Upload folder using huggingface_hub

.env

@@ -0,0 +1 @@
+REPLICATE_API_TOKEN=r8_DUSwbx0Ce5hM8rsZmDyizDX4vRrM3W00j2TjA

01_eddacc_exploration.ipynb

@@ -0,0 +1,554 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "d54feda5",
+   "metadata": {},
+   "source": [
+    "## EddAcc Exploration\n",
+    "We visualise the speaker identities/accents in the EddAcc test and dev splits (those are the splits used to avoid splits potentially seen by SLMs)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "58b9fbe9",
+   "metadata": {
+    "vscode": {
+     "languageId": "bat"
+    }
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "ERROR: Python with virtualenvwrapper module not found!\n",
+      "Either, install virtualenvwrapper module for the default python3 interpreter\n",
+      "or set VIRTUALENVWRAPPER_PYTHON to the interpreter to use.\n",
+      "Requirement already satisfied: requests in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 1)) (2.32.5)\n",
+      "Requirement already satisfied: datasets in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 2)) (3.6.0)\n",
+      "Requirement already satisfied: tqdm in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 3)) (4.67.1)\n",
+      "Requirement already satisfied: ipywidgets in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 4)) (8.1.8)\n",
+      "Requirement already satisfied: torchaudio in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 5)) (2.2.2)\n",
+      "Requirement already satisfied: torch in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 6)) (2.2.2)\n",
+      "Requirement already satisfied: torchcodec in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 8)) (0.8.1)\n",
+      "Requirement already satisfied: deepmultilingualpunctuation in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from -r requirements.txt (line 9)) (1.0.1)\n",
+      "Requirement already satisfied: idna<4,>=2.5 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from requests->-r requirements.txt (line 1)) (3.10)\n",
+      "Requirement already satisfied: urllib3<3,>=1.21.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from requests->-r requirements.txt (line 1)) (2.5.0)\n",
+      "Requirement already satisfied: certifi>=2017.4.17 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from requests->-r requirements.txt (line 1)) (2025.8.3)\n",
+      "Requirement already satisfied: charset_normalizer<4,>=2 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from requests->-r requirements.txt (line 1)) (3.4.3)\n",
+      "Requirement already satisfied: huggingface-hub>=0.24.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (0.35.1)\n",
+      "Requirement already satisfied: pandas in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (2.3.2)\n",
+      "Requirement already satisfied: xxhash in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (3.6.0)\n",
+      "Requirement already satisfied: pyarrow>=15.0.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (21.0.0)\n",
+      "Requirement already satisfied: filelock in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (3.19.1)\n",
+      "Requirement already satisfied: numpy>=1.17 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (1.26.4)\n",
+      "Requirement already satisfied: pyyaml>=5.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (6.0.3)\n",
+      "Requirement already satisfied: dill<0.3.9,>=0.3.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (0.3.8)\n",
+      "Requirement already satisfied: packaging in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (25.0)\n",
+      "Requirement already satisfied: fsspec[http]<=2025.3.0,>=2023.1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (2025.3.0)\n",
+      "Requirement already satisfied: multiprocess<0.70.17 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (0.70.16)\n",
+      "Requirement already satisfied: comm>=0.1.3 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipywidgets->-r requirements.txt (line 4)) (0.2.3)\n",
+      "Requirement already satisfied: traitlets>=4.3.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipywidgets->-r requirements.txt (line 4)) (5.14.3)\n",
+      "Requirement already satisfied: ipython>=6.1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipywidgets->-r requirements.txt (line 4)) (8.37.0)\n",
+      "Requirement already satisfied: jupyterlab_widgets~=3.0.15 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipywidgets->-r requirements.txt (line 4)) (3.0.16)\n",
+      "Requirement already satisfied: widgetsnbextension~=4.0.14 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipywidgets->-r requirements.txt (line 4)) (4.0.15)\n",
+      "Requirement already satisfied: networkx in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from torch->-r requirements.txt (line 6)) (3.4.2)\n",
+      "Requirement already satisfied: jinja2 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from torch->-r requirements.txt (line 6)) (3.1.6)\n",
+      "Requirement already satisfied: sympy in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from torch->-r requirements.txt (line 6)) (1.14.0)\n",
+      "Requirement already satisfied: typing-extensions>=4.8.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from torch->-r requirements.txt (line 6)) (4.15.0)\n",
+      "Requirement already satisfied: librosa in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (0.11.0)\n",
+      "Requirement already satisfied: soundfile>=0.12.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (0.13.1)\n",
+      "Requirement already satisfied: soxr>=0.4.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from datasets->-r requirements.txt (line 2)) (1.0.0)\n",
+      "Requirement already satisfied: transformers in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from deepmultilingualpunctuation->-r requirements.txt (line 9)) (4.56.2)\n",
+      "Requirement already satisfied: aiohttp!=4.0.0a0,!=4.0.0a1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (3.12.15)\n",
+      "Requirement already satisfied: hf-xet<2.0.0,>=1.1.3 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from huggingface-hub>=0.24.0->datasets->-r requirements.txt (line 2)) (1.1.10)\n",
+      "Requirement already satisfied: pygments>=2.4.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (2.19.2)\n",
+      "Requirement already satisfied: pexpect>4.3 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (4.9.0)\n",
+      "Requirement already satisfied: jedi>=0.16 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.19.2)\n",
+      "Requirement already satisfied: prompt_toolkit<3.1.0,>=3.0.41 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (3.0.52)\n",
+      "Requirement already satisfied: matplotlib-inline in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.2.1)\n",
+      "Requirement already satisfied: decorator in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (5.2.1)\n",
+      "Requirement already satisfied: stack_data in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.6.3)\n",
+      "Requirement already satisfied: exceptiongroup in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (1.3.0)\n",
+      "Requirement already satisfied: cffi>=1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from soundfile>=0.12.1->datasets->-r requirements.txt (line 2)) (2.0.0)\n",
+      "Requirement already satisfied: MarkupSafe>=2.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from jinja2->torch->-r requirements.txt (line 6)) (2.1.5)\n",
+      "Requirement already satisfied: msgpack>=1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (1.1.1)\n",
+      "Requirement already satisfied: lazy_loader>=0.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (0.4)\n",
+      "Requirement already satisfied: joblib>=1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (1.5.2)\n",
+      "Requirement already satisfied: audioread>=2.1.9 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (3.0.1)\n",
+      "Requirement already satisfied: scipy>=1.6.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (1.15.3)\n",
+      "Requirement already satisfied: scikit-learn>=1.1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (1.7.2)\n",
+      "Requirement already satisfied: numba>=0.51.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (0.62.0)\n",
+      "Requirement already satisfied: pooch>=1.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from librosa->datasets->-r requirements.txt (line 2)) (1.8.2)\n",
+      "Requirement already satisfied: pytz>=2020.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from pandas->datasets->-r requirements.txt (line 2)) (2025.2)\n",
+      "Requirement already satisfied: tzdata>=2022.7 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from pandas->datasets->-r requirements.txt (line 2)) (2025.2)\n",
+      "Requirement already satisfied: python-dateutil>=2.8.2 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from pandas->datasets->-r requirements.txt (line 2)) (2.9.0.post0)\n",
+      "Requirement already satisfied: mpmath<1.4,>=1.1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from sympy->torch->-r requirements.txt (line 6)) (1.3.0)\n",
+      "Requirement already satisfied: regex!=2019.12.17 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from transformers->deepmultilingualpunctuation->-r requirements.txt (line 9)) (2025.9.18)\n",
+      "Requirement already satisfied: tokenizers<=0.23.0,>=0.22.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from transformers->deepmultilingualpunctuation->-r requirements.txt (line 9)) (0.22.1)\n",
+      "Requirement already satisfied: safetensors>=0.4.3 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from transformers->deepmultilingualpunctuation->-r requirements.txt (line 9)) (0.6.2)\n",
+      "Requirement already satisfied: attrs>=17.3.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (25.3.0)\n",
+      "Requirement already satisfied: multidict<7.0,>=4.5 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (6.6.4)\n",
+      "Requirement already satisfied: async-timeout<6.0,>=4.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (5.0.1)\n",
+      "Requirement already satisfied: aiosignal>=1.4.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (1.4.0)\n",
+      "Requirement already satisfied: frozenlist>=1.1.1 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (1.7.0)\n",
+      "Requirement already satisfied: propcache>=0.2.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (0.3.2)\n",
+      "Requirement already satisfied: yarl<2.0,>=1.17.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (1.20.1)\n",
+      "Requirement already satisfied: aiohappyeyeballs>=2.5.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from aiohttp!=4.0.0a0,!=4.0.0a1->fsspec[http]<=2025.3.0,>=2023.1.0->datasets->-r requirements.txt (line 2)) (2.6.1)\n",
+      "Requirement already satisfied: pycparser in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from cffi>=1.0->soundfile>=0.12.1->datasets->-r requirements.txt (line 2)) (2.23)\n",
+      "Requirement already satisfied: parso<0.9.0,>=0.8.4 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from jedi>=0.16->ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.8.5)\n",
+      "Requirement already satisfied: llvmlite<0.46,>=0.45.0dev0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from numba>=0.51.0->librosa->datasets->-r requirements.txt (line 2)) (0.45.0)\n",
+      "Requirement already satisfied: ptyprocess>=0.5 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from pexpect>4.3->ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.7.0)\n",
+      "Requirement already satisfied: platformdirs>=2.5.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from pooch>=1.1->librosa->datasets->-r requirements.txt (line 2)) (4.4.0)\n",
+      "Requirement already satisfied: wcwidth in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from prompt_toolkit<3.1.0,>=3.0.41->ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.2.14)\n",
+      "Requirement already satisfied: six>=1.5 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from python-dateutil>=2.8.2->pandas->datasets->-r requirements.txt (line 2)) (1.17.0)\n",
+      "Requirement already satisfied: threadpoolctl>=3.1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from scikit-learn>=1.1.0->librosa->datasets->-r requirements.txt (line 2)) (3.6.0)\n",
+      "Requirement already satisfied: asttokens>=2.1.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from stack_data->ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (3.0.1)\n",
+      "Requirement already satisfied: pure-eval in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from stack_data->ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (0.2.3)\n",
+      "Requirement already satisfied: executing>=1.2.0 in /Users/cminixho/.pyenv/versions/3.10.13/lib/python3.10/site-packages (from stack_data->ipython>=6.1.0->ipywidgets->-r requirements.txt (line 4)) (2.2.1)\n",
+      "\n",
+      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.0.1\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m25.3\u001b[0m\n",
+      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49m/Users/cminixho/.pyenv/versions/3.10.13/bin/python -m pip install --upgrade pip\u001b[0m\n",
+      "Note: you may need to restart the kernel to use updated packages.\n"
+     ]
+    }
+   ],
+   "source": [
+    "%pip install -r requirements.txt"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "62334e1e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset, concatenate_datasets\n",
+    "\n",
+    "ds = load_dataset(\"edinburghcstr/edacc\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "695479b7",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ds_all = concatenate_datasets([ds[\"validation\"], ds[\"test\"]])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 35,
+   "id": "afd697e1",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "531f921a89bc45edb1b57e0ffcd6cb25",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/19137 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "import torch\n",
+    "from tqdm.auto import tqdm\n",
+    "from transformers import AutoModelForCausalLM, AutoTokenizer\n",
+    "from pathlib import Path\n",
+    "import torchaudio\n",
+    "import json\n",
+    "\n",
+    "checkpoint = \"HuggingFaceTB/SmolLM2-135M\"\n",
+    "\n",
+    "tokenizer = AutoTokenizer.from_pretrained(checkpoint)\n",
+    "# Use bfloat16 for efficiency\n",
+    "model = AutoModelForCausalLM.from_pretrained(checkpoint)\n",
+    "\n",
+    "# Get the token ID for \"?\"\n",
+    "question_mark_token = tokenizer.encode(\"?\", add_special_tokens=False)[0]\n",
+    "\n",
+    "questions = []\n",
+    "\n",
+    "Path(\"questions\").mkdir(exist_ok=True)\n",
+    "\n",
+    "for i in tqdm(range(len(ds_all))):\n",
+    "    # Use lowercase text as input\n",
+    "    text = ds_all[i][\"text\"].lower()\n",
+    "    inputs = tokenizer.encode(text, return_tensors=\"pt\").to(model.device)\n",
+    "\n",
+    "    # Get model predictions for next token\n",
+    "    with torch.no_grad():\n",
+    "        outputs = model(inputs)\n",
+    "        # Get logits for the last position (next token prediction)\n",
+    "        next_token_logits = outputs.logits[0, -1, :]\n",
+    "        # Get the most likely next token\n",
+    "        predicted_token_id = torch.argmax(next_token_logits).item()\n",
+    "\n",
+    "    # Check if predicted next token is a question mark\n",
+    "    if predicted_token_id == question_mark_token:\n",
+    "        torchaudio.save(\n",
+    "            f\"questions/{i}.wav\",\n",
+    "            torch.from_numpy(ds_all[i][\"audio\"][\"array\"]).unsqueeze(0).float(),\n",
+    "            ds_all[i][\"audio\"][\"sampling_rate\"],\n",
+    "        )\n",
+    "        with open(f\"questions/{i}.json\", \"w\") as f:\n",
+    "            json_out = ds_all[i].copy()\n",
+    "            del json_out[\"audio\"]\n",
+    "            json.dump(json_out, f)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "5999fd61",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from dotenv import load_dotenv\n",
+    "from pathlib import Path\n",
+    "import json\n",
+    "\n",
+    "load_dotenv()\n",
+    "\n",
+    "import replicate"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "42d1c793",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "accents_gender = {}\n",
+    "for item in sorted(Path(\"questions\").glob(\"*.json\")):\n",
+    "    with open(item, \"r\") as f:\n",
+    "        data = json.load(f)\n",
+    "    if len(data[\"text\"]) > 50:\n",
+    "        if data[\"accent\"] == \"Mainstream US English\" and data[\"l1\"] != \"Mainstream US English\":\n",
+    "            continue\n",
+    "        if data[\"speaker\"] == \"EDACC-C19-A\":\n",
+    "            continue\n",
+    "        accent_gender_key = f\"{data['accent']}-{data['gender']}\"\n",
+    "        if accent_gender_key not in accents_gender:\n",
+    "            accents_gender[accent_gender_key] = item.stem\n",
+    "    else:\n",
+    "        pass"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "id": "80ed0d46",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "German-female\n",
+      "Bulgarian-male\n",
+      "Romanian-female\n",
+      "Lithuanian-male\n",
+      "Ghanain English-female\n",
+      "Japanese-female\n",
+      "Indonesian-female\n",
+      "Egyptian-male\n",
+      "Latin-female\n",
+      "Brazilian-male\n"
+     ]
+    }
+   ],
+   "source": [
+    "accent_gender_both = {}\n",
+    "for k in accents_gender:\n",
+    "    if \"-male\" in k:\n",
+    "        if  k.replace(\"-male\", \"-female\") in accents_gender:\n",
+    "            accent_gender_both[k] = accents_gender[k]\n",
+    "        else:\n",
+    "            print(k)\n",
+    "    elif \"-female\" in k:\n",
+    "        if k.replace(\"-female\", \"-male\") in accents_gender:\n",
+    "            accent_gender_both[k] = accents_gender[k]\n",
+    "        else:\n",
+    "            print(k)\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "3df3e24f",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "{'Irish English', 'Eastern European', 'Catalan', 'Indonesian English', 'Vietnamese', 'French', 'Southern British English', 'Mainstream US English', 'South African English', 'Jamaican English', 'Spanish', 'Latin American', 'Chinese', 'Kenyan English', \"Don't know\", 'European', 'Italian', 'Indian English', 'Nigerian English', 'Scottish English'}\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(set([x.split(\"-\")[0] for x in accent_gender_both.keys()]))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "id": "f09bc7e8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "keys_to_keep = [\n",
+    "    \"Mainstream US\",\n",
+    "    \"Southern British\",\n",
+    "    \"Indian\",\n",
+    "    \"Chinese\",\n",
+    "    \"Latin American\",\n",
+    "    \"Eastern European\",\n",
+    "]\n",
+    "\n",
+    "final_dict = {}\n",
+    "for k in accent_gender_both:\n",
+    "    if any(x in k for x in keys_to_keep):\n",
+    "        final_dict[k] = accent_gender_both[k]\n",
+    "    else:\n",
+    "        pass\n",
+    "\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "id": "13247f3b",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'Eastern European-male': '10153',\n",
+       " 'Southern British English-female': '1028',\n",
+       " 'Mainstream US English-female': '10293',\n",
+       " 'Chinese-male': '10314',\n",
+       " 'Indian English-female': '10892',\n",
+       " 'Latin American-female': '11303',\n",
+       " 'Latin American-male': '11310',\n",
+       " 'Indian English-male': '13170',\n",
+       " 'Southern British English-male': '16172',\n",
+       " 'Chinese-female': '16385',\n",
+       " 'Mainstream US English-male': '18708',\n",
+       " 'Eastern European-female': '4252'}"
+      ]
+     },
+     "execution_count": 20,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "final_dict"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b33d29a0",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Loaded as API: https://cdminix-megatts3-voice-cloning.hf.space ✔\n",
+      "Total prompts: 80 (40 with hesitation, 40 without)\n",
+      "Total files to generate: 960\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "e7073169b40f4bb2a947d00dc962fe04",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Generating audio:   0%|          | 0/960 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "from gradio_client import Client, handle_file\n",
+    "import shutil\n",
+    "from tqdm.auto import tqdm\n",
+    "import time\n",
+    "\n",
+    "# Initialize MegaTTS3 client\n",
+    "client = Client(\"cdminix/MegaTTS3-Voice-Cloning\")\n",
+    "\n",
+    "# Retry helper function with exponential backoff\n",
+    "def generate_with_retry(client, speaker_audio_path, text, max_retries=5, initial_delay=1):\n",
+    "    \"\"\"Generate speech with retry logic and exponential backoff.\"\"\"\n",
+    "    for attempt in range(max_retries):\n",
+    "        try:\n",
+    "            result = client.predict(\n",
+    "                inp_audio=handle_file(speaker_audio_path),\n",
+    "                inp_text=text,\n",
+    "                infer_timestep=32,\n",
+    "                p_w=1.4,\n",
+    "                t_w=3,\n",
+    "                api_name=\"/generate_speech\"\n",
+    "            )\n",
+    "            return result\n",
+    "        except Exception as e:\n",
+    "            if attempt == max_retries - 1:\n",
+    "                # Last attempt failed, raise the exception\n",
+    "                raise\n",
+    "            # Calculate backoff delay: 1s, 2s, 4s, 8s, 16s\n",
+    "            delay = initial_delay * (2 ** attempt)\n",
+    "            print(f\"\\nRetry {attempt + 1}/{max_retries} after error: {str(e)[:100]}\")\n",
+    "            print(f\"Waiting {delay}s before retrying...\")\n",
+    "            time.sleep(delay)\n",
+    "\n",
+    "df = pd.read_csv(\"tts_prompts.csv\")\n",
+    "df_hesitation = df[df[\"has_disfluency\"] == True]\n",
+    "df_no_hesitation = df[df[\"has_disfluency\"] == False]\n",
+    "\n",
+    "print(f\"Total prompts: {len(df)} ({len(df_hesitation)} with hesitation, {len(df_no_hesitation)} without)\")\n",
+    "\n",
+    "result_rows = []\n",
+    "\n",
+    "# Calculate total number of files to generate\n",
+    "total_files = len(final_dict.keys()) * (len(df_hesitation) + len(df_no_hesitation))\n",
+    "print(f\"Total files to generate: {total_files}\")\n",
+    "\n",
+    "pbar = tqdm(total=total_files, desc=\"Generating audio\")\n",
+    "\n",
+    "for k in sorted(final_dict.keys()):\n",
+    "    accent = k.split(\"-\")[0]\n",
+    "    accent = accent.replace(\" \", \"_\").lower()\n",
+    "    gender = k.split(\"-\")[1]\n",
+    "    Path(f\"dataset/{accent}/{gender}/hesitation\").mkdir(parents=True, exist_ok=True)\n",
+    "    \n",
+    "    speaker_audio_path = \"questions/\" + final_dict[k] + \".wav\"\n",
+    "\n",
+    "    for idx, row in df_hesitation.iterrows():\n",
+    "        _id = f\"{idx}-{accent}-{gender}-{final_dict[k]}-hesitation\"\n",
+    "        output_path = Path(f\"dataset/{accent}/{gender}/hesitation/{_id}.wav\")\n",
+    "        if output_path.exists():\n",
+    "            result_rows.append(\n",
+    "                {\n",
+    "                    \"id\": _id,\n",
+    "                    \"accent\": accent,\n",
+    "                    \"gender\": gender,\n",
+    "                    \"speaker_audio\": speaker_audio_path,\n",
+    "                }\n",
+    "            )\n",
+    "            pbar.update(1)\n",
+    "            continue\n",
+    "        \n",
+    "        # Generate with retry logic\n",
+    "        result = generate_with_retry(client, speaker_audio_path, row[\"text\"])\n",
+    "        \n",
+    "        # Copy the result file to output path\n",
+    "        shutil.copy(result, output_path)\n",
+    "        \n",
+    "        result_rows.append(\n",
+    "            {\n",
+    "                \"id\": _id,\n",
+    "                \"accent\": accent,\n",
+    "                \"gender\": gender,\n",
+    "                \"speaker_audio\": speaker_audio_path,\n",
+    "            }\n",
+    "        )\n",
+    "        pbar.update(1)\n",
+    "\n",
+    "    Path(f\"dataset/{accent}/{gender}/no_hesitation\").mkdir(parents=True, exist_ok=True)\n",
+    "\n",
+    "    for idx, row in df_no_hesitation.iterrows():\n",
+    "        _id = f\"{idx}-{accent}-{gender}-{final_dict[k]}-no_hesitation\"\n",
+    "        output_path = Path(f\"dataset/{accent}/{gender}/no_hesitation/{_id}.wav\")\n",
+    "        if output_path.exists():\n",
+    "            result_rows.append(\n",
+    "                {\n",
+    "                    \"id\": _id,\n",
+    "                    \"accent\": accent,\n",
+    "                    \"gender\": gender,\n",
+    "                    \"speaker_audio\": speaker_audio_path,\n",
+    "                }\n",
+    "            )\n",
+    "            pbar.update(1)\n",
+    "            continue\n",
+    "        \n",
+    "        # Generate with retry logic\n",
+    "        result = generate_with_retry(client, speaker_audio_path, row[\"text\"])\n",
+    "        \n",
+    "        # Copy the result file to output path\n",
+    "        shutil.copy(result, output_path)\n",
+    "        \n",
+    "        result_rows.append(\n",
+    "            {\n",
+    "                \"id\": _id,\n",
+    "                \"accent\": accent,\n",
+    "                \"gender\": gender,\n",
+    "                \"speaker_audio\": speaker_audio_path,\n",
+    "            }\n",
+    "        )\n",
+    "        pbar.update(1)\n",
+    "\n",
+    "pbar.close()\n",
+    "\n",
+    "df_result = pd.DataFrame(result_rows)\n",
+    "df_result.to_csv(\"tts_prompts_result.csv\", index=False)\n",
+    "print(f\"Completed! Generated {len(result_rows)} audio files.\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6c97f22b",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.13"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

dataset/chinese/female/hesitation/40-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58269a6da4ae8ac5e6bbcec1b94bfa56f53b3212c00fa14e1e210ba0300627d9
+size 1152044

dataset/chinese/female/hesitation/41-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0503571dce7d0c2b6ed08384882ed2f2c8bcb30882f8681657f7369f2410b94c
+size 1186604

dataset/chinese/female/hesitation/42-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac710badfe973a0116be5a0e6c960c4c4feb984f56b34396bfbeaa9af3905713
+size 963884

dataset/chinese/female/hesitation/43-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6026bb5e60a79a9a5486fac701790cda6d3036b66252f4876c0c435213f6c69a
+size 894764

dataset/chinese/female/hesitation/44-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5566c15cb607e36c3bd01bb528b43c741205eba694f65f500b644f331624c043
+size 887084

dataset/chinese/female/hesitation/45-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4478d98ef7b19febc490327f301eedcb76dd6e8450e16265c2506ab95bdab4fb
+size 1278764

dataset/chinese/female/hesitation/46-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15cfcc3f0ed8907e641ab87356ce0b657ec1c85d3423ec77de9d9ba965c8d4f0
+size 1052204

dataset/chinese/female/hesitation/47-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cff6e98074cf9fa4da0f762fe8c40e9810a6ba2e1bc7c8596421864557b8871
+size 1217324

dataset/chinese/female/hesitation/48-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9339dce6439fd436b31e1691fa59ec497a8d9dc04a87288f13d42bbb2accc144
+size 1159724

dataset/chinese/female/hesitation/49-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe5ae510d4940bb5a19bd4dfb77c98b8e60185893cd6629cee2a5007eec983de
+size 1098284

dataset/chinese/female/hesitation/50-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f93ad87b67c94732800620551ae5683714f154a0746df9ecd9ed2a5c9ab74675
+size 986924

dataset/chinese/female/hesitation/51-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bac31997c86883a2a8eff859217067f5fe38cd8a3911fe448b71762164981c0
+size 841004

dataset/chinese/female/hesitation/52-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d44bf302ab80c325b72165514132f3de4fda12d00010e1e1d8c3e206c3751b76
+size 883244

dataset/chinese/female/hesitation/53-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12777ea41afd71b804c2b0aacc552e8ff560421e7f1ff734d33f69a27647fcea
+size 725804

dataset/chinese/female/hesitation/54-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6721029e11e2f0d2701bf18f6061c6e1f92429d67bca18a830476b923eec5141
+size 933164

dataset/chinese/female/hesitation/55-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fef2eb93f20932bf7210c5449c0ca6ae602876c902fcba51a901275b1ac4a093
+size 1194284

dataset/chinese/female/hesitation/56-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c57b4a7555b68dd726ab04cfce7e1b6f9dde9306290de78864ecb165d65d520d
+size 1059884

dataset/chinese/female/hesitation/57-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4f7c965644da4a525582c014d3fec0095db1d2857638f2c23aef7710161c028
+size 1075244

dataset/chinese/female/hesitation/58-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:695a87be77dfbc37c5ad40f06e267fe24e64119be46ba79e91cab693001b0608
+size 1098284

dataset/chinese/female/hesitation/59-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8289481b6dc9dfadf7b009bb19fb1e82ca905ee07658772d0e02c1d7f7b94599
+size 1098284

dataset/chinese/female/hesitation/60-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3126998789bfb5d92a884ec5cbc69b9a65073b26b3f7b29c423b842cb34e7b7a
+size 975404

dataset/chinese/female/hesitation/61-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c93f45f057b6bc3a3917b0c2326259f0b084ad48ad7509e6b45e750356261686
+size 879404

dataset/chinese/female/hesitation/62-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1b4c958e342ddaa609bed013695d5acbebae0279e295fc5c6c01f2a04392491
+size 929324

dataset/chinese/female/hesitation/63-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eacf7169dfeb7b6c35423304e3afea5ff07f963108d5cf80d5ad12dcac64b44d
+size 1140524

dataset/chinese/female/hesitation/64-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d386f93c3506b1b52d45e4a3cf8bbd247f11efd8c98ba2b3278cd2611a39052
+size 967724

dataset/chinese/female/hesitation/65-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc9306fbd97dd819909b5cd18f8d06deee1bbee4b85d72aa0363a12512ea8b25
+size 967724

dataset/chinese/female/hesitation/66-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f98ad81c644132957c56ead5fd601d2b247920089d1417ef68e10941c735752d
+size 860204

dataset/chinese/female/hesitation/67-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:021b516a8c82bbd95b3713bc0e4f7ff1dde9909d9740883f767332959c809d47
+size 1017644

dataset/chinese/female/hesitation/68-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2dd4dd75eb029d05851853e6eda9fe2f5674cf661c215158aaecd5fb70067cef
+size 986924

dataset/chinese/female/hesitation/69-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66cba5ac92867f2791b6ec315339f61d542df7d60c30ec1f8b6e6fcc51d5d445
+size 990764

dataset/chinese/female/hesitation/70-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd4fb48dbedea5418d854b3c669bfb61d00106617624f952650d9b1dce99fc7d
+size 967724

dataset/chinese/female/hesitation/71-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc7f5c08dc6649f10a7890505c1350e2af0273aa2630f74c6fe393d2c048511c
+size 952364

dataset/chinese/female/hesitation/72-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:317f44df5fbd2a29b06a1a61048cce798d40ecbceed9bafc7b5494022ea38343
+size 1198124

dataset/chinese/female/hesitation/73-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cbb3cd68848053928db1b177e88e5bb156aff5c6b2872cc3d0891b77c1ad386
+size 1136684

dataset/chinese/female/hesitation/74-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3128c10a41faebe77dcf2e20c7aa8f5ab2e458d12440287e32d95b36114c0313
+size 1090604

dataset/chinese/female/hesitation/75-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ae20a3b90a7be5db41f562252a8e54c0ba8444ea5e51b95948f241c02a89b14
+size 979244

dataset/chinese/female/hesitation/76-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7535824a21ef4015ccc02861d93fe3985ade4a56af4987d0c0ce10157c65271
+size 967724

dataset/chinese/female/hesitation/77-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:240afc2f2916fe59937c1425f1556d0fe875f2924f9aebfc0ca454851408ebf7
+size 1082924

dataset/chinese/female/hesitation/78-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:141a5f4f6e2be384d78690102352ecddcc9f04ea14552d75712b439a6628de31
+size 994604

dataset/chinese/female/hesitation/79-chinese-female-16385-hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:688e21d1b86beb3c5e1d09a81e60518d55f741c0ab79548b9362ea28952022ab
+size 975404

dataset/chinese/female/no_hesitation/0-chinese-female-16385-no_hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a812bbdc9cf27e14b99a3c5348281c1d20fa0d38658475bd65055bc91786f57
+size 844844

dataset/chinese/female/no_hesitation/1-chinese-female-16385-no_hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bf091fc7c90a7ebf74cc1fe0405835fc1b2aab62cea9a03d1ec2dad5db74b9b
+size 902444

dataset/chinese/female/no_hesitation/10-chinese-female-16385-no_hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:473fe3a12f03817bb4efe1489e6372c1697cc43a360ba908c9efee1efd4bbe59
+size 890924

dataset/chinese/female/no_hesitation/11-chinese-female-16385-no_hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c2ebed5d3b91a97cfdac0764da3e72d397caf587968dd9ab275286735de8a5a
+size 794924

dataset/chinese/female/no_hesitation/12-chinese-female-16385-no_hesitation.wav

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52b87590d1d081c803deacda0e8d8267527f2326f70f64fe72dccdb29929bcf3
+size 775724