "nbytes": 117360969,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0006.pt",
-      "sha256": "53654fb97aca55d7533a3a610ca8328821d021c1f6a6b16d20e04bff609a7c19"
-    },
-    {
-      "name": "core_block_0007.pt",
-      "nbytes": 117602004,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0007.pt",
-      "sha256": "81d04a1e9ab1ceb42b07b3967b152f174a39dd7d3c5c28670db38bc1564f7f7d"
-    },
-    {
-      "name": "core_block_0008.pt",
-      "nbytes": 117452281,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0008.pt",
-      "sha256": "02917716cf79f1e63c308f2baeca7b87e79197ec4c184cedba9f6aa4f3eb57ac"
-    },
-    {
-      "name": "core_block_0009.pt",
-      "nbytes": 117285114,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0009.pt",
-      "sha256": "0e5562662aeb1a21d2ecb9709c4dd61a724b226de00f369ff6b31c34d7091b3d"
-    },
-    {
-      "name": "core_block_0010.pt",
-      "nbytes": 117414116,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0010.pt",
-      "sha256": "25a08d491583d1818cca6363af11f3504f73f35553e4b1ae12f702ca6e13e6de"
-    },
-    {
-      "name": "core_block_0011.pt",
-      "nbytes": 117411087,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0011.pt",
-      "sha256": "055b5a40e2b5cc593af84299b44998e4b4a4bacc3ff9a3b90c3b3415aeed66de"
-    },
-    {
-      "name": "core_block_0012.pt",
-      "nbytes": 123854669,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0012.pt",
-      "sha256": "54855f7be882be524d9640b52cfb7d208023392f4a3f1339be5fe27bd6d1290f"
-    },
-    {
-      "name": "core_block_0013.pt",
-      "nbytes": 120975183,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0013.pt",
-      "sha256": "d7bace8cceb145ef0074fd793238020549f974f92ffa169cf0edebdcb689de19"
-    },
-    {
-      "name": "core_block_0014.pt",
-      "nbytes": 117793475,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0014.pt",
-      "sha256": "2962bffe129b4ac9ef82cbe7506674818398637c4f1b6ba475f5a023bb067cd3"
-    },
-    {
-      "name": "core_block_0015.pt",
-      "nbytes": 116989512,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0015.pt",
-      "sha256": "2392bffa37a7149136409b4115f9454cc0bc85ee18bce5fe349625841bb18f62"
-    },
-    {
-      "name": "core_block_0016.pt",
-      "nbytes": 117232767,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0016.pt",
-      "sha256": "d39bdb00165ec54651ab1b870423714e2db18d2cad4e9050dadf43ba1b6e7ef4"
-    },
-    {
-      "name": "core_block_0017.pt",
-      "nbytes": 116948233,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0017.pt",
-      "sha256": "b4e5f854b93cf21a5c463c2d4ddb055e555dcd2dc497e8bcfb73ce1714cde7af"
-    },
-    {
-      "name": "core_block_0018.pt",
-      "nbytes": 117107701,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0018.pt",
-      "sha256": "9dbc36dd65532b845bbb819ad5de426a60ec7fee8e981efd071a69aeaed6db03"
-    },
-    {
-      "name": "core_block_0019.pt",
-      "nbytes": 117002475,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0019.pt",
-      "sha256": "e8aa8ea266ed218b394eaa7432890ef8c88af4bcc4c749c615100aa24be17f57"
-    },
-    {
-      "name": "core_block_0020.pt",
-      "nbytes": 117026596,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0020.pt",
-      "sha256": "daee0d1fbcee62c5d1ff28aef279e5dc5e4a08758e41b4a54cd86f54cfc866ba"
-    },
-    {
-      "name": "core_block_0021.pt",
-      "nbytes": 116873728,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0021.pt",
-      "sha256": "4bd0858ce2c7fb25031bcd6d2012690a6723daa38efad9543de4046ba6e52c50"
-    },
-    {
-      "name": "core_block_0022.pt",
-      "nbytes": 117047033,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0022.pt",
-      "sha256": "21f56b75ef33c33b8e2c3de4ca28bf5782bff1be4a00a52d6cb84ca1c4ab7d1b"
-    },
-    {
-      "name": "core_block_0023.pt",
-      "nbytes": 117315323,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0023.pt",
-      "sha256": "2dc3d6575da8236e590d8688ff9900a7cf26b4d9867b800aa636a059cce874ed"
-    },
-    {
-      "name": "core_block_0024.pt",
-      "nbytes": 121466755,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0024.pt",
-      "sha256": "6c9c669a8246c6b35de2641865babc5cb493a4868e714da4c3e30b9b7cb668f2"
-    },
-    {
-      "name": "core_block_0025.pt",
-      "nbytes": 121496154,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0025.pt",
-      "sha256": "24fb9aa8ad2a994268fe3985b0055ff682bdb21e1d5f0e0fead587c7c476d00d"
-    },
-    {
-      "name": "core_block_0026.pt",
-      "nbytes": 121424479,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0026.pt",
-      "sha256": "1711fa0e274c9291d2ed1c66431d0061f08de5ff55ee1174ecf319d9e92261bc"
-    },
-    {
-      "name": "core_block_0027.pt",
-      "nbytes": 121617836,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/core_block_0027.pt",
-      "sha256": "3eb33cdb8599d01c150811b86a2920bd18fac5aab20004a54dd99392e9995d8f"
-    },
-    {
-      "name": "ar.pt",
-      "nbytes": 608333815,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/ar.pt",
-      "sha256": "0a47cc5c88863f4d428f07a013cdb1077997a6334c653483f2cfd4d8a264e02c"
-    },
-    {
-      "name": "sat.pt",
-      "nbytes": 608343430,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/sat.pt",
-      "sha256": "035b0f7d9c9a9345e78c9824392230e939b0835baed36c853e8030213703eac1"
-    },
-    {
-      "name": "nat.pt",
-      "nbytes": 608333815,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/nat.pt",
-      "sha256": "cf569d40c6aec7f5fce6034a28c27afb4d792d15f39038e0fa6d708982b0e666"
-    },
-    {
-      "name": "opt.pt",
-      "nbytes": 1569900662,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/opt.pt",
-      "sha256": "a17ba07575a39189157df24e8522963a5f748a64091d08c14f9a07424e68d214"
-    },
-    {
-      "name": "scaler.pt",
-      "nbytes": 149,
-      "path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.shards/scaler.pt",
-      "sha256": "c2110dfd7b7a6c034fb62244a6f8f76a22d9936d55b6d0fe133e43e65b38631c"
-    }
-  ],
-  "step": 2456209,
-  "tokenizer_sidecar_nbytes": 17465482,
-  "tokenizer_sidecar_path": "/workspace/agillm43_production_v14/production_quality_v18_hard_spike_gates/checkpoints_continue400B_from200B/pretrain_step02456209_from02436057_20260828T0414Z.pt.tokenizer.json",
-  "tokenizer_sidecar_sha256": "8e21b2b5a9f2723ad4668e746d37c512a91aa2e07590d2dca029adaf4f4ba227"
-}