Skip to content

Feat/create graphs by keys #58

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
mary-silence wants to merge 4 commits into
base: dev
Choose a base branch
from
Open

Feat/create graphs by keys #58

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
3a5e743
add experiments keys to dialog
mary-silence May 5, 2025
ccbc0e7
change paths
mary-silence May 5, 2025
c96ca0a
fix errors
mary-silence May 5, 2025
4428d4f
fix md-file
mary-silence May 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
209 changes: 209 additions & 0 deletions experiments/exp2025_03_27_create_graphs_by_keys/Research_notebook.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,209 @@
{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"provenance": []
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"name": "python"
}
},
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "vS5M-x8G7jV2"
},
"outputs": [],
"source": [
"!pip install -qU openai tiktoken langchain_openai pydantic_settings"
]
},
{
"cell_type": "code",
"source": [
"import os\n",
"\n",
"os.environ[\"OPENAI_API_KEY\"] = \"...\"\n",
"os.environ[\"OPENAI_API_BASE\"] = \"...\""
],
"metadata": {
"id": "n-816XuR8Bwg"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from experiments.exp2025_03_27_create_graphs_by_keys.keys2graph.pipeline import Pipeline\n",
"import json\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt"
],
"metadata": {
"id": "ina5eQDt8OV7"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"key_list = [\"topic\", \"sub_topic\", \"bot_goal\", \"success_criteria\", \"context_info\", \"language\", \"formality_level\", \"emotional_tone\", \"lexical_diversity\", \"use_of_jargon\",\n",
"\"max_dialog_depth\", \"max_branching_factor\", \"mandatory_nodes\", \"optional_nodes\", \"start_node\", \"user_intents\", \"intent_hierarchy\", \"required_slots\",\n",
"\"follow_up_questions\", \"fallback_strategy\", \"confirmation_needed\", \"max_dialog_length\", \"alternate_paths\", \"escalation_policy\", \"user_feedback_collection\",\n",
"\"user_persona\", \"dynamic_content\"]"
],
"metadata": {
"id": "KGwz8DxY7oVZ"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"#remove keys for the experiment\n",
"key_list.remove(\"mandatory_nodes\")"
],
"metadata": {
"id": "kMA7M1z77tXp"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"ROOT = \"/content/experiments/exp2025_03_27_create_graphs_by_keys\"\n",
"src_graphs = f\"{ROOT}/test_original_graphs/graphs_level_1.json\"\n",
"similar_json = f\"{ROOT}/triplet_similarity.json\""
],
"metadata": {
"id": "Q7PeZPWw73rn"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"pipe = Pipeline()"
],
"metadata": {
"id": "wyD-XTTJ8QqL"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"#select a file with original dialog graphs\n",
"all_data = pipe.load_dialog_graphs(src_graphs)"
],
"metadata": {
"id": "WtivXnRx8gqI"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"#load annotated data (if it's ready)\n",
"with open(\"/content/annotated_data.json\", encoding=\"utf-8\") as f:\n",
" pipe._original_graphs_annotation = json.load(f)"
],
"metadata": {
"id": "qmKxNvST8eqc"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"#select original dialog graphs: \"all\" or list if ids\n",
"test_graphs = pipe.add_graphs_to_test(\"all\", src_graphs)"
],
"metadata": {
"id": "ug1TFTOa8vSy"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"# generate graphs by keys and save data\n",
"gen_out = \"generated_data.json\"\n",
"pipe.generate_graphs_by_keys(\n",
" #keys=[\"topic\", \"sub_topic\"], # or \"all\"\n",
" keys=key_list,\n",
" model_name=\"o1-mini\",\n",
" temperature=0.0, #set 0 for \"o1-mini\"\n",
" output_file=gen_out\n",
")"
],
"metadata": {
"id": "vUTpKbD28_-W"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"#calculate metrics\n",
"pipe.calculate_graphs_similarity(similar_json)"
],
"metadata": {
"id": "ie8gXaZZ-PMf"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"#metrics visualisation\n",
"with open(similar_json, encoding=\"utf-8\") as f:\n",
" data = json.load(f)\n",
"\n",
"similarity_values = [p[\"similarity_avg\"] for p in data.get(\"pairs\", [])]\n",
"\n",
"mean_val = np.mean(similarity_values) if similarity_values else 0.0\n",
"var_val = np.var(similarity_values) if similarity_values else 0.0\n",
"\n",
"print(f\"Average similarity_avg : {mean_val:.4f}\")\n",
"print(f\"Dispersion : {var_val:.6f}\")\n",
"\n",
"plt.figure(figsize=(8, 4))\n",
"bins = max(5, min(10, len(similarity_values)//2 or 5))\n",
"plt.hist(similarity_values,\n",
" bins=bins,\n",
" edgecolor=\"black\", # контур столбца\n",
" linewidth=1.2,\n",
" alpha=0.85)\n",
"plt.title(\"Distribution of similarity_avg across graph pairs\")\n",
"plt.xlabel(\"similarity_avg\")\n",
"plt.ylabel(\"Frequency\")\n",
"plt.tight_layout()\n",
"plt.show()\n"
],
"metadata": {
"id": "BLqeOzB3-VQg"
},
"execution_count": null,
"outputs": []
}
]
}
Loading
Loading