Publish RIDE-RL reproducibility artifacts

LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

README.md

@@ -0,0 +1,43 @@
+---
+license: apache-2.0
+library_name: pytorch
+tags:
+- rna
+- inverse-folding
+- reinforcement-learning
+- diffusion
+- grpo
+---
+# RIDE-RL v6f step 390
+
+This repository contains the selected RIDE-RL post-training checkpoint used in the 153-target evaluation.
+
+## Files
+
+- `policy_step000390.pt`: PyTorch checkpoint with a single top-level `model` state dictionary.
+- `config/rl_v6f_t75.yaml`: exact long-training configuration used for run v6f.
+- `SHA256SUMS`: file checksums.
+
+## Model and training
+
+The policy starts from the pretrained RIDE backbone and is post-trained with grouped policy updates and a geometric structural reward. The v6f run used 527 targets, 3 epochs, group size 8, 4 targets per update, learning rate `2e-4`, and 395 optimizer updates. Validation was run every 10 updates on 153 targets.
+
+The selected snapshot is step 390 (the last scheduled validation snapshot; training ended at step 395). On the 153-target, candidate-0 evaluation it achieved GDT-TS 0.3827, C4'-TM 0.3400, RMSD 7.9196 Å, C1'-TM 0.3855 (n=126), success rate 0.3007, and sequence recovery 0.5056. See the project repository for evaluation code and metric definitions.
+
+## Loading
+
+```python
+import torch
+checkpoint = torch.load("policy_step000390.pt", map_location="cpu")
+state_dict = checkpoint["model"]
+```
+
+The architecture and runtime are provided by the RIDE-RL code repository: https://github.com/Gabrile166/RIDE-RL
+
+## Dependencies and license
+
+This checkpoint is a derivative of the Apache-2.0 RIDER/RIDE implementation and is released under Apache-2.0. It does not include RhoFold+, USalign, EternaFold, or other third-party weights/binaries; obtain those from their original sources under their own terms.
+
+## Citation
+
+Please cite the upstream RIDER work and this repository when using the checkpoint.

SHA256SUMS

@@ -0,0 +1,4 @@
+dcebb89289181a0946116b899ef4eafd75cd64637b321e93a350e6c12fdb7cbf  policy_step000390.pt
+3907ad0b8615595024f82dd505759e452e1ff0493b2ae5389d39ea042d04384d  config/rl_v6f_t75.yaml
+c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4  LICENSE
+d804550e4ff4ad2785d7a1f58e3b6d159ae835d64baa243e16ad6ae383e9f376  README.md

config/rl_v6f_t75.yaml

@@ -0,0 +1,179 @@
+# ===== 跑5 (v6f): 基于跑4(v6e) events.jsonl 实测动态调参 =====
+# 与跑2/3/4 相比只改 4 项, reward preset / group_size / batch / nft_beta /
+# reference_weight / z_clip / seed / shuffle_seed 一律不动。
+#
+#  1) lr 1e-4 -> 2e-4
+#     依据: v6e 全程 grad_norm 0.058-0.086, 而 grad_clip_norm=1.0 -- 比阈值小
+#     一个数量级, 裁剪一次都没触发。翻倍后预期 ~0.12-0.17, 仍远低于 1.0。
+#  2) lr_min_ratio 0.1 -> 0.3
+#     依据: v6e 末段 lr 已衰到 1e-5, 但 batch/good_rate 在 230-263 段达到全程
+#     最高 0.4744, reward_mean 仍在 0.52 高位 -- 训练是被 lr 掐死的, 不是收敛的。
+#     0.3 让 lr 底线抬到 6e-5。
+#  3) epochs 2 -> 3, max_outer_steps / lr_total_steps 263 -> 395
+#     依据: reward_mean 分段 0.4231->0.5233 单调上行未饱和, advantage 未退化。
+#  4) validation.keep_runs 3 -> 4
+#     依据: v6e 曲线真峰值在 step130 (GDT 0.376947), 但 top3 预算把它挤掉了,
+#     峰值权重不存在、无法评测。
+# ============================================================
+data:
+  pool_manifest: artifacts/evaluation_pools/train_final_v4.pt
+  ride_checkpoint: saved_models/checkpoint.h5
+  processed_data: ../data/processed.pt
+  split_file: ../data/das_split.pt
+  condition_noise_scale: 0.0
+  condition_snapshot_enabled: false
+  x0_representation: one_hot_01
+ride_model:
+  node_in_dim:
+  - 15
+  - 4
+  node_h_dim:
+  - 256
+  - 24
+  edge_in_dim:
+  - 131
+  - 3
+  edge_h_dim:
+  - 128
+  - 4
+  num_layers: 5
+  drop_rate: 0.5
+  out_dim: 4
+  time_cond: true
+noise_schedule:
+  schedule: linear
+  continuous_beta_0: 0.1
+  continuous_beta_1: 20.0
+  eps: 0.001
+rollout:
+  group_size: 8
+  sampler:
+    name: x0_renoise
+    n_steps: 50
+    version: v1
+  temperature:
+    initial: 1.0
+    min: 0.2
+    max: 1.5
+    step: 0.1
+  rescue:
+    max_rounds: 1
+    min_unique_sequences: 2
+    min_valid_sequences: 2
+reward:
+  oracle:
+    backend: local_jsonl_worker
+    python: ${RHOFOLD_PYTHON:-python}
+    worker_script: scripts/rhofold_oracle_worker.py
+    ckpt: ${RHOFOLD_CHECKPOINT:-artifacts/models/rhofold/RhoFold_pretrained.pt}
+    output_dir: outputs/rl_v6f/oracle
+    timeout_sec: 600
+    device: cuda:0
+    fake: false
+  metric_cache:
+    dir: ${RIDE_RL_CACHE:-cache/rl_metrics}
+    metric_version: ride_evaluator_v2_geo6
+    alignment_config: c4p_kabsch_centered
+  composer:
+    # === A/B 实验的唯一变量 ===
+    # A 组: balanced_cosine_v1 = 三分量加权算术平均
+    # B 组: geometric_v3       = 六分量加权几何平均, 权重由
+    #       scripts/reward/reward_design_v3.py 在 1224 条真实候选上求解得到
+    #       (gdt .25 / tm .30 / rmsd .30 / scmcc .05 / edefect .05 / composition .05)
+    #       alignment 0.9608 (门槛 0.70), discrimination 1.2061 (门槛 1.20, 唯一紧约束)
+    # good 门槛由 preset 内部的 threshold_position=0.60 推出, 不再从这里读,
+    # 保留字段只为与 A 组配置结构完全对齐。
+    preset: geometric_v3
+    good:
+      gdt_ts: 0.5
+      tm_score: 0.5
+      rmsd: 4.0
+advantage:
+  group_normalize: true
+  z_clip: 2.0
+  scale:
+    kind: ema
+    decay: 0.95
+    epsilon: 1.0e-06
+    min_valid: 2
+trainer:
+  loss:
+    nft_beta: 0.1
+    reference_weight: 0.02
+    k_t: 8
+    weighting: uniform_eps
+    residual_normalization: none
+    divide_by_beta: true
+  optim:
+    lr: 0.0002
+    weight_decay: 0.0
+    grad_clip_norm: 1.0
+    gradient_accumulation_steps: 1
+    amp: false
+    microbatch_size: 1
+    lr_schedule: warmup_cosine
+    lr_warmup_ratio: 0.05
+    lr_min_ratio: 0.3
+    lr_total_steps: 395
+  ema:
+    decay_max: 0.999
+    update_interval: 1
+runtime:
+  seed: 0
+  device: cuda
+  mode: train
+  ranks_per_target: 1
+  # 527 unique targets * 3 passes / 4 per batch = 395 updates (1581 // 4).
+  # engine.py: total_steps = len(targets) // batch_targets, so max_outer_steps
+  # must equal that or the walk clamps it.
+  epochs: 3
+  shuffle_targets: true
+  shuffle_seed: 20260826
+  budget:
+    max_outer_steps: 395
+    max_oracle_calls: 200000
+    max_wall_hours: 72.0
+    checkpoint_every_oracle_calls: 128
+validation:
+  enabled: true
+  # Held-out 153, disjoint from the 527 training pool (verified zero sequence overlap).
+  pool_manifest: artifacts/evaluation_pools/test_final_v3_std.pt
+  config: configs/diffv3_te0.yaml
+  # t75 only has 4 usable cards (PCI 3E:00.0 / 3F:00.0 are disabled -- they drop
+  # the driver under load), so validation shares them with training instead of
+  # getting spare cards of its own. This is safe here because the worker is
+  # fully detached (start_new_session, torchrun env stripped) and never joins the
+  # process group: the earlier stall came from a mismatched barrier on the
+  # critical path, not from co-residency. Memory fits -- RhoFold+ peaks under
+  # 11 GiB and training holds roughly 10 GiB of a 24 GiB 3090.
+  # 4 shards over 4 cards keeps every card busy during the fold phase.
+  devices: "cuda:0,cuda:1,cuda:2,cuda:3"
+  test_freq: 10
+  # 4 instead of 3: v6e 的真峰值在 step130, 按 top-2+latest 的 top3 预算被挤掉。
+  keep_runs: 4
+  before_train: true
+  n_samples: 8
+  timeout_sec: 7200.0
+  usalign_binary: ../third_party/USalign/USalign
+logging:
+  jsonl: outputs/rl_v6f/train.jsonl
+  wandb:
+    enabled: true
+    project: ride-rl
+    entity: gabriel166714-fudan-university
+    name: v6f-t75-geometric-v3-lr2e4-min03-3ep-395step
+    tags:
+    - v6
+    - 4gpu
+    - ddp
+    - batch4
+    - grpo
+    - geometric_v3
+    - train527
+    - test153
+checkpoint:
+  dir: outputs/rl_v6f/checkpoints
+  save_latest: true
+evaluation:
+  deterministic_anchor_temperature: 0.0
+  group_size: 2

policy_step000390.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcebb89289181a0946116b899ef4eafd75cd64637b321e93a350e6c12fdb7cbf
+size 41025394