first commit

Pipeline.ipynb

@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 41,
+   "execution_count": 92,
    "id": "edc2e2ff",
    "metadata": {},
    "outputs": [],
@@ -19,7 +19,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 93,
    "id": "76f25cc3",
    "metadata": {},
    "outputs": [
@@ -27,7 +27,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "11/08/2022 14:17:47 - INFO - huggingsound.speech_recognition.model - Loading model...\n"
+      "12/06/2022 13:42:19 - INFO - huggingsound.speech_recognition.model - Loading model...\n"
      ]
     },
     {
@@ -41,8 +41,8 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "11/08/2022 14:17:49 - WARNING - root - bos_token <s> not in provided tokens. It will be added to the list of tokens\n",
-      "11/08/2022 14:17:49 - WARNING - root - eos_token </s> not in provided tokens. It will be added to the list of tokens\n"
+      "12/06/2022 13:42:23 - WARNING - root - bos_token <s> not in provided tokens. It will be added to the list of tokens\n",
+      "12/06/2022 13:42:23 - WARNING - root - eos_token </s> not in provided tokens. It will be added to the list of tokens\n"
      ]
     }
    ],
@@ -55,7 +55,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 38,
+   "execution_count": 94,
    "id": "3b142546",
    "metadata": {},
    "outputs": [],
@@ -82,8 +82,8 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 39,
-   "id": "48bed0f8",
+   "execution_count": 95,
+   "id": "b0fae1dd",
    "metadata": {},
    "outputs": [
     {
@@ -98,27 +98,34 @@
      "output_type": "stream",
      "text": [
       "Detected language is English\n",
-      "WITHOUT THE DATA SET THE ARTICLE IS USELESS\n"
+      "NISHE JUAN FANMA HE MOVED ABOUT INVISIBLE BUT EVERYONE COULD HEAR HIM\n"
      ]
     }
    ],
    "source": [
-    "class Speech_to_Text(Pipeline):\n",
-    "    def postprocess(self,model_outputs):\n",
-    "        if prediction[3][0] == 'zh: Chinese':\n",
-    "        "
+    "start = time.time()\n",
+    "pipeline('combine.wav')\n",
+    "end = time.time()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 51,
-   "id": "b0fae1dd",
+   "execution_count": 96,
+   "id": "1e0321b5",
    "metadata": {},
    "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Detected Language is Chinese\n"
+     ]
+    },
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
+      "100%|█████████████████████████████████████████████| 1/1 [00:00<00:00,  1.28it/s]\n",
       "It is strongly recommended to pass the ``sampling_rate`` argument to this function. Failing to do so can result in silent errors that might be hard to debug.\n"
      ]
     },
@@ -126,36 +133,57 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Detected language is English\n",
-      "WITHOUT THE DATA SET THE ARTICLE IS USELESS\n"
+      "你喜欢饭吗\n",
+      "Detected language is English\n"
      ]
-    }
-   ],
-   "source": [
-    "start = time.time()\n",
-    "pipeline('english.wav')\n",
-    "end = time.time()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 49,
-   "id": "1e0321b5",
-   "metadata": {},
-   "outputs": [
+    },
     {
-     "data": {
-      "text/plain": [
-       "0.5424931049346924"
-      ]
-     },
-     "execution_count": 49,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "It is strongly recommended to pass the ``sampling_rate`` argument to this function. Failing to do so can result in silent errors that might be hard to debug.\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "Detected language is English\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "It is strongly recommended to pass the ``sampling_rate`` argument to this function. Failing to do so can result in silent errors that might be hard to debug.\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "HE MOVED ABOUT\n",
+      "Detected language is English\n",
+      "INVISIBLE BUT EVERYONE COULD HEAR HIM\n"
+     ]
     }
    ],
    "source": [
-    "end - start"
+    "from pydub import AudioSegment\n",
+    "from pydub.silence import split_on_silence\n",
+    "\n",
+    "sound_file = AudioSegment.from_wav(\"combine.wav\")\n",
+    "audio_chunks = split_on_silence(sound_file, \n",
+    "    min_silence_len=100,\n",
+    "    silence_thresh=-50\n",
+    ")\n",
+    "\n",
+    "for i, chunk in enumerate(audio_chunks):\n",
+    "\n",
+    "    out_file = \"./chunk{0}.wav\".format(i)\n",
+    "    chunk.export(out_file, format=\"wav\")\n",
+    "    pipeline(out_file)"
    ]
   },
   {

tmp/classifier.ckpt

@@ -0,0 +1 @@
+/Users/varun/.cache/huggingface/hub/models--speechbrain--lang-id-voxlingua107-ecapa/snapshots/d771b530cec097adc0088b4dbd173e242f895464/classifier.ckpt

tmp/embedding_model.ckpt

@@ -0,0 +1 @@
+/Users/varun/.cache/huggingface/hub/models--speechbrain--lang-id-voxlingua107-ecapa/snapshots/d771b530cec097adc0088b4dbd173e242f895464/embedding_model.ckpt

tmp/hyperparams.yaml

@@ -0,0 +1 @@
+/Users/varun/.cache/huggingface/hub/models--speechbrain--lang-id-voxlingua107-ecapa/snapshots/d771b530cec097adc0088b4dbd173e242f895464/hyperparams.yaml

tmp/label_encoder.ckpt

@@ -0,0 +1 @@
+/Users/varun/.cache/huggingface/hub/models--speechbrain--lang-id-voxlingua107-ecapa/snapshots/d771b530cec097adc0088b4dbd173e242f895464/label_encoder.txt