Upload bulgarian_training_script.ipynb

bulgarian_training_script.ipynb

@@ -212,7 +212,8 @@
      "base_uri": "https://localhost:8080/"
     },
     "id": "Mz4bubhxxsad",
-    "outputId": "23398525-cc19-43c2-9fec-497e06214f29"
+    "outputId": "23398525-cc19-43c2-9fec-497e06214f29",
+    "scrolled": true
    },
    "outputs": [
     {
@@ -1898,58 +1899,139 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 25,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/ba/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n"
+      "Reusing dataset common_voice (/workspace/.cache/huggingface/datasets/mozilla-foundation___common_voice/bg/7.0.0/33e08856cfa0d0665e837bcad73ffd920a0bc713ce8c5fffb55dbdf1c084d5ba)\n"
      ]
     },
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "400e81b47cf84505b84756bd0cdb853f",
+       "model_id": "16eaba3c65494a12921df272388262b6",
        "version_major": 2,
        "version_minor": 0
       },
       "text/plain": [
-       "  0%|          | 0/10 [00:00<?, ?ex/s]"
+       "Downloading:   0%|          | 0.00/1.99k [00:00<?, ?B/s]"
       ]
      },
      "metadata": {},
      "output_type": "display_data"
     },
     {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "['шағир дуҫымдың һүҙҙәрен ишеткәс ҡыуанып киттем', 'йөҙөҙ һағышлы', 'тик ниңәлер ул һуңғы йылдарҙа йыш ауырый башланы', 'нимә һас йөрөйгегөҙм әллә', 'оятпым хурлыҡ', 'ҡараһаң атай әсәйҙең ҙур фортаһын ҡулына алған да текәлеп тиҙ бултыра', 'һин дә хан балаһы түгел', 'юҡ миңә маддоннаның уң яҡҡабырғаһында бәүелеп', 'ә бына иляста кәше итеп ҡуйы бик шәп иткәнһең', 'беҙ быларҙы кемгә генә маҡтанып күрһәтмәнек']\n"
-     ]
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "a7d77fd025284c198de8148a66ac6e23",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Downloading:   0%|          | 0.00/212 [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "b42cb83436dc43e5aec83d4bc1b20317",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Downloading:   0%|          | 0.00/1.18G [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
     },
     {
      "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "0dc58dd4cf5542a4b5627d0748dd1574",
+       "version_major": 2,
+       "version_minor": 0
+      },
       "text/plain": [
-       "'Шағир дуҫымдың һүҙҙәрен ишеткәс, ҡыуанып киттем.'"
+       "Downloading:   0%|          | 0.00/260 [00:00<?, ?B/s]"
       ]
      },
-     "execution_count": 25,
      "metadata": {},
-     "output_type": "execute_result"
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "c39a37f06df548caa547615d9edc8450",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Downloading:   0%|          | 0.00/384 [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "90f133429122486ebac0db594fc133a2",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Downloading:   0%|          | 0.00/23.0 [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "2ac0d8c0256d4470a28d8d7a2c451990",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Downloading:   0%|          | 0.00/309 [00:00<?, ?B/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "7a6b55c283d44d1ca6100799737eccbc",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "  0%|          | 0/10 [00:00<?, ?ex/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
     }
    ],
    "source": [
     "from datasets import Audio, Dataset, load_dataset, load_metric\n",
     "from transformers import AutoFeatureExtractor, pipeline\n",
     "\n",
-    "dataset = load_dataset(\"mozilla-foundation/common_voice_7_0\", \"ba\", use_auth_token=True, split=\"train+validation\")\n",
+    "dataset = load_dataset(\"mozilla-foundation/common_voice_7_0\", \"bg\", use_auth_token=True, split=\"train+validation\")\n",
     "\n",
     "# for testing: only process the first two examples as a test\n",
     "dataset = dataset.select(range(10))\n",
     "\n",
-    "repo_name = 'infinitejoy/wav2vec2-large-xls-r-300m-bashkir'\n",
+    "repo_name = 'infinitejoy/wav2vec2-large-xls-r-300m-bulgarian'\n",
     "\n",
     "# load processor\n",
     "feature_extractor = AutoFeatureExtractor.from_pretrained(repo_name)\n",