Add files using upload-large-folder tool
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- .gitattributes +2 -0
- README.md +104 -0
- THIRD_PARTY_NOTICES.md +33 -0
- canonical/mind2web/default--train--0000.jsonl.gz +3 -0
- canonical/mind2web/default--train--0001.jsonl.gz +3 -0
- canonical/mind2web/default--train--0002.jsonl.gz +3 -0
- canonical/mind2web/default--train--0003.jsonl.gz +3 -0
- canonical/mind2web/default--train--0005.jsonl.gz +3 -0
- canonical/mind2web/default--train--0007.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0002.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0003.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0005.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0007.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0010.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0011.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0012.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0014.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0016.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0017.jsonl.gz +3 -0
- canonical/nvidia_open_swe_traces/v1.2--openhands--0018.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0000.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0001.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0002.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0003.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0004.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0005.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0006.jsonl.gz +3 -0
- canonical/swesmith_trajectories/default--tool--0007.jsonl.gz +3 -0
- docs/dataset_adapter.md +352 -0
- docs/licensing_and_federation.md +68 -0
- docs/nemotron_sft_demo.md +48 -0
- duplicate_provenance.jsonl.gz +3 -0
- evaluation/model_imitation.json +5 -0
- evaluation/per_trajectory.jsonl.gz +3 -0
- evaluation/run_config.json +77 -0
- evaluation/summary.json +2901 -0
- examples.html +3 -0
- examples.selection.json +142 -0
- examples/nemotron_sft_demo.json +191 -0
- index.sqlite +3 -0
- manifest.json +1277 -0
- rankings.jsonl.gz +3 -0
- registry/permissions.json +215 -0
- registry/sources.json +465 -0
- staging_report.json +26 -0
- structurally_eligible.index.jsonl.gz +3 -0
- successful.index.jsonl.gz +3 -0
- task_splits.jsonl.gz +3 -0
- training_eligible.index.jsonl.gz +3 -0
- unknown.index.jsonl.gz +3 -0
.gitattributes
CHANGED
|
@@ -58,3 +58,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 58 |
# Video files - compressed
|
| 59 |
*.mp4 filter=lfs diff=lfs merge=lfs -text
|
| 60 |
*.webm filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
| 58 |
# Video files - compressed
|
| 59 |
*.mp4 filter=lfs diff=lfs merge=lfs -text
|
| 60 |
*.webm filter=lfs diff=lfs merge=lfs -text
|
| 61 |
+
examples.html filter=lfs diff=lfs merge=lfs -text
|
| 62 |
+
index.sqlite filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,104 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
pretty_name: SFT Benchmark Trajectories
|
| 4 |
+
task_categories:
|
| 5 |
+
- text-generation
|
| 6 |
+
tags:
|
| 7 |
+
- agent-trajectories
|
| 8 |
+
- sft
|
| 9 |
+
- tool-use
|
| 10 |
+
- software-engineering
|
| 11 |
+
size_categories:
|
| 12 |
+
- 100K<n<1M
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
# SFT Benchmark Trajectories
|
| 16 |
+
|
| 17 |
+
A canonical, model-agnostic inventory of recorded expert demonstrations and agent completion trajectories. This is a **federated release**: the complete 155,278-trajectory metadata inventory, scores, lineage and source pointers are public, while canonical event payloads are hosted only where the source-specific redistribution audit supports it.
|
| 18 |
+
|
| 19 |
+
## Release summary
|
| 20 |
+
|
| 21 |
+
- Complete canonical inventory after exact-record deduplication: **155,278**
|
| 22 |
+
- Canonical payloads hosted here: **136,373**
|
| 23 |
+
- Metadata-only records recoverable from pinned upstream sources: **18,905**
|
| 24 |
+
- Unique tasks: **67,610**
|
| 25 |
+
- Successful / unsuccessful / unknown outcomes: **54,296 / 79,973 / 21,009**
|
| 26 |
+
- Structurally eligible: **143,999**
|
| 27 |
+
- Registry training-eligible before use-rights filtering: **110,208**
|
| 28 |
+
- Native environment replays performed by this project: **0**
|
| 29 |
+
- Scores rebuilt for this publication: **no** (persisted scores were copied)
|
| 30 |
+
- Source code commit: `59e8576bfe3372b6a601a898c980b5db31eb1789`
|
| 31 |
+
|
| 32 |
+
| Canonical source | Inventory | Public payload | Training permission | Transformed redistribution |
|
| 33 |
+
|---|---:|---|---|---|
|
| 34 |
+
| `agentsuite_tau2` | 8,340 | metadata_only | uncertain | uncertain |
|
| 35 |
+
| `browsecomp_rollouts` | 5,063 | metadata_only | uncertain | uncertain |
|
| 36 |
+
| `mind2web` | 1,009 | hosted | permitted_with_conditions | permitted |
|
| 37 |
+
| `nvidia_open_swe_traces` | 51,626 | hosted | permitted | permitted |
|
| 38 |
+
| `official_tau3` | 200 | hosted | permitted_with_conditions | permitted |
|
| 39 |
+
| `swebench_multilingual_trajectories` | 129 | hosted | permitted_with_conditions | permitted_with_conditions |
|
| 40 |
+
| `swebench_verified_trajectories` | 4,544 | hosted | permitted_with_conditions | permitted_with_conditions |
|
| 41 |
+
| `swegym_openhands` | 491 | hosted | permitted | permitted |
|
| 42 |
+
| `swegym_sampled` | 5,502 | metadata_only | uncertain | uncertain |
|
| 43 |
+
| `swesmith_trajectories` | 24,100 | hosted | permitted_with_conditions | permitted_with_conditions |
|
| 44 |
+
| `terminalbench_trajectories` | 34,263 | hosted | permitted_with_conditions | permitted_with_conditions |
|
| 45 |
+
| `terminaltraj` | 20,000 | hosted | permitted | permitted_with_conditions |
|
| 46 |
+
| `toolbench` | 11 | hosted | permitted_with_conditions | permitted_with_conditions |
|
| 47 |
+
|
| 48 |
+
`permitted_with_conditions` is not the same as unrestricted use. Inspect `registry/permissions.json` for verification status, commercial-training status, conditions, evidence and pinned URLs. `THIRD_PARTY_NOTICES.md` contains attribution and notice guidance. This is an engineering audit, not legal advice.
|
| 49 |
+
|
| 50 |
+
## Download and consume
|
| 51 |
+
|
| 52 |
+
```bash
|
| 53 |
+
pip install 'git+https://github.com/heisenger/nvidia_benchmark.git@59e8576bfe3372b6a601a898c980b5db31eb1789'
|
| 54 |
+
pip install huggingface_hub
|
| 55 |
+
huggingface-cli download Heisenger/sft-benchmark-trajectories --repo-type dataset --local-dir ./sft-benchmark-trajectories
|
| 56 |
+
```
|
| 57 |
+
|
| 58 |
+
```python
|
| 59 |
+
from agent_trajectories import CanonicalDatasetAdapter, DatasetQuery
|
| 60 |
+
|
| 61 |
+
dataset = CanonicalDatasetAdapter("./sft-benchmark-trajectories")
|
| 62 |
+
query = DatasetQuery.for_training(
|
| 63 |
+
splits=("train",),
|
| 64 |
+
content_availability=("hosted",),
|
| 65 |
+
training_permissions=("permitted", "permitted_with_conditions"),
|
| 66 |
+
training_verifications=("verified",),
|
| 67 |
+
)
|
| 68 |
+
for sample in dataset.iter_examples(query, view="full_trajectory"):
|
| 69 |
+
print(sample.trajectory_id, sample.events)
|
| 70 |
+
break
|
| 71 |
+
```
|
| 72 |
+
|
| 73 |
+
Normal iteration reads persisted canonical data and never contacts upstream sources. If a consumer is authorized to retrieve a metadata-only source, they can run the explicit pinned source adapter and combine the resulting canonical shard without changing the public snapshot:
|
| 74 |
+
|
| 75 |
+
```python
|
| 76 |
+
dataset = CanonicalDatasetAdapter(
|
| 77 |
+
"./sft-benchmark-trajectories",
|
| 78 |
+
content_overlays={"agentsuite_tau2": "./authorized/agentsuite_tau2"},
|
| 79 |
+
)
|
| 80 |
+
```
|
| 81 |
+
|
| 82 |
+
The adapter rejects a payload query with a source-specific `MissingTrajectoryContentError` when only metadata is available. Metadata and permission queries remain available for the entire inventory.
|
| 83 |
+
|
| 84 |
+
## Layout
|
| 85 |
+
|
| 86 |
+
- `manifest.json`: portable relative paths, complete inventory counts, content availability and hashes.
|
| 87 |
+
- `canonical/`: hosted canonical gzip shards.
|
| 88 |
+
- `index.sqlite`: all 155,278 metadata/ranking/outcome records, including metadata-only sources.
|
| 89 |
+
- `registry/permissions.json`: queryable multi-dimensional permission audit.
|
| 90 |
+
- `rankings.jsonl.gz`, `task_splits.jsonl.gz`, outcome/eligibility indexes and `evaluation/`: persisted full-inventory derivations.
|
| 91 |
+
- `examples.html`: human-readable trajectory examples.
|
| 92 |
+
|
| 93 |
+
## Outcomes and scores
|
| 94 |
+
|
| 95 |
+
Native rewards and success labels are preserved from upstream records; they were not regenerated. The project computed structural completeness, action/observation, known-outcome, verification-reliability and SFT-eligibility quality scores, plus deterministic within-task ranks. It did **not** replay upstream environments or compute model NLL. Historical labels are not scores for a new target model.
|
| 96 |
+
|
| 97 |
+
## Licenses and limitations
|
| 98 |
+
|
| 99 |
+
This aggregate uses source-specific terms and therefore has repository license `other`. Each record retains source ID, immutable revision and record locator. Payload hosting does not erase embedded repository, model-provider, website-content, attribution or benchmark-contamination obligations. Sources with unresolved transformed-redistribution rights remain metadata-only rather than being dropped or falsely labelled permissive.
|
| 100 |
+
|
| 101 |
+
- Outcomes are predominantly source-reported, not independently replayed.
|
| 102 |
+
- Some actions and observations are encoded in text rather than typed tool-call fields.
|
| 103 |
+
- The inventory contains the imported subset at each pinned revision, not every upstream row.
|
| 104 |
+
- `training_eligible` is a structural/source-quality gate; combine it with permission filters for a compliant subset.
|
THIRD_PARTY_NOTICES.md
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Third-party notices and attribution
|
| 2 |
+
|
| 3 |
+
The canonical library is an aggregate. The package source code license does not
|
| 4 |
+
replace source-dataset terms. Every record retains `source_id`, immutable
|
| 5 |
+
revision and `record_locator`; the evidence and conditions used for publication
|
| 6 |
+
are machine-readable in `data/registry/permissions.json`.
|
| 7 |
+
|
| 8 |
+
Publicly hosted payloads use the following source terms:
|
| 9 |
+
|
| 10 |
+
| Source | Terms and attribution |
|
| 11 |
+
|---|---|
|
| 12 |
+
| `swegym_openhands` | [SWE-Gym/OpenHands-SFT-Trajectories](https://huggingface.co/datasets/SWE-Gym/OpenHands-SFT-Trajectories/tree/4aaa5a4a4b5861f4799d2336908760c190ac3b17), MIT. Credit the SWE-Gym authors and retain the MIT notice. |
|
| 13 |
+
| `nvidia_open_swe_traces` | [NVIDIA Open-SWE-Traces](https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586), CC-BY-4.0. Credit NVIDIA/Open-SWE-Traces, link CC-BY-4.0, identify canonical transformations, and retain the per-record repository license. |
|
| 14 |
+
| `terminaltraj` | [m-a-p/TerminalTraj](https://huggingface.co/datasets/m-a-p/TerminalTraj/tree/5c1823f4a8b9ca0cf02c27d2db52c5b35b53a308), linked publisher project under Apache-2.0. Credit the TerminalTraj authors and honor embedded MIT/Apache/BSD repository terms. |
|
| 15 |
+
| `official_tau3` | [Sierra Research tau2-bench](https://github.com/sierra-research/tau2-bench/tree/dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd), MIT. Copyright Sierra Research; retain the MIT notice. Training on benchmark tasks/rewards must be disclosed as evaluation overlap. |
|
| 16 |
+
| `mind2web` | [OSU-NLP Mind2Web](https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474), CC-BY-4.0. Credit the Mind2Web authors, link CC-BY-4.0, identify transformations, and retain the research-purpose/harm disclaimer. |
|
| 17 |
+
| `toolbench` | [OpenBMB ToolBench](https://github.com/OpenBMB/ToolBench/tree/d56fdd89faf8c91fa135090b212bb9057ee5cfc2), Apache-2.0 with research-and-education intended use. Credit OpenBMB/ToolBench and consider third-party API content rights. |
|
| 18 |
+
| `swebench_verified_trajectories` | [tarsur385/swebench-verified-trajectories](https://huggingface.co/datasets/tarsur385/swebench-verified-trajectories/tree/773748a7c1222e8a642a7059821498e14293562a), MIT. Retain the dataset notice; task repositories retain their own terms. |
|
| 19 |
+
| `swebench_multilingual_trajectories` | [AlienKevin/SWE-bench-Multilingual-trajectories](https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories/tree/7221c09c6cc0ed16d51c9cb9886a1ab484465670), MIT. Retain the dataset notice; task repositories retain their own terms. |
|
| 20 |
+
| `terminalbench_trajectories` | [yoonholee/terminalbench-trajectories](https://huggingface.co/datasets/yoonholee/terminalbench-trajectories/tree/04e8940f5b6736a7ce8d22224fe2f2af74163ed2), Apache-2.0. Credit the dataset publisher and Terminal-Bench; consider the terms of model outputs represented in public leaderboard data. |
|
| 21 |
+
| `swesmith_trajectories` | [SWE-bench/SWE-smith-trajectories](https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4), MIT. Credit SWE-smith and retain generated task repository terms. |
|
| 22 |
+
|
| 23 |
+
License references: [MIT](https://opensource.org/license/mit),
|
| 24 |
+
[Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0), and
|
| 25 |
+
[CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/legalcode).
|
| 26 |
+
|
| 27 |
+
The raw payloads for `agentsuite_tau2`, `swegym_sampled`, and
|
| 28 |
+
`browsecomp_rollouts` are metadata-only in the public aggregate because the
|
| 29 |
+
separate trajectory repositories do not establish sufficiently clear
|
| 30 |
+
transformed-redistribution terms. Their absence is not a claim that access,
|
| 31 |
+
research use, or every form of training is prohibited.
|
| 32 |
+
|
| 33 |
+
This file is an engineering attribution aid, not legal advice.
|
canonical/mind2web/default--train--0000.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:654357077e6249cd030e516a74bc3ac7551e26553809581bbebb56a4f87e4c37
|
| 3 |
+
size 60960835
|
canonical/mind2web/default--train--0001.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea62d59ed99fa6632c59bb62cf339054fe72e717c38d36109526414bb9ce2f60
|
| 3 |
+
size 61759916
|
canonical/mind2web/default--train--0002.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:77d3d9b0ef99cf5ff8300c7e0681847780752ce60e2c2e0887e247b491b48e22
|
| 3 |
+
size 61958505
|
canonical/mind2web/default--train--0003.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3807d46087383372f43ca0e3ed95d1b14e9ceca917940c652d136eca5157d7c
|
| 3 |
+
size 64515683
|
canonical/mind2web/default--train--0005.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd075e514c9006f2781cec84a0bbb816e1bc6ccdf22217885925e2ab5d02df8b
|
| 3 |
+
size 59608293
|
canonical/mind2web/default--train--0007.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa20a9fa7022a87f523148839fbbccae8576eecd011eb8a8b7a12be2472c8e5e
|
| 3 |
+
size 116176682
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0002.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3067df96c93975319e290aa0b5f7b4be0ceecb6275f761cea33db9b6b3c54ffd
|
| 3 |
+
size 305310880
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0003.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b3324ce204a000df83ad9e207b4318af57ed19d36c11d7cbbae7bf5db90b453
|
| 3 |
+
size 290056585
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0005.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92b92e2aba80e79cbb07e9ba071db946fe4aa885deef4aefa9d934ae3bff193e
|
| 3 |
+
size 5161644
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0007.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af1f7ed30b9037df06ebf7e7f0d146913ad813511e065e8f735a520b646b74bf
|
| 3 |
+
size 371128455
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0010.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2eff407239d58e2d9d525aa70402dcf01501ef3d3007d6407a69765c9c4936b5
|
| 3 |
+
size 370224921
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0011.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a68e3249212dbf0661db6e7d7a5b9afc6a2a4bf0172d0f32e9ad0cbf272abbc
|
| 3 |
+
size 83589148
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0012.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d54f3aebffe6bfdd1889488db25bf851210f000d6f6ca2fa2bebf83155fea40a
|
| 3 |
+
size 243965767
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0014.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e964d62766afd638a8526c5e44fca9581e2cea0d3ec0ffd0a14beeefc1e16c6c
|
| 3 |
+
size 249402952
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0016.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95b32f5071fe69c1949e66623413b9e02970973ea501040336955b56d6de0b16
|
| 3 |
+
size 242546761
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0017.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c17ad85f0ef74b315314fc6929e3efe7875f5d88f319208bc07fbbbc621c296
|
| 3 |
+
size 74960935
|
canonical/nvidia_open_swe_traces/v1.2--openhands--0018.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:331b6e4d1611cc71c643b7a1af41744218d25aa5d6be33675f3ca2bcc20ab183
|
| 3 |
+
size 373422381
|
canonical/swesmith_trajectories/default--tool--0000.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:909cd6654ba6b253bf83874168fcfef57eb7ae700d94d39d4e10da15e2d4df22
|
| 3 |
+
size 208736286
|
canonical/swesmith_trajectories/default--tool--0001.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8efdc46253958b9bfdba85da004f1de435709881773472d627a21372f197ba78
|
| 3 |
+
size 128599828
|
canonical/swesmith_trajectories/default--tool--0002.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80d7a0105bee1b9cde168524b9cd702005786014464a3d5adcb40dcb7a0603ff
|
| 3 |
+
size 115440687
|
canonical/swesmith_trajectories/default--tool--0003.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48df7f6939fd5619d1376c3e6909cfb71d88db56c11ce0511989b51620e44a9c
|
| 3 |
+
size 147332694
|
canonical/swesmith_trajectories/default--tool--0004.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da2007c58f68ec2de3736cf7dbb6b1356e1d1657bfe8ef5f03ecb77b03a7e227
|
| 3 |
+
size 93517927
|
canonical/swesmith_trajectories/default--tool--0005.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97e66d5f2c7e74a53d2d56528d9972ad3ad455ba7472900f8ba474e7bfecba9a
|
| 3 |
+
size 147229136
|
canonical/swesmith_trajectories/default--tool--0006.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fcf23d8b7b0f907ce65e99fd25983b479e099e69506eaeef4c6794f6a663b1cd
|
| 3 |
+
size 162574019
|
canonical/swesmith_trajectories/default--tool--0007.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff5c4100eb4fa6610b91012d93d071989a7cb33510f9b5e273e19e41a05d7b39
|
| 3 |
+
size 141721185
|
docs/dataset_adapter.md
ADDED
|
@@ -0,0 +1,352 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Consumer-facing dataset adapter
|
| 2 |
+
|
| 3 |
+
`agent_trajectories` has two intentionally separate adapter layers:
|
| 4 |
+
|
| 5 |
+
1. **Source adapters** acquire a pinned external release and normalize its
|
| 6 |
+
benchmark-specific rows into the canonical `Trajectory` schema. They run
|
| 7 |
+
during ingestion or an explicit refresh.
|
| 8 |
+
2. **The consumer adapter** reads only the persisted canonical library and its
|
| 9 |
+
read-only metadata index. Training and evaluation jobs never import
|
| 10 |
+
benchmark parsers, contact upstream hosts, or rewrite dataset files.
|
| 11 |
+
|
| 12 |
+
The stable entry point is `CanonicalDatasetAdapter`. Its API version is `1.1`;
|
| 13 |
+
the package version is `0.3.0`.
|
| 14 |
+
|
| 15 |
+
```python
|
| 16 |
+
from agent_trajectories import CanonicalDatasetAdapter
|
| 17 |
+
|
| 18 |
+
dataset = CanonicalDatasetAdapter("data/library")
|
| 19 |
+
print(dataset.available_benchmarks())
|
| 20 |
+
print(dataset.available_sources())
|
| 21 |
+
```
|
| 22 |
+
|
| 23 |
+
The adapter validates `manifest.json`, requires the persisted `index.sqlite`
|
| 24 |
+
and available canonical shards, and opens SQLite in read-only mode. A federated
|
| 25 |
+
release can contain full-inventory metadata while marking some payloads as
|
| 26 |
+
metadata-only. It keeps only lightweight selected index rows in memory.
|
| 27 |
+
Unsampled event payloads are parsed one trajectory at a time.
|
| 28 |
+
|
| 29 |
+
## Query contract
|
| 30 |
+
|
| 31 |
+
`DatasetQuery` fields use OR semantics within a field and AND semantics across
|
| 32 |
+
fields:
|
| 33 |
+
|
| 34 |
+
- `benchmarks`: stable families such as `swe-bench`, `swe-gym`, `swe-smith`,
|
| 35 |
+
`terminal-bench`, `tau2-bench`, `tau3-bench`, `browsecomp`, `mind2web`, and
|
| 36 |
+
`toolbench`;
|
| 37 |
+
- `source_ids`: exact canonical sources such as `swegym_sampled` or
|
| 38 |
+
`agentsuite_tau2`;
|
| 39 |
+
- `source_versions`: exact immutable source revisions;
|
| 40 |
+
- `agent_models` and `agent_scaffolds`;
|
| 41 |
+
- `trajectory_ids`, `task_keys`, and compatible `environment_keys`;
|
| 42 |
+
- task-level library `splits`: `train`, `validation`, `test`, or
|
| 43 |
+
`protected_evaluation`;
|
| 44 |
+
- `quality_tiers` and persisted task/environment `quality_ranks`;
|
| 45 |
+
- dynamic `content_availability`: `hosted`, `overlay`, or `metadata_only`;
|
| 46 |
+
- audited `research_permissions`, `training_permissions`,
|
| 47 |
+
`training_verifications`, `commercial_training_permissions`, and
|
| 48 |
+
transformed `redistribution_permissions`;
|
| 49 |
+
- `structurally_eligible` and `training_eligible` gates;
|
| 50 |
+
- `selection`: `all_attempts`, `all_successful`, or `best_verified`.
|
| 51 |
+
|
| 52 |
+
The concise string aliases `all`, `successful`, and `best` are also accepted,
|
| 53 |
+
as are `full`/`full-trajectory` and `turn`/`assistant-turn` view names. Enum
|
| 54 |
+
members are preferred when static typing is available.
|
| 55 |
+
|
| 56 |
+
`DatasetQuery.for_training(...)` enables both structural and training-source
|
| 57 |
+
gates. This quality gate is not a legal-permission decision; combine it with
|
| 58 |
+
the permission fields for the downstream project's policy. `best_verified`
|
| 59 |
+
returns at most one successful source-evaluator or
|
| 60 |
+
independently reproduced trajectory per selected task. Apply an
|
| 61 |
+
`environment_keys` filter when alternatives from different environment
|
| 62 |
+
versions must not compete. Quality rank itself is always calculated within a
|
| 63 |
+
compatible task/environment group over the full canonical population; source
|
| 64 |
+
filtering never silently rewrites the stored rank.
|
| 65 |
+
|
| 66 |
+
## Full-trajectory SFT
|
| 67 |
+
|
| 68 |
+
```python
|
| 69 |
+
from agent_trajectories import (
|
| 70 |
+
CanonicalDatasetAdapter,
|
| 71 |
+
DatasetQuery,
|
| 72 |
+
TrainingView,
|
| 73 |
+
TrajectorySelection,
|
| 74 |
+
)
|
| 75 |
+
|
| 76 |
+
dataset = CanonicalDatasetAdapter("data/library")
|
| 77 |
+
query = DatasetQuery.for_training(
|
| 78 |
+
benchmarks=("swe-gym", "swe-smith"),
|
| 79 |
+
splits=("train",),
|
| 80 |
+
selection=TrajectorySelection.ALL_SUCCESSFUL,
|
| 81 |
+
)
|
| 82 |
+
|
| 83 |
+
for example in dataset.iter_examples(query, view=TrainingView.FULL_TRAJECTORY):
|
| 84 |
+
row = example.to_dict()
|
| 85 |
+
model_input = row["model_input"]
|
| 86 |
+
loss_mask = row["loss_mask"]
|
| 87 |
+
lineage = row["metadata"]["lineage"]
|
| 88 |
+
# Apply a model-specific chat template/tokenizer downstream.
|
| 89 |
+
```
|
| 90 |
+
|
| 91 |
+
Each full example has the complete event sequence, assistant-only
|
| 92 |
+
`target_event_ids`, and an aligned `loss_mask`. Outcomes, verification evidence
|
| 93 |
+
and ranks are in metadata, never in `model_input`.
|
| 94 |
+
|
| 95 |
+
## Assistant-turn SFT
|
| 96 |
+
|
| 97 |
+
```python
|
| 98 |
+
query = DatasetQuery.for_training(
|
| 99 |
+
source_ids=("agentsuite_tau2",),
|
| 100 |
+
splits=("train",),
|
| 101 |
+
)
|
| 102 |
+
|
| 103 |
+
for turn in dataset.iter_examples(query, view=TrainingView.ASSISTANT_TURN):
|
| 104 |
+
row = turn.to_dict()
|
| 105 |
+
assert len(row["context"]) == row["target_index"]
|
| 106 |
+
assert row["target_event_id"] == row["metadata"]["lineage"]["turn_id"]
|
| 107 |
+
model_input = row["model_input"] # causal prefix only
|
| 108 |
+
supervised_target = row["supervised_target"]
|
| 109 |
+
```
|
| 110 |
+
|
| 111 |
+
Full and turn views derive from the same canonical record, retain the same task
|
| 112 |
+
split and source lineage, and have equivalent assistant targets. Turn context
|
| 113 |
+
contains no target event, future observation, later action, outcome, or answer.
|
| 114 |
+
|
| 115 |
+
## Best trajectory and multiple completions
|
| 116 |
+
|
| 117 |
+
Select the best verified trajectory for every selected task:
|
| 118 |
+
|
| 119 |
+
```python
|
| 120 |
+
best = DatasetQuery(
|
| 121 |
+
benchmarks=("swe-bench",),
|
| 122 |
+
splits=("protected_evaluation",),
|
| 123 |
+
selection=TrajectorySelection.BEST_VERIFIED,
|
| 124 |
+
)
|
| 125 |
+
for record in dataset.iter_records(best):
|
| 126 |
+
print(record.metadata.task_key, record.metadata.quality_rank)
|
| 127 |
+
```
|
| 128 |
+
|
| 129 |
+
Request one exact canonical trajectory:
|
| 130 |
+
|
| 131 |
+
```python
|
| 132 |
+
record = dataset.get_trajectory("trajectory:...")
|
| 133 |
+
print(record.trajectory.events)
|
| 134 |
+
print(record.lineage())
|
| 135 |
+
```
|
| 136 |
+
|
| 137 |
+
Collapsed duplicate IDs are resolved to their retained canonical trajectory;
|
| 138 |
+
`contributing_trajectory_ids` preserves every original ID.
|
| 139 |
+
|
| 140 |
+
Enumerate all completions for a task, in quality order:
|
| 141 |
+
|
| 142 |
+
```python
|
| 143 |
+
first = next(dataset.metadata(DatasetQuery(source_ids=("agentsuite_tau2",))))
|
| 144 |
+
attempts = dataset.alternatives(first.task_key)
|
| 145 |
+
for attempt in attempts:
|
| 146 |
+
print(
|
| 147 |
+
attempt.metadata.trajectory_id,
|
| 148 |
+
attempt.metadata.quality_rank,
|
| 149 |
+
attempt.metadata.outcome_category,
|
| 150 |
+
attempt.metadata.agent_model,
|
| 151 |
+
)
|
| 152 |
+
```
|
| 153 |
+
|
| 154 |
+
`environment_key=` can restrict alternatives to one compatible environment.
|
| 155 |
+
|
| 156 |
+
## Deterministic sampling
|
| 157 |
+
|
| 158 |
+
```python
|
| 159 |
+
from agent_trajectories import SamplingConfig, SamplingStrategy
|
| 160 |
+
|
| 161 |
+
population = DatasetQuery.for_training(splits=("train",))
|
| 162 |
+
|
| 163 |
+
# Uniform trajectory sampling.
|
| 164 |
+
random_sample = SamplingConfig(count=1_000, seed=42)
|
| 165 |
+
|
| 166 |
+
# Uniform task sampling; the best matching attempt represents each drawn task.
|
| 167 |
+
task_sample = SamplingConfig(
|
| 168 |
+
count=1_000, seed=42, strategy=SamplingStrategy.TASK
|
| 169 |
+
)
|
| 170 |
+
|
| 171 |
+
# Round-robin sources, then sample within each source.
|
| 172 |
+
balanced_sample = SamplingConfig(
|
| 173 |
+
count=1_000, seed=42, strategy=SamplingStrategy.SOURCE_BALANCED
|
| 174 |
+
)
|
| 175 |
+
|
| 176 |
+
for record in dataset.iter_records(population, sampling=balanced_sample):
|
| 177 |
+
print(record.metadata.trajectory_id, record.metadata.source_id)
|
| 178 |
+
```
|
| 179 |
+
|
| 180 |
+
Set `with_replacement=True` for repeated draws. The same persisted library,
|
| 181 |
+
query, seed and sampling configuration produce the same ordered IDs.
|
| 182 |
+
Task-level sampling prevents tasks with many recorded attempts from dominating.
|
| 183 |
+
Source-balanced sampling distributes draws round-robin across non-empty sources
|
| 184 |
+
and deterministically redistributes unused capacity when a source is exhausted.
|
| 185 |
+
|
| 186 |
+
## Streaming and batching
|
| 187 |
+
|
| 188 |
+
```python
|
| 189 |
+
for batch in dataset.iter_batches(
|
| 190 |
+
8,
|
| 191 |
+
DatasetQuery.for_training(splits=("train",)),
|
| 192 |
+
view=TrainingView.ASSISTANT_TURN,
|
| 193 |
+
):
|
| 194 |
+
rows = [example.to_dict() for example in batch]
|
| 195 |
+
print(len(rows), rows[0]["sample_id"])
|
| 196 |
+
```
|
| 197 |
+
|
| 198 |
+
These are bounded Python batches, not tensors. Canonical storage stays
|
| 199 |
+
framework-independent and downstream code retains control of templates,
|
| 200 |
+
tokenization, truncation, packing and collation.
|
| 201 |
+
|
| 202 |
+
PyTorch is optional and imported lazily:
|
| 203 |
+
|
| 204 |
+
```python
|
| 205 |
+
torch_dataset = dataset.as_torch_iterable(
|
| 206 |
+
DatasetQuery.for_training(splits=("train",)),
|
| 207 |
+
view=TrainingView.ASSISTANT_TURN,
|
| 208 |
+
)
|
| 209 |
+
|
| 210 |
+
# Standard torch DataLoader worker processes are sharded by item index.
|
| 211 |
+
from torch.utils.data import DataLoader
|
| 212 |
+
loader = DataLoader(torch_dataset, batch_size=None, num_workers=2)
|
| 213 |
+
```
|
| 214 |
+
|
| 215 |
+
The core package has no ML-framework dependency. Install the optional extra
|
| 216 |
+
only when needed: `pip install 'agent-trajectory-datasets[torch]'`.
|
| 217 |
+
|
| 218 |
+
## Evaluation over all trajectories
|
| 219 |
+
|
| 220 |
+
Task splits are assigned once by canonical `task_key`, then joined into every
|
| 221 |
+
source and both derived views. A task cannot appear in more than one of
|
| 222 |
+
`train`, `validation`, `test`, and `protected_evaluation`.
|
| 223 |
+
|
| 224 |
+
```python
|
| 225 |
+
evaluation = DatasetQuery(
|
| 226 |
+
benchmarks=("tau2-bench",),
|
| 227 |
+
splits=("validation", "test", "protected_evaluation"),
|
| 228 |
+
selection=TrajectorySelection.ALL_ATTEMPTS,
|
| 229 |
+
)
|
| 230 |
+
|
| 231 |
+
for record in dataset.iter_records(evaluation):
|
| 232 |
+
print(record.metadata.task_key, record.trajectory.outcome.success)
|
| 233 |
+
```
|
| 234 |
+
|
| 235 |
+
For all references, source verdicts and verification evidence associated with
|
| 236 |
+
one task:
|
| 237 |
+
|
| 238 |
+
```python
|
| 239 |
+
task = next(dataset.metadata(evaluation)).task_key
|
| 240 |
+
bundle = dataset.evaluation_bundle(task, query=evaluation)
|
| 241 |
+
for reference in bundle.references:
|
| 242 |
+
print(
|
| 243 |
+
reference.trajectory_id,
|
| 244 |
+
reference.outcome.success,
|
| 245 |
+
reference.verification.method,
|
| 246 |
+
reference.verification.evidence,
|
| 247 |
+
reference.supervised_targets,
|
| 248 |
+
)
|
| 249 |
+
```
|
| 250 |
+
|
| 251 |
+
Historical source outcomes are evaluation references, not scores for a new
|
| 252 |
+
model. Native environment success still requires running the corresponding
|
| 253 |
+
benchmark evaluator. Unknown outcomes remain unknown rather than being counted
|
| 254 |
+
as failures.
|
| 255 |
+
|
| 256 |
+
## Metadata and lineage
|
| 257 |
+
|
| 258 |
+
Every `DatasetRecord` carries `TrajectoryMetadata`, including:
|
| 259 |
+
|
| 260 |
+
- canonical task, trajectory and environment IDs;
|
| 261 |
+
- benchmark family, source ID, immutable revision, upstream record locator and
|
| 262 |
+
canonical/source split;
|
| 263 |
+
- agent model and scaffold;
|
| 264 |
+
- quality tier, numeric score and within-task quality rank;
|
| 265 |
+
- structural/training eligibility;
|
| 266 |
+
- current payload availability and the complete source permission audit,
|
| 267 |
+
including verification state, commercial-training status and conditions;
|
| 268 |
+
- outcome category, reward, verification method/strength and independent-replay
|
| 269 |
+
status;
|
| 270 |
+
- every duplicate source trajectory ID collapsed into the canonical record.
|
| 271 |
+
|
| 272 |
+
`record.lineage()` adds the complete `SourceRef`, producer provenance, source
|
| 273 |
+
outcome and full verification evidence. Derived examples store that object at
|
| 274 |
+
`example.metadata["lineage"]`; assistant-turn examples also store `turn_id` and
|
| 275 |
+
full examples store all supervised `turn_ids`.
|
| 276 |
+
|
| 277 |
+
## Federated payloads and authorized overlays
|
| 278 |
+
|
| 279 |
+
`available_sources()` enumerates the complete indexed inventory;
|
| 280 |
+
`content_sources()` lists sources whose event payload is currently readable.
|
| 281 |
+
Inspect one permission record without loading trajectories:
|
| 282 |
+
|
| 283 |
+
```python
|
| 284 |
+
rights = dataset.permissions("agentsuite_tau2")
|
| 285 |
+
print(rights.dimensions["model_training"])
|
| 286 |
+
print(dataset.content_availability("agentsuite_tau2"))
|
| 287 |
+
```
|
| 288 |
+
|
| 289 |
+
When a project is separately authorized to retrieve a metadata-only source,
|
| 290 |
+
run its explicit pinned source adapter and provide the resulting canonical
|
| 291 |
+
file or directory as an overlay:
|
| 292 |
+
|
| 293 |
+
```python
|
| 294 |
+
dataset = CanonicalDatasetAdapter(
|
| 295 |
+
"./sft-benchmark-trajectories",
|
| 296 |
+
content_overlays={
|
| 297 |
+
"agentsuite_tau2": "./authorized/canonical/agentsuite_tau2",
|
| 298 |
+
},
|
| 299 |
+
)
|
| 300 |
+
|
| 301 |
+
query = DatasetQuery(
|
| 302 |
+
source_ids=("agentsuite_tau2",),
|
| 303 |
+
content_availability=("overlay",),
|
| 304 |
+
)
|
| 305 |
+
for record in dataset.iter_records(query):
|
| 306 |
+
...
|
| 307 |
+
```
|
| 308 |
+
|
| 309 |
+
The overlay is read-only and is not copied or uploaded. Canonical IDs join it
|
| 310 |
+
to the hosted full-inventory index. Selecting records whose payload is still
|
| 311 |
+
absent raises `MissingTrajectoryContentError` with the missing source IDs;
|
| 312 |
+
metadata queries never silently discard them.
|
| 313 |
+
|
| 314 |
+
## Initial ingestion and explicit refresh
|
| 315 |
+
|
| 316 |
+
Normal consumption performs no downloads. Initial ingestion is a separate,
|
| 317 |
+
explicit operation. For a hosted release:
|
| 318 |
+
|
| 319 |
+
```bash
|
| 320 |
+
.venv/bin/python scripts/import_hf_release.py SOURCE_ID DATASET_ID \
|
| 321 |
+
--revision IMMUTABLE_SHA \
|
| 322 |
+
--output data/full/canonical/OUTPUT \
|
| 323 |
+
--cache data/cache/full/OUTPUT \
|
| 324 |
+
--config CONFIG --split SPLIT
|
| 325 |
+
|
| 326 |
+
.venv/bin/python scripts/build_complete_library.py \
|
| 327 |
+
--input-dir data/full/canonical/OUTPUT \
|
| 328 |
+
--output data/library
|
| 329 |
+
```
|
| 330 |
+
|
| 331 |
+
The complete multi-source build command and source-specific revisions are in
|
| 332 |
+
the root README and `docs/trajectory_sources.md`.
|
| 333 |
+
|
| 334 |
+
A refresh is never implicit. Choose and audit a new upstream immutable SHA,
|
| 335 |
+
write it to a new canonical output directory, then deliberately rebuild the
|
| 336 |
+
library with the old or new directory selected:
|
| 337 |
+
|
| 338 |
+
```bash
|
| 339 |
+
.venv/bin/python scripts/import_hf_release.py SOURCE_ID DATASET_ID \
|
| 340 |
+
--revision NEW_IMMUTABLE_SHA \
|
| 341 |
+
--output data/full/canonical/OUTPUT_NEW_SHA \
|
| 342 |
+
--cache data/cache/full/OUTPUT_NEW_SHA \
|
| 343 |
+
--config CONFIG --split SPLIT
|
| 344 |
+
|
| 345 |
+
# Review release-manifest.json and rejects before replacing the old input.
|
| 346 |
+
.venv/bin/python scripts/build_complete_library.py \
|
| 347 |
+
--input-dir data/full/canonical/OUTPUT_NEW_SHA \
|
| 348 |
+
--output data/library-refreshed
|
| 349 |
+
```
|
| 350 |
+
|
| 351 |
+
This separation makes dataset changes reviewable and keeps training runs pinned
|
| 352 |
+
to the manifest and shard hashes they actually consumed.
|
docs/licensing_and_federation.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Permission audit and federated content
|
| 2 |
+
|
| 3 |
+
Permission is not a single Boolean. The registry records, for every canonical
|
| 4 |
+
source, separate findings for public access, local research use, model
|
| 5 |
+
training, commercial training, redistribution of original data, and
|
| 6 |
+
redistribution of transformed canonical trajectories. Each finding has both a
|
| 7 |
+
permission value and a verification value, plus source-specific conditions and
|
| 8 |
+
evidence URLs.
|
| 9 |
+
|
| 10 |
+
The audit vocabulary is:
|
| 11 |
+
|
| 12 |
+
- permission: `permitted`, `permitted_with_conditions`, `uncertain`, or
|
| 13 |
+
`restricted`;
|
| 14 |
+
- verification: `verified`, `inferred`, or `uncertain`;
|
| 15 |
+
- public payload: `hosted` or `metadata_only`.
|
| 16 |
+
|
| 17 |
+
The complete local inventory remains 155,278 trajectories. Publication does
|
| 18 |
+
not change scores, outcomes, ranks, splits, lineage or canonical hashes.
|
| 19 |
+
Instead, the public Hugging Face release contains the complete metadata index
|
| 20 |
+
and the payloads supported by the audit. Metadata-only sources retain pinned
|
| 21 |
+
upstream URLs and canonical shard hashes.
|
| 22 |
+
|
| 23 |
+
The initial public payload decision is:
|
| 24 |
+
|
| 25 |
+
- hosted: `swegym_openhands`, `nvidia_open_swe_traces`, `terminaltraj`,
|
| 26 |
+
`official_tau3`, `mind2web`, `toolbench`,
|
| 27 |
+
`swebench_verified_trajectories`,
|
| 28 |
+
`swebench_multilingual_trajectories`, `terminalbench_trajectories`, and
|
| 29 |
+
`swesmith_trajectories`;
|
| 30 |
+
- metadata-only: `agentsuite_tau2`, `swegym_sampled`, and
|
| 31 |
+
`browsecomp_rollouts`.
|
| 32 |
+
|
| 33 |
+
This is not a blanket declaration that hosted data is unrestricted or that
|
| 34 |
+
metadata-only data cannot be used. For example, Mind2Web combines CC-BY-4.0
|
| 35 |
+
metadata with a research-purpose statement; tau benchmark training creates
|
| 36 |
+
evaluation contamination that must be disclosed; and software-engineering
|
| 37 |
+
records can embed code governed by per-repository licenses.
|
| 38 |
+
|
| 39 |
+
Downstream projects can select a policy directly:
|
| 40 |
+
|
| 41 |
+
```python
|
| 42 |
+
from agent_trajectories import CanonicalDatasetAdapter, DatasetQuery
|
| 43 |
+
|
| 44 |
+
dataset = CanonicalDatasetAdapter("./sft-benchmark-trajectories")
|
| 45 |
+
strict_research = DatasetQuery.for_training(
|
| 46 |
+
content_availability=("hosted",),
|
| 47 |
+
training_permissions=("permitted", "permitted_with_conditions"),
|
| 48 |
+
training_verifications=("verified",),
|
| 49 |
+
commercial_training_permissions=("permitted", "permitted_with_conditions"),
|
| 50 |
+
)
|
| 51 |
+
for metadata in dataset.metadata(strict_research):
|
| 52 |
+
print(metadata.source_id, metadata.source_permissions.to_dict())
|
| 53 |
+
```
|
| 54 |
+
|
| 55 |
+
An authorized user can explicitly ingest a metadata-only source at its pinned
|
| 56 |
+
revision and overlay the canonical shards:
|
| 57 |
+
|
| 58 |
+
```python
|
| 59 |
+
dataset = CanonicalDatasetAdapter(
|
| 60 |
+
"./sft-benchmark-trajectories",
|
| 61 |
+
content_overlays={"agentsuite_tau2": "./canonical/agentsuite_tau2"},
|
| 62 |
+
)
|
| 63 |
+
```
|
| 64 |
+
|
| 65 |
+
The overlay is local and is never uploaded by the adapter. Canonical trajectory
|
| 66 |
+
IDs are matched against the persisted full-inventory index. Queries that select
|
| 67 |
+
missing content fail with `MissingTrajectoryContentError` and the relevant
|
| 68 |
+
source IDs instead of silently omitting records.
|
docs/nemotron_sft_demo.md
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Nemotron SFT tokenization demonstration
|
| 2 |
+
|
| 3 |
+
`scripts/demo_nemotron_sft.py` reads a real hosted canonical record through the
|
| 4 |
+
consumer adapter, derives full-trajectory and causal assistant-turn views,
|
| 5 |
+
applies the pinned native Nemotron chat template, builds assistant-only token
|
| 6 |
+
loss masks, and pads both samples into one framework-neutral model batch.
|
| 7 |
+
|
| 8 |
+
The demonstrated tokenizer is
|
| 9 |
+
`nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16` at immutable revision
|
| 10 |
+
`a9904d24bcc1d289a1950fa9d2b978c47cf903b9`. No model weights or training are
|
| 11 |
+
required.
|
| 12 |
+
|
| 13 |
+
```bash
|
| 14 |
+
pip install '.[tokenization]'
|
| 15 |
+
HF_HOME=.cache/huggingface PYTHONPATH=src python \
|
| 16 |
+
scripts/demo_nemotron_sft.py \
|
| 17 |
+
--library ./sft-benchmark-trajectories \
|
| 18 |
+
--output examples/nemotron_sft_demo.json
|
| 19 |
+
```
|
| 20 |
+
|
| 21 |
+
The real successful tau trajectory
|
| 22 |
+
`trajectory:f9cf8588a314fa9804d57402` has 13 events and six supervised assistant
|
| 23 |
+
targets. Results from the pinned tokenizer are:
|
| 24 |
+
|
| 25 |
+
| View | Total tokens | Supervised assistant tokens | Ignored tokens |
|
| 26 |
+
|---|---:|---:|---:|
|
| 27 |
+
| Full trajectory | 3,174 | 696 | 2,478 |
|
| 28 |
+
| One causal assistant turn | 1,877 | 60 | 1,817 |
|
| 29 |
+
|
| 30 |
+
For the turn view, 80 tokens belonging to prior assistant turns are explicitly
|
| 31 |
+
masked out; only the selected target at event index 5 contributes to loss. The
|
| 32 |
+
causal prefix is 1,817 tokens and does not include the target or any future
|
| 33 |
+
observation.
|
| 34 |
+
|
| 35 |
+
The resulting batch has fields `input_ids`, `attention_mask`, and `labels`,
|
| 36 |
+
shape `[2, 3174]`, right padding, and label ignore index `-100`. Every context
|
| 37 |
+
and padding label is `-100`; exactly 696 and 60 label positions are supervised
|
| 38 |
+
in the two rows. Its deterministic JSON representation has SHA-256
|
| 39 |
+
`d680ae7e4ea605b3db96370de6760d86ed227411dac901d74a511f72c3346ae8`.
|
| 40 |
+
|
| 41 |
+
Nemotron's pinned template does not ship Transformers `{% generation %}` spans.
|
| 42 |
+
The script makes no semantic template changes: it wraps the template's native
|
| 43 |
+
assistant-rendering branch with those spans so Transformers can return exact
|
| 44 |
+
token-aligned assistant masks. It disables historical-thinking truncation to
|
| 45 |
+
keep causal prefixes byte-for-byte stable. The script checks mask/label
|
| 46 |
+
alignment, target-prefix stability, non-empty supervision, and ignored padding
|
| 47 |
+
before emitting the report. See `examples/nemotron_sft_demo.json` for the event
|
| 48 |
+
preview, lineage, token previews, lengths, mask checks, and batch checksum.
|
duplicate_provenance.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea4cd16eac95b5fa34bc0413feb4a078347a214bcb2af15cfd8a03ea42482625
|
| 3 |
+
size 76971
|
evaluation/model_imitation.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"implementation": "The canonical full and assistant-turn exports, exact structured-action scorer, and a pinned Nemotron tokenizer/mask/batch demonstration are available; NLL requires model weights and inference.",
|
| 3 |
+
"reason": "No trained language model weights or inference accelerator were used. Tokenizer-derived lengths and assistant-only masks are reported separately and are not NLL.",
|
| 4 |
+
"status": "not-run"
|
| 5 |
+
}
|
evaluation/per_trajectory.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0dfc8d4e0517311818a245c86b83086557740cd6fce9cafc1acc2cde717042fe
|
| 3 |
+
size 15275530
|
evaluation/run_config.json
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"evaluation_errors": [],
|
| 3 |
+
"evaluator_versions": {
|
| 4 |
+
"canonical_validation": "1.0",
|
| 5 |
+
"exact_record_deduplication": "1.0",
|
| 6 |
+
"quality_ranking": "1.0",
|
| 7 |
+
"structured_action_exact_match": "1.0",
|
| 8 |
+
"task_split": "trajectory-library-v1",
|
| 9 |
+
"trajectory_quality": "1.0"
|
| 10 |
+
},
|
| 11 |
+
"finalization_runtime_seconds": 25.363,
|
| 12 |
+
"inputs": [
|
| 13 |
+
"metadata-only:agentsuite_tau2",
|
| 14 |
+
"metadata-only:browsecomp_rollouts",
|
| 15 |
+
"metadata-only:browsecomp_rollouts",
|
| 16 |
+
"canonical/mind2web/default--train--0000.jsonl.gz",
|
| 17 |
+
"canonical/mind2web/default--train--0001.jsonl.gz",
|
| 18 |
+
"canonical/mind2web/default--train--0002.jsonl.gz",
|
| 19 |
+
"canonical/mind2web/default--train--0003.jsonl.gz",
|
| 20 |
+
"canonical/mind2web/default--train--0004.jsonl.gz",
|
| 21 |
+
"canonical/mind2web/default--train--0005.jsonl.gz",
|
| 22 |
+
"canonical/mind2web/default--train--0006.jsonl.gz",
|
| 23 |
+
"canonical/mind2web/default--train--0007.jsonl.gz",
|
| 24 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0000.jsonl.gz",
|
| 25 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0001.jsonl.gz",
|
| 26 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0002.jsonl.gz",
|
| 27 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0003.jsonl.gz",
|
| 28 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0004.jsonl.gz",
|
| 29 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0005.jsonl.gz",
|
| 30 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0006.jsonl.gz",
|
| 31 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0007.jsonl.gz",
|
| 32 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0008.jsonl.gz",
|
| 33 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0009.jsonl.gz",
|
| 34 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0010.jsonl.gz",
|
| 35 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0011.jsonl.gz",
|
| 36 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0012.jsonl.gz",
|
| 37 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0013.jsonl.gz",
|
| 38 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0014.jsonl.gz",
|
| 39 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0015.jsonl.gz",
|
| 40 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0016.jsonl.gz",
|
| 41 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0017.jsonl.gz",
|
| 42 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0018.jsonl.gz",
|
| 43 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0019.jsonl.gz",
|
| 44 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0020.jsonl.gz",
|
| 45 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0021.jsonl.gz",
|
| 46 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0022.jsonl.gz",
|
| 47 |
+
"canonical/nvidia_open_swe_traces/v1.2--openhands--0023.jsonl.gz",
|
| 48 |
+
"canonical/official_tau3+toolbench/official_tau3_toolbench.jsonl.gz",
|
| 49 |
+
"canonical/swebench_multilingual_trajectories/default--Multi_SWE_agent_32B_Rust--0000.jsonl.gz",
|
| 50 |
+
"canonical/swebench_multilingual_trajectories/default--Qwen2.5_Coder_32B_Instruct--0001.jsonl.gz",
|
| 51 |
+
"canonical/swebench_multilingual_trajectories/default--SWE_agent_LM_32B--0002.jsonl.gz",
|
| 52 |
+
"canonical/swebench_verified_trajectories/all.jsonl.gz",
|
| 53 |
+
"canonical/swegym_openhands/default--train.success.oss--0000.jsonl.gz",
|
| 54 |
+
"metadata-only:swegym_sampled",
|
| 55 |
+
"metadata-only:swegym_sampled",
|
| 56 |
+
"metadata-only:swegym_sampled",
|
| 57 |
+
"canonical/swesmith_trajectories/default--tool--0000.jsonl.gz",
|
| 58 |
+
"canonical/swesmith_trajectories/default--tool--0001.jsonl.gz",
|
| 59 |
+
"canonical/swesmith_trajectories/default--tool--0002.jsonl.gz",
|
| 60 |
+
"canonical/swesmith_trajectories/default--tool--0003.jsonl.gz",
|
| 61 |
+
"canonical/swesmith_trajectories/default--tool--0004.jsonl.gz",
|
| 62 |
+
"canonical/swesmith_trajectories/default--tool--0005.jsonl.gz",
|
| 63 |
+
"canonical/swesmith_trajectories/default--tool--0006.jsonl.gz",
|
| 64 |
+
"canonical/swesmith_trajectories/default--tool--0007.jsonl.gz",
|
| 65 |
+
"canonical/terminalbench_trajectories/default--train--0000.jsonl.gz",
|
| 66 |
+
"canonical/terminalbench_trajectories/default--train--0001.jsonl.gz",
|
| 67 |
+
"canonical/terminaltraj/default--train--0000.jsonl.gz",
|
| 68 |
+
"canonical/terminaltraj/default--train--0001.jsonl.gz"
|
| 69 |
+
],
|
| 70 |
+
"model_imitation_status": "not-run",
|
| 71 |
+
"native_environment_replays": 0,
|
| 72 |
+
"platform": "Linux-6.18.44-x86_64-with-glibc2.41",
|
| 73 |
+
"python": "3.12.14 (main, Aug 25 2026, 14:00:49) [Clang 22.1.3 ]",
|
| 74 |
+
"registry": "registry/sources.json",
|
| 75 |
+
"runtime_seconds": 953.287,
|
| 76 |
+
"schema_version": 1
|
| 77 |
+
}
|
evaluation/summary.json
ADDED
|
@@ -0,0 +1,2901 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"by_agent_scaffold": {
|
| 3 |
+
"AgentSuite tau2 benchmark rollout": {
|
| 4 |
+
"cost_samples": 0,
|
| 5 |
+
"duration_samples": 0,
|
| 6 |
+
"failure": 4965,
|
| 7 |
+
"input_token_samples": 0,
|
| 8 |
+
"mean_actions": 17.801798561151077,
|
| 9 |
+
"mean_cost_source_native_units": null,
|
| 10 |
+
"mean_duration_seconds": null,
|
| 11 |
+
"mean_events": 45.433093525179856,
|
| 12 |
+
"mean_input_tokens": null,
|
| 13 |
+
"mean_output_tokens": null,
|
| 14 |
+
"mean_supervised_assistant_characters": 5750.385491606715,
|
| 15 |
+
"output_token_samples": 0,
|
| 16 |
+
"structurally_eligible": 8337,
|
| 17 |
+
"success": 3375,
|
| 18 |
+
"success_rate_known_outcomes": 0.40467625899280574,
|
| 19 |
+
"training_eligible": 8337,
|
| 20 |
+
"trajectories": 8340,
|
| 21 |
+
"unique_tasks": 228,
|
| 22 |
+
"unknown": 0
|
| 23 |
+
},
|
| 24 |
+
"Mind2Web human annotator": {
|
| 25 |
+
"cost_samples": 0,
|
| 26 |
+
"duration_samples": 0,
|
| 27 |
+
"failure": 0,
|
| 28 |
+
"input_token_samples": 0,
|
| 29 |
+
"mean_actions": 7.705649157581764,
|
| 30 |
+
"mean_cost_source_native_units": null,
|
| 31 |
+
"mean_duration_seconds": null,
|
| 32 |
+
"mean_events": 16.411298315163528,
|
| 33 |
+
"mean_input_tokens": null,
|
| 34 |
+
"mean_output_tokens": null,
|
| 35 |
+
"mean_supervised_assistant_characters": 256.4281466798811,
|
| 36 |
+
"output_token_samples": 0,
|
| 37 |
+
"structurally_eligible": 1009,
|
| 38 |
+
"success": 0,
|
| 39 |
+
"success_rate_known_outcomes": null,
|
| 40 |
+
"training_eligible": 1009,
|
| 41 |
+
"trajectories": 1009,
|
| 42 |
+
"unique_tasks": 1009,
|
| 43 |
+
"unknown": 1009
|
| 44 |
+
},
|
| 45 |
+
"NVIDIA Open-SWE trace generator": {
|
| 46 |
+
"cost_samples": 0,
|
| 47 |
+
"duration_samples": 0,
|
| 48 |
+
"failure": 26883,
|
| 49 |
+
"input_token_samples": 0,
|
| 50 |
+
"mean_actions": 58.35106341765777,
|
| 51 |
+
"mean_cost_source_native_units": null,
|
| 52 |
+
"mean_duration_seconds": null,
|
| 53 |
+
"mean_events": 124.11902917134778,
|
| 54 |
+
"mean_input_tokens": null,
|
| 55 |
+
"mean_output_tokens": null,
|
| 56 |
+
"mean_supervised_assistant_characters": 4650.015089296091,
|
| 57 |
+
"output_token_samples": 0,
|
| 58 |
+
"structurally_eligible": 51626,
|
| 59 |
+
"success": 24743,
|
| 60 |
+
"success_rate_known_outcomes": 0.4792740092201604,
|
| 61 |
+
"training_eligible": 51626,
|
| 62 |
+
"trajectories": 51626,
|
| 63 |
+
"unique_tasks": 33444,
|
| 64 |
+
"unknown": 0
|
| 65 |
+
},
|
| 66 |
+
"OpenHands": {
|
| 67 |
+
"cost_samples": 0,
|
| 68 |
+
"duration_samples": 0,
|
| 69 |
+
"failure": 5011,
|
| 70 |
+
"input_token_samples": 0,
|
| 71 |
+
"mean_actions": 20.382406397673574,
|
| 72 |
+
"mean_cost_source_native_units": null,
|
| 73 |
+
"mean_duration_seconds": null,
|
| 74 |
+
"mean_events": 42.287168302435475,
|
| 75 |
+
"mean_input_tokens": null,
|
| 76 |
+
"mean_output_tokens": null,
|
| 77 |
+
"mean_supervised_assistant_characters": 7133.47528171574,
|
| 78 |
+
"output_token_samples": 0,
|
| 79 |
+
"structurally_eligible": 5502,
|
| 80 |
+
"success": 491,
|
| 81 |
+
"success_rate_known_outcomes": 0.08924027626317703,
|
| 82 |
+
"training_eligible": 5502,
|
| 83 |
+
"trajectories": 5502,
|
| 84 |
+
"unique_tasks": 2438,
|
| 85 |
+
"unknown": 0
|
| 86 |
+
},
|
| 87 |
+
"SWE-agent": {
|
| 88 |
+
"cost_samples": 0,
|
| 89 |
+
"duration_samples": 0,
|
| 90 |
+
"failure": 14673,
|
| 91 |
+
"input_token_samples": 0,
|
| 92 |
+
"mean_actions": 31.585477178423236,
|
| 93 |
+
"mean_cost_source_native_units": null,
|
| 94 |
+
"mean_duration_seconds": null,
|
| 95 |
+
"mean_events": 64.17095435684648,
|
| 96 |
+
"mean_input_tokens": null,
|
| 97 |
+
"mean_output_tokens": null,
|
| 98 |
+
"mean_supervised_assistant_characters": 17664.133236514524,
|
| 99 |
+
"output_token_samples": 0,
|
| 100 |
+
"structurally_eligible": 24100,
|
| 101 |
+
"success": 9427,
|
| 102 |
+
"success_rate_known_outcomes": 0.39116182572614105,
|
| 103 |
+
"training_eligible": 24100,
|
| 104 |
+
"trajectories": 24100,
|
| 105 |
+
"unique_tasks": 10637,
|
| 106 |
+
"unknown": 0
|
| 107 |
+
},
|
| 108 |
+
"SWE-agent-compatible scaffold": {
|
| 109 |
+
"cost_samples": 0,
|
| 110 |
+
"duration_samples": 0,
|
| 111 |
+
"failure": 121,
|
| 112 |
+
"input_token_samples": 0,
|
| 113 |
+
"mean_actions": 47.37984496124031,
|
| 114 |
+
"mean_cost_source_native_units": null,
|
| 115 |
+
"mean_duration_seconds": null,
|
| 116 |
+
"mean_events": 96.75968992248062,
|
| 117 |
+
"mean_input_tokens": null,
|
| 118 |
+
"mean_output_tokens": null,
|
| 119 |
+
"mean_supervised_assistant_characters": 36004.64341085271,
|
| 120 |
+
"output_token_samples": 0,
|
| 121 |
+
"structurally_eligible": 110,
|
| 122 |
+
"success": 8,
|
| 123 |
+
"success_rate_known_outcomes": 0.06201550387596899,
|
| 124 |
+
"training_eligible": 0,
|
| 125 |
+
"trajectories": 129,
|
| 126 |
+
"unique_tasks": 43,
|
| 127 |
+
"unknown": 0
|
| 128 |
+
},
|
| 129 |
+
"ToolBench answer generation": {
|
| 130 |
+
"cost_samples": 0,
|
| 131 |
+
"duration_samples": 0,
|
| 132 |
+
"failure": 4,
|
| 133 |
+
"input_token_samples": 0,
|
| 134 |
+
"mean_actions": 3.8181818181818183,
|
| 135 |
+
"mean_cost_source_native_units": null,
|
| 136 |
+
"mean_duration_seconds": null,
|
| 137 |
+
"mean_events": 9.090909090909092,
|
| 138 |
+
"mean_input_tokens": null,
|
| 139 |
+
"mean_output_tokens": null,
|
| 140 |
+
"mean_supervised_assistant_characters": 834.3636363636364,
|
| 141 |
+
"output_token_samples": 0,
|
| 142 |
+
"structurally_eligible": 0,
|
| 143 |
+
"success": 7,
|
| 144 |
+
"success_rate_known_outcomes": 0.6363636363636364,
|
| 145 |
+
"training_eligible": 0,
|
| 146 |
+
"trajectories": 11,
|
| 147 |
+
"unique_tasks": 11,
|
| 148 |
+
"unknown": 0
|
| 149 |
+
},
|
| 150 |
+
"claude-code": {
|
| 151 |
+
"cost_samples": 2517,
|
| 152 |
+
"duration_samples": 2517,
|
| 153 |
+
"failure": 1418,
|
| 154 |
+
"input_token_samples": 1204,
|
| 155 |
+
"mean_actions": 54.15454906634883,
|
| 156 |
+
"mean_cost_source_native_units": 66.30842242871526,
|
| 157 |
+
"mean_duration_seconds": 545.8310086178694,
|
| 158 |
+
"mean_events": 89.69288835915772,
|
| 159 |
+
"mean_input_tokens": 2248965.17358804,
|
| 160 |
+
"mean_output_tokens": 2596.156146179402,
|
| 161 |
+
"mean_supervised_assistant_characters": 9204.064759634486,
|
| 162 |
+
"output_token_samples": 1204,
|
| 163 |
+
"structurally_eligible": 21,
|
| 164 |
+
"success": 1099,
|
| 165 |
+
"success_rate_known_outcomes": 0.43663090981326974,
|
| 166 |
+
"training_eligible": 0,
|
| 167 |
+
"trajectories": 2517,
|
| 168 |
+
"unique_tasks": 88,
|
| 169 |
+
"unknown": 0
|
| 170 |
+
},
|
| 171 |
+
"codex": {
|
| 172 |
+
"cost_samples": 1385,
|
| 173 |
+
"duration_samples": 1385,
|
| 174 |
+
"failure": 908,
|
| 175 |
+
"input_token_samples": 1383,
|
| 176 |
+
"mean_actions": 17.857761732851987,
|
| 177 |
+
"mean_cost_source_native_units": 9.600910829986692,
|
| 178 |
+
"mean_duration_seconds": 377.06245802180433,
|
| 179 |
+
"mean_events": 34.25270758122743,
|
| 180 |
+
"mean_input_tokens": 365547.114967462,
|
| 181 |
+
"mean_output_tokens": 9098.387563268258,
|
| 182 |
+
"mean_supervised_assistant_characters": 4321.786281588447,
|
| 183 |
+
"output_token_samples": 1383,
|
| 184 |
+
"structurally_eligible": 23,
|
| 185 |
+
"success": 477,
|
| 186 |
+
"success_rate_known_outcomes": 0.3444043321299639,
|
| 187 |
+
"training_eligible": 0,
|
| 188 |
+
"trajectories": 1385,
|
| 189 |
+
"unique_tasks": 89,
|
| 190 |
+
"unknown": 0
|
| 191 |
+
},
|
| 192 |
+
"deepagent-harbor": {
|
| 193 |
+
"cost_samples": 380,
|
| 194 |
+
"duration_samples": 380,
|
| 195 |
+
"failure": 98,
|
| 196 |
+
"input_token_samples": 0,
|
| 197 |
+
"mean_actions": 29.263157894736842,
|
| 198 |
+
"mean_cost_source_native_units": 0.0,
|
| 199 |
+
"mean_duration_seconds": 705.5991674925151,
|
| 200 |
+
"mean_events": 59.747368421052634,
|
| 201 |
+
"mean_input_tokens": null,
|
| 202 |
+
"mean_output_tokens": null,
|
| 203 |
+
"mean_supervised_assistant_characters": 4741.5447368421055,
|
| 204 |
+
"output_token_samples": 0,
|
| 205 |
+
"structurally_eligible": 337,
|
| 206 |
+
"success": 282,
|
| 207 |
+
"success_rate_known_outcomes": 0.7421052631578947,
|
| 208 |
+
"training_eligible": 0,
|
| 209 |
+
"trajectories": 380,
|
| 210 |
+
"unique_tasks": 89,
|
| 211 |
+
"unknown": 0
|
| 212 |
+
},
|
| 213 |
+
"gemini-cli": {
|
| 214 |
+
"cost_samples": 755,
|
| 215 |
+
"duration_samples": 755,
|
| 216 |
+
"failure": 604,
|
| 217 |
+
"input_token_samples": 755,
|
| 218 |
+
"mean_actions": 17.34834437086093,
|
| 219 |
+
"mean_cost_source_native_units": 27.581561065845143,
|
| 220 |
+
"mean_duration_seconds": 509.88918244822923,
|
| 221 |
+
"mean_events": 34.70728476821192,
|
| 222 |
+
"mean_input_tokens": 613653.8317880795,
|
| 223 |
+
"mean_output_tokens": 22646.129801324503,
|
| 224 |
+
"mean_supervised_assistant_characters": 2940.1867549668873,
|
| 225 |
+
"output_token_samples": 755,
|
| 226 |
+
"structurally_eligible": 568,
|
| 227 |
+
"success": 151,
|
| 228 |
+
"success_rate_known_outcomes": 0.2,
|
| 229 |
+
"training_eligible": 0,
|
| 230 |
+
"trajectories": 755,
|
| 231 |
+
"unique_tasks": 88,
|
| 232 |
+
"unknown": 0
|
| 233 |
+
},
|
| 234 |
+
"judy": {
|
| 235 |
+
"cost_samples": 157,
|
| 236 |
+
"duration_samples": 157,
|
| 237 |
+
"failure": 37,
|
| 238 |
+
"input_token_samples": 0,
|
| 239 |
+
"mean_actions": 32.611464968152866,
|
| 240 |
+
"mean_cost_source_native_units": 0.0,
|
| 241 |
+
"mean_duration_seconds": 792.0475469212623,
|
| 242 |
+
"mean_events": 65.35031847133757,
|
| 243 |
+
"mean_input_tokens": null,
|
| 244 |
+
"mean_output_tokens": null,
|
| 245 |
+
"mean_supervised_assistant_characters": 4949.356687898089,
|
| 246 |
+
"output_token_samples": 0,
|
| 247 |
+
"structurally_eligible": 14,
|
| 248 |
+
"success": 120,
|
| 249 |
+
"success_rate_known_outcomes": 0.7643312101910829,
|
| 250 |
+
"training_eligible": 0,
|
| 251 |
+
"trajectories": 157,
|
| 252 |
+
"unique_tasks": 83,
|
| 253 |
+
"unknown": 0
|
| 254 |
+
},
|
| 255 |
+
"mini-swe-agent": {
|
| 256 |
+
"cost_samples": 6437,
|
| 257 |
+
"duration_samples": 6437,
|
| 258 |
+
"failure": 4936,
|
| 259 |
+
"input_token_samples": 6421,
|
| 260 |
+
"mean_actions": 21.314276837035887,
|
| 261 |
+
"mean_cost_source_native_units": 29.442191969641353,
|
| 262 |
+
"mean_duration_seconds": 491.84549342628515,
|
| 263 |
+
"mean_events": 43.85194966599347,
|
| 264 |
+
"mean_input_tokens": 743809.1583865442,
|
| 265 |
+
"mean_output_tokens": 19476.359601308206,
|
| 266 |
+
"mean_supervised_assistant_characters": 11188.48003728445,
|
| 267 |
+
"output_token_samples": 6421,
|
| 268 |
+
"structurally_eligible": 6339,
|
| 269 |
+
"success": 1501,
|
| 270 |
+
"success_rate_known_outcomes": 0.23318315985707627,
|
| 271 |
+
"training_eligible": 0,
|
| 272 |
+
"trajectories": 6437,
|
| 273 |
+
"unique_tasks": 89,
|
| 274 |
+
"unknown": 0
|
| 275 |
+
},
|
| 276 |
+
"minisweagent.run.benchmarks.swebench.ProgressTrackingAgent": {
|
| 277 |
+
"cost_samples": 0,
|
| 278 |
+
"duration_samples": 0,
|
| 279 |
+
"failure": 1152,
|
| 280 |
+
"input_token_samples": 0,
|
| 281 |
+
"mean_actions": 47.387984154929576,
|
| 282 |
+
"mean_cost_source_native_units": null,
|
| 283 |
+
"mean_duration_seconds": null,
|
| 284 |
+
"mean_events": 97.00748239436619,
|
| 285 |
+
"mean_input_tokens": null,
|
| 286 |
+
"mean_output_tokens": null,
|
| 287 |
+
"mean_supervised_assistant_characters": 12330.321082746479,
|
| 288 |
+
"output_token_samples": 0,
|
| 289 |
+
"structurally_eligible": 4532,
|
| 290 |
+
"success": 3392,
|
| 291 |
+
"success_rate_known_outcomes": 0.7464788732394366,
|
| 292 |
+
"training_eligible": 0,
|
| 293 |
+
"trajectories": 4544,
|
| 294 |
+
"unique_tasks": 500,
|
| 295 |
+
"unknown": 0
|
| 296 |
+
},
|
| 297 |
+
"openhands": {
|
| 298 |
+
"cost_samples": 6033,
|
| 299 |
+
"duration_samples": 6033,
|
| 300 |
+
"failure": 4296,
|
| 301 |
+
"input_token_samples": 5593,
|
| 302 |
+
"mean_actions": 82.31791811702304,
|
| 303 |
+
"mean_cost_source_native_units": 161.66761875649664,
|
| 304 |
+
"mean_duration_seconds": 885.4345594482321,
|
| 305 |
+
"mean_events": 144.69766285430134,
|
| 306 |
+
"mean_input_tokens": 1835762.452887538,
|
| 307 |
+
"mean_output_tokens": 27097.81995351332,
|
| 308 |
+
"mean_supervised_assistant_characters": 18699.04243328361,
|
| 309 |
+
"output_token_samples": 5593,
|
| 310 |
+
"structurally_eligible": 284,
|
| 311 |
+
"success": 1737,
|
| 312 |
+
"success_rate_known_outcomes": 0.28791645947289907,
|
| 313 |
+
"training_eligible": 0,
|
| 314 |
+
"trajectories": 6033,
|
| 315 |
+
"unique_tasks": 89,
|
| 316 |
+
"unknown": 0
|
| 317 |
+
},
|
| 318 |
+
"rl-rag BrowseComp research scaffold": {
|
| 319 |
+
"cost_samples": 0,
|
| 320 |
+
"duration_samples": 5063,
|
| 321 |
+
"failure": 3852,
|
| 322 |
+
"input_token_samples": 0,
|
| 323 |
+
"mean_actions": 27.13786292711831,
|
| 324 |
+
"mean_cost_source_native_units": null,
|
| 325 |
+
"mean_duration_seconds": 108.07624142800492,
|
| 326 |
+
"mean_events": 54.28303377444203,
|
| 327 |
+
"mean_input_tokens": null,
|
| 328 |
+
"mean_output_tokens": null,
|
| 329 |
+
"mean_supervised_assistant_characters": 15385.260912502468,
|
| 330 |
+
"output_token_samples": 0,
|
| 331 |
+
"structurally_eligible": 5063,
|
| 332 |
+
"success": 1211,
|
| 333 |
+
"success_rate_known_outcomes": 0.23918625320955955,
|
| 334 |
+
"training_eligible": 0,
|
| 335 |
+
"trajectories": 5063,
|
| 336 |
+
"unique_tasks": 1266,
|
| 337 |
+
"unknown": 0
|
| 338 |
+
},
|
| 339 |
+
"tau benchmark agent and user simulator": {
|
| 340 |
+
"cost_samples": 0,
|
| 341 |
+
"duration_samples": 0,
|
| 342 |
+
"failure": 100,
|
| 343 |
+
"input_token_samples": 0,
|
| 344 |
+
"mean_actions": 14.5,
|
| 345 |
+
"mean_cost_source_native_units": null,
|
| 346 |
+
"mean_duration_seconds": null,
|
| 347 |
+
"mean_events": 30.0,
|
| 348 |
+
"mean_input_tokens": null,
|
| 349 |
+
"mean_output_tokens": null,
|
| 350 |
+
"mean_supervised_assistant_characters": 4221.82,
|
| 351 |
+
"output_token_samples": 0,
|
| 352 |
+
"structurally_eligible": 0,
|
| 353 |
+
"success": 100,
|
| 354 |
+
"success_rate_known_outcomes": 0.5,
|
| 355 |
+
"training_eligible": 0,
|
| 356 |
+
"trajectories": 200,
|
| 357 |
+
"unique_tasks": 50,
|
| 358 |
+
"unknown": 0
|
| 359 |
+
},
|
| 360 |
+
"terminus-2": {
|
| 361 |
+
"cost_samples": 15921,
|
| 362 |
+
"duration_samples": 15921,
|
| 363 |
+
"failure": 10743,
|
| 364 |
+
"input_token_samples": 14581,
|
| 365 |
+
"mean_actions": 25.161610451604798,
|
| 366 |
+
"mean_cost_source_native_units": 42.898824807674124,
|
| 367 |
+
"mean_duration_seconds": 743.9324663575665,
|
| 368 |
+
"mean_events": 52.535456315558065,
|
| 369 |
+
"mean_input_tokens": 612642.0770866196,
|
| 370 |
+
"mean_output_tokens": 16685.474110143336,
|
| 371 |
+
"mean_supervised_assistant_characters": 17296.68670309654,
|
| 372 |
+
"output_token_samples": 14581,
|
| 373 |
+
"structurally_eligible": 15822,
|
| 374 |
+
"success": 5178,
|
| 375 |
+
"success_rate_known_outcomes": 0.32523082720934615,
|
| 376 |
+
"training_eligible": 0,
|
| 377 |
+
"trajectories": 15921,
|
| 378 |
+
"unique_tasks": 89,
|
| 379 |
+
"unknown": 0
|
| 380 |
+
},
|
| 381 |
+
"terminus-3-3": {
|
| 382 |
+
"cost_samples": 678,
|
| 383 |
+
"duration_samples": 678,
|
| 384 |
+
"failure": 172,
|
| 385 |
+
"input_token_samples": 678,
|
| 386 |
+
"mean_actions": 29.125368731563423,
|
| 387 |
+
"mean_cost_source_native_units": 74.12365559350073,
|
| 388 |
+
"mean_duration_seconds": 664.1559937288628,
|
| 389 |
+
"mean_events": 59.250737463126846,
|
| 390 |
+
"mean_input_tokens": 990659.5044247787,
|
| 391 |
+
"mean_output_tokens": 25339.938053097347,
|
| 392 |
+
"mean_supervised_assistant_characters": 5243.02802359882,
|
| 393 |
+
"output_token_samples": 678,
|
| 394 |
+
"structurally_eligible": 678,
|
| 395 |
+
"success": 506,
|
| 396 |
+
"success_rate_known_outcomes": 0.7463126843657817,
|
| 397 |
+
"training_eligible": 0,
|
| 398 |
+
"trajectories": 678,
|
| 399 |
+
"unique_tasks": 89,
|
| 400 |
+
"unknown": 0
|
| 401 |
+
},
|
| 402 |
+
"unknown": {
|
| 403 |
+
"cost_samples": 0,
|
| 404 |
+
"duration_samples": 0,
|
| 405 |
+
"failure": 0,
|
| 406 |
+
"input_token_samples": 0,
|
| 407 |
+
"mean_actions": 16.883314625933338,
|
| 408 |
+
"mean_cost_source_native_units": null,
|
| 409 |
+
"mean_duration_seconds": null,
|
| 410 |
+
"mean_events": 33.79059099116685,
|
| 411 |
+
"mean_input_tokens": null,
|
| 412 |
+
"mean_output_tokens": null,
|
| 413 |
+
"mean_supervised_assistant_characters": 10559.201747108487,
|
| 414 |
+
"output_token_samples": 0,
|
| 415 |
+
"structurally_eligible": 19634,
|
| 416 |
+
"success": 491,
|
| 417 |
+
"success_rate_known_outcomes": 1.0,
|
| 418 |
+
"training_eligible": 19634,
|
| 419 |
+
"trajectories": 20491,
|
| 420 |
+
"unique_tasks": 18080,
|
| 421 |
+
"unknown": 20000
|
| 422 |
+
}
|
| 423 |
+
},
|
| 424 |
+
"by_model": {
|
| 425 |
+
"ChatGPT (repository-described; record has no exact version)": {
|
| 426 |
+
"cost_samples": 0,
|
| 427 |
+
"duration_samples": 0,
|
| 428 |
+
"failure": 4,
|
| 429 |
+
"input_token_samples": 0,
|
| 430 |
+
"mean_actions": 3.8181818181818183,
|
| 431 |
+
"mean_cost_source_native_units": null,
|
| 432 |
+
"mean_duration_seconds": null,
|
| 433 |
+
"mean_events": 9.090909090909092,
|
| 434 |
+
"mean_input_tokens": null,
|
| 435 |
+
"mean_output_tokens": null,
|
| 436 |
+
"mean_supervised_assistant_characters": 834.3636363636364,
|
| 437 |
+
"output_token_samples": 0,
|
| 438 |
+
"structurally_eligible": 0,
|
| 439 |
+
"success": 7,
|
| 440 |
+
"success_rate_known_outcomes": 0.6363636363636364,
|
| 441 |
+
"training_eligible": 0,
|
| 442 |
+
"trajectories": 11,
|
| 443 |
+
"unique_tasks": 11,
|
| 444 |
+
"unknown": 0
|
| 445 |
+
},
|
| 446 |
+
"Multi-SWE-agent-32B-Rust": {
|
| 447 |
+
"cost_samples": 0,
|
| 448 |
+
"duration_samples": 0,
|
| 449 |
+
"failure": 42,
|
| 450 |
+
"input_token_samples": 0,
|
| 451 |
+
"mean_actions": 70.65116279069767,
|
| 452 |
+
"mean_cost_source_native_units": null,
|
| 453 |
+
"mean_duration_seconds": null,
|
| 454 |
+
"mean_events": 143.30232558139534,
|
| 455 |
+
"mean_input_tokens": null,
|
| 456 |
+
"mean_output_tokens": null,
|
| 457 |
+
"mean_supervised_assistant_characters": 26940.906976744187,
|
| 458 |
+
"output_token_samples": 0,
|
| 459 |
+
"structurally_eligible": 41,
|
| 460 |
+
"success": 1,
|
| 461 |
+
"success_rate_known_outcomes": 0.023255813953488372,
|
| 462 |
+
"training_eligible": 0,
|
| 463 |
+
"trajectories": 43,
|
| 464 |
+
"unique_tasks": 43,
|
| 465 |
+
"unknown": 0
|
| 466 |
+
},
|
| 467 |
+
"Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8@together_ai": {
|
| 468 |
+
"cost_samples": 1798,
|
| 469 |
+
"duration_samples": 1798,
|
| 470 |
+
"failure": 1348,
|
| 471 |
+
"input_token_samples": 1794,
|
| 472 |
+
"mean_actions": 72.29365962180201,
|
| 473 |
+
"mean_cost_source_native_units": 255.6923547450772,
|
| 474 |
+
"mean_duration_seconds": 648.0904410365426,
|
| 475 |
+
"mean_events": 119.13681868743048,
|
| 476 |
+
"mean_input_tokens": 1270502.5462653288,
|
| 477 |
+
"mean_output_tokens": 10809.755852842809,
|
| 478 |
+
"mean_supervised_assistant_characters": 27958.512791991103,
|
| 479 |
+
"output_token_samples": 1794,
|
| 480 |
+
"structurally_eligible": 928,
|
| 481 |
+
"success": 450,
|
| 482 |
+
"success_rate_known_outcomes": 0.25027808676307006,
|
| 483 |
+
"training_eligible": 0,
|
| 484 |
+
"trajectories": 1798,
|
| 485 |
+
"unique_tasks": 89,
|
| 486 |
+
"unknown": 0
|
| 487 |
+
},
|
| 488 |
+
"Qwen2.5-Coder-32B-Instruct": {
|
| 489 |
+
"cost_samples": 0,
|
| 490 |
+
"duration_samples": 0,
|
| 491 |
+
"failure": 43,
|
| 492 |
+
"input_token_samples": 0,
|
| 493 |
+
"mean_actions": 29.74418604651163,
|
| 494 |
+
"mean_cost_source_native_units": null,
|
| 495 |
+
"mean_duration_seconds": null,
|
| 496 |
+
"mean_events": 61.48837209302326,
|
| 497 |
+
"mean_input_tokens": null,
|
| 498 |
+
"mean_output_tokens": null,
|
| 499 |
+
"mean_supervised_assistant_characters": 57204.72093023256,
|
| 500 |
+
"output_token_samples": 0,
|
| 501 |
+
"structurally_eligible": 27,
|
| 502 |
+
"success": 0,
|
| 503 |
+
"success_rate_known_outcomes": 0.0,
|
| 504 |
+
"training_eligible": 0,
|
| 505 |
+
"trajectories": 43,
|
| 506 |
+
"unique_tasks": 43,
|
| 507 |
+
"unknown": 0
|
| 508 |
+
},
|
| 509 |
+
"Qwen3.8-27B": {
|
| 510 |
+
"cost_samples": 0,
|
| 511 |
+
"duration_samples": 0,
|
| 512 |
+
"failure": 13666,
|
| 513 |
+
"input_token_samples": 0,
|
| 514 |
+
"mean_actions": 62.70772295792804,
|
| 515 |
+
"mean_cost_source_native_units": null,
|
| 516 |
+
"mean_duration_seconds": null,
|
| 517 |
+
"mean_events": 134.19887495605298,
|
| 518 |
+
"mean_input_tokens": null,
|
| 519 |
+
"mean_output_tokens": null,
|
| 520 |
+
"mean_supervised_assistant_characters": 6554.125161139107,
|
| 521 |
+
"output_token_samples": 0,
|
| 522 |
+
"structurally_eligible": 25599,
|
| 523 |
+
"success": 11933,
|
| 524 |
+
"success_rate_known_outcomes": 0.4661510215242783,
|
| 525 |
+
"training_eligible": 25599,
|
| 526 |
+
"trajectories": 25599,
|
| 527 |
+
"unique_tasks": 25599,
|
| 528 |
+
"unknown": 0
|
| 529 |
+
},
|
| 530 |
+
"Qwen3.8-Flash-Next": {
|
| 531 |
+
"cost_samples": 0,
|
| 532 |
+
"duration_samples": 0,
|
| 533 |
+
"failure": 13217,
|
| 534 |
+
"input_token_samples": 0,
|
| 535 |
+
"mean_actions": 54.06604679755638,
|
| 536 |
+
"mean_cost_source_native_units": null,
|
| 537 |
+
"mean_duration_seconds": null,
|
| 538 |
+
"mean_events": 114.20494102278403,
|
| 539 |
+
"mean_input_tokens": null,
|
| 540 |
+
"mean_output_tokens": null,
|
| 541 |
+
"mean_supervised_assistant_characters": 2777.217082260729,
|
| 542 |
+
"output_token_samples": 0,
|
| 543 |
+
"structurally_eligible": 26027,
|
| 544 |
+
"success": 12810,
|
| 545 |
+
"success_rate_known_outcomes": 0.49218119644984054,
|
| 546 |
+
"training_eligible": 26027,
|
| 547 |
+
"trajectories": 26027,
|
| 548 |
+
"unique_tasks": 26027,
|
| 549 |
+
"unknown": 0
|
| 550 |
+
},
|
| 551 |
+
"SWE-agent-LM-32B": {
|
| 552 |
+
"cost_samples": 0,
|
| 553 |
+
"duration_samples": 0,
|
| 554 |
+
"failure": 36,
|
| 555 |
+
"input_token_samples": 0,
|
| 556 |
+
"mean_actions": 41.74418604651163,
|
| 557 |
+
"mean_cost_source_native_units": null,
|
| 558 |
+
"mean_duration_seconds": null,
|
| 559 |
+
"mean_events": 85.48837209302326,
|
| 560 |
+
"mean_input_tokens": null,
|
| 561 |
+
"mean_output_tokens": null,
|
| 562 |
+
"mean_supervised_assistant_characters": 23868.302325581397,
|
| 563 |
+
"output_token_samples": 0,
|
| 564 |
+
"structurally_eligible": 42,
|
| 565 |
+
"success": 7,
|
| 566 |
+
"success_rate_known_outcomes": 0.16279069767441862,
|
| 567 |
+
"training_eligible": 0,
|
| 568 |
+
"trajectories": 43,
|
| 569 |
+
"unique_tasks": 43,
|
| 570 |
+
"unknown": 0
|
| 571 |
+
},
|
| 572 |
+
"accounts/fireworks/models/glm-4p6@fireworks_ai": {
|
| 573 |
+
"cost_samples": 904,
|
| 574 |
+
"duration_samples": 904,
|
| 575 |
+
"failure": 680,
|
| 576 |
+
"input_token_samples": 904,
|
| 577 |
+
"mean_actions": 41.66814159292036,
|
| 578 |
+
"mean_cost_source_native_units": 11.691257349520395,
|
| 579 |
+
"mean_duration_seconds": 476.20702301295455,
|
| 580 |
+
"mean_events": 85.45575221238938,
|
| 581 |
+
"mean_input_tokens": 1122957.5884955751,
|
| 582 |
+
"mean_output_tokens": 11417.679203539823,
|
| 583 |
+
"mean_supervised_assistant_characters": 30965.702433628318,
|
| 584 |
+
"output_token_samples": 904,
|
| 585 |
+
"structurally_eligible": 904,
|
| 586 |
+
"success": 224,
|
| 587 |
+
"success_rate_known_outcomes": 0.24778761061946902,
|
| 588 |
+
"training_eligible": 0,
|
| 589 |
+
"trajectories": 904,
|
| 590 |
+
"unique_tasks": 89,
|
| 591 |
+
"unknown": 0
|
| 592 |
+
},
|
| 593 |
+
"accounts/fireworks/models/minimax-m2@fireworks_ai": {
|
| 594 |
+
"cost_samples": 898,
|
| 595 |
+
"duration_samples": 898,
|
| 596 |
+
"failure": 622,
|
| 597 |
+
"input_token_samples": 894,
|
| 598 |
+
"mean_actions": 35.66369710467706,
|
| 599 |
+
"mean_cost_source_native_units": 32.10366033558591,
|
| 600 |
+
"mean_duration_seconds": 931.9153214180656,
|
| 601 |
+
"mean_events": 75.79955456570156,
|
| 602 |
+
"mean_input_tokens": 1006714.9149888143,
|
| 603 |
+
"mean_output_tokens": 17048.77852348993,
|
| 604 |
+
"mean_supervised_assistant_characters": 24994.038975501113,
|
| 605 |
+
"output_token_samples": 894,
|
| 606 |
+
"structurally_eligible": 888,
|
| 607 |
+
"success": 276,
|
| 608 |
+
"success_rate_known_outcomes": 0.30734966592427615,
|
| 609 |
+
"training_eligible": 0,
|
| 610 |
+
"trajectories": 898,
|
| 611 |
+
"unique_tasks": 89,
|
| 612 |
+
"unknown": 0
|
| 613 |
+
},
|
| 614 |
+
"anthropic/claude-4-opus-thinking-off": {
|
| 615 |
+
"cost_samples": 0,
|
| 616 |
+
"duration_samples": 0,
|
| 617 |
+
"failure": 134,
|
| 618 |
+
"input_token_samples": 0,
|
| 619 |
+
"mean_actions": 15.327338129496402,
|
| 620 |
+
"mean_cost_source_native_units": null,
|
| 621 |
+
"mean_duration_seconds": null,
|
| 622 |
+
"mean_events": 41.37410071942446,
|
| 623 |
+
"mean_input_tokens": null,
|
| 624 |
+
"mean_output_tokens": null,
|
| 625 |
+
"mean_supervised_assistant_characters": 5759.190647482014,
|
| 626 |
+
"output_token_samples": 0,
|
| 627 |
+
"structurally_eligible": 278,
|
| 628 |
+
"success": 144,
|
| 629 |
+
"success_rate_known_outcomes": 0.5179856115107914,
|
| 630 |
+
"training_eligible": 278,
|
| 631 |
+
"trajectories": 278,
|
| 632 |
+
"unique_tasks": 228,
|
| 633 |
+
"unknown": 0
|
| 634 |
+
},
|
| 635 |
+
"anthropic/claude-4-opus-thinking-on-10k": {
|
| 636 |
+
"cost_samples": 0,
|
| 637 |
+
"duration_samples": 0,
|
| 638 |
+
"failure": 135,
|
| 639 |
+
"input_token_samples": 0,
|
| 640 |
+
"mean_actions": 14.964028776978417,
|
| 641 |
+
"mean_cost_source_native_units": null,
|
| 642 |
+
"mean_duration_seconds": null,
|
| 643 |
+
"mean_events": 40.18345323741007,
|
| 644 |
+
"mean_input_tokens": null,
|
| 645 |
+
"mean_output_tokens": null,
|
| 646 |
+
"mean_supervised_assistant_characters": 5505.651079136691,
|
| 647 |
+
"output_token_samples": 0,
|
| 648 |
+
"structurally_eligible": 278,
|
| 649 |
+
"success": 143,
|
| 650 |
+
"success_rate_known_outcomes": 0.5143884892086331,
|
| 651 |
+
"training_eligible": 278,
|
| 652 |
+
"trajectories": 278,
|
| 653 |
+
"unique_tasks": 228,
|
| 654 |
+
"unknown": 0
|
| 655 |
+
},
|
| 656 |
+
"anthropic/claude-4-sonnet-thinking-off": {
|
| 657 |
+
"cost_samples": 0,
|
| 658 |
+
"duration_samples": 0,
|
| 659 |
+
"failure": 150,
|
| 660 |
+
"input_token_samples": 0,
|
| 661 |
+
"mean_actions": 14.158273381294965,
|
| 662 |
+
"mean_cost_source_native_units": null,
|
| 663 |
+
"mean_duration_seconds": null,
|
| 664 |
+
"mean_events": 36.330935251798564,
|
| 665 |
+
"mean_input_tokens": null,
|
| 666 |
+
"mean_output_tokens": null,
|
| 667 |
+
"mean_supervised_assistant_characters": 5403.838129496403,
|
| 668 |
+
"output_token_samples": 0,
|
| 669 |
+
"structurally_eligible": 278,
|
| 670 |
+
"success": 128,
|
| 671 |
+
"success_rate_known_outcomes": 0.460431654676259,
|
| 672 |
+
"training_eligible": 278,
|
| 673 |
+
"trajectories": 278,
|
| 674 |
+
"unique_tasks": 228,
|
| 675 |
+
"unknown": 0
|
| 676 |
+
},
|
| 677 |
+
"anthropic/claude-4-sonnet-thinking-on-10k": {
|
| 678 |
+
"cost_samples": 0,
|
| 679 |
+
"duration_samples": 0,
|
| 680 |
+
"failure": 133,
|
| 681 |
+
"input_token_samples": 0,
|
| 682 |
+
"mean_actions": 13.611510791366907,
|
| 683 |
+
"mean_cost_source_native_units": null,
|
| 684 |
+
"mean_duration_seconds": null,
|
| 685 |
+
"mean_events": 34.989208633093526,
|
| 686 |
+
"mean_input_tokens": null,
|
| 687 |
+
"mean_output_tokens": null,
|
| 688 |
+
"mean_supervised_assistant_characters": 5727.251798561151,
|
| 689 |
+
"output_token_samples": 0,
|
| 690 |
+
"structurally_eligible": 278,
|
| 691 |
+
"success": 145,
|
| 692 |
+
"success_rate_known_outcomes": 0.5215827338129496,
|
| 693 |
+
"training_eligible": 278,
|
| 694 |
+
"trajectories": 278,
|
| 695 |
+
"unique_tasks": 228,
|
| 696 |
+
"unknown": 0
|
| 697 |
+
},
|
| 698 |
+
"anthropic/claude-4.5-sonnet-thinking-off": {
|
| 699 |
+
"cost_samples": 0,
|
| 700 |
+
"duration_samples": 0,
|
| 701 |
+
"failure": 141,
|
| 702 |
+
"input_token_samples": 0,
|
| 703 |
+
"mean_actions": 15.31294964028777,
|
| 704 |
+
"mean_cost_source_native_units": null,
|
| 705 |
+
"mean_duration_seconds": null,
|
| 706 |
+
"mean_events": 42.37410071942446,
|
| 707 |
+
"mean_input_tokens": null,
|
| 708 |
+
"mean_output_tokens": null,
|
| 709 |
+
"mean_supervised_assistant_characters": 5430.76618705036,
|
| 710 |
+
"output_token_samples": 0,
|
| 711 |
+
"structurally_eligible": 278,
|
| 712 |
+
"success": 137,
|
| 713 |
+
"success_rate_known_outcomes": 0.49280575539568344,
|
| 714 |
+
"training_eligible": 278,
|
| 715 |
+
"trajectories": 278,
|
| 716 |
+
"unique_tasks": 228,
|
| 717 |
+
"unknown": 0
|
| 718 |
+
},
|
| 719 |
+
"anthropic/claude-4.5-sonnet-thinking-on-10k": {
|
| 720 |
+
"cost_samples": 0,
|
| 721 |
+
"duration_samples": 0,
|
| 722 |
+
"failure": 126,
|
| 723 |
+
"input_token_samples": 0,
|
| 724 |
+
"mean_actions": 13.507194244604317,
|
| 725 |
+
"mean_cost_source_native_units": null,
|
| 726 |
+
"mean_duration_seconds": null,
|
| 727 |
+
"mean_events": 39.55755395683453,
|
| 728 |
+
"mean_input_tokens": null,
|
| 729 |
+
"mean_output_tokens": null,
|
| 730 |
+
"mean_supervised_assistant_characters": 5573.884892086331,
|
| 731 |
+
"output_token_samples": 0,
|
| 732 |
+
"structurally_eligible": 278,
|
| 733 |
+
"success": 152,
|
| 734 |
+
"success_rate_known_outcomes": 0.5467625899280576,
|
| 735 |
+
"training_eligible": 278,
|
| 736 |
+
"trajectories": 278,
|
| 737 |
+
"unique_tasks": 228,
|
| 738 |
+
"unknown": 0
|
| 739 |
+
},
|
| 740 |
+
"claude-3-5-sonnet-20241022": {
|
| 741 |
+
"cost_samples": 0,
|
| 742 |
+
"duration_samples": 0,
|
| 743 |
+
"failure": 3567,
|
| 744 |
+
"input_token_samples": 0,
|
| 745 |
+
"mean_actions": 21.48913043478261,
|
| 746 |
+
"mean_cost_source_native_units": null,
|
| 747 |
+
"mean_duration_seconds": null,
|
| 748 |
+
"mean_events": 43.97826086956522,
|
| 749 |
+
"mean_input_tokens": null,
|
| 750 |
+
"mean_output_tokens": null,
|
| 751 |
+
"mean_supervised_assistant_characters": 8249.628260869566,
|
| 752 |
+
"output_token_samples": 0,
|
| 753 |
+
"structurally_eligible": 5980,
|
| 754 |
+
"success": 2413,
|
| 755 |
+
"success_rate_known_outcomes": 0.4035117056856187,
|
| 756 |
+
"training_eligible": 5980,
|
| 757 |
+
"trajectories": 5980,
|
| 758 |
+
"unique_tasks": 2194,
|
| 759 |
+
"unknown": 0
|
| 760 |
+
},
|
| 761 |
+
"claude-3-7-sonnet-20250219": {
|
| 762 |
+
"cost_samples": 0,
|
| 763 |
+
"duration_samples": 0,
|
| 764 |
+
"failure": 10909,
|
| 765 |
+
"input_token_samples": 0,
|
| 766 |
+
"mean_actions": 34.72726765280491,
|
| 767 |
+
"mean_cost_source_native_units": null,
|
| 768 |
+
"mean_duration_seconds": null,
|
| 769 |
+
"mean_events": 70.45453530560982,
|
| 770 |
+
"mean_input_tokens": null,
|
| 771 |
+
"mean_output_tokens": null,
|
| 772 |
+
"mean_supervised_assistant_characters": 20883.35668434273,
|
| 773 |
+
"output_token_samples": 0,
|
| 774 |
+
"structurally_eligible": 17715,
|
| 775 |
+
"success": 7006,
|
| 776 |
+
"success_rate_known_outcomes": 0.3910689366452693,
|
| 777 |
+
"training_eligible": 17715,
|
| 778 |
+
"trajectories": 17915,
|
| 779 |
+
"unique_tasks": 9193,
|
| 780 |
+
"unknown": 0
|
| 781 |
+
},
|
| 782 |
+
"claude-4.5-haiku-high": {
|
| 783 |
+
"cost_samples": 0,
|
| 784 |
+
"duration_samples": 0,
|
| 785 |
+
"failure": 167,
|
| 786 |
+
"input_token_samples": 0,
|
| 787 |
+
"mean_actions": 66.118,
|
| 788 |
+
"mean_cost_source_native_units": null,
|
| 789 |
+
"mean_duration_seconds": null,
|
| 790 |
+
"mean_events": 133.272,
|
| 791 |
+
"mean_input_tokens": null,
|
| 792 |
+
"mean_output_tokens": null,
|
| 793 |
+
"mean_supervised_assistant_characters": 21335.944,
|
| 794 |
+
"output_token_samples": 0,
|
| 795 |
+
"structurally_eligible": 500,
|
| 796 |
+
"success": 333,
|
| 797 |
+
"success_rate_known_outcomes": 0.666,
|
| 798 |
+
"training_eligible": 0,
|
| 799 |
+
"trajectories": 500,
|
| 800 |
+
"unique_tasks": 500,
|
| 801 |
+
"unknown": 0
|
| 802 |
+
},
|
| 803 |
+
"claude-4.5-opus-high": {
|
| 804 |
+
"cost_samples": 0,
|
| 805 |
+
"duration_samples": 0,
|
| 806 |
+
"failure": 116,
|
| 807 |
+
"input_token_samples": 0,
|
| 808 |
+
"mean_actions": 32.892,
|
| 809 |
+
"mean_cost_source_native_units": null,
|
| 810 |
+
"mean_duration_seconds": null,
|
| 811 |
+
"mean_events": 72.41,
|
| 812 |
+
"mean_input_tokens": null,
|
| 813 |
+
"mean_output_tokens": null,
|
| 814 |
+
"mean_supervised_assistant_characters": 11651.378,
|
| 815 |
+
"output_token_samples": 0,
|
| 816 |
+
"structurally_eligible": 500,
|
| 817 |
+
"success": 384,
|
| 818 |
+
"success_rate_known_outcomes": 0.768,
|
| 819 |
+
"training_eligible": 0,
|
| 820 |
+
"trajectories": 500,
|
| 821 |
+
"unique_tasks": 500,
|
| 822 |
+
"unknown": 0
|
| 823 |
+
},
|
| 824 |
+
"claude-haiku-4-5-20251001@anthropic": {
|
| 825 |
+
"cost_samples": 1656,
|
| 826 |
+
"duration_samples": 1656,
|
| 827 |
+
"failure": 1230,
|
| 828 |
+
"input_token_samples": 1655,
|
| 829 |
+
"mean_actions": 80.83393719806763,
|
| 830 |
+
"mean_cost_source_native_units": 207.72488678684053,
|
| 831 |
+
"mean_duration_seconds": 743.0986861180568,
|
| 832 |
+
"mean_events": 202.96497584541063,
|
| 833 |
+
"mean_input_tokens": 3116007.3438066463,
|
| 834 |
+
"mean_output_tokens": 18089.96193353474,
|
| 835 |
+
"mean_supervised_assistant_characters": 18744.347222222223,
|
| 836 |
+
"output_token_samples": 1655,
|
| 837 |
+
"structurally_eligible": 965,
|
| 838 |
+
"success": 426,
|
| 839 |
+
"success_rate_known_outcomes": 0.2572463768115942,
|
| 840 |
+
"training_eligible": 0,
|
| 841 |
+
"trajectories": 1656,
|
| 842 |
+
"unique_tasks": 89,
|
| 843 |
+
"unknown": 0
|
| 844 |
+
},
|
| 845 |
+
"claude-opus-4-1-20250805@anthropic": {
|
| 846 |
+
"cost_samples": 1712,
|
| 847 |
+
"duration_samples": 1712,
|
| 848 |
+
"failure": 1072,
|
| 849 |
+
"input_token_samples": 1710,
|
| 850 |
+
"mean_actions": 47.4696261682243,
|
| 851 |
+
"mean_cost_source_native_units": 255.39316846193555,
|
| 852 |
+
"mean_duration_seconds": 620.2949460087535,
|
| 853 |
+
"mean_events": 80.42815420560747,
|
| 854 |
+
"mean_input_tokens": 1029506.9678362573,
|
| 855 |
+
"mean_output_tokens": 9701.605847953217,
|
| 856 |
+
"mean_supervised_assistant_characters": 12596.474299065421,
|
| 857 |
+
"output_token_samples": 1710,
|
| 858 |
+
"structurally_eligible": 872,
|
| 859 |
+
"success": 640,
|
| 860 |
+
"success_rate_known_outcomes": 0.37383177570093457,
|
| 861 |
+
"training_eligible": 0,
|
| 862 |
+
"trajectories": 1712,
|
| 863 |
+
"unique_tasks": 89,
|
| 864 |
+
"unknown": 0
|
| 865 |
+
},
|
| 866 |
+
"claude-opus-4-5-20251101@anthropic": {
|
| 867 |
+
"cost_samples": 1307,
|
| 868 |
+
"duration_samples": 1307,
|
| 869 |
+
"failure": 594,
|
| 870 |
+
"input_token_samples": 445,
|
| 871 |
+
"mean_actions": 41.711553175210405,
|
| 872 |
+
"mean_cost_source_native_units": 19.595637352118167,
|
| 873 |
+
"mean_duration_seconds": 664.1525678860841,
|
| 874 |
+
"mean_events": 78.16296863045142,
|
| 875 |
+
"mean_input_tokens": 943597.584269663,
|
| 876 |
+
"mean_output_tokens": 14243.811235955056,
|
| 877 |
+
"mean_supervised_assistant_characters": 13112.324407039021,
|
| 878 |
+
"output_token_samples": 445,
|
| 879 |
+
"structurally_eligible": 601,
|
| 880 |
+
"success": 713,
|
| 881 |
+
"success_rate_known_outcomes": 0.5455241009946442,
|
| 882 |
+
"training_eligible": 0,
|
| 883 |
+
"trajectories": 1307,
|
| 884 |
+
"unique_tasks": 89,
|
| 885 |
+
"unknown": 0
|
| 886 |
+
},
|
| 887 |
+
"claude-opus-4-6@anthropic": {
|
| 888 |
+
"cost_samples": 1127,
|
| 889 |
+
"duration_samples": 1127,
|
| 890 |
+
"failure": 403,
|
| 891 |
+
"input_token_samples": 692,
|
| 892 |
+
"mean_actions": 29.628216503992903,
|
| 893 |
+
"mean_cost_source_native_units": 35.3747195851623,
|
| 894 |
+
"mean_duration_seconds": 667.7565543810244,
|
| 895 |
+
"mean_events": 54.184560780834076,
|
| 896 |
+
"mean_input_tokens": 543390.0708092486,
|
| 897 |
+
"mean_output_tokens": 16201.849710982659,
|
| 898 |
+
"mean_supervised_assistant_characters": 8093.924578527063,
|
| 899 |
+
"output_token_samples": 692,
|
| 900 |
+
"structurally_eligible": 689,
|
| 901 |
+
"success": 724,
|
| 902 |
+
"success_rate_known_outcomes": 0.642413487133984,
|
| 903 |
+
"training_eligible": 0,
|
| 904 |
+
"trajectories": 1127,
|
| 905 |
+
"unique_tasks": 89,
|
| 906 |
+
"unknown": 0
|
| 907 |
+
},
|
| 908 |
+
"claude-opus-4.6": {
|
| 909 |
+
"cost_samples": 0,
|
| 910 |
+
"duration_samples": 0,
|
| 911 |
+
"failure": 114,
|
| 912 |
+
"input_token_samples": 0,
|
| 913 |
+
"mean_actions": 28.932,
|
| 914 |
+
"mean_cost_source_native_units": null,
|
| 915 |
+
"mean_duration_seconds": null,
|
| 916 |
+
"mean_events": 60.844,
|
| 917 |
+
"mean_input_tokens": null,
|
| 918 |
+
"mean_output_tokens": null,
|
| 919 |
+
"mean_supervised_assistant_characters": 6521.846,
|
| 920 |
+
"output_token_samples": 0,
|
| 921 |
+
"structurally_eligible": 500,
|
| 922 |
+
"success": 386,
|
| 923 |
+
"success_rate_known_outcomes": 0.772,
|
| 924 |
+
"training_eligible": 0,
|
| 925 |
+
"trajectories": 500,
|
| 926 |
+
"unique_tasks": 500,
|
| 927 |
+
"unknown": 0
|
| 928 |
+
},
|
| 929 |
+
"claude-opus-4.6@anthropic": {
|
| 930 |
+
"cost_samples": 157,
|
| 931 |
+
"duration_samples": 157,
|
| 932 |
+
"failure": 37,
|
| 933 |
+
"input_token_samples": 0,
|
| 934 |
+
"mean_actions": 32.611464968152866,
|
| 935 |
+
"mean_cost_source_native_units": 0.0,
|
| 936 |
+
"mean_duration_seconds": 792.0475469212623,
|
| 937 |
+
"mean_events": 65.35031847133757,
|
| 938 |
+
"mean_input_tokens": null,
|
| 939 |
+
"mean_output_tokens": null,
|
| 940 |
+
"mean_supervised_assistant_characters": 4949.356687898089,
|
| 941 |
+
"output_token_samples": 0,
|
| 942 |
+
"structurally_eligible": 14,
|
| 943 |
+
"success": 120,
|
| 944 |
+
"success_rate_known_outcomes": 0.7643312101910829,
|
| 945 |
+
"training_eligible": 0,
|
| 946 |
+
"trajectories": 157,
|
| 947 |
+
"unique_tasks": 83,
|
| 948 |
+
"unknown": 0
|
| 949 |
+
},
|
| 950 |
+
"claude-sonnet-4-5-20250929@anthropic": {
|
| 951 |
+
"cost_samples": 1696,
|
| 952 |
+
"duration_samples": 1696,
|
| 953 |
+
"failure": 958,
|
| 954 |
+
"input_token_samples": 1692,
|
| 955 |
+
"mean_actions": 56.375,
|
| 956 |
+
"mean_cost_source_native_units": 66.14636548995128,
|
| 957 |
+
"mean_duration_seconds": 662.6581952335699,
|
| 958 |
+
"mean_events": 96.6745283018868,
|
| 959 |
+
"mean_input_tokens": 1397487.5827423169,
|
| 960 |
+
"mean_output_tokens": 11408.179669030733,
|
| 961 |
+
"mean_supervised_assistant_characters": 16556.10554245283,
|
| 962 |
+
"output_token_samples": 1692,
|
| 963 |
+
"structurally_eligible": 868,
|
| 964 |
+
"success": 738,
|
| 965 |
+
"success_rate_known_outcomes": 0.43514150943396224,
|
| 966 |
+
"training_eligible": 0,
|
| 967 |
+
"trajectories": 1696,
|
| 968 |
+
"unique_tasks": 89,
|
| 969 |
+
"unknown": 0
|
| 970 |
+
},
|
| 971 |
+
"deepseek-ai/DeepSeek-R1-0528": {
|
| 972 |
+
"cost_samples": 0,
|
| 973 |
+
"duration_samples": 0,
|
| 974 |
+
"failure": 157,
|
| 975 |
+
"input_token_samples": 0,
|
| 976 |
+
"mean_actions": 13.305755395683454,
|
| 977 |
+
"mean_cost_source_native_units": null,
|
| 978 |
+
"mean_duration_seconds": null,
|
| 979 |
+
"mean_events": 36.50359712230216,
|
| 980 |
+
"mean_input_tokens": null,
|
| 981 |
+
"mean_output_tokens": null,
|
| 982 |
+
"mean_supervised_assistant_characters": 5477.802158273381,
|
| 983 |
+
"output_token_samples": 0,
|
| 984 |
+
"structurally_eligible": 277,
|
| 985 |
+
"success": 121,
|
| 986 |
+
"success_rate_known_outcomes": 0.4352517985611511,
|
| 987 |
+
"training_eligible": 277,
|
| 988 |
+
"trajectories": 278,
|
| 989 |
+
"unique_tasks": 228,
|
| 990 |
+
"unknown": 0
|
| 991 |
+
},
|
| 992 |
+
"deepseek-ai/DeepSeek-V3-0324": {
|
| 993 |
+
"cost_samples": 0,
|
| 994 |
+
"duration_samples": 0,
|
| 995 |
+
"failure": 168,
|
| 996 |
+
"input_token_samples": 0,
|
| 997 |
+
"mean_actions": 14.212230215827338,
|
| 998 |
+
"mean_cost_source_native_units": null,
|
| 999 |
+
"mean_duration_seconds": null,
|
| 1000 |
+
"mean_events": 40.07194244604317,
|
| 1001 |
+
"mean_input_tokens": null,
|
| 1002 |
+
"mean_output_tokens": null,
|
| 1003 |
+
"mean_supervised_assistant_characters": 8366.03237410072,
|
| 1004 |
+
"output_token_samples": 0,
|
| 1005 |
+
"structurally_eligible": 278,
|
| 1006 |
+
"success": 110,
|
| 1007 |
+
"success_rate_known_outcomes": 0.39568345323741005,
|
| 1008 |
+
"training_eligible": 278,
|
| 1009 |
+
"trajectories": 278,
|
| 1010 |
+
"unique_tasks": 228,
|
| 1011 |
+
"unknown": 0
|
| 1012 |
+
},
|
| 1013 |
+
"deepseek-ai/DeepSeek-V3.2-Exp-thinking-off": {
|
| 1014 |
+
"cost_samples": 0,
|
| 1015 |
+
"duration_samples": 0,
|
| 1016 |
+
"failure": 151,
|
| 1017 |
+
"input_token_samples": 0,
|
| 1018 |
+
"mean_actions": 15.356115107913668,
|
| 1019 |
+
"mean_cost_source_native_units": null,
|
| 1020 |
+
"mean_duration_seconds": null,
|
| 1021 |
+
"mean_events": 40.989208633093526,
|
| 1022 |
+
"mean_input_tokens": null,
|
| 1023 |
+
"mean_output_tokens": null,
|
| 1024 |
+
"mean_supervised_assistant_characters": 5493.4604316546765,
|
| 1025 |
+
"output_token_samples": 0,
|
| 1026 |
+
"structurally_eligible": 278,
|
| 1027 |
+
"success": 127,
|
| 1028 |
+
"success_rate_known_outcomes": 0.4568345323741007,
|
| 1029 |
+
"training_eligible": 278,
|
| 1030 |
+
"trajectories": 278,
|
| 1031 |
+
"unique_tasks": 228,
|
| 1032 |
+
"unknown": 0
|
| 1033 |
+
},
|
| 1034 |
+
"deepseek-ai/DeepSeek-V3.2-Exp-thinking-on": {
|
| 1035 |
+
"cost_samples": 0,
|
| 1036 |
+
"duration_samples": 0,
|
| 1037 |
+
"failure": 154,
|
| 1038 |
+
"input_token_samples": 0,
|
| 1039 |
+
"mean_actions": 14.676258992805755,
|
| 1040 |
+
"mean_cost_source_native_units": null,
|
| 1041 |
+
"mean_duration_seconds": null,
|
| 1042 |
+
"mean_events": 39.856115107913666,
|
| 1043 |
+
"mean_input_tokens": null,
|
| 1044 |
+
"mean_output_tokens": null,
|
| 1045 |
+
"mean_supervised_assistant_characters": 4574.9604316546765,
|
| 1046 |
+
"output_token_samples": 0,
|
| 1047 |
+
"structurally_eligible": 278,
|
| 1048 |
+
"success": 124,
|
| 1049 |
+
"success_rate_known_outcomes": 0.4460431654676259,
|
| 1050 |
+
"training_eligible": 278,
|
| 1051 |
+
"trajectories": 278,
|
| 1052 |
+
"unique_tasks": 228,
|
| 1053 |
+
"unknown": 0
|
| 1054 |
+
},
|
| 1055 |
+
"deepseek-v3.2@deepseek": {
|
| 1056 |
+
"cost_samples": 373,
|
| 1057 |
+
"duration_samples": 373,
|
| 1058 |
+
"failure": 220,
|
| 1059 |
+
"input_token_samples": 373,
|
| 1060 |
+
"mean_actions": 22.630026809651476,
|
| 1061 |
+
"mean_cost_source_native_units": 1.1921455508981889,
|
| 1062 |
+
"mean_duration_seconds": 724.4749572385732,
|
| 1063 |
+
"mean_events": 46.265415549597854,
|
| 1064 |
+
"mean_input_tokens": 517430.51742627344,
|
| 1065 |
+
"mean_output_tokens": 15099.662198391421,
|
| 1066 |
+
"mean_supervised_assistant_characters": 14214.538873994637,
|
| 1067 |
+
"output_token_samples": 373,
|
| 1068 |
+
"structurally_eligible": 373,
|
| 1069 |
+
"success": 153,
|
| 1070 |
+
"success_rate_known_outcomes": 0.4101876675603217,
|
| 1071 |
+
"training_eligible": 0,
|
| 1072 |
+
"trajectories": 373,
|
| 1073 |
+
"unique_tasks": 87,
|
| 1074 |
+
"unknown": 0
|
| 1075 |
+
},
|
| 1076 |
+
"gemini-2.5-flash@gemini": {
|
| 1077 |
+
"cost_samples": 1678,
|
| 1078 |
+
"duration_samples": 1678,
|
| 1079 |
+
"failure": 1386,
|
| 1080 |
+
"input_token_samples": 1677,
|
| 1081 |
+
"mean_actions": 42.11442193087009,
|
| 1082 |
+
"mean_cost_source_native_units": 20.479215825333995,
|
| 1083 |
+
"mean_duration_seconds": 630.2167853126936,
|
| 1084 |
+
"mean_events": 73.4523241954708,
|
| 1085 |
+
"mean_input_tokens": 2427590.1562313656,
|
| 1086 |
+
"mean_output_tokens": 37394.98270721526,
|
| 1087 |
+
"mean_supervised_assistant_characters": 12389.981525625744,
|
| 1088 |
+
"output_token_samples": 1677,
|
| 1089 |
+
"structurally_eligible": 1151,
|
| 1090 |
+
"success": 292,
|
| 1091 |
+
"success_rate_known_outcomes": 0.17401668653158522,
|
| 1092 |
+
"training_eligible": 0,
|
| 1093 |
+
"trajectories": 1678,
|
| 1094 |
+
"unique_tasks": 89,
|
| 1095 |
+
"unknown": 0
|
| 1096 |
+
},
|
| 1097 |
+
"gemini-2.5-pro@gemini": {
|
| 1098 |
+
"cost_samples": 1687,
|
| 1099 |
+
"duration_samples": 1687,
|
| 1100 |
+
"failure": 1270,
|
| 1101 |
+
"input_token_samples": 1685,
|
| 1102 |
+
"mean_actions": 33.0,
|
| 1103 |
+
"mean_cost_source_native_units": 53.944440666818025,
|
| 1104 |
+
"mean_duration_seconds": 612.0814591494268,
|
| 1105 |
+
"mean_events": 57.75874333135744,
|
| 1106 |
+
"mean_input_tokens": 1062349.9721068249,
|
| 1107 |
+
"mean_output_tokens": 28035.420771513353,
|
| 1108 |
+
"mean_supervised_assistant_characters": 9635.45880260818,
|
| 1109 |
+
"output_token_samples": 1685,
|
| 1110 |
+
"structurally_eligible": 1149,
|
| 1111 |
+
"success": 417,
|
| 1112 |
+
"success_rate_known_outcomes": 0.24718435091879076,
|
| 1113 |
+
"training_eligible": 0,
|
| 1114 |
+
"trajectories": 1687,
|
| 1115 |
+
"unique_tasks": 89,
|
| 1116 |
+
"unknown": 0
|
| 1117 |
+
},
|
| 1118 |
+
"gemini-3-flash-high": {
|
| 1119 |
+
"cost_samples": 0,
|
| 1120 |
+
"duration_samples": 0,
|
| 1121 |
+
"failure": 121,
|
| 1122 |
+
"input_token_samples": 0,
|
| 1123 |
+
"mean_actions": 56.108,
|
| 1124 |
+
"mean_cost_source_native_units": null,
|
| 1125 |
+
"mean_duration_seconds": null,
|
| 1126 |
+
"mean_events": 113.236,
|
| 1127 |
+
"mean_input_tokens": null,
|
| 1128 |
+
"mean_output_tokens": null,
|
| 1129 |
+
"mean_supervised_assistant_characters": 68234.066,
|
| 1130 |
+
"output_token_samples": 0,
|
| 1131 |
+
"structurally_eligible": 498,
|
| 1132 |
+
"success": 379,
|
| 1133 |
+
"success_rate_known_outcomes": 0.758,
|
| 1134 |
+
"training_eligible": 0,
|
| 1135 |
+
"trajectories": 500,
|
| 1136 |
+
"unique_tasks": 500,
|
| 1137 |
+
"unknown": 0
|
| 1138 |
+
},
|
| 1139 |
+
"gemini-3-flash-preview@gemini": {
|
| 1140 |
+
"cost_samples": 444,
|
| 1141 |
+
"duration_samples": 444,
|
| 1142 |
+
"failure": 214,
|
| 1143 |
+
"input_token_samples": 0,
|
| 1144 |
+
"mean_actions": 19.763513513513512,
|
| 1145 |
+
"mean_cost_source_native_units": 0.0,
|
| 1146 |
+
"mean_duration_seconds": 545.6824201635412,
|
| 1147 |
+
"mean_events": 40.556306306306304,
|
| 1148 |
+
"mean_input_tokens": null,
|
| 1149 |
+
"mean_output_tokens": null,
|
| 1150 |
+
"mean_supervised_assistant_characters": 11352.623873873874,
|
| 1151 |
+
"output_token_samples": 0,
|
| 1152 |
+
"structurally_eligible": 444,
|
| 1153 |
+
"success": 230,
|
| 1154 |
+
"success_rate_known_outcomes": 0.5180180180180181,
|
| 1155 |
+
"training_eligible": 0,
|
| 1156 |
+
"trajectories": 444,
|
| 1157 |
+
"unique_tasks": 89,
|
| 1158 |
+
"unknown": 0
|
| 1159 |
+
},
|
| 1160 |
+
"gemini-3-pro": {
|
| 1161 |
+
"cost_samples": 0,
|
| 1162 |
+
"duration_samples": 0,
|
| 1163 |
+
"failure": 0,
|
| 1164 |
+
"input_token_samples": 0,
|
| 1165 |
+
"mean_actions": 51.256,
|
| 1166 |
+
"mean_cost_source_native_units": null,
|
| 1167 |
+
"mean_duration_seconds": null,
|
| 1168 |
+
"mean_events": 103.722,
|
| 1169 |
+
"mean_input_tokens": null,
|
| 1170 |
+
"mean_output_tokens": null,
|
| 1171 |
+
"mean_supervised_assistant_characters": 203.868,
|
| 1172 |
+
"output_token_samples": 0,
|
| 1173 |
+
"structurally_eligible": 490,
|
| 1174 |
+
"success": 500,
|
| 1175 |
+
"success_rate_known_outcomes": 1.0,
|
| 1176 |
+
"training_eligible": 0,
|
| 1177 |
+
"trajectories": 500,
|
| 1178 |
+
"unique_tasks": 500,
|
| 1179 |
+
"unknown": 0
|
| 1180 |
+
},
|
| 1181 |
+
"gemini-3-pro-preview@gemini": {
|
| 1182 |
+
"cost_samples": 445,
|
| 1183 |
+
"duration_samples": 445,
|
| 1184 |
+
"failure": 192,
|
| 1185 |
+
"input_token_samples": 445,
|
| 1186 |
+
"mean_actions": 13.262921348314606,
|
| 1187 |
+
"mean_cost_source_native_units": 45.99529477558779,
|
| 1188 |
+
"mean_duration_seconds": 548.5905832183495,
|
| 1189 |
+
"mean_events": 27.557303370786517,
|
| 1190 |
+
"mean_input_tokens": 320152.2426966292,
|
| 1191 |
+
"mean_output_tokens": 24330.34831460674,
|
| 1192 |
+
"mean_supervised_assistant_characters": 7961.357303370786,
|
| 1193 |
+
"output_token_samples": 445,
|
| 1194 |
+
"structurally_eligible": 445,
|
| 1195 |
+
"success": 253,
|
| 1196 |
+
"success_rate_known_outcomes": 0.5685393258426966,
|
| 1197 |
+
"training_eligible": 0,
|
| 1198 |
+
"trajectories": 445,
|
| 1199 |
+
"unique_tasks": 89,
|
| 1200 |
+
"unknown": 0
|
| 1201 |
+
},
|
| 1202 |
+
"gemini-3.1-pro-preview@Google": {
|
| 1203 |
+
"cost_samples": 431,
|
| 1204 |
+
"duration_samples": 431,
|
| 1205 |
+
"failure": 111,
|
| 1206 |
+
"input_token_samples": 431,
|
| 1207 |
+
"mean_actions": 26.454756380510442,
|
| 1208 |
+
"mean_cost_source_native_units": 75.75997570551189,
|
| 1209 |
+
"mean_duration_seconds": 632.5132716946702,
|
| 1210 |
+
"mean_events": 53.909512761020885,
|
| 1211 |
+
"mean_input_tokens": 992285.0626450116,
|
| 1212 |
+
"mean_output_tokens": 29053.654292343388,
|
| 1213 |
+
"mean_supervised_assistant_characters": 4961.839907192575,
|
| 1214 |
+
"output_token_samples": 431,
|
| 1215 |
+
"structurally_eligible": 431,
|
| 1216 |
+
"success": 320,
|
| 1217 |
+
"success_rate_known_outcomes": 0.7424593967517401,
|
| 1218 |
+
"training_eligible": 0,
|
| 1219 |
+
"trajectories": 431,
|
| 1220 |
+
"unique_tasks": 89,
|
| 1221 |
+
"unknown": 0
|
| 1222 |
+
},
|
| 1223 |
+
"glm-5-high": {
|
| 1224 |
+
"cost_samples": 0,
|
| 1225 |
+
"duration_samples": 0,
|
| 1226 |
+
"failure": 136,
|
| 1227 |
+
"input_token_samples": 0,
|
| 1228 |
+
"mean_actions": 76.026,
|
| 1229 |
+
"mean_cost_source_native_units": null,
|
| 1230 |
+
"mean_duration_seconds": null,
|
| 1231 |
+
"mean_events": 154.152,
|
| 1232 |
+
"mean_input_tokens": null,
|
| 1233 |
+
"mean_output_tokens": null,
|
| 1234 |
+
"mean_supervised_assistant_characters": 350.692,
|
| 1235 |
+
"output_token_samples": 0,
|
| 1236 |
+
"structurally_eligible": 500,
|
| 1237 |
+
"success": 364,
|
| 1238 |
+
"success_rate_known_outcomes": 0.728,
|
| 1239 |
+
"training_eligible": 0,
|
| 1240 |
+
"trajectories": 500,
|
| 1241 |
+
"unique_tasks": 500,
|
| 1242 |
+
"unknown": 0
|
| 1243 |
+
},
|
| 1244 |
+
"glm-5@z-ai": {
|
| 1245 |
+
"cost_samples": 376,
|
| 1246 |
+
"duration_samples": 376,
|
| 1247 |
+
"failure": 165,
|
| 1248 |
+
"input_token_samples": 376,
|
| 1249 |
+
"mean_actions": 28.388297872340427,
|
| 1250 |
+
"mean_cost_source_native_units": 0.0,
|
| 1251 |
+
"mean_duration_seconds": 809.2636813914522,
|
| 1252 |
+
"mean_events": 57.81382978723404,
|
| 1253 |
+
"mean_input_tokens": 0.0,
|
| 1254 |
+
"mean_output_tokens": 0.0,
|
| 1255 |
+
"mean_supervised_assistant_characters": 14967.672872340425,
|
| 1256 |
+
"output_token_samples": 376,
|
| 1257 |
+
"structurally_eligible": 376,
|
| 1258 |
+
"success": 211,
|
| 1259 |
+
"success_rate_known_outcomes": 0.5611702127659575,
|
| 1260 |
+
"training_eligible": 0,
|
| 1261 |
+
"trajectories": 376,
|
| 1262 |
+
"unique_tasks": 88,
|
| 1263 |
+
"unknown": 0
|
| 1264 |
+
},
|
| 1265 |
+
"google/gemini-2.5-flash-thinking-off": {
|
| 1266 |
+
"cost_samples": 0,
|
| 1267 |
+
"duration_samples": 0,
|
| 1268 |
+
"failure": 168,
|
| 1269 |
+
"input_token_samples": 0,
|
| 1270 |
+
"mean_actions": 11.456834532374101,
|
| 1271 |
+
"mean_cost_source_native_units": null,
|
| 1272 |
+
"mean_duration_seconds": null,
|
| 1273 |
+
"mean_events": 25.154676258992804,
|
| 1274 |
+
"mean_input_tokens": null,
|
| 1275 |
+
"mean_output_tokens": null,
|
| 1276 |
+
"mean_supervised_assistant_characters": 1610.4496402877699,
|
| 1277 |
+
"output_token_samples": 0,
|
| 1278 |
+
"structurally_eligible": 278,
|
| 1279 |
+
"success": 110,
|
| 1280 |
+
"success_rate_known_outcomes": 0.39568345323741005,
|
| 1281 |
+
"training_eligible": 278,
|
| 1282 |
+
"trajectories": 278,
|
| 1283 |
+
"unique_tasks": 228,
|
| 1284 |
+
"unknown": 0
|
| 1285 |
+
},
|
| 1286 |
+
"google/gemini-2.5-flash-thinking-on": {
|
| 1287 |
+
"cost_samples": 0,
|
| 1288 |
+
"duration_samples": 0,
|
| 1289 |
+
"failure": 176,
|
| 1290 |
+
"input_token_samples": 0,
|
| 1291 |
+
"mean_actions": 15.902877697841726,
|
| 1292 |
+
"mean_cost_source_native_units": null,
|
| 1293 |
+
"mean_duration_seconds": null,
|
| 1294 |
+
"mean_events": 39.02517985611511,
|
| 1295 |
+
"mean_input_tokens": null,
|
| 1296 |
+
"mean_output_tokens": null,
|
| 1297 |
+
"mean_supervised_assistant_characters": 2495.3920863309354,
|
| 1298 |
+
"output_token_samples": 0,
|
| 1299 |
+
"structurally_eligible": 278,
|
| 1300 |
+
"success": 102,
|
| 1301 |
+
"success_rate_known_outcomes": 0.3669064748201439,
|
| 1302 |
+
"training_eligible": 278,
|
| 1303 |
+
"trajectories": 278,
|
| 1304 |
+
"unique_tasks": 228,
|
| 1305 |
+
"unknown": 0
|
| 1306 |
+
},
|
| 1307 |
+
"google/gemini-2.5-pro-thinking-on": {
|
| 1308 |
+
"cost_samples": 0,
|
| 1309 |
+
"duration_samples": 0,
|
| 1310 |
+
"failure": 165,
|
| 1311 |
+
"input_token_samples": 0,
|
| 1312 |
+
"mean_actions": 15.02158273381295,
|
| 1313 |
+
"mean_cost_source_native_units": null,
|
| 1314 |
+
"mean_duration_seconds": null,
|
| 1315 |
+
"mean_events": 37.53237410071942,
|
| 1316 |
+
"mean_input_tokens": null,
|
| 1317 |
+
"mean_output_tokens": null,
|
| 1318 |
+
"mean_supervised_assistant_characters": 2018.86690647482,
|
| 1319 |
+
"output_token_samples": 0,
|
| 1320 |
+
"structurally_eligible": 278,
|
| 1321 |
+
"success": 113,
|
| 1322 |
+
"success_rate_known_outcomes": 0.4064748201438849,
|
| 1323 |
+
"training_eligible": 278,
|
| 1324 |
+
"trajectories": 278,
|
| 1325 |
+
"unique_tasks": 228,
|
| 1326 |
+
"unknown": 0
|
| 1327 |
+
},
|
| 1328 |
+
"gpt-4o-2024-08-06": {
|
| 1329 |
+
"cost_samples": 0,
|
| 1330 |
+
"duration_samples": 0,
|
| 1331 |
+
"failure": 5308,
|
| 1332 |
+
"input_token_samples": 0,
|
| 1333 |
+
"mean_actions": 21.264601320467243,
|
| 1334 |
+
"mean_cost_source_native_units": null,
|
| 1335 |
+
"mean_duration_seconds": null,
|
| 1336 |
+
"mean_events": 44.01574403250381,
|
| 1337 |
+
"mean_input_tokens": null,
|
| 1338 |
+
"mean_output_tokens": null,
|
| 1339 |
+
"mean_supervised_assistant_characters": 7167.808362959201,
|
| 1340 |
+
"output_token_samples": 0,
|
| 1341 |
+
"structurally_eligible": 5907,
|
| 1342 |
+
"success": 599,
|
| 1343 |
+
"success_rate_known_outcomes": 0.10140511257829693,
|
| 1344 |
+
"training_eligible": 5907,
|
| 1345 |
+
"trajectories": 5907,
|
| 1346 |
+
"unique_tasks": 2649,
|
| 1347 |
+
"unknown": 0
|
| 1348 |
+
},
|
| 1349 |
+
"gpt-5-codex@openai": {
|
| 1350 |
+
"cost_samples": 1027,
|
| 1351 |
+
"duration_samples": 1027,
|
| 1352 |
+
"failure": 579,
|
| 1353 |
+
"input_token_samples": 1025,
|
| 1354 |
+
"mean_actions": 23.137293086660176,
|
| 1355 |
+
"mean_cost_source_native_units": 34.65822178652798,
|
| 1356 |
+
"mean_duration_seconds": 894.1482539265158,
|
| 1357 |
+
"mean_events": 47.249269717624145,
|
| 1358 |
+
"mean_input_tokens": 404544.2819512195,
|
| 1359 |
+
"mean_output_tokens": 24892.82731707317,
|
| 1360 |
+
"mean_supervised_assistant_characters": 9759.669912366115,
|
| 1361 |
+
"output_token_samples": 1025,
|
| 1362 |
+
"structurally_eligible": 882,
|
| 1363 |
+
"success": 448,
|
| 1364 |
+
"success_rate_known_outcomes": 0.436222005842259,
|
| 1365 |
+
"training_eligible": 0,
|
| 1366 |
+
"trajectories": 1027,
|
| 1367 |
+
"unique_tasks": 89,
|
| 1368 |
+
"unknown": 0
|
| 1369 |
+
},
|
| 1370 |
+
"gpt-5-mini": {
|
| 1371 |
+
"cost_samples": 0,
|
| 1372 |
+
"duration_samples": 0,
|
| 1373 |
+
"failure": 219,
|
| 1374 |
+
"input_token_samples": 0,
|
| 1375 |
+
"mean_actions": 20.108,
|
| 1376 |
+
"mean_cost_source_native_units": null,
|
| 1377 |
+
"mean_duration_seconds": null,
|
| 1378 |
+
"mean_events": 41.45,
|
| 1379 |
+
"mean_input_tokens": null,
|
| 1380 |
+
"mean_output_tokens": null,
|
| 1381 |
+
"mean_supervised_assistant_characters": 1133.53,
|
| 1382 |
+
"output_token_samples": 0,
|
| 1383 |
+
"structurally_eligible": 500,
|
| 1384 |
+
"success": 281,
|
| 1385 |
+
"success_rate_known_outcomes": 0.562,
|
| 1386 |
+
"training_eligible": 0,
|
| 1387 |
+
"trajectories": 500,
|
| 1388 |
+
"unique_tasks": 500,
|
| 1389 |
+
"unknown": 0
|
| 1390 |
+
},
|
| 1391 |
+
"gpt-5-mini@openai": {
|
| 1392 |
+
"cost_samples": 1705,
|
| 1393 |
+
"duration_samples": 1705,
|
| 1394 |
+
"failure": 1231,
|
| 1395 |
+
"input_token_samples": 1703,
|
| 1396 |
+
"mean_actions": 20.30909090909091,
|
| 1397 |
+
"mean_cost_source_native_units": 6.237601142212792,
|
| 1398 |
+
"mean_duration_seconds": 738.1692138985105,
|
| 1399 |
+
"mean_events": 36.47214076246334,
|
| 1400 |
+
"mean_input_tokens": 362939.1297709924,
|
| 1401 |
+
"mean_output_tokens": 20953.19553728714,
|
| 1402 |
+
"mean_supervised_assistant_characters": 7645.26568914956,
|
| 1403 |
+
"output_token_samples": 1703,
|
| 1404 |
+
"structurally_eligible": 857,
|
| 1405 |
+
"success": 474,
|
| 1406 |
+
"success_rate_known_outcomes": 0.2780058651026393,
|
| 1407 |
+
"training_eligible": 0,
|
| 1408 |
+
"trajectories": 1705,
|
| 1409 |
+
"unique_tasks": 89,
|
| 1410 |
+
"unknown": 0
|
| 1411 |
+
},
|
| 1412 |
+
"gpt-5-nano@openai": {
|
| 1413 |
+
"cost_samples": 1707,
|
| 1414 |
+
"duration_samples": 1707,
|
| 1415 |
+
"failure": 1546,
|
| 1416 |
+
"input_token_samples": 1703,
|
| 1417 |
+
"mean_actions": 27.11833626244874,
|
| 1418 |
+
"mean_cost_source_native_units": 2.833243009294413,
|
| 1419 |
+
"mean_duration_seconds": 795.4362061986046,
|
| 1420 |
+
"mean_events": 49.25834797891037,
|
| 1421 |
+
"mean_input_tokens": 561219.6964180857,
|
| 1422 |
+
"mean_output_tokens": 53272.514386376984,
|
| 1423 |
+
"mean_supervised_assistant_characters": 11135.304042179261,
|
| 1424 |
+
"output_token_samples": 1703,
|
| 1425 |
+
"structurally_eligible": 861,
|
| 1426 |
+
"success": 161,
|
| 1427 |
+
"success_rate_known_outcomes": 0.09431751611013474,
|
| 1428 |
+
"training_eligible": 0,
|
| 1429 |
+
"trajectories": 1707,
|
| 1430 |
+
"unique_tasks": 89,
|
| 1431 |
+
"unknown": 0
|
| 1432 |
+
},
|
| 1433 |
+
"gpt-5.1-codex@openai": {
|
| 1434 |
+
"cost_samples": 441,
|
| 1435 |
+
"duration_samples": 441,
|
| 1436 |
+
"failure": 277,
|
| 1437 |
+
"input_token_samples": 440,
|
| 1438 |
+
"mean_actions": 22.573696145124718,
|
| 1439 |
+
"mean_cost_source_native_units": 31.287152531770083,
|
| 1440 |
+
"mean_duration_seconds": 1209.0573483940695,
|
| 1441 |
+
"mean_events": 46.14965986394558,
|
| 1442 |
+
"mean_input_tokens": 481445.4818181818,
|
| 1443 |
+
"mean_output_tokens": 20365.454545454544,
|
| 1444 |
+
"mean_supervised_assistant_characters": 11924.947845804989,
|
| 1445 |
+
"output_token_samples": 440,
|
| 1446 |
+
"structurally_eligible": 429,
|
| 1447 |
+
"success": 164,
|
| 1448 |
+
"success_rate_known_outcomes": 0.37188208616780044,
|
| 1449 |
+
"training_eligible": 0,
|
| 1450 |
+
"trajectories": 441,
|
| 1451 |
+
"unique_tasks": 89,
|
| 1452 |
+
"unknown": 0
|
| 1453 |
+
},
|
| 1454 |
+
"gpt-5.1@openai": {
|
| 1455 |
+
"cost_samples": 445,
|
| 1456 |
+
"duration_samples": 445,
|
| 1457 |
+
"failure": 233,
|
| 1458 |
+
"input_token_samples": 444,
|
| 1459 |
+
"mean_actions": 8.55056179775281,
|
| 1460 |
+
"mean_cost_source_native_units": 41.71665708241838,
|
| 1461 |
+
"mean_duration_seconds": 845.7475696006518,
|
| 1462 |
+
"mean_events": 18.10112359550562,
|
| 1463 |
+
"mean_input_tokens": 134251.63288288287,
|
| 1464 |
+
"mean_output_tokens": 37152.01351351351,
|
| 1465 |
+
"mean_supervised_assistant_characters": 6334.330337078652,
|
| 1466 |
+
"output_token_samples": 444,
|
| 1467 |
+
"structurally_eligible": 438,
|
| 1468 |
+
"success": 212,
|
| 1469 |
+
"success_rate_known_outcomes": 0.4764044943820225,
|
| 1470 |
+
"training_eligible": 0,
|
| 1471 |
+
"trajectories": 445,
|
| 1472 |
+
"unique_tasks": 89,
|
| 1473 |
+
"unknown": 0
|
| 1474 |
+
},
|
| 1475 |
+
"gpt-5.2-codex": {
|
| 1476 |
+
"cost_samples": 0,
|
| 1477 |
+
"duration_samples": 0,
|
| 1478 |
+
"failure": 86,
|
| 1479 |
+
"input_token_samples": 0,
|
| 1480 |
+
"mean_actions": 35.49090909090909,
|
| 1481 |
+
"mean_cost_source_native_units": null,
|
| 1482 |
+
"mean_duration_seconds": null,
|
| 1483 |
+
"mean_events": 73.75636363636363,
|
| 1484 |
+
"mean_input_tokens": null,
|
| 1485 |
+
"mean_output_tokens": null,
|
| 1486 |
+
"mean_supervised_assistant_characters": 949.08,
|
| 1487 |
+
"output_token_samples": 0,
|
| 1488 |
+
"structurally_eligible": 275,
|
| 1489 |
+
"success": 189,
|
| 1490 |
+
"success_rate_known_outcomes": 0.6872727272727273,
|
| 1491 |
+
"training_eligible": 0,
|
| 1492 |
+
"trajectories": 275,
|
| 1493 |
+
"unique_tasks": 275,
|
| 1494 |
+
"unknown": 0
|
| 1495 |
+
},
|
| 1496 |
+
"gpt-5.2-codex@openai": {
|
| 1497 |
+
"cost_samples": 380,
|
| 1498 |
+
"duration_samples": 380,
|
| 1499 |
+
"failure": 98,
|
| 1500 |
+
"input_token_samples": 0,
|
| 1501 |
+
"mean_actions": 29.263157894736842,
|
| 1502 |
+
"mean_cost_source_native_units": 0.0,
|
| 1503 |
+
"mean_duration_seconds": 705.5991674925151,
|
| 1504 |
+
"mean_events": 59.747368421052634,
|
| 1505 |
+
"mean_input_tokens": null,
|
| 1506 |
+
"mean_output_tokens": null,
|
| 1507 |
+
"mean_supervised_assistant_characters": 4741.5447368421055,
|
| 1508 |
+
"output_token_samples": 0,
|
| 1509 |
+
"structurally_eligible": 337,
|
| 1510 |
+
"success": 282,
|
| 1511 |
+
"success_rate_known_outcomes": 0.7421052631578947,
|
| 1512 |
+
"training_eligible": 0,
|
| 1513 |
+
"trajectories": 380,
|
| 1514 |
+
"unique_tasks": 89,
|
| 1515 |
+
"unknown": 0
|
| 1516 |
+
},
|
| 1517 |
+
"gpt-5.2-high": {
|
| 1518 |
+
"cost_samples": 0,
|
| 1519 |
+
"duration_samples": 0,
|
| 1520 |
+
"failure": 72,
|
| 1521 |
+
"input_token_samples": 0,
|
| 1522 |
+
"mean_actions": 35.806691449814124,
|
| 1523 |
+
"mean_cost_source_native_units": null,
|
| 1524 |
+
"mean_duration_seconds": null,
|
| 1525 |
+
"mean_events": 74.45724907063197,
|
| 1526 |
+
"mean_input_tokens": null,
|
| 1527 |
+
"mean_output_tokens": null,
|
| 1528 |
+
"mean_supervised_assistant_characters": 943.1858736059479,
|
| 1529 |
+
"output_token_samples": 0,
|
| 1530 |
+
"structurally_eligible": 269,
|
| 1531 |
+
"success": 197,
|
| 1532 |
+
"success_rate_known_outcomes": 0.7323420074349443,
|
| 1533 |
+
"training_eligible": 0,
|
| 1534 |
+
"trajectories": 269,
|
| 1535 |
+
"unique_tasks": 269,
|
| 1536 |
+
"unknown": 0
|
| 1537 |
+
},
|
| 1538 |
+
"gpt-5.2@openai": {
|
| 1539 |
+
"cost_samples": 442,
|
| 1540 |
+
"duration_samples": 442,
|
| 1541 |
+
"failure": 203,
|
| 1542 |
+
"input_token_samples": 0,
|
| 1543 |
+
"mean_actions": 8.339366515837105,
|
| 1544 |
+
"mean_cost_source_native_units": 0.0,
|
| 1545 |
+
"mean_duration_seconds": 812.1008663954239,
|
| 1546 |
+
"mean_events": 17.67873303167421,
|
| 1547 |
+
"mean_input_tokens": null,
|
| 1548 |
+
"mean_output_tokens": null,
|
| 1549 |
+
"mean_supervised_assistant_characters": 9987.174208144796,
|
| 1550 |
+
"output_token_samples": 0,
|
| 1551 |
+
"structurally_eligible": 435,
|
| 1552 |
+
"success": 239,
|
| 1553 |
+
"success_rate_known_outcomes": 0.5407239819004525,
|
| 1554 |
+
"training_eligible": 0,
|
| 1555 |
+
"trajectories": 442,
|
| 1556 |
+
"unique_tasks": 89,
|
| 1557 |
+
"unknown": 0
|
| 1558 |
+
},
|
| 1559 |
+
"gpt-5@openai": {
|
| 1560 |
+
"cost_samples": 1707,
|
| 1561 |
+
"duration_samples": 1707,
|
| 1562 |
+
"failure": 989,
|
| 1563 |
+
"input_token_samples": 1705,
|
| 1564 |
+
"mean_actions": 18.446983011130637,
|
| 1565 |
+
"mean_cost_source_native_units": 32.87692377919262,
|
| 1566 |
+
"mean_duration_seconds": 789.6129793145884,
|
| 1567 |
+
"mean_events": 34.26479203280609,
|
| 1568 |
+
"mean_input_tokens": 288354.0152492669,
|
| 1569 |
+
"mean_output_tokens": 26094.951906158356,
|
| 1570 |
+
"mean_supervised_assistant_characters": 6976.230228471002,
|
| 1571 |
+
"output_token_samples": 1705,
|
| 1572 |
+
"structurally_eligible": 854,
|
| 1573 |
+
"success": 718,
|
| 1574 |
+
"success_rate_known_outcomes": 0.4206209724663152,
|
| 1575 |
+
"training_eligible": 0,
|
| 1576 |
+
"trajectories": 1707,
|
| 1577 |
+
"unique_tasks": 89,
|
| 1578 |
+
"unknown": 0
|
| 1579 |
+
},
|
| 1580 |
+
"gpt-oss-120b": {
|
| 1581 |
+
"cost_samples": 0,
|
| 1582 |
+
"duration_samples": 5063,
|
| 1583 |
+
"failure": 3852,
|
| 1584 |
+
"input_token_samples": 0,
|
| 1585 |
+
"mean_actions": 27.13786292711831,
|
| 1586 |
+
"mean_cost_source_native_units": null,
|
| 1587 |
+
"mean_duration_seconds": 108.07624142800492,
|
| 1588 |
+
"mean_events": 54.28303377444203,
|
| 1589 |
+
"mean_input_tokens": null,
|
| 1590 |
+
"mean_output_tokens": null,
|
| 1591 |
+
"mean_supervised_assistant_characters": 15385.260912502468,
|
| 1592 |
+
"output_token_samples": 0,
|
| 1593 |
+
"structurally_eligible": 5063,
|
| 1594 |
+
"success": 1211,
|
| 1595 |
+
"success_rate_known_outcomes": 0.23918625320955955,
|
| 1596 |
+
"training_eligible": 0,
|
| 1597 |
+
"trajectories": 5063,
|
| 1598 |
+
"unique_tasks": 1266,
|
| 1599 |
+
"unknown": 0
|
| 1600 |
+
},
|
| 1601 |
+
"grok-4-0709@xai": {
|
| 1602 |
+
"cost_samples": 1264,
|
| 1603 |
+
"duration_samples": 1264,
|
| 1604 |
+
"failure": 927,
|
| 1605 |
+
"input_token_samples": 1254,
|
| 1606 |
+
"mean_actions": 24.593354430379748,
|
| 1607 |
+
"mean_cost_source_native_units": -2.574454700680379,
|
| 1608 |
+
"mean_duration_seconds": 1095.5605030421968,
|
| 1609 |
+
"mean_events": 43.932753164556964,
|
| 1610 |
+
"mean_input_tokens": 292064.3724082935,
|
| 1611 |
+
"mean_output_tokens": 3859.8508771929824,
|
| 1612 |
+
"mean_supervised_assistant_characters": 7819.378955696203,
|
| 1613 |
+
"output_token_samples": 1254,
|
| 1614 |
+
"structurally_eligible": 821,
|
| 1615 |
+
"success": 337,
|
| 1616 |
+
"success_rate_known_outcomes": 0.2666139240506329,
|
| 1617 |
+
"training_eligible": 0,
|
| 1618 |
+
"trajectories": 1264,
|
| 1619 |
+
"unique_tasks": 89,
|
| 1620 |
+
"unknown": 0
|
| 1621 |
+
},
|
| 1622 |
+
"grok-code-fast-1@xai": {
|
| 1623 |
+
"cost_samples": 880,
|
| 1624 |
+
"duration_samples": 880,
|
| 1625 |
+
"failure": 702,
|
| 1626 |
+
"input_token_samples": 878,
|
| 1627 |
+
"mean_actions": 15.525,
|
| 1628 |
+
"mean_cost_source_native_units": 1.126558439504482,
|
| 1629 |
+
"mean_duration_seconds": 348.8899872476404,
|
| 1630 |
+
"mean_events": 32.86022727272727,
|
| 1631 |
+
"mean_input_tokens": 258086.56719817768,
|
| 1632 |
+
"mean_output_tokens": 2870.0159453302963,
|
| 1633 |
+
"mean_supervised_assistant_characters": 7485.394318181819,
|
| 1634 |
+
"output_token_samples": 878,
|
| 1635 |
+
"structurally_eligible": 876,
|
| 1636 |
+
"success": 178,
|
| 1637 |
+
"success_rate_known_outcomes": 0.20227272727272727,
|
| 1638 |
+
"training_eligible": 0,
|
| 1639 |
+
"trajectories": 880,
|
| 1640 |
+
"unique_tasks": 89,
|
| 1641 |
+
"unknown": 0
|
| 1642 |
+
},
|
| 1643 |
+
"minimax-m2.1@minimax": {
|
| 1644 |
+
"cost_samples": 440,
|
| 1645 |
+
"duration_samples": 440,
|
| 1646 |
+
"failure": 277,
|
| 1647 |
+
"input_token_samples": 0,
|
| 1648 |
+
"mean_actions": 47.49318181818182,
|
| 1649 |
+
"mean_cost_source_native_units": 0.0,
|
| 1650 |
+
"mean_duration_seconds": 779.6249267968265,
|
| 1651 |
+
"mean_events": 85.37727272727273,
|
| 1652 |
+
"mean_input_tokens": null,
|
| 1653 |
+
"mean_output_tokens": null,
|
| 1654 |
+
"mean_supervised_assistant_characters": 8243.768181818183,
|
| 1655 |
+
"output_token_samples": 0,
|
| 1656 |
+
"structurally_eligible": 1,
|
| 1657 |
+
"success": 163,
|
| 1658 |
+
"success_rate_known_outcomes": 0.3704545454545455,
|
| 1659 |
+
"training_eligible": 0,
|
| 1660 |
+
"trajectories": 440,
|
| 1661 |
+
"unique_tasks": 88,
|
| 1662 |
+
"unknown": 0
|
| 1663 |
+
},
|
| 1664 |
+
"minimax-m2.1@openai": {
|
| 1665 |
+
"cost_samples": 438,
|
| 1666 |
+
"duration_samples": 438,
|
| 1667 |
+
"failure": 308,
|
| 1668 |
+
"input_token_samples": 0,
|
| 1669 |
+
"mean_actions": 30.929223744292237,
|
| 1670 |
+
"mean_cost_source_native_units": 0.0,
|
| 1671 |
+
"mean_duration_seconds": 916.7981030538202,
|
| 1672 |
+
"mean_events": 62.858447488584474,
|
| 1673 |
+
"mean_input_tokens": null,
|
| 1674 |
+
"mean_output_tokens": null,
|
| 1675 |
+
"mean_supervised_assistant_characters": 19007.906392694065,
|
| 1676 |
+
"output_token_samples": 0,
|
| 1677 |
+
"structurally_eligible": 438,
|
| 1678 |
+
"success": 130,
|
| 1679 |
+
"success_rate_known_outcomes": 0.2968036529680365,
|
| 1680 |
+
"training_eligible": 0,
|
| 1681 |
+
"trajectories": 438,
|
| 1682 |
+
"unique_tasks": 89,
|
| 1683 |
+
"unknown": 0
|
| 1684 |
+
},
|
| 1685 |
+
"minimax-m2.5-high": {
|
| 1686 |
+
"cost_samples": 0,
|
| 1687 |
+
"duration_samples": 0,
|
| 1688 |
+
"failure": 121,
|
| 1689 |
+
"input_token_samples": 0,
|
| 1690 |
+
"mean_actions": 60.438,
|
| 1691 |
+
"mean_cost_source_native_units": null,
|
| 1692 |
+
"mean_duration_seconds": null,
|
| 1693 |
+
"mean_events": 121.894,
|
| 1694 |
+
"mean_input_tokens": null,
|
| 1695 |
+
"mean_output_tokens": null,
|
| 1696 |
+
"mean_supervised_assistant_characters": 1597.206,
|
| 1697 |
+
"output_token_samples": 0,
|
| 1698 |
+
"structurally_eligible": 500,
|
| 1699 |
+
"success": 379,
|
| 1700 |
+
"success_rate_known_outcomes": 0.758,
|
| 1701 |
+
"training_eligible": 0,
|
| 1702 |
+
"trajectories": 500,
|
| 1703 |
+
"unique_tasks": 500,
|
| 1704 |
+
"unknown": 0
|
| 1705 |
+
},
|
| 1706 |
+
"minimax-m2.5@Minimax": {
|
| 1707 |
+
"cost_samples": 411,
|
| 1708 |
+
"duration_samples": 411,
|
| 1709 |
+
"failure": 230,
|
| 1710 |
+
"input_token_samples": 411,
|
| 1711 |
+
"mean_actions": 29.394160583941606,
|
| 1712 |
+
"mean_cost_source_native_units": 0.0,
|
| 1713 |
+
"mean_duration_seconds": 1034.1070254434924,
|
| 1714 |
+
"mean_events": 60.02676399026764,
|
| 1715 |
+
"mean_input_tokens": 181155.46228710463,
|
| 1716 |
+
"mean_output_tokens": 12802.311435523114,
|
| 1717 |
+
"mean_supervised_assistant_characters": 12416.776155717762,
|
| 1718 |
+
"output_token_samples": 411,
|
| 1719 |
+
"structurally_eligible": 410,
|
| 1720 |
+
"success": 181,
|
| 1721 |
+
"success_rate_known_outcomes": 0.44038929440389296,
|
| 1722 |
+
"training_eligible": 0,
|
| 1723 |
+
"trajectories": 411,
|
| 1724 |
+
"unique_tasks": 88,
|
| 1725 |
+
"unknown": 0
|
| 1726 |
+
},
|
| 1727 |
+
"moonshotai/Kimi-K2-Instruct-0905@together_ai": {
|
| 1728 |
+
"cost_samples": 1783,
|
| 1729 |
+
"duration_samples": 1783,
|
| 1730 |
+
"failure": 1287,
|
| 1731 |
+
"input_token_samples": 1783,
|
| 1732 |
+
"mean_actions": 69.88951205832866,
|
| 1733 |
+
"mean_cost_source_native_units": 117.50347072651596,
|
| 1734 |
+
"mean_duration_seconds": 679.7632403584992,
|
| 1735 |
+
"mean_events": 114.4823331463825,
|
| 1736 |
+
"mean_input_tokens": 1144773.9444756028,
|
| 1737 |
+
"mean_output_tokens": 10086.920919798093,
|
| 1738 |
+
"mean_supervised_assistant_characters": 29183.99046550757,
|
| 1739 |
+
"output_token_samples": 1783,
|
| 1740 |
+
"structurally_eligible": 923,
|
| 1741 |
+
"success": 496,
|
| 1742 |
+
"success_rate_known_outcomes": 0.2781828379136287,
|
| 1743 |
+
"training_eligible": 0,
|
| 1744 |
+
"trajectories": 1783,
|
| 1745 |
+
"unique_tasks": 89,
|
| 1746 |
+
"unknown": 0
|
| 1747 |
+
},
|
| 1748 |
+
"moonshotai/Kimi-K2-Thinking@together_ai": {
|
| 1749 |
+
"cost_samples": 892,
|
| 1750 |
+
"duration_samples": 892,
|
| 1751 |
+
"failure": 572,
|
| 1752 |
+
"input_token_samples": 890,
|
| 1753 |
+
"mean_actions": 34.10538116591928,
|
| 1754 |
+
"mean_cost_source_native_units": 119.29089524606952,
|
| 1755 |
+
"mean_duration_seconds": 971.4054908068191,
|
| 1756 |
+
"mean_events": 70.29820627802691,
|
| 1757 |
+
"mean_input_tokens": 949991.6898876404,
|
| 1758 |
+
"mean_output_tokens": 18533.206741573034,
|
| 1759 |
+
"mean_supervised_assistant_characters": 29966.461883408072,
|
| 1760 |
+
"output_token_samples": 890,
|
| 1761 |
+
"structurally_eligible": 892,
|
| 1762 |
+
"success": 320,
|
| 1763 |
+
"success_rate_known_outcomes": 0.35874439461883406,
|
| 1764 |
+
"training_eligible": 0,
|
| 1765 |
+
"trajectories": 892,
|
| 1766 |
+
"unique_tasks": 89,
|
| 1767 |
+
"unknown": 0
|
| 1768 |
+
},
|
| 1769 |
+
"openai/gpt-4.1": {
|
| 1770 |
+
"cost_samples": 0,
|
| 1771 |
+
"duration_samples": 0,
|
| 1772 |
+
"failure": 153,
|
| 1773 |
+
"input_token_samples": 0,
|
| 1774 |
+
"mean_actions": 14.446043165467627,
|
| 1775 |
+
"mean_cost_source_native_units": null,
|
| 1776 |
+
"mean_duration_seconds": null,
|
| 1777 |
+
"mean_events": 40.118705035971225,
|
| 1778 |
+
"mean_input_tokens": null,
|
| 1779 |
+
"mean_output_tokens": null,
|
| 1780 |
+
"mean_supervised_assistant_characters": 3550.039568345324,
|
| 1781 |
+
"output_token_samples": 0,
|
| 1782 |
+
"structurally_eligible": 278,
|
| 1783 |
+
"success": 125,
|
| 1784 |
+
"success_rate_known_outcomes": 0.44964028776978415,
|
| 1785 |
+
"training_eligible": 278,
|
| 1786 |
+
"trajectories": 278,
|
| 1787 |
+
"unique_tasks": 228,
|
| 1788 |
+
"unknown": 0
|
| 1789 |
+
},
|
| 1790 |
+
"openai/gpt-4.1-mini": {
|
| 1791 |
+
"cost_samples": 0,
|
| 1792 |
+
"duration_samples": 0,
|
| 1793 |
+
"failure": 158,
|
| 1794 |
+
"input_token_samples": 0,
|
| 1795 |
+
"mean_actions": 19.776978417266186,
|
| 1796 |
+
"mean_cost_source_native_units": null,
|
| 1797 |
+
"mean_duration_seconds": null,
|
| 1798 |
+
"mean_events": 50.32014388489208,
|
| 1799 |
+
"mean_input_tokens": null,
|
| 1800 |
+
"mean_output_tokens": null,
|
| 1801 |
+
"mean_supervised_assistant_characters": 3496.9460431654675,
|
| 1802 |
+
"output_token_samples": 0,
|
| 1803 |
+
"structurally_eligible": 278,
|
| 1804 |
+
"success": 120,
|
| 1805 |
+
"success_rate_known_outcomes": 0.4316546762589928,
|
| 1806 |
+
"training_eligible": 278,
|
| 1807 |
+
"trajectories": 278,
|
| 1808 |
+
"unique_tasks": 228,
|
| 1809 |
+
"unknown": 0
|
| 1810 |
+
},
|
| 1811 |
+
"openai/gpt-4.1-nano": {
|
| 1812 |
+
"cost_samples": 0,
|
| 1813 |
+
"duration_samples": 0,
|
| 1814 |
+
"failure": 238,
|
| 1815 |
+
"input_token_samples": 0,
|
| 1816 |
+
"mean_actions": 33.71223021582734,
|
| 1817 |
+
"mean_cost_source_native_units": null,
|
| 1818 |
+
"mean_duration_seconds": null,
|
| 1819 |
+
"mean_events": 77.18705035971223,
|
| 1820 |
+
"mean_input_tokens": null,
|
| 1821 |
+
"mean_output_tokens": null,
|
| 1822 |
+
"mean_supervised_assistant_characters": 4114.230215827338,
|
| 1823 |
+
"output_token_samples": 0,
|
| 1824 |
+
"structurally_eligible": 278,
|
| 1825 |
+
"success": 40,
|
| 1826 |
+
"success_rate_known_outcomes": 0.14388489208633093,
|
| 1827 |
+
"training_eligible": 278,
|
| 1828 |
+
"trajectories": 278,
|
| 1829 |
+
"unique_tasks": 228,
|
| 1830 |
+
"unknown": 0
|
| 1831 |
+
},
|
| 1832 |
+
"openai/gpt-4o-20240806": {
|
| 1833 |
+
"cost_samples": 0,
|
| 1834 |
+
"duration_samples": 0,
|
| 1835 |
+
"failure": 170,
|
| 1836 |
+
"input_token_samples": 0,
|
| 1837 |
+
"mean_actions": 17.758992805755394,
|
| 1838 |
+
"mean_cost_source_native_units": null,
|
| 1839 |
+
"mean_duration_seconds": null,
|
| 1840 |
+
"mean_events": 48.302158273381295,
|
| 1841 |
+
"mean_input_tokens": null,
|
| 1842 |
+
"mean_output_tokens": null,
|
| 1843 |
+
"mean_supervised_assistant_characters": 3614.1582733812947,
|
| 1844 |
+
"output_token_samples": 0,
|
| 1845 |
+
"structurally_eligible": 278,
|
| 1846 |
+
"success": 108,
|
| 1847 |
+
"success_rate_known_outcomes": 0.38848920863309355,
|
| 1848 |
+
"training_eligible": 278,
|
| 1849 |
+
"trajectories": 278,
|
| 1850 |
+
"unique_tasks": 228,
|
| 1851 |
+
"unknown": 0
|
| 1852 |
+
},
|
| 1853 |
+
"openai/gpt-4o-mini": {
|
| 1854 |
+
"cost_samples": 0,
|
| 1855 |
+
"duration_samples": 0,
|
| 1856 |
+
"failure": 205,
|
| 1857 |
+
"input_token_samples": 0,
|
| 1858 |
+
"mean_actions": 45.21223021582734,
|
| 1859 |
+
"mean_cost_source_native_units": null,
|
| 1860 |
+
"mean_duration_seconds": null,
|
| 1861 |
+
"mean_events": 93.01079136690647,
|
| 1862 |
+
"mean_input_tokens": null,
|
| 1863 |
+
"mean_output_tokens": null,
|
| 1864 |
+
"mean_supervised_assistant_characters": 1972.478417266187,
|
| 1865 |
+
"output_token_samples": 0,
|
| 1866 |
+
"structurally_eligible": 278,
|
| 1867 |
+
"success": 73,
|
| 1868 |
+
"success_rate_known_outcomes": 0.26258992805755393,
|
| 1869 |
+
"training_eligible": 278,
|
| 1870 |
+
"trajectories": 278,
|
| 1871 |
+
"unique_tasks": 228,
|
| 1872 |
+
"unknown": 0
|
| 1873 |
+
},
|
| 1874 |
+
"openai/gpt-5": {
|
| 1875 |
+
"cost_samples": 0,
|
| 1876 |
+
"duration_samples": 0,
|
| 1877 |
+
"failure": 138,
|
| 1878 |
+
"input_token_samples": 0,
|
| 1879 |
+
"mean_actions": 12.305755395683454,
|
| 1880 |
+
"mean_cost_source_native_units": null,
|
| 1881 |
+
"mean_duration_seconds": null,
|
| 1882 |
+
"mean_events": 40.26618705035971,
|
| 1883 |
+
"mean_input_tokens": null,
|
| 1884 |
+
"mean_output_tokens": null,
|
| 1885 |
+
"mean_supervised_assistant_characters": 4689.446043165468,
|
| 1886 |
+
"output_token_samples": 0,
|
| 1887 |
+
"structurally_eligible": 278,
|
| 1888 |
+
"success": 140,
|
| 1889 |
+
"success_rate_known_outcomes": 0.5035971223021583,
|
| 1890 |
+
"training_eligible": 278,
|
| 1891 |
+
"trajectories": 278,
|
| 1892 |
+
"unique_tasks": 228,
|
| 1893 |
+
"unknown": 0
|
| 1894 |
+
},
|
| 1895 |
+
"openai/gpt-5-nano": {
|
| 1896 |
+
"cost_samples": 0,
|
| 1897 |
+
"duration_samples": 0,
|
| 1898 |
+
"failure": 183,
|
| 1899 |
+
"input_token_samples": 0,
|
| 1900 |
+
"mean_actions": 24.07913669064748,
|
| 1901 |
+
"mean_cost_source_native_units": null,
|
| 1902 |
+
"mean_duration_seconds": null,
|
| 1903 |
+
"mean_events": 61.83453237410072,
|
| 1904 |
+
"mean_input_tokens": null,
|
| 1905 |
+
"mean_output_tokens": null,
|
| 1906 |
+
"mean_supervised_assistant_characters": 8785.273381294965,
|
| 1907 |
+
"output_token_samples": 0,
|
| 1908 |
+
"structurally_eligible": 278,
|
| 1909 |
+
"success": 95,
|
| 1910 |
+
"success_rate_known_outcomes": 0.34172661870503596,
|
| 1911 |
+
"training_eligible": 278,
|
| 1912 |
+
"trajectories": 278,
|
| 1913 |
+
"unique_tasks": 228,
|
| 1914 |
+
"unknown": 0
|
| 1915 |
+
},
|
| 1916 |
+
"openai/gpt-oss-120b@together_ai": {
|
| 1917 |
+
"cost_samples": 1805,
|
| 1918 |
+
"duration_samples": 1805,
|
| 1919 |
+
"failure": 1502,
|
| 1920 |
+
"input_token_samples": 1801,
|
| 1921 |
+
"mean_actions": 10.353462603878116,
|
| 1922 |
+
"mean_cost_source_native_units": 2.2321994562228302,
|
| 1923 |
+
"mean_duration_seconds": 383.20367774012345,
|
| 1924 |
+
"mean_events": 22.829362880886425,
|
| 1925 |
+
"mean_input_tokens": 118470.27984453081,
|
| 1926 |
+
"mean_output_tokens": 7668.382565241533,
|
| 1927 |
+
"mean_supervised_assistant_characters": 10867.516897506925,
|
| 1928 |
+
"output_token_samples": 1801,
|
| 1929 |
+
"structurally_eligible": 1789,
|
| 1930 |
+
"success": 303,
|
| 1931 |
+
"success_rate_known_outcomes": 0.16786703601108033,
|
| 1932 |
+
"training_eligible": 0,
|
| 1933 |
+
"trajectories": 1805,
|
| 1934 |
+
"unique_tasks": 89,
|
| 1935 |
+
"unknown": 0
|
| 1936 |
+
},
|
| 1937 |
+
"openai/gpt-oss-20b@together_ai": {
|
| 1938 |
+
"cost_samples": 1807,
|
| 1939 |
+
"duration_samples": 1807,
|
| 1940 |
+
"failure": 1749,
|
| 1941 |
+
"input_token_samples": 1805,
|
| 1942 |
+
"mean_actions": 22.045379081350305,
|
| 1943 |
+
"mean_cost_source_native_units": 2.7351576161561053,
|
| 1944 |
+
"mean_duration_seconds": 225.4812311870193,
|
| 1945 |
+
"mean_events": 47.460431654676256,
|
| 1946 |
+
"mean_input_tokens": 503665.8925207756,
|
| 1947 |
+
"mean_output_tokens": 10992.939612188366,
|
| 1948 |
+
"mean_supervised_assistant_characters": 16832.06917542889,
|
| 1949 |
+
"output_token_samples": 1805,
|
| 1950 |
+
"structurally_eligible": 1745,
|
| 1951 |
+
"success": 58,
|
| 1952 |
+
"success_rate_known_outcomes": 0.032097399003873824,
|
| 1953 |
+
"training_eligible": 0,
|
| 1954 |
+
"trajectories": 1807,
|
| 1955 |
+
"unique_tasks": 89,
|
| 1956 |
+
"unknown": 0
|
| 1957 |
+
},
|
| 1958 |
+
"openai/o3-high": {
|
| 1959 |
+
"cost_samples": 0,
|
| 1960 |
+
"duration_samples": 0,
|
| 1961 |
+
"failure": 136,
|
| 1962 |
+
"input_token_samples": 0,
|
| 1963 |
+
"mean_actions": 15.010791366906474,
|
| 1964 |
+
"mean_cost_source_native_units": null,
|
| 1965 |
+
"mean_duration_seconds": null,
|
| 1966 |
+
"mean_events": 42.780575539568346,
|
| 1967 |
+
"mean_input_tokens": null,
|
| 1968 |
+
"mean_output_tokens": null,
|
| 1969 |
+
"mean_supervised_assistant_characters": 3309.9172661870502,
|
| 1970 |
+
"output_token_samples": 0,
|
| 1971 |
+
"structurally_eligible": 278,
|
| 1972 |
+
"success": 142,
|
| 1973 |
+
"success_rate_known_outcomes": 0.5107913669064749,
|
| 1974 |
+
"training_eligible": 278,
|
| 1975 |
+
"trajectories": 278,
|
| 1976 |
+
"unique_tasks": 228,
|
| 1977 |
+
"unknown": 0
|
| 1978 |
+
},
|
| 1979 |
+
"openai/o4-mini-high": {
|
| 1980 |
+
"cost_samples": 0,
|
| 1981 |
+
"duration_samples": 0,
|
| 1982 |
+
"failure": 162,
|
| 1983 |
+
"input_token_samples": 0,
|
| 1984 |
+
"mean_actions": 14.733812949640289,
|
| 1985 |
+
"mean_cost_source_native_units": null,
|
| 1986 |
+
"mean_duration_seconds": null,
|
| 1987 |
+
"mean_events": 39.410071942446045,
|
| 1988 |
+
"mean_input_tokens": null,
|
| 1989 |
+
"mean_output_tokens": null,
|
| 1990 |
+
"mean_supervised_assistant_characters": 2741.8057553956833,
|
| 1991 |
+
"output_token_samples": 0,
|
| 1992 |
+
"structurally_eligible": 278,
|
| 1993 |
+
"success": 116,
|
| 1994 |
+
"success_rate_known_outcomes": 0.4172661870503597,
|
| 1995 |
+
"training_eligible": 278,
|
| 1996 |
+
"trajectories": 278,
|
| 1997 |
+
"unique_tasks": 228,
|
| 1998 |
+
"unknown": 0
|
| 1999 |
+
},
|
| 2000 |
+
"togetherai/Qwen/Qwen3-235B-A22B-FP8": {
|
| 2001 |
+
"cost_samples": 0,
|
| 2002 |
+
"duration_samples": 0,
|
| 2003 |
+
"failure": 181,
|
| 2004 |
+
"input_token_samples": 0,
|
| 2005 |
+
"mean_actions": 14.18705035971223,
|
| 2006 |
+
"mean_cost_source_native_units": null,
|
| 2007 |
+
"mean_duration_seconds": null,
|
| 2008 |
+
"mean_events": 35.55755395683453,
|
| 2009 |
+
"mean_input_tokens": null,
|
| 2010 |
+
"mean_output_tokens": null,
|
| 2011 |
+
"mean_supervised_assistant_characters": 3833.568345323741,
|
| 2012 |
+
"output_token_samples": 0,
|
| 2013 |
+
"structurally_eligible": 278,
|
| 2014 |
+
"success": 97,
|
| 2015 |
+
"success_rate_known_outcomes": 0.3489208633093525,
|
| 2016 |
+
"training_eligible": 278,
|
| 2017 |
+
"trajectories": 278,
|
| 2018 |
+
"unique_tasks": 228,
|
| 2019 |
+
"unknown": 0
|
| 2020 |
+
},
|
| 2021 |
+
"togetherai/Qwen/Qwen3-235B-A22B-Instruct-2507-FP8": {
|
| 2022 |
+
"cost_samples": 0,
|
| 2023 |
+
"duration_samples": 0,
|
| 2024 |
+
"failure": 154,
|
| 2025 |
+
"input_token_samples": 0,
|
| 2026 |
+
"mean_actions": 16.37410071942446,
|
| 2027 |
+
"mean_cost_source_native_units": null,
|
| 2028 |
+
"mean_duration_seconds": null,
|
| 2029 |
+
"mean_events": 41.82374100719424,
|
| 2030 |
+
"mean_input_tokens": null,
|
| 2031 |
+
"mean_output_tokens": null,
|
| 2032 |
+
"mean_supervised_assistant_characters": 3558.928057553957,
|
| 2033 |
+
"output_token_samples": 0,
|
| 2034 |
+
"structurally_eligible": 278,
|
| 2035 |
+
"success": 124,
|
| 2036 |
+
"success_rate_known_outcomes": 0.4460431654676259,
|
| 2037 |
+
"training_eligible": 278,
|
| 2038 |
+
"trajectories": 278,
|
| 2039 |
+
"unique_tasks": 228,
|
| 2040 |
+
"unknown": 0
|
| 2041 |
+
},
|
| 2042 |
+
"togetherai/Qwen/Qwen3-235B-A22B-Thinking-2507-FP8": {
|
| 2043 |
+
"cost_samples": 0,
|
| 2044 |
+
"duration_samples": 0,
|
| 2045 |
+
"failure": 162,
|
| 2046 |
+
"input_token_samples": 0,
|
| 2047 |
+
"mean_actions": 15.913669064748202,
|
| 2048 |
+
"mean_cost_source_native_units": null,
|
| 2049 |
+
"mean_duration_seconds": null,
|
| 2050 |
+
"mean_events": 40.17985611510792,
|
| 2051 |
+
"mean_input_tokens": null,
|
| 2052 |
+
"mean_output_tokens": null,
|
| 2053 |
+
"mean_supervised_assistant_characters": 7153.723021582734,
|
| 2054 |
+
"output_token_samples": 0,
|
| 2055 |
+
"structurally_eligible": 278,
|
| 2056 |
+
"success": 116,
|
| 2057 |
+
"success_rate_known_outcomes": 0.4172661870503597,
|
| 2058 |
+
"training_eligible": 278,
|
| 2059 |
+
"trajectories": 278,
|
| 2060 |
+
"unique_tasks": 228,
|
| 2061 |
+
"unknown": 0
|
| 2062 |
+
},
|
| 2063 |
+
"togetherai/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8": {
|
| 2064 |
+
"cost_samples": 0,
|
| 2065 |
+
"duration_samples": 0,
|
| 2066 |
+
"failure": 218,
|
| 2067 |
+
"input_token_samples": 0,
|
| 2068 |
+
"mean_actions": 14.129496402877697,
|
| 2069 |
+
"mean_cost_source_native_units": null,
|
| 2070 |
+
"mean_duration_seconds": null,
|
| 2071 |
+
"mean_events": 38.697841726618705,
|
| 2072 |
+
"mean_input_tokens": null,
|
| 2073 |
+
"mean_output_tokens": null,
|
| 2074 |
+
"mean_supervised_assistant_characters": 3825.9172661870502,
|
| 2075 |
+
"output_token_samples": 0,
|
| 2076 |
+
"structurally_eligible": 278,
|
| 2077 |
+
"success": 60,
|
| 2078 |
+
"success_rate_known_outcomes": 0.2158273381294964,
|
| 2079 |
+
"training_eligible": 278,
|
| 2080 |
+
"trajectories": 278,
|
| 2081 |
+
"unique_tasks": 228,
|
| 2082 |
+
"unknown": 0
|
| 2083 |
+
},
|
| 2084 |
+
"togetherai/moonshotai/Kimi-K2-Instruct": {
|
| 2085 |
+
"cost_samples": 0,
|
| 2086 |
+
"duration_samples": 0,
|
| 2087 |
+
"failure": 141,
|
| 2088 |
+
"input_token_samples": 0,
|
| 2089 |
+
"mean_actions": 14.881294964028777,
|
| 2090 |
+
"mean_cost_source_native_units": null,
|
| 2091 |
+
"mean_duration_seconds": null,
|
| 2092 |
+
"mean_events": 40.35251798561151,
|
| 2093 |
+
"mean_input_tokens": null,
|
| 2094 |
+
"mean_output_tokens": null,
|
| 2095 |
+
"mean_supervised_assistant_characters": 5261.179856115108,
|
| 2096 |
+
"output_token_samples": 0,
|
| 2097 |
+
"structurally_eligible": 276,
|
| 2098 |
+
"success": 137,
|
| 2099 |
+
"success_rate_known_outcomes": 0.49280575539568344,
|
| 2100 |
+
"training_eligible": 276,
|
| 2101 |
+
"trajectories": 278,
|
| 2102 |
+
"unique_tasks": 228,
|
| 2103 |
+
"unknown": 0
|
| 2104 |
+
},
|
| 2105 |
+
"togetherai/moonshotai/Kimi-K2-Instruct-0905": {
|
| 2106 |
+
"cost_samples": 0,
|
| 2107 |
+
"duration_samples": 0,
|
| 2108 |
+
"failure": 139,
|
| 2109 |
+
"input_token_samples": 0,
|
| 2110 |
+
"mean_actions": 16.007194244604317,
|
| 2111 |
+
"mean_cost_source_native_units": null,
|
| 2112 |
+
"mean_duration_seconds": null,
|
| 2113 |
+
"mean_events": 42.42446043165467,
|
| 2114 |
+
"mean_input_tokens": null,
|
| 2115 |
+
"mean_output_tokens": null,
|
| 2116 |
+
"mean_supervised_assistant_characters": 5449.546762589928,
|
| 2117 |
+
"output_token_samples": 0,
|
| 2118 |
+
"structurally_eligible": 278,
|
| 2119 |
+
"success": 139,
|
| 2120 |
+
"success_rate_known_outcomes": 0.5,
|
| 2121 |
+
"training_eligible": 278,
|
| 2122 |
+
"trajectories": 278,
|
| 2123 |
+
"unique_tasks": 228,
|
| 2124 |
+
"unknown": 0
|
| 2125 |
+
},
|
| 2126 |
+
"togetherai/openai/gpt-oss-120b": {
|
| 2127 |
+
"cost_samples": 0,
|
| 2128 |
+
"duration_samples": 0,
|
| 2129 |
+
"failure": 221,
|
| 2130 |
+
"input_token_samples": 0,
|
| 2131 |
+
"mean_actions": 13.420863309352518,
|
| 2132 |
+
"mean_cost_source_native_units": null,
|
| 2133 |
+
"mean_duration_seconds": null,
|
| 2134 |
+
"mean_events": 35.97122302158273,
|
| 2135 |
+
"mean_input_tokens": null,
|
| 2136 |
+
"mean_output_tokens": null,
|
| 2137 |
+
"mean_supervised_assistant_characters": 14511.31654676259,
|
| 2138 |
+
"output_token_samples": 0,
|
| 2139 |
+
"structurally_eligible": 278,
|
| 2140 |
+
"success": 57,
|
| 2141 |
+
"success_rate_known_outcomes": 0.20503597122302158,
|
| 2142 |
+
"training_eligible": 278,
|
| 2143 |
+
"trajectories": 278,
|
| 2144 |
+
"unique_tasks": 228,
|
| 2145 |
+
"unknown": 0
|
| 2146 |
+
},
|
| 2147 |
+
"togetherai/openai/gpt-oss-20b": {
|
| 2148 |
+
"cost_samples": 0,
|
| 2149 |
+
"duration_samples": 0,
|
| 2150 |
+
"failure": 248,
|
| 2151 |
+
"input_token_samples": 0,
|
| 2152 |
+
"mean_actions": 45.29136690647482,
|
| 2153 |
+
"mean_cost_source_native_units": null,
|
| 2154 |
+
"mean_duration_seconds": null,
|
| 2155 |
+
"mean_events": 100.81294964028777,
|
| 2156 |
+
"mean_input_tokens": null,
|
| 2157 |
+
"mean_output_tokens": null,
|
| 2158 |
+
"mean_supervised_assistant_characters": 29205.543165467625,
|
| 2159 |
+
"output_token_samples": 0,
|
| 2160 |
+
"structurally_eligible": 278,
|
| 2161 |
+
"success": 30,
|
| 2162 |
+
"success_rate_known_outcomes": 0.1079136690647482,
|
| 2163 |
+
"training_eligible": 278,
|
| 2164 |
+
"trajectories": 278,
|
| 2165 |
+
"unique_tasks": 228,
|
| 2166 |
+
"unknown": 0
|
| 2167 |
+
},
|
| 2168 |
+
"unknown": {
|
| 2169 |
+
"cost_samples": 0,
|
| 2170 |
+
"duration_samples": 0,
|
| 2171 |
+
"failure": 0,
|
| 2172 |
+
"input_token_samples": 0,
|
| 2173 |
+
"mean_actions": 16.45260465116279,
|
| 2174 |
+
"mean_cost_source_native_units": null,
|
| 2175 |
+
"mean_duration_seconds": null,
|
| 2176 |
+
"mean_events": 32.974976744186044,
|
| 2177 |
+
"mean_input_tokens": null,
|
| 2178 |
+
"mean_output_tokens": null,
|
| 2179 |
+
"mean_supervised_assistant_characters": 10075.690186046511,
|
| 2180 |
+
"output_token_samples": 0,
|
| 2181 |
+
"structurally_eligible": 20643,
|
| 2182 |
+
"success": 491,
|
| 2183 |
+
"success_rate_known_outcomes": 1.0,
|
| 2184 |
+
"training_eligible": 20643,
|
| 2185 |
+
"trajectories": 21500,
|
| 2186 |
+
"unique_tasks": 19089,
|
| 2187 |
+
"unknown": 21009
|
| 2188 |
+
}
|
| 2189 |
+
},
|
| 2190 |
+
"by_outcome": {
|
| 2191 |
+
"failure": {
|
| 2192 |
+
"cost_samples": 23212,
|
| 2193 |
+
"duration_samples": 27064,
|
| 2194 |
+
"failure": 79973,
|
| 2195 |
+
"input_token_samples": 21454,
|
| 2196 |
+
"mean_actions": 45.62608630412765,
|
| 2197 |
+
"mean_cost_source_native_units": 75.64188306777397,
|
| 2198 |
+
"mean_duration_seconds": 701.5190378383018,
|
| 2199 |
+
"mean_events": 94.278981656309,
|
| 2200 |
+
"mean_input_tokens": 1156962.542835835,
|
| 2201 |
+
"mean_output_tokens": 21745.191386221682,
|
| 2202 |
+
"mean_supervised_assistant_characters": 12323.303715003814,
|
| 2203 |
+
"output_token_samples": 21454,
|
| 2204 |
+
"structurally_eligible": 73054,
|
| 2205 |
+
"success": 0,
|
| 2206 |
+
"success_rate_known_outcomes": 0.0,
|
| 2207 |
+
"training_eligible": 51532,
|
| 2208 |
+
"trajectories": 79973,
|
| 2209 |
+
"unique_tasks": 28927,
|
| 2210 |
+
"unknown": 0
|
| 2211 |
+
},
|
| 2212 |
+
"success": {
|
| 2213 |
+
"cost_samples": 11051,
|
| 2214 |
+
"duration_samples": 12262,
|
| 2215 |
+
"failure": 0,
|
| 2216 |
+
"input_token_samples": 9161,
|
| 2217 |
+
"mean_actions": 36.86015544423162,
|
| 2218 |
+
"mean_cost_source_native_units": 31.06763293005613,
|
| 2219 |
+
"mean_duration_seconds": 410.78903144437595,
|
| 2220 |
+
"mean_events": 76.32702224841609,
|
| 2221 |
+
"mean_input_tokens": 382399.21984499507,
|
| 2222 |
+
"mean_output_tokens": 11283.977950005457,
|
| 2223 |
+
"mean_supervised_assistant_characters": 7551.789303079417,
|
| 2224 |
+
"output_token_samples": 9161,
|
| 2225 |
+
"structurally_eligible": 50793,
|
| 2226 |
+
"success": 54296,
|
| 2227 |
+
"success_rate_known_outcomes": 1.0,
|
| 2228 |
+
"training_eligible": 38524,
|
| 2229 |
+
"trajectories": 54296,
|
| 2230 |
+
"unique_tasks": 23353,
|
| 2231 |
+
"unknown": 0
|
| 2232 |
+
},
|
| 2233 |
+
"unknown": {
|
| 2234 |
+
"cost_samples": 0,
|
| 2235 |
+
"duration_samples": 0,
|
| 2236 |
+
"failure": 0,
|
| 2237 |
+
"input_token_samples": 0,
|
| 2238 |
+
"mean_actions": 16.3787424437146,
|
| 2239 |
+
"mean_cost_source_native_units": null,
|
| 2240 |
+
"mean_duration_seconds": null,
|
| 2241 |
+
"mean_events": 32.805511923461374,
|
| 2242 |
+
"mean_input_tokens": null,
|
| 2243 |
+
"mean_output_tokens": null,
|
| 2244 |
+
"mean_supervised_assistant_characters": 10051.800371269455,
|
| 2245 |
+
"output_token_samples": 0,
|
| 2246 |
+
"structurally_eligible": 20152,
|
| 2247 |
+
"success": 0,
|
| 2248 |
+
"success_rate_known_outcomes": null,
|
| 2249 |
+
"training_eligible": 20152,
|
| 2250 |
+
"trajectories": 21009,
|
| 2251 |
+
"unique_tasks": 18796,
|
| 2252 |
+
"unknown": 21009
|
| 2253 |
+
}
|
| 2254 |
+
},
|
| 2255 |
+
"by_quality_tier": {
|
| 2256 |
+
"3": {
|
| 2257 |
+
"cost_samples": 7668,
|
| 2258 |
+
"duration_samples": 8879,
|
| 2259 |
+
"failure": 0,
|
| 2260 |
+
"input_token_samples": 6699,
|
| 2261 |
+
"mean_actions": 35.69987645264661,
|
| 2262 |
+
"mean_cost_source_native_units": 23.312137899828038,
|
| 2263 |
+
"mean_duration_seconds": 387.87317761634836,
|
| 2264 |
+
"mean_events": 75.29464113354696,
|
| 2265 |
+
"mean_input_tokens": 242325.22913867742,
|
| 2266 |
+
"mean_output_tokens": 11504.635617256306,
|
| 2267 |
+
"mean_supervised_assistant_characters": 7296.183294081309,
|
| 2268 |
+
"output_token_samples": 6699,
|
| 2269 |
+
"structurally_eligible": 51802,
|
| 2270 |
+
"success": 50793,
|
| 2271 |
+
"success_rate_known_outcomes": 1.0,
|
| 2272 |
+
"training_eligible": 39533,
|
| 2273 |
+
"trajectories": 51802,
|
| 2274 |
+
"unique_tasks": 24316,
|
| 2275 |
+
"unknown": 1009
|
| 2276 |
+
},
|
| 2277 |
+
"4": {
|
| 2278 |
+
"cost_samples": 0,
|
| 2279 |
+
"duration_samples": 0,
|
| 2280 |
+
"failure": 0,
|
| 2281 |
+
"input_token_samples": 0,
|
| 2282 |
+
"mean_actions": 17.524369221125216,
|
| 2283 |
+
"mean_cost_source_native_units": null,
|
| 2284 |
+
"mean_duration_seconds": null,
|
| 2285 |
+
"mean_events": 35.04873844225043,
|
| 2286 |
+
"mean_input_tokens": null,
|
| 2287 |
+
"mean_output_tokens": null,
|
| 2288 |
+
"mean_supervised_assistant_characters": 10991.896829128142,
|
| 2289 |
+
"output_token_samples": 0,
|
| 2290 |
+
"structurally_eligible": 19143,
|
| 2291 |
+
"success": 0,
|
| 2292 |
+
"success_rate_known_outcomes": null,
|
| 2293 |
+
"training_eligible": 19143,
|
| 2294 |
+
"trajectories": 19143,
|
| 2295 |
+
"unique_tasks": 16936,
|
| 2296 |
+
"unknown": 19143
|
| 2297 |
+
},
|
| 2298 |
+
"5": {
|
| 2299 |
+
"cost_samples": 26595,
|
| 2300 |
+
"duration_samples": 30447,
|
| 2301 |
+
"failure": 79973,
|
| 2302 |
+
"input_token_samples": 23916,
|
| 2303 |
+
"mean_actions": 45.172364317645524,
|
| 2304 |
+
"mean_cost_source_native_units": 72.20794615015373,
|
| 2305 |
+
"mean_duration_seconds": 675.8984596042691,
|
| 2306 |
+
"mean_events": 92.5128360191147,
|
| 2307 |
+
"mean_input_tokens": 1116461.6548335843,
|
| 2308 |
+
"mean_output_tokens": 20606.468640240844,
|
| 2309 |
+
"mean_supervised_assistant_characters": 12075.546061446883,
|
| 2310 |
+
"output_token_samples": 23916,
|
| 2311 |
+
"structurally_eligible": 73054,
|
| 2312 |
+
"success": 3503,
|
| 2313 |
+
"success_rate_known_outcomes": 0.041964157362595234,
|
| 2314 |
+
"training_eligible": 51532,
|
| 2315 |
+
"trajectories": 84333,
|
| 2316 |
+
"unique_tasks": 29808,
|
| 2317 |
+
"unknown": 857
|
| 2318 |
+
}
|
| 2319 |
+
},
|
| 2320 |
+
"by_source": {
|
| 2321 |
+
"agentsuite_tau2": {
|
| 2322 |
+
"cost_samples": 0,
|
| 2323 |
+
"duration_samples": 0,
|
| 2324 |
+
"failure": 4965,
|
| 2325 |
+
"input_token_samples": 0,
|
| 2326 |
+
"mean_actions": 17.801798561151077,
|
| 2327 |
+
"mean_cost_source_native_units": null,
|
| 2328 |
+
"mean_duration_seconds": null,
|
| 2329 |
+
"mean_events": 45.433093525179856,
|
| 2330 |
+
"mean_input_tokens": null,
|
| 2331 |
+
"mean_output_tokens": null,
|
| 2332 |
+
"mean_supervised_assistant_characters": 5750.385491606715,
|
| 2333 |
+
"output_token_samples": 0,
|
| 2334 |
+
"structurally_eligible": 8337,
|
| 2335 |
+
"success": 3375,
|
| 2336 |
+
"success_rate_known_outcomes": 0.40467625899280574,
|
| 2337 |
+
"training_eligible": 8337,
|
| 2338 |
+
"trajectories": 8340,
|
| 2339 |
+
"unique_tasks": 228,
|
| 2340 |
+
"unknown": 0
|
| 2341 |
+
},
|
| 2342 |
+
"browsecomp_rollouts": {
|
| 2343 |
+
"cost_samples": 0,
|
| 2344 |
+
"duration_samples": 5063,
|
| 2345 |
+
"failure": 3852,
|
| 2346 |
+
"input_token_samples": 0,
|
| 2347 |
+
"mean_actions": 27.13786292711831,
|
| 2348 |
+
"mean_cost_source_native_units": null,
|
| 2349 |
+
"mean_duration_seconds": 108.07624142800492,
|
| 2350 |
+
"mean_events": 54.28303377444203,
|
| 2351 |
+
"mean_input_tokens": null,
|
| 2352 |
+
"mean_output_tokens": null,
|
| 2353 |
+
"mean_supervised_assistant_characters": 15385.260912502468,
|
| 2354 |
+
"output_token_samples": 0,
|
| 2355 |
+
"structurally_eligible": 5063,
|
| 2356 |
+
"success": 1211,
|
| 2357 |
+
"success_rate_known_outcomes": 0.23918625320955955,
|
| 2358 |
+
"training_eligible": 0,
|
| 2359 |
+
"trajectories": 5063,
|
| 2360 |
+
"unique_tasks": 1266,
|
| 2361 |
+
"unknown": 0
|
| 2362 |
+
},
|
| 2363 |
+
"mind2web": {
|
| 2364 |
+
"cost_samples": 0,
|
| 2365 |
+
"duration_samples": 0,
|
| 2366 |
+
"failure": 0,
|
| 2367 |
+
"input_token_samples": 0,
|
| 2368 |
+
"mean_actions": 7.705649157581764,
|
| 2369 |
+
"mean_cost_source_native_units": null,
|
| 2370 |
+
"mean_duration_seconds": null,
|
| 2371 |
+
"mean_events": 16.411298315163528,
|
| 2372 |
+
"mean_input_tokens": null,
|
| 2373 |
+
"mean_output_tokens": null,
|
| 2374 |
+
"mean_supervised_assistant_characters": 256.4281466798811,
|
| 2375 |
+
"output_token_samples": 0,
|
| 2376 |
+
"structurally_eligible": 1009,
|
| 2377 |
+
"success": 0,
|
| 2378 |
+
"success_rate_known_outcomes": null,
|
| 2379 |
+
"training_eligible": 1009,
|
| 2380 |
+
"trajectories": 1009,
|
| 2381 |
+
"unique_tasks": 1009,
|
| 2382 |
+
"unknown": 1009
|
| 2383 |
+
},
|
| 2384 |
+
"nvidia_open_swe_traces": {
|
| 2385 |
+
"cost_samples": 0,
|
| 2386 |
+
"duration_samples": 0,
|
| 2387 |
+
"failure": 26883,
|
| 2388 |
+
"input_token_samples": 0,
|
| 2389 |
+
"mean_actions": 58.35106341765777,
|
| 2390 |
+
"mean_cost_source_native_units": null,
|
| 2391 |
+
"mean_duration_seconds": null,
|
| 2392 |
+
"mean_events": 124.11902917134778,
|
| 2393 |
+
"mean_input_tokens": null,
|
| 2394 |
+
"mean_output_tokens": null,
|
| 2395 |
+
"mean_supervised_assistant_characters": 4650.015089296091,
|
| 2396 |
+
"output_token_samples": 0,
|
| 2397 |
+
"structurally_eligible": 51626,
|
| 2398 |
+
"success": 24743,
|
| 2399 |
+
"success_rate_known_outcomes": 0.4792740092201604,
|
| 2400 |
+
"training_eligible": 51626,
|
| 2401 |
+
"trajectories": 51626,
|
| 2402 |
+
"unique_tasks": 33444,
|
| 2403 |
+
"unknown": 0
|
| 2404 |
+
},
|
| 2405 |
+
"official_tau3": {
|
| 2406 |
+
"cost_samples": 0,
|
| 2407 |
+
"duration_samples": 0,
|
| 2408 |
+
"failure": 100,
|
| 2409 |
+
"input_token_samples": 0,
|
| 2410 |
+
"mean_actions": 14.5,
|
| 2411 |
+
"mean_cost_source_native_units": null,
|
| 2412 |
+
"mean_duration_seconds": null,
|
| 2413 |
+
"mean_events": 30.0,
|
| 2414 |
+
"mean_input_tokens": null,
|
| 2415 |
+
"mean_output_tokens": null,
|
| 2416 |
+
"mean_supervised_assistant_characters": 4221.82,
|
| 2417 |
+
"output_token_samples": 0,
|
| 2418 |
+
"structurally_eligible": 0,
|
| 2419 |
+
"success": 100,
|
| 2420 |
+
"success_rate_known_outcomes": 0.5,
|
| 2421 |
+
"training_eligible": 0,
|
| 2422 |
+
"trajectories": 200,
|
| 2423 |
+
"unique_tasks": 50,
|
| 2424 |
+
"unknown": 0
|
| 2425 |
+
},
|
| 2426 |
+
"swebench_multilingual_trajectories": {
|
| 2427 |
+
"cost_samples": 0,
|
| 2428 |
+
"duration_samples": 0,
|
| 2429 |
+
"failure": 121,
|
| 2430 |
+
"input_token_samples": 0,
|
| 2431 |
+
"mean_actions": 47.37984496124031,
|
| 2432 |
+
"mean_cost_source_native_units": null,
|
| 2433 |
+
"mean_duration_seconds": null,
|
| 2434 |
+
"mean_events": 96.75968992248062,
|
| 2435 |
+
"mean_input_tokens": null,
|
| 2436 |
+
"mean_output_tokens": null,
|
| 2437 |
+
"mean_supervised_assistant_characters": 36004.64341085271,
|
| 2438 |
+
"output_token_samples": 0,
|
| 2439 |
+
"structurally_eligible": 110,
|
| 2440 |
+
"success": 8,
|
| 2441 |
+
"success_rate_known_outcomes": 0.06201550387596899,
|
| 2442 |
+
"training_eligible": 0,
|
| 2443 |
+
"trajectories": 129,
|
| 2444 |
+
"unique_tasks": 43,
|
| 2445 |
+
"unknown": 0
|
| 2446 |
+
},
|
| 2447 |
+
"swebench_verified_trajectories": {
|
| 2448 |
+
"cost_samples": 0,
|
| 2449 |
+
"duration_samples": 0,
|
| 2450 |
+
"failure": 1152,
|
| 2451 |
+
"input_token_samples": 0,
|
| 2452 |
+
"mean_actions": 47.387984154929576,
|
| 2453 |
+
"mean_cost_source_native_units": null,
|
| 2454 |
+
"mean_duration_seconds": null,
|
| 2455 |
+
"mean_events": 97.00748239436619,
|
| 2456 |
+
"mean_input_tokens": null,
|
| 2457 |
+
"mean_output_tokens": null,
|
| 2458 |
+
"mean_supervised_assistant_characters": 12330.321082746479,
|
| 2459 |
+
"output_token_samples": 0,
|
| 2460 |
+
"structurally_eligible": 4532,
|
| 2461 |
+
"success": 3392,
|
| 2462 |
+
"success_rate_known_outcomes": 0.7464788732394366,
|
| 2463 |
+
"training_eligible": 0,
|
| 2464 |
+
"trajectories": 4544,
|
| 2465 |
+
"unique_tasks": 500,
|
| 2466 |
+
"unknown": 0
|
| 2467 |
+
},
|
| 2468 |
+
"swegym_openhands": {
|
| 2469 |
+
"cost_samples": 0,
|
| 2470 |
+
"duration_samples": 0,
|
| 2471 |
+
"failure": 0,
|
| 2472 |
+
"input_token_samples": 0,
|
| 2473 |
+
"mean_actions": 19.613034623217924,
|
| 2474 |
+
"mean_cost_source_native_units": null,
|
| 2475 |
+
"mean_duration_seconds": null,
|
| 2476 |
+
"mean_events": 40.22606924643585,
|
| 2477 |
+
"mean_input_tokens": null,
|
| 2478 |
+
"mean_output_tokens": null,
|
| 2479 |
+
"mean_supervised_assistant_characters": 11097.892057026476,
|
| 2480 |
+
"output_token_samples": 0,
|
| 2481 |
+
"structurally_eligible": 491,
|
| 2482 |
+
"success": 491,
|
| 2483 |
+
"success_rate_known_outcomes": 1.0,
|
| 2484 |
+
"training_eligible": 491,
|
| 2485 |
+
"trajectories": 491,
|
| 2486 |
+
"unique_tasks": 293,
|
| 2487 |
+
"unknown": 0
|
| 2488 |
+
},
|
| 2489 |
+
"swegym_sampled": {
|
| 2490 |
+
"cost_samples": 0,
|
| 2491 |
+
"duration_samples": 0,
|
| 2492 |
+
"failure": 5011,
|
| 2493 |
+
"input_token_samples": 0,
|
| 2494 |
+
"mean_actions": 20.382406397673574,
|
| 2495 |
+
"mean_cost_source_native_units": null,
|
| 2496 |
+
"mean_duration_seconds": null,
|
| 2497 |
+
"mean_events": 42.287168302435475,
|
| 2498 |
+
"mean_input_tokens": null,
|
| 2499 |
+
"mean_output_tokens": null,
|
| 2500 |
+
"mean_supervised_assistant_characters": 7133.47528171574,
|
| 2501 |
+
"output_token_samples": 0,
|
| 2502 |
+
"structurally_eligible": 5502,
|
| 2503 |
+
"success": 491,
|
| 2504 |
+
"success_rate_known_outcomes": 0.08924027626317703,
|
| 2505 |
+
"training_eligible": 5502,
|
| 2506 |
+
"trajectories": 5502,
|
| 2507 |
+
"unique_tasks": 2438,
|
| 2508 |
+
"unknown": 0
|
| 2509 |
+
},
|
| 2510 |
+
"swesmith_trajectories": {
|
| 2511 |
+
"cost_samples": 0,
|
| 2512 |
+
"duration_samples": 0,
|
| 2513 |
+
"failure": 14673,
|
| 2514 |
+
"input_token_samples": 0,
|
| 2515 |
+
"mean_actions": 31.585477178423236,
|
| 2516 |
+
"mean_cost_source_native_units": null,
|
| 2517 |
+
"mean_duration_seconds": null,
|
| 2518 |
+
"mean_events": 64.17095435684648,
|
| 2519 |
+
"mean_input_tokens": null,
|
| 2520 |
+
"mean_output_tokens": null,
|
| 2521 |
+
"mean_supervised_assistant_characters": 17664.133236514524,
|
| 2522 |
+
"output_token_samples": 0,
|
| 2523 |
+
"structurally_eligible": 24100,
|
| 2524 |
+
"success": 9427,
|
| 2525 |
+
"success_rate_known_outcomes": 0.39116182572614105,
|
| 2526 |
+
"training_eligible": 24100,
|
| 2527 |
+
"trajectories": 24100,
|
| 2528 |
+
"unique_tasks": 10637,
|
| 2529 |
+
"unknown": 0
|
| 2530 |
+
},
|
| 2531 |
+
"terminalbench_trajectories": {
|
| 2532 |
+
"cost_samples": 34263,
|
| 2533 |
+
"duration_samples": 34263,
|
| 2534 |
+
"failure": 23212,
|
| 2535 |
+
"input_token_samples": 30615,
|
| 2536 |
+
"mean_actions": 36.32335172051484,
|
| 2537 |
+
"mean_cost_source_native_units": 61.26514903187753,
|
| 2538 |
+
"mean_duration_seconds": 685.1652316865642,
|
| 2539 |
+
"mean_events": 69.00125499810291,
|
| 2540 |
+
"mean_input_tokens": 925188.0988731014,
|
| 2541 |
+
"mean_output_tokens": 18614.85735750449,
|
| 2542 |
+
"mean_supervised_assistant_characters": 14526.395032542392,
|
| 2543 |
+
"output_token_samples": 30615,
|
| 2544 |
+
"structurally_eligible": 24086,
|
| 2545 |
+
"success": 11051,
|
| 2546 |
+
"success_rate_known_outcomes": 0.32253451244783005,
|
| 2547 |
+
"training_eligible": 0,
|
| 2548 |
+
"trajectories": 34263,
|
| 2549 |
+
"unique_tasks": 89,
|
| 2550 |
+
"unknown": 0
|
| 2551 |
+
},
|
| 2552 |
+
"terminaltraj": {
|
| 2553 |
+
"cost_samples": 0,
|
| 2554 |
+
"duration_samples": 0,
|
| 2555 |
+
"failure": 0,
|
| 2556 |
+
"input_token_samples": 0,
|
| 2557 |
+
"mean_actions": 16.8163,
|
| 2558 |
+
"mean_cost_source_native_units": null,
|
| 2559 |
+
"mean_duration_seconds": null,
|
| 2560 |
+
"mean_events": 33.6326,
|
| 2561 |
+
"mean_input_tokens": null,
|
| 2562 |
+
"mean_output_tokens": null,
|
| 2563 |
+
"mean_supervised_assistant_characters": 10545.9769,
|
| 2564 |
+
"output_token_samples": 0,
|
| 2565 |
+
"structurally_eligible": 19143,
|
| 2566 |
+
"success": 0,
|
| 2567 |
+
"success_rate_known_outcomes": null,
|
| 2568 |
+
"training_eligible": 19143,
|
| 2569 |
+
"trajectories": 20000,
|
| 2570 |
+
"unique_tasks": 17787,
|
| 2571 |
+
"unknown": 20000
|
| 2572 |
+
},
|
| 2573 |
+
"toolbench": {
|
| 2574 |
+
"cost_samples": 0,
|
| 2575 |
+
"duration_samples": 0,
|
| 2576 |
+
"failure": 4,
|
| 2577 |
+
"input_token_samples": 0,
|
| 2578 |
+
"mean_actions": 3.8181818181818183,
|
| 2579 |
+
"mean_cost_source_native_units": null,
|
| 2580 |
+
"mean_duration_seconds": null,
|
| 2581 |
+
"mean_events": 9.090909090909092,
|
| 2582 |
+
"mean_input_tokens": null,
|
| 2583 |
+
"mean_output_tokens": null,
|
| 2584 |
+
"mean_supervised_assistant_characters": 834.3636363636364,
|
| 2585 |
+
"output_token_samples": 0,
|
| 2586 |
+
"structurally_eligible": 0,
|
| 2587 |
+
"success": 7,
|
| 2588 |
+
"success_rate_known_outcomes": 0.6363636363636364,
|
| 2589 |
+
"training_eligible": 0,
|
| 2590 |
+
"trajectories": 11,
|
| 2591 |
+
"unique_tasks": 11,
|
| 2592 |
+
"unknown": 0
|
| 2593 |
+
}
|
| 2594 |
+
},
|
| 2595 |
+
"by_task_family": {
|
| 2596 |
+
"software-engineering": {
|
| 2597 |
+
"cost_samples": 0,
|
| 2598 |
+
"duration_samples": 0,
|
| 2599 |
+
"failure": 47840,
|
| 2600 |
+
"input_token_samples": 0,
|
| 2601 |
+
"mean_actions": 47.65324335586629,
|
| 2602 |
+
"mean_cost_source_native_units": null,
|
| 2603 |
+
"mean_duration_seconds": null,
|
| 2604 |
+
"mean_events": 100.24062413186407,
|
| 2605 |
+
"mean_input_tokens": null,
|
| 2606 |
+
"mean_output_tokens": null,
|
| 2607 |
+
"mean_supervised_assistant_characters": 8926.038452634502,
|
| 2608 |
+
"output_token_samples": 0,
|
| 2609 |
+
"structurally_eligible": 86361,
|
| 2610 |
+
"success": 38552,
|
| 2611 |
+
"success_rate_known_outcomes": 0.4462450226872859,
|
| 2612 |
+
"training_eligible": 81719,
|
| 2613 |
+
"trajectories": 86392,
|
| 2614 |
+
"unique_tasks": 47170,
|
| 2615 |
+
"unknown": 0
|
| 2616 |
+
},
|
| 2617 |
+
"terminal-agent": {
|
| 2618 |
+
"cost_samples": 34263,
|
| 2619 |
+
"duration_samples": 34263,
|
| 2620 |
+
"failure": 23212,
|
| 2621 |
+
"input_token_samples": 30615,
|
| 2622 |
+
"mean_actions": 29.13353482114885,
|
| 2623 |
+
"mean_cost_source_native_units": 61.26514903187753,
|
| 2624 |
+
"mean_duration_seconds": 685.1652316865642,
|
| 2625 |
+
"mean_events": 55.965243351823524,
|
| 2626 |
+
"mean_input_tokens": 925188.0988731014,
|
| 2627 |
+
"mean_output_tokens": 18614.85735750449,
|
| 2628 |
+
"mean_supervised_assistant_characters": 13059.311335532499,
|
| 2629 |
+
"output_token_samples": 30615,
|
| 2630 |
+
"structurally_eligible": 43229,
|
| 2631 |
+
"success": 11051,
|
| 2632 |
+
"success_rate_known_outcomes": 0.32253451244783005,
|
| 2633 |
+
"training_eligible": 19143,
|
| 2634 |
+
"trajectories": 54263,
|
| 2635 |
+
"unique_tasks": 17876,
|
| 2636 |
+
"unknown": 20000
|
| 2637 |
+
},
|
| 2638 |
+
"tool-agent-user-simulation": {
|
| 2639 |
+
"cost_samples": 0,
|
| 2640 |
+
"duration_samples": 0,
|
| 2641 |
+
"failure": 5065,
|
| 2642 |
+
"input_token_samples": 0,
|
| 2643 |
+
"mean_actions": 17.724473067915692,
|
| 2644 |
+
"mean_cost_source_native_units": null,
|
| 2645 |
+
"mean_duration_seconds": null,
|
| 2646 |
+
"mean_events": 45.071662763466044,
|
| 2647 |
+
"mean_input_tokens": null,
|
| 2648 |
+
"mean_output_tokens": null,
|
| 2649 |
+
"mean_supervised_assistant_characters": 5714.587704918033,
|
| 2650 |
+
"output_token_samples": 0,
|
| 2651 |
+
"structurally_eligible": 8337,
|
| 2652 |
+
"success": 3475,
|
| 2653 |
+
"success_rate_known_outcomes": 0.4069086651053864,
|
| 2654 |
+
"training_eligible": 8337,
|
| 2655 |
+
"trajectories": 8540,
|
| 2656 |
+
"unique_tasks": 278,
|
| 2657 |
+
"unknown": 0
|
| 2658 |
+
},
|
| 2659 |
+
"tool-use": {
|
| 2660 |
+
"cost_samples": 0,
|
| 2661 |
+
"duration_samples": 0,
|
| 2662 |
+
"failure": 4,
|
| 2663 |
+
"input_token_samples": 0,
|
| 2664 |
+
"mean_actions": 3.8181818181818183,
|
| 2665 |
+
"mean_cost_source_native_units": null,
|
| 2666 |
+
"mean_duration_seconds": null,
|
| 2667 |
+
"mean_events": 9.090909090909092,
|
| 2668 |
+
"mean_input_tokens": null,
|
| 2669 |
+
"mean_output_tokens": null,
|
| 2670 |
+
"mean_supervised_assistant_characters": 834.3636363636364,
|
| 2671 |
+
"output_token_samples": 0,
|
| 2672 |
+
"structurally_eligible": 0,
|
| 2673 |
+
"success": 7,
|
| 2674 |
+
"success_rate_known_outcomes": 0.6363636363636364,
|
| 2675 |
+
"training_eligible": 0,
|
| 2676 |
+
"trajectories": 11,
|
| 2677 |
+
"unique_tasks": 11,
|
| 2678 |
+
"unknown": 0
|
| 2679 |
+
},
|
| 2680 |
+
"web-navigation": {
|
| 2681 |
+
"cost_samples": 0,
|
| 2682 |
+
"duration_samples": 0,
|
| 2683 |
+
"failure": 0,
|
| 2684 |
+
"input_token_samples": 0,
|
| 2685 |
+
"mean_actions": 7.705649157581764,
|
| 2686 |
+
"mean_cost_source_native_units": null,
|
| 2687 |
+
"mean_duration_seconds": null,
|
| 2688 |
+
"mean_events": 16.411298315163528,
|
| 2689 |
+
"mean_input_tokens": null,
|
| 2690 |
+
"mean_output_tokens": null,
|
| 2691 |
+
"mean_supervised_assistant_characters": 256.4281466798811,
|
| 2692 |
+
"output_token_samples": 0,
|
| 2693 |
+
"structurally_eligible": 1009,
|
| 2694 |
+
"success": 0,
|
| 2695 |
+
"success_rate_known_outcomes": null,
|
| 2696 |
+
"training_eligible": 1009,
|
| 2697 |
+
"trajectories": 1009,
|
| 2698 |
+
"unique_tasks": 1009,
|
| 2699 |
+
"unknown": 1009
|
| 2700 |
+
},
|
| 2701 |
+
"web-research": {
|
| 2702 |
+
"cost_samples": 0,
|
| 2703 |
+
"duration_samples": 5063,
|
| 2704 |
+
"failure": 3852,
|
| 2705 |
+
"input_token_samples": 0,
|
| 2706 |
+
"mean_actions": 27.13786292711831,
|
| 2707 |
+
"mean_cost_source_native_units": null,
|
| 2708 |
+
"mean_duration_seconds": 108.07624142800492,
|
| 2709 |
+
"mean_events": 54.28303377444203,
|
| 2710 |
+
"mean_input_tokens": null,
|
| 2711 |
+
"mean_output_tokens": null,
|
| 2712 |
+
"mean_supervised_assistant_characters": 15385.260912502468,
|
| 2713 |
+
"output_token_samples": 0,
|
| 2714 |
+
"structurally_eligible": 5063,
|
| 2715 |
+
"success": 1211,
|
| 2716 |
+
"success_rate_known_outcomes": 0.23918625320955955,
|
| 2717 |
+
"training_eligible": 0,
|
| 2718 |
+
"trajectories": 5063,
|
| 2719 |
+
"unique_tasks": 1266,
|
| 2720 |
+
"unknown": 0
|
| 2721 |
+
}
|
| 2722 |
+
},
|
| 2723 |
+
"compatible_task_environments": 67660,
|
| 2724 |
+
"evaluated_trajectories": 155278,
|
| 2725 |
+
"evaluation_scope": "all imported canonical trajectories, including every alternative attempt",
|
| 2726 |
+
"identical_duplicate_records": 1208,
|
| 2727 |
+
"independently_reproduced": 0,
|
| 2728 |
+
"label_policy": {
|
| 2729 |
+
"cost_units_are_source_native_and_not_globally_comparable": true,
|
| 2730 |
+
"historical_outcomes_are_target_model_scores": false,
|
| 2731 |
+
"independent_reproduction_requires_explicit_evidence": true,
|
| 2732 |
+
"unknown_is_failure": false
|
| 2733 |
+
},
|
| 2734 |
+
"outcomes": {
|
| 2735 |
+
"failure": 79973,
|
| 2736 |
+
"known_denominator": 134269,
|
| 2737 |
+
"success": 54296,
|
| 2738 |
+
"success_rate_all_trajectories": 0.34966962480196806,
|
| 2739 |
+
"success_rate_known_outcomes": 0.40438224757762403,
|
| 2740 |
+
"unknown": 21009
|
| 2741 |
+
},
|
| 2742 |
+
"quality_tiers": {
|
| 2743 |
+
"3": 51802,
|
| 2744 |
+
"4": 19143,
|
| 2745 |
+
"5": 84333
|
| 2746 |
+
},
|
| 2747 |
+
"repeated_attempts": {
|
| 2748 |
+
"attempt_count_distribution": {
|
| 2749 |
+
"1": 40296,
|
| 2750 |
+
"10": 135,
|
| 2751 |
+
"11": 58,
|
| 2752 |
+
"117": 1,
|
| 2753 |
+
"12": 112,
|
| 2754 |
+
"13": 20,
|
| 2755 |
+
"139": 1,
|
| 2756 |
+
"14": 31,
|
| 2757 |
+
"15": 6,
|
| 2758 |
+
"16": 11,
|
| 2759 |
+
"17": 2,
|
| 2760 |
+
"18": 5,
|
| 2761 |
+
"2": 22077,
|
| 2762 |
+
"20": 6,
|
| 2763 |
+
"21": 3,
|
| 2764 |
+
"22": 2,
|
| 2765 |
+
"23": 1,
|
| 2766 |
+
"27": 2,
|
| 2767 |
+
"28": 1,
|
| 2768 |
+
"29": 2,
|
| 2769 |
+
"3": 825,
|
| 2770 |
+
"30": 278,
|
| 2771 |
+
"31": 1,
|
| 2772 |
+
"321": 1,
|
| 2773 |
+
"345": 2,
|
| 2774 |
+
"346": 1,
|
| 2775 |
+
"358": 2,
|
| 2776 |
+
"361": 1,
|
| 2777 |
+
"362": 2,
|
| 2778 |
+
"366": 2,
|
| 2779 |
+
"369": 1,
|
| 2780 |
+
"371": 1,
|
| 2781 |
+
"372": 1,
|
| 2782 |
+
"373": 2,
|
| 2783 |
+
"375": 2,
|
| 2784 |
+
"377": 4,
|
| 2785 |
+
"378": 1,
|
| 2786 |
+
"379": 2,
|
| 2787 |
+
"380": 1,
|
| 2788 |
+
"383": 2,
|
| 2789 |
+
"384": 1,
|
| 2790 |
+
"385": 1,
|
| 2791 |
+
"386": 4,
|
| 2792 |
+
"387": 8,
|
| 2793 |
+
"388": 6,
|
| 2794 |
+
"389": 2,
|
| 2795 |
+
"39": 1,
|
| 2796 |
+
"390": 1,
|
| 2797 |
+
"391": 5,
|
| 2798 |
+
"392": 8,
|
| 2799 |
+
"393": 4,
|
| 2800 |
+
"394": 5,
|
| 2801 |
+
"395": 5,
|
| 2802 |
+
"396": 2,
|
| 2803 |
+
"397": 1,
|
| 2804 |
+
"398": 3,
|
| 2805 |
+
"399": 1,
|
| 2806 |
+
"4": 1867,
|
| 2807 |
+
"401": 1,
|
| 2808 |
+
"403": 1,
|
| 2809 |
+
"404": 1,
|
| 2810 |
+
"405": 1,
|
| 2811 |
+
"412": 1,
|
| 2812 |
+
"46": 1,
|
| 2813 |
+
"5": 593,
|
| 2814 |
+
"6": 271,
|
| 2815 |
+
"67": 1,
|
| 2816 |
+
"7": 161,
|
| 2817 |
+
"8": 217,
|
| 2818 |
+
"9": 582
|
| 2819 |
+
},
|
| 2820 |
+
"empirical_success_at_k": [
|
| 2821 |
+
{
|
| 2822 |
+
"eligible_tasks": 48864,
|
| 2823 |
+
"empirical_success_at_k": 0.4352079240340537,
|
| 2824 |
+
"k": 1,
|
| 2825 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2826 |
+
"solved_tasks": 21266
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"eligible_tasks": 26897,
|
| 2830 |
+
"empirical_success_at_k": 0.5153362828568242,
|
| 2831 |
+
"k": 2,
|
| 2832 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2833 |
+
"solved_tasks": 13861
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"eligible_tasks": 5057,
|
| 2837 |
+
"empirical_success_at_k": 0.4601542416452442,
|
| 2838 |
+
"k": 3,
|
| 2839 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2840 |
+
"solved_tasks": 2327
|
| 2841 |
+
},
|
| 2842 |
+
{
|
| 2843 |
+
"eligible_tasks": 4313,
|
| 2844 |
+
"empirical_success_at_k": 0.4817992116856017,
|
| 2845 |
+
"k": 4,
|
| 2846 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2847 |
+
"solved_tasks": 2078
|
| 2848 |
+
},
|
| 2849 |
+
{
|
| 2850 |
+
"eligible_tasks": 2487,
|
| 2851 |
+
"empirical_success_at_k": 0.5158825894652191,
|
| 2852 |
+
"k": 5,
|
| 2853 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2854 |
+
"solved_tasks": 1283
|
| 2855 |
+
},
|
| 2856 |
+
{
|
| 2857 |
+
"eligible_tasks": 1919,
|
| 2858 |
+
"empirical_success_at_k": 0.6242834809796769,
|
| 2859 |
+
"k": 6,
|
| 2860 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2861 |
+
"solved_tasks": 1198
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"eligible_tasks": 1663,
|
| 2865 |
+
"empirical_success_at_k": 0.6915213469633193,
|
| 2866 |
+
"k": 7,
|
| 2867 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2868 |
+
"solved_tasks": 1150
|
| 2869 |
+
},
|
| 2870 |
+
{
|
| 2871 |
+
"eligible_tasks": 1511,
|
| 2872 |
+
"empirical_success_at_k": 0.7346128391793514,
|
| 2873 |
+
"k": 8,
|
| 2874 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2875 |
+
"solved_tasks": 1110
|
| 2876 |
+
},
|
| 2877 |
+
{
|
| 2878 |
+
"eligible_tasks": 1303,
|
| 2879 |
+
"empirical_success_at_k": 0.7705295471987721,
|
| 2880 |
+
"k": 9,
|
| 2881 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2882 |
+
"solved_tasks": 1004
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"eligible_tasks": 724,
|
| 2886 |
+
"empirical_success_at_k": 0.649171270718232,
|
| 2887 |
+
"k": 10,
|
| 2888 |
+
"ordering": "source record locator then trajectory ID; empirical, not estimated pass@k",
|
| 2889 |
+
"solved_tasks": 470
|
| 2890 |
+
}
|
| 2891 |
+
],
|
| 2892 |
+
"fraction_solved_at_least_once": 0.47888015717092336,
|
| 2893 |
+
"tasks_solved_at_least_once": 23400,
|
| 2894 |
+
"tasks_with_known_outcomes": 48864,
|
| 2895 |
+
"tasks_with_multiple_trajectories": 27364
|
| 2896 |
+
},
|
| 2897 |
+
"schema_version": 1,
|
| 2898 |
+
"structurally_eligible": 143999,
|
| 2899 |
+
"training_eligible": 110208,
|
| 2900 |
+
"unique_tasks": 67610
|
| 2901 |
+
}
|
examples.html
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b543762180caad13dcc65fced9398d7afbe791dee86c0d27d9772d0a6b7fda19
|
| 3 |
+
size 14119755
|
examples.selection.json
ADDED
|
@@ -0,0 +1,142 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"comparison": {
|
| 3 |
+
"attempts": 412,
|
| 4 |
+
"environment_key": "Terminal-Bench@2.0",
|
| 5 |
+
"task_key": "terminalbench:count-dataset-tokens"
|
| 6 |
+
},
|
| 7 |
+
"records": [
|
| 8 |
+
{
|
| 9 |
+
"reason": "browsecomp_rollouts/success/tier-3",
|
| 10 |
+
"trajectory_id": "trajectory:f6afa2d095e1a4ebbe1de80b"
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"reason": "browsecomp_rollouts/failure/tier-5",
|
| 14 |
+
"trajectory_id": "trajectory:aad1409d32b2ad5bfa3717fb"
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"reason": "mind2web/unknown/tier-3",
|
| 18 |
+
"trajectory_id": "trajectory:37ff002c486f5bbc671c7f05"
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"reason": "swebench_multilingual_trajectories/success/tier-3",
|
| 22 |
+
"trajectory_id": "trajectory:ceb8c403bec585635db5c885"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"reason": "swebench_multilingual_trajectories/failure/tier-5",
|
| 26 |
+
"trajectory_id": "trajectory:230dd0fadb9aa207896b2745"
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"reason": "swegym_sampled/failure/tier-5",
|
| 30 |
+
"trajectory_id": "trajectory:17b01c2476ae7525876e3ca3"
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"reason": "nvidia_open_swe_traces/success/tier-3",
|
| 34 |
+
"trajectory_id": "trajectory:dc2a7493e851fb4205dabf60"
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"reason": "swebench_verified_trajectories/failure/tier-5",
|
| 38 |
+
"trajectory_id": "trajectory:1939c1be6fbe482a9fd74b4b"
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"reason": "swegym_sampled/success/tier-3",
|
| 42 |
+
"trajectory_id": "trajectory:465fdac1a1b3856392aa77c7"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"reason": "nvidia_open_swe_traces/failure/tier-5",
|
| 46 |
+
"trajectory_id": "trajectory:803ce21a16aed837c0db037b"
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"reason": "swegym_openhands/success/tier-3",
|
| 50 |
+
"trajectory_id": "trajectory:679220a4ab9e2685d09d8b8f"
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"reason": "swebench_verified_trajectories/success/tier-5",
|
| 54 |
+
"trajectory_id": "trajectory:7cd76b774e3d1f674cd36ca9"
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"reason": "swebench_verified_trajectories/success/tier-3",
|
| 58 |
+
"trajectory_id": "trajectory:c39124375e6ebd36c9f65512"
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"reason": "swesmith_trajectories/failure/tier-5",
|
| 62 |
+
"trajectory_id": "trajectory:b506991f4199d7819fd6fe87"
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"reason": "swesmith_trajectories/success/tier-3",
|
| 66 |
+
"trajectory_id": "trajectory:08cddd9336ac3a944e153a35"
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"reason": "agentsuite_tau2/success/tier-3",
|
| 70 |
+
"trajectory_id": "trajectory:0cbb52ef305b68ca6a0e28b8"
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"reason": "agentsuite_tau2/success/tier-5",
|
| 74 |
+
"trajectory_id": "trajectory:8f3680cf4f445e8511d91a17"
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"reason": "agentsuite_tau2/failure/tier-5",
|
| 78 |
+
"trajectory_id": "trajectory:ab0d84b3dc737fca1610ac4d"
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"reason": "official_tau3/failure/tier-5",
|
| 82 |
+
"trajectory_id": "trajectory:a2dc8e895192d065455cf275"
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"reason": "official_tau3/success/tier-5",
|
| 86 |
+
"trajectory_id": "trajectory:7c7d376b6e7e99e06a878784"
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"reason": "terminaltraj/unknown/tier-4",
|
| 90 |
+
"trajectory_id": "trajectory:049aa1b8a9751320c05978f9"
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"reason": "terminaltraj/unknown/tier-5",
|
| 94 |
+
"trajectory_id": "trajectory:29a619875a0a331fc3fe3515"
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"reason": "terminalbench_trajectories/success/tier-5",
|
| 98 |
+
"trajectory_id": "trajectory:42e80eb22663e48fd9b2b497"
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"reason": "terminalbench_trajectories/failure/tier-5",
|
| 102 |
+
"trajectory_id": "trajectory:afd3c68daf9d09b9cd03d3fd"
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"reason": "repeated-attempt-comparison",
|
| 106 |
+
"trajectory_id": "trajectory:01bf48ea0f6c3beb1162bf22"
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"reason": "repeated-attempt-comparison",
|
| 110 |
+
"trajectory_id": "trajectory:0698279b8ea54e0e439e4e74"
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"reason": "repeated-attempt-comparison",
|
| 114 |
+
"trajectory_id": "trajectory:0ea8a532668fdf63eadc25c7"
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"reason": "repeated-attempt-comparison",
|
| 118 |
+
"trajectory_id": "trajectory:8969e9a90720972236678351"
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"reason": "repeated-attempt-comparison",
|
| 122 |
+
"trajectory_id": "trajectory:d0ce4ac1d315c8c143c8a915"
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"reason": "repeated-attempt-comparison",
|
| 126 |
+
"trajectory_id": "trajectory:e4618556d3f972e3974f7b5a"
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"reason": "terminalbench_trajectories/success/tier-3",
|
| 130 |
+
"trajectory_id": "trajectory:abce4a0d6188f4510f4d0bc0"
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"reason": "toolbench/success/tier-5",
|
| 134 |
+
"trajectory_id": "trajectory:5653af79ecc03fead3f540ff"
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"reason": "toolbench/failure/tier-5",
|
| 138 |
+
"trajectory_id": "trajectory:e3784d724ae3e7dd55befd52"
|
| 139 |
+
}
|
| 140 |
+
],
|
| 141 |
+
"selection_policy": "lexicographically first trajectory per source/outcome/tier plus the largest non-BFCL repeated-attempt group"
|
| 142 |
+
}
|
examples/nemotron_sft_demo.json
ADDED
|
@@ -0,0 +1,191 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"examples": [
|
| 3 |
+
{
|
| 4 |
+
"assistant_loss_tokens": 696,
|
| 5 |
+
"assistant_only_mask": true,
|
| 6 |
+
"ignored_context_tokens": 2478,
|
| 7 |
+
"input_ids_preview": [
|
| 8 |
+
10,
|
| 9 |
+
25708,
|
| 10 |
+
1010,
|
| 11 |
+
1035,
|
| 12 |
+
7466,
|
| 13 |
+
2839,
|
| 14 |
+
38114,
|
| 15 |
+
28093,
|
| 16 |
+
1595,
|
| 17 |
+
13404,
|
| 18 |
+
5010,
|
| 19 |
+
28036,
|
| 20 |
+
2794,
|
| 21 |
+
1595,
|
| 22 |
+
11,
|
| 23 |
+
1010
|
| 24 |
+
],
|
| 25 |
+
"labels_preview": [
|
| 26 |
+
-100,
|
| 27 |
+
-100,
|
| 28 |
+
-100,
|
| 29 |
+
-100,
|
| 30 |
+
-100,
|
| 31 |
+
-100,
|
| 32 |
+
-100,
|
| 33 |
+
-100,
|
| 34 |
+
-100,
|
| 35 |
+
-100,
|
| 36 |
+
-100,
|
| 37 |
+
-100,
|
| 38 |
+
-100,
|
| 39 |
+
-100,
|
| 40 |
+
-100,
|
| 41 |
+
-100
|
| 42 |
+
],
|
| 43 |
+
"sample_id": "full:736e63752d3cff92f0d883d1",
|
| 44 |
+
"token_length": 3174,
|
| 45 |
+
"view": "full_trajectory"
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"assistant_loss_tokens": 60,
|
| 49 |
+
"assistant_only_mask": true,
|
| 50 |
+
"causal_context_events": 5,
|
| 51 |
+
"causal_context_tokens": 1817,
|
| 52 |
+
"ignored_context_tokens": 1817,
|
| 53 |
+
"input_ids_preview": [
|
| 54 |
+
10,
|
| 55 |
+
25708,
|
| 56 |
+
1010,
|
| 57 |
+
1035,
|
| 58 |
+
7466,
|
| 59 |
+
2839,
|
| 60 |
+
38114,
|
| 61 |
+
28093,
|
| 62 |
+
1561,
|
| 63 |
+
1885,
|
| 64 |
+
5165,
|
| 65 |
+
1561,
|
| 66 |
+
15,
|
| 67 |
+
1010,
|
| 68 |
+
11,
|
| 69 |
+
1010
|
| 70 |
+
],
|
| 71 |
+
"labels_preview": [
|
| 72 |
+
-100,
|
| 73 |
+
-100,
|
| 74 |
+
-100,
|
| 75 |
+
-100,
|
| 76 |
+
-100,
|
| 77 |
+
-100,
|
| 78 |
+
-100,
|
| 79 |
+
-100,
|
| 80 |
+
1561,
|
| 81 |
+
1885,
|
| 82 |
+
5165,
|
| 83 |
+
1561,
|
| 84 |
+
15,
|
| 85 |
+
1010,
|
| 86 |
+
11,
|
| 87 |
+
1010
|
| 88 |
+
],
|
| 89 |
+
"prior_assistant_tokens_masked_out": 80,
|
| 90 |
+
"sample_id": "turn:61eec58539653d88c1971b68",
|
| 91 |
+
"target_event_id": "event:f8253b7938b844a48ae3ab91",
|
| 92 |
+
"target_event_index": 5,
|
| 93 |
+
"token_length": 1877,
|
| 94 |
+
"view": "assistant_turn"
|
| 95 |
+
}
|
| 96 |
+
],
|
| 97 |
+
"model_ready_batch": {
|
| 98 |
+
"all_padding_labels_ignored": true,
|
| 99 |
+
"fields": [
|
| 100 |
+
"input_ids",
|
| 101 |
+
"attention_mask",
|
| 102 |
+
"labels"
|
| 103 |
+
],
|
| 104 |
+
"framework_neutral_python_batch": true,
|
| 105 |
+
"label_ignore_index": -100,
|
| 106 |
+
"non_padding_tokens": [
|
| 107 |
+
3174,
|
| 108 |
+
1877
|
| 109 |
+
],
|
| 110 |
+
"right_padded": true,
|
| 111 |
+
"sha256": "d680ae7e4ea605b3db96370de6760d86ed227411dac901d74a511f72c3346ae8",
|
| 112 |
+
"shape": [
|
| 113 |
+
2,
|
| 114 |
+
3174
|
| 115 |
+
],
|
| 116 |
+
"supervised_label_tokens": [
|
| 117 |
+
696,
|
| 118 |
+
60
|
| 119 |
+
]
|
| 120 |
+
},
|
| 121 |
+
"real_trajectory": {
|
| 122 |
+
"assistant_target_count": 6,
|
| 123 |
+
"event_count": 13,
|
| 124 |
+
"outcome_success": true,
|
| 125 |
+
"preview": [
|
| 126 |
+
{
|
| 127 |
+
"content": "# Airline Agent Policy\n\nThe current time is 2024-05-15 15:00:00 EST.\n\nAs an airline agent, you can help users **book**, **modify**, or **cancel** flight reservations. You also handle **refunds and compensation**.\n\nBefore taking any actions ",
|
| 128 |
+
"role": "system",
|
| 129 |
+
"supervised": false,
|
| 130 |
+
"tool_calls": []
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"content": "Hi! How can I help you today?",
|
| 134 |
+
"role": "assistant",
|
| 135 |
+
"supervised": true,
|
| 136 |
+
"tool_calls": []
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"content": "Hi, I’d like to cancel my reservation with the code EHGLP3. Can you help me with that?",
|
| 140 |
+
"role": "user",
|
| 141 |
+
"supervised": false,
|
| 142 |
+
"tool_calls": []
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"content": "I'd be happy to help you cancel your reservation. Before proceeding, I'll need a few pieces of information.\n\nCould you please provide your user ID? Also, may I ask what is the reason for cancellation (change of plan, airline cancelled fligh",
|
| 146 |
+
"role": "assistant",
|
| 147 |
+
"supervised": true,
|
| 148 |
+
"tool_calls": []
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"content": "Sure, my user ID is emma_kim_9957. The reason is that my plans changed unexpectedly.",
|
| 152 |
+
"role": "user",
|
| 153 |
+
"supervised": false,
|
| 154 |
+
"tool_calls": []
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"content": "Thank you for providing your user ID. Let me check the details of your reservation before proceeding with the cancellation.",
|
| 158 |
+
"role": "assistant",
|
| 159 |
+
"supervised": true,
|
| 160 |
+
"tool_calls": [
|
| 161 |
+
"get_reservation_details"
|
| 162 |
+
]
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"content": "{\"reservation_id\": \"EHGLP3\", \"user_id\": \"emma_kim_9957\", \"origin\": \"PHX\", \"destination\": \"JFK\", \"flight_type\": \"one_way\", \"cabin\": \"basic_economy\", \"flights\": [{\"flight_number\": \"HAT156\", \"origin\": \"PHX\", \"destination\": \"SEA\", \"date\": \"2024",
|
| 166 |
+
"role": "tool",
|
| 167 |
+
"supervised": false,
|
| 168 |
+
"tool_calls": []
|
| 169 |
+
}
|
| 170 |
+
],
|
| 171 |
+
"record_locator": "data/cache/official_tau3/source.json#simulations[1]",
|
| 172 |
+
"source_id": "official_tau3",
|
| 173 |
+
"source_revision": "dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd",
|
| 174 |
+
"task_id": "0",
|
| 175 |
+
"task_key": "tau:airline:0",
|
| 176 |
+
"trajectory_id": "trajectory:f9cf8588a314fa9804d57402",
|
| 177 |
+
"verification_method": "tau environment reward"
|
| 178 |
+
},
|
| 179 |
+
"schema_version": 1,
|
| 180 |
+
"tokenizer": {
|
| 181 |
+
"assistant_mask_method": "pinned native chat template with Transformers generation spans around native assistant rendering",
|
| 182 |
+
"class": "TokenizersBackend",
|
| 183 |
+
"eos_token_id": 11,
|
| 184 |
+
"model": "nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16",
|
| 185 |
+
"model_max_length": 262144,
|
| 186 |
+
"pad_token_id": 11,
|
| 187 |
+
"revision": "a9904d24bcc1d289a1950fa9d2b978c47cf903b9",
|
| 188 |
+
"truncate_history_thinking": false,
|
| 189 |
+
"vocabulary_size": 131072
|
| 190 |
+
}
|
| 191 |
+
}
|
index.sqlite
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:25b97793c7b002995cdbea6daddb4e0068b81b40aaeb11179a2915f9810857fd
|
| 3 |
+
size 178581504
|
manifest.json
ADDED
|
@@ -0,0 +1,1277 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"content_complete": false,
|
| 3 |
+
"duplicate_provenance_records": 1208,
|
| 4 |
+
"duplicate_trajectory_ids_skipped": 0,
|
| 5 |
+
"evaluated": 155278,
|
| 6 |
+
"files": {
|
| 7 |
+
"all": null,
|
| 8 |
+
"duplicate_provenance": "duplicate_provenance.jsonl.gz",
|
| 9 |
+
"partition_indices": {
|
| 10 |
+
"structurally_eligible": "structurally_eligible.index.jsonl.gz",
|
| 11 |
+
"successful": "successful.index.jsonl.gz",
|
| 12 |
+
"training_eligible": "training_eligible.index.jsonl.gz",
|
| 13 |
+
"unknown": "unknown.index.jsonl.gz",
|
| 14 |
+
"unsuccessful": "unsuccessful.index.jsonl.gz"
|
| 15 |
+
}
|
| 16 |
+
},
|
| 17 |
+
"hosted_source_counts": {
|
| 18 |
+
"mind2web": 1009,
|
| 19 |
+
"nvidia_open_swe_traces": 51626,
|
| 20 |
+
"official_tau3": 200,
|
| 21 |
+
"swebench_multilingual_trajectories": 129,
|
| 22 |
+
"swebench_verified_trajectories": 4544,
|
| 23 |
+
"swegym_openhands": 491,
|
| 24 |
+
"swesmith_trajectories": 24100,
|
| 25 |
+
"terminalbench_trajectories": 34263,
|
| 26 |
+
"terminaltraj": 20000,
|
| 27 |
+
"toolbench": 11
|
| 28 |
+
},
|
| 29 |
+
"hosted_trajectories": 136373,
|
| 30 |
+
"identical_records_collapsed": 1208,
|
| 31 |
+
"inputs": [
|
| 32 |
+
{
|
| 33 |
+
"path": "canonical/mind2web/default--train--0000.jsonl.gz",
|
| 34 |
+
"records": 100,
|
| 35 |
+
"sha256": "654357077e6249cd030e516a74bc3ac7551e26553809581bbebb56a4f87e4c37",
|
| 36 |
+
"size_bytes": 60960835,
|
| 37 |
+
"source_counts": {
|
| 38 |
+
"mind2web": 100
|
| 39 |
+
}
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"path": "canonical/mind2web/default--train--0001.jsonl.gz",
|
| 43 |
+
"records": 100,
|
| 44 |
+
"sha256": "ea62d59ed99fa6632c59bb62cf339054fe72e717c38d36109526414bb9ce2f60",
|
| 45 |
+
"size_bytes": 61759916,
|
| 46 |
+
"source_counts": {
|
| 47 |
+
"mind2web": 100
|
| 48 |
+
}
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"path": "canonical/mind2web/default--train--0002.jsonl.gz",
|
| 52 |
+
"records": 109,
|
| 53 |
+
"sha256": "77d3d9b0ef99cf5ff8300c7e0681847780752ce60e2c2e0887e247b491b48e22",
|
| 54 |
+
"size_bytes": 61958505,
|
| 55 |
+
"source_counts": {
|
| 56 |
+
"mind2web": 109
|
| 57 |
+
}
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"path": "canonical/mind2web/default--train--0003.jsonl.gz",
|
| 61 |
+
"records": 100,
|
| 62 |
+
"sha256": "f3807d46087383372f43ca0e3ed95d1b14e9ceca917940c652d136eca5157d7c",
|
| 63 |
+
"size_bytes": 64515683,
|
| 64 |
+
"source_counts": {
|
| 65 |
+
"mind2web": 100
|
| 66 |
+
}
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"path": "canonical/mind2web/default--train--0004.jsonl.gz",
|
| 70 |
+
"records": 100,
|
| 71 |
+
"sha256": "f4e03e9b52f92aecac56c551d2a83e1af8fe26933e8366a26cfb90249c9b1f5f",
|
| 72 |
+
"size_bytes": 59759759,
|
| 73 |
+
"source_counts": {
|
| 74 |
+
"mind2web": 100
|
| 75 |
+
}
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"path": "canonical/mind2web/default--train--0005.jsonl.gz",
|
| 79 |
+
"records": 100,
|
| 80 |
+
"sha256": "cd075e514c9006f2781cec84a0bbb816e1bc6ccdf22217885925e2ab5d02df8b",
|
| 81 |
+
"size_bytes": 59608293,
|
| 82 |
+
"source_counts": {
|
| 83 |
+
"mind2web": 100
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"path": "canonical/mind2web/default--train--0006.jsonl.gz",
|
| 88 |
+
"records": 200,
|
| 89 |
+
"sha256": "fa7ab83febb58ab7a51ecdee07a03369ccc55b8b8ae811dd5679693f98a70c52",
|
| 90 |
+
"size_bytes": 115862382,
|
| 91 |
+
"source_counts": {
|
| 92 |
+
"mind2web": 200
|
| 93 |
+
}
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"path": "canonical/mind2web/default--train--0007.jsonl.gz",
|
| 97 |
+
"records": 200,
|
| 98 |
+
"sha256": "aa20a9fa7022a87f523148839fbbccae8576eecd011eb8a8b7a12be2472c8e5e",
|
| 99 |
+
"size_bytes": 116176682,
|
| 100 |
+
"source_counts": {
|
| 101 |
+
"mind2web": 200
|
| 102 |
+
}
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0000.jsonl.gz",
|
| 106 |
+
"records": 2500,
|
| 107 |
+
"sha256": "22d874f857dd91bef6f539e5c5809b5f654bb72c310caad8d42cd1d552189c45",
|
| 108 |
+
"size_bytes": 303785042,
|
| 109 |
+
"source_counts": {
|
| 110 |
+
"nvidia_open_swe_traces": 2500
|
| 111 |
+
}
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0001.jsonl.gz",
|
| 115 |
+
"records": 2500,
|
| 116 |
+
"sha256": "be0c9efbdfb9947734634f75102ad4e2be154b4ab63770a748ebcfb0a5677927",
|
| 117 |
+
"size_bytes": 295831492,
|
| 118 |
+
"source_counts": {
|
| 119 |
+
"nvidia_open_swe_traces": 2500
|
| 120 |
+
}
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0002.jsonl.gz",
|
| 124 |
+
"records": 2500,
|
| 125 |
+
"sha256": "3067df96c93975319e290aa0b5f7b4be0ceecb6275f761cea33db9b6b3c54ffd",
|
| 126 |
+
"size_bytes": 305310880,
|
| 127 |
+
"source_counts": {
|
| 128 |
+
"nvidia_open_swe_traces": 2500
|
| 129 |
+
}
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0003.jsonl.gz",
|
| 133 |
+
"records": 2500,
|
| 134 |
+
"sha256": "2b3324ce204a000df83ad9e207b4318af57ed19d36c11d7cbbae7bf5db90b453",
|
| 135 |
+
"size_bytes": 290056585,
|
| 136 |
+
"source_counts": {
|
| 137 |
+
"nvidia_open_swe_traces": 2500
|
| 138 |
+
}
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0004.jsonl.gz",
|
| 142 |
+
"records": 2500,
|
| 143 |
+
"sha256": "5373e01f7f6527a0611ae933bb73b92d6e2ae312ea81d0d3aef85b1b241e55a7",
|
| 144 |
+
"size_bytes": 287437480,
|
| 145 |
+
"source_counts": {
|
| 146 |
+
"nvidia_open_swe_traces": 2500
|
| 147 |
+
}
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0005.jsonl.gz",
|
| 151 |
+
"records": 41,
|
| 152 |
+
"sha256": "92b92e2aba80e79cbb07e9ba071db946fe4aa885deef4aefa9d934ae3bff193e",
|
| 153 |
+
"size_bytes": 5161644,
|
| 154 |
+
"source_counts": {
|
| 155 |
+
"nvidia_open_swe_traces": 41
|
| 156 |
+
}
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0006.jsonl.gz",
|
| 160 |
+
"records": 2500,
|
| 161 |
+
"sha256": "065ba7c18fbf854df437e352fe79cf54237f22d9290ef0b55283ff7f5c9e7e77",
|
| 162 |
+
"size_bytes": 368841972,
|
| 163 |
+
"source_counts": {
|
| 164 |
+
"nvidia_open_swe_traces": 2500
|
| 165 |
+
}
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0007.jsonl.gz",
|
| 169 |
+
"records": 2500,
|
| 170 |
+
"sha256": "af1f7ed30b9037df06ebf7e7f0d146913ad813511e065e8f735a520b646b74bf",
|
| 171 |
+
"size_bytes": 371128455,
|
| 172 |
+
"source_counts": {
|
| 173 |
+
"nvidia_open_swe_traces": 2500
|
| 174 |
+
}
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0008.jsonl.gz",
|
| 178 |
+
"records": 2500,
|
| 179 |
+
"sha256": "1b8247219e917730ef629eb0e66f2ec19d91ea404ea546c1ed10961eba7e0d30",
|
| 180 |
+
"size_bytes": 367495181,
|
| 181 |
+
"source_counts": {
|
| 182 |
+
"nvidia_open_swe_traces": 2500
|
| 183 |
+
}
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0009.jsonl.gz",
|
| 187 |
+
"records": 2500,
|
| 188 |
+
"sha256": "0aec09bafc26fcd3254794c23b358cb597e94a492d8e0fd153650e859f5997be",
|
| 189 |
+
"size_bytes": 370761485,
|
| 190 |
+
"source_counts": {
|
| 191 |
+
"nvidia_open_swe_traces": 2500
|
| 192 |
+
}
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0010.jsonl.gz",
|
| 196 |
+
"records": 2500,
|
| 197 |
+
"sha256": "2eff407239d58e2d9d525aa70402dcf01501ef3d3007d6407a69765c9c4936b5",
|
| 198 |
+
"size_bytes": 370224921,
|
| 199 |
+
"source_counts": {
|
| 200 |
+
"nvidia_open_swe_traces": 2500
|
| 201 |
+
}
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0011.jsonl.gz",
|
| 205 |
+
"records": 558,
|
| 206 |
+
"sha256": "2a68e3249212dbf0661db6e7d7a5b9afc6a2a4bf0172d0f32e9ad0cbf272abbc",
|
| 207 |
+
"size_bytes": 83589148,
|
| 208 |
+
"source_counts": {
|
| 209 |
+
"nvidia_open_swe_traces": 558
|
| 210 |
+
}
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0012.jsonl.gz",
|
| 214 |
+
"records": 2500,
|
| 215 |
+
"sha256": "d54f3aebffe6bfdd1889488db25bf851210f000d6f6ca2fa2bebf83155fea40a",
|
| 216 |
+
"size_bytes": 243965767,
|
| 217 |
+
"source_counts": {
|
| 218 |
+
"nvidia_open_swe_traces": 2500
|
| 219 |
+
}
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0013.jsonl.gz",
|
| 223 |
+
"records": 2500,
|
| 224 |
+
"sha256": "454932733cdc2ce5d6570d064a239ebbe497b0b1915c51f1ea1c3837edc5fae3",
|
| 225 |
+
"size_bytes": 241910439,
|
| 226 |
+
"source_counts": {
|
| 227 |
+
"nvidia_open_swe_traces": 2500
|
| 228 |
+
}
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0014.jsonl.gz",
|
| 232 |
+
"records": 2500,
|
| 233 |
+
"sha256": "e964d62766afd638a8526c5e44fca9581e2cea0d3ec0ffd0a14beeefc1e16c6c",
|
| 234 |
+
"size_bytes": 249402952,
|
| 235 |
+
"source_counts": {
|
| 236 |
+
"nvidia_open_swe_traces": 2500
|
| 237 |
+
}
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0015.jsonl.gz",
|
| 241 |
+
"records": 2500,
|
| 242 |
+
"sha256": "97fee226a8dc416dadb5673cace864585c36a933d42238abd640526814a5b2c5",
|
| 243 |
+
"size_bytes": 240747809,
|
| 244 |
+
"source_counts": {
|
| 245 |
+
"nvidia_open_swe_traces": 2500
|
| 246 |
+
}
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0016.jsonl.gz",
|
| 250 |
+
"records": 2500,
|
| 251 |
+
"sha256": "95b32f5071fe69c1949e66623413b9e02970973ea501040336955b56d6de0b16",
|
| 252 |
+
"size_bytes": 242546761,
|
| 253 |
+
"source_counts": {
|
| 254 |
+
"nvidia_open_swe_traces": 2500
|
| 255 |
+
}
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0017.jsonl.gz",
|
| 259 |
+
"records": 771,
|
| 260 |
+
"sha256": "2c17ad85f0ef74b315314fc6929e3efe7875f5d88f319208bc07fbbbc621c296",
|
| 261 |
+
"size_bytes": 74960935,
|
| 262 |
+
"source_counts": {
|
| 263 |
+
"nvidia_open_swe_traces": 771
|
| 264 |
+
}
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0018.jsonl.gz",
|
| 268 |
+
"records": 2500,
|
| 269 |
+
"sha256": "331b6e4d1611cc71c643b7a1af41744218d25aa5d6be33675f3ca2bcc20ab183",
|
| 270 |
+
"size_bytes": 373422381,
|
| 271 |
+
"source_counts": {
|
| 272 |
+
"nvidia_open_swe_traces": 2500
|
| 273 |
+
}
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0019.jsonl.gz",
|
| 277 |
+
"records": 2500,
|
| 278 |
+
"sha256": "a42a718e33c3b46b9bc01867f57b2166a3ef6de0063e41d6dfb419bb55df971f",
|
| 279 |
+
"size_bytes": 371590204,
|
| 280 |
+
"source_counts": {
|
| 281 |
+
"nvidia_open_swe_traces": 2500
|
| 282 |
+
}
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0020.jsonl.gz",
|
| 286 |
+
"records": 2500,
|
| 287 |
+
"sha256": "e412984c9e629a4621083e68f60fa1713ca1678f7eab46c8d0ee25813f967f90",
|
| 288 |
+
"size_bytes": 376298397,
|
| 289 |
+
"source_counts": {
|
| 290 |
+
"nvidia_open_swe_traces": 2500
|
| 291 |
+
}
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0021.jsonl.gz",
|
| 295 |
+
"records": 2500,
|
| 296 |
+
"sha256": "22ae50d9032b0850fcf93080fcc81143d6af8f6ecad74c732c6d0428a6b4000c",
|
| 297 |
+
"size_bytes": 376587752,
|
| 298 |
+
"source_counts": {
|
| 299 |
+
"nvidia_open_swe_traces": 2500
|
| 300 |
+
}
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0022.jsonl.gz",
|
| 304 |
+
"records": 2500,
|
| 305 |
+
"sha256": "379caa74f2bedf81325a98874fc804c40fd8741f7bf6571b66d72a4b905b6911",
|
| 306 |
+
"size_bytes": 380085424,
|
| 307 |
+
"source_counts": {
|
| 308 |
+
"nvidia_open_swe_traces": 2500
|
| 309 |
+
}
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0023.jsonl.gz",
|
| 313 |
+
"records": 256,
|
| 314 |
+
"sha256": "863309aa10f41341676f0fd71e7e2d68db2d50b3d8451b5b8adab83ff44c515c",
|
| 315 |
+
"size_bytes": 38763239,
|
| 316 |
+
"source_counts": {
|
| 317 |
+
"nvidia_open_swe_traces": 256
|
| 318 |
+
}
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"path": "canonical/official_tau3+toolbench/official_tau3_toolbench.jsonl.gz",
|
| 322 |
+
"records": 211,
|
| 323 |
+
"sha256": "0e08775a1a2e47543fb6c3ddd1789660ece1d9fe3bbf6189e457fc06f7bcc594",
|
| 324 |
+
"size_bytes": 2412172,
|
| 325 |
+
"source_counts": {
|
| 326 |
+
"official_tau3": 200,
|
| 327 |
+
"toolbench": 11
|
| 328 |
+
}
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"path": "canonical/swebench_multilingual_trajectories/default--Multi_SWE_agent_32B_Rust--0000.jsonl.gz",
|
| 332 |
+
"records": 43,
|
| 333 |
+
"sha256": "0034093e5c0b717c4957b481d85c9381f9cba025ad16d1b0fcac9152c9698656",
|
| 334 |
+
"size_bytes": 17147283,
|
| 335 |
+
"source_counts": {
|
| 336 |
+
"swebench_multilingual_trajectories": 43
|
| 337 |
+
}
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"path": "canonical/swebench_multilingual_trajectories/default--Qwen2.5_Coder_32B_Instruct--0001.jsonl.gz",
|
| 341 |
+
"records": 43,
|
| 342 |
+
"sha256": "146462000c0168746f0ca6fccec7f0cf859db13353ccd7504d8e264964eb4441",
|
| 343 |
+
"size_bytes": 36236356,
|
| 344 |
+
"source_counts": {
|
| 345 |
+
"swebench_multilingual_trajectories": 43
|
| 346 |
+
}
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"path": "canonical/swebench_multilingual_trajectories/default--SWE_agent_LM_32B--0002.jsonl.gz",
|
| 350 |
+
"records": 43,
|
| 351 |
+
"sha256": "bf4d1265c6f2928fc1b8c188299fc5e46892bc3624ca377249c9855bf3e766fa",
|
| 352 |
+
"size_bytes": 18456716,
|
| 353 |
+
"source_counts": {
|
| 354 |
+
"swebench_multilingual_trajectories": 43
|
| 355 |
+
}
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"path": "canonical/swebench_verified_trajectories/all.jsonl.gz",
|
| 359 |
+
"records": 5000,
|
| 360 |
+
"sha256": "7da41ea495e4f899daacf55989fcd5be6ef103c870f1f22e2aa11c0a14ce8982",
|
| 361 |
+
"size_bytes": 336638978,
|
| 362 |
+
"source_counts": {
|
| 363 |
+
"swebench_verified_trajectories": 5000
|
| 364 |
+
}
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"path": "canonical/swegym_openhands/default--train.success.oss--0000.jsonl.gz",
|
| 368 |
+
"records": 491,
|
| 369 |
+
"sha256": "8d5a0fb7c17af0eefdf7ed0748ded69aef38fab9e9c6d661bf384915847777f9",
|
| 370 |
+
"size_bytes": 13990587,
|
| 371 |
+
"source_counts": {
|
| 372 |
+
"swegym_openhands": 491
|
| 373 |
+
}
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"path": "canonical/swesmith_trajectories/default--tool--0000.jsonl.gz",
|
| 377 |
+
"records": 3013,
|
| 378 |
+
"sha256": "909cd6654ba6b253bf83874168fcfef57eb7ae700d94d39d4e10da15e2d4df22",
|
| 379 |
+
"size_bytes": 208736286,
|
| 380 |
+
"source_counts": {
|
| 381 |
+
"swesmith_trajectories": 3013
|
| 382 |
+
}
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"path": "canonical/swesmith_trajectories/default--tool--0001.jsonl.gz",
|
| 386 |
+
"records": 3013,
|
| 387 |
+
"sha256": "8efdc46253958b9bfdba85da004f1de435709881773472d627a21372f197ba78",
|
| 388 |
+
"size_bytes": 128599828,
|
| 389 |
+
"source_counts": {
|
| 390 |
+
"swesmith_trajectories": 3013
|
| 391 |
+
}
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"path": "canonical/swesmith_trajectories/default--tool--0002.jsonl.gz",
|
| 395 |
+
"records": 3013,
|
| 396 |
+
"sha256": "80d7a0105bee1b9cde168524b9cd702005786014464a3d5adcb40dcb7a0603ff",
|
| 397 |
+
"size_bytes": 115440687,
|
| 398 |
+
"source_counts": {
|
| 399 |
+
"swesmith_trajectories": 3013
|
| 400 |
+
}
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"path": "canonical/swesmith_trajectories/default--tool--0003.jsonl.gz",
|
| 404 |
+
"records": 3013,
|
| 405 |
+
"sha256": "48df7f6939fd5619d1376c3e6909cfb71d88db56c11ce0511989b51620e44a9c",
|
| 406 |
+
"size_bytes": 147332694,
|
| 407 |
+
"source_counts": {
|
| 408 |
+
"swesmith_trajectories": 3013
|
| 409 |
+
}
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"path": "canonical/swesmith_trajectories/default--tool--0004.jsonl.gz",
|
| 413 |
+
"records": 3012,
|
| 414 |
+
"sha256": "da2007c58f68ec2de3736cf7dbb6b1356e1d1657bfe8ef5f03ecb77b03a7e227",
|
| 415 |
+
"size_bytes": 93517927,
|
| 416 |
+
"source_counts": {
|
| 417 |
+
"swesmith_trajectories": 3012
|
| 418 |
+
}
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"path": "canonical/swesmith_trajectories/default--tool--0005.jsonl.gz",
|
| 422 |
+
"records": 3012,
|
| 423 |
+
"sha256": "97e66d5f2c7e74a53d2d56528d9972ad3ad455ba7472900f8ba474e7bfecba9a",
|
| 424 |
+
"size_bytes": 147229136,
|
| 425 |
+
"source_counts": {
|
| 426 |
+
"swesmith_trajectories": 3012
|
| 427 |
+
}
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"path": "canonical/swesmith_trajectories/default--tool--0006.jsonl.gz",
|
| 431 |
+
"records": 3012,
|
| 432 |
+
"sha256": "fcf23d8b7b0f907ce65e99fd25983b479e099e69506eaeef4c6794f6a663b1cd",
|
| 433 |
+
"size_bytes": 162574019,
|
| 434 |
+
"source_counts": {
|
| 435 |
+
"swesmith_trajectories": 3012
|
| 436 |
+
}
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"path": "canonical/swesmith_trajectories/default--tool--0007.jsonl.gz",
|
| 440 |
+
"records": 3012,
|
| 441 |
+
"sha256": "ff5c4100eb4fa6610b91012d93d071989a7cb33510f9b5e273e19e41a05d7b39",
|
| 442 |
+
"size_bytes": 141721185,
|
| 443 |
+
"source_counts": {
|
| 444 |
+
"swesmith_trajectories": 3012
|
| 445 |
+
}
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"path": "canonical/terminalbench_trajectories/default--train--0000.jsonl.gz",
|
| 449 |
+
"records": 17432,
|
| 450 |
+
"sha256": "ab68edea96136704624b82ec2b8de28bf07135b8cb93da708e6f199ee873e06c",
|
| 451 |
+
"size_bytes": 167147418,
|
| 452 |
+
"source_counts": {
|
| 453 |
+
"terminalbench_trajectories": 17432
|
| 454 |
+
}
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"path": "canonical/terminalbench_trajectories/default--train--0001.jsonl.gz",
|
| 458 |
+
"records": 17030,
|
| 459 |
+
"sha256": "8cacfd4217b7cdfd2e53e481d8a8845a73925f4f44759da29e2c6a550f4f9e63",
|
| 460 |
+
"size_bytes": 154916466,
|
| 461 |
+
"source_counts": {
|
| 462 |
+
"terminalbench_trajectories": 17030
|
| 463 |
+
}
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"path": "canonical/terminaltraj/default--train--0000.jsonl.gz",
|
| 467 |
+
"records": 10000,
|
| 468 |
+
"sha256": "6e17b247614ddfd84792e72de91ad2a5217f2f86086cb05afbcf01a9cfd9e5d3",
|
| 469 |
+
"size_bytes": 89960134,
|
| 470 |
+
"source_counts": {
|
| 471 |
+
"terminaltraj": 10000
|
| 472 |
+
}
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"path": "canonical/terminaltraj/default--train--0001.jsonl.gz",
|
| 476 |
+
"records": 10000,
|
| 477 |
+
"sha256": "f01da77f7ac09de890a7d8b0cb208c3ddc23ef5f2fe125b0d5237bbf1f4b120f",
|
| 478 |
+
"size_bytes": 91844522,
|
| 479 |
+
"source_counts": {
|
| 480 |
+
"terminaltraj": 10000
|
| 481 |
+
}
|
| 482 |
+
}
|
| 483 |
+
],
|
| 484 |
+
"inventory_inputs": [
|
| 485 |
+
{
|
| 486 |
+
"canonical_sha256": "c17428d18c66788e5bbb8c705467f7df49bfa3807c18479c788d3b784ae8cc23",
|
| 487 |
+
"content_availability": "metadata_only",
|
| 488 |
+
"path": null,
|
| 489 |
+
"records": 8340,
|
| 490 |
+
"size_bytes": 116362991,
|
| 491 |
+
"source_counts": {
|
| 492 |
+
"agentsuite_tau2": 8340
|
| 493 |
+
},
|
| 494 |
+
"upstream": [
|
| 495 |
+
"https://huggingface.co/datasets/AgentSuite/tau2-bench-trajectories/tree/07264f1d253a618fc20966670eddc44246212916"
|
| 496 |
+
]
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"canonical_sha256": "4329332610ac06539794aa2781b452c95ffef9a5b633d82c97e083b1ef1961d6",
|
| 500 |
+
"content_availability": "metadata_only",
|
| 501 |
+
"path": null,
|
| 502 |
+
"records": 2531,
|
| 503 |
+
"size_bytes": 217366029,
|
| 504 |
+
"source_counts": {
|
| 505 |
+
"browsecomp_rollouts": 2531
|
| 506 |
+
},
|
| 507 |
+
"upstream": [
|
| 508 |
+
"https://huggingface.co/datasets/rl-rag/browsecomp-gpt-oss-120b-260222/tree/ac4854be0eb527cdf90f6c7d09446eeb7a3d1dbd"
|
| 509 |
+
]
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"canonical_sha256": "a4ba3ccfa48f233d4be820e50a3c1d0c5e8c0213c5af4c6a4dd7f5eea30b2753",
|
| 513 |
+
"content_availability": "metadata_only",
|
| 514 |
+
"path": null,
|
| 515 |
+
"records": 2532,
|
| 516 |
+
"size_bytes": 223325709,
|
| 517 |
+
"source_counts": {
|
| 518 |
+
"browsecomp_rollouts": 2532
|
| 519 |
+
},
|
| 520 |
+
"upstream": [
|
| 521 |
+
"https://huggingface.co/datasets/rl-rag/browsecomp-gpt-oss-120b-260222/tree/ac4854be0eb527cdf90f6c7d09446eeb7a3d1dbd"
|
| 522 |
+
]
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"canonical_sha256": "654357077e6249cd030e516a74bc3ac7551e26553809581bbebb56a4f87e4c37",
|
| 526 |
+
"content_availability": "hosted",
|
| 527 |
+
"path": "canonical/mind2web/default--train--0000.jsonl.gz",
|
| 528 |
+
"records": 100,
|
| 529 |
+
"size_bytes": 60960835,
|
| 530 |
+
"source_counts": {
|
| 531 |
+
"mind2web": 100
|
| 532 |
+
},
|
| 533 |
+
"upstream": [
|
| 534 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 535 |
+
]
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"canonical_sha256": "ea62d59ed99fa6632c59bb62cf339054fe72e717c38d36109526414bb9ce2f60",
|
| 539 |
+
"content_availability": "hosted",
|
| 540 |
+
"path": "canonical/mind2web/default--train--0001.jsonl.gz",
|
| 541 |
+
"records": 100,
|
| 542 |
+
"size_bytes": 61759916,
|
| 543 |
+
"source_counts": {
|
| 544 |
+
"mind2web": 100
|
| 545 |
+
},
|
| 546 |
+
"upstream": [
|
| 547 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 548 |
+
]
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"canonical_sha256": "77d3d9b0ef99cf5ff8300c7e0681847780752ce60e2c2e0887e247b491b48e22",
|
| 552 |
+
"content_availability": "hosted",
|
| 553 |
+
"path": "canonical/mind2web/default--train--0002.jsonl.gz",
|
| 554 |
+
"records": 109,
|
| 555 |
+
"size_bytes": 61958505,
|
| 556 |
+
"source_counts": {
|
| 557 |
+
"mind2web": 109
|
| 558 |
+
},
|
| 559 |
+
"upstream": [
|
| 560 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 561 |
+
]
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"canonical_sha256": "f3807d46087383372f43ca0e3ed95d1b14e9ceca917940c652d136eca5157d7c",
|
| 565 |
+
"content_availability": "hosted",
|
| 566 |
+
"path": "canonical/mind2web/default--train--0003.jsonl.gz",
|
| 567 |
+
"records": 100,
|
| 568 |
+
"size_bytes": 64515683,
|
| 569 |
+
"source_counts": {
|
| 570 |
+
"mind2web": 100
|
| 571 |
+
},
|
| 572 |
+
"upstream": [
|
| 573 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 574 |
+
]
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"canonical_sha256": "f4e03e9b52f92aecac56c551d2a83e1af8fe26933e8366a26cfb90249c9b1f5f",
|
| 578 |
+
"content_availability": "hosted",
|
| 579 |
+
"path": "canonical/mind2web/default--train--0004.jsonl.gz",
|
| 580 |
+
"records": 100,
|
| 581 |
+
"size_bytes": 59759759,
|
| 582 |
+
"source_counts": {
|
| 583 |
+
"mind2web": 100
|
| 584 |
+
},
|
| 585 |
+
"upstream": [
|
| 586 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 587 |
+
]
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"canonical_sha256": "cd075e514c9006f2781cec84a0bbb816e1bc6ccdf22217885925e2ab5d02df8b",
|
| 591 |
+
"content_availability": "hosted",
|
| 592 |
+
"path": "canonical/mind2web/default--train--0005.jsonl.gz",
|
| 593 |
+
"records": 100,
|
| 594 |
+
"size_bytes": 59608293,
|
| 595 |
+
"source_counts": {
|
| 596 |
+
"mind2web": 100
|
| 597 |
+
},
|
| 598 |
+
"upstream": [
|
| 599 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 600 |
+
]
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"canonical_sha256": "fa7ab83febb58ab7a51ecdee07a03369ccc55b8b8ae811dd5679693f98a70c52",
|
| 604 |
+
"content_availability": "hosted",
|
| 605 |
+
"path": "canonical/mind2web/default--train--0006.jsonl.gz",
|
| 606 |
+
"records": 200,
|
| 607 |
+
"size_bytes": 115862382,
|
| 608 |
+
"source_counts": {
|
| 609 |
+
"mind2web": 200
|
| 610 |
+
},
|
| 611 |
+
"upstream": [
|
| 612 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 613 |
+
]
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"canonical_sha256": "aa20a9fa7022a87f523148839fbbccae8576eecd011eb8a8b7a12be2472c8e5e",
|
| 617 |
+
"content_availability": "hosted",
|
| 618 |
+
"path": "canonical/mind2web/default--train--0007.jsonl.gz",
|
| 619 |
+
"records": 200,
|
| 620 |
+
"size_bytes": 116176682,
|
| 621 |
+
"source_counts": {
|
| 622 |
+
"mind2web": 200
|
| 623 |
+
},
|
| 624 |
+
"upstream": [
|
| 625 |
+
"https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474"
|
| 626 |
+
]
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"canonical_sha256": "22d874f857dd91bef6f539e5c5809b5f654bb72c310caad8d42cd1d552189c45",
|
| 630 |
+
"content_availability": "hosted",
|
| 631 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0000.jsonl.gz",
|
| 632 |
+
"records": 2500,
|
| 633 |
+
"size_bytes": 303785042,
|
| 634 |
+
"source_counts": {
|
| 635 |
+
"nvidia_open_swe_traces": 2500
|
| 636 |
+
},
|
| 637 |
+
"upstream": [
|
| 638 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 639 |
+
]
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"canonical_sha256": "be0c9efbdfb9947734634f75102ad4e2be154b4ab63770a748ebcfb0a5677927",
|
| 643 |
+
"content_availability": "hosted",
|
| 644 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0001.jsonl.gz",
|
| 645 |
+
"records": 2500,
|
| 646 |
+
"size_bytes": 295831492,
|
| 647 |
+
"source_counts": {
|
| 648 |
+
"nvidia_open_swe_traces": 2500
|
| 649 |
+
},
|
| 650 |
+
"upstream": [
|
| 651 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 652 |
+
]
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"canonical_sha256": "3067df96c93975319e290aa0b5f7b4be0ceecb6275f761cea33db9b6b3c54ffd",
|
| 656 |
+
"content_availability": "hosted",
|
| 657 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0002.jsonl.gz",
|
| 658 |
+
"records": 2500,
|
| 659 |
+
"size_bytes": 305310880,
|
| 660 |
+
"source_counts": {
|
| 661 |
+
"nvidia_open_swe_traces": 2500
|
| 662 |
+
},
|
| 663 |
+
"upstream": [
|
| 664 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 665 |
+
]
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"canonical_sha256": "2b3324ce204a000df83ad9e207b4318af57ed19d36c11d7cbbae7bf5db90b453",
|
| 669 |
+
"content_availability": "hosted",
|
| 670 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0003.jsonl.gz",
|
| 671 |
+
"records": 2500,
|
| 672 |
+
"size_bytes": 290056585,
|
| 673 |
+
"source_counts": {
|
| 674 |
+
"nvidia_open_swe_traces": 2500
|
| 675 |
+
},
|
| 676 |
+
"upstream": [
|
| 677 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 678 |
+
]
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"canonical_sha256": "5373e01f7f6527a0611ae933bb73b92d6e2ae312ea81d0d3aef85b1b241e55a7",
|
| 682 |
+
"content_availability": "hosted",
|
| 683 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0004.jsonl.gz",
|
| 684 |
+
"records": 2500,
|
| 685 |
+
"size_bytes": 287437480,
|
| 686 |
+
"source_counts": {
|
| 687 |
+
"nvidia_open_swe_traces": 2500
|
| 688 |
+
},
|
| 689 |
+
"upstream": [
|
| 690 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 691 |
+
]
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"canonical_sha256": "92b92e2aba80e79cbb07e9ba071db946fe4aa885deef4aefa9d934ae3bff193e",
|
| 695 |
+
"content_availability": "hosted",
|
| 696 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0005.jsonl.gz",
|
| 697 |
+
"records": 41,
|
| 698 |
+
"size_bytes": 5161644,
|
| 699 |
+
"source_counts": {
|
| 700 |
+
"nvidia_open_swe_traces": 41
|
| 701 |
+
},
|
| 702 |
+
"upstream": [
|
| 703 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 704 |
+
]
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"canonical_sha256": "065ba7c18fbf854df437e352fe79cf54237f22d9290ef0b55283ff7f5c9e7e77",
|
| 708 |
+
"content_availability": "hosted",
|
| 709 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0006.jsonl.gz",
|
| 710 |
+
"records": 2500,
|
| 711 |
+
"size_bytes": 368841972,
|
| 712 |
+
"source_counts": {
|
| 713 |
+
"nvidia_open_swe_traces": 2500
|
| 714 |
+
},
|
| 715 |
+
"upstream": [
|
| 716 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 717 |
+
]
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"canonical_sha256": "af1f7ed30b9037df06ebf7e7f0d146913ad813511e065e8f735a520b646b74bf",
|
| 721 |
+
"content_availability": "hosted",
|
| 722 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0007.jsonl.gz",
|
| 723 |
+
"records": 2500,
|
| 724 |
+
"size_bytes": 371128455,
|
| 725 |
+
"source_counts": {
|
| 726 |
+
"nvidia_open_swe_traces": 2500
|
| 727 |
+
},
|
| 728 |
+
"upstream": [
|
| 729 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 730 |
+
]
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"canonical_sha256": "1b8247219e917730ef629eb0e66f2ec19d91ea404ea546c1ed10961eba7e0d30",
|
| 734 |
+
"content_availability": "hosted",
|
| 735 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0008.jsonl.gz",
|
| 736 |
+
"records": 2500,
|
| 737 |
+
"size_bytes": 367495181,
|
| 738 |
+
"source_counts": {
|
| 739 |
+
"nvidia_open_swe_traces": 2500
|
| 740 |
+
},
|
| 741 |
+
"upstream": [
|
| 742 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 743 |
+
]
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"canonical_sha256": "0aec09bafc26fcd3254794c23b358cb597e94a492d8e0fd153650e859f5997be",
|
| 747 |
+
"content_availability": "hosted",
|
| 748 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0009.jsonl.gz",
|
| 749 |
+
"records": 2500,
|
| 750 |
+
"size_bytes": 370761485,
|
| 751 |
+
"source_counts": {
|
| 752 |
+
"nvidia_open_swe_traces": 2500
|
| 753 |
+
},
|
| 754 |
+
"upstream": [
|
| 755 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 756 |
+
]
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"canonical_sha256": "2eff407239d58e2d9d525aa70402dcf01501ef3d3007d6407a69765c9c4936b5",
|
| 760 |
+
"content_availability": "hosted",
|
| 761 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0010.jsonl.gz",
|
| 762 |
+
"records": 2500,
|
| 763 |
+
"size_bytes": 370224921,
|
| 764 |
+
"source_counts": {
|
| 765 |
+
"nvidia_open_swe_traces": 2500
|
| 766 |
+
},
|
| 767 |
+
"upstream": [
|
| 768 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 769 |
+
]
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"canonical_sha256": "2a68e3249212dbf0661db6e7d7a5b9afc6a2a4bf0172d0f32e9ad0cbf272abbc",
|
| 773 |
+
"content_availability": "hosted",
|
| 774 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0011.jsonl.gz",
|
| 775 |
+
"records": 558,
|
| 776 |
+
"size_bytes": 83589148,
|
| 777 |
+
"source_counts": {
|
| 778 |
+
"nvidia_open_swe_traces": 558
|
| 779 |
+
},
|
| 780 |
+
"upstream": [
|
| 781 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 782 |
+
]
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"canonical_sha256": "d54f3aebffe6bfdd1889488db25bf851210f000d6f6ca2fa2bebf83155fea40a",
|
| 786 |
+
"content_availability": "hosted",
|
| 787 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0012.jsonl.gz",
|
| 788 |
+
"records": 2500,
|
| 789 |
+
"size_bytes": 243965767,
|
| 790 |
+
"source_counts": {
|
| 791 |
+
"nvidia_open_swe_traces": 2500
|
| 792 |
+
},
|
| 793 |
+
"upstream": [
|
| 794 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 795 |
+
]
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"canonical_sha256": "454932733cdc2ce5d6570d064a239ebbe497b0b1915c51f1ea1c3837edc5fae3",
|
| 799 |
+
"content_availability": "hosted",
|
| 800 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0013.jsonl.gz",
|
| 801 |
+
"records": 2500,
|
| 802 |
+
"size_bytes": 241910439,
|
| 803 |
+
"source_counts": {
|
| 804 |
+
"nvidia_open_swe_traces": 2500
|
| 805 |
+
},
|
| 806 |
+
"upstream": [
|
| 807 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 808 |
+
]
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"canonical_sha256": "e964d62766afd638a8526c5e44fca9581e2cea0d3ec0ffd0a14beeefc1e16c6c",
|
| 812 |
+
"content_availability": "hosted",
|
| 813 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0014.jsonl.gz",
|
| 814 |
+
"records": 2500,
|
| 815 |
+
"size_bytes": 249402952,
|
| 816 |
+
"source_counts": {
|
| 817 |
+
"nvidia_open_swe_traces": 2500
|
| 818 |
+
},
|
| 819 |
+
"upstream": [
|
| 820 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 821 |
+
]
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"canonical_sha256": "97fee226a8dc416dadb5673cace864585c36a933d42238abd640526814a5b2c5",
|
| 825 |
+
"content_availability": "hosted",
|
| 826 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0015.jsonl.gz",
|
| 827 |
+
"records": 2500,
|
| 828 |
+
"size_bytes": 240747809,
|
| 829 |
+
"source_counts": {
|
| 830 |
+
"nvidia_open_swe_traces": 2500
|
| 831 |
+
},
|
| 832 |
+
"upstream": [
|
| 833 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 834 |
+
]
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"canonical_sha256": "95b32f5071fe69c1949e66623413b9e02970973ea501040336955b56d6de0b16",
|
| 838 |
+
"content_availability": "hosted",
|
| 839 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0016.jsonl.gz",
|
| 840 |
+
"records": 2500,
|
| 841 |
+
"size_bytes": 242546761,
|
| 842 |
+
"source_counts": {
|
| 843 |
+
"nvidia_open_swe_traces": 2500
|
| 844 |
+
},
|
| 845 |
+
"upstream": [
|
| 846 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 847 |
+
]
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"canonical_sha256": "2c17ad85f0ef74b315314fc6929e3efe7875f5d88f319208bc07fbbbc621c296",
|
| 851 |
+
"content_availability": "hosted",
|
| 852 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0017.jsonl.gz",
|
| 853 |
+
"records": 771,
|
| 854 |
+
"size_bytes": 74960935,
|
| 855 |
+
"source_counts": {
|
| 856 |
+
"nvidia_open_swe_traces": 771
|
| 857 |
+
},
|
| 858 |
+
"upstream": [
|
| 859 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 860 |
+
]
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"canonical_sha256": "331b6e4d1611cc71c643b7a1af41744218d25aa5d6be33675f3ca2bcc20ab183",
|
| 864 |
+
"content_availability": "hosted",
|
| 865 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0018.jsonl.gz",
|
| 866 |
+
"records": 2500,
|
| 867 |
+
"size_bytes": 373422381,
|
| 868 |
+
"source_counts": {
|
| 869 |
+
"nvidia_open_swe_traces": 2500
|
| 870 |
+
},
|
| 871 |
+
"upstream": [
|
| 872 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 873 |
+
]
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"canonical_sha256": "a42a718e33c3b46b9bc01867f57b2166a3ef6de0063e41d6dfb419bb55df971f",
|
| 877 |
+
"content_availability": "hosted",
|
| 878 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0019.jsonl.gz",
|
| 879 |
+
"records": 2500,
|
| 880 |
+
"size_bytes": 371590204,
|
| 881 |
+
"source_counts": {
|
| 882 |
+
"nvidia_open_swe_traces": 2500
|
| 883 |
+
},
|
| 884 |
+
"upstream": [
|
| 885 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 886 |
+
]
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"canonical_sha256": "e412984c9e629a4621083e68f60fa1713ca1678f7eab46c8d0ee25813f967f90",
|
| 890 |
+
"content_availability": "hosted",
|
| 891 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0020.jsonl.gz",
|
| 892 |
+
"records": 2500,
|
| 893 |
+
"size_bytes": 376298397,
|
| 894 |
+
"source_counts": {
|
| 895 |
+
"nvidia_open_swe_traces": 2500
|
| 896 |
+
},
|
| 897 |
+
"upstream": [
|
| 898 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 899 |
+
]
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"canonical_sha256": "22ae50d9032b0850fcf93080fcc81143d6af8f6ecad74c732c6d0428a6b4000c",
|
| 903 |
+
"content_availability": "hosted",
|
| 904 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0021.jsonl.gz",
|
| 905 |
+
"records": 2500,
|
| 906 |
+
"size_bytes": 376587752,
|
| 907 |
+
"source_counts": {
|
| 908 |
+
"nvidia_open_swe_traces": 2500
|
| 909 |
+
},
|
| 910 |
+
"upstream": [
|
| 911 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 912 |
+
]
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"canonical_sha256": "379caa74f2bedf81325a98874fc804c40fd8741f7bf6571b66d72a4b905b6911",
|
| 916 |
+
"content_availability": "hosted",
|
| 917 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0022.jsonl.gz",
|
| 918 |
+
"records": 2500,
|
| 919 |
+
"size_bytes": 380085424,
|
| 920 |
+
"source_counts": {
|
| 921 |
+
"nvidia_open_swe_traces": 2500
|
| 922 |
+
},
|
| 923 |
+
"upstream": [
|
| 924 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 925 |
+
]
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"canonical_sha256": "863309aa10f41341676f0fd71e7e2d68db2d50b3d8451b5b8adab83ff44c515c",
|
| 929 |
+
"content_availability": "hosted",
|
| 930 |
+
"path": "canonical/nvidia_open_swe_traces/v1.2--openhands--0023.jsonl.gz",
|
| 931 |
+
"records": 256,
|
| 932 |
+
"size_bytes": 38763239,
|
| 933 |
+
"source_counts": {
|
| 934 |
+
"nvidia_open_swe_traces": 256
|
| 935 |
+
},
|
| 936 |
+
"upstream": [
|
| 937 |
+
"https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586"
|
| 938 |
+
]
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"canonical_sha256": "0e08775a1a2e47543fb6c3ddd1789660ece1d9fe3bbf6189e457fc06f7bcc594",
|
| 942 |
+
"content_availability": "hosted",
|
| 943 |
+
"path": "canonical/official_tau3+toolbench/official_tau3_toolbench.jsonl.gz",
|
| 944 |
+
"records": 211,
|
| 945 |
+
"size_bytes": 2412172,
|
| 946 |
+
"source_counts": {
|
| 947 |
+
"official_tau3": 200,
|
| 948 |
+
"toolbench": 11
|
| 949 |
+
},
|
| 950 |
+
"upstream": [
|
| 951 |
+
"https://github.com/sierra-research/tau2-bench/tree/dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd",
|
| 952 |
+
"https://github.com/OpenBMB/ToolBench/tree/d56fdd89faf8c91fa135090b212bb9057ee5cfc2"
|
| 953 |
+
]
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"canonical_sha256": "0034093e5c0b717c4957b481d85c9381f9cba025ad16d1b0fcac9152c9698656",
|
| 957 |
+
"content_availability": "hosted",
|
| 958 |
+
"path": "canonical/swebench_multilingual_trajectories/default--Multi_SWE_agent_32B_Rust--0000.jsonl.gz",
|
| 959 |
+
"records": 43,
|
| 960 |
+
"size_bytes": 17147283,
|
| 961 |
+
"source_counts": {
|
| 962 |
+
"swebench_multilingual_trajectories": 43
|
| 963 |
+
},
|
| 964 |
+
"upstream": [
|
| 965 |
+
"https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories/tree/7221c09c6cc0ed16d51c9cb9886a1ab484465670"
|
| 966 |
+
]
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"canonical_sha256": "146462000c0168746f0ca6fccec7f0cf859db13353ccd7504d8e264964eb4441",
|
| 970 |
+
"content_availability": "hosted",
|
| 971 |
+
"path": "canonical/swebench_multilingual_trajectories/default--Qwen2.5_Coder_32B_Instruct--0001.jsonl.gz",
|
| 972 |
+
"records": 43,
|
| 973 |
+
"size_bytes": 36236356,
|
| 974 |
+
"source_counts": {
|
| 975 |
+
"swebench_multilingual_trajectories": 43
|
| 976 |
+
},
|
| 977 |
+
"upstream": [
|
| 978 |
+
"https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories/tree/7221c09c6cc0ed16d51c9cb9886a1ab484465670"
|
| 979 |
+
]
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"canonical_sha256": "bf4d1265c6f2928fc1b8c188299fc5e46892bc3624ca377249c9855bf3e766fa",
|
| 983 |
+
"content_availability": "hosted",
|
| 984 |
+
"path": "canonical/swebench_multilingual_trajectories/default--SWE_agent_LM_32B--0002.jsonl.gz",
|
| 985 |
+
"records": 43,
|
| 986 |
+
"size_bytes": 18456716,
|
| 987 |
+
"source_counts": {
|
| 988 |
+
"swebench_multilingual_trajectories": 43
|
| 989 |
+
},
|
| 990 |
+
"upstream": [
|
| 991 |
+
"https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories/tree/7221c09c6cc0ed16d51c9cb9886a1ab484465670"
|
| 992 |
+
]
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"canonical_sha256": "7da41ea495e4f899daacf55989fcd5be6ef103c870f1f22e2aa11c0a14ce8982",
|
| 996 |
+
"content_availability": "hosted",
|
| 997 |
+
"path": "canonical/swebench_verified_trajectories/all.jsonl.gz",
|
| 998 |
+
"records": 5000,
|
| 999 |
+
"size_bytes": 336638978,
|
| 1000 |
+
"source_counts": {
|
| 1001 |
+
"swebench_verified_trajectories": 5000
|
| 1002 |
+
},
|
| 1003 |
+
"upstream": [
|
| 1004 |
+
"https://huggingface.co/datasets/tarsur385/swebench-verified-trajectories/tree/773748a7c1222e8a642a7059821498e14293562a"
|
| 1005 |
+
]
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"canonical_sha256": "8d5a0fb7c17af0eefdf7ed0748ded69aef38fab9e9c6d661bf384915847777f9",
|
| 1009 |
+
"content_availability": "hosted",
|
| 1010 |
+
"path": "canonical/swegym_openhands/default--train.success.oss--0000.jsonl.gz",
|
| 1011 |
+
"records": 491,
|
| 1012 |
+
"size_bytes": 13990587,
|
| 1013 |
+
"source_counts": {
|
| 1014 |
+
"swegym_openhands": 491
|
| 1015 |
+
},
|
| 1016 |
+
"upstream": [
|
| 1017 |
+
"https://huggingface.co/datasets/SWE-Gym/OpenHands-SFT-Trajectories/tree/4aaa5a4a4b5861f4799d2336908760c190ac3b17"
|
| 1018 |
+
]
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"canonical_sha256": "7a13d09314190f2caefd8b6ccfac49a74c81cb3ff444a62d5b6a38cd89300772",
|
| 1022 |
+
"content_availability": "metadata_only",
|
| 1023 |
+
"path": null,
|
| 1024 |
+
"records": 2019,
|
| 1025 |
+
"size_bytes": 65321488,
|
| 1026 |
+
"source_counts": {
|
| 1027 |
+
"swegym_sampled": 2019
|
| 1028 |
+
},
|
| 1029 |
+
"upstream": [
|
| 1030 |
+
"https://huggingface.co/datasets/SWE-Gym/OpenHands-Sampled-Trajectories/tree/baf3a4e4bff514d48ddc08a93a2ade5c126212c7"
|
| 1031 |
+
]
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"canonical_sha256": "a2addae2568f38226261c33bd83106899f1e27e5e87a0cea0b832bad29ec659e",
|
| 1035 |
+
"content_availability": "metadata_only",
|
| 1036 |
+
"path": null,
|
| 1037 |
+
"records": 2018,
|
| 1038 |
+
"size_bytes": 67493699,
|
| 1039 |
+
"source_counts": {
|
| 1040 |
+
"swegym_sampled": 2018
|
| 1041 |
+
},
|
| 1042 |
+
"upstream": [
|
| 1043 |
+
"https://huggingface.co/datasets/SWE-Gym/OpenHands-Sampled-Trajectories/tree/baf3a4e4bff514d48ddc08a93a2ade5c126212c7"
|
| 1044 |
+
]
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"canonical_sha256": "a472b24ce0f1d557c42c07f20a136c04ec160d77fbf342a0125daf4ff5da25ca",
|
| 1048 |
+
"content_availability": "metadata_only",
|
| 1049 |
+
"path": null,
|
| 1050 |
+
"records": 2018,
|
| 1051 |
+
"size_bytes": 117810358,
|
| 1052 |
+
"source_counts": {
|
| 1053 |
+
"swegym_sampled": 2018
|
| 1054 |
+
},
|
| 1055 |
+
"upstream": [
|
| 1056 |
+
"https://huggingface.co/datasets/SWE-Gym/OpenHands-Sampled-Trajectories/tree/baf3a4e4bff514d48ddc08a93a2ade5c126212c7"
|
| 1057 |
+
]
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"canonical_sha256": "909cd6654ba6b253bf83874168fcfef57eb7ae700d94d39d4e10da15e2d4df22",
|
| 1061 |
+
"content_availability": "hosted",
|
| 1062 |
+
"path": "canonical/swesmith_trajectories/default--tool--0000.jsonl.gz",
|
| 1063 |
+
"records": 3013,
|
| 1064 |
+
"size_bytes": 208736286,
|
| 1065 |
+
"source_counts": {
|
| 1066 |
+
"swesmith_trajectories": 3013
|
| 1067 |
+
},
|
| 1068 |
+
"upstream": [
|
| 1069 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1070 |
+
]
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"canonical_sha256": "8efdc46253958b9bfdba85da004f1de435709881773472d627a21372f197ba78",
|
| 1074 |
+
"content_availability": "hosted",
|
| 1075 |
+
"path": "canonical/swesmith_trajectories/default--tool--0001.jsonl.gz",
|
| 1076 |
+
"records": 3013,
|
| 1077 |
+
"size_bytes": 128599828,
|
| 1078 |
+
"source_counts": {
|
| 1079 |
+
"swesmith_trajectories": 3013
|
| 1080 |
+
},
|
| 1081 |
+
"upstream": [
|
| 1082 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1083 |
+
]
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"canonical_sha256": "80d7a0105bee1b9cde168524b9cd702005786014464a3d5adcb40dcb7a0603ff",
|
| 1087 |
+
"content_availability": "hosted",
|
| 1088 |
+
"path": "canonical/swesmith_trajectories/default--tool--0002.jsonl.gz",
|
| 1089 |
+
"records": 3013,
|
| 1090 |
+
"size_bytes": 115440687,
|
| 1091 |
+
"source_counts": {
|
| 1092 |
+
"swesmith_trajectories": 3013
|
| 1093 |
+
},
|
| 1094 |
+
"upstream": [
|
| 1095 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1096 |
+
]
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"canonical_sha256": "48df7f6939fd5619d1376c3e6909cfb71d88db56c11ce0511989b51620e44a9c",
|
| 1100 |
+
"content_availability": "hosted",
|
| 1101 |
+
"path": "canonical/swesmith_trajectories/default--tool--0003.jsonl.gz",
|
| 1102 |
+
"records": 3013,
|
| 1103 |
+
"size_bytes": 147332694,
|
| 1104 |
+
"source_counts": {
|
| 1105 |
+
"swesmith_trajectories": 3013
|
| 1106 |
+
},
|
| 1107 |
+
"upstream": [
|
| 1108 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1109 |
+
]
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"canonical_sha256": "da2007c58f68ec2de3736cf7dbb6b1356e1d1657bfe8ef5f03ecb77b03a7e227",
|
| 1113 |
+
"content_availability": "hosted",
|
| 1114 |
+
"path": "canonical/swesmith_trajectories/default--tool--0004.jsonl.gz",
|
| 1115 |
+
"records": 3012,
|
| 1116 |
+
"size_bytes": 93517927,
|
| 1117 |
+
"source_counts": {
|
| 1118 |
+
"swesmith_trajectories": 3012
|
| 1119 |
+
},
|
| 1120 |
+
"upstream": [
|
| 1121 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1122 |
+
]
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"canonical_sha256": "97e66d5f2c7e74a53d2d56528d9972ad3ad455ba7472900f8ba474e7bfecba9a",
|
| 1126 |
+
"content_availability": "hosted",
|
| 1127 |
+
"path": "canonical/swesmith_trajectories/default--tool--0005.jsonl.gz",
|
| 1128 |
+
"records": 3012,
|
| 1129 |
+
"size_bytes": 147229136,
|
| 1130 |
+
"source_counts": {
|
| 1131 |
+
"swesmith_trajectories": 3012
|
| 1132 |
+
},
|
| 1133 |
+
"upstream": [
|
| 1134 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1135 |
+
]
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"canonical_sha256": "fcf23d8b7b0f907ce65e99fd25983b479e099e69506eaeef4c6794f6a663b1cd",
|
| 1139 |
+
"content_availability": "hosted",
|
| 1140 |
+
"path": "canonical/swesmith_trajectories/default--tool--0006.jsonl.gz",
|
| 1141 |
+
"records": 3012,
|
| 1142 |
+
"size_bytes": 162574019,
|
| 1143 |
+
"source_counts": {
|
| 1144 |
+
"swesmith_trajectories": 3012
|
| 1145 |
+
},
|
| 1146 |
+
"upstream": [
|
| 1147 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1148 |
+
]
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"canonical_sha256": "ff5c4100eb4fa6610b91012d93d071989a7cb33510f9b5e273e19e41a05d7b39",
|
| 1152 |
+
"content_availability": "hosted",
|
| 1153 |
+
"path": "canonical/swesmith_trajectories/default--tool--0007.jsonl.gz",
|
| 1154 |
+
"records": 3012,
|
| 1155 |
+
"size_bytes": 141721185,
|
| 1156 |
+
"source_counts": {
|
| 1157 |
+
"swesmith_trajectories": 3012
|
| 1158 |
+
},
|
| 1159 |
+
"upstream": [
|
| 1160 |
+
"https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4"
|
| 1161 |
+
]
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"canonical_sha256": "ab68edea96136704624b82ec2b8de28bf07135b8cb93da708e6f199ee873e06c",
|
| 1165 |
+
"content_availability": "hosted",
|
| 1166 |
+
"path": "canonical/terminalbench_trajectories/default--train--0000.jsonl.gz",
|
| 1167 |
+
"records": 17432,
|
| 1168 |
+
"size_bytes": 167147418,
|
| 1169 |
+
"source_counts": {
|
| 1170 |
+
"terminalbench_trajectories": 17432
|
| 1171 |
+
},
|
| 1172 |
+
"upstream": [
|
| 1173 |
+
"https://huggingface.co/datasets/yoonholee/terminalbench-trajectories/tree/04e8940f5b6736a7ce8d22224fe2f2af74163ed2"
|
| 1174 |
+
]
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"canonical_sha256": "8cacfd4217b7cdfd2e53e481d8a8845a73925f4f44759da29e2c6a550f4f9e63",
|
| 1178 |
+
"content_availability": "hosted",
|
| 1179 |
+
"path": "canonical/terminalbench_trajectories/default--train--0001.jsonl.gz",
|
| 1180 |
+
"records": 17030,
|
| 1181 |
+
"size_bytes": 154916466,
|
| 1182 |
+
"source_counts": {
|
| 1183 |
+
"terminalbench_trajectories": 17030
|
| 1184 |
+
},
|
| 1185 |
+
"upstream": [
|
| 1186 |
+
"https://huggingface.co/datasets/yoonholee/terminalbench-trajectories/tree/04e8940f5b6736a7ce8d22224fe2f2af74163ed2"
|
| 1187 |
+
]
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"canonical_sha256": "6e17b247614ddfd84792e72de91ad2a5217f2f86086cb05afbcf01a9cfd9e5d3",
|
| 1191 |
+
"content_availability": "hosted",
|
| 1192 |
+
"path": "canonical/terminaltraj/default--train--0000.jsonl.gz",
|
| 1193 |
+
"records": 10000,
|
| 1194 |
+
"size_bytes": 89960134,
|
| 1195 |
+
"source_counts": {
|
| 1196 |
+
"terminaltraj": 10000
|
| 1197 |
+
},
|
| 1198 |
+
"upstream": [
|
| 1199 |
+
"https://huggingface.co/datasets/m-a-p/TerminalTraj/tree/5c1823f4a8b9ca0cf02c27d2db52c5b35b53a308"
|
| 1200 |
+
]
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"canonical_sha256": "f01da77f7ac09de890a7d8b0cb208c3ddc23ef5f2fe125b0d5237bbf1f4b120f",
|
| 1204 |
+
"content_availability": "hosted",
|
| 1205 |
+
"path": "canonical/terminaltraj/default--train--0001.jsonl.gz",
|
| 1206 |
+
"records": 10000,
|
| 1207 |
+
"size_bytes": 91844522,
|
| 1208 |
+
"source_counts": {
|
| 1209 |
+
"terminaltraj": 10000
|
| 1210 |
+
},
|
| 1211 |
+
"upstream": [
|
| 1212 |
+
"https://huggingface.co/datasets/m-a-p/TerminalTraj/tree/5c1823f4a8b9ca0cf02c27d2db52c5b35b53a308"
|
| 1213 |
+
]
|
| 1214 |
+
}
|
| 1215 |
+
],
|
| 1216 |
+
"metadata_only_source_counts": {
|
| 1217 |
+
"agentsuite_tau2": 8340,
|
| 1218 |
+
"browsecomp_rollouts": 5063,
|
| 1219 |
+
"swegym_sampled": 5502
|
| 1220 |
+
},
|
| 1221 |
+
"metadata_only_trajectories": 18905,
|
| 1222 |
+
"outcome_counts": {
|
| 1223 |
+
"failure": 79973,
|
| 1224 |
+
"success": 54296,
|
| 1225 |
+
"unknown": 21009
|
| 1226 |
+
},
|
| 1227 |
+
"partition_counts": {
|
| 1228 |
+
"structurally_eligible": 143999,
|
| 1229 |
+
"successful": 54296,
|
| 1230 |
+
"training_eligible": 110208,
|
| 1231 |
+
"unknown": 21009,
|
| 1232 |
+
"unsuccessful": 79973
|
| 1233 |
+
},
|
| 1234 |
+
"permissions": "registry/permissions.json",
|
| 1235 |
+
"ranked": 155278,
|
| 1236 |
+
"schema_version": 1,
|
| 1237 |
+
"source_counts": {
|
| 1238 |
+
"agentsuite_tau2": 8340,
|
| 1239 |
+
"browsecomp_rollouts": 5063,
|
| 1240 |
+
"mind2web": 1009,
|
| 1241 |
+
"nvidia_open_swe_traces": 51626,
|
| 1242 |
+
"official_tau3": 200,
|
| 1243 |
+
"swebench_multilingual_trajectories": 129,
|
| 1244 |
+
"swebench_verified_trajectories": 4544,
|
| 1245 |
+
"swegym_openhands": 491,
|
| 1246 |
+
"swegym_sampled": 5502,
|
| 1247 |
+
"swesmith_trajectories": 24100,
|
| 1248 |
+
"terminalbench_trajectories": 34263,
|
| 1249 |
+
"terminaltraj": 20000,
|
| 1250 |
+
"toolbench": 11
|
| 1251 |
+
},
|
| 1252 |
+
"source_record_counts": {
|
| 1253 |
+
"agentsuite_tau2": 8340,
|
| 1254 |
+
"browsecomp_rollouts": 5063,
|
| 1255 |
+
"mind2web": 1009,
|
| 1256 |
+
"nvidia_open_swe_traces": 51626,
|
| 1257 |
+
"official_tau3": 200,
|
| 1258 |
+
"swebench_multilingual_trajectories": 129,
|
| 1259 |
+
"swebench_verified_trajectories": 5000,
|
| 1260 |
+
"swegym_openhands": 491,
|
| 1261 |
+
"swegym_sampled": 6055,
|
| 1262 |
+
"swesmith_trajectories": 24100,
|
| 1263 |
+
"terminalbench_trajectories": 34462,
|
| 1264 |
+
"terminaltraj": 20000,
|
| 1265 |
+
"toolbench": 11
|
| 1266 |
+
},
|
| 1267 |
+
"source_records": 156486,
|
| 1268 |
+
"structurally_eligible": 143999,
|
| 1269 |
+
"task_split_counts": {
|
| 1270 |
+
"protected_evaluation": 1948,
|
| 1271 |
+
"test": 6442,
|
| 1272 |
+
"train": 52592,
|
| 1273 |
+
"validation": 6628
|
| 1274 |
+
},
|
| 1275 |
+
"training_eligible": 110208,
|
| 1276 |
+
"trajectories": 155278
|
| 1277 |
+
}
|
rankings.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a56a9ae57c860130af600fb534afaaa337124c15646573ec9f8dc86b950fa41
|
| 3 |
+
size 10959026
|
registry/permissions.json
ADDED
|
@@ -0,0 +1,215 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"schema_version": 1,
|
| 3 |
+
"audit_date": "2026-10-14",
|
| 4 |
+
"vocabulary": {
|
| 5 |
+
"permission": ["permitted", "permitted_with_conditions", "uncertain", "restricted"],
|
| 6 |
+
"verification": ["verified", "inferred", "uncertain"],
|
| 7 |
+
"public_payload": ["hosted", "metadata_only"]
|
| 8 |
+
},
|
| 9 |
+
"scope": "Source-level engineering audit for the 13 sources represented in the 155,278-record canonical inventory. Permissions apply to the identified upstream revision and can differ for embedded repository code, web content, model outputs, or later releases.",
|
| 10 |
+
"policy": "Keep the complete inventory and derived metadata. Host canonical payload only when redistribution is supported by source-specific evidence; otherwise publish metadata, hashes, pinned upstream locators, and authorized-overlay instructions.",
|
| 11 |
+
"disclaimer": "This is an evidence-based engineering audit, not legal advice. Downstream users remain responsible for their use case and third-party rights.",
|
| 12 |
+
"sources": [
|
| 13 |
+
{
|
| 14 |
+
"source_id": "swegym_openhands",
|
| 15 |
+
"upstream": "https://huggingface.co/datasets/SWE-Gym/OpenHands-SFT-Trajectories/tree/4aaa5a4a4b5861f4799d2336908760c190ac3b17",
|
| 16 |
+
"declared_license": "MIT",
|
| 17 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 18 |
+
"local_research_use": {"permission": "permitted", "verification": "verified"},
|
| 19 |
+
"model_training": {"permission": "permitted", "verification": "verified", "conditions": ["retain MIT notice"]},
|
| 20 |
+
"commercial_training": {"permission": "permitted", "verification": "verified", "conditions": ["retain MIT notice"]},
|
| 21 |
+
"redistribution_original": {"permission": "permitted", "verification": "verified", "conditions": ["retain MIT notice"]},
|
| 22 |
+
"redistribution_transformed": {"permission": "permitted", "verification": "verified", "conditions": ["retain MIT notice and source attribution"]},
|
| 23 |
+
"public_payload": "hosted",
|
| 24 |
+
"evidence": [
|
| 25 |
+
{"finding": "Pinned dataset card declares MIT.", "url": "https://huggingface.co/datasets/SWE-Gym/OpenHands-SFT-Trajectories/blob/4aaa5a4a4b5861f4799d2336908760c190ac3b17/README.md"},
|
| 26 |
+
{"finding": "The SWE-Gym project explicitly documents this dataset for model training.", "url": "https://github.com/SWE-Gym/SWE-Gym/blob/b681068ca20628c6987b7416cc4cf03f06b77ba5/docs/OpenHands.md"}
|
| 27 |
+
]
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"source_id": "nvidia_open_swe_traces",
|
| 31 |
+
"upstream": "https://huggingface.co/datasets/nvidia/Open-SWE-Traces/tree/6d0a83373ae0d83abe3d363f1f18cb1822e16586",
|
| 32 |
+
"declared_license": "CC-BY-4.0 plus per-repository permissive licenses",
|
| 33 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 34 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["provide attribution"]},
|
| 35 |
+
"model_training": {"permission": "permitted", "verification": "verified", "conditions": ["provide attribution", "honor the per-record repository license"]},
|
| 36 |
+
"commercial_training": {"permission": "permitted", "verification": "verified", "conditions": ["provide attribution", "honor the per-record repository license"]},
|
| 37 |
+
"redistribution_original": {"permission": "permitted", "verification": "verified", "conditions": ["CC-BY-4.0 attribution", "retain per-record repository license metadata"]},
|
| 38 |
+
"redistribution_transformed": {"permission": "permitted", "verification": "verified", "conditions": ["CC-BY-4.0 attribution", "identify canonical transformations", "retain per-record repository license metadata"]},
|
| 39 |
+
"public_payload": "hosted",
|
| 40 |
+
"evidence": [
|
| 41 |
+
{"finding": "Card declares CC-BY-4.0, commercial/non-commercial readiness, permissive task repositories, and SFT/distillation as intended uses.", "url": "https://huggingface.co/datasets/nvidia/Open-SWE-Traces/blob/6d0a83373ae0d83abe3d363f1f18cb1822e16586/README.md"}
|
| 42 |
+
]
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"source_id": "terminaltraj",
|
| 46 |
+
"upstream": "https://huggingface.co/datasets/m-a-p/TerminalTraj/tree/5c1823f4a8b9ca0cf02c27d2db52c5b35b53a308",
|
| 47 |
+
"declared_license": "Apache-2.0 project license; dataset card omits a license field",
|
| 48 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 49 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["honor embedded repository licenses"]},
|
| 50 |
+
"model_training": {"permission": "permitted", "verification": "verified", "conditions": ["honor embedded repository licenses"]},
|
| 51 |
+
"commercial_training": {"permission": "permitted_with_conditions", "verification": "inferred", "conditions": ["confirm embedded repository obligations for the selected records"]},
|
| 52 |
+
"redistribution_original": {"permission": "permitted_with_conditions", "verification": "inferred", "conditions": ["Apache-2.0 notice", "honor embedded MIT/Apache-2.0/BSD-2-Clause/BSD-3-Clause repository terms"]},
|
| 53 |
+
"redistribution_transformed": {"permission": "permitted_with_conditions", "verification": "inferred", "conditions": ["Apache-2.0 notice", "source attribution", "honor embedded repository terms"]},
|
| 54 |
+
"public_payload": "hosted",
|
| 55 |
+
"evidence": [
|
| 56 |
+
{"finding": "The publisher's project links this exact dataset as its released training dataset, carries Apache-2.0, and says embedded repositories are limited to permissive MIT/Apache/BSD licenses.", "url": "https://github.com/multimodal-art-projection/TerminalTraj"},
|
| 57 |
+
{"finding": "Pinned dataset card links the same paper but does not repeat the project license.", "url": "https://huggingface.co/datasets/m-a-p/TerminalTraj/blob/5c1823f4a8b9ca0cf02c27d2db52c5b35b53a308/README.md"}
|
| 58 |
+
]
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"source_id": "agentsuite_tau2",
|
| 62 |
+
"upstream": "https://huggingface.co/datasets/AgentSuite/tau2-bench-trajectories/tree/07264f1d253a618fc20966670eddc44246212916",
|
| 63 |
+
"declared_license": null,
|
| 64 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 65 |
+
"local_research_use": {"permission": "permitted_with_conditions", "verification": "inferred", "conditions": ["public-release research context", "honor tau2-bench MIT terms and model-provider terms"]},
|
| 66 |
+
"model_training": {"permission": "uncertain", "verification": "uncertain", "conditions": ["obtain or confirm trajectory publisher and model-output rights"]},
|
| 67 |
+
"commercial_training": {"permission": "uncertain", "verification": "uncertain"},
|
| 68 |
+
"redistribution_original": {"permission": "uncertain", "verification": "uncertain"},
|
| 69 |
+
"redistribution_transformed": {"permission": "uncertain", "verification": "uncertain"},
|
| 70 |
+
"public_payload": "metadata_only",
|
| 71 |
+
"evidence": [
|
| 72 |
+
{"finding": "Pinned dataset is explicitly a public release but supplies no license or usage terms.", "url": "https://huggingface.co/datasets/AgentSuite/tau2-bench-trajectories/blob/07264f1d253a618fc20966670eddc44246212916/README.md"},
|
| 73 |
+
{"finding": "AgentSuite vendors tau2-bench under MIT, but that does not by itself license separately hosted model outputs.", "url": "https://github.com/Agent-Suite/AgentSuite/tree/0f9eac1c1a376a411ad807bd974555055f08e6c5/tau2-bench"}
|
| 74 |
+
]
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"source_id": "official_tau3",
|
| 78 |
+
"upstream": "https://github.com/sierra-research/tau2-bench/tree/dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd",
|
| 79 |
+
"declared_license": "MIT",
|
| 80 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 81 |
+
"local_research_use": {"permission": "permitted", "verification": "verified"},
|
| 82 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["training on benchmark tasks or rewards creates evaluation overlap and must be disclosed as a custom submission"]},
|
| 83 |
+
"commercial_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "disclose benchmark contamination"]},
|
| 84 |
+
"redistribution_original": {"permission": "permitted", "verification": "verified", "conditions": ["retain MIT notice"]},
|
| 85 |
+
"redistribution_transformed": {"permission": "permitted", "verification": "verified", "conditions": ["retain MIT notice and source attribution"]},
|
| 86 |
+
"public_payload": "hosted",
|
| 87 |
+
"evidence": [
|
| 88 |
+
{"finding": "The exact result JSON files are bundled in the pinned MIT-licensed repository.", "url": "https://github.com/sierra-research/tau2-bench/tree/dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd/data/tau2/results/final"},
|
| 89 |
+
{"finding": "Training is supported, while leaderboard rules classify domain/task overlap as a custom submission requiring disclosure.", "url": "https://github.com/sierra-research/tau2-bench/blob/dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd/docs/leaderboard-submission.md"}
|
| 90 |
+
]
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"source_id": "mind2web",
|
| 94 |
+
"upstream": "https://huggingface.co/datasets/osunlp/Mind2Web/tree/17ece8eb89862368edc0cc806acee6fca5163474",
|
| 95 |
+
"declared_license": "CC-BY-4.0 with a research-only intended-use statement",
|
| 96 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 97 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["provide attribution"]},
|
| 98 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["research use", "provide attribution", "avoid harmful use"]},
|
| 99 |
+
"commercial_training": {"permission": "restricted", "verification": "verified", "conditions": ["dataset card states it was released solely for research purposes despite CC-BY metadata"]},
|
| 100 |
+
"redistribution_original": {"permission": "permitted", "verification": "verified", "conditions": ["CC-BY-4.0 attribution", "retain research-use disclaimer"]},
|
| 101 |
+
"redistribution_transformed": {"permission": "permitted", "verification": "verified", "conditions": ["CC-BY-4.0 attribution", "identify canonical transformations", "retain research-use disclaimer"]},
|
| 102 |
+
"public_payload": "hosted",
|
| 103 |
+
"evidence": [
|
| 104 |
+
{"finding": "Pinned card declares CC-BY-4.0 and separately states research-only release intent.", "url": "https://huggingface.co/datasets/osunlp/Mind2Web/blob/17ece8eb89862368edc0cc806acee6fca5163474/README.md"}
|
| 105 |
+
]
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"source_id": "toolbench",
|
| 109 |
+
"upstream": "https://github.com/OpenBMB/ToolBench/tree/d56fdd89faf8c91fa135090b212bb9057ee5cfc2",
|
| 110 |
+
"declared_license": "Apache-2.0 with research-and-education intended use",
|
| 111 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 112 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["research or education", "consider third-party API rights"]},
|
| 113 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["research or education", "consider third-party API rights"]},
|
| 114 |
+
"commercial_training": {"permission": "restricted", "verification": "verified", "conditions": ["README states ToolBench is intended solely for research and educational purposes"]},
|
| 115 |
+
"redistribution_original": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["Apache-2.0 notice", "research/education intended use", "third-party API content remains subject to third-party rights"]},
|
| 116 |
+
"redistribution_transformed": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["Apache-2.0 notice", "source attribution", "research/education intended use", "third-party API content remains subject to third-party rights"]},
|
| 117 |
+
"public_payload": "hosted",
|
| 118 |
+
"evidence": [
|
| 119 |
+
{"finding": "Pinned README says the dataset is distributed under Apache-2.0, is intended solely for research and education, and provides SFT data/training instructions.", "url": "https://github.com/OpenBMB/ToolBench/blob/d56fdd89faf8c91fa135090b212bb9057ee5cfc2/README.md"}
|
| 120 |
+
]
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"source_id": "swegym_sampled",
|
| 124 |
+
"upstream": "https://huggingface.co/datasets/SWE-Gym/OpenHands-Sampled-Trajectories/tree/baf3a4e4bff514d48ddc08a93a2ade5c126212c7",
|
| 125 |
+
"declared_license": null,
|
| 126 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 127 |
+
"local_research_use": {"permission": "permitted_with_conditions", "verification": "inferred", "conditions": ["SWE-Gym project is Apache-2.0 and research/training oriented", "confirm dataset-specific terms"]},
|
| 128 |
+
"model_training": {"permission": "uncertain", "verification": "uncertain"},
|
| 129 |
+
"commercial_training": {"permission": "uncertain", "verification": "uncertain"},
|
| 130 |
+
"redistribution_original": {"permission": "uncertain", "verification": "uncertain"},
|
| 131 |
+
"redistribution_transformed": {"permission": "uncertain", "verification": "uncertain"},
|
| 132 |
+
"public_payload": "metadata_only",
|
| 133 |
+
"evidence": [
|
| 134 |
+
{"finding": "Pinned dataset card has no license or usage statement.", "url": "https://huggingface.co/datasets/SWE-Gym/OpenHands-Sampled-Trajectories/blob/baf3a4e4bff514d48ddc08a93a2ade5c126212c7/README.md"},
|
| 135 |
+
{"finding": "The related SWE-Gym project is Apache-2.0, but its training documentation references a different SFT trajectory dataset and does not establish terms for this separate repository.", "url": "https://github.com/SWE-Gym/SWE-Gym/tree/b681068ca20628c6987b7416cc4cf03f06b77ba5"}
|
| 136 |
+
]
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"source_id": "swebench_verified_trajectories",
|
| 140 |
+
"upstream": "https://huggingface.co/datasets/tarsur385/swebench-verified-trajectories/tree/773748a7c1222e8a642a7059821498e14293562a",
|
| 141 |
+
"declared_license": "MIT",
|
| 142 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 143 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["honor underlying task repository licenses"]},
|
| 144 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "honor underlying task repository licenses", "avoid evaluation leakage"]},
|
| 145 |
+
"commercial_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["honor underlying task repository and model-provider terms"]},
|
| 146 |
+
"redistribution_original": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "underlying task repositories retain their licenses"]},
|
| 147 |
+
"redistribution_transformed": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice and attribution", "underlying task repositories retain their licenses"]},
|
| 148 |
+
"public_payload": "hosted",
|
| 149 |
+
"evidence": [
|
| 150 |
+
{"finding": "Pinned dataset card declares MIT and identifies the trajectory provenance.", "url": "https://huggingface.co/datasets/tarsur385/swebench-verified-trajectories/blob/773748a7c1222e8a642a7059821498e14293562a/README.md"}
|
| 151 |
+
]
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"source_id": "swebench_multilingual_trajectories",
|
| 155 |
+
"upstream": "https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories/tree/7221c09c6cc0ed16d51c9cb9886a1ab484465670",
|
| 156 |
+
"declared_license": "MIT",
|
| 157 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 158 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["honor underlying task repository licenses"]},
|
| 159 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "honor underlying task repository licenses", "avoid evaluation leakage"]},
|
| 160 |
+
"commercial_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["honor underlying task repository and model-provider terms"]},
|
| 161 |
+
"redistribution_original": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "underlying task repositories retain their licenses"]},
|
| 162 |
+
"redistribution_transformed": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice and attribution", "underlying task repositories retain their licenses"]},
|
| 163 |
+
"public_payload": "hosted",
|
| 164 |
+
"evidence": [
|
| 165 |
+
{"finding": "Pinned dataset card declares MIT and describes the three trajectory splits/models.", "url": "https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories/blob/7221c09c6cc0ed16d51c9cb9886a1ab484465670/README.md"}
|
| 166 |
+
]
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"source_id": "terminalbench_trajectories",
|
| 170 |
+
"upstream": "https://huggingface.co/datasets/yoonholee/terminalbench-trajectories/tree/04e8940f5b6736a7ce8d22224fe2f2af74163ed2",
|
| 171 |
+
"declared_license": "Apache-2.0",
|
| 172 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 173 |
+
"local_research_use": {"permission": "permitted", "verification": "verified"},
|
| 174 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain Apache-2.0 notice", "avoid or disclose Terminal-Bench evaluation leakage"]},
|
| 175 |
+
"commercial_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["review model-provider terms represented in the scraped public leaderboard trajectories"]},
|
| 176 |
+
"redistribution_original": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain Apache-2.0 notice and attribution", "source says trajectories were scraped from public leaderboard data"]},
|
| 177 |
+
"redistribution_transformed": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain Apache-2.0 notice and attribution", "identify canonical transformations"]},
|
| 178 |
+
"public_payload": "hosted",
|
| 179 |
+
"evidence": [
|
| 180 |
+
{"finding": "Pinned dataset card declares Apache-2.0, describes complete trajectories, and identifies the public leaderboard source.", "url": "https://huggingface.co/datasets/yoonholee/terminalbench-trajectories/blob/04e8940f5b6736a7ce8d22224fe2f2af74163ed2/README.md"}
|
| 181 |
+
]
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"source_id": "browsecomp_rollouts",
|
| 185 |
+
"upstream": "https://huggingface.co/datasets/rl-rag/browsecomp-gpt-oss-120b-260222/tree/ac4854be0eb527cdf90f6c7d09446eeb7a3d1dbd",
|
| 186 |
+
"declared_license": null,
|
| 187 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 188 |
+
"local_research_use": {"permission": "permitted_with_conditions", "verification": "inferred", "conditions": ["underlying BrowseComp data is MIT", "web observations retain third-party rights"]},
|
| 189 |
+
"model_training": {"permission": "uncertain", "verification": "uncertain", "conditions": ["trajectory publisher terms absent", "web observations may contain third-party content"]},
|
| 190 |
+
"commercial_training": {"permission": "uncertain", "verification": "uncertain"},
|
| 191 |
+
"redistribution_original": {"permission": "uncertain", "verification": "uncertain"},
|
| 192 |
+
"redistribution_transformed": {"permission": "uncertain", "verification": "uncertain"},
|
| 193 |
+
"public_payload": "metadata_only",
|
| 194 |
+
"evidence": [
|
| 195 |
+
{"finding": "Pinned trajectory card has no license and includes full model conversations plus search/open results from third-party sites.", "url": "https://huggingface.co/datasets/rl-rag/browsecomp-gpt-oss-120b-260222/blob/ac4854be0eb527cdf90f6c7d09446eeb7a3d1dbd/README.md"},
|
| 196 |
+
{"finding": "Underlying BrowseComp questions are MIT in OpenAI simple-evals; that does not automatically license the separate rollouts or retrieved web content.", "url": "https://github.com/openai/simple-evals/tree/652c89d0ca9df547706735883097e9537d40dc47"}
|
| 197 |
+
]
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"source_id": "swesmith_trajectories",
|
| 201 |
+
"upstream": "https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/tree/08e109b4a59eaeebf80e4675cd125d42e7ac99a4",
|
| 202 |
+
"declared_license": "MIT",
|
| 203 |
+
"public_access": {"permission": "permitted", "verification": "verified"},
|
| 204 |
+
"local_research_use": {"permission": "permitted", "verification": "verified", "conditions": ["honor generated task repository licenses"]},
|
| 205 |
+
"model_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "honor generated task repository licenses"]},
|
| 206 |
+
"commercial_training": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["honor generated task repository and model-provider terms"]},
|
| 207 |
+
"redistribution_original": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice", "generated task repositories retain their licenses"]},
|
| 208 |
+
"redistribution_transformed": {"permission": "permitted_with_conditions", "verification": "verified", "conditions": ["retain MIT notice and attribution", "generated task repositories retain their licenses"]},
|
| 209 |
+
"public_payload": "hosted",
|
| 210 |
+
"evidence": [
|
| 211 |
+
{"finding": "Pinned dataset card declares MIT and states the trajectories were used to fine-tune SWE-agent-LM.", "url": "https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories/blob/08e109b4a59eaeebf80e4675cd125d42e7ac99a4/README.md"}
|
| 212 |
+
]
|
| 213 |
+
}
|
| 214 |
+
]
|
| 215 |
+
}
|
registry/sources.json
ADDED
|
@@ -0,0 +1,465 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"registry_version": 1,
|
| 3 |
+
"audit_date": "2026-10-14",
|
| 4 |
+
"sources": [
|
| 5 |
+
{
|
| 6 |
+
"id": "swegym_openhands",
|
| 7 |
+
"name": "SWE-Gym OpenHands SFT Trajectories",
|
| 8 |
+
"url": "https://huggingface.co/datasets/SWE-Gym/OpenHands-SFT-Trajectories",
|
| 9 |
+
"pinned_revision": "4aaa5a4a4b5861f4799d2336908760c190ac3b17",
|
| 10 |
+
"related_code_revision": "b681068ca20628c6987b7416cc4cf03f06b77ba5",
|
| 11 |
+
"available_splits": ["train.success.oss"],
|
| 12 |
+
"upstream_record_count": 491,
|
| 13 |
+
"qualifying_trajectory_count": 491,
|
| 14 |
+
"verified_record_count": 491,
|
| 15 |
+
"reported_statistics": "Dataset server reports 491 rows; all 491 were imported and passed the sequential-action/context gate.",
|
| 16 |
+
"schema_summary": "Messages-only conversations with system/user/assistant turns; assistant tool actions and execution results are encoded in message content.",
|
| 17 |
+
"full_trajectory": "yes for all 491 audited rows",
|
| 18 |
+
"provenance": "OpenHands agent rollouts; exact model is not supplied per row",
|
| 19 |
+
"verification": "Published train.success.oss split label; not independently replayed",
|
| 20 |
+
"eligibility": "conditionally-usable",
|
| 21 |
+
"license": "MIT",
|
| 22 |
+
"task_id": "Stable hash of the initial user task because rows do not expose an instance ID",
|
| 23 |
+
"duplicate_risks": ["SWE-bench task overlap", "multiple attempts per instance"],
|
| 24 |
+
"limitations": ["Tool calls/results are content-encoded rather than structured fields", "Success is source-label-derived rather than independently replayed"]
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"id": "nvidia_open_swe_traces",
|
| 28 |
+
"name": "NVIDIA Open-SWE-Traces",
|
| 29 |
+
"url": "https://huggingface.co/datasets/nvidia/Open-SWE-Traces",
|
| 30 |
+
"pinned_revision": "6d0a83373ae0d83abe3d363f1f18cb1822e16586",
|
| 31 |
+
"available_splits": ["v1.0: openhands/sweagent", "v1.1: openhands/sweagent/minisweagent", "v1.2: openhands/sweagent/minisweagent"],
|
| 32 |
+
"upstream_record_count": 677815,
|
| 33 |
+
"qualifying_trajectory_count": 677815,
|
| 34 |
+
"not_imported_due_resource_limit": 626189,
|
| 35 |
+
"rejected_before_library": 0,
|
| 36 |
+
"verified_record_count": 51626,
|
| 37 |
+
"reported_statistics": "The pinned release enumerates 677,815 rows (v1.0 151,219; v1.1 263,226; v1.2 263,370). All 51,626 v1.2 OpenHands rows were imported; 626,189 rows in the other version/scaffold splits remain unimported.",
|
| 38 |
+
"schema_summary": "Verified fields include trajectory_id, instance_id, repo, resolved, messages, tools, metadata, language and license; messages/tools/metadata are JSON-encoded strings.",
|
| 39 |
+
"full_trajectory": "yes",
|
| 40 |
+
"provenance": "Model agent traces with model/scaffold metadata preserved from each row",
|
| 41 |
+
"verification": "Published resolved field; not independently replayed here",
|
| 42 |
+
"eligibility": "conditionally-usable",
|
| 43 |
+
"license": "CC-BY-4.0",
|
| 44 |
+
"task_id": "Native instance_id",
|
| 45 |
+
"duplicate_risks": ["SWE-bench and SWE-Gym overlap", "multiple attempts"],
|
| 46 |
+
"limitations": ["Only v1.2/OpenHands is persisted locally", "The remaining 626,189 rows comprise 55,056,684,304 bytes of source Parquet", "Source outcomes were not independently replayed"]
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"id": "terminaltraj",
|
| 50 |
+
"name": "TerminalTraj",
|
| 51 |
+
"url": "https://huggingface.co/datasets/m-a-p/TerminalTraj",
|
| 52 |
+
"pinned_revision": "5c1823f4a8b9ca0cf02c27d2db52c5b35b53a308",
|
| 53 |
+
"available_splits": ["train"],
|
| 54 |
+
"upstream_record_count": 20000,
|
| 55 |
+
"qualifying_trajectory_count": 19143,
|
| 56 |
+
"verified_record_count": 20000,
|
| 57 |
+
"reported_statistics": "All 20,000 train trajectories imported; 19,143 contain at least two fully contextualized assistant actions.",
|
| 58 |
+
"schema_summary": "Alternating user terminal-output and assistant JSON action messages containing analysis, plan, commands and task_complete.",
|
| 59 |
+
"full_trajectory": "yes for audited rows",
|
| 60 |
+
"provenance": "Terminal agent command/output trajectories; exact model is not supplied per row",
|
| 61 |
+
"verification": "Paired command/output transcript; no trustworthy task-level success field",
|
| 62 |
+
"eligibility": "conditionally-usable",
|
| 63 |
+
"license": "unverified at dataset revision",
|
| 64 |
+
"task_id": "Native task ID required",
|
| 65 |
+
"duplicate_risks": ["Terminal-Bench overlap", "repeated attempts"],
|
| 66 |
+
"limitations": ["No trustworthy binary success indicator", "857 imported records are short or lack full intervening context and remain tier-5 diagnostics"]
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"id": "agentsuite_tau2",
|
| 70 |
+
"name": "AgentSuite tau2-bench trajectories",
|
| 71 |
+
"url": "https://huggingface.co/datasets/AgentSuite/tau2-bench-trajectories",
|
| 72 |
+
"pinned_revision": "07264f1d253a618fc20966670eddc44246212916",
|
| 73 |
+
"available_splits": ["train"],
|
| 74 |
+
"upstream_record_count": 8340,
|
| 75 |
+
"qualifying_trajectory_count": 8337,
|
| 76 |
+
"verified_record_count": 8340,
|
| 77 |
+
"reported_statistics": "All 8,340 public trajectories (30 models x 278 tasks) imported from the pinned native JSONL files.",
|
| 78 |
+
"schema_summary": "System/user/assistant/tool messages plus eval_result.score/db_match and meta.is_correct/reward_details/finish_reason.",
|
| 79 |
+
"full_trajectory": "yes for all 8,340 imported rows",
|
| 80 |
+
"provenance": "Model agent plus model user simulator rollouts",
|
| 81 |
+
"verification": "AgentSuite evaluator fields; meta.is_correct is used for the binary outcome",
|
| 82 |
+
"eligibility": "conditionally-usable",
|
| 83 |
+
"license": "unverified at dataset revision",
|
| 84 |
+
"task_id": "Domain plus native task ID",
|
| 85 |
+
"duplicate_risks": ["official tau task overlap", "tau2/tau3 version mismatch"],
|
| 86 |
+
"limitations": ["Tool definitions are absent", "Three records are explicitly incomplete because a source tool call has no returned observation", "Must not be treated as version-matched tau3 demonstrations", "Evaluator results were not independently replayed"]
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"id": "official_tau3",
|
| 90 |
+
"name": "Official tau3-bench",
|
| 91 |
+
"url": "https://github.com/sierra-research/tau2-bench",
|
| 92 |
+
"pinned_revision": "dd8757d4c8791dad7b3cdaf02a105a3d5f9d86bd",
|
| 93 |
+
"available_splits": ["base and repository-defined evaluation splits"],
|
| 94 |
+
"upstream_record_count": 200,
|
| 95 |
+
"qualifying_trajectory_count": 200,
|
| 96 |
+
"verified_record_count": 200,
|
| 97 |
+
"verified_file_hashes": [
|
| 98 |
+
{
|
| 99 |
+
"path": "data/tau2/results/final/claude-3-7-sonnet-20250219_airline_default_gpt-4.1-2025-04-14_4trials.json",
|
| 100 |
+
"sha256": "40a2c6a246eab27db5cdefda895fbe7f44be78a23d300817248534aa606d66de"
|
| 101 |
+
}
|
| 102 |
+
],
|
| 103 |
+
"reported_statistics": null,
|
| 104 |
+
"schema_summary": "Results object with tasks and simulations; each simulation has task_id, messages, reward_info, trial and seed.",
|
| 105 |
+
"full_trajectory": "yes for audited text simulations",
|
| 106 |
+
"provenance": "Model agent plus model user simulator",
|
| 107 |
+
"verification": "Environment/action/communication reward evidence",
|
| 108 |
+
"eligibility": "evaluation-only",
|
| 109 |
+
"license": "MIT code; benchmark evaluation constraints apply",
|
| 110 |
+
"task_id": "domain plus task_id",
|
| 111 |
+
"duplicate_risks": ["multiple trials", "AgentSuite tau2 overlap"],
|
| 112 |
+
"limitations": ["Only one checked-in result file counted", "reference actions are not demonstrations"]
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"id": "mind2web",
|
| 116 |
+
"name": "Mind2Web",
|
| 117 |
+
"url": "https://huggingface.co/datasets/osunlp/Mind2Web",
|
| 118 |
+
"pinned_revision": "17ece8eb89862368edc0cc806acee6fca5163474",
|
| 119 |
+
"available_splits": ["train"],
|
| 120 |
+
"upstream_record_count": 1009,
|
| 121 |
+
"qualifying_trajectory_count": 1009,
|
| 122 |
+
"verified_record_count": 1009,
|
| 123 |
+
"reported_statistics": "All 1,009 rows exposed by the dataset-server Parquet conversion imported with every human action and pre-action DOM observation.",
|
| 124 |
+
"schema_summary": "Human action sequence with confirmed_task and, for every action, operation, action representation, target candidates and pre-action cleaned/raw DOM.",
|
| 125 |
+
"full_trajectory": "yes for audited action/context sequences",
|
| 126 |
+
"provenance": "Human web demonstrations",
|
| 127 |
+
"verification": "Per-action DOM context and human annotation are present; no explicit binary execution-success field",
|
| 128 |
+
"eligibility": "conditionally-usable",
|
| 129 |
+
"license": "CC-BY-4.0",
|
| 130 |
+
"task_id": "Native annotation/task ID required",
|
| 131 |
+
"duplicate_risks": ["same website/task across splits"],
|
| 132 |
+
"limitations": ["No explicit binary success indicator", "DOM snapshots make full records large"]
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"id": "toolbench",
|
| 136 |
+
"name": "ToolBench",
|
| 137 |
+
"url": "https://github.com/OpenBMB/ToolBench",
|
| 138 |
+
"pinned_revision": "d56fdd89faf8c91fa135090b212bb9057ee5cfc2",
|
| 139 |
+
"available_splits": ["G1", "G2", "G3 examples"],
|
| 140 |
+
"upstream_record_count": 15,
|
| 141 |
+
"qualifying_trajectory_count": 11,
|
| 142 |
+
"rejected_before_library": 4,
|
| 143 |
+
"verified_record_count": 15,
|
| 144 |
+
"verified_file_hashes": [
|
| 145 |
+
{
|
| 146 |
+
"path": "data_example/answer/G1_answer/10_ChatGPT_DFS_woFilter_w2.json",
|
| 147 |
+
"sha256": "76a77721640045e07142c669008611e6b82b25007c73427aec771adbf9dd9841"
|
| 148 |
+
}
|
| 149 |
+
],
|
| 150 |
+
"reported_statistics": "Full dataset statistics are upstream-reported and not verified here.",
|
| 151 |
+
"schema_summary": "answer_generation with function definitions and cumulative train_messages; final list entry is the complete function-call conversation.",
|
| 152 |
+
"full_trajectory": "yes in checked-in examples",
|
| 153 |
+
"provenance": "ChatGPT-generated search/tool paths",
|
| 154 |
+
"verification": "valid_data/win generation status; no independent semantic proof",
|
| 155 |
+
"eligibility": "conditionally-usable",
|
| 156 |
+
"license": "Repository states Apache-2.0 and research/educational intent; third-party API content needs review",
|
| 157 |
+
"task_id": "query_id or stable hash of query",
|
| 158 |
+
"duplicate_risks": ["same query across answer methods", "cumulative message prefixes"],
|
| 159 |
+
"limitations": ["13 of 15 examples adapted; two lacked a complete cumulative history", "11 passed the strict sequential context gate", "win/valid_data are not independent semantic success proof"]
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"id": "alfworld",
|
| 163 |
+
"name": "ALFWorld",
|
| 164 |
+
"url": "https://github.com/alfworld/alfworld",
|
| 165 |
+
"pinned_revision": "release-0.2.2-sha256-25171f16e20ad7b048c47275c45b0babf3aa1cbab29cec97387922350a9844bc",
|
| 166 |
+
"related_code_revision": "aaba6870f86c5be6a08a491f32a50b906227bc3e",
|
| 167 |
+
"available_splits": ["train (6374)", "valid_seen (251)", "valid_train (200)", "valid_unseen (255)"],
|
| 168 |
+
"upstream_record_count": 7080,
|
| 169 |
+
"qualifying_trajectory_count": 0,
|
| 170 |
+
"rejected_before_library": 7080,
|
| 171 |
+
"verified_record_count": 7080,
|
| 172 |
+
"verified_file_hashes": [
|
| 173 |
+
{
|
| 174 |
+
"path": "json_2.1.1_json.zip",
|
| 175 |
+
"sha256": "25171f16e20ad7b048c47275c45b0babf3aa1cbab29cec97387922350a9844bc"
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"path": "json_2.1.1/valid_seen/pick_cool_then_place_in_recep-AppleSliced-None-GarbageCan-25/trial_T20190908_193525_796595/traj_data.json",
|
| 179 |
+
"sha256": "1c8c8dc3b72bea45bcd3e6714705c257f94f5cc7636094f74a4816ae90a0856a"
|
| 180 |
+
}
|
| 181 |
+
],
|
| 182 |
+
"reported_statistics": "README reports 3,553 train, 140 seen-validation and 134 unseen-validation games; the verified release archive contains different counts and includes valid_train.",
|
| 183 |
+
"schema_summary": "Verified traj_data.json with task/scene, three Turk annotations, planner high_pddl, low actions, and image references.",
|
| 184 |
+
"full_trajectory": "planner action trajectory; not a chat transcript",
|
| 185 |
+
"provenance": "PDDL planner actions plus human language annotations",
|
| 186 |
+
"verification": "Official release ZIP integrity passed; one record mapped. Planner actions were not replayed in an environment.",
|
| 187 |
+
"eligibility": "conditionally-usable",
|
| 188 |
+
"license": "MIT repository; downloaded data terms still require confirmation",
|
| 189 |
+
"task_id": "task_id/trial ID",
|
| 190 |
+
"duplicate_risks": ["language annotations over the same plan"],
|
| 191 |
+
"limitations": ["Only one of 7,080 records received adapter smoke validation", "tool observations are not present as chat events", "image payload availability was not validated"]
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"id": "swebench_verified",
|
| 195 |
+
"name": "SWE-bench Verified",
|
| 196 |
+
"url": "https://huggingface.co/datasets/princeton-nlp/SWE-bench_Verified",
|
| 197 |
+
"pinned_revision": null,
|
| 198 |
+
"related_code_revision": "02e7a74ffd0b707aab73d203fe87bdc7c76afc8e",
|
| 199 |
+
"available_splits": [],
|
| 200 |
+
"verified_record_count": null,
|
| 201 |
+
"reported_statistics": null,
|
| 202 |
+
"schema_summary": "Benchmark instances, not expert trajectories.",
|
| 203 |
+
"full_trajectory": "no",
|
| 204 |
+
"provenance": "Curated software issue benchmark",
|
| 205 |
+
"verification": "Benchmark tests",
|
| 206 |
+
"eligibility": "evaluation-only",
|
| 207 |
+
"license": "Per-instance repository licenses require review",
|
| 208 |
+
"task_id": "instance_id",
|
| 209 |
+
"duplicate_risks": ["overlap with SWE agent trajectory sources"],
|
| 210 |
+
"limitations": ["Dataset host blocked", "gold patches must never become model-visible context"]
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"id": "swebench_multilingual",
|
| 214 |
+
"name": "SWE-bench Multilingual",
|
| 215 |
+
"url": "https://huggingface.co/datasets/SWE-bench/SWE-bench_Multilingual",
|
| 216 |
+
"pinned_revision": null,
|
| 217 |
+
"available_splits": [],
|
| 218 |
+
"verified_record_count": null,
|
| 219 |
+
"reported_statistics": null,
|
| 220 |
+
"schema_summary": "Benchmark instances, not expert trajectories; unverified locally.",
|
| 221 |
+
"full_trajectory": "no",
|
| 222 |
+
"provenance": "Software issue benchmark",
|
| 223 |
+
"verification": "Benchmark tests",
|
| 224 |
+
"eligibility": "evaluation-only",
|
| 225 |
+
"license": "Per-instance repository licenses require review",
|
| 226 |
+
"task_id": "instance_id",
|
| 227 |
+
"duplicate_risks": ["cross-dataset repository/issue overlap"],
|
| 228 |
+
"limitations": ["Dataset host blocked", "reference patches must remain hidden"]
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"id": "terminal_bench",
|
| 232 |
+
"name": "Terminal-Bench",
|
| 233 |
+
"url": "https://github.com/laude-institute/terminal-bench",
|
| 234 |
+
"pinned_revision": "d28711d0da2675d0bb1d56de45ae5df6082438a3",
|
| 235 |
+
"available_splits": ["repository task registry"],
|
| 236 |
+
"verified_record_count": 241,
|
| 237 |
+
"reported_statistics": null,
|
| 238 |
+
"schema_summary": "Task manifests, tests and reference scripts; not complete conversational demonstrations.",
|
| 239 |
+
"full_trajectory": "no in audited repository",
|
| 240 |
+
"provenance": "Protected benchmark tasks",
|
| 241 |
+
"verification": "Container tests",
|
| 242 |
+
"eligibility": "evaluation-only",
|
| 243 |
+
"license": "Apache-2.0 repository plus explicit anti-training instructions",
|
| 244 |
+
"task_id": "task directory/registry ID",
|
| 245 |
+
"duplicate_risks": ["TerminalTraj task overlap"],
|
| 246 |
+
"limitations": ["966 files carried anti-training canaries", "reference scripts are not trajectories"]
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"id": "deepswe_datacurve",
|
| 250 |
+
"name": "DeepSWE (Datacurve)",
|
| 251 |
+
"url": "https://www.datacurve.ai/",
|
| 252 |
+
"pinned_revision": null,
|
| 253 |
+
"available_splits": [],
|
| 254 |
+
"verified_record_count": null,
|
| 255 |
+
"reported_statistics": null,
|
| 256 |
+
"schema_summary": "Exact requested artifact was not resolvable from available evidence.",
|
| 257 |
+
"full_trajectory": "unverified",
|
| 258 |
+
"provenance": "unverified",
|
| 259 |
+
"verification": "unverified",
|
| 260 |
+
"eligibility": "evaluation-only",
|
| 261 |
+
"license": "unverified",
|
| 262 |
+
"task_id": "unverified",
|
| 263 |
+
"duplicate_risks": ["SWE benchmark overlap"],
|
| 264 |
+
"limitations": ["Treat as protected evaluation data pending an exact source URL and anti-training review"]
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"id": "browsecomp",
|
| 268 |
+
"name": "BrowseComp",
|
| 269 |
+
"url": "https://github.com/openai/simple-evals",
|
| 270 |
+
"pinned_revision": "652c89d0ca9df547706735883097e9537d40dc47",
|
| 271 |
+
"available_splits": ["evaluation implementation"],
|
| 272 |
+
"verified_record_count": null,
|
| 273 |
+
"reported_statistics": null,
|
| 274 |
+
"schema_summary": "Question/answer evaluation benchmark, not trajectories.",
|
| 275 |
+
"full_trajectory": "no",
|
| 276 |
+
"provenance": "Evaluation benchmark",
|
| 277 |
+
"verification": "Answer grading",
|
| 278 |
+
"eligibility": "evaluation-only",
|
| 279 |
+
"license": "Repository license; dataset terms require separate review",
|
| 280 |
+
"task_id": "benchmark example ID",
|
| 281 |
+
"duplicate_risks": ["web question leakage"],
|
| 282 |
+
"limitations": ["No rollout data inspected"]
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"id": "bfcl",
|
| 286 |
+
"name": "Berkeley Function Calling Leaderboard",
|
| 287 |
+
"url": "https://github.com/ShishirPatil/gorilla",
|
| 288 |
+
"pinned_revision": "6ea57973c7a6097fd7c5915698c54c17c5b1b6c8",
|
| 289 |
+
"available_splits": ["evaluation implementation"],
|
| 290 |
+
"verified_record_count": null,
|
| 291 |
+
"reported_statistics": null,
|
| 292 |
+
"schema_summary": "Function-calling evaluation cases and expected outputs, not full expert trajectories.",
|
| 293 |
+
"full_trajectory": "no",
|
| 294 |
+
"provenance": "Evaluation benchmark",
|
| 295 |
+
"verification": "Leaderboard validators",
|
| 296 |
+
"eligibility": "evaluation-only",
|
| 297 |
+
"license": "Repository/dataset terms require split-level review",
|
| 298 |
+
"task_id": "benchmark case ID",
|
| 299 |
+
"duplicate_risks": ["ToolBench/function-call overlap"],
|
| 300 |
+
"limitations": ["Expected calls must not be exported as training targets"]
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"id": "artificial_analysis_index",
|
| 304 |
+
"name": "Artificial Analysis Intelligence Index",
|
| 305 |
+
"url": "https://artificialanalysis.ai/",
|
| 306 |
+
"pinned_revision": null,
|
| 307 |
+
"available_splits": [],
|
| 308 |
+
"verified_record_count": null,
|
| 309 |
+
"reported_statistics": null,
|
| 310 |
+
"schema_summary": "Public leaderboard/index, not a downloadable trajectory dataset.",
|
| 311 |
+
"full_trajectory": "no",
|
| 312 |
+
"provenance": "Aggregated evaluation index",
|
| 313 |
+
"verification": "Provider methodology not inspected in this environment",
|
| 314 |
+
"eligibility": "unavailable",
|
| 315 |
+
"license": "No training-data license verified",
|
| 316 |
+
"task_id": "not applicable",
|
| 317 |
+
"duplicate_risks": [],
|
| 318 |
+
"limitations": ["No qualifying public trajectory records identified"]
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"id": "swegym_sampled",
|
| 322 |
+
"name": "SWE-Gym OpenHands Sampled Trajectories",
|
| 323 |
+
"url": "https://huggingface.co/datasets/SWE-Gym/OpenHands-Sampled-Trajectories",
|
| 324 |
+
"pinned_revision": "baf3a4e4bff514d48ddc08a93a2ade5c126212c7",
|
| 325 |
+
"available_splits": ["train.raw"],
|
| 326 |
+
"upstream_record_count": 6055,
|
| 327 |
+
"qualifying_trajectory_count": 6055,
|
| 328 |
+
"verified_record_count": 6055,
|
| 329 |
+
"reported_statistics": "6,055/6,055 rows imported, including every successful and failed sampled attempt; 553 exact interaction/outcome duplicates collapse into retained canonical provenance groups.",
|
| 330 |
+
"schema_summary": "Full OpenHands messages, structured tools, run ID, resolved verdict and execution report.",
|
| 331 |
+
"full_trajectory": "yes",
|
| 332 |
+
"provenance": "OpenHands model rollouts",
|
| 333 |
+
"verification": "Source execution test result; not independently replayed here",
|
| 334 |
+
"eligibility": "conditionally-usable",
|
| 335 |
+
"license": "Dataset card does not declare a license",
|
| 336 |
+
"task_id": "instance_id",
|
| 337 |
+
"duplicate_risks": ["multiple run_id attempts per instance", "SWE-Gym SFT overlap"],
|
| 338 |
+
"limitations": ["Source evaluator was not independently reproduced"]
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"id": "swebench_verified_trajectories",
|
| 342 |
+
"name": "SWE-bench Verified mini-swe-agent Trajectories",
|
| 343 |
+
"url": "https://huggingface.co/datasets/tarsur385/swebench-verified-trajectories",
|
| 344 |
+
"pinned_revision": "773748a7c1222e8a642a7059821498e14293562a",
|
| 345 |
+
"available_splits": ["10 models x 500 SWE-bench Verified tasks"],
|
| 346 |
+
"upstream_record_count": 5000,
|
| 347 |
+
"qualifying_trajectory_count": 5000,
|
| 348 |
+
"verified_record_count": 5000,
|
| 349 |
+
"reported_statistics": "All 5,000 native .traj.json files imported; 456 exact duplicates collapse into retained provenance groups and 4,532 retained interactions pass the strict gate.",
|
| 350 |
+
"schema_summary": "Native mini-swe-agent messages, tool calls/results, model/run IDs and resolved score.",
|
| 351 |
+
"full_trajectory": "yes",
|
| 352 |
+
"provenance": "Ten model families using mini-swe-agent",
|
| 353 |
+
"verification": "Published SWE-bench execution result; protected evaluation split",
|
| 354 |
+
"eligibility": "evaluation-only",
|
| 355 |
+
"license": "MIT dataset card; underlying repositories retain their own licenses",
|
| 356 |
+
"task_id": "instance_id",
|
| 357 |
+
"duplicate_risks": ["exactly ten attempts per benchmark task", "cross-source SWE-bench overlap"],
|
| 358 |
+
"limitations": ["Native tests were not independently rerun in this environment"]
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"id": "swebench_multilingual_trajectories",
|
| 362 |
+
"name": "SWE-bench Multilingual Trajectories",
|
| 363 |
+
"url": "https://huggingface.co/datasets/AlienKevin/SWE-bench-Multilingual-trajectories",
|
| 364 |
+
"pinned_revision": "7221c09c6cc0ed16d51c9cb9886a1ab484465670",
|
| 365 |
+
"available_splits": ["Multi_SWE_agent_32B_Rust", "Qwen2.5_Coder_32B_Instruct", "SWE_agent_LM_32B"],
|
| 366 |
+
"upstream_record_count": 129,
|
| 367 |
+
"qualifying_trajectory_count": 110,
|
| 368 |
+
"verified_record_count": 129,
|
| 369 |
+
"reported_statistics": "All 129 trajectories (43 tasks x three models) imported.",
|
| 370 |
+
"schema_summary": "Action, response/thought, terminal observation and state for every step.",
|
| 371 |
+
"full_trajectory": "yes",
|
| 372 |
+
"provenance": "Three SWE-agent-compatible model releases",
|
| 373 |
+
"verification": "Published resolved/completed/errored fields",
|
| 374 |
+
"eligibility": "evaluation-only",
|
| 375 |
+
"license": "MIT",
|
| 376 |
+
"task_id": "id",
|
| 377 |
+
"duplicate_risks": ["three model attempts for each of 43 Rust tasks"],
|
| 378 |
+
"limitations": ["Source verdicts were not independently replayed"]
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"id": "terminalbench_trajectories",
|
| 382 |
+
"name": "Terminal-Bench 2.0 Recorded Trajectories",
|
| 383 |
+
"url": "https://huggingface.co/datasets/yoonholee/terminalbench-trajectories",
|
| 384 |
+
"pinned_revision": "04e8940f5b6736a7ce8d22224fe2f2af74163ed2",
|
| 385 |
+
"available_splits": ["train"],
|
| 386 |
+
"upstream_record_count": 52104,
|
| 387 |
+
"qualifying_trajectory_count": 24171,
|
| 388 |
+
"rejected_before_library": 17642,
|
| 389 |
+
"verified_record_count": 34462,
|
| 390 |
+
"reported_statistics": "52,104 trials enumerated; 34,462 have steps and were imported; 17,642 have null steps. Exact deduplication retains 34,263 interactions, of which 24,086 pass the strict gate.",
|
| 391 |
+
"schema_summary": "Full step sequence with agent message, tool bundle, returned observation, reward, model, scaffold, duration, tokens and cost.",
|
| 392 |
+
"full_trajectory": "yes for 34,462 rows; absent for 17,642 rows",
|
| 393 |
+
"provenance": "26 agent scaffolds and 49 underlying models",
|
| 394 |
+
"verification": "Source-reported Terminal-Bench reward",
|
| 395 |
+
"eligibility": "evaluation-only",
|
| 396 |
+
"license": "Apache-2.0 dataset card",
|
| 397 |
+
"task_id": "task_name with unique trial_id per attempt",
|
| 398 |
+
"duplicate_risks": ["typically five trials per task/agent combination"],
|
| 399 |
+
"limitations": ["Observations are source-truncated to 5,000 characters", "17,642 leaderboard trials contain no trajectory steps"]
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"id": "agentsuite_bfcl",
|
| 403 |
+
"name": "AgentSuite BFCL Outputs",
|
| 404 |
+
"url": "https://huggingface.co/datasets/AgentSuite/BFCL-trajectories",
|
| 405 |
+
"pinned_revision": "a86612c6bb68e6566e5e7d206f7b031af889edc0",
|
| 406 |
+
"available_splits": ["train"],
|
| 407 |
+
"upstream_record_count": 115466,
|
| 408 |
+
"qualifying_trajectory_count": 0,
|
| 409 |
+
"diagnostic_records_examined": 115466,
|
| 410 |
+
"rejected_before_library": 115466,
|
| 411 |
+
"verified_record_count": 115466,
|
| 412 |
+
"reported_statistics": "All 115,466 records imported for diagnostics; zero pass the full-trajectory action-observation gate.",
|
| 413 |
+
"schema_summary": "User prompt and one or more model function-call answers, evaluator score and latency; no returned tool observations.",
|
| 414 |
+
"full_trajectory": "no; final function-call outputs only",
|
| 415 |
+
"provenance": "26 model outputs through AgentSuite",
|
| 416 |
+
"verification": "BFCL evaluator score",
|
| 417 |
+
"eligibility": "evaluation-only",
|
| 418 |
+
"license": "No license declared on the dataset card",
|
| 419 |
+
"task_id": "meta.id",
|
| 420 |
+
"duplicate_risks": ["one attempt per model/task"],
|
| 421 |
+
"limitations": ["No causal tool results; retained only as tier-5 diagnostics, never as trajectory SFT"]
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"id": "browsecomp_rollouts",
|
| 425 |
+
"name": "BrowseComp GPT-OSS-120B Rollouts",
|
| 426 |
+
"url": "https://huggingface.co/datasets/rl-rag/browsecomp-gpt-oss-120b-260222",
|
| 427 |
+
"pinned_revision": "ac4854be0eb527cdf90f6c7d09446eeb7a3d1dbd",
|
| 428 |
+
"available_splits": ["train"],
|
| 429 |
+
"upstream_record_count": 5064,
|
| 430 |
+
"qualifying_trajectory_count": 5063,
|
| 431 |
+
"rejected_before_library": 1,
|
| 432 |
+
"verified_record_count": 5064,
|
| 433 |
+
"reported_statistics": "5,063 full conversations imported; one malformed row lacks a conversation.",
|
| 434 |
+
"schema_summary": "Four research-agent attempts per question with tool calls, full tool responses, answer judge verdict and latency.",
|
| 435 |
+
"full_trajectory": "yes for 5,063/5,064 rows",
|
| 436 |
+
"provenance": "gpt-oss-120b research-agent rollouts",
|
| 437 |
+
"verification": "Source-reported GPT-4o answer judge; not independently reproduced",
|
| 438 |
+
"eligibility": "evaluation-only",
|
| 439 |
+
"license": "No license declared on the dataset card",
|
| 440 |
+
"task_id": "qid",
|
| 441 |
+
"duplicate_risks": ["exactly four attempts per question"],
|
| 442 |
+
"limitations": ["One malformed record", "Reference answers remain raw provenance and are excluded from model_input"]
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"id": "swesmith_trajectories",
|
| 446 |
+
"name": "SWE-smith Trajectories",
|
| 447 |
+
"url": "https://huggingface.co/datasets/SWE-bench/SWE-smith-trajectories",
|
| 448 |
+
"pinned_revision": "08e109b4a59eaeebf80e4675cd125d42e7ac99a4",
|
| 449 |
+
"available_splits": ["tool", "xml", "ticks"],
|
| 450 |
+
"upstream_record_count": 24100,
|
| 451 |
+
"qualifying_trajectory_count": 24100,
|
| 452 |
+
"verified_record_count": 24100,
|
| 453 |
+
"reported_statistics": "The tool-format population has 24,100 rows; XML/ticks are alternate serializations, not independent rollouts.",
|
| 454 |
+
"schema_summary": "Full SWE-agent messages, task/trajectory IDs, resolved outcome and patch.",
|
| 455 |
+
"full_trajectory": "yes",
|
| 456 |
+
"provenance": "SWE-agent with Claude-generated trajectories",
|
| 457 |
+
"verification": "Published resolved field",
|
| 458 |
+
"eligibility": "conditionally-usable",
|
| 459 |
+
"license": "MIT",
|
| 460 |
+
"task_id": "instance_id",
|
| 461 |
+
"duplicate_risks": ["tool/xml/ticks serialization variants", "SWE-bench task overlap"],
|
| 462 |
+
"limitations": ["Tool definitions are not published; names and arguments remain in messages"]
|
| 463 |
+
}
|
| 464 |
+
]
|
| 465 |
+
}
|
staging_report.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"hosted_shards": 50,
|
| 3 |
+
"hosted_sources": [
|
| 4 |
+
"mind2web",
|
| 5 |
+
"nvidia_open_swe_traces",
|
| 6 |
+
"official_tau3",
|
| 7 |
+
"swebench_multilingual_trajectories",
|
| 8 |
+
"swebench_verified_trajectories",
|
| 9 |
+
"swegym_openhands",
|
| 10 |
+
"swesmith_trajectories",
|
| 11 |
+
"terminalbench_trajectories",
|
| 12 |
+
"terminaltraj",
|
| 13 |
+
"toolbench"
|
| 14 |
+
],
|
| 15 |
+
"hosted_trajectories": 136373,
|
| 16 |
+
"inventory_shards": 56,
|
| 17 |
+
"inventory_trajectories": 155278,
|
| 18 |
+
"metadata_only_sources": [
|
| 19 |
+
"agentsuite_tau2",
|
| 20 |
+
"browsecomp_rollouts",
|
| 21 |
+
"swegym_sampled"
|
| 22 |
+
],
|
| 23 |
+
"metadata_only_trajectories": 18905,
|
| 24 |
+
"schema_version": 1,
|
| 25 |
+
"scores_recomputed": false
|
| 26 |
+
}
|
structurally_eligible.index.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7abe69bdb9cd3d07b48a282c0bf371f7bcff26776ad3caba307e22b45faf40f8
|
| 3 |
+
size 4326891
|
successful.index.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30b4ec2dd4faacb804e78a2826df649af7737807f542ddc6f667e314ae08c3e1
|
| 3 |
+
size 1561257
|
task_splits.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f932f6d111636fa2306d1b2fef94a6a5b98e9fdab426c95f0da78fced367f776
|
| 3 |
+
size 582841
|
training_eligible.index.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd0619062d95f44824710e32922c7fc330c44c367f41c2adcd3fb2002308a00f
|
| 3 |
+
size 3426523
|
unknown.index.jsonl.gz
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a11f7e31b6bf29ee513709a4118ad99823a8634b6f0b3f5ecd8cf605b58cc45
|
| 3 |
+
size 663458
|