AdithyaSK HF Staff commited on
Commit
07eac47
·
1 Parent(s): d899c70

SmolDataEnvs: rename, new README, banner

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. README.md +70 -25
  2. banner.png +3 -0
  3. registry.json +2 -2
  4. tasks/0000_886_886039_qa_5/task.toml +3 -2
  5. tasks/0000_901_901543_qa_2/task.toml +3 -2
  6. tasks/0000_901_901543_qa_3/task.toml +3 -2
  7. tasks/0000_903_903504_qa_1/task.toml +3 -2
  8. tasks/0000_939_939639_qa_2/task.toml +3 -2
  9. tasks/0000_939_939639_qa_3/task.toml +3 -2
  10. tasks/0000_949_949403_qa_3/task.toml +3 -2
  11. tasks/0000_975_975906_qa_1/task.toml +3 -2
  12. tasks/0000_986_986420_qa_2/task.toml +3 -2
  13. tasks/0000_986_986420_qa_4/task.toml +3 -2
  14. tasks/0000_988_988342_qa_4/task.toml +3 -2
  15. tasks/0000_988_988657_qa_4/task.toml +3 -2
  16. tasks/0000_991_991010_qa_1/task.toml +3 -2
  17. tasks/0000_991_991010_qa_2/task.toml +3 -2
  18. tasks/0000_991_991010_qa_3/task.toml +3 -2
  19. tasks/0000_994_994384_qa_5/task.toml +3 -2
  20. tasks/0000_994_994443_qa_4/task.toml +3 -2
  21. tasks/0001_004_1004716_qa_2/task.toml +3 -2
  22. tasks/0001_024_1024990_qa_2/task.toml +3 -2
  23. tasks/0001_027_1027372_qa_2/task.toml +3 -2
  24. tasks/0001_029_1029150_qa_2/task.toml +3 -2
  25. tasks/0001_036_1036744_qa_1/task.toml +3 -2
  26. tasks/0001_042_1042725_qa_2/task.toml +3 -2
  27. tasks/0001_044_1044175_qa_3/task.toml +3 -2
  28. tasks/0001_062_1062946_qa_2/task.toml +3 -2
  29. tasks/0001_065_1065379_qa_3/task.toml +3 -2
  30. tasks/0001_067_1067596_qa_5/task.toml +3 -2
  31. tasks/0001_068_1068916_qa_1/task.toml +3 -2
  32. tasks/0001_068_1068916_qa_5/task.toml +3 -2
  33. tasks/0001_078_1078204_qa_2/task.toml +3 -2
  34. tasks/0001_079_1079149_qa_2/task.toml +3 -2
  35. tasks/0001_082_1082305_qa_1/task.toml +3 -2
  36. tasks/0001_083_1083006_qa_5/task.toml +3 -2
  37. tasks/0001_091_1091047_qa_4/task.toml +3 -2
  38. tasks/0001_091_1091047_qa_5/task.toml +3 -2
  39. tasks/0001_095_1095486_qa_2/task.toml +3 -2
  40. tasks/0001_102_1102505_qa_2/task.toml +3 -2
  41. tasks/0001_104_1104901_qa_2/task.toml +3 -2
  42. tasks/0001_104_1104901_qa_3/task.toml +3 -2
  43. tasks/0001_104_1104959_qa_2/task.toml +3 -2
  44. tasks/0001_105_1105529_qa_5/task.toml +3 -2
  45. tasks/0001_105_1105892_qa_3/task.toml +3 -2
  46. tasks/0001_105_1105892_qa_5/task.toml +3 -2
  47. tasks/0001_116_1116342_qa_3/task.toml +3 -2
  48. tasks/0001_116_1116342_qa_5/task.toml +3 -2
  49. tasks/0001_123_1123075_qa_4/task.toml +3 -2
  50. tasks/0001_124_1124736_qa_1/task.toml +3 -2
README.md CHANGED
@@ -3,49 +3,94 @@ license: mit
3
  task_categories:
4
  - other
5
  tags:
 
6
  - rl-environment
7
  - agent
8
  - data-analysis
 
9
  - code-agent
10
  - harbor
11
  - openenv
12
  ---
13
 
14
- [![View tasks in Harbor Visualiser](https://img.shields.io/badge/%F0%9F%A4%97%20Harbor%20Visualiser-View%20tasks-FFD21F?style=for-the-badge)](https://huggingface.co/spaces/HuggingFaceH4/harbor-visualiser?dataset=FineEnvs/data-agent-harbor-eval)
15
- # 🧪 Data Agent — Harbor (eval)
16
 
17
- A small, **difficulty-balanced validation split** — **144 tasks** — perfect for quick checkpoints
18
- while you train. Same idea as the rest of the family: your agent gets a real dataset and a
19
- question, explores and answers, and everything is graded **deterministically, no LLM judge**.
20
 
21
- Packaged in [**Harbor**](https://github.com/huggingface/OpenEnv) format.
22
 
23
- ## Where it comes from
24
- Built from the [**jupyter-agent dataset**](https://huggingface.co/datasets/jupyter-agent/jupyter-agent-dataset)
25
- (real notebooks over Kaggle datasets). Every task was **verified** — a strong agent must reproduce
26
- the gold answer in a sandbox under deterministic grading — so **each task is known-solvable and
27
- unambiguously gradable.** Held out from training.
 
 
 
 
 
 
 
 
 
28
 
29
  ## What's inside
30
- - **144 verified tasks**
31
- - **Difficulty** — easy **16** · medium **74** · hard **54** (`difficulty_tier`; `difficulty_level` 1–4)
32
- - **Answer types** — numeric 83 · short-label 56 · yes/no 5
33
 
34
  ## How a task is laid out
 
 
 
 
 
 
 
 
 
35
  ```
36
- tasks/<task_id>/ → task.toml · instruction.md · environment/ · tests/
37
- registry.json · manifest.parquet
 
 
 
 
 
 
38
  ```
39
- Input files land in `/home/user/input/` at task start.
40
 
41
- ## How grading works
42
- Answer goes to `/workdir/answer.txt`; `grader.py` scores it deterministically —
43
- **exact → numeric tolerance → list/percent normalization → symbolic (math-verify)** — as `1.0` or
44
- `0.0`. No model, no network.
45
 
46
- ## Run it
47
  ```bash
48
- openenv harbor info --dataset HuggingEnvs/data-agent-harbor-eval
49
- openenv harbor run --dataset HuggingEnvs/data-agent-harbor-eval --model <your-model>
 
 
50
  ```
51
- Any tool-calling model works; grading is model-agnostic and offline.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  task_categories:
4
  - other
5
  tags:
6
+ - smoldataenvs
7
  - rl-environment
8
  - agent
9
  - data-analysis
10
+ - reinforcement-learning
11
  - code-agent
12
  - harbor
13
  - openenv
14
  ---
15
 
16
+ <div align="center">
 
17
 
18
+ <img src="https://huggingface.co/datasets/FineEnvs/SmolDataEnvs-harbor-eval/resolve/main/banner.png" alt="SmolDataEnvs" width="100%">
 
 
19
 
20
+ # 📊 SmolDataEnvs — Harbor (eval)
21
 
22
+ **5.5K+ RL tasks for hill-climbing small models in code and data science.**
23
+
24
+ [![Collection](https://img.shields.io/badge/%F0%9F%A4%97%20Collection-SmolDataEnvs-FFD21E?style=for-the-badge&labelColor=1a1a1a)](https://huggingface.co/collections/FineEnvs/smoldataenvs)
25
+ [![Harbor Visualiser](https://img.shields.io/badge/%F0%9F%A4%97%20Harbor%20Visualiser-Browse%20tasks-FFD21E?style=for-the-badge&labelColor=1a1a1a)](https://huggingface.co/spaces/HuggingFaceH4/harbor-visualiser?dataset=FineEnvs/SmolDataEnvs-harbor-eval)
26
+
27
+ </div>
28
+
29
+
30
+ The validation suite: **144 tasks**, small enough to run every few hundred training steps without
31
+ the eval becoming the expensive part of the loop.
32
+
33
+
34
+ Packaged in [Harbor](https://github.com/huggingface/OpenEnv) format, so each task is a ready-made
35
+ agentic environment: its own container, its own data, its own verifier.
36
 
37
  ## What's inside
38
+
39
+ - **144 verified tasks** — quick validation during a run
40
+ - **Difficulty** — easy **16** · medium **74** · hard **54** (`difficulty_tier`, plus `difficulty_level` 1–5)
41
 
42
  ## How a task is laid out
43
+
44
+ ```
45
+ tasks/<task_id>/
46
+ task.toml # metadata, the question, the gold answer, grading tolerances
47
+ instruction.md # the prompt the agent sees
48
+ environment/ # Dockerfile (shared base image) + the data-pull hook
49
+ tests/ # grader.py (deterministic) + test.sh
50
+ registry.json # the suite manifest
51
+ manifest.parquet # one row per task, for filtering without walking the tree
52
  ```
53
+
54
+ ## Serve it
55
+
56
+ ```bash
57
+ openenv harbor serve \
58
+ --dataset FineEnvs/SmolDataEnvs-harbor-eval \
59
+ --llm-url http://127.0.0.1:8000/v1 --model <your-model> \
60
+ --port 8000 --capture-port 8100
61
  ```
 
62
 
63
+ Pass several with `--dataset a,b` and each arrives as its own split, which is how you train against
64
+ `-train` and validate against `-eval` from one server.
65
+
66
+ ## One rollout, no trainer
67
 
 
68
  ```bash
69
+ openenv harbor rollout \
70
+ --dataset FineEnvs/SmolDataEnvs-harbor-eval \
71
+ --llm-url http://127.0.0.1:8000/v1 --model <your-model> \
72
+ --harness opencode --sandbox e2b --task-index 0
73
  ```
74
+
75
+ ## Where it comes from
76
+
77
+ Built from the [jupyter-agent dataset](https://huggingface.co/datasets/jupyter-agent/jupyter-agent-dataset)
78
+ — real data-science notebooks over 471 Kaggle datasets. Every question–answer pair was extracted and
79
+ then **verified**: strong agent models had to solve the task in a live sandbox and reproduce the gold
80
+ answer under deterministic grading. Anything ambiguous or un-checkable was dropped. So every task
81
+ here is known-solvable and unambiguously gradable.
82
+
83
+ **Verified by a checker, not judged by a model.** Grading is an exact comparison against a known
84
+ answer, through a ladder of checks: exact match → numeric with tolerances → list and percent
85
+ normalisation → symbolic equivalence. No LLM sits in the reward path, so the signal does not drift
86
+ when you change the grader's model, because there isn't one.
87
+
88
+ ## The family
89
+
90
+ | Repo | What it is |
91
+ |---|---|
92
+ | [`SmolDataEnvs`](https://huggingface.co/datasets/FineEnvs/SmolDataEnvs) | the tasks as plain rows — load it and prompt any model |
93
+ | [`SmolDataEnvs-sft`](https://huggingface.co/datasets/FineEnvs/SmolDataEnvs-sft) | 4,677 verified agent trajectories, TRL-ready |
94
+ | [`SmolDataEnvs-harbor-train`](https://huggingface.co/datasets/FineEnvs/SmolDataEnvs-harbor-train) | 5,000 tasks as Harbor environments |
95
+ | [`SmolDataEnvs-harbor-test`](https://huggingface.co/datasets/FineEnvs/SmolDataEnvs-harbor-test) | 250 held-out, deliberately harder |
96
+ | [`SmolDataEnvs-harbor-eval`](https://huggingface.co/datasets/FineEnvs/SmolDataEnvs-harbor-eval) | 144 for quick validation during a run |
banner.png ADDED

Git LFS Details

  • SHA256: b5a8a6613c85b436d5624937cbbd61d6cf827df7cb0db04905cc1d9c7c5d8081
  • Pointer size: 132 Bytes
  • Size of remote file: 1.99 MB
registry.json CHANGED
@@ -1,6 +1,6 @@
1
  [
2
  {
3
- "name": "data-agent-harbor-eval",
4
  "version": "2.0",
5
  "description": "144 deterministic data-analysis tasks (no LLM judge).",
6
  "tasks": [
@@ -582,4 +582,4 @@
582
  }
583
  ]
584
  }
585
- ]
 
1
  [
2
  {
3
+ "name": "smoldataenvs-harbor-eval",
4
  "version": "2.0",
5
  "description": "144 deterministic data-analysis tasks (no LLM judge).",
6
  "tasks": [
 
582
  }
583
  ]
584
  }
585
+ ]
tasks/0000_886_886039_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_886_886039_qa_5"
6
  description = "How many times did Balon/Euron Greyjoy defeat Robb Stark in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "4"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_886_886039_qa_5"
6
  description = "How many times did Balon/Euron Greyjoy defeat Robb Stark in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_901_901543_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0000_901_901543_qa_2"
6
  description = "What numerical value was assigned to the 'no_employees' category 'More than 1000' during preprocessing?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "6"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_901_901543_qa_2"
6
  description = "What numerical value was assigned to the 'no_employees' category 'More than 1000' during preprocessing?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_901_901543_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_901_901543_qa_3"
6
  description = "How many columns are present in the dataset after removing the 'Timestamp' column?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "26"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_901_901543_qa_3"
6
  description = "How many columns are present in the dataset after removing the 'Timestamp' column?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0000_903_903504_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-eval-v1/0000_903_903504_qa_1"
6
  description = "What is the correlation coefficient between Air Temperature and Sea Surface Temperature in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "0.93"
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_903_903504_qa_1"
6
  description = "What is the correlation coefficient between Air Temperature and Sea Surface Temperature in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_939_939639_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_939_939639_qa_2"
6
  description = "What is the total number of rows in the dataset after using the pd.melt() function to transform the data for the swarmplot visualization?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "4800"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_939_939639_qa_2"
6
  description = "What is the total number of rows in the dataset after using the pd.melt() function to transform the data for the swarmplot visualization?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_939_939639_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0000_939_939639_qa_3"
6
  description = "What is the number of unique statistical categories (e.g., HP, Attack, Defense, etc.) included in the transformed dataset used for the swarmplot visualization?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "6"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_939_939639_qa_3"
6
  description = "What is the number of unique statistical categories (e.g., HP, Attack, Defense, etc.) included in the transformed dataset used for the swarmplot visualization?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_949_949403_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_949_949403_qa_3"
6
  description = "Which publisher is associated with the most games in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Electronic Arts"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_949_949403_qa_3"
6
  description = "Which publisher is associated with the most games in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0000_975_975906_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0000_975_975906_qa_1"
6
  description = "What is the most common weapon used in unsolved homicides in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Handgun"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_975_975906_qa_1"
6
  description = "What is the most common weapon used in unsolved homicides in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_986_986420_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0000_986_986420_qa_2"
6
  description = "What is the accuracy of a model that predicts all transactions as non-fraudulent in this dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "99.83"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_986_986420_qa_2"
6
  description = "What is the accuracy of a model that predicts all transactions as non-fraudulent in this dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_986_986420_qa_4/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0000_986_986420_qa_4"
6
  description = "What percentage of the transactions in the dataset are classified as non-fraudulent?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "99.83"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_986_986420_qa_4"
6
  description = "What percentage of the transactions in the dataset are classified as non-fraudulent?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0000_988_988342_qa_4/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-eval-v1/0000_988_988342_qa_4"
6
  description = "What is the most common cause of tsunamis in the dataset when excluding 'Unknown' and 'Questionable Earthquake' classifications?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Earthquake"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 0
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_988_988342_qa_4"
6
  description = "What is the most common cause of tsunamis in the dataset when excluding 'Unknown' and 'Questionable Earthquake' classifications?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 0
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_988_988657_qa_4/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_988_988657_qa_4"
6
  description = "How many individuals from India have been selected as Person of the Year in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "1"
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_988_988657_qa_4"
6
  description = "How many individuals from India have been selected as Person of the Year in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_991_991010_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0000_991_991010_qa_1"
6
  description = "Which game has the highest total cumulative playing time in hours across all users?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Dota 2"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 3
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_991_991010_qa_1"
6
  description = "Which game has the highest total cumulative playing time in hours across all users?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 3
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_991_991010_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_991_991010_qa_2"
6
  description = "What is the average number of games played per user in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "6.21"
15
  reward_mode_initial = "numeric"
16
  package_tier = 3
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_991_991010_qa_2"
6
  description = "What is the average number of games played per user in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 3
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_991_991010_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0000_991_991010_qa_3"
6
  description = "What percentage of all games played that exceed 1,000 hours are dedicated to Dota 2?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "49.1987"
15
  reward_mode_initial = "numeric"
16
  package_tier = 3
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_991_991010_qa_3"
6
  description = "What percentage of all games played that exceed 1,000 hours are dedicated to Dota 2?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 3
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_994_994384_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0000_994_994384_qa_5"
6
  description = "How many users have played at least one game in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "11350"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_994_994384_qa_5"
6
  description = "How many users have played at least one game in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0000_994_994443_qa_4/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0000_994_994443_qa_4"
6
  description = "How many spam messages in the dataset contain phone numbers with four or more digits?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "623"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0000_994_994443_qa_4"
6
  description = "How many spam messages in the dataset contain phone numbers with four or more digits?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_004_1004716_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_004_1004716_qa_2"
6
  description = "What is the most frequent category in the 'Type' column of the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Type 2"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_004_1004716_qa_2"
6
  description = "What is the most frequent category in the 'Type' column of the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_024_1024990_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_024_1024990_qa_2"
6
  description = "What is the total number of samples across all glass types in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "214"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_024_1024990_qa_2"
6
  description = "What is the total number of samples across all glass types in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_027_1027372_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_027_1027372_qa_2"
6
  description = "How many Pokémon in the dataset are classified as dual-type (have a non-null Type 2)?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "67"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_027_1027372_qa_2"
6
  description = "How many Pokémon in the dataset are classified as dual-type (have a non-null Type 2)?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_029_1029150_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_029_1029150_qa_2"
6
  description = "Which state has the highest total number of homicide crimes recorded in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "California"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_029_1029150_qa_2"
6
  description = "Which state has the highest total number of homicide crimes recorded in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_036_1036744_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_036_1036744_qa_1"
6
  description = "Which ride purpose has the highest average miles driven per ride in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Commute"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_036_1036744_qa_1"
6
  description = "Which ride purpose has the highest average miles driven per ride in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_042_1042725_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_042_1042725_qa_2"
6
  description = "Which country has the highest number of top 100 female marathon runners in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "USA"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 3
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_042_1042725_qa_2"
6
  description = "Which country has the highest number of top 100 female marathon runners in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 3
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_044_1044175_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_044_1044175_qa_3"
6
  description = "What is the total number of miles driven by the Uber driver according to the dataset, as indicated in the \"Totals\" row before it was removed?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "12204.7"
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_044_1044175_qa_3"
6
  description = "What is the total number of miles driven by the Uber driver according to the dataset, as indicated in the \"Totals\" row before it was removed?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_062_1062946_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_062_1062946_qa_2"
6
  description = "Which team has the highest number of games played as the blue team in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "SKT"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_062_1062946_qa_2"
6
  description = "Which team has the highest number of games played as the blue team in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_065_1065379_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_065_1065379_qa_3"
6
  description = "What is the most frequently occurring activity class in the training dataset based on the distribution visualization?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "A"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_065_1065379_qa_3"
6
  description = "What is the most frequently occurring activity class in the training dataset based on the distribution visualization?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_067_1067596_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_067_1067596_qa_5"
6
  description = "How many channels are present in each image pair in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "2"
15
  reward_mode_initial = "numeric"
16
  package_tier = 3
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_067_1067596_qa_5"
6
  description = "How many channels are present in each image pair in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 3
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_068_1068916_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-eval-v1/0001_068_1068916_qa_1"
6
  description = "What is the average spam score across all entries in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "0.026"
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_068_1068916_qa_1"
6
  description = "What is the average spam score across all entries in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_068_1068916_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_068_1068916_qa_5"
6
  description = "What is the average number of comments per entry in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "0.038"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_068_1068916_qa_5"
6
  description = "What is the average number of comments per entry in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_078_1078204_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_078_1078204_qa_2"
6
  description = "What is the average duration of games in which SKT T1 participated?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "37.79"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_078_1078204_qa_2"
6
  description = "What is the average duration of games in which SKT T1 participated?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_079_1079149_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_079_1079149_qa_2"
6
  description = "What is the average word count difference between spam and ham messages?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "9.57"
15
  reward_mode_initial = "numeric"
16
  package_tier = 2
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_079_1079149_qa_2"
6
  description = "What is the average word count difference between spam and ham messages?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 2
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_082_1082305_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_082_1082305_qa_1"
6
  description = "What is the proportion of edible mushrooms (class 0) in the dataset compared to the total number of mushrooms?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "51.8"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_082_1082305_qa_1"
6
  description = "What is the proportion of edible mushrooms (class 0) in the dataset compared to the total number of mushrooms?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_083_1083006_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_083_1083006_qa_5"
6
  description = "Which crime type is most frequently associated with the \"Invest Cont\" status in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "TRAFFIC DR #"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 0
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_083_1083006_qa_5"
6
  description = "Which crime type is most frequently associated with the \"Invest Cont\" status in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 0
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_091_1091047_qa_4/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-eval-v1/0001_091_1091047_qa_4"
6
  description = "What is the median trip duration in seconds after removing the top 0.5% of outlier durations?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "515"
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_091_1091047_qa_4"
6
  description = "What is the median trip duration in seconds after removing the top 0.5% of outlier durations?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_091_1091047_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_091_1091047_qa_5"
6
  description = "What is the total number of unique days represented in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "733"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_091_1091047_qa_5"
6
  description = "What is the total number of unique days represented in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_095_1095486_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_095_1095486_qa_2"
6
  description = "Which day of the week has the lowest number of traffic-related 911 calls according to the analysis?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Sunday"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_095_1095486_qa_2"
6
  description = "Which day of the week has the lowest number of traffic-related 911 calls according to the analysis?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_102_1102505_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_102_1102505_qa_2"
6
  description = "Which political party secured the highest total number of votes in the UP election based on the dataset analysis?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "BJP+"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_102_1102505_qa_2"
6
  description = "Which political party secured the highest total number of votes in the UP election based on the dataset analysis?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_104_1104901_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_104_1104901_qa_2"
6
  description = "What is the most common reason cited for negative tweets in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Customer Service Issue"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_104_1104901_qa_2"
6
  description = "What is the most common reason cited for negative tweets in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_104_1104901_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_104_1104901_qa_3"
6
  description = "Which airline received the highest number of neutral sentiment tweets in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Delta"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_104_1104901_qa_3"
6
  description = "Which airline received the highest number of neutral sentiment tweets in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_104_1104959_qa_2/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_104_1104959_qa_2"
6
  description = "What is the most common sentiment category in the dataset based on tweet distribution?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Negative"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 3
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_104_1104959_qa_2"
6
  description = "What is the most common sentiment category in the dataset based on tweet distribution?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 3
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_105_1105529_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_105_1105529_qa_5"
6
  description = "What is the strongest negative correlation with the Happiness Score in the 2015 dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Happiness Rank"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_105_1105529_qa_5"
6
  description = "What is the strongest negative correlation with the Happiness Score in the 2015 dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_105_1105892_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_105_1105892_qa_3"
6
  description = "What is the maximum number of times a single player was considered for Hall of Fame induction across all years?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "20"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_105_1105892_qa_3"
6
  description = "What is the maximum number of times a single player was considered for Hall of Fame induction across all years?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_105_1105892_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-eval-v1/0001_105_1105892_qa_5"
6
  description = "What is the average number of awards received by players who received any awards in the dataset?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "4.59"
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_105_1105892_qa_5"
6
  description = "What is the average number of awards received by players who received any awards in the dataset?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 0
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_116_1116342_qa_3/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "data-agent-train-v1/0001_116_1116342_qa_3"
6
  description = "What is the most frequent cap surface type in the original dataset before encoding?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "y"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_116_1116342_qa_3"
6
  description = "What is the most frequent cap surface type in the original dataset before encoding?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_116_1116342_qa_5/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_116_1116342_qa_5"
6
  description = "What is the most common cap color category in the original dataset before encoding?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "n"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_116_1116342_qa_5"
6
  description = "What is the most common cap color category in the original dataset before encoding?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]
tasks/0001_123_1123075_qa_4/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify/0001_123_1123075_qa_4"
6
  description = "Which constituency (ac) had the highest total votes secured by BJP+ candidates?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "Sahibabad"
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
 
18
  difficulty_tier = "medium"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_123_1123075_qa_4"
6
  description = "Which constituency (ac) had the highest total votes secured by BJP+ candidates?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "exact_short"
16
  package_tier = 1
17
  difficulty_level = 2
18
+ difficulty = "medium"
19
  difficulty_tier = "medium"
20
 
21
  [environment]
tasks/0001_124_1124736_qa_1/task.toml CHANGED
@@ -2,10 +2,10 @@ schema_version = "1.2"
2
  artifacts = []
3
 
4
  [task]
5
- name = "train-verify2/0001_124_1124736_qa_1"
6
  description = "What percentage of transactions in the dataset are classified as fraudulent?"
7
  authors = []
8
- keywords = ["data-agent", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
@@ -15,6 +15,7 @@ gold_answer = "0.1727"
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
 
18
  difficulty_tier = "easy"
19
 
20
  [environment]
 
2
  artifacts = []
3
 
4
  [task]
5
+ name = "smoldataenvs-eval/0001_124_1124736_qa_1"
6
  description = "What percentage of transactions in the dataset are classified as fraudulent?"
7
  authors = []
8
+ keywords = ["smoldataenvs", "data-analysis", "kaggle"]
9
 
10
  [metadata]
11
  source_dataset = "jupyter-agent/jupyter-agent-dataset"
 
15
  reward_mode_initial = "numeric"
16
  package_tier = 1
17
  difficulty_level = 1
18
+ difficulty = "easy"
19
  difficulty_tier = "easy"
20
 
21
  [environment]