Archive original ant pi05 profile bundle

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/README.md

@@ -0,0 +1,31 @@
+# ant / qwenpi_v3
+
+Training condition: `latency-aware`. Run: `ant_pi05_profile_h1_5fps_g128_20260920`.
+
+Copied from [Standard-Pipeline](https://huggingface.co/latency-sensitive-bench/Standard-Pipeline/tree/2208875f92b21e3f8ec2363a30a6705dfe56cc04/ant/profile_latency/Pi05) at `2208875f92b21e3f8ec2363a30a6705dfe56cc04`.
+
+- Checkpoint: `checkpoints/model.pt`
+- Selection: only published VLA checkpoint
+- Checkpoint SHA256: `01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f`
+- Configuration, FPS, actions, normalization and latency semantics retain their source values.
+- This import does not run evaluations or change training/evaluation/acceptance status.
+
+Action horizon is 1. Use `config.yaml` and `dataset_statistics.json` with the matching
+prepared backbone. `task_contract.json` is extracted from the original training configuration.
+Original machine paths in historical configurations are provenance: bind checkpoint,
+backbone, manifest and profile paths to downloaded local files when running inference.
+
+See [the original model notes](source/README.md) for experiment results and limitations.
+
+[Original provenance](source/provenance.json) retains dataset references and recorded experiment evidence.
+
+## Download
+
+```python
+from huggingface_hub import snapshot_download
+prefix = "latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920"
+snapshot_download("latency-sensitive-bench/benchmark-models", revision="<commit>",
+                  allow_patterns=[prefix + "/**"])
+```
+
+Use the verified import commit listed in the repository's import report.

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/checkpoints/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f
+size 10922682525

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/config.full.yaml

@@ -0,0 +1,280 @@
+framework:
+  name: QwenPI_v3
+  qwenvl:
+    base_vlm: ${oc.env:PI05_BACKBONE_DIR}
+    attn_implementation: flash_attention_2
+    vl_hidden_dim: 2560
+    num_vl_layers: 36
+    enable_gradient_checkpointing: false
+  action_model:
+    action_model_type: LayerwiseFM
+    action_dim: 8
+    state_dim: 27
+    state_encoding: continuous_projector
+    task_objective: null
+    action_horizon: 1
+    repeated_diffusion_steps: 2
+    num_inference_timesteps: 4
+    add_pos_embed: true
+    max_seq_len: 1024
+    num_target_vision_tokens: 32
+    noise_beta_alpha: 1.5
+    noise_beta_beta: 1.0
+    noise_s: 0.999
+    num_timestep_buckets: 1000
+    diffusion_model_cfg:
+      action_dit_hidden_dim: 1024
+      dropout: 0.2
+      final_dropout: true
+      interleave_self_attention: true
+      norm_type: ada_norm
+      positional_embeddings: null
+      attention_head_dim: 64
+      num_layers: 36
+      input_embedding_dim: 1024
+      cross_attention_dim: 1024
+      output_dim: 1024
+      num_attention_heads: 16
+    action_env_dim: 8
+    future_action_window_size: 0
+    past_action_window_size: 0
+datasets:
+  vla_data:
+    dataset_py: lerobot_datasets
+    include_state: true
+    data_root_dir: /
+    data_mix: train
+    eval_data_mix: validation
+    custom_mixtures_path: ${PI05_RUN_DIR}/profile_latency/vla/mixture.json
+    action_type: continuous
+    sequential_step_sampling: true
+    eval_sequential_step_sampling: null
+    num_workers: 4
+    eval_num_workers: 0
+    prefetch_factor: 2
+    persistent_workers: true
+    pin_memory: true
+    shuffle: true
+    action_balance:
+      enabled: false
+      strategy: balanced_epoch
+      action_key: action_id
+      target_flap_fraction: 0.3
+      noop_id: 0
+      flap_id: 1
+    latency_curriculum:
+      enabled: false
+      strategy: exclusive
+      latencies: null
+      phase_steps: null
+      phase_distributions: null
+      new_latency_passes: 1.0
+      replay_passes: 0.25
+      target_total_passes: 2.0
+      final_equalization: true
+      step_budget_mode: auto
+      eval_at_phase_end: false
+      save_at_phase_end: false
+      computed_plan: null
+    per_device_batch_size: 32
+    load_all_data_for_training: true
+    num_obs_frames: 1
+    image_mode: single
+    prompt_mode: raw
+    stitch_grid:
+    - 2
+    - 2
+    obs_image_size:
+    - 224
+    - 224
+    video_backend: pyav
+    lerobot_version: v2.0
+    gymnasium_task_contract:
+      action_space:
+        dtype: float32
+        high:
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        labels:
+        - back_right_hip_torque
+        - back_right_ankle_torque
+        - front_left_hip_torque
+        - front_left_ankle_torque
+        - front_right_hip_torque
+        - front_right_ankle_torque
+        - back_left_hip_torque
+        - back_left_ankle_torque
+        low:
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        type: box
+      base_prompt: Make the Ant move forward as fast as possible without falling.
+        Predict eight continuous torques in [-1, 1] ordered as back right hip, back
+        right ankle, front left hip, front left ankle, front right hip, front right
+        ankle, back left hip, and back left ankle.
+      env_fps: 5.0
+      env_id: LatencyBench/AntContinuous-v0
+      frame_stack: 1
+      make_kwargs:
+        base_env_id: Ant-v4
+        base_make_kwargs:
+          exclude_current_positions_from_observation: true
+          use_contact_forces: false
+        render_mode: rgb_array
+      noop_action:
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      obs_fps: 5.0
+      registration_imports:
+      - latency_bench.envs.gymnasium_ant
+      state_space:
+        labels:
+        - torso_z
+        - torso_quaternion_w
+        - torso_quaternion_x
+        - torso_quaternion_y
+        - torso_quaternion_z
+        - front_left_hip_angle
+        - front_left_ankle_angle
+        - front_right_hip_angle
+        - front_right_ankle_angle
+        - back_left_hip_angle
+        - back_left_ankle_angle
+        - back_right_hip_angle
+        - back_right_ankle_angle
+        - torso_x_velocity
+        - torso_y_velocity
+        - torso_z_velocity
+        - torso_angular_velocity_x
+        - torso_angular_velocity_y
+        - torso_angular_velocity_z
+        - front_left_hip_angular_velocity
+        - front_left_ankle_angular_velocity
+        - front_right_hip_angular_velocity
+        - front_right_ankle_angular_velocity
+        - back_left_hip_angular_velocity
+        - back_left_ankle_angular_velocity
+        - back_right_hip_angular_velocity
+        - back_right_ankle_angular_velocity
+      task_name: ant
+    active_action_dim: 8
+dataset:
+  source_hf: ''
+  config_name: null
+  source_subdir: null
+  converted_name: flappy_train
+  single_source_hf: ''
+  mixed_source_hf: ''
+  single_converted_name: flappy_train
+  mixed_converted_name: flappy_mixed_latency_train
+  single_latency_filter: null
+  mixed_latency_filter: null
+  force_download: false
+  setup_force: false
+  skip_verification: false
+  target_latency_unit: observation_steps
+  verify_rows: 200
+  max_episodes: null
+  episodes_per_latency: null
+  latency_filter: null
+  debug_subset:
+    enabled: false
+    max_episodes: 5
+    suffix: debug
+base_model:
+  repo_id: Qwen/Qwen3-VL-4B-Instruct
+initialization:
+  checkpoint_local_dir: null
+  checkpoint_hf_repo_id: null
+  checkpoint_filename: null
+trainer:
+  max_train_steps: 5000
+  num_warmup_steps: 100
+  save_interval: 500
+  eval_interval: 500
+  eval_num_batches: 200
+  per_latency_eval_num_batches: null
+  eval_action_classification: false
+  eval_action_classification_interval: null
+  cc_f1_tolerance: 1
+  learning_rate:
+    base: 2.0e-05
+    qwen_vl_interface: 1.0e-05
+    action_model: 0.0001
+  lr_scheduler_type: cosine_with_min_lr
+  scheduler_specific_kwargs:
+    min_lr: 1.0e-06
+  freeze_modules: ''
+  freeze_vit: false
+  freeze_tied_embedding: false
+  freeze_llm_layers: []
+  loss_scale:
+    vla: 1.0
+    vlm: 0.1
+  max_grad_norm: 1.0
+  weight_decay: 0.0
+  logging_frequency: 1
+  profile_timing:
+    enabled: false
+    log_interval: 10
+  gradient_clipping: 1.0
+  gradient_accumulation_steps: 2
+  distributed_backend: deepspeed
+  is_resume: false
+  pretrained_checkpoint: null
+  resume_step: 0
+  reload_modules: null
+  optimizer:
+    name: AdamW
+    betas:
+    - 0.9
+    - 0.95
+    eps: 1.0e-08
+    weight_decay: 1.0e-08
+    fused: true
+  save_format: pt
+run_id: ant_pi05_profile_h1_5fps_g128_20260920
+run_root_dir: ${PI05_RUN_DIR}/profile_latency/vla/training
+seed: 42
+is_debug: false
+version_id: '0.21'
+wandb_project: latency-sensitive-bench
+wandb_entity: dongqianyu99-zhejiang-university
+wandb_name: ant_pi05_profile_h1_5fps_g128_20260920
+wandb_group: pi05-seven-env
+wandb_tags:
+- ant
+- profile_latency
+- Pi05
+- h1
+training_latency_condition: profile_latency
+backbone:
+  repo_id: Qwen/Qwen3-VL-4B-Instruct
+  revision: ebb281ec70b05090aa6165b016eac8ec08e71b17
+checkpoint:
+  save_best_model: false
+  save_final_model: true
+  save_pt_file: true
+  save_training_state: true
+  local:
+    keep_last_n: 1
+config_yaml: ${PI05_RUN_DIR}/profile_latency/train.yaml
+output_dir: ${PI05_RUN_DIR}/profile_latency/vla/training/ant_pi05_profile_h1_5fps_g128_20260920

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/config.yaml

@@ -0,0 +1,157 @@
+framework:
+  name: QwenPI_v3
+  qwenvl:
+    base_vlm: ${oc.env:PI05_BACKBONE_DIR}
+    attn_implementation: flash_attention_2
+    vl_hidden_dim: 2560
+    num_vl_layers: 36
+    enable_gradient_checkpointing: false
+  action_model:
+    action_model_type: LayerwiseFM
+    action_dim: 8
+    state_dim: 27
+    state_encoding: continuous_projector
+    task_objective: null
+    action_horizon: 1
+    repeated_diffusion_steps: 2
+    num_inference_timesteps: 4
+    add_pos_embed: true
+    max_seq_len: 1024
+    num_target_vision_tokens: 32
+    noise_beta_alpha: 1.5
+    noise_beta_beta: 1.0
+    noise_s: 0.999
+    num_timestep_buckets: 1000
+    diffusion_model_cfg:
+      action_dit_hidden_dim: 1024
+      dropout: 0.2
+      final_dropout: true
+      interleave_self_attention: true
+      norm_type: ada_norm
+      positional_embeddings: null
+      attention_head_dim: 64
+      num_layers: 36
+      input_embedding_dim: 1024
+      cross_attention_dim: 1024
+      output_dim: 1024
+      num_attention_heads: 16
+    action_env_dim: 8
+    future_action_window_size: 0
+    past_action_window_size: 0
+version_id: '0.21'
+datasets:
+  vla_data:
+    include_state: true
+    obs_image_size:
+    - 224
+    - 224
+    image_mode: single
+    stitch_grid:
+    - 2
+    - 2
+    num_obs_frames: 1
+    gymnasium_task_contract: &id001
+      action_space:
+        dtype: float32
+        high:
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        - 1.0
+        labels:
+        - back_right_hip_torque
+        - back_right_ankle_torque
+        - front_left_hip_torque
+        - front_left_ankle_torque
+        - front_right_hip_torque
+        - front_right_ankle_torque
+        - back_left_hip_torque
+        - back_left_ankle_torque
+        low:
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        - -1.0
+        type: box
+      base_prompt: Make the Ant move forward as fast as possible without falling.
+        Predict eight continuous torques in [-1, 1] ordered as back right hip, back
+        right ankle, front left hip, front left ankle, front right hip, front right
+        ankle, back left hip, and back left ankle.
+      env_fps: 5.0
+      env_id: LatencyBench/AntContinuous-v0
+      frame_stack: 1
+      make_kwargs:
+        base_env_id: Ant-v4
+        base_make_kwargs:
+          exclude_current_positions_from_observation: true
+          use_contact_forces: false
+        render_mode: rgb_array
+      noop_action:
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      - 0.0
+      obs_fps: 5.0
+      registration_imports:
+      - latency_bench.envs.gymnasium_ant
+      state_space:
+        labels:
+        - torso_z
+        - torso_quaternion_w
+        - torso_quaternion_x
+        - torso_quaternion_y
+        - torso_quaternion_z
+        - front_left_hip_angle
+        - front_left_ankle_angle
+        - front_right_hip_angle
+        - front_right_ankle_angle
+        - back_left_hip_angle
+        - back_left_ankle_angle
+        - back_right_hip_angle
+        - back_right_ankle_angle
+        - torso_x_velocity
+        - torso_y_velocity
+        - torso_z_velocity
+        - torso_angular_velocity_x
+        - torso_angular_velocity_y
+        - torso_angular_velocity_z
+        - front_left_hip_angular_velocity
+        - front_left_ankle_angular_velocity
+        - front_right_hip_angular_velocity
+        - front_right_ankle_angular_velocity
+        - back_left_hip_angular_velocity
+        - back_left_ankle_angular_velocity
+        - back_right_hip_angular_velocity
+        - back_right_ankle_angular_velocity
+      task_name: ant
+trainer:
+  pretrained_checkpoint: null
+  profile_timing:
+    enabled: false
+    log_interval: 10
+backbone:
+  repo_id: Qwen/Qwen3-VL-4B-Instruct
+  revision: ebb281ec70b05090aa6165b016eac8ec08e71b17
+rl_games:
+  model_alias: pi05
+  task: gymnasium
+  action_carrier: native
+  gymnasium:
+    task_contract: *id001
+  env_eval:
+    image_size: 224
+    frameskip: 1
+    image_transform: raw_rgb
+    prompt_mode: raw

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/dataset_statistics.json

@@ -0,0 +1,254 @@
+{
+  "new_embodiment": {
+    "action": {
+      "mean": [
+        -0.21258775889873505,
+        0.0060735042206943035,
+        0.09963339567184448,
+        -0.14281702041625977,
+        -0.18596597015857697,
+        0.16119267046451569,
+        -0.1288464218378067,
+        -0.007791953627020121
+      ],
+      "std": [
+        0.09634324163198471,
+        0.31038498878479004,
+        0.41215693950653076,
+        0.28469160199165344,
+        0.09189476817846298,
+        0.050707682967185974,
+        0.3488219678401947,
+        0.3428356945514679
+      ],
+      "max": [
+        0.32485297322273254,
+        0.8219043016433716,
+        0.9785619378089905,
+        0.4289686977863312,
+        0.3769160807132721,
+        0.3570922017097473,
+        0.6995806694030762,
+        0.9399116635322571
+      ],
+      "min": [
+        -0.5221534371376038,
+        -0.6670061945915222,
+        -0.7619105577468872,
+        -0.6787745952606201,
+        -0.9190491437911987,
+        -0.1269017904996872,
+        -0.9429033994674683,
+        -0.6738496422767639
+      ],
+      "q01": [
+        -0.4226320269703865,
+        -0.5816266393661499,
+        -0.5641161489486695,
+        -0.6360065984725952,
+        -0.4370796835422516,
+        0.032235059924423694,
+        -0.8411082011461258,
+        -0.42670102924108505
+      ],
+      "q99": [
+        0.018601863235234804,
+        0.5683065080642696,
+        0.8988946610689162,
+        0.31603450089693047,
+        0.041702924370765025,
+        0.25051053106784804,
+        0.5144581693410865,
+        0.8364450520277023
+      ],
+      "mask": [
+        true,
+        true,
+        true,
+        true,
+        true,
+        true,
+        true,
+        true
+      ]
+    },
+    "state": {
+      "mean": [
+        -0.22882035374641418,
+        0.47941550612449646,
+        -0.1400502622127533,
+        -0.03265565261244774,
+        0.20965337753295898,
+        -0.16846208274364471,
+        0.1846654862165451,
+        -0.8101072311401367,
+        0.09434358030557632,
+        0.04786616191267967,
+        -0.4570174515247345,
+        -0.8917045593261719,
+        0.4862286448478699,
+        0.35206252336502075,
+        0.015424875542521477,
+        -0.038056906312704086,
+        0.06804008036851883,
+        -0.01555140595883131,
+        -0.11194365471601486,
+        -0.1391102522611618,
+        0.00826873816549778,
+        0.2900799512863159,
+        0.3209729492664337,
+        0.09138798713684082,
+        0.02823597937822342,
+        0.08164655417203903,
+        -0.06988196820020676
+      ],
+      "std": [
+        0.36191973090171814,
+        0.12071270495653152,
+        0.21179752051830292,
+        0.2269587516784668,
+        0.14413319528102875,
+        0.6609764099121094,
+        0.33706650137901306,
+        0.0639922022819519,
+        0.07227976620197296,
+        0.6279513239860535,
+        0.40478047728538513,
+        0.07100282609462738,
+        0.4061044752597809,
+        0.2651371955871582,
+        0.19818426668643951,
+        0.3602597117424011,
+        0.20922724902629852,
+        0.272988498210907,
+        0.1799745410680771,
+        0.42945030331611633,
+        0.3028114140033722,
+        0.04903984069824219,
+        0.0550890788435936,
+        0.39300021529197693,
+        0.3226502239704132,
+        0.042170703411102295,
+        0.3688967823982239
+      ],
+      "max": [
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0
+      ],
+      "min": [
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0
+      ],
+      "q01": [
+        -0.8241139680147171,
+        0.04648104071617127,
+        -0.6652810025215149,
+        -0.5314224624633789,
+        -0.19820078015327453,
+        -0.9496272021532058,
+        -0.20859015226364136,
+        -0.8253665775060653,
+        0.09396191000938416,
+        -0.9315658015012741,
+        -0.9077210491895675,
+        -0.9014575052261352,
+        -0.22581478476524353,
+        -0.5506806242465973,
+        -0.42719365894794464,
+        -0.7398031288385392,
+        -0.36758168041706085,
+        -0.5742466670274734,
+        -0.6432376319169998,
+        -0.8104029732942581,
+        -0.7943315953016281,
+        0.2125857985019684,
+        0.3195498716831207,
+        -0.817195657491684,
+        -0.4413272815942764,
+        0.05512574553489685,
+        -0.8615017014741898
+      ],
+      "q99": [
+        0.6039263105392445,
+        0.6871944534778593,
+        0.3730386269092557,
+        0.42124163150787025,
+        0.6027398216724373,
+        0.9456771314144133,
+        0.9311108589172361,
+        -0.6299777722358707,
+        0.09945510149002075,
+        0.9258959579467773,
+        0.24740778565406782,
+        -0.8100068545341509,
+        0.9577416539192198,
+        0.7944118678569789,
+        0.5177989375591278,
+        0.7203878808021538,
+        0.6027004325389856,
+        0.6471473491191858,
+        0.2969542503356919,
+        0.8709750354290008,
+        0.4969204437732693,
+        0.3660900735855095,
+        0.32227142453193663,
+        0.7621688938140869,
+        0.8692629277706144,
+        0.12456895589828483,
+        0.6941961252689341
+      ]
+    },
+    "num_transitions": 89808,
+    "num_trajectories": 90
+  }
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/evaluation/comparison.json

@@ -0,0 +1,419 @@
+{
+  "state": "MATCHED_B_F20_AUDIT_PASSED",
+  "verified_at_utc": "2026-09-22T03:35:27.392265+00:00",
+  "B": {
+    "state": "B20_SPLIT_SESSION_DERIVED_AUDIT_PASSED",
+    "episodes": 20,
+    "seeds": [
+      42,
+      43,
+      44,
+      45,
+      46,
+      47,
+      48,
+      49,
+      50,
+      51,
+      52,
+      53,
+      54,
+      55,
+      56,
+      57,
+      58,
+      59,
+      60,
+      61
+    ],
+    "returns": [
+      32.501595490013216,
+      56.46892726482151,
+      96.83510560036257,
+      95.12004972167394,
+      68.4796112625599,
+      9.427925579208752,
+      180.91789918774197,
+      126.02311329849451,
+      239.58810433012508,
+      3.683059784739528,
+      291.6266216954973,
+      297.4161642880803,
+      50.97806575612095,
+      48.653726214514066,
+      243.76432224961033,
+      78.96932272634479,
+      252.69957930958387,
+      142.4877330153016,
+      86.73317256303835,
+      229.67420439888363
+    ],
+    "mean_return": 131.6024151868358,
+    "population_std_return": 93.71548520884977,
+    "raw_steps": 1444,
+    "actions": 1424,
+    "invalid_actions": 0,
+    "dropped_actions": 0,
+    "dropped_observations": 0,
+    "model_revision": "95dedf8583692024daa890fe3499b5540167eb77",
+    "checkpoint_sha256": "ed340a88cfef3522405265e8646b97e26d61106d011f0911d78f8d276a1e8843"
+  },
+  "F": {
+    "state": "F20_FULL_STEP_ACTION_AUDIT_PASSED",
+    "task": "ant",
+    "verified_at_utc": "2026-09-22T03:35:27.392265+00:00",
+    "episodes": 20,
+    "seeds": [
+      42,
+      43,
+      44,
+      45,
+      46,
+      47,
+      48,
+      49,
+      50,
+      51,
+      52,
+      53,
+      54,
+      55,
+      56,
+      57,
+      58,
+      59,
+      60,
+      61
+    ],
+    "returns": [
+      2215.21984877659,
+      764.2694680291046,
+      2290.599554986267,
+      1840.29021669272,
+      2231.3439323113707,
+      2373.9321151344743,
+      2372.1426528938036,
+      166.30488169019765,
+      2302.3785949510366,
+      2367.2167682154372,
+      2310.4893684982335,
+      246.45578812725222,
+      677.2055059936616,
+      2090.486226750155,
+      2344.0109866336898,
+      1810.7094305991875,
+      2392.8136264328004,
+      2337.4267695838707,
+      104.35524685972497,
+      2256.5825535088743
+    ],
+    "lengths": [
+      924,
+      382,
+      1000,
+      774,
+      1000,
+      1000,
+      1000,
+      99,
+      1000,
+      1000,
+      1000,
+      123,
+      332,
+      910,
+      1000,
+      806,
+      1000,
+      1000,
+      63,
+      1000
+    ],
+    "mean_return": 1774.7116768334226,
+    "population_std_return": 824.6118449109582,
+    "mean_length": 770.65,
+    "raw_steps": 15413,
+    "actions": 15393,
+    "invalid_actions": 0,
+    "dropped_actions": 0,
+    "dropped_observations": 0,
+    "submitted_observations": 15413,
+    "actual_env_fps_episode_mean": 4.912483565125906,
+    "mean_of_episode_latency_ms": 142.64479040267122,
+    "action_weighted_mean_latency_ms": 142.5879202059613,
+    "action_latency_p95_ms": 148.24638906866312,
+    "within_episode_wall_interval_count": 15393,
+    "within_episode_wall_interval_mean_ms": 200.00094455385926,
+    "within_episode_wall_interval_min_ms": 198.1171239167452,
+    "within_episode_wall_interval_max_ms": 201.79568137973547,
+    "actual_env_fps_note": "Per-episode actual_env_fps includes startup overhead; within-episode wall intervals are reported separately.",
+    "model_revision": "20f7452df3e281b6aea0f5a66a1602eb70297490",
+    "checkpoint_sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+    "config_sha256": "50304b94913641a189e1481915ff5540773a016b2233d71a7bc45443418eb877",
+    "prompt_map_sha256": "310e240a812bcb44139959d28a331ad14fa0f54bb01bd9cfdcd8656b7df99401",
+    "matched_B_config": true,
+    "native_manifest_binding": {
+      "uses_state": true,
+      "state_dim": 27,
+      "action_dim": 8,
+      "action_carrier": "native",
+      "state_normalization": {
+        "type": "min_max",
+        "min": [
+          0.45712053775787354,
+          0.42918211221694946,
+          -0.28085193037986755,
+          -0.4074253737926483,
+          -0.5388316512107849,
+          -0.615333616733551,
+          -0.09917011857032776,
+          -0.6332460641860962,
+          -1.2830065488815308,
+          -0.6662936806678772,
+          -1.334958553314209,
+          -0.5831336379051208,
+          -0.09470116347074509,
+          -1.3858237266540527,
+          -3.0080575942993164,
+          -2.8010413646698,
+          -5.880337715148926,
+          -6.067046165466309,
+          -4.152688026428223,
+          -11.533856391906738,
+          -11.273722648620605,
+          -10.24732494354248,
+          -11.152595520019531,
+          -14.95035457611084,
+          -13.3801851272583,
+          -7.523056507110596,
+          -10.32906723022461
+        ],
+        "max": [
+          0.9959571361541748,
+          1.098469853401184,
+          0.19786669313907623,
+          0.22940033674240112,
+          0.9005154967308044,
+          0.664892852306366,
+          1.2915066480636597,
+          0.5571166276931763,
+          0.09998779743909836,
+          0.630243182182312,
+          0.09576962888240814,
+          0.5611315369606018,
+          1.318502426147461,
+          6.299008369445801,
+          3.3342177867889404,
+          3.019660234451294,
+          5.154758930206299,
+          6.224544525146484,
+          5.348186016082764,
+          15.20273494720459,
+          11.13831615447998,
+          5.624480724334717,
+          5.72410774230957,
+          12.494281768798828,
+          12.612875938415527,
+          6.370715618133545,
+          11.95276927947998
+        ]
+      }
+    },
+    "post_evaluation_checkpoint_sha256_verified": true,
+    "post_evaluation_checkpoint_hash": {
+      "state": "COMPLETED_F_CHECKPOINT_STREAM_HASH_VERIFIED",
+      "path": "/dev/shm/lsb-ant-profile-f-20f7452d/ant/profile_latency/Pi05/checkpoints/model.pt",
+      "bytes": 10922682525,
+      "mtime_epoch": 1790005295,
+      "sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+      "hash_source": "streaming_sha256_of_F_yaml_policy_checkpoint_path",
+      "gpu0_protection_check": "empty_before_hash",
+      "hash_started_utc": "2026-09-22T03:26:04.687079982Z",
+      "hash_finished_utc": "2026-09-22T03:26:10.917816690Z"
+    },
+    "prompt_key": 1,
+    "prompt_sha256": "310e240a812bcb44139959d28a331ad14fa0f54bb01bd9cfdcd8656b7df99401",
+    "protocol": {
+      "env_fps": 5,
+      "obs_fps": 5,
+      "raw_step_cap": 1000,
+      "mode": "realtime",
+      "latency": "measured-only sync_cuda",
+      "workers": 1,
+      "H": 1,
+      "state_dim": 27,
+      "action_dim": 8
+    },
+    "preclip_commands_requiring_native_clip": 5,
+    "preclip_range": [
+      -0.9922769069671631,
+      1.013676643371582
+    ],
+    "state_note": "State is bound to the F manifest/config and prior loader verification; per-step state vectors were not logged.",
+    "capacity_drop_note": "Observation and action drop counts are taken from the actual F step/attempt/action joins and episode metadata.",
+    "baseline": {
+      "source": "B-recovered-audit.json",
+      "state": "B20_SPLIT_SESSION_DERIVED_AUDIT_PASSED",
+      "split_session": {
+        "original_ids": [
+          0,
+          1,
+          2,
+          3,
+          4,
+          5,
+          6,
+          7,
+          8,
+          9,
+          10,
+          11,
+          12,
+          13,
+          14,
+          15,
+          16,
+          17
+        ],
+        "recovery_ids": [
+          18,
+          19
+        ],
+        "container_restarted": true,
+        "diffusion_rng_stream_reset": true,
+        "continuous_session_equivalence": false
+      },
+      "rng_caveat": "The B recovery session reset the diffusion RNG stream; it is retained as recorded."
+    }
+  },
+  "paired": [
+    {
+      "seed": 42,
+      "B": 32.501595490013216,
+      "F": 2215.21984877659,
+      "delta": 2182.718253286577
+    },
+    {
+      "seed": 43,
+      "B": 56.46892726482151,
+      "F": 764.2694680291046,
+      "delta": 707.800540764283
+    },
+    {
+      "seed": 44,
+      "B": 96.83510560036257,
+      "F": 2290.599554986267,
+      "delta": 2193.7644493859043
+    },
+    {
+      "seed": 45,
+      "B": 95.12004972167394,
+      "F": 1840.29021669272,
+      "delta": 1745.170166971046
+    },
+    {
+      "seed": 46,
+      "B": 68.4796112625599,
+      "F": 2231.3439323113707,
+      "delta": 2162.864321048811
+    },
+    {
+      "seed": 47,
+      "B": 9.427925579208752,
+      "F": 2373.9321151344743,
+      "delta": 2364.5041895552654
+    },
+    {
+      "seed": 48,
+      "B": 180.91789918774197,
+      "F": 2372.1426528938036,
+      "delta": 2191.2247537060616
+    },
+    {
+      "seed": 49,
+      "B": 126.02311329849451,
+      "F": 166.30488169019765,
+      "delta": 40.28176839170314
+    },
+    {
+      "seed": 50,
+      "B": 239.58810433012508,
+      "F": 2302.3785949510366,
+      "delta": 2062.7904906209114
+    },
+    {
+      "seed": 51,
+      "B": 3.683059784739528,
+      "F": 2367.2167682154372,
+      "delta": 2363.5337084306975
+    },
+    {
+      "seed": 52,
+      "B": 291.6266216954973,
+      "F": 2310.4893684982335,
+      "delta": 2018.8627468027362
+    },
+    {
+      "seed": 53,
+      "B": 297.4161642880803,
+      "F": 246.45578812725222,
+      "delta": -50.960376160828105
+    },
+    {
+      "seed": 54,
+      "B": 50.97806575612095,
+      "F": 677.2055059936616,
+      "delta": 626.2274402375407
+    },
+    {
+      "seed": 55,
+      "B": 48.653726214514066,
+      "F": 2090.486226750155,
+      "delta": 2041.8325005356407
+    },
+    {
+      "seed": 56,
+      "B": 243.76432224961033,
+      "F": 2344.0109866336898,
+      "delta": 2100.2466643840794
+    },
+    {
+      "seed": 57,
+      "B": 78.96932272634479,
+      "F": 1810.7094305991875,
+      "delta": 1731.7401078728426
+    },
+    {
+      "seed": 58,
+      "B": 252.69957930958387,
+      "F": 2392.8136264328004,
+      "delta": 2140.1140471232166
+    },
+    {
+      "seed": 59,
+      "B": 142.4877330153016,
+      "F": 2337.4267695838707,
+      "delta": 2194.939036568569
+    },
+    {
+      "seed": 60,
+      "B": 86.73317256303835,
+      "F": 104.35524685972497,
+      "delta": 17.62207429668662
+    },
+    {
+      "seed": 61,
+      "B": 229.67420439888363,
+      "F": 2256.5825535088743,
+      "delta": 2026.9083491099907
+    }
+  ],
+  "mean_gain": 1643.1092616465867,
+  "population_std_gain": 824.5230676778788,
+  "wins": 19,
+  "ties": 0,
+  "losses": 1,
+  "gain_percent": 1248.5403549120786,
+  "protocol": "Ant native state27/action8, seeds42-61, cap1000, env/obs5/5FPS, one worker, measured-only CUDA-synchronized latency, B prompt0 and F prompt1, hold/FIFO; no IID delay.",
+  "causal_scope": "Descriptive paired comparison under the two measured-latency runs; no equal-delay or architecture-ranking claim."
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/evaluation/status.json

@@ -0,0 +1,47 @@
+{
+  "state": "PROFILE_MODEL_F20_DEPLOYMENT_AUDIT_VERIFIED",
+  "model_revision": "20f7452df3e281b6aea0f5a66a1602eb70297490",
+  "checkpoint_sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+  "training_steps": 5000,
+  "formal_source_state": "COMPLETED",
+  "profile_validation_history": {
+    "state": "CPU_STATE_DICT_AND_BUNDLE_HASH_VERIFIED",
+    "saved_model_gpu_forward": "NOT_RUN",
+    "closed_loop_evaluation": "NOT_RUN"
+  },
+  "F_deployed_inference_audit": "F20_FULL_STEP_ACTION_AUDIT_PASSED",
+  "F_independent_heldout_gpu_forward": "NOT_RUN",
+  "F": {
+    "F_publication_revision": "591db41843567bf4a1946ddfe66d2f51aaf8d7c6",
+    "F_model_revision": "20f7452df3e281b6aea0f5a66a1602eb70297490",
+    "F_checkpoint_sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+    "F_audit_state": "F20_FULL_STEP_ACTION_AUDIT_PASSED",
+    "F_verification_published_sha256": "bf9d8c4605b5a4406dea883b8bf498aa13c1d88d7c7cc83293374ccdd36c3777",
+    "F_comparison_published_sha256": "bbdce3420304948b37554a7c4f2b37097ee92ac49baa08bc08b4c1d2a9266150",
+    "F_config_published_sha256": "387c32ea04ca864934ad10e1b31e41490e526cae71b08cc84e0eb4057ae6313a",
+    "F_checkpoint_hash_published_sha256": "acccd0cfcd9c034dda2ba318ea86aa80b55fdca8691bef71e9f66d88f3817c3b",
+    "seeds": "42-61",
+    "episodes": 20,
+    "raw_step_cap": 1000,
+    "raw_steps": 15413,
+    "actions": 15393,
+    "invalid_actions": 0,
+    "dropped_actions": 0,
+    "mean_length": 770.65,
+    "F_mean_return": 1774.7116768334226,
+    "F_population_std": 824.6118449109582,
+    "action_latency_mean_ms": 142.5879202059613,
+    "action_latency_p95_ms": 148.24638906866312,
+    "paired_gain": 1643.1092616465867,
+    "paired_gain_percent": 1248.5403549120786,
+    "wins_ties_losses": [
+      19,
+      0,
+      1
+    ],
+    "baseline_caveat": "The B recovery session reset the diffusion RNG stream; it is retained as recorded.",
+    "causal_scope": "Descriptive paired comparison under the two measured-latency runs; no equal-delay or architecture-ranking claim."
+  },
+  "weights_changed": false,
+  "metadata_only_closeout": true
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/latency_prompt_map.json

@@ -0,0 +1,7 @@
+{
+  "1": {
+    "prompt": "Make the Ant move forward as fast as possible without falling. Predict eight continuous torques in [-1, 1] ordered as back right hip, back right ankle, front left hip, front left ankle, front right hip, front right ankle, back left hip, and back left ankle. Current action latency is 1 raw frames (200.00 ms).  The environment runs at 5 FPS and observations are emitted at 5 FPS. Choose the best next action.",
+    "latency_raw_frames": 1,
+    "latency_ms": 200.0
+  }
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/manifest.json

@@ -0,0 +1,232 @@
+{
+  "dataset_name": "ant_h1_5fps_profile",
+  "env_name": "ant",
+  "episodes": 90,
+  "frames": 89808,
+  "task_prompts": [
+    "Make the Ant move forward as fast as possible without falling. Predict eight continuous torques in [-1, 1] ordered as back right hip, back right ankle, front left hip, front left ankle, front right hip, front right ankle, back left hip, and back left ankle. Current action latency is 1 raw frames (200.00 ms).  The environment runs at 5 FPS and observations are emitted at 5 FPS. Choose the best next action."
+  ],
+  "format": "lerobot_v2_image_parquet",
+  "source": "${PI05_RUN_DIR}/profile_latency/data/raw_profile",
+  "integration_name": "gymnasium",
+  "task_name": "ant",
+  "action_layout": "gymnasium_continuous_v1",
+  "action_labels": [
+    "back_right_hip_torque",
+    "back_right_ankle_torque",
+    "front_left_hip_torque",
+    "front_left_ankle_torque",
+    "front_right_hip_torque",
+    "front_right_ankle_torque",
+    "back_left_hip_torque",
+    "back_left_ankle_torque"
+  ],
+  "carrier_action_labels": [
+    "back_right_hip_torque",
+    "back_right_ankle_torque",
+    "front_left_hip_torque",
+    "front_left_ankle_torque",
+    "front_right_hip_torque",
+    "front_right_ankle_torque",
+    "back_left_hip_torque",
+    "back_left_ankle_torque"
+  ],
+  "action_dim": 8,
+  "active_action_dim": 8,
+  "action_carrier": "native",
+  "rows_unit": "decision_step",
+  "obs_fps": 5.0,
+  "obs_stride_raw_frames": 1,
+  "uses_state": true,
+  "state_dim": 27,
+  "state_labels": [
+    "torso_z",
+    "torso_quaternion_w",
+    "torso_quaternion_x",
+    "torso_quaternion_y",
+    "torso_quaternion_z",
+    "front_left_hip_angle",
+    "front_left_ankle_angle",
+    "front_right_hip_angle",
+    "front_right_ankle_angle",
+    "back_left_hip_angle",
+    "back_left_ankle_angle",
+    "back_right_hip_angle",
+    "back_right_ankle_angle",
+    "torso_x_velocity",
+    "torso_y_velocity",
+    "torso_z_velocity",
+    "torso_angular_velocity_x",
+    "torso_angular_velocity_y",
+    "torso_angular_velocity_z",
+    "front_left_hip_angular_velocity",
+    "front_left_ankle_angular_velocity",
+    "front_right_hip_angular_velocity",
+    "front_right_ankle_angular_velocity",
+    "back_left_hip_angular_velocity",
+    "back_left_ankle_angular_velocity",
+    "back_right_hip_angular_velocity",
+    "back_right_ankle_angular_velocity"
+  ],
+  "state_normalization": {
+    "type": "min_max",
+    "min": [
+      0.45712053775787354,
+      0.42918211221694946,
+      -0.28085193037986755,
+      -0.4074253737926483,
+      -0.5388316512107849,
+      -0.615333616733551,
+      -0.09917011857032776,
+      -0.6332460641860962,
+      -1.2830065488815308,
+      -0.6662936806678772,
+      -1.334958553314209,
+      -0.5831336379051208,
+      -0.09470116347074509,
+      -1.3858237266540527,
+      -3.0080575942993164,
+      -2.8010413646698,
+      -5.880337715148926,
+      -6.067046165466309,
+      -4.152688026428223,
+      -11.533856391906738,
+      -11.273722648620605,
+      -10.24732494354248,
+      -11.152595520019531,
+      -14.95035457611084,
+      -13.3801851272583,
+      -7.523056507110596,
+      -10.32906723022461
+    ],
+    "max": [
+      0.9959571361541748,
+      1.098469853401184,
+      0.19786669313907623,
+      0.22940033674240112,
+      0.9005154967308044,
+      0.664892852306366,
+      1.2915066480636597,
+      0.5571166276931763,
+      0.09998779743909836,
+      0.630243182182312,
+      0.09576962888240814,
+      0.5611315369606018,
+      1.318502426147461,
+      6.299008369445801,
+      3.3342177867889404,
+      3.019660234451294,
+      5.154758930206299,
+      6.224544525146484,
+      5.348186016082764,
+      15.20273494720459,
+      11.13831615447998,
+      5.624480724334717,
+      5.72410774230957,
+      12.494281768798828,
+      12.612875938415527,
+      6.370715618133545,
+      11.95276927947998
+    ]
+  },
+  "robot_type": "rl_games_gymnasium",
+  "latency_prompt_map_path": "${PI05_RUN_DIR}/profile_latency/data/lerobot/ant_h1_5fps_profile/latency_prompt_map.json",
+  "custom_mixtures_path": "${PI05_RUN_DIR}/profile_latency/data/lerobot/_generated_mixtures/ant_h1_5fps_profile.json",
+  "gymnasium_task": {
+    "action_space": {
+      "dtype": "float32",
+      "high": [
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0,
+        1.0
+      ],
+      "labels": [
+        "back_right_hip_torque",
+        "back_right_ankle_torque",
+        "front_left_hip_torque",
+        "front_left_ankle_torque",
+        "front_right_hip_torque",
+        "front_right_ankle_torque",
+        "back_left_hip_torque",
+        "back_left_ankle_torque"
+      ],
+      "low": [
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0,
+        -1.0
+      ],
+      "type": "box"
+    },
+    "base_prompt": "Make the Ant move forward as fast as possible without falling. Predict eight continuous torques in [-1, 1] ordered as back right hip, back right ankle, front left hip, front left ankle, front right hip, front right ankle, back left hip, and back left ankle.",
+    "env_fps": 5.0,
+    "env_id": "LatencyBench/AntContinuous-v0",
+    "frame_stack": 1,
+    "make_kwargs": {
+      "base_env_id": "Ant-v4",
+      "base_make_kwargs": {
+        "exclude_current_positions_from_observation": true,
+        "use_contact_forces": false
+      },
+      "render_mode": "rgb_array"
+    },
+    "noop_action": [
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0
+    ],
+    "obs_fps": 5.0,
+    "registration_imports": [
+      "latency_bench.envs.gymnasium_ant"
+    ],
+    "state_space": {
+      "labels": [
+        "torso_z",
+        "torso_quaternion_w",
+        "torso_quaternion_x",
+        "torso_quaternion_y",
+        "torso_quaternion_z",
+        "front_left_hip_angle",
+        "front_left_ankle_angle",
+        "front_right_hip_angle",
+        "front_right_ankle_angle",
+        "back_left_hip_angle",
+        "back_left_ankle_angle",
+        "back_right_hip_angle",
+        "back_right_ankle_angle",
+        "torso_x_velocity",
+        "torso_y_velocity",
+        "torso_z_velocity",
+        "torso_angular_velocity_x",
+        "torso_angular_velocity_y",
+        "torso_angular_velocity_z",
+        "front_left_hip_angular_velocity",
+        "front_left_ankle_angular_velocity",
+        "front_right_hip_angular_velocity",
+        "front_right_ankle_angular_velocity",
+        "back_left_hip_angular_velocity",
+        "back_left_ankle_angular_velocity",
+        "back_right_hip_angular_velocity",
+        "back_right_ankle_angular_velocity"
+      ]
+    },
+    "task_name": "ant"
+  },
+  "validation_dataset_name": "ant_h1_5fps_profile__val",
+  "validation_episodes": 10,
+  "validation_frames": 9757
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/provenance.json

@@ -0,0 +1,34 @@
+{
+  "task": "ant",
+  "model": "qwenpi_v3",
+  "training_condition": "latency-aware",
+  "training_run_id": "ant_pi05_profile_h1_5fps_g128_20260920",
+  "source": {
+    "repo_id": "latency-sensitive-bench/Standard-Pipeline",
+    "revision": "2208875f92b21e3f8ec2363a30a6705dfe56cc04",
+    "path": "ant/profile_latency/Pi05",
+    "url": "https://huggingface.co/latency-sensitive-bench/Standard-Pipeline/tree/2208875f92b21e3f8ec2363a30a6705dfe56cc04/ant/profile_latency/Pi05"
+  },
+  "hf_repo_id": "latency-sensitive-bench/benchmark-models",
+  "hf_path": "latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920",
+  "checkpoint": {
+    "source_file": "ant/profile_latency/Pi05/checkpoints/model.pt",
+    "source_sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+    "source_bytes": 10922682525,
+    "file": "checkpoints/model.pt",
+    "selection_rule": "only published VLA checkpoint",
+    "method": "copy",
+    "sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+    "bytes": 10922682525
+  },
+  "config_source": "ant/profile_latency/Pi05/config.yaml",
+  "task_contract_file": "task_contract.json",
+  "missing_source_files": [],
+  "migration": {
+    "operation": "copy_selected_inference_bundle",
+    "source_modified": false,
+    "evaluation_performed": false,
+    "experiment_status": "preserved_in_source_evidence"
+  },
+  "action_horizon": 1
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/source/README.md

@@ -0,0 +1,5 @@
+# Ant Pi0.5 — profile latency closeout
+
+This profile-latency model is the fresh QwenPI_v3 Ant H1/5FPS training result, with native state27/action8 and the published checkpoint SHA `01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f`. The original package validation remains CPU tensor and bundle verification; that historical record continues to state that an independent saved-model GPU forward was not run.
+
+A separate F20 deployment audit exercised this fixed published bundle on seeds42–61 at measured 5/5FPS latency, cap1000, native state/action contracts. It audited 20 complete traces and is published at dataset revision `591db41843567bf4a1946ddfe66d2f51aaf8d7c6` with F mean `1774.711677` and B/F paired comparison in `evaluation/comparison.json`. The F audit has no equal-delay or architecture-ranking causal scope.

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/source/provenance.json

@@ -0,0 +1,562 @@
+{
+  "task": "ant",
+  "backbone": {
+    "repo_id": "Qwen/Qwen3-VL-4B-Instruct",
+    "revision": "ebb281ec70b05090aa6165b016eac8ec08e71b17"
+  },
+  "checkpoint": "steps_5000_pytorch_model.pt",
+  "training_steps": 5000,
+  "seed": 42,
+  "global_batch": 128,
+  "training_run_id": "ant_pi05_profile_h1_5fps_g128_20260920",
+  "condition": "profile_latency",
+  "source": {
+    "task": "ant",
+    "backbone": {
+      "repo_id": "Qwen/Qwen3-VL-4B-Instruct",
+      "revision": "ebb281ec70b05090aa6165b016eac8ec08e71b17"
+    },
+    "teacher": {
+      "state": "TEACHER_FULL_TRACE_AUDIT_PASSED",
+      "verified_at": "2026-09-21T06:03:32.767075+00:00",
+      "run": "h1_5fps_g128_20260920",
+      "env_fps": 5,
+      "obs_fps": 5,
+      "raw_frames_per_training_decision": 1,
+      "budget_counted_env_steps": 10000000,
+      "actual_env_steps": 10002432,
+      "seed": 2222,
+      "profile_sha256": "19534f5cce184dd4987b8234b7d712851caef92c2ea7ca0451d8d5aa90cd850c",
+      "source_profile_capture_fps": [
+        5,
+        5
+      ],
+      "profile_ms_rescaled": false,
+      "selection": {
+        "selected": "final",
+        "candidates": {
+          "final": {
+            "checkpoint": "${PI05_RUN_DIR}/profile_latency/teacher/checkpoints/ant_pi05_profile_appo_h1_5fps_g128_20260920/checkpoint_p0/checkpoint_000019536_10002432.pth",
+            "sha256": "bd40c91b4d7cbcd74b60b240638b1587e3269e93ace6fe41166c25c60da4181f",
+            "episodes": 20,
+            "seeds": [
+              104771,
+              104772,
+              104773,
+              104774,
+              104775,
+              104776,
+              104777,
+              104778,
+              104779,
+              104780,
+              104781,
+              104782,
+              104783,
+              104784,
+              104785,
+              104786,
+              104787,
+              104788,
+              104789,
+              104790
+            ],
+            "returns": [
+              4387.684316148426,
+              4506.74232190309,
+              4479.172451874021,
+              4389.610530796052,
+              4527.969076051201,
+              4532.91391579832,
+              4480.349285462922,
+              4648.747994777685,
+              4536.536483122084,
+              4485.51685443897,
+              4460.708050489044,
+              4419.17757232338,
+              4393.322009171424,
+              4492.807308077739,
+              4510.353408400214,
+              4407.056274506973,
+              4576.498658344057,
+              4466.795111615093,
+              4342.463588897995,
+              4449.385739559862
+            ],
+            "mean": 4474.6905475879275,
+            "population_std": 71.05590293801662,
+            "strict_gt3000": 20
+          },
+          "training_best": {
+            "checkpoint": "${PI05_RUN_DIR}/profile_latency/teacher/checkpoints/ant_pi05_profile_appo_h1_5fps_g128_20260920/checkpoint_p0/best_000019536_10002432_reward_3737.439.pth",
+            "sha256": "4ffd8b0969e7c68515561cef98f5364cf0535a4d1edd46d9d15fc4b645d4aa4d",
+            "episodes": 20,
+            "seeds": [
+              104771,
+              104772,
+              104773,
+              104774,
+              104775,
+              104776,
+              104777,
+              104778,
+              104779,
+              104780,
+              104781,
+              104782,
+              104783,
+              104784,
+              104785,
+              104786,
+              104787,
+              104788,
+              104789,
+              104790
+            ],
+            "returns": [
+              4387.684316148426,
+              4506.74232190309,
+              4479.172451874021,
+              4389.610530796052,
+              4527.969076051201,
+              4532.91391579832,
+              4480.349285462922,
+              4648.747994777685,
+              4536.536483122084,
+              4485.51685443897,
+              4460.708050489044,
+              4419.17757232338,
+              4393.322009171424,
+              4492.807308077739,
+              4510.353408400214,
+              4407.056274506973,
+              4576.498658344057,
+              4466.795111615093,
+              4342.463588897995,
+              4449.385739559862
+            ],
+            "mean": 4474.6905475879275,
+            "population_std": 71.05590293801662,
+            "strict_gt3000": 20
+          }
+        },
+        "completed_at": "2026-09-21T05:54:05.270163+00:00"
+      },
+      "evaluations": {
+        "selection_final": {
+          "episodes": 20,
+          "seeds": [
+            104771,
+            104772,
+            104773,
+            104774,
+            104775,
+            104776,
+            104777,
+            104778,
+            104779,
+            104780,
+            104781,
+            104782,
+            104783,
+            104784,
+            104785,
+            104786,
+            104787,
+            104788,
+            104789,
+            104790
+          ],
+          "returns": [
+            4387.684316148426,
+            4506.74232190309,
+            4479.172451874021,
+            4389.610530796052,
+            4527.969076051201,
+            4532.91391579832,
+            4480.349285462922,
+            4648.747994777685,
+            4536.536483122084,
+            4485.51685443897,
+            4460.708050489044,
+            4419.17757232338,
+            4393.322009171424,
+            4492.807308077739,
+            4510.353408400214,
+            4407.056274506973,
+            4576.498658344057,
+            4466.795111615093,
+            4342.463588897995,
+            4449.385739559862
+          ],
+          "lengths": [
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000
+          ],
+          "mean_return": 4474.6905475879275,
+          "population_std_return": 71.05590293801663,
+          "mean_length": 1000.0,
+          "raw_steps": 20000,
+          "cap_episodes": 20,
+          "mean_of_episode_latency_ms": 145.46123989918118,
+          "action_weighted_mean_latency_ms": 145.46123989918115,
+          "action_latency_p95_ms": 151.58370965278118,
+          "actions": 20000,
+          "invalid_actions": 0,
+          "dropped_actions": 0,
+          "submitted_observations": 20000,
+          "dropped_observations": 0,
+          "observation_opportunities": 20000,
+          "observation_drop_fraction": 0.0,
+          "raw_rewards_lengths_and_admission_counters_match": true,
+          "raw_step_indices_and_clock_verified": true,
+          "mode": "profile_sample",
+          "config_sha256": "33ad9f282ca42be2a0bcacaffeb8d480a24be09caa2dac044e2699f1d48c2b54",
+          "strict_gt3000": 20,
+          "commands_requiring_native_clip": 0,
+          "state_contract": "State27 bound to native config/checkpoint and prior environment preflight; no per-step state vector claim."
+        },
+        "selection_training_best": {
+          "episodes": 20,
+          "seeds": [
+            104771,
+            104772,
+            104773,
+            104774,
+            104775,
+            104776,
+            104777,
+            104778,
+            104779,
+            104780,
+            104781,
+            104782,
+            104783,
+            104784,
+            104785,
+            104786,
+            104787,
+            104788,
+            104789,
+            104790
+          ],
+          "returns": [
+            4387.684316148426,
+            4506.74232190309,
+            4479.172451874021,
+            4389.610530796052,
+            4527.969076051201,
+            4532.91391579832,
+            4480.349285462922,
+            4648.747994777685,
+            4536.536483122084,
+            4485.51685443897,
+            4460.708050489044,
+            4419.17757232338,
+            4393.322009171424,
+            4492.807308077739,
+            4510.353408400214,
+            4407.056274506973,
+            4576.498658344057,
+            4466.795111615093,
+            4342.463588897995,
+            4449.385739559862
+          ],
+          "lengths": [
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000
+          ],
+          "mean_return": 4474.6905475879275,
+          "population_std_return": 71.05590293801663,
+          "mean_length": 1000.0,
+          "raw_steps": 20000,
+          "cap_episodes": 20,
+          "mean_of_episode_latency_ms": 145.46123989918118,
+          "action_weighted_mean_latency_ms": 145.46123989918115,
+          "action_latency_p95_ms": 151.58370965278118,
+          "actions": 20000,
+          "invalid_actions": 0,
+          "dropped_actions": 0,
+          "submitted_observations": 20000,
+          "dropped_observations": 0,
+          "observation_opportunities": 20000,
+          "observation_drop_fraction": 0.0,
+          "raw_rewards_lengths_and_admission_counters_match": true,
+          "raw_step_indices_and_clock_verified": true,
+          "mode": "profile_sample",
+          "config_sha256": "9b8c6ef91a10ed0a709a3a18a52a863417f8459416f9de2d2674005fc885d42a",
+          "strict_gt3000": 20,
+          "commands_requiring_native_clip": 0,
+          "state_contract": "State27 bound to native config/checkpoint and prior environment preflight; no per-step state vector claim."
+        },
+        "E10": {
+          "episodes": 10,
+          "seeds": [
+            104771,
+            104772,
+            104773,
+            104774,
+            104775,
+            104776,
+            104777,
+            104778,
+            104779,
+            104780
+          ],
+          "returns": [
+            4387.684316148426,
+            4506.74232190309,
+            4479.172451874021,
+            4389.610530796052,
+            4527.969076051201,
+            4532.91391579832,
+            4480.349285462922,
+            4648.747994777685,
+            4536.536483122084,
+            4485.51685443897
+          ],
+          "lengths": [
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000,
+            1000
+          ],
+          "mean_return": 4497.524323037277,
+          "population_std_return": 71.51292984931624,
+          "mean_length": 1000.0,
+          "raw_steps": 10000,
+          "cap_episodes": 10,
+          "mean_of_episode_latency_ms": 145.47968413719246,
+          "action_weighted_mean_latency_ms": 145.47968413719244,
+          "action_latency_p95_ms": 151.63293832470646,
+          "actions": 10000,
+          "invalid_actions": 0,
+          "dropped_actions": 0,
+          "submitted_observations": 10000,
+          "dropped_observations": 0,
+          "observation_opportunities": 10000,
+          "observation_drop_fraction": 0.0,
+          "raw_rewards_lengths_and_admission_counters_match": true,
+          "raw_step_indices_and_clock_verified": true,
+          "mode": "profile_sample",
+          "config_sha256": "61915ecb228b9ade648570a82164f7007ba7071f138c39b2fcc975bd9442b22d",
+          "strict_gt3000": 10,
+          "commands_requiring_native_clip": 0,
+          "state_contract": "State27 bound to native config/checkpoint and prior environment preflight; no per-step state vector claim."
+        }
+      },
+      "probe": {
+        "source_state_sha256": "55b41a8afec1f04cc93fcd0ce889e5862d4ded18a0a9ed0c12b6a79af0f9d589",
+        "attempts": 12,
+        "accepted": 12,
+        "rejected": 0,
+        "accepted_specs": [
+          {
+            "attempt_idx": 0,
+            "episode_idx": 0,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4303.197367310524,
+            "seed": 2222,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 1,
+            "episode_idx": 1,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4538.047294437885,
+            "seed": 2223,
+            "split": "val"
+          },
+          {
+            "attempt_idx": 2,
+            "episode_idx": 2,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4552.908017247915,
+            "seed": 2224,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 3,
+            "episode_idx": 3,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4452.552011370659,
+            "seed": 2225,
+            "split": "val"
+          },
+          {
+            "attempt_idx": 4,
+            "episode_idx": 4,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4484.449267148972,
+            "seed": 2226,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 5,
+            "episode_idx": 5,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4352.745992147713,
+            "seed": 2227,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 6,
+            "episode_idx": 6,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4649.359137028456,
+            "seed": 2228,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 7,
+            "episode_idx": 7,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4498.849343210459,
+            "seed": 2229,
+            "split": "val"
+          },
+          {
+            "attempt_idx": 8,
+            "episode_idx": 8,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4381.613434404135,
+            "seed": 2230,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 9,
+            "episode_idx": 9,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4592.18594032526,
+            "seed": 2231,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 10,
+            "episode_idx": 10,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4324.289011418819,
+            "seed": 2232,
+            "split": "train"
+          },
+          {
+            "attempt_idx": 11,
+            "episode_idx": 11,
+            "episode_length": 1000,
+            "episode_raw_frames": 1000,
+            "episode_raw_return": 4577.22168815136,
+            "seed": 2233,
+            "split": "train"
+          }
+        ]
+      },
+      "strict_gate": ">3000",
+      "E10_seed_caveat": "E10 repeats the first10selection seeds, not independent evaluation",
+      "raw_budget_caveat": "Actual counted envsteps are read from Final checkpoint, not W&B telemetry; target10M atstride1.",
+      "downstream": "After positive12attemptprobe,100accepted strict>3000 within120TOTALincludingprobe12,actualreplayQA,freshPi05global128/micro32/acc2/GCfalse/ZeRO2/5000updates,andRTX3090F20; otherwise scientificstop",
+      "environment_preflight_sha256": "a6aab1ec52c552f83efdac6cf5d24901f5a3e42479aa7a00ee5a37550c3e2b59",
+      "source_P_verification_sha256": "008bf81ff19b73e6c6ca81752be4d5cab95be2626536ee20835b23880cd278fa",
+      "gate": "PASSED"
+    },
+    "profile_sha256": "19534f5cce184dd4987b8234b7d712851caef92c2ea7ca0451d8d5aa90cd850c",
+    "probe_anchors_sha256": "738da45fe09135d0357c35f9c9e49cf5b8c9bedd8e2ca366d065fb2d46051b1f",
+    "dataset": {
+      "attempts": 101,
+      "accepted": 100,
+      "max_total_attempts": 120,
+      "splits": {
+        "train": {
+          "rows": 89808,
+          "episodes": 90,
+          "return_min": 3643.214549936354,
+          "action_min": -0.9429033994674683,
+          "action_max": 0.9785619378089905
+        },
+        "val": {
+          "rows": 9757,
+          "episodes": 10,
+          "return_min": 3314.8423104584217,
+          "action_min": -0.9211073517799377,
+          "action_max": 0.9520043134689331
+        }
+      },
+      "sha256": {
+        "metadata.json": "8d3f1b02ee54ad97283b03392bd5529d72a4b2dfa9f94229db7e9c4e628d88d6",
+        "train.parquet": "5bbd64c7d8052722cff5a8efd5369d5ab9078db9e0307105e202fc7b15bfb8f6",
+        "val.parquet": "8e6b6c4bdb705d273e1e56ca6e94db22f835d969a391fa6a8f0a7045d0f8ddac"
+      }
+    },
+    "evaluation_prompt_keys": [
+      "1"
+    ],
+    "global_batch": 128,
+    "micro_batch": 32,
+    "accumulation": 2,
+    "gradient_checkpointing": false,
+    "zero_stage": 2,
+    "hardware_allocation": {
+      "physical_gpu_indices": [
+        4,
+        5
+      ],
+      "gpu_type": "H200",
+      "gpus": 2,
+      "reason": "Coordinator reallocated same2H200 recipe from occupied6/7 to idle4/5; scientifichyperparameters unchanged"
+    }
+  },
+  "training_config_sha256": "8fcc1fdca9a567462f38a896ed3d0bf6c830b1bf4842236a07b2381d943962c2",
+  "dataset_manifest_sha256": "e48b48b325cdf8253c1559bd411e23ef7961531c442df69f8a47be0a2cead11b"
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/task_contract.json

@@ -0,0 +1,94 @@
+{
+  "action_space": {
+    "dtype": "float32",
+    "high": [
+      1.0,
+      1.0,
+      1.0,
+      1.0,
+      1.0,
+      1.0,
+      1.0,
+      1.0
+    ],
+    "labels": [
+      "back_right_hip_torque",
+      "back_right_ankle_torque",
+      "front_left_hip_torque",
+      "front_left_ankle_torque",
+      "front_right_hip_torque",
+      "front_right_ankle_torque",
+      "back_left_hip_torque",
+      "back_left_ankle_torque"
+    ],
+    "low": [
+      -1.0,
+      -1.0,
+      -1.0,
+      -1.0,
+      -1.0,
+      -1.0,
+      -1.0,
+      -1.0
+    ],
+    "type": "box"
+  },
+  "base_prompt": "Make the Ant move forward as fast as possible without falling. Predict eight continuous torques in [-1, 1] ordered as back right hip, back right ankle, front left hip, front left ankle, front right hip, front right ankle, back left hip, and back left ankle.",
+  "env_fps": 5.0,
+  "env_id": "LatencyBench/AntContinuous-v0",
+  "frame_stack": 1,
+  "make_kwargs": {
+    "base_env_id": "Ant-v4",
+    "base_make_kwargs": {
+      "exclude_current_positions_from_observation": true,
+      "use_contact_forces": false
+    },
+    "render_mode": "rgb_array"
+  },
+  "noop_action": [
+    0.0,
+    0.0,
+    0.0,
+    0.0,
+    0.0,
+    0.0,
+    0.0,
+    0.0
+  ],
+  "obs_fps": 5.0,
+  "registration_imports": [
+    "latency_bench.envs.gymnasium_ant"
+  ],
+  "state_space": {
+    "labels": [
+      "torso_z",
+      "torso_quaternion_w",
+      "torso_quaternion_x",
+      "torso_quaternion_y",
+      "torso_quaternion_z",
+      "front_left_hip_angle",
+      "front_left_ankle_angle",
+      "front_right_hip_angle",
+      "front_right_ankle_angle",
+      "back_left_hip_angle",
+      "back_left_ankle_angle",
+      "back_right_hip_angle",
+      "back_right_ankle_angle",
+      "torso_x_velocity",
+      "torso_y_velocity",
+      "torso_z_velocity",
+      "torso_angular_velocity_x",
+      "torso_angular_velocity_y",
+      "torso_angular_velocity_z",
+      "front_left_hip_angular_velocity",
+      "front_left_ankle_angular_velocity",
+      "front_right_hip_angular_velocity",
+      "front_right_ankle_angular_velocity",
+      "back_left_hip_angular_velocity",
+      "back_left_ankle_angular_velocity",
+      "back_right_hip_angular_velocity",
+      "back_right_ankle_angular_velocity"
+    ]
+  },
+  "task_name": "ant"
+}

archive/issued-action-fix-20261010/latency-aware/ant/vla/starvla-qwenpi_v3-h1/ant_pi05_profile_h1_5fps_g128_20260920/validation.json

@@ -0,0 +1,11 @@
+{
+  "state": "CPU_STATE_DICT_AND_BUNDLE_HASH_VERIFIED",
+  "checkpoint_sha256": "01ce038a96b6479d74680235df1edd6ce586ee040dd1cfa55bca9a9dfa60e70f",
+  "tensor_count": 1386,
+  "all_floating_tensors_finite": true,
+  "training_steps": 5000,
+  "global_batch": 128,
+  "saved_model_gpu_forward": "NOT_RUN",
+  "closed_loop_evaluation": "NOT_RUN",
+  "verified_at": "2026-09-21T15:24:41.867755+00:00"
+}