diff --git a/README.md b/README.md index 46ec0a54..c6322143 100644 --- a/README.md +++ b/README.md @@ -154,7 +154,7 @@ We are constantly working to make UpTrain better. Want a new feature or need any # License 💻 -This repo is published under Apache 2.0 license. We're currently focused on developing non-enterprise offerings that should cover most use cases by adding more features and extending to more models. We also working towards adding a hosted offering - [contact us](mailto:sourabh@insane.ai) if you are interested. +This repo is published under Apache 2.0 license, with the exception of the ee directory which will contain premium features requiring an enterprise license in the future. We're currently focused on developing non-enterprise offerings that should cover most use cases by adding more features and extending to more models. We also working towards adding a hosted offering - [contact us](mailto:sourabh@insane.ai) if you are interested. # Stay Updated ☎️ We are continuously adding tons of features and use cases. Please support us by giving the project a star ⭐! diff --git a/examples/speech_to_text/run.ipynb b/examples/speech_to_text/run.ipynb new file mode 100644 index 00000000..f36b46a2 --- /dev/null +++ b/examples/speech_to_text/run.ipynb @@ -0,0 +1,346 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "56f37485-2944-4b4e-b37a-911539fd624b", + "metadata": { + "tags": [] + }, + "source": [ + "

\n", + " \n", + " \"uptrain\"\n", + " \n", + "

" + ] + }, + { + "cell_type": "markdown", + "id": "480e699e-bb65-44dd-8bff-b5f0d18886f3", + "metadata": { + "tags": [] + }, + "source": [ + "

Collecting Failure cases for ASR

" + ] + }, + { + "cell_type": "markdown", + "id": "7d1ebc89-c40b-40cb-8a25-f7638a0f785a", + "metadata": { + "tags": [] + }, + "source": [ + "**Objective**: Collect failure cases to improve the Speech to text model.\n", + "\n", + "**Model**: We are working on the `facebook/s2t-small-librispeech-asr model` which is a Speech to Text Transformer (S2T) model trained for automatic speech recognition (ASR). The S2T model was proposed in this [paper](https://arxiv.org/abs/2010.05171) and released in this [repository](https://github.com/facebookresearch/fairseq/tree/main/examples/speech_to_text)\n", + "\n", + "**Dataset**: The model is trained on [LibriSpeech ASR Corpus](https://www.openslr.org/12), a dataset consisting of approximately 1000 hours of 16kHz read English speech.\n", + "\n", + "**Method**: We use the faster-whisper model to identify failure cases" + ] + }, + { + "cell_type": "markdown", + "id": "619871be-388d-4f52-981a-244b26fb0d72", + "metadata": {}, + "source": [ + "#### Install required packages" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "f5bf030a-fd6d-4ae8-be74-31f0d421b7ae", + "metadata": {}, + "outputs": [], + "source": [ + "# Installation steps: https://huggingface.co/docs/transformers/installation\n", + "# Model borrowed from: https://huggingface.co/docs/transformers/model_doc/speech_to_text\n", + "# pip install datasets\n", + "# https://github.com/google/sentencepiece#installation\n", + "# pip install soundfile librosa" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "89047854-f88c-48e9-b0ec-6f61a6533260", + "metadata": { + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/sourabhagrawal/miniconda3/envs/prod_dev2/lib/python3.10/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", + " from .autonotebook import tqdm as notebook_tqdm\n" + ] + } + ], + "source": [ + "import torch\n", + "from transformers import Speech2TextProcessor, Speech2TextForConditionalGeneration, pipeline, AutoTokenizer, AutoModel\n", + "from datasets import load_dataset, Audio\n", + "import warnings\n", + "import pandas as pd\n", + "import numpy as np\n", + "\n", + "import uptrain\n", + "\n", + "warnings.simplefilter('ignore')" + ] + }, + { + "cell_type": "markdown", + "id": "fec6e2e2-84cf-417c-a7a9-82529717535b", + "metadata": {}, + "source": [ + "#### Define our model and datasets" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "c6cd6e69-807b-4cdf-aeaf-f9be914f375b", + "metadata": { + "tags": [] + }, + "outputs": [], + "source": [ + "model = Speech2TextForConditionalGeneration.from_pretrained(\"facebook/s2t-small-librispeech-asr\")\n", + "transcriber = pipeline(\"automatic-speech-recognition\", model=\"facebook/s2t-small-librispeech-asr\")" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "5212a96e-24ea-4496-b57c-640a98a7e5d8", + "metadata": { + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Found cached dataset minds14 (/Users/sourabhagrawal/.cache/huggingface/datasets/PolyAI___minds14/en-US/1.0.0/65c7e0f3be79e18a6ffaf879a083daf706312d421ac90d25718459cbf3c42696)\n", + "Loading cached shuffled indices for dataset at /Users/sourabhagrawal/.cache/huggingface/datasets/PolyAI___minds14/en-US/1.0.0/65c7e0f3be79e18a6ffaf879a083daf706312d421ac90d25718459cbf3c42696/cache-cfa7a2ed5f85e6a7.arrow\n" + ] + } + ], + "source": [ + "def process_dataset(dataset):\n", + " dataset = dataset.cast_column(\"audio\", Audio(sampling_rate=16000))\n", + " return dataset\n", + "\n", + "dataset = process_dataset(load_dataset(\"PolyAI/minds14\", name=\"en-US\", split=\"train\").shuffle(seed=42).train_test_split(test_size=50))['test']" + ] + }, + { + "cell_type": "markdown", + "id": "839dd89d-1b8e-45d3-bacb-9126634e0613", + "metadata": {}, + "source": [ + "Let's define UpTrain config. We will use Monitor.OUTPUT_COMPARISON to compare our model's output against output generated by the [Whisper model](https://github.com/guillaumekln/faster-whisper). We will use RogueL as the metric for comparison of the two model outputs. " + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "id": "b468e3f1-b5b2-456a-9cd8-2329aa74f8af", + "metadata": { + "tags": [] + }, + "outputs": [ + { + "data": { + "text/plain": [ + "Dataset({\n", + " features: ['path', 'audio', 'transcription', 'english_transcription', 'intent_class', 'lang_id'],\n", + " num_rows: 50\n", + "})" + ] + }, + "execution_count": 5, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "dataset" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "id": "3a11b5e5-0b1e-4be5-b8b6-7dbfffbd93de", + "metadata": { + "tags": [] + }, + "outputs": [], + "source": [ + "cfg = {\n", + " \"checks\": [{\n", + " 'type': uptrain.Monitor.OUTPUT_COMPARISON,\n", + " \"measurable_args\": {\n", + " 'type': uptrain.MeasurableType.PREDICTION,\n", + " },\n", + " \"comparison_model\": uptrain.ComparisonModel.FASTER_WHISPER,\n", + " \"comparison_metric\": uptrain.ComparisonMetric.ROGUE_L_F1,\n", + " \"comparison_model_input_args\": {\n", + " \"type\": uptrain.MeasurableType.INPUT_FEATURE,\n", + " \"feature_name\": \"audio_file\"\n", + " },\n", + " \"threshold\": 0.6\n", + " }],\n", + "\n", + " \"logging_args\": {\n", + " \"st_logging\": True\n", + " }\n", + "}" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "id": "6c88d529-c060-4f8a-9cea-d6a9648968aa", + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "objc[49409]: Class AVFFrameReceiver is implemented in both /Users/sourabhagrawal/miniconda3/envs/prod_dev2/lib/libavdevice.58.8.100.dylib (0x2a96dc798) and /Users/sourabhagrawal/miniconda3/envs/prod_dev2/lib/python3.10/site-packages/av/.dylibs/libavdevice.59.7.100.dylib (0x2abfe8778). One of the two will be used. Which one is undefined.\n", + "objc[49409]: Class AVFAudioReceiver is implemented in both /Users/sourabhagrawal/miniconda3/envs/prod_dev2/lib/libavdevice.58.8.100.dylib (0x2a96dc7e8) and /Users/sourabhagrawal/miniconda3/envs/prod_dev2/lib/python3.10/site-packages/av/.dylibs/libavdevice.59.7.100.dylib (0x2abfe87c8). One of the two will be used. Which one is undefined.\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Deleting the folder: uptrain_smart_data\n", + "Deleting the folder: uptrain_logs\n" + ] + }, + { + "ename": "TypeError", + "evalue": "function() argument 'code' must be code, not str", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mTypeError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[7], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m framework \u001b[38;5;241m=\u001b[39m \u001b[43muptrain\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mFramework\u001b[49m\u001b[43m(\u001b[49m\u001b[43mcfg_dict\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcfg\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 3\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m idx \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(\u001b[38;5;28mlen\u001b[39m(dataset)):\n\u001b[1;32m 4\u001b[0m inputs \u001b[38;5;241m=\u001b[39m {\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124maudio_file\u001b[39m\u001b[38;5;124m\"\u001b[39m: [dataset[idx][\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124maudio\u001b[39m\u001b[38;5;124m\"\u001b[39m][\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mpath\u001b[39m\u001b[38;5;124m\"\u001b[39m]]}\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/classes/framework.py:103\u001b[0m, in \u001b[0;36mFramework.__init__\u001b[0;34m(self, cfg_dict)\u001b[0m\n\u001b[1;32m 101\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdataset_handler \u001b[38;5;241m=\u001b[39m DatasetHandler(framework\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m, cfg\u001b[38;5;241m=\u001b[39mcfg)\n\u001b[1;32m 102\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmodel_handler \u001b[38;5;241m=\u001b[39m ModelHandler()\n\u001b[0;32m--> 103\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcheck_manager \u001b[38;5;241m=\u001b[39m \u001b[43mCheckManager\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mchecks\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 104\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mreset_retraining()\n\u001b[1;32m 106\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m training_args\u001b[38;5;241m.\u001b[39mdata_transformation_func:\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/classes/managers/check_manager.py:33\u001b[0m, in \u001b[0;36mCheckManager.__init__\u001b[0;34m(self, framework, checks)\u001b[0m\n\u001b[1;32m 31\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m check \u001b[38;5;129;01min\u001b[39;00m checks:\n\u001b[1;32m 32\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m check[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtype\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;129;01min\u001b[39;00m Monitor:\n\u001b[0;32m---> 33\u001b[0m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43madd_monitor\u001b[49m\u001b[43m(\u001b[49m\u001b[43mcheck\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 34\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m check[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtype\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;129;01min\u001b[39;00m Statistic:\n\u001b[1;32m 35\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39madd_statistic(check)\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/classes/managers/check_manager.py:96\u001b[0m, in \u001b[0;36mCheckManager.add_monitor\u001b[0;34m(self, check)\u001b[0m\n\u001b[1;32m 94\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmonitors_to_check\u001b[38;5;241m.\u001b[39mextend(integrity_managers)\n\u001b[1;32m 95\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m check[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtype\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;241m==\u001b[39m Monitor\u001b[38;5;241m.\u001b[39mOUTPUT_COMPARISON:\n\u001b[0;32m---> 96\u001b[0m comparison_monitor \u001b[38;5;241m=\u001b[39m \u001b[43mOutputComparison\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfw\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 97\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmonitors_to_check\u001b[38;5;241m.\u001b[39mappend(comparison_monitor)\n\u001b[1;32m 98\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/classes/monitors/abstract_check.py:33\u001b[0m, in \u001b[0;36mAbstractCheck.__init__\u001b[0;34m(self, fw, check)\u001b[0m\n\u001b[1;32m 31\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mchildren\u001b[38;5;241m.\u001b[39mappend(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__class__\u001b[39m(fw, check_copy))\n\u001b[1;32m 32\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m---> 33\u001b[0m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbase_init\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfw\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcheck\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 35\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m check\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfeat_slicing\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mFalse\u001b[39;00m):\n\u001b[1;32m 36\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpath_dashboard_data \u001b[38;5;241m=\u001b[39m os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39mjoin(fw\u001b[38;5;241m.\u001b[39mfold_name, \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdashboard_name\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m.csv\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/classes/monitors/output_comparison.py:13\u001b[0m, in \u001b[0;36mOutputComparison.base_init\u001b[0;34m(self, fw, check)\u001b[0m\n\u001b[1;32m 11\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mbase_init\u001b[39m(\u001b[38;5;28mself\u001b[39m, fw, check):\n\u001b[1;32m 12\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcomparison_model_base \u001b[38;5;241m=\u001b[39m check[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcomparison_model\u001b[39m\u001b[38;5;124m'\u001b[39m]\n\u001b[0;32m---> 13\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcomparison_model_resolved \u001b[38;5;241m=\u001b[39m \u001b[43mComparisonModelResolver\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mresolve\u001b[49m\u001b[43m(\u001b[49m\u001b[43mcheck\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mcomparison_model\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 14\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcomparison_model_inputs \u001b[38;5;241m=\u001b[39m MeasurableResolver(check\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcomparison_model_input_args\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;28;01mNone\u001b[39;00m))\u001b[38;5;241m.\u001b[39mresolve(fw)\n\u001b[1;32m 15\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mcomparison_metric_base \u001b[38;5;241m=\u001b[39m check[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcomparison_metric\u001b[39m\u001b[38;5;124m'\u001b[39m]\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/classes/monitors/output_comparison.py:66\u001b[0m, in \u001b[0;36mComparisonModelResolver.resolve\u001b[0;34m(self, model)\u001b[0m\n\u001b[1;32m 64\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mresolve\u001b[39m(\u001b[38;5;28mself\u001b[39m, model):\n\u001b[1;32m 65\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m model \u001b[38;5;241m==\u001b[39m ComparisonModel\u001b[38;5;241m.\u001b[39mFASTER_WHISPER:\n\u001b[0;32m---> 66\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01muptrain\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mee\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mlib\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01malgorithms\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m faster_whisper_speech_to_text\n\u001b[1;32m 67\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m faster_whisper_speech_to_text\n\u001b[1;32m 68\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/ee/lib/algorithms.py:15\u001b[0m\n\u001b[1;32m 11\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m:\n\u001b[1;32m 12\u001b[0m rouge \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m 14\u001b[0m \u001b[38;5;129;43m@dependency_required\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43mfaster_whisper\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mfaster_whisper\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[0;32m---> 15\u001b[0m \u001b[38;5;28;43;01mdef\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;21;43mfaster_whisper_speech_to_text\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43maudio_files\u001b[49m\u001b[43m)\u001b[49m\u001b[43m:\u001b[49m\n\u001b[1;32m 16\u001b[0m \u001b[43m \u001b[49m\u001b[43mmodel_size\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mlarge-v2\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\n\u001b[1;32m 17\u001b[0m \u001b[43m \u001b[49m\u001b[43mmodel\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m \u001b[49m\u001b[43mfaster_whisper\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mWhisperModel\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmodel_size\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdevice\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mcpu\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcompute_type\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mint8\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/Desktop/codes/dev/uptrain/uptrain/core/lib/helper_funcs.py:260\u001b[0m, in \u001b[0;36mdependency_required..class_decorator\u001b[0;34m(cls)\u001b[0m\n\u001b[1;32m 258\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mclass_decorator\u001b[39m(\u001b[38;5;28mcls\u001b[39m):\n\u001b[1;32m 259\u001b[0m \u001b[38;5;129m@functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(\u001b[38;5;28mcls\u001b[39m, updated\u001b[38;5;241m=\u001b[39m())\n\u001b[0;32m--> 260\u001b[0m \u001b[38;5;28;01mclass\u001b[39;00m \u001b[38;5;21;01mWrappedClass\u001b[39;00m(\u001b[38;5;28mcls\u001b[39m):\n\u001b[1;32m 261\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m__init__\u001b[39m(\u001b[38;5;28mself\u001b[39m, \u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[1;32m 262\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m dependency_name \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n", + "\u001b[0;31mTypeError\u001b[0m: function() argument 'code' must be code, not str" + ] + } + ], + "source": [ + "framework = uptrain.Framework(cfg_dict=cfg)\n", + "\n", + "for idx in range(len(dataset)):\n", + " inputs = {\"audio_file\": [dataset[idx][\"audio\"][\"path\"]]}\n", + " preds = [x['text'] for x in transcriber(inputs[\"audio_file\"])]\n", + " framework.log(inputs=inputs, outputs=preds)" + ] + }, + { + "attachments": { + "736a5a17-c1c9-4322-8a6b-1250a506c110.png": { + "image/png": "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" + } + }, + "cell_type": "markdown", + "id": "9e4092ba-125f-4d0a-b0bb-28410cdc9e56", + "metadata": {}, + "source": [ + "This is how the UpTrain dashboard looks like.\n", + "![Screenshot 2023-05-03 at 8.24.55 PM.png](attachment:736a5a17-c1c9-4322-8a6b-1250a506c110.png)\n", + "\n", + "As we can see we have several cases where our model disagrees with Whisper's outputs. Let's see what failure cases are collected by the UpTrain framework" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "3b069cca-8c08-4ffa-8b1a-039d92903584", + "metadata": { + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + " You can now view your Streamlit app in your browser.\n", + "\n", + " Local URL: http://localhost:8503\n", + " Network URL: http://192.168.151.48:8503\n", + "\n", + " For better performance, install the Watchdog module:\n", + "\n", + " $ xcode-select --install\n", + " $ pip install watchdog\n", + " \n" + ] + } + ], + "source": [ + "import pandas as pd\n", + "pd.read_csv(\"uptrain_smart_data/1/smart_data.csv\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "063ff5e7-1749-4a39-b5a4-7dcf0396a940", + "metadata": {}, + "outputs": [], + "source": [] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8d63c050-5ca2-4c6e-bd6f-47523908d86f", + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "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": "3.10.9" + }, + "vscode": { + "interpreter": { + "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6" + } + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/uptrain/__init__.py b/uptrain/__init__.py index 5c188058..11b5627c 100644 --- a/uptrain/__init__.py +++ b/uptrain/__init__.py @@ -11,5 +11,7 @@ PlotType, Statistic, Visual, + ComparisonMetric, + ComparisonModel ) from uptrain.core.encoders import UpTrainEncoder diff --git a/uptrain/constants.py b/uptrain/constants.py index 06645f96..6d9a837e 100644 --- a/uptrain/constants.py +++ b/uptrain/constants.py @@ -54,6 +54,7 @@ class Monitor(str, Enum): CONCEPT_DRIFT = "concept_drift" POPULARITY_BIAS = "popularity_bias" DATA_INTEGRITY = "data_integrity" + OUTPUT_COMPARISON = "output_comparison" class PlotType(str, Enum): @@ -76,3 +77,11 @@ class Visual(str, Enum): TSNE = "t-SNE" SHAP = "SHAP" PLOT = "PLOT" + + +class ComparisonModel(str, Enum): + FASTER_WHISPER = "faster_whisper" + + +class ComparisonMetric(str, Enum): + ROGUE_L_F1 = "rogue-l-f" \ No newline at end of file diff --git a/uptrain/core/classes/framework.py b/uptrain/core/classes/framework.py index be77425c..f54c5650 100644 --- a/uptrain/core/classes/framework.py +++ b/uptrain/core/classes/framework.py @@ -232,14 +232,15 @@ def check_and_add_data(self, inputs, outputs, gts=None, extra_args={}): } ) - if self.log_data: - # Log all the data-points into all_data warehouse - add_data_to_warehouse(deepcopy(data), self.path_all_data) - # Check for any monitors self.check(data, extra_args) self.predicted_count += self.batch_size + if self.log_data: + data.update(extra_args) + # Log all the data-points into all_data warehouse + add_data_to_warehouse(deepcopy(data), self.path_all_data) + # Smartly add data for retraining self.smartly_add_data(data, extra_args) self.extra_args = extra_args diff --git a/uptrain/core/classes/managers/check_manager.py b/uptrain/core/classes/managers/check_manager.py index f27680a1..d46b5e6b 100644 --- a/uptrain/core/classes/managers/check_manager.py +++ b/uptrain/core/classes/managers/check_manager.py @@ -11,6 +11,7 @@ ModelBias, DataIntegrity, EdgeCase, + OutputComparison ) from uptrain.core.classes.statistics import ( Distance, @@ -99,6 +100,9 @@ def add_monitor(self, check): ) integrity_managers.append(DataIntegrity(self.fw, check_copy)) self.monitors_to_check.extend(integrity_managers) + elif check["type"] == Monitor.OUTPUT_COMPARISON: + comparison_monitor = OutputComparison(self.fw, check) + self.monitors_to_check.append(comparison_monitor) else: raise Exception("Monitor type not Supported") diff --git a/uptrain/core/classes/measurables/measurable_resolver.py b/uptrain/core/classes/measurables/measurable_resolver.py index 15727da3..1accdf9c 100644 --- a/uptrain/core/classes/measurables/measurable_resolver.py +++ b/uptrain/core/classes/measurables/measurable_resolver.py @@ -1,5 +1,7 @@ from uptrain.core.classes.measurables import ( Measurable, + InputFeatureMeasurable, + OutputFeatureMeasurable, FeatureMeasurable, FeatureConcatMeasurable, ConditionMeasurable, @@ -35,11 +37,11 @@ def resolve(self, framework) -> Measurable: resolve_args = self._args measurable_type = resolve_args["type"] if measurable_type == MeasurableType.INPUT_FEATURE: - return FeatureMeasurable(framework, resolve_args["feature_name"], "inputs") + return InputFeatureMeasurable(framework, resolve_args["feature_name"]) elif measurable_type == MeasurableType.FEATURE_CONCAT: return FeatureConcatMeasurable(framework, resolve_args["feat_name_list"]) elif measurable_type == MeasurableType.PREDICTION: - return FeatureMeasurable(framework, resolve_args["feature_name"], "outputs") + return OutputFeatureMeasurable(framework) elif measurable_type == MeasurableType.CUSTOM: return CustomMeasurable(framework, resolve_args) elif measurable_type == MeasurableType.ACCURACY: @@ -51,13 +53,13 @@ def resolve(self, framework) -> Measurable: elif measurable_type == MeasurableType.CONDITION_ON_INPUT: return ConditionMeasurable( framework, - FeatureMeasurable(framework, resolve_args["feature_name"], "inputs"), + InputFeatureMeasurable(framework, resolve_args["feature_name"]), resolve_args["condition_args"], ) elif measurable_type == MeasurableType.CONDITION_ON_PREDICTION: return ConditionMeasurable( framework, - FeatureMeasurable(framework, resolve_args["feature_name"], "outputs"), + OutputFeatureMeasurable(framework), resolve_args["condition_args"], ) elif measurable_type == MeasurableType.SCALAR_FROM_EMBEDDING: diff --git a/uptrain/core/classes/measurables/output_feature.py b/uptrain/core/classes/measurables/output_feature.py index 54674913..9027f413 100644 --- a/uptrain/core/classes/measurables/output_feature.py +++ b/uptrain/core/classes/measurables/output_feature.py @@ -6,9 +6,8 @@ class OutputFeatureMeasurable(Measurable): """Class that returns the output feature corresponding to the feature name.""" - def __init__(self, framework, feature_name) -> None: + def __init__(self, framework) -> None: super().__init__(framework) - self.feature_name = feature_name def _compute(self, inputs=None, outputs=None, gts=None, extra=None) -> Any: return outputs diff --git a/uptrain/core/classes/monitors/__init__.py b/uptrain/core/classes/monitors/__init__.py index c6d5ba7e..4b315bcc 100644 --- a/uptrain/core/classes/monitors/__init__.py +++ b/uptrain/core/classes/monitors/__init__.py @@ -8,3 +8,4 @@ from .edge_case import EdgeCase from .model_bias import ModelBias from .data_integrity import DataIntegrity +from .output_comparison import OutputComparison diff --git a/uptrain/core/classes/monitors/data_integrity.py b/uptrain/core/classes/monitors/data_integrity.py index 12643031..5335651e 100644 --- a/uptrain/core/classes/monitors/data_integrity.py +++ b/uptrain/core/classes/monitors/data_integrity.py @@ -33,6 +33,8 @@ def base_check(self, inputs, outputs, gts=None, extra_args={}): has_issue = signal_value == None elif self.integrity_type == "less_than": has_issue = signal_value > self.threshold + elif self.integrity_type == "equal_to": + has_issue = signal_value == self.threshold elif self.integrity_type == "greater_than": has_issue = signal_value < self.threshold elif self.integrity_type == "minus_one": diff --git a/uptrain/core/classes/monitors/output_comparison.py b/uptrain/core/classes/monitors/output_comparison.py new file mode 100644 index 00000000..8ccc6ebb --- /dev/null +++ b/uptrain/core/classes/monitors/output_comparison.py @@ -0,0 +1,79 @@ +import numpy as np +from uptrain.core.classes.monitors import AbstractMonitor +from uptrain.core.classes.measurables import MeasurableResolver +from uptrain.constants import Monitor, ComparisonModel, ComparisonMetric + + +class OutputComparison(AbstractMonitor): + dashboard_name = "output_comparison" + monitor_type = Monitor.OUTPUT_COMPARISON + + def base_init(self, fw, check): + self.comparison_model_base = check['comparison_model'] + self.comparison_model_resolved = ComparisonModelResolver().resolve(check['comparison_model']) + self.comparison_model_inputs = MeasurableResolver(check.get("comparison_model_input_args", None)).resolve(fw) + self.comparison_metric_base = check['comparison_metric'] + self.comparison_metric_resolved = ComparisonMetricResolver().resolve(check['comparison_metric']) + self.threshold = check['threshold'] + self.count = 0 + + def base_check(self, inputs, outputs, gts=None, extra_args={}): + vals = self.measurable.compute_and_log( + inputs, outputs, gts=gts, extra=extra_args + ) + + comparison_model_inputs = self.comparison_model_inputs.compute_and_log( + inputs, outputs, gts=gts, extra=extra_args + ) + + comparison_model_outputs = self.comparison_model_resolved(comparison_model_inputs) + batch_metrics = self.comparison_metric_resolved(vals, comparison_model_outputs) + self.batch_metrics = batch_metrics + + extra_args.update({self.comparison_model_base + " outputs": comparison_model_outputs, self.comparison_metric_base: batch_metrics}) + + feat_name = self.comparison_metric_base + plot_name = f"{feat_name} Comparison - Production vs {self.comparison_model_base}" + self.count += len(extra_args['id']) + + self.log_handler.add_scalars( + plot_name, + {"y_" + feat_name: np.mean(batch_metrics)}, + self.count, + self.dashboard_name, + file_name=plot_name, + ) + + def need_ground_truth(self): + return False + + def base_is_data_interesting(self, inputs, outputs, gts=None, extra_args={}): + reasons = ["None"] * len(extra_args["id"]) + is_interesting = self.batch_metrics < self.threshold + reasons = [] + for idx in range(len(extra_args["id"])): + if is_interesting[idx] == 0: + reasons.append("None") + else: + reasons.append(f"Different output compared to {self.comparison_model_base}") + return is_interesting, reasons + + +class ComparisonModelResolver: + + def resolve(self, model): + if model == ComparisonModel.FASTER_WHISPER: + from uptrain.ee.lib.algorithms import faster_whisper_speech_to_text + return faster_whisper_speech_to_text + else: + raise Exception(f"{model} can't be resolved") + + +class ComparisonMetricResolver: + + def resolve(self, metric): + if metric == ComparisonMetric.ROGUE_L_F1: + from uptrain.ee.lib.algorithms import rogue_l_similarity + return rogue_l_similarity + else: + raise Exception(f"{metric} can't be resolved") diff --git a/uptrain/ee/lib/algorithms.py b/uptrain/ee/lib/algorithms.py new file mode 100644 index 00000000..a76eb31f --- /dev/null +++ b/uptrain/ee/lib/algorithms.py @@ -0,0 +1,31 @@ +import numpy as np +from uptrain.core.lib.helper_funcs import fn_dependency_required + +try: + import faster_whisper +except: + faster_whisper = None + +try: + import rouge +except: + rouge = None + +@fn_dependency_required(faster_whisper, "faster_whisper") +def faster_whisper_speech_to_text(audio_files): + model_size = "large-v2" + model = faster_whisper.WhisperModel(model_size, device="cpu", compute_type="int8") + prescribed_texts = [] + for audio_file in audio_files: + segments, _ = model.transcribe(audio_file, beam_size=5) + prescribed_text = '' + for segment in segments: + prescribed_text += segment.text + prescribed_texts.append(prescribed_text) + return prescribed_texts + +@fn_dependency_required(rouge, "rouge") +def rogue_l_similarity(text1_list, text2_list): + r = rouge.Rouge() + res = r.get_scores([x.lower() for x in text1_list],[x.lower() for x in text2_list]) + return np.array([x['rouge-l']['f'] for x in res])