initial commit

README.md

@@ -1,3 +1,156 @@
----
-license: cc-by-4.0
----
+---
+license: cc-by-4.0
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+  - split: validation
+    path: data/validation-*
+  - split: test
+    path: data/test-*
+dataset_info:
+  features:
+  - name: system
+    dtype: string
+  - name: in
+    dtype: string
+  - name: out
+    dtype: string
+  splits:
+  - name: train
+    num_bytes: 3795873
+    num_examples: 2384
+  - name: validation
+    num_bytes: 799795
+    num_examples: 500
+  - name: test
+    num_bytes: 798213
+    num_examples: 500
+  download_size: 1055715
+  dataset_size: 5393881
+task_categories:
+- text-generation
+language:
+- en
+tags:
+- code
+- dsl
+- structured-output
+- tool-calling
+pretty_name: DIA Intent Sequencer for simple calculator
+size_categories:
+- 1K<n<10K
+---
+
+# dia-intent-sequencer-calculator-dataset
+
+This dataset provides simple natural language math expressions paired with structured DSL function call trees using the **DIA intent sequencer** format. It serves two complementary purposes:
+
+- 🧪 As a **demonstration and testbed**, it showcases the model's ability to handle nested, composable intent calls and to generalize across randomized function and parameter names—focusing on structural reasoning rather than memorization.  
+- 🧱 As a **foundational training resource**, it offers a rich set of examples that can be extended or combined with domain-specific data to bootstrap a larger, **`DIA`** task-specific intent sequencing model in a user's own application.
+
+It was uploaded using the [`fifo-tool-datasets`](https://github.com/gh9869827/fifo-tool-datasets) utility with the `dsl` adapter, which supports structured `system → in → out` triplets.
+
+The dataset was programmatically generated using [`calculator_synthetic_data_generator.py`](https://github.com/gh9869827/fifo-dev-dsl/blob/main/fifo_dev_dsl/dia/demo/calculator_synthetic_data_generator.py), which builds symbolic expressions and randomized tool mappings for robust generalization.
+
+## 🧩 Related Datasets
+
+This dataset is part of a broader collection of **`DIA`** datasets, each demonstrating different capabilities of intent sequencing and tool-based reasoning:
+
+- [`a6188466/dia-intent-sequencer-robot-arm-dataset`](https://huggingface.co/datasets/a6188466/dia-intent-sequencer-robot-arm-dataset)  
+  Designed for training a model to control a robotic arm handling an inventory of screws. It illustrates:
+  - Mapping of user intent to tool calling  
+  - Dynamic resolution of missing information from runtime-accessible resources  
+  - Interactive clarification of missing or conflicting information via user interaction
+
+- [`a6188466/dia-intent-sequencer-calendar-dataset`](https://huggingface.co/datasets/a6188466/dia-intent-sequencer-calendar-dataset)  
+  Designed for training a task/reminder scheduling agent. It illustrates:
+  - Mapping of user intent to tool calling  
+  - Nesting of intents
+  - Dynamic resolution of missing information from runtime-accessible resources  
+  - Interactive clarification of missing or conflicting information via user interaction
+
+---
+
+## 📦 Format
+
+The dataset uses a **wide format** with the following columns:
+
+- `system` — the system prompt defining the model's task  
+- `in` — a natural language math prompt with a simple arithmetic expression (e.g. `"what is (2 + 3) × 4?"`)  
+- `out` — a symbolic DSL representation composed of nested tool calls (e.g. `multiply(a=add(a=2, b=3), b=4)` — real examples use randomized names)
+
+Each entry is structured as:
+
+```
+system → in → out
+```
+
+---
+
+## 📋 System Prompt Format
+
+The `system` prompt is a structured instruction that tells the model to translate the user input into one or more **atomic DIA DSL function calls** based on a list of available arithmetic function: `add`, `subtract`, `divide` and `multiply`.
+
+Each example includes:
+- A list of available intents (i.e. functions), each with a **randomized name**  
+- Randomized parameter names to prevent overfitting  
+- Descriptions of function behavior and argument roles  
+
+This ensures the model cannot memorize specific function names like `add`, `multiply`, or `divide`, and must instead rely on **semantic understanding** of the prompt and function definitions.
+
+The system prompt always ends with:
+
+```
+QUERY_FILL cannot be used as no information can be retrieved at runtime.
+```
+
+This instruction is included because, unlike scenarios involving real-world environments (e.g. a robotic arm querying an inventory of screws), the calculator use case contains **no missing information** that needs to be retrieved. Every user input provides all the operands explicitly, so there's nothing for the model to "fill in" via external queries. This requires the model to rely entirely on the input structure without any lookup behavior.
+
+> 💡 Example: Instead of hardcoding `add(a=x, b=y)`, a model may need to generate something like `csisz(o=2, u=3)` where `csisz` is the randomized function name for `add`.
+
+---
+
+## 🧠 DIA DSL Project
+
+The DSL used in this dataset is defined in the [DIA module](https://github.com/gh9869827/fifo-dev-dsl/tree/main/fifo_dev_dsl/dia), which implements:
+
+- 🧠 A symbolic DSL for **goal-driven, composable intent execution**  
+- 🧩 Nested function calls and argument mapping  
+- 🔄 Support for randomized function and parameter names to simulate tool diversity  
+- 🤖 Integration with LLMs for automatic parsing and resolution
+
+For a complete example of tool loading, DSL resolution, and evaluation, see [calculator.py](https://github.com/gh9869827/fifo-dev-dsl/blob/main/fifo_dev_dsl/dia/demo/calculator.py).
+
+---
+
+## 🚀 Using the Dataset
+
+To load the dataset:
+
+```python
+from fifo_tool_datasets.sdk.hf_dataset_adapters.dsl import DSLAdapter
+dataset_dict = DSLAdapter().from_hub_to_dataset_dict("a6188466/dia-intent-sequencer-calculator-dataset")
+print(dataset_dict["train"])
+print(dataset_dict["validation"])
+```
+
+This dataset is ideal for:
+- Testing nested intent sequencing  
+- Validating function argument mapping  
+- Evaluating generalization to unseen tool/parameter names  
+
+---
+
+## 🔧 Upload & Editing Tools
+
+Uploaded via [`fifo-tool-datasets`](https://github.com/gh9869827/fifo-tool-datasets), which supports `.dat` ↔ `datasets.Dataset` conversion.
+
+You can modify, re-randomize, or regenerate examples using the CLI or SDK.
+
+---
+
+## 🪪 License
+
+This dataset is licensed under **CC BY 4.0**. See [LICENSE](LICENSE) for details.