id
int64
0
190k
prompt
stringlengths
21
13.4M
docstring
stringlengths
1
12k
0
from setuptools import find_packages, setup import os import subprocess import time def readme(): with open('README.md', encoding='utf-8') as f: content = f.read() return content
null
1
from setuptools import find_packages, setup import os import subprocess import time version_file = 'realesrgan/version.py' def get_hash(): if os.path.exists('.git'): sha = get_git_hash()[:7] else: sha = 'unknown' return sha def write_version_py(): content = """# GENERATED VERSION FILE # TIME: {} __version__ = '{}' __gitsha__ = '{}' version_info = ({}) """ sha = get_hash() with open('VERSION', 'r') as f: SHORT_VERSION = f.read().strip() VERSION_INFO = ', '.join([x if x.isdigit() else f'"{x}"' for x in SHORT_VERSION.split('.')]) version_file_str = content.format(time.asctime(), SHORT_VERSION, sha, VERSION_INFO) with open(version_file, 'w') as f: f.write(version_file_str)
null
2
from setuptools import find_packages, setup import os import subprocess import time version_file = 'realesrgan/version.py' def get_version(): with open(version_file, 'r') as f: exec(compile(f.read(), version_file, 'exec')) return locals()['__version__']
null
3
from setuptools import find_packages, setup import os import subprocess import time def get_requirements(filename='requirements.txt'): here = os.path.dirname(os.path.realpath(__file__)) with open(os.path.join(here, filename), 'r') as f: requires = [line.replace('\n', '') for line in f.readlines()] return requires
null
4
import argparse import cv2 import glob import mimetypes import numpy as np import os import shutil import subprocess import torch from basicsr.archs.rrdbnet_arch import RRDBNet from basicsr.utils.download_util import load_file_from_url from os import path as osp from tqdm import tqdm from realesrgan import RealESRGANer from realesrgan.archs.srvgg_arch import SRVGGNetCompact def get_video_meta_info(video_path): ret = {} probe = ffmpeg.probe(video_path) video_streams = [stream for stream in probe['streams'] if stream['codec_type'] == 'video'] has_audio = any(stream['codec_type'] == 'audio' for stream in probe['streams']) ret['width'] = video_streams[0]['width'] ret['height'] = video_streams[0]['height'] ret['fps'] = eval(video_streams[0]['avg_frame_rate']) ret['audio'] = ffmpeg.input(video_path).audio if has_audio else None ret['nb_frames'] = int(video_streams[0]['nb_frames']) return ret def get_sub_video(args, num_process, process_idx): if num_process == 1: return args.input meta = get_video_meta_info(args.input) duration = int(meta['nb_frames'] / meta['fps']) part_time = duration // num_process print(f'duration: {duration}, part_time: {part_time}') os.makedirs(osp.join(args.output, f'{args.video_name}_inp_tmp_videos'), exist_ok=True) out_path = osp.join(args.output, f'{args.video_name}_inp_tmp_videos', f'{process_idx:03d}.mp4') cmd = [ args.ffmpeg_bin, f'-i {args.input}', '-ss', f'{part_time * process_idx}', f'-to {part_time * (process_idx + 1)}' if process_idx != num_process - 1 else '', '-async 1', out_path, '-y' ] print(' '.join(cmd)) subprocess.call(' '.join(cmd), shell=True) return out_path
null
5
import argparse import cv2 import glob import mimetypes import numpy as np import os import shutil import subprocess import torch from basicsr.archs.rrdbnet_arch import RRDBNet from basicsr.utils.download_util import load_file_from_url from os import path as osp from tqdm import tqdm from realesrgan import RealESRGANer from realesrgan.archs.srvgg_arch import SRVGGNetCompact def inference_video(args, video_save_path, device=None, total_workers=1, worker_idx=0): def run(args): args.video_name = osp.splitext(os.path.basename(args.input))[0] video_save_path = osp.join(args.output, f'{args.video_name}_{args.suffix}.mp4') if args.extract_frame_first: tmp_frames_folder = osp.join(args.output, f'{args.video_name}_inp_tmp_frames') os.makedirs(tmp_frames_folder, exist_ok=True) os.system(f'ffmpeg -i {args.input} -qscale:v 1 -qmin 1 -qmax 1 -vsync 0 {tmp_frames_folder}/frame%08d.png') args.input = tmp_frames_folder num_gpus = torch.cuda.device_count() num_process = num_gpus * args.num_process_per_gpu if num_process == 1: inference_video(args, video_save_path) return ctx = torch.multiprocessing.get_context('spawn') pool = ctx.Pool(num_process) os.makedirs(osp.join(args.output, f'{args.video_name}_out_tmp_videos'), exist_ok=True) pbar = tqdm(total=num_process, unit='sub_video', desc='inference') for i in range(num_process): sub_video_save_path = osp.join(args.output, f'{args.video_name}_out_tmp_videos', f'{i:03d}.mp4') pool.apply_async( inference_video, args=(args, sub_video_save_path, torch.device(i % num_gpus), num_process, i), callback=lambda arg: pbar.update(1)) pool.close() pool.join() # combine sub videos # prepare vidlist.txt with open(f'{args.output}/{args.video_name}_vidlist.txt', 'w') as f: for i in range(num_process): f.write(f'file \'{args.video_name}_out_tmp_videos/{i:03d}.mp4\'\n') cmd = [ args.ffmpeg_bin, '-f', 'concat', '-safe', '0', '-i', f'{args.output}/{args.video_name}_vidlist.txt', '-c', 'copy', f'{video_save_path}' ] print(' '.join(cmd)) subprocess.call(cmd) shutil.rmtree(osp.join(args.output, f'{args.video_name}_out_tmp_videos')) if osp.exists(osp.join(args.output, f'{args.video_name}_inp_tmp_videos')): shutil.rmtree(osp.join(args.output, f'{args.video_name}_inp_tmp_videos')) os.remove(f'{args.output}/{args.video_name}_vidlist.txt')
null
6
import os os.system('pip install gfpgan') os.system('python setup.py develop') import cv2 import shutil import tempfile import torch from basicsr.archs.rrdbnet_arch import RRDBNet from basicsr.archs.srvgg_arch import SRVGGNetCompact from realesrgan.utils import RealESRGANer def clean_folder(folder): for filename in os.listdir(folder): file_path = os.path.join(folder, filename) try: if os.path.isfile(file_path) or os.path.islink(file_path): os.unlink(file_path) elif os.path.isdir(file_path): shutil.rmtree(file_path) except Exception as e: print(f'Failed to delete {file_path}. Reason: {e}')
null
7
import argparse import cv2 import numpy as np import os import sys from basicsr.utils import scandir from multiprocessing import Pool from os import path as osp from tqdm import tqdm def worker(path, opt): """Worker for each process. Args: path (str): Image path. opt (dict): Configuration dict. It contains: crop_size (int): Crop size. step (int): Step for overlapped sliding window. thresh_size (int): Threshold size. Patches whose size is lower than thresh_size will be dropped. save_folder (str): Path to save folder. compression_level (int): for cv2.IMWRITE_PNG_COMPRESSION. Returns: process_info (str): Process information displayed in progress bar. """ crop_size = opt['crop_size'] step = opt['step'] thresh_size = opt['thresh_size'] img_name, extension = osp.splitext(osp.basename(path)) # remove the x2, x3, x4 and x8 in the filename for DIV2K img_name = img_name.replace('x2', '').replace('x3', '').replace('x4', '').replace('x8', '') img = cv2.imread(path, cv2.IMREAD_UNCHANGED) h, w = img.shape[0:2] h_space = np.arange(0, h - crop_size + 1, step) if h - (h_space[-1] + crop_size) > thresh_size: h_space = np.append(h_space, h - crop_size) w_space = np.arange(0, w - crop_size + 1, step) if w - (w_space[-1] + crop_size) > thresh_size: w_space = np.append(w_space, w - crop_size) index = 0 for x in h_space: for y in w_space: index += 1 cropped_img = img[x:x + crop_size, y:y + crop_size, ...] cropped_img = np.ascontiguousarray(cropped_img) cv2.imwrite( osp.join(opt['save_folder'], f'{img_name}_s{index:03d}{extension}'), cropped_img, [cv2.IMWRITE_PNG_COMPRESSION, opt['compression_level']]) process_info = f'Processing {img_name} ...' return process_info The provided code snippet includes necessary dependencies for implementing the `extract_subimages` function. Write a Python function `def extract_subimages(opt)` to solve the following problem: Crop images to subimages. Args: opt (dict): Configuration dict. It contains: input_folder (str): Path to the input folder. save_folder (str): Path to save folder. n_thread (int): Thread number. Here is the function: def extract_subimages(opt): """Crop images to subimages. Args: opt (dict): Configuration dict. It contains: input_folder (str): Path to the input folder. save_folder (str): Path to save folder. n_thread (int): Thread number. """ input_folder = opt['input_folder'] save_folder = opt['save_folder'] if not osp.exists(save_folder): os.makedirs(save_folder) print(f'mkdir {save_folder} ...') else: print(f'Folder {save_folder} already exists. Exit.') sys.exit(1) # scan all images img_list = list(scandir(input_folder, full_path=True)) pbar = tqdm(total=len(img_list), unit='image', desc='Extract') pool = Pool(opt['n_thread']) for path in img_list: pool.apply_async(worker, args=(path, opt), callback=lambda arg: pbar.update(1)) pool.close() pool.join() pbar.close() print('All processes done.')
Crop images to subimages. Args: opt (dict): Configuration dict. It contains: input_folder (str): Path to the input folder. save_folder (str): Path to save folder. n_thread (int): Thread number.
8
import cv2 import numpy as np from PIL import Image def rotate_array(image: np.ndarray, angle: float) -> np.ndarray: if angle == 0: return image h, w = image.shape[:2] center = (w // 2, h // 2) M = cv2.getRotationMatrix2D(center, angle, 1.0) return cv2.warpAffine(image, M, (w, h)) def rotate_image(image: Image, angle: float) -> Image: if angle == 0: return image return Image.fromarray(rotate_array(np.array(image), angle))
null
9
import traceback from typing import Dict from scripts.io.util import load_classes_from_directory from scripts.use_cases.face_detector import FaceDetector from scripts.use_cases.face_processor import FaceProcessor from scripts.use_cases.mask_generator import MaskGenerator def create(all_classes, type: str) -> Dict: d = {} for cls in all_classes: try: c = cls() d[c.name().lower()] = c except Exception as e: print(traceback.format_exc()) print(f"Face Editor: {cls}, Error: {e}") return d def load_classes_from_directory(base_class: Type, installer: bool = False) -> List[Type]: if not installer: all_classes = load_classes_from_directory_(base_class, inferencers_dir, False) else: all_classes = [] for component in shared.opts.data.get("face_editor_additional_components", []): all_classes.extend( load_classes_from_directory_(base_class, os.path.join(inferencers_dir, component), installer) ) return all_classes class FaceDetector(ABC): def name(self) -> str: pass def detect_faces(self, image: Image, **kwargs) -> List[Rect]: pass def load_face_detector() -> Dict[str, FaceDetector]: return create(load_classes_from_directory(FaceDetector), "FaceDetector")
null
10
import traceback from typing import Dict from scripts.io.util import load_classes_from_directory from scripts.use_cases.face_detector import FaceDetector from scripts.use_cases.face_processor import FaceProcessor from scripts.use_cases.mask_generator import MaskGenerator def create(all_classes, type: str) -> Dict: def load_classes_from_directory(base_class: Type, installer: bool = False) -> List[Type]: class FaceProcessor(ABC): def name(self) -> str: def process(self, face: Face, p: StableDiffusionProcessingImg2Img, **kwargs) -> Image: def load_face_processor() -> Dict[str, FaceProcessor]: return create(load_classes_from_directory(FaceProcessor), "FaceProcessor")
null
11
import traceback from typing import Dict from scripts.io.util import load_classes_from_directory from scripts.use_cases.face_detector import FaceDetector from scripts.use_cases.face_processor import FaceProcessor from scripts.use_cases.mask_generator import MaskGenerator def create(all_classes, type: str) -> Dict: d = {} for cls in all_classes: try: c = cls() d[c.name().lower()] = c except Exception as e: print(traceback.format_exc()) print(f"Face Editor: {cls}, Error: {e}") return d def load_classes_from_directory(base_class: Type, installer: bool = False) -> List[Type]: if not installer: all_classes = load_classes_from_directory_(base_class, inferencers_dir, False) else: all_classes = [] for component in shared.opts.data.get("face_editor_additional_components", []): all_classes.extend( load_classes_from_directory_(base_class, os.path.join(inferencers_dir, component), installer) ) return all_classes class MaskGenerator(ABC): def name(self) -> str: pass def generate_mask( self, face_image: np.ndarray, face_area_on_image: Tuple[int, int, int, int], **kwargs, ) -> np.ndarray: pass def mask_non_face_areas(image: np.ndarray, face_area_on_image: Tuple[int, int, int, int]) -> np.ndarray: left, top, right, bottom = face_area_on_image image = image.copy() image[:top, :] = 0 image[bottom:, :] = 0 image[:, :left] = 0 image[:, right:] = 0 return image def calculate_mask_coverage(mask: np.ndarray): gray_mask = cv2.cvtColor(mask, cv2.COLOR_RGB2GRAY) non_black_pixels = np.count_nonzero(gray_mask) total_pixels = gray_mask.size return non_black_pixels / total_pixels def load_mask_generator() -> Dict[str, MaskGenerator]: return create(load_classes_from_directory(MaskGenerator), "MaskGenerator")
null
12
import operator from typing import Dict from lark import Lark, Tree def starts_with(a, b): return a.startswith(b)
null
13
import operator from typing import Dict from lark import Lark, Tree def ends_with(a, b): return a.endswith(b)
null
14
import operator from typing import Dict from lark import Lark, Tree def contains(a, b): return b in a
null
15
import operator from typing import Dict from lark import Lark, Tree def not_contains(a, b): return b not in a
null
16
import operator from typing import Dict from lark import Lark, Tree def evaluate(query: str, attributes: Dict[str, str]) -> bool: def validate(query: str): return evaluate(query, {})
null
17
from typing import Dict, List, Optional, Union from pydantic import BaseModel, root_validator, validator class Worker(BaseModel): name: str params: Optional[Dict] def default_params(cls, values): if "params" not in values or values["params"] is None: values["params"] = {} return values def lowercase_name(cls, v): return v.lower() def parse_worker_field(value: Union[str, Dict, Worker]) -> Worker: if isinstance(value, Dict): return Worker(**value) if isinstance(value, str): return Worker(name=value) return value
null
18
import os import gradio as gr from modules import script_callbacks, shared from scripts.entities.option import Option from scripts.io.util import inferencers_dir from scripts.ui import workflow_editor from scripts.ui.param_value_parser import ParamValueParser inferencers_dir = os.path.join(get_path("scripts", "inferencers")) def on_ui_settings(): section = ("face_editor", "Face Editor") shared.opts.add_option( "face_editor_search_subdirectories", shared.OptionInfo(False, "Search workflows in subdirectories", gr.Checkbox, section=section), ) additional_components = [] with os.scandir(inferencers_dir) as entries: for entry in entries: if entry.is_dir() and entry.name[0].isalnum(): additional_components.append(entry.name) shared.opts.add_option( "face_editor_additional_components", shared.OptionInfo( [], "Additional components", gr.CheckboxGroup, {"choices": additional_components}, section=section ), ) shared.opts.add_option( "face_editor_save_original_on_detection_fail", shared.OptionInfo(True, "Save original image if face detection fails", gr.Checkbox, section=section), ) shared.opts.add_option( "face_editor_correct_tilt", shared.OptionInfo(False, "Adjust tilt for detected faces", gr.Checkbox, section=section), ) shared.opts.add_option( "face_editor_auto_face_size_by_model", shared.OptionInfo(False, "Auto face size adjustment by model", gr.Checkbox, section=section), ) shared.opts.add_option( "face_editor_script_index", shared.OptionInfo( 99, "The position in postprocess at which this script will be executed; " "0 means it will be executed before any scripts, 99 means it will probably be executed last.", gr.Slider, {"minimum": 0, "maximum": 99, "step": 1}, section=section, ), )
null
19
import json import os from typing import Any, Dict, List import gradio as gr from modules import shared from pydantic import ValidationError from scripts.io.util import workflows_dir from scripts.use_cases.workflow_manager import WorkflowManager def load_workflow(file: str) -> str: if file is not None: filepath = os.path.join(workflows_dir, file + ".json") if os.path.isfile(filepath): return open(filepath).read() return "" def get_filename(file: str) -> str: if file == "default": return "" return file def sync_selection(file: str) -> str: return file def save_workflow(name: str, workflow: str) -> str: if name is None or len(name) == 0: return "" with open(os.path.join(workflows_dir, name + ".json"), "w") as file: file.write(workflow) return f"Saved to {name}.json" def get_files() -> List[str]: search_subdirectories = shared.opts.data.get("face_editor_search_subdirectories", False) files = [] for root, _, filenames in os.walk(workflows_dir): if not search_subdirectories and not os.path.samefile(root, workflows_dir): continue for filename in filenames: if filename.endswith(".json"): relative_path, _ = os.path.splitext(os.path.relpath(os.path.join(root, filename), workflows_dir)) files.append(relative_path) return files def refresh_files(workflow: str, file: str) -> dict: files = get_files() kwargs: Dict[str, Any] = {"choices": files} if workflow: for file in files: if load_workflow(file) == workflow: kwargs["value"] = file break return gr.update(**kwargs) def validate_workflow(workflow: str) -> str: try: json.loads(workflow) WorkflowManager.get(workflow) return "No errors found in the Workflow." except json.JSONDecodeError as e: return f"Error in JSON: {str(e)}" except ValidationError as e: errors = e.errors() if len(errors) == 0: return f"{str(e)}" err = errors[-1] return f"{' -> '.join(str(er) for er in err['loc'])} {err['msg']}\n--\n{str(e)}" except Exception as e: return f"{str(e)}" def build(workflow_selector: gr.Dropdown): with gr.Blocks(title="Workflow"): with gr.Row(): filename_dropdown = gr.Dropdown( choices=get_files(), label="Choose a Workflow", value="default", scale=2, min_width=400, show_label=False, ) refresh_button = gr.Button(value="🔄", scale=0, size="sm", elem_classes="tool") with gr.Row(): filename_input = gr.Textbox(scale=2, show_label=False, placeholder="Save as") save_button = gr.Button(value="💾", scale=0, size="sm", elem_classes="tool") workflow_editor = gr.Code(language="json", label="Workflow", value=load_workflow("default")) with gr.Row(): json_status = gr.Textbox(scale=2, show_label=False) validate_button = gr.Button(value="✅", scale=0, size="sm", elem_classes="tool") filename_dropdown.input(load_workflow, inputs=[filename_dropdown], outputs=[workflow_editor]) filename_dropdown.change(get_filename, inputs=[filename_dropdown], outputs=[filename_input]) filename_dropdown.input(sync_selection, inputs=[filename_dropdown], outputs=[workflow_selector]) workflow_selector.input(load_workflow, inputs=[workflow_selector], outputs=[workflow_editor]) workflow_selector.input(get_filename, inputs=[workflow_selector], outputs=[filename_input]) workflow_selector.input(sync_selection, inputs=[workflow_selector], outputs=[filename_dropdown]) save_button.click(validate_workflow, inputs=[workflow_editor], outputs=[json_status]) save_button.click(save_workflow, inputs=[filename_input, workflow_editor]) refresh_button.click(refresh_files, inputs=[workflow_editor, filename_dropdown], outputs=[filename_dropdown]) refresh_button.click(refresh_files, inputs=[workflow_editor, filename_dropdown], outputs=[workflow_selector]) validate_button.click(validate_workflow, inputs=[workflow_editor], outputs=[json_status]) return workflow_editor
null
20
import cv2 import numpy as np from modules.processing import StableDiffusionProcessingImg2Img from PIL import Image from scripts.entities.face import Face from scripts.use_cases.face_processor import FaceProcessor def color_generator(colors): while True: for color in colors: yield color
null
21
import importlib.util import inspect import os from typing import List, Type import modules.scripts as scripts from modules import shared def get_path(*p: str) -> str: dir = os.path.join(scripts.basedir(), *p) if not os.path.isdir(dir): dir = os.path.join(scripts.basedir(), "extensions", "sd-face-editor", *p) if not os.path.isdir(dir): raise RuntimeError(f"not found:{dir}") return dir
null
22
import seqio import t5.data from t5.data.glue_utils import get_glue_weight_mapping from t5.data.glue_utils import get_super_glue_weight_mapping from t5.data.glue_utils import get_super_glue_weight_mapping_sentinel import t5.data.tasks _GLUE_WEIGHT_MAPPING = get_glue_weight_mapping() _SUPER_GLUE_WEIGHT_MAPPING = get_super_glue_weight_mapping() def _dedupe(name): rate = None if name in _GLUE_WEIGHT_MAPPING: rate = _GLUE_WEIGHT_MAPPING[name] elif name in _SUPER_GLUE_WEIGHT_MAPPING: rate = _SUPER_GLUE_WEIGHT_MAPPING[name] if rate is None: return t5.data.rate_num_examples if "glue" in name and "rte" in name: rate *= 0.5 return rate
null
23
import seqio import t5.data from t5.data.glue_utils import get_glue_weight_mapping from t5.data.glue_utils import get_super_glue_weight_mapping from t5.data.glue_utils import get_super_glue_weight_mapping_sentinel import t5.data.tasks _GLUE_WEIGHT_MAPPING = get_glue_weight_mapping() _SUPER_GLUE_WEIGHT_MAPPING = get_super_glue_weight_mapping() def assign_weight_or_rate_num_examples(name): if name in _GLUE_WEIGHT_MAPPING: return _GLUE_WEIGHT_MAPPING[name] elif name in _SUPER_GLUE_WEIGHT_MAPPING: return _SUPER_GLUE_WEIGHT_MAPPING[name] else: return t5.data.rate_num_examples
null
24
import gin import seqio DEFAULT_SPM_PATH = "gs://t5-data/vocabs/cc_all.32000/sentencepiece.model" DEFAULT_EXTRA_IDS = 100 def get_default_vocabulary(): return seqio.SentencePieceVocabulary(DEFAULT_SPM_PATH, DEFAULT_EXTRA_IDS)
null
25
import gin import seqio The provided code snippet includes necessary dependencies for implementing the `rate_num_examples` function. Write a Python function `def rate_num_examples( task, maximum=None, temperature=1.0, scale=1.0, fallback_to_num_input_examples=True)` to solve the following problem: Mixing rate equal to the number of examples for the task. Here is the function: def rate_num_examples( task, maximum=None, temperature=1.0, scale=1.0, fallback_to_num_input_examples=True): """Mixing rate equal to the number of examples for the task.""" return seqio.mixing_rate_num_examples( task=task, maximum=maximum, scale=scale, temperature=temperature, fallback_to_num_input_examples=fallback_to_num_input_examples)
Mixing rate equal to the number of examples for the task.
26
import gin import seqio The provided code snippet includes necessary dependencies for implementing the `rate_unsupervised` function. Write a Python function `def rate_unsupervised(task, value=1e6)` to solve the following problem: Gin-configurable mixing rate for the unsupervised co-training task. Here is the function: def rate_unsupervised(task, value=1e6): """Gin-configurable mixing rate for the unsupervised co-training task.""" del task return value
Gin-configurable mixing rate for the unsupervised co-training task.
27
import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `lower_text` function. Write a Python function `def lower_text(string, **unused_kwargs)` to solve the following problem: Lowercases text. Here is the function: def lower_text(string, **unused_kwargs): """Lowercases text.""" return string.lower()
Lowercases text.
28
import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `qa` function. Write a Python function `def qa(answer, example=None, is_target=False)` to solve the following problem: Returns answer, or all answers if the full example is provided. Here is the function: def qa(answer, example=None, is_target=False): """Returns answer, or all answers if the full example is provided.""" if is_target: return [tf.compat.as_text(a) for a in example["answers"]] return answer
Returns answer, or all answers if the full example is provided.
29
import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `span_qa` function. Write a Python function `def span_qa(answer, example=None, is_target=False)` to solve the following problem: Returns answer, or a dict with answers and context if the example is provided. Here is the function: def span_qa(answer, example=None, is_target=False): """Returns answer, or a dict with answers and context if the example is provided.""" if is_target: return { "answers": [tf.compat.as_text(a) for a in example["answers"]], "context": tf.compat.as_text(example["context"]) } return answer
Returns answer, or a dict with answers and context if the example is provided.
30
import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `wsc_simple` function. Write a Python function `def wsc_simple(prediction, example=None, is_target=False)` to solve the following problem: Sees whether we predicted the referent or not. Here is the function: def wsc_simple(prediction, example=None, is_target=False): """Sees whether we predicted the referent or not.""" if is_target: return example["label"] determiners = { "a", "an", "few", "her", "his", "each", "every", "many", "much", "my", "our", "some", "that", "the", "their", "these", "this", "those", "which", "whose", "your" } def clean(s): """Ignore capitalization and determiners.""" s = tf.compat.as_text(s).strip().lower() return " ".join([w for w in s.split(" ") if w not in determiners]) prediction = clean(prediction) if not prediction: # We don't want an empty prediction to accidentally return 0 and spuriously # match the label. return -1 # We aren't using the label but rather using the extracted referent so that we # can see if the prediction is equivalent to the referent. referent = clean(example["targets_pretokenized"]) if ("'" in prediction) != ("'" in referent): # Make sure we don't mark cases where the prediction is "Bob" and the # referent is "Bob's hat" as predicting the referent. predicted_referent = False else: prediction_words = set(prediction.split(" ")) referent_words = set(referent.split(" ")) # Handle cases where the prediction is "fuzzy bunny" and the referent is # "bunny". predicted_referent = prediction_words.issubset( referent_words) or referent_words.issubset(prediction_words) return int(predicted_referent)
Sees whether we predicted the referent or not.
31
import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `rank_classification` function. Write a Python function `def rank_classification(score, example=None, is_target=False, passthrough_feature_keys=None)` to solve the following problem: A postprocessor for the `rank_classification` preprocessor and metric. Here is the function: def rank_classification(score, example=None, is_target=False, passthrough_feature_keys=None): """A postprocessor for the `rank_classification` preprocessor and metric.""" if is_target: outputs = [ tuple(example["idx"]), example["is_correct"], example.get("weight", 1.0), len(example["targets"]) ] if passthrough_feature_keys: for key in passthrough_feature_keys: outputs.append(example[key]) return tuple(outputs) else: return score
A postprocessor for the `rank_classification` preprocessor and metric.
32
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics GLUE_WEIGHT_MAPPING = { "glue_cola_v002": 8_551., "glue_sst2_v002": 67_349., "glue_mrpc_v002": 3_668., "glue_qqp_v002": 363_849., "glue_stsb_v002": 5_749., "glue_mnli_v002": 392_702., "glue_qnli_v002": 104_743., "glue_rte_v002": 2_490., "glue_mnli_mismatched_v002": 0., "glue_mnli_matched_v002": 0., "glue_ax_v002": 0., } def get_glue_weight_mapping(): return GLUE_WEIGHT_MAPPING
null
33
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics SUPER_GLUE_WEIGHT_MAPPING = { "dpr_v001_simple": 1_322., "super_glue_wsc_v102_simple_train": 259., "super_glue_wsc_v102_simple_eval": 0., "super_glue_boolq_v102": 9_427., "super_glue_cb_v102": 250., "super_glue_copa_v102": 400., "super_glue_multirc_v102": 27_243., "super_glue_record_v102": 138_854., "super_glue_rte_v102": 2_490., "super_glue_wic_v102": 5_428., "super_glue_axb_v102": 0., "super_glue_axg_v102": 0., } def get_super_glue_weight_mapping(): return SUPER_GLUE_WEIGHT_MAPPING
null
34
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics SUPER_GLUE_WEIGHT_MAPPING_SENTINEL = { "dpr_v001_simple_1_sentinel": 1_322., "super_glue_wsc_v102_simple_1_sentinel_train": 259., "super_glue_wsc_v102_simple_1_sentinel_eval": 0., "super_glue_boolq_v102_1_sentinel": 9_427., "super_glue_cb_v102_1_sentinel": 250., "super_glue_copa_v102_1_sentinel": 400., "super_glue_multirc_v102_1_sentinel": 27_243., "super_glue_record_v102_1_sentinel": 138_854., "super_glue_rte_v102_1_sentinel": 2_490., "super_glue_wic_v102_1_sentinel": 5_428., "super_glue_axb_v102_1_sentinel": 0., "super_glue_axg_v102_1_sentinel": 0., } def get_super_glue_weight_mapping_sentinel(): return SUPER_GLUE_WEIGHT_MAPPING_SENTINEL
null
35
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics The provided code snippet includes necessary dependencies for implementing the `get_glue_text_preprocessor` function. Write a Python function `def get_glue_text_preprocessor(builder_config)` to solve the following problem: Return the glue preprocessor. Args: builder_config: a BuilderConfig Returns: a preprocessor function Here is the function: def get_glue_text_preprocessor(builder_config): """Return the glue preprocessor. Args: builder_config: a BuilderConfig Returns: a preprocessor function """ # stsb uses a floating point target, so use special preprocessor if builder_config.name == "stsb": return preprocessors.stsb elif builder_config.name == "wsc.fixed": return preprocessors.wsc elif builder_config.name == "record": return preprocessors.record else: if "mnli" in builder_config.name or builder_config.name == "ax": # Cast the GLUE diagnostic task as MNLI. benchmark_name = "mnli" elif builder_config.name in ["axb", "axg"]: # Cast the SuperGLUE diagnostic tasks as RTE. benchmark_name = "rte" else: benchmark_name = builder_config.name if builder_config.name == "multirc": feature_names = ("question", "answer", "paragraph") elif builder_config.name == "wic": # This ignores the start/end indices which show where in each sentence the # word appears. # TODO(craffel): Investigate using those indices. feature_names = ("sentence1", "sentence2", "word") else: feature_names = None return functools.partial( preprocessors.glue, benchmark_name=benchmark_name, label_names=builder_config.label_classes, feature_names=feature_names)
Return the glue preprocessor. Args: builder_config: a BuilderConfig Returns: a preprocessor function
36
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics def get_glue_postprocess_fn(builder_config): if builder_config.name == "stsb": return postprocessors.string_to_float elif builder_config.name == "multirc": return postprocessors.multirc elif builder_config.name == "record": return postprocessors.record else: return functools.partial( postprocessors.string_label_to_class_id, label_classes=builder_config.label_classes, )
null
37
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics GLUE_METRICS = collections.OrderedDict([ ("cola", [metrics.sklearn_metrics_wrapper( "matthews_corrcoef", metric_post_process_fn=lambda x: 100 * x)]), ("sst2", [metrics.accuracy]), ("mrpc", [metrics.f1_score_with_invalid, metrics.accuracy]), ("stsb", [metrics.pearson_corrcoef, metrics.spearman_corrcoef]), ("qqp", [metrics.f1_score_with_invalid, metrics.accuracy]), ("mnli", [metrics.accuracy]), ("mnli_matched", [metrics.accuracy]), ("mnli_mismatched", [metrics.accuracy]), ("qnli", [metrics.accuracy]), ("rte", [metrics.accuracy]), ("wnli", [metrics.accuracy]), ("ax", []), # Only test set available. ]) def get_glue_metric(task_name): return GLUE_METRICS[task_name]
null
38
import collections import functools from t5.data import postprocessors from t5.data import preprocessors from t5.evaluation import metrics SUPERGLUE_METRICS = collections.OrderedDict([ ("boolq", [metrics.accuracy]), ("cb", [metrics.mean_multiclass_f1(num_classes=3), metrics.accuracy]), ("copa", [metrics.accuracy]), ("multirc", [ metrics.multirc_f1_over_all_answers, metrics.mean_group_metric(metrics.all_match) ]), ("record", [metrics.deduplicate_metric(metrics.squad)]), ("rte", [metrics.accuracy]), ("wic", [metrics.accuracy]), ("axb", []), # Only test set available. ("axg", []), # Only test set available. ]) def get_super_glue_metric(task_name): return SUPERGLUE_METRICS[task_name]
null
39
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `summarize` function. Write a Python function `def summarize(x, article_key, summary_key)` to solve the following problem: Convert a summarization dataset to a text2text pair. For example, say the dataset returns examples of this format: {'article': <article>, 'highlights': <summary>} If article_key = 'article', summary_key = 'highlights', then the outputs will have the format: {'inputs': 'summarize': <article>, 'targets': <summary>} Args: x: an example to process. article_key: the feature key for the article to summarize. summary_key: the feature key for the target summary. Returns: A preprocessed example with the format listed above. Here is the function: def summarize(x, article_key, summary_key): """Convert a summarization dataset to a text2text pair. For example, say the dataset returns examples of this format: {'article': <article>, 'highlights': <summary>} If article_key = 'article', summary_key = 'highlights', then the outputs will have the format: {'inputs': 'summarize': <article>, 'targets': <summary>} Args: x: an example to process. article_key: the feature key for the article to summarize. summary_key: the feature key for the target summary. Returns: A preprocessed example with the format listed above. """ strs_to_join = ['summarize:', x[article_key]] return { 'inputs': tf.strings.join(strs_to_join, separator=' '), 'targets': x[summary_key], }
Convert a summarization dataset to a text2text pair. For example, say the dataset returns examples of this format: {'article': <article>, 'highlights': <summary>} If article_key = 'article', summary_key = 'highlights', then the outputs will have the format: {'inputs': 'summarize': <article>, 'targets': <summary>} Args: x: an example to process. article_key: the feature key for the article to summarize. summary_key: the feature key for the target summary. Returns: A preprocessed example with the format listed above.
40
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf NON_SPACED_LANGUAGE_RANGES = ( '\u1000-\u104f', # Burmese '\u4e00-\u9fff', # CJK Unified Ideographs '\u3400-\u4dbf', # CJK Unified Ideographs Extension A '\uf900-\ufaff', # CJK Compatibility Ideographs '\u2e80-\u2eff', # CJK Radicals Supplement '\u31c0-\u31ef', # CJK Strokes '\u3000-\u303f', # CJK Symbols and Punctuation '\u3040-\u309f', # Japanese Hiragana '\u30a0-\u30ff', # Japanese Katakana '\ua980-\ua9df', # Javanese '\u1780-\u17ff', # Khmer '\u19e0-\u19ff', # Khmer Symbols '\u0e80-\u0eff', # Lao '\u1980-\u19df', # Tai Lue '\u1a20-\u1aaf', # Tai Tham '\u0e00-\u0e7f', # Thai '\u0f00-\u0fff', # Tibetan ) The provided code snippet includes necessary dependencies for implementing the `pad_nonspaced_languages` function. Write a Python function `def pad_nonspaced_languages(x, text_key='text')` to solve the following problem: Pad non-spaced languages with spaces around each character. Args: x: an example to process. text_key: a string, the key for the text feature to preprocess in the dataset examples. Returns: A preprocessed example. Here is the function: def pad_nonspaced_languages(x, text_key='text'): """Pad non-spaced languages with spaces around each character. Args: x: an example to process. text_key: a string, the key for the text feature to preprocess in the dataset examples. Returns: A preprocessed example. """ res = dict(x) text = res[text_key] # Add spaces around any character from a non-spaced language. pattern = ''.join(NON_SPACED_LANGUAGE_RANGES) text = tf.strings.regex_replace(text, u'([{}])'.format(pattern), r' \1 ') # Collapse consecutive whitespace into one space. text = tf.strings.regex_replace(text, r'\s+', ' ') res[text_key] = text return res
Pad non-spaced languages with spaces around each character. Args: x: an example to process. text_key: a string, the key for the text feature to preprocess in the dataset examples. Returns: A preprocessed example.
41
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf AUTOTUNE = tf.data.experimental.AUTOTUNE def _pad_punctuation(text): """Adds spaces around punctuation.""" # Add space around punctuation. text = tf.strings.regex_replace(text, r'([[:punct:]])', r' \1 ') # Collapse consecutive whitespace into one space. text = tf.strings.regex_replace(text, r'\s+', ' ') return text def _string_join(lst): # Join on space, but collapse consecutive spaces. out = tf.strings.join(lst, separator=' ') return tf.strings.regex_replace(out, r'\s+', ' ') The provided code snippet includes necessary dependencies for implementing the `trivia_qa` function. Write a Python function `def trivia_qa(dataset)` to solve the following problem: Convert a TriviaQA example to multiple flattened examples. TriviaQA produces examples with this form: {'entity_pages': {dict of wiki entities}, 'search_results': <dict of web search results>, 'answer': {dict of all answers}, 'question': <question>, 'question_id': <question_id>, 'question_source': <question_source>} This function will return flattend examples of the format: {'inputs': 'question: <question> context: <article>' 'targets': 'answer: <sampled answer>'} Args: dataset: a tf.data.Dataset to process. Returns: A preprocessed tf.data.Dataset with the format listed above. Here is the function: def trivia_qa(dataset): """Convert a TriviaQA example to multiple flattened examples. TriviaQA produces examples with this form: {'entity_pages': {dict of wiki entities}, 'search_results': <dict of web search results>, 'answer': {dict of all answers}, 'question': <question>, 'question_id': <question_id>, 'question_source': <question_source>} This function will return flattend examples of the format: {'inputs': 'question: <question> context: <article>' 'targets': 'answer: <sampled answer>'} Args: dataset: a tf.data.Dataset to process. Returns: A preprocessed tf.data.Dataset with the format listed above. """ def triviaqa_question_answer_context(x): """Extracts matched contexts and answers. Returns all matched (question-context, answer) pairs. Args: x: A tfds sample. Returns: Flattened samples: (question-context, answer). """ contexts = [] if 'entity_pages' in x: contexts.append(x['entity_pages']['wiki_context']) if 'search_results' in x: contexts.append(x['search_results']['search_context']) contexts = tf.concat(contexts, 0) q = _pad_punctuation(x['question']) answers = x['answer']['normalized_aliases'] combination_size = tf.size(answers)*tf.size(contexts) find_answers = tf.TensorArray( tf.bool, size=combination_size, dynamic_size=True) selected_answers = tf.TensorArray( tf.string, size=combination_size, dynamic_size=True) join_q_c = tf.TensorArray( tf.string, size=combination_size, dynamic_size=True) def cond_fn(i, find_answers, selected_answers, join_q_c): del find_answers, selected_answers, join_q_c # Unused return tf.less(i, combination_size) def body_fn(i, find_answers, selected_answers, join_q_c): """Find answers from contexts and join.""" context_idx = tf.math.floordiv(i, tf.size(answers)) answer_idx = tf.math.mod(i, tf.size(answers)) a = _pad_punctuation(answers[answer_idx]) a_ = tf.strings.join(['.*', a, '.*']) c = _pad_punctuation(contexts[context_idx]) find_a = tf.strings.regex_full_match( tf.strings.lower(c), tf.strings.lower(a_)) find_answers = find_answers.write(i, find_a) selected_answers = selected_answers.write(i, a) join_q_c_str = _string_join(['question:', q, 'context:', c]) join_q_c = join_q_c.write(i, join_q_c_str) return (i + 1, find_answers, selected_answers, join_q_c) _, find_answers, selected_answers, join_q_c = tf.while_loop( cond_fn, body_fn, loop_vars=[ tf.constant(0), find_answers, selected_answers, join_q_c ]) find_answers = find_answers.stack() selected_answers = selected_answers.stack() join_q_c = join_q_c.stack() selected_answers = tf.boolean_mask(selected_answers, find_answers) selected_join_q_c = tf.boolean_mask(join_q_c, find_answers) return selected_join_q_c, selected_answers def my_fn(x): """Create TriviaQA example.""" join_q_c, a = triviaqa_question_answer_context(x) return { 'inputs': join_q_c, 'targets': a } dataset = dataset.map(my_fn, num_parallel_calls=AUTOTUNE) return dataset.unbatch()
Convert a TriviaQA example to multiple flattened examples. TriviaQA produces examples with this form: {'entity_pages': {dict of wiki entities}, 'search_results': <dict of web search results>, 'answer': {dict of all answers}, 'question': <question>, 'question_id': <question_id>, 'question_source': <question_source>} This function will return flattend examples of the format: {'inputs': 'question: <question> context: <article>' 'targets': 'answer: <sampled answer>'} Args: dataset: a tf.data.Dataset to process. Returns: A preprocessed tf.data.Dataset with the format listed above.
42
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf AUTOTUNE = tf.data.experimental.AUTOTUNE def squad(x, include_context=True): """Convert SQuAD examples to a text2text pair. SQuAD produces examples with this form: {'id': <id>, context': <article>, 'question': <question>, 'answers': { 'text': [<n answers>] }} This function will return examples of the format: {'inputs': 'question: <question> context: <article>', 'targets': '<answer_0>', 'id': <id>, 'question': <question>, 'context': <context>, 'answers': [<n answers>]}, Args: x: an example to process. include_context: a boolean Returns: A preprocessed example with the format listed above. """ a = _pad_punctuation(x['answers']['text']) q = _pad_punctuation(x['question']) c = _pad_punctuation(x['context']) if include_context: inputs = _string_join(['question:', q, 'context:', c]) else: inputs = _string_join(['squad trivia question:', q]) return { 'inputs': inputs, 'targets': a[0], 'id': x['id'], 'context': c, 'question': q, 'answers': a } def _span_answer(context, answer_text): """Finds start/end indices of answer_text in context after space tokenization. If answer_tokens is not a sublist of context_tokens, returns empty string. Args: context: 0-d string tensor answer_text: 0-d string Returns: A string tensor. """ def space_tok(s): """Replace non-word chars with space then split on space.""" s = tf.strings.regex_replace(s, r'\W', ' ') return tf.strings.split(input=[s], sep=' ').values def find_subseq(n, h): """Finds index of needle subsequence inside haystack. Args: n: 1-d tensor h: 1-d tensor same type as n Returns: Index of start of n if found found; otherwise -1. """ l_n = tf.size(n) l_h = tf.size(h) found = -1 for i in tf.range(0, l_h - l_n): if tf.reduce_all(tf.equal(h[i:i+l_n], n)): found = i break return found answer_tokens = space_tok(answer_text) context_tokens = space_tok(context) start = find_subseq(answer_tokens, context_tokens) end = start + tf.size(answer_tokens) - 1 # Just take the first candidate that matches exactly. if tf.equal(start, -1): return '' return tf.strings.format('start: {} end: {}', [start, end]) The provided code snippet includes necessary dependencies for implementing the `squad_span_space_tokenized` function. Write a Python function `def squad_span_space_tokenized(dataset)` to solve the following problem: Convert SQuAD examples to a text2text pair with span output. SQuAD produces examples with this form: {'context': <article>, 'question': <question>, 'answers': { 'text': [<all answers>] }} This function returns examples with the format {'inputs': 'context: <article> question: <question>', 'targets': 'start: <start_index> end: <end_index>'} where <start_index> and <end_index> specify the space-tokenized span start/end indices. Both <start_index> and <end_index> are included in the answer. In the case where the tokenized answer is not found in the tokenized context, the example is skipped. Args: dataset: a tf.data.Dataset to process. Returns: A preprocessed tf.data.Dataset with the format listed above. Here is the function: def squad_span_space_tokenized(dataset): """Convert SQuAD examples to a text2text pair with span output. SQuAD produces examples with this form: {'context': <article>, 'question': <question>, 'answers': { 'text': [<all answers>] }} This function returns examples with the format {'inputs': 'context: <article> question: <question>', 'targets': 'start: <start_index> end: <end_index>'} where <start_index> and <end_index> specify the space-tokenized span start/end indices. Both <start_index> and <end_index> are included in the answer. In the case where the tokenized answer is not found in the tokenized context, the example is skipped. Args: dataset: a tf.data.Dataset to process. Returns: A preprocessed tf.data.Dataset with the format listed above. """ def my_fn(x): """Create squad example as in squad_span_char, but tokenized on spaces.""" res = dict(x) res['targets'] = _span_answer(x['context'], x['targets']) return res dataset = squad(dataset) dataset = dataset.map(my_fn, num_parallel_calls=AUTOTUNE) return dataset.filter(lambda x: tf.strings.length(x['targets']) > 0)
Convert SQuAD examples to a text2text pair with span output. SQuAD produces examples with this form: {'context': <article>, 'question': <question>, 'answers': { 'text': [<all answers>] }} This function returns examples with the format {'inputs': 'context: <article> question: <question>', 'targets': 'start: <start_index> end: <end_index>'} where <start_index> and <end_index> specify the space-tokenized span start/end indices. Both <start_index> and <end_index> are included in the answer. In the case where the tokenized answer is not found in the tokenized context, the example is skipped. Args: dataset: a tf.data.Dataset to process. Returns: A preprocessed tf.data.Dataset with the format listed above.
43
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `random_split_text` function. Write a Python function `def random_split_text(dataset, text_key='text', min_words_per_segment=16, max_words_per_segment=512, max_words_total=8192)` to solve the following problem: Randomly split single-string examples into multiple examples each. Segment lengths are chosen according to a log-uniform distribution. Each incoming string is chopped into multiple equal-length examples with the last one possibly being shorter. If the input string is longer than max_words_total, then we use one random chunk and discard the rest. This may help with model stability. The intended use case is to break up long text examples for use in unsupervised transfer-learning. We don't really want to use this preprocessor for any dataset which has a well-defined evaluation procedure. If apply this preprocessor e.g. in an MT component, then the evaluation job will randomly split text when evaluating and the BLEU will get funky. Args: dataset: a tf.data.Dataset with dictionaries containing the key text_key text_key: a string min_words_per_segment: an integer max_words_per_segment: an integer max_words_total: an integer Returns: a dataset Here is the function: def random_split_text(dataset, text_key='text', min_words_per_segment=16, max_words_per_segment=512, max_words_total=8192): """Randomly split single-string examples into multiple examples each. Segment lengths are chosen according to a log-uniform distribution. Each incoming string is chopped into multiple equal-length examples with the last one possibly being shorter. If the input string is longer than max_words_total, then we use one random chunk and discard the rest. This may help with model stability. The intended use case is to break up long text examples for use in unsupervised transfer-learning. We don't really want to use this preprocessor for any dataset which has a well-defined evaluation procedure. If apply this preprocessor e.g. in an MT component, then the evaluation job will randomly split text when evaluating and the BLEU will get funky. Args: dataset: a tf.data.Dataset with dictionaries containing the key text_key text_key: a string min_words_per_segment: an integer max_words_per_segment: an integer max_words_total: an integer Returns: a dataset """ def random_chunk(x, chunk_size, seed): """Pick a random chunk of a 1d Tensor. The tensor is divided into chunks of length chunk_size, with the last chunk being potentially smaller. A random chunk is returned. Args: x: a 1d tf.Tensor. chunk_size: an integer. seed: int32 [2]-Tensor, the random seed. Returns: a 1d tf.Tensor with length <= chunk_size. """ size = tf.size(x) num_chunks = tf.maximum(1, (size - 1) // chunk_size + 1) chunk_num = tf.random.stateless_uniform( [], seed=seed, minval=0, maxval=num_chunks, dtype=tf.int32) return x[chunk_size * chunk_num:chunk_size * (chunk_num + 1)] @seqio.map_over_dataset(num_seeds=2) def my_fn(x, seeds): """Split one string into multiple strings. Args: x: a feature dictionary seeds: an int32 Tensor, shaped (2, 2), the random seeds. Returns: a feature dictionary """ text = x[text_key] words = tf.strings.split([text]).values if max_words_total: words = random_chunk(words, max_words_total, seed=seeds[0]) n_words = tf.size(words) # first pick a length (number of words per segment) length = tf.cast( tf.exp( tf.random.stateless_uniform( [], minval=math.log(min_words_per_segment), maxval=math.log(max_words_per_segment), seed=seeds[1], ) ), tf.int32) # Pad to a multiple of length, then use tf.reshape to split up the words # into num_segments segments each of the given length. num_segments = tf.cast( tf.math.ceil( tf.cast(n_words, tf.float32) / tf.cast(length, tf.float32) ), tf.int32) padding = num_segments * length - n_words words = tf.pad(words, [[0, padding]]) words = tf.reshape(words, [-1, length]) # Finally, join with spaces and strip. The padding turns into a bunch of # spaces that get stripped out. words = tf.strings.reduce_join(words, axis=1, separator=' ') return {text_key: tf.strings.strip(words)} return my_fn(dataset).unbatch()
Randomly split single-string examples into multiple examples each. Segment lengths are chosen according to a log-uniform distribution. Each incoming string is chopped into multiple equal-length examples with the last one possibly being shorter. If the input string is longer than max_words_total, then we use one random chunk and discard the rest. This may help with model stability. The intended use case is to break up long text examples for use in unsupervised transfer-learning. We don't really want to use this preprocessor for any dataset which has a well-defined evaluation procedure. If apply this preprocessor e.g. in an MT component, then the evaluation job will randomly split text when evaluating and the BLEU will get funky. Args: dataset: a tf.data.Dataset with dictionaries containing the key text_key text_key: a string min_words_per_segment: an integer max_words_per_segment: an integer max_words_total: an integer Returns: a dataset
44
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def split_text_to_words(dataset, text_key='text', min_num_words=2): """Split text to words and filter out examples with too few words.""" def split(x): res = dict(x) res['words'] = tf.strings.split([x[text_key]]).values return res dataset = dataset.map(split, num_parallel_calls=AUTOTUNE) return dataset.filter(lambda x: tf.size(x['words']) >= min_num_words) The provided code snippet includes necessary dependencies for implementing the `fill_in_the_blank` function. Write a Python function `def fill_in_the_blank(dataset, text_key='text', label='fill: ')` to solve the following problem: Create a dataset consisting of fill-in-the-blank text examples. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'inputs' and 'targets'. The input string is split on whitespace to form a sequence of words. This sequence is chopped randomly into segments of one or more words. Alternate segments are included in the inputs and targets, with a special word 'X' marking a missing segment. The given label is prepended to the inputs. Each input string produces two examples - one the inverse of the other. Inputs with less than two words are dropped. EXAMPLE: input: { 'text': 'The fat cat sat on the mat.' } outputs: { 'inputs': 'fill: The fat X the X' 'targets': 'X cat sat on X mat.' } { 'inputs': 'fill: X cat sat on X mat.' 'targets': 'The fat X the X' } Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. label: a string, the label to prepend to the inputs. Returns: a tf.data.Dataset Here is the function: def fill_in_the_blank(dataset, text_key='text', label='fill: '): """Create a dataset consisting of fill-in-the-blank text examples. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'inputs' and 'targets'. The input string is split on whitespace to form a sequence of words. This sequence is chopped randomly into segments of one or more words. Alternate segments are included in the inputs and targets, with a special word 'X' marking a missing segment. The given label is prepended to the inputs. Each input string produces two examples - one the inverse of the other. Inputs with less than two words are dropped. EXAMPLE: input: { 'text': 'The fat cat sat on the mat.' } outputs: { 'inputs': 'fill: The fat X the X' 'targets': 'X cat sat on X mat.' } { 'inputs': 'fill: X cat sat on X mat.' 'targets': 'The fat X the X' } Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. label: a string, the label to prepend to the inputs. Returns: a tf.data.Dataset """ @seqio.map_over_dataset(num_seeds=3) def my_fn(x, seeds): """Generates two preprocessed examples that are roughly inverses. Args: x: an example dict with text pre-split in `words` feature. seeds: an int32 Tensor, shaped (3, 2), the random seeds. Returns: an example dict with two inputs and two targets, one for each resulting preprocessed example. """ words = x['words'] n_words = tf.size(words) # First select the break probability. We pick this on a log-uniform # distribution between 1/(n_words + 1) and 1/2. This means that some # sequences will be chopped roughly and others finely. min_log_p_break = -tf.math.log(tf.cast(n_words, tf.float32) + 2.0) max_log_p_break = -tf.math.log(2.0) p_break = tf.exp( tf.random.stateless_uniform( [], minval=min_log_p_break, maxval=max_log_p_break, seed=seeds[0]) ) # craffel@ says that there may be bugs in random.uniform making it not # really uniform. This doesn't seem horribly important here, but may # need another look. breaks = tf.less( tf.random.stateless_uniform([n_words - 1], seed=seeds[1]), p_break) def one_random_break(): pos = tf.random.stateless_uniform( [], minval=0, maxval=n_words - 1, dtype=tf.int32, seed=seeds[2]) return tf.one_hot(pos, n_words - 1, dtype=tf.bool, on_value=True, off_value=False) breaks = tf.cond( tf.math.reduce_any(breaks), lambda: breaks, one_random_break) breaks = tf.concat([[True], breaks], axis=0) word_to_seq_id = tf.math.mod(tf.math.cumsum(tf.cast(breaks, tf.int32)), 2) # separators: # if in your segment: ' ' # if break to other segment: ' X' # else: '' results = [] for seq_id in [0, 1]: in_my_seq = tf.equal(word_to_seq_id, seq_id) separator_strings = tf.where( in_my_seq, ' ', tf.where(breaks, ' X', '') ) word_strings = tf.where(in_my_seq, words, '') all_strings = tf.stack([separator_strings, word_strings], axis=1) results.append(tf.strings.substr( tf.strings.reduce_join(all_strings), 1, tf.int32.max)) inputs = tf.stack([tf.strings.join([label, results[0]]), tf.strings.join([label, results[1]])]) targets = tf.stack([results[1], results[0]]) return {'inputs': inputs, 'targets': targets} dataset = split_text_to_words(dataset, text_key, min_num_words=2) return my_fn(dataset).unbatch()
Create a dataset consisting of fill-in-the-blank text examples. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'inputs' and 'targets'. The input string is split on whitespace to form a sequence of words. This sequence is chopped randomly into segments of one or more words. Alternate segments are included in the inputs and targets, with a special word 'X' marking a missing segment. The given label is prepended to the inputs. Each input string produces two examples - one the inverse of the other. Inputs with less than two words are dropped. EXAMPLE: input: { 'text': 'The fat cat sat on the mat.' } outputs: { 'inputs': 'fill: The fat X the X' 'targets': 'X cat sat on X mat.' } { 'inputs': 'fill: X cat sat on X mat.' 'targets': 'The fat X the X' } Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. label: a string, the label to prepend to the inputs. Returns: a tf.data.Dataset
45
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def split_text_to_words(dataset, text_key='text', min_num_words=2): """Split text to words and filter out examples with too few words.""" def split(x): res = dict(x) res['words'] = tf.strings.split([x[text_key]]).values return res dataset = dataset.map(split, num_parallel_calls=AUTOTUNE) return dataset.filter(lambda x: tf.size(x['words']) >= min_num_words) The provided code snippet includes necessary dependencies for implementing the `fill_in_the_blank_sized` function. Write a Python function `def fill_in_the_blank_sized( dataset, size_bins=(1, 2, 4, 8, 16, 32, 64, 128, 256, 512), text_key='text', label='fill: ')` to solve the following problem: Fill in the blank preprocessor that labels blank with a binned size. The actual blank size is sampled uniformly from the inclusive range of the min and max bin. The blank is then filled in with the closest bin size to the actual blank size. Args: dataset: a tf.data.Dataset, the dataset to preprocess. size_bins: a list, a list of blank sizes to select from when labelling the blank. text_key: a string, the key for the text feature to preprocess in the dataset examples. label: a string, the label to prepend to the inputs. Returns: a tf.data.Dataset Here is the function: def fill_in_the_blank_sized( dataset, size_bins=(1, 2, 4, 8, 16, 32, 64, 128, 256, 512), text_key='text', label='fill: '): """Fill in the blank preprocessor that labels blank with a binned size. The actual blank size is sampled uniformly from the inclusive range of the min and max bin. The blank is then filled in with the closest bin size to the actual blank size. Args: dataset: a tf.data.Dataset, the dataset to preprocess. size_bins: a list, a list of blank sizes to select from when labelling the blank. text_key: a string, the key for the text feature to preprocess in the dataset examples. label: a string, the label to prepend to the inputs. Returns: a tf.data.Dataset """ bins = sorted(size_bins) @seqio.map_over_dataset(num_seeds=2) def my_fn(x, seeds): """Apply transformation.""" words = x['words'] n_words = tf.size(words) blank_size = tf.random.stateless_uniform( [], minval=bins[0], maxval=tf.math.minimum(n_words, bins[-1]), dtype=tf.dtypes.int32, seed=seeds[0]) bin_delta = tf.math.abs(bins - blank_size) bin_ = tf.gather(bins, tf.argmin(bin_delta)) blank_start = tf.random.stateless_uniform( [], minval=0, maxval=tf.math.maximum(0, n_words-blank_size) + 1, dtype=tf.dtypes.int32, seed=seeds[1]) pre_blank = tf.strings.reduce_join(words[0:blank_start], separator=' ') post_blank = tf.strings.reduce_join( words[blank_start+blank_size:], separator=' ') blank = tf.strings.format('_{}_', bin_) # We strip to handle cases where blank is at beginning or end. input_ = tf.strings.strip( tf.strings.join([pre_blank, blank, post_blank], ' ')) input_ = tf.strings.join([label, input_]) target = tf.strings.reduce_join( words[blank_start:blank_start+blank_size], separator=' ') return { 'inputs': tf.strings.strip(input_), 'targets': tf.strings.strip(target)} dataset = split_text_to_words(dataset, text_key, min_num_words=2) # Filter out examples with fewer words than the minimum. dataset = dataset.filter(lambda x: tf.size(x['words']) >= bins[0]) return my_fn(dataset)
Fill in the blank preprocessor that labels blank with a binned size. The actual blank size is sampled uniformly from the inclusive range of the min and max bin. The blank is then filled in with the closest bin size to the actual blank size. Args: dataset: a tf.data.Dataset, the dataset to preprocess. size_bins: a list, a list of blank sizes to select from when labelling the blank. text_key: a string, the key for the text feature to preprocess in the dataset examples. label: a string, the label to prepend to the inputs. Returns: a tf.data.Dataset
46
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf AUTOTUNE = tf.data.experimental.AUTOTUNE def translate(x, source_language, target_language): """Convert a translation dataset to a text2text pair. For example, say the dataset returns examples of this format: {'de': 'Das ist gut.', 'en': 'That is good.'} If source_language = 'de', target_language = 'en', then the outputs will have the format: {'inputs': 'translate German to English: Das ist gut.', 'targets': 'That is good.'} Args: x: an example to process. source_language: source language code (e.g. 'en') to translate from. target_language: target language code (e.g. 'de') to translate to. Returns: A preprocessed example with the format listed above. """ # Language codes like zh-cn are not supported; use only the first 2 chars for language in (source_language, target_language): if language != language[:2]: logging.warning( 'Extended language code %s not supported. Falling back on %s.', language, language[:2] ) lang_id_to_string = { source_language: babel.Locale(source_language[:2]).english_name, target_language: babel.Locale(target_language[:2]).english_name, } src_str = 'translate {}'.format(lang_id_to_string[source_language]) tgt_str = ' to {}: '.format(lang_id_to_string[target_language]) return { 'inputs': tf.strings.join([src_str, tgt_str, x[source_language]]), 'targets': x[target_language], } The provided code snippet includes necessary dependencies for implementing the `multi_translate` function. Write a Python function `def multi_translate(dataset, source_language, target_language)` to solve the following problem: Convert a multi-translate dataset to a text2text pair. For example, say the dataset returns examples which have a 'translations' feature key so that examples have the following format: { ... 'translations': { 'language': ['de', 'fr', 'en'], 'translation': ['Das ist gut.', 'Ca c'est bon', 'That is good.'] }, ... } If source_language = 'de', target_language = 'en', then this function will return examples of the format: {'inputs': 'translate German to English: Das is gut.', 'targets': 'That is good.'} Any other languages present in the dataset will be filtered out. Args: dataset: a tf.data.Dataset to process. source_language: source language code (e.g. 'en') to translate from. target_language: target language code (e.g. 'de') to translate to. Returns: A preprocessed tf.data.Dataset with the format listed above. Here is the function: def multi_translate(dataset, source_language, target_language): """Convert a multi-translate dataset to a text2text pair. For example, say the dataset returns examples which have a 'translations' feature key so that examples have the following format: { ... 'translations': { 'language': ['de', 'fr', 'en'], 'translation': ['Das ist gut.', 'Ca c'est bon', 'That is good.'] }, ... } If source_language = 'de', target_language = 'en', then this function will return examples of the format: {'inputs': 'translate German to English: Das is gut.', 'targets': 'That is good.'} Any other languages present in the dataset will be filtered out. Args: dataset: a tf.data.Dataset to process. source_language: source language code (e.g. 'en') to translate from. target_language: target language code (e.g. 'de') to translate to. Returns: A preprocessed tf.data.Dataset with the format listed above. """ def filter_fn(x): langs = x['translations']['language'] # Test whether both source/target_language appear in the language list source_in_langs = tf.reduce_any(tf.equal(source_language, langs)) target_in_langs = tf.reduce_any(tf.equal(target_language, langs)) return tf.logical_and(source_in_langs, target_in_langs) def map_fn(x): langs = x['translations']['language'] # Retrieve the index in langs where source/target_language appears src_idx = tf.squeeze(tf.where(tf.equal(langs, source_language))) tgt_idx = tf.squeeze(tf.where(tf.equal(langs, target_language))) return { source_language: x['translations']['translation'][src_idx], target_language: x['translations']['translation'][tgt_idx], } dataset = dataset.filter(filter_fn) dataset = dataset.map(map_fn, num_parallel_calls=AUTOTUNE) return translate(dataset, source_language, target_language)
Convert a multi-translate dataset to a text2text pair. For example, say the dataset returns examples which have a 'translations' feature key so that examples have the following format: { ... 'translations': { 'language': ['de', 'fr', 'en'], 'translation': ['Das ist gut.', 'Ca c'est bon', 'That is good.'] }, ... } If source_language = 'de', target_language = 'en', then this function will return examples of the format: {'inputs': 'translate German to English: Das is gut.', 'targets': 'That is good.'} Any other languages present in the dataset will be filtered out. Args: dataset: a tf.data.Dataset to process. source_language: source language code (e.g. 'en') to translate from. target_language: target language code (e.g. 'de') to translate to. Returns: A preprocessed tf.data.Dataset with the format listed above.
47
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `definite_pronoun_resolution_simple` function. Write a Python function `def definite_pronoun_resolution_simple(x, label='wsc:')` to solve the following problem: Converts DPR examples to a simple text to text format. A typical example from the definite pronoun resolution dataset might look like { 'sentence': 'Bob asked Tom if he can lend some money.', 'pronoun': 'he', 'candidates': ['Bob', 'Tom'], 'label': 1, } This will be transformed to { 'inputs': 'wsc: Bob asked Tom if *he* can lend some money.' 'targets': 'Tom', } Args: x: an example to process. label: a string, the label to prepend to the inputs. Returns: A preprocessed example. Here is the function: def definite_pronoun_resolution_simple(x, label='wsc:'): """Converts DPR examples to a simple text to text format. A typical example from the definite pronoun resolution dataset might look like { 'sentence': 'Bob asked Tom if he can lend some money.', 'pronoun': 'he', 'candidates': ['Bob', 'Tom'], 'label': 1, } This will be transformed to { 'inputs': 'wsc: Bob asked Tom if *he* can lend some money.' 'targets': 'Tom', } Args: x: an example to process. label: a string, the label to prepend to the inputs. Returns: A preprocessed example. """ # If there are multiple instances of the pronoun in the sentence, the first # one is the one that needs to be resolved. inputs = [ label, tf.strings.regex_replace( x['sentence'], tf.strings.join([r' (', x['pronoun'], r')( |\.|,)']), r' *\1*\2', replace_global=False, ), ] return { 'inputs': tf.strings.join(inputs, separator=' '), 'targets': x['candidates'][x['label']], }
Converts DPR examples to a simple text to text format. A typical example from the definite pronoun resolution dataset might look like { 'sentence': 'Bob asked Tom if he can lend some money.', 'pronoun': 'he', 'candidates': ['Bob', 'Tom'], 'label': 1, } This will be transformed to { 'inputs': 'wsc: Bob asked Tom if *he* can lend some money.' 'targets': 'Tom', } Args: x: an example to process. label: a string, the label to prepend to the inputs. Returns: A preprocessed example.
48
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def neighboring_pairs(dataset, text_key='text', reuse_sentences=True): """Create a dataset consisting of neighboring sentence pairs. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'first' and 'second'. We only take sentence pairs from within the same line since lines seem to represent paragraph-like structures in our text datasets. Empty lines and 1-sentence lines will thus be ignored. The argument reuse_sentences determines whether a sentence can be used as both the first and last element in the pair. For example, the input with sentences A,B,C,D will return (A,B),(B,C),(C,D) if reuse_sentences is True and (A,B),(C,D) if reuse_sentences is False. Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. reuse_sentences: a boolean Returns: a tf.data.Dataset """ def split_by_lines(dataset): """Splits text in dataset by line, removing empty lines.""" def my_fn(text): lines = tf.strings.split([text], sep='\n').values return tf.strings.strip(lines) dataset = dataset.map(my_fn, num_parallel_calls=AUTOTUNE) dataset = dataset.unbatch() return dataset.filter(lambda x: tf.strings.length(x) > 0) def split_into_pairs(line): """Split a given text example into pairs of neighboring sentences.""" # TODO(mmatena): Use better sentence segmentation. sep = str(uuid.uuid4()) sentences = tf.strings.regex_replace(line, r'((?:\.|\!|\?)+)', r'\1' + sep) sentences = tf.strings.strip(tf.strings.split([sentences], sep).values) if reuse_sentences: firsts = sentences[:-1] seconds = sentences[1:] else: firsts = sentences[:-1:2] seconds = sentences[1::2] return { 'first': firsts, 'second': seconds, } def example_len(x): return tf.math.minimum( tf.strings.length(x['first']), tf.strings.length(x['second'])) # Split by lines. dataset = dataset.map(lambda x: x[text_key], num_parallel_calls=AUTOTUNE) dataset = split_by_lines(dataset) # Get pairs of neighboring sentences. dataset = dataset.map(split_into_pairs, num_parallel_calls=AUTOTUNE) dataset = dataset.unbatch() # Remove examples with empty strings. dataset = dataset.filter(lambda x: example_len(x) > 0) return dataset The provided code snippet includes necessary dependencies for implementing the `next_sentence_prediction` function. Write a Python function `def next_sentence_prediction(dataset, text_key='text', reuse_sentences=True, label_sentences=False, p_neighbors=0.5, label='nsp: ', buffer_size=50000)` to solve the following problem: Create a dataset containing a next sentence prediction objective. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'inputs' and 'targets'. EXAMPLE OUTPUTS: { input: "nsp: sentence1: The man went to the store. sentence2: Penguins are " "flightless birds.", target: "not_next" } The "sentence1:" and "sentence2:" labels will be omitted if label_sentences is False. Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. reuse_sentences: a boolean, see docs for `neighboring_pairs` for more info. label_sentences: a boolean p_neighbors: a float between 0 and 1, the probability that a sentence pair will be neighbors. label: a string, the label to prepend to the inputs. buffer_size: an int, the size of the shuffle buffer used to get non-neighboring sentences. Returns: a tf.data.Dataset Here is the function: def next_sentence_prediction(dataset, text_key='text', reuse_sentences=True, label_sentences=False, p_neighbors=0.5, label='nsp: ', buffer_size=50000): """Create a dataset containing a next sentence prediction objective. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'inputs' and 'targets'. EXAMPLE OUTPUTS: { input: "nsp: sentence1: The man went to the store. sentence2: Penguins are " "flightless birds.", target: "not_next" } The "sentence1:" and "sentence2:" labels will be omitted if label_sentences is False. Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. reuse_sentences: a boolean, see docs for `neighboring_pairs` for more info. label_sentences: a boolean p_neighbors: a float between 0 and 1, the probability that a sentence pair will be neighbors. label: a string, the label to prepend to the inputs. buffer_size: an int, the size of the shuffle buffer used to get non-neighboring sentences. Returns: a tf.data.Dataset """ sentence1_label, sentence2_label = '', '' if label_sentences: sentence1_label, sentence2_label = 'sentence1: ', 'sentence2: ' empty = tf.constant('', dtype=tf.string, shape=[1]) dataset = neighboring_pairs( dataset, text_key=text_key, reuse_sentences=reuse_sentences) dataset = dataset.shuffle(buffer_size).batch(2, drop_remainder=True) def some_are_empty(*tensors): """See if at least one tensor has shape [0].""" empty = [tf.equal(tf.size(t), 0) for t in tensors] return tf.reduce_any(empty) @seqio.map_over_dataset(num_seeds=1) def my_fn(x, seed): """Function to be applied to each example in dataset.""" use_neighbors = ( tf.random.stateless_uniform(shape=[], seed=seed) < p_neighbors ) firsts, seconds = tf.cond( use_neighbors, lambda: (x['first'], x['second']), lambda: (x['first'], tf.stack([x['second'][1], x['second'][0]])), ) relation_label = tf.cond( use_neighbors, lambda: 'next', lambda: 'not_next', ) inputs = [] for i in range(2): first_inputs = firsts[i] second_inputs = seconds[i] def create_examples(first_i=first_inputs, second_i=second_inputs): return tf.strings.join([ label, sentence1_label, first_i, ' ', sentence2_label, second_i, ]) inpt = tf.cond( some_are_empty(first_inputs, second_inputs), lambda: empty, create_examples, ) inputs.append(tf.strings.strip(inpt)) inputs = tf.reshape(inputs, [-1]) targets = tf.reshape(2 * [relation_label], [-1]) return {'inputs': inputs, 'targets': targets} dataset = my_fn(dataset).unbatch() def example_len(x): return tf.math.minimum( tf.strings.length(x['inputs']), tf.strings.length(x['targets'])) # Remove examples with empty strings. return dataset.filter(lambda x: example_len(x) > 0)
Create a dataset containing a next sentence prediction objective. The input examples should have a key text_key associated with a tf.string value. The output examples have keys 'inputs' and 'targets'. EXAMPLE OUTPUTS: { input: "nsp: sentence1: The man went to the store. sentence2: Penguins are " "flightless birds.", target: "not_next" } The "sentence1:" and "sentence2:" labels will be omitted if label_sentences is False. Args: dataset: a tf.data.Dataset text_key: a string, the key for the text feature to preprocess in the dataset examples. reuse_sentences: a boolean, see docs for `neighboring_pairs` for more info. label_sentences: a boolean p_neighbors: a float between 0 and 1, the probability that a sentence pair will be neighbors. label: a string, the label to prepend to the inputs. buffer_size: an int, the size of the shuffle buffer used to get non-neighboring sentences. Returns: a tf.data.Dataset
49
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `lm` function. Write a Python function `def lm(x)` to solve the following problem: Basic language modeling objective for text - empty inputs. Given inputs with the format: {"text": "Here is some text."} This preprocess produces examples with the format {"inputs": "", "targets": "Here is some text."} Args: x: an example to process. Returns: A preprocessed example. Here is the function: def lm(x): """Basic language modeling objective for text - empty inputs. Given inputs with the format: {"text": "Here is some text."} This preprocess produces examples with the format {"inputs": "", "targets": "Here is some text."} Args: x: an example to process. Returns: A preprocessed example. """ return {'inputs': '', 'targets': x['text']}
Basic language modeling objective for text - empty inputs. Given inputs with the format: {"text": "Here is some text."} This preprocess produces examples with the format {"inputs": "", "targets": "Here is some text."} Args: x: an example to process. Returns: A preprocessed example.
50
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf AUTOTUNE = tf.data.experimental.AUTOTUNE def _wsc_inputs(x): """Given an example from SuperGLUE WSC, compute the 'inputs' value. The output will look like a fill in the blank with the pronoun blanked out. For example, the text 'Mitchell asked Tom if he could lend some money.' would be transformed to 'Mitchell asked Tom if X could lend some money.' Args: x: A dict that is an example from the WSC task of SuperGLUE. Returns: A scalar string tensor. """ words = tf.strings.split([x['text']], sep=' ').values # We would need some special logic to handle the case where the pronoun is the # first or last word in the text. None of the examples in WSC seem to have # this, so we are ignoring these cases. with tf.control_dependencies([ tf.assert_greater(x['span2_index'], 0), tf.assert_less(x['span2_index'], tf.size(words)), ]): pronoun_index = tf.identity(x['span2_index']) def create_input(): with tf.control_dependencies( [tf.assert_equal(words[pronoun_index], x['span2_text'])]): return tf.strings.join( [ tf.strings.reduce_join(words[:pronoun_index], separator=' '), 'X', tf.strings.reduce_join( words[pronoun_index + 1:], separator=' '), ], separator=' ', ) # Handle some special cases. if tf.equal( x['text'], 'The boy continued to whip the pony , and eventually the pony threw him over. John laughed out quite loud. \"Good for him,\" he said. ' ): return ( 'The boy continued to whip the pony , and eventually the pony threw ' 'him over. John laughed out quite loud. "Good for X ," he said.' ) # Using the span2_index, we get 'use' instead of 'it'. if tf.equal( x['text'], 'When they had eventually calmed down a bit , and had gotten home, Mr. Farley put the magic pebble in an iron safe . Some day they might want to use it , but really for now, what more could they wish for?' ): return ( 'When they had eventually calmed down a bit , and had gotten home, ' 'Mr. Farley put the magic pebble in an iron safe . Some day they might ' 'want to use X , but really for now, what more could they wish for?' ) return create_input() The provided code snippet includes necessary dependencies for implementing the `wsc_simple` function. Write a Python function `def wsc_simple(dataset, label='wsc:', correct_referent_only=False)` to solve the following problem: Converts SuperGLUE WSC examples to a simple text to text format. A typical example from SuperGLUE WSC might look like { 'text': 'Mitchell asked Tom if he could lend some money.', 'span1_text': 'Tom', 'span2_text': 'he', 'span2_index': 4, } This will be transformed to { 'inputs': 'wsc: Bob asked Tom if *he* can lend some money.' 'targets': 'Tom', } The targets will always be the text of the referent regardless of whether it is the correct referrent of the pronoun. Thus for training purposes, please set `correct_referent_only` to be True. Args: dataset: a tf.data.Dataset label: a string, the label to prepend to the inputs. correct_referent_only: a bool, whether to filter out examples for which the targets is not the correct referent of the pronoun. Returns: a tf.data.Dataset Here is the function: def wsc_simple(dataset, label='wsc:', correct_referent_only=False): """Converts SuperGLUE WSC examples to a simple text to text format. A typical example from SuperGLUE WSC might look like { 'text': 'Mitchell asked Tom if he could lend some money.', 'span1_text': 'Tom', 'span2_text': 'he', 'span2_index': 4, } This will be transformed to { 'inputs': 'wsc: Bob asked Tom if *he* can lend some money.' 'targets': 'Tom', } The targets will always be the text of the referent regardless of whether it is the correct referrent of the pronoun. Thus for training purposes, please set `correct_referent_only` to be True. Args: dataset: a tf.data.Dataset label: a string, the label to prepend to the inputs. correct_referent_only: a bool, whether to filter out examples for which the targets is not the correct referent of the pronoun. Returns: a tf.data.Dataset """ def map_fn(x): """Function to be called for every example in dataset.""" inputs = [ label, tf.strings.regex_replace( _wsc_inputs(x), r' X ', ' *' + x['span2_text'] + '* '), ] referent = x['span1_text'] return { 'inputs': tf.strings.join(inputs, separator=' '), # The reshape is necessary as otherwise the tensor has unknown rank. 'targets': tf.reshape(referent, shape=[]), 'label': x.get('label', 0), 'idx': x['idx'], } if correct_referent_only: dataset = dataset.filter(lambda x: tf.cast(x.get('label', False), tf.bool)) return dataset.map(map_fn, num_parallel_calls=AUTOTUNE)
Converts SuperGLUE WSC examples to a simple text to text format. A typical example from SuperGLUE WSC might look like { 'text': 'Mitchell asked Tom if he could lend some money.', 'span1_text': 'Tom', 'span2_text': 'he', 'span2_index': 4, } This will be transformed to { 'inputs': 'wsc: Bob asked Tom if *he* can lend some money.' 'targets': 'Tom', } The targets will always be the text of the referent regardless of whether it is the correct referrent of the pronoun. Thus for training purposes, please set `correct_referent_only` to be True. Args: dataset: a tf.data.Dataset label: a string, the label to prepend to the inputs. correct_referent_only: a bool, whether to filter out examples for which the targets is not the correct referent of the pronoun. Returns: a tf.data.Dataset
51
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `wnli_simple` function. Write a Python function `def wnli_simple(x, label='wsc:')` to solve the following problem: Converts GLUE WNLI examples to a simple text to text format. A typical example from WNLI might look like: { 'sentence1': 'The fish ate the worm. It was tasty.', 'sentence2': 'The worm was tasty.', 'label': 1, } This will be transformed to: { 'inputs': 'wsc: The fish ate the worm. *It* was tasty.', 'targets': 'The worm', 'premise': 'The fish ate the worm. It was tasty., 'hypothesis': 'The worm was tasty.', 'label': 1, } This preprocessor has been manually verified to produce reasonable WSC examples for the dev and test sets. Tasks using this preprocessor should only be used eval and not train. Args: x: an example to process. label: a string, the label to prepend to the inputs. Returns: A preprocessed example. Here is the function: def wnli_simple(x, label='wsc:'): """Converts GLUE WNLI examples to a simple text to text format. A typical example from WNLI might look like: { 'sentence1': 'The fish ate the worm. It was tasty.', 'sentence2': 'The worm was tasty.', 'label': 1, } This will be transformed to: { 'inputs': 'wsc: The fish ate the worm. *It* was tasty.', 'targets': 'The worm', 'premise': 'The fish ate the worm. It was tasty., 'hypothesis': 'The worm was tasty.', 'label': 1, } This preprocessor has been manually verified to produce reasonable WSC examples for the dev and test sets. Tasks using this preprocessor should only be used eval and not train. Args: x: an example to process. label: a string, the label to prepend to the inputs. Returns: A preprocessed example. """ pronouns = ['he', 'she', 'they', 'it', 'her', 'his', 'their', 'them', 'him'] PronounMatch = collections.namedtuple( # pylint: disable=invalid-name 'PronounMatch', ['score', 'index_in_premise', 'candidate']) def split_clean(s): """Returns array of words with punctuation and capitalization removed.""" words = [ re.sub(r'(\.|,|\?|\!)$', '', w) for w in s.strip().lower().split(' ') ] return [w for w in words if w] def get_all_pronoun_indices(s): return [i for i, w in enumerate(s) if w in pronouns] def get_post_match_size(hypothesis, words): """Returns len of largest prefix of words that is substr of hypothesis.""" hypothesis = ' '.join(hypothesis) for i in range(len(words)): if ' '.join(words[:i + 1]) not in hypothesis: return i return len(words) def get_pre_match_size(hypothesis, words): """Returns len of largest suffix of words that is substr of hypothesis.""" return get_post_match_size(hypothesis[::-1], words[::-1]) def get_pronoun_match(premise, hypothesis, index): """Return the PronounMatch for the pronoun at `index` in premise.""" pre, post = premise[:index], premise[index + 1:] pre_match_size = get_pre_match_size(hypothesis, pre) post_match_size = get_post_match_size(hypothesis, post) score = pre_match_size + post_match_size candidate = '' if score: pre_match = pre[-pre_match_size or len(pre):] post_match = post[:post_match_size] m = re.search(' '.join(pre_match + [r'(.+)'] + post_match), ' '.join(hypothesis)) if not m: # Handle cases where the candidate is at the start of the hypthesis. m = re.search(' '.join([r'^(.+)'] + post_match), ' '.join(hypothesis)) if not m: # Handle cases where the candidate is at the end of the hypthesis. m = re.search(' '.join(pre_match + [r'(.+)$']), ' '.join(hypothesis)) if m: candidate = m.group(1) return PronounMatch( score=score, index_in_premise=index, candidate=candidate) def get_best_pronoun_match(premise, hypothesis): """Returns the match for the pronoun in the premise to disambiguate.""" pronoun_indices = get_all_pronoun_indices(premise) scoredpronouns = [ get_pronoun_match(premise, hypothesis, index) for index in pronoun_indices ] return max(scoredpronouns, key=lambda x: x.score) def highlight(sentence, index): words = sentence.split(' ') word = words[index] if word[-1] in ['.', ',', '!', '?']: highlighted = '*{}* {}'.format(word[:-1], word[-1]) else: highlighted = '*{}*'.format(word) return ' '.join(words[:index] + [highlighted] + words[index + 1:]) def make_nonpossessive(word): # WSC simple targets will never be possessive, even when the pronoun is # possesive. if word.endswith("'"): return word[:-1] elif word.endswith("'s"): return word[:-2] else: return word def clean_up(candidate): words = candidate.split(' ') # Sometimes the candidate extraction messes up, and the candidate will start # with the start of the hypothesis and extend to the correct candidate. We # can try to clean up the candidate in some cases by removing everything up # to the last article in the sentence. article_index = max( [words.index(art) for art in {'a', 'an', 'the'} if art in words] or [0]) return ' '.join(words[article_index:]) def process_candidate(candidate, hypothesis): """Handles special cases and adds proper punctuation/capitalization.""" candidate = clean_up(candidate) pattern = '({})'.format(' '.join([ r'{}(?:\.|,|\?|\!)?'.format(re.escape(c)) for c in candidate.split(' ') ])) m = re.search(pattern, hypothesis, re.IGNORECASE) if not m: raise ValueError( 'Unable to find candidate "{}" in hypothesis "{}".'.format( candidate, hypothesis)) candidate = m.group(1) if candidate and candidate[-1] in ['.', ',', '!', '?']: candidate = candidate[:-1] return make_nonpossessive(candidate) def compute_inputs_and_targets(premise, hypothesis): """Compute inputs and targets for WNLI simple.""" premise = tf.compat.as_text(premise.numpy()) hypothesis = tf.compat.as_text(hypothesis.numpy()) match = get_best_pronoun_match( split_clean(premise), split_clean(hypothesis)) targets = process_candidate(match.candidate, hypothesis) inputs = '{} {}'.format(label, highlight(premise, match.index_in_premise)) return inputs, targets inputs, targets = tf.py_function( compute_inputs_and_targets, inp=[x['sentence1'], x['sentence2']], Tout=[tf.string, tf.string]) return { # The reshape is necessary as otherwise the tensor has unknown rank. 'inputs': tf.reshape(inputs, shape=[]), 'targets': tf.reshape(targets, shape=[]), 'premise': x['sentence1'], 'hypothesis': x['sentence2'], 'label': x.get('label', 0), 'idx': x['idx'], }
Converts GLUE WNLI examples to a simple text to text format. A typical example from WNLI might look like: { 'sentence1': 'The fish ate the worm. It was tasty.', 'sentence2': 'The worm was tasty.', 'label': 1, } This will be transformed to: { 'inputs': 'wsc: The fish ate the worm. *It* was tasty.', 'targets': 'The worm', 'premise': 'The fish ate the worm. It was tasty., 'hypothesis': 'The worm was tasty.', 'label': 1, } This preprocessor has been manually verified to produce reasonable WSC examples for the dev and test sets. Tasks using this preprocessor should only be used eval and not train. Args: x: an example to process. label: a string, the label to prepend to the inputs. Returns: A preprocessed example.
52
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def rank_classification( ds: tf.data.Dataset, inputs_fn: Callable[[FeatureType], tf.Tensor], targets_fn: Callable[[FeatureType], tf.Tensor], is_correct_fn: Callable[[FeatureType], tf.Tensor], weight_fn: Optional[Callable[[FeatureType], tf.Tensor]] = None, mode: str = 'eval', passthrough_feature_keys: Optional[Sequence[str]] = None, ) -> tf.data.Dataset: """Prepare dataset for rank classification scoring. Intended to be used with `rank_classification` postprocessor and metric. `inputs_fn` and `targets_fn` must return the 'inputs' and 'targets' features, respectively, for each possible class label given the raw example features. 'is_correct_fn' must return the 'is_correct' feature, a boolean for whether each label is matching with the ground truth target before the examples are expanded. In 'train' mode, only the inputs / targets marked correct will be produced. In 'eval' mode, all inputs / targets will be produced. In 'fewshot_eval', all inputs / targets will be produced as a single batch. Each output example will also be given a unique 'idx' feature. The first dim is a sequential index for the input example and the second is the index of the generated output for it. E.g., the second output example from the fourth input example would be `[3, 1]`. To be clear, consider the following arguments: inputs_fn=lambda ex: ex['prefix'], targets_fn=lambda ex: ex['suffix'], is_correct_fn=lambda ex: tf.one_hot(ex['label'], num_classes) weight_fn=lambda ex: ex['weight'] Given the following example: { 'prefix': ['The farmland needed ', 'The farmland wanted '], 'suffix': ['water', 'cows'], 'label': 0, 'weight': 1.0, } the preprocessor would return: [{ 'idx': [0, 0], 'inputs': 'The farmland needed ', 'targets': 'water', 'is_correct': True, 'weight': 1.0 }, { 'idx': [0, 1], 'inputs': 'The farmland wanted ', 'targets': 'cows', 'is_correct': False, 'weight': 1.0 }] With mode set to 'train', it would return only the first example, since it uses the correct label. With mode set to 'fewshot_eval', it would return both examples in a single batch. Args: ds: a tf.data.Dataset to preprocess. inputs_fn: a callable that returns the 'inputs' features for each label given the input example. targets_fn: a callable that returns the 'targets' features for each label given the input example. is_correct_fn: a callable that returns the 'label' feature. May be an int32 scalar or 1-D Tensor. weight_fn: a callable that returns the 'weight' feature (float32 scalar). mode: A string, one of 'train' or'eval 'train' produces only the correct example(s) based on the label value(s). 'eval' produces an example for every possible class value, sequentially. 'fewshot_eval' produces an example for every possible class value, batched together for each input example. passthrough_feature_keys: a sequence of feature names that should be passed through to the output of this preprocessor. eg: ["starburst", "tokens"] Returns: A tf.data.Dataset containing 'idx', inputs', 'targets', and 'is_correct'. """ if mode not in ('train', 'eval', 'fewshot_eval'): raise ValueError( "Mode must be one of 'train', 'eval', or 'fewshot_eval'. " f"Got '{mode}'.") def make_examples(idx, ex): inputs = inputs_fn(ex) targets = targets_fn(ex) is_correct = tf.cast(is_correct_fn(ex), tf.bool) tf.debugging.assert_equal( tf.size(is_correct), [tf.size(inputs), tf.size(targets)], '`inputs_fn`, `targets_fn`, and `is_correct_fn` must return the same ' 'size tensors.') num_out = tf.size(is_correct) in_idx = tf.fill([num_out], tf.cast(idx, tf.int32)) out_idx = tf.range(num_out) output = { 'idx': tf.stack([in_idx, out_idx], 1), 'inputs': inputs, 'targets': targets, 'is_correct': is_correct, } if weight_fn is not None: output['weight'] = tf.fill(tf.shape(is_correct), weight_fn(ex)) output['weight'] = tf.cast(output['weight'], tf.float32) for feature_name in passthrough_feature_keys or []: if feature_name in output: raise ValueError( f'The feature {feature_name} to pass through, already exists' 'in the preprocessed output. Try renaming it to something else.' ) tiled_shape = tf.concat( [ tf.expand_dims(tf.shape(targets)[0], axis=0), tf.ones(len(ex[feature_name].shape), dtype=tf.int32), ], axis=0, ) output[feature_name] = tf.tile( tf.expand_dims(ex[feature_name], axis=0), tiled_shape ) return output ds = ds.enumerate() ds = ds.map(make_examples, num_parallel_calls=AUTOTUNE) if mode != 'fewshot_eval': ds = ds.unbatch() if mode == 'train': ds = ds.filter(lambda ex: ex['is_correct']) return ds The provided code snippet includes necessary dependencies for implementing the `rank_classification_formatter` function. Write a Python function `def rank_classification_formatter( ds: tf.data.Dataset, inputs_formats: Union[str, Sequence[str]], targets_formats: Union[str, Sequence[str]], mode: str = 'eval', label_key: str = 'label', weight_key: Optional[str] = None) -> tf.data.Dataset` to solve the following problem: Create 'inputs' and 'targets' strings for ranking classification. Intended to be used with `rank_classification` postprocessor and metric. Inputs will be formatted by filling in the feature values in the `inputs_formats` and `targets_formats` strings. Nested features can be accessed by concatenating the features using forward slash. For eg: if sub-sub-key is nested under sub-key, which is nested under key, then sub-sub-key can be accessed using key/sub-key/sub-sub-key. In 'eval' mode, a separate example will be produced for each targets / inputs format string. These can then be scored to find the one with the highest likelihood. The `rank_classification` postprocessor and metric allow you to evaluate with this technique. In 'train' mode, only the targets / inputs format string indexed by the label(s) will be produced. In 'eval' mode, all inputs / targets will be produced. Each input example will also be given a unique, sequential index called 'idx'. For example, with arguments: ``` inputs_format='{premise} What is the {question}? X', targets_formats=[ 'I think {choice1}.', 'I think {choice2}.' ], mode='eval' ``` given the input: { 'premise': 'The farmland needed irrigation.', 'question': 'effect', 'choice1' : 'a canal was constructed', 'choice2': 'the crops grew tall', 'label': 0, } the preprocessor would return: [{ 'idx': 0, 'inputs': 'The farmland needed irrigation. What is the effect? X', 'targets': 'I think a canal was constructed.', 'is_correct': True }, { 'idx': 0, 'inputs': 'The farmland needed irrigation. What is the effect? X', 'targets': 'I think the crops grew tall.', 'is_correct': False }] With `mode='train'`, it would return only the first example, since it uses the correct label. With `mode='fewshot_eval'`, it would return both examples in a single batch. Args: ds: a tf.data.Dataset to preprocess. inputs_formats: A string or a list of strings to format with feature values to produce 'inputs'. Feature keys should be surrounded by curly braces to be replaced. targets_formats: A string or a list of strings to format with feature values to produce 'targets', one for each possible class value. Feature keys should be surrounded by curly braces to be replaced. mode: A string, one of 'train', 'eval', or 'fewshot_train') 'train' produces only the correct example(s) based on the label value(s). 'eval' produces an example for every possible class value, sequentially. 'fewshot_eval': produces an example for every possible class value, batched together for each input example. label_key: A string, the feature key for the integer label value(s). weight_key: A string, the feature key for the float example weight. Returns: A tf.data.Dataset containing 'idx', inputs', 'targets', and 'is_correct'. Here is the function: def rank_classification_formatter( ds: tf.data.Dataset, inputs_formats: Union[str, Sequence[str]], targets_formats: Union[str, Sequence[str]], mode: str = 'eval', label_key: str = 'label', weight_key: Optional[str] = None) -> tf.data.Dataset: """Create 'inputs' and 'targets' strings for ranking classification. Intended to be used with `rank_classification` postprocessor and metric. Inputs will be formatted by filling in the feature values in the `inputs_formats` and `targets_formats` strings. Nested features can be accessed by concatenating the features using forward slash. For eg: if sub-sub-key is nested under sub-key, which is nested under key, then sub-sub-key can be accessed using key/sub-key/sub-sub-key. In 'eval' mode, a separate example will be produced for each targets / inputs format string. These can then be scored to find the one with the highest likelihood. The `rank_classification` postprocessor and metric allow you to evaluate with this technique. In 'train' mode, only the targets / inputs format string indexed by the label(s) will be produced. In 'eval' mode, all inputs / targets will be produced. Each input example will also be given a unique, sequential index called 'idx'. For example, with arguments: ``` inputs_format='{premise} What is the {question}? X', targets_formats=[ 'I think {choice1}.', 'I think {choice2}.' ], mode='eval' ``` given the input: { 'premise': 'The farmland needed irrigation.', 'question': 'effect', 'choice1' : 'a canal was constructed', 'choice2': 'the crops grew tall', 'label': 0, } the preprocessor would return: [{ 'idx': 0, 'inputs': 'The farmland needed irrigation. What is the effect? X', 'targets': 'I think a canal was constructed.', 'is_correct': True }, { 'idx': 0, 'inputs': 'The farmland needed irrigation. What is the effect? X', 'targets': 'I think the crops grew tall.', 'is_correct': False }] With `mode='train'`, it would return only the first example, since it uses the correct label. With `mode='fewshot_eval'`, it would return both examples in a single batch. Args: ds: a tf.data.Dataset to preprocess. inputs_formats: A string or a list of strings to format with feature values to produce 'inputs'. Feature keys should be surrounded by curly braces to be replaced. targets_formats: A string or a list of strings to format with feature values to produce 'targets', one for each possible class value. Feature keys should be surrounded by curly braces to be replaced. mode: A string, one of 'train', 'eval', or 'fewshot_train') 'train' produces only the correct example(s) based on the label value(s). 'eval' produces an example for every possible class value, sequentially. 'fewshot_eval': produces an example for every possible class value, batched together for each input example. label_key: A string, the feature key for the integer label value(s). weight_key: A string, the feature key for the float example weight. Returns: A tf.data.Dataset containing 'idx', inputs', 'targets', and 'is_correct'. """ if (isinstance(inputs_formats, (list, tuple)) and isinstance(targets_formats, (list, tuple))): if len(inputs_formats) != len(targets_formats): raise ValueError( f'The inputs_formats ({len(inputs_formats)}) and ' f'targets_formats ({len(targets_formats)}) are both instances ' 'of list or tuple, but do not have matching lengths.') elif isinstance(inputs_formats, (list, tuple)): num_classes = len(inputs_formats) targets_formats = [targets_formats] * num_classes elif isinstance(targets_formats, (list, tuple)): num_classes = len(targets_formats) inputs_formats = [inputs_formats] * num_classes else: raise ValueError( 'One of the inputs_formats and targets_formats has to ' f'be a list or tuple, inputs_formats: {inputs_formats}, ' f'target_formats: {targets_formats}.') def _format_str(features, fmt): keys = set(re.findall(r'{(\S+)}', fmt)) s = fmt for k in keys: value = features for subkey in k.split('/'): value = value[subkey] if not isinstance(value, tf.Tensor): raise ValueError( f'Final value of key \'{k}\' must be a tf.string. ' f'Got: {type(value).__name__}') tf.debugging.assert_type( value, tf.string, f'Final value of key \'{k}\' must be a tf.string. ' f'Got: {value.dtype.name}') s = tf.strings.regex_replace(s, '{%s}' % k, value) return s def _apply_formats(features, fmts): return [_format_str(features, fmt) for fmt in fmts] def _is_correct_fn(ex): labels = ex[label_key] is_correct = tf.one_hot(labels, num_classes, on_value=True, off_value=False) if labels.shape.rank: is_correct = tf.math.reduce_any(is_correct, axis=0) return is_correct def _weight_fn(ex): return ex[weight_key] return rank_classification( ds, inputs_fn=functools.partial(_apply_formats, fmts=inputs_formats), targets_fn=functools.partial(_apply_formats, fmts=targets_formats), is_correct_fn=_is_correct_fn, weight_fn=None if weight_key is None else _weight_fn, mode=mode)
Create 'inputs' and 'targets' strings for ranking classification. Intended to be used with `rank_classification` postprocessor and metric. Inputs will be formatted by filling in the feature values in the `inputs_formats` and `targets_formats` strings. Nested features can be accessed by concatenating the features using forward slash. For eg: if sub-sub-key is nested under sub-key, which is nested under key, then sub-sub-key can be accessed using key/sub-key/sub-sub-key. In 'eval' mode, a separate example will be produced for each targets / inputs format string. These can then be scored to find the one with the highest likelihood. The `rank_classification` postprocessor and metric allow you to evaluate with this technique. In 'train' mode, only the targets / inputs format string indexed by the label(s) will be produced. In 'eval' mode, all inputs / targets will be produced. Each input example will also be given a unique, sequential index called 'idx'. For example, with arguments: ``` inputs_format='{premise} What is the {question}? X', targets_formats=[ 'I think {choice1}.', 'I think {choice2}.' ], mode='eval' ``` given the input: { 'premise': 'The farmland needed irrigation.', 'question': 'effect', 'choice1' : 'a canal was constructed', 'choice2': 'the crops grew tall', 'label': 0, } the preprocessor would return: [{ 'idx': 0, 'inputs': 'The farmland needed irrigation. What is the effect? X', 'targets': 'I think a canal was constructed.', 'is_correct': True }, { 'idx': 0, 'inputs': 'The farmland needed irrigation. What is the effect? X', 'targets': 'I think the crops grew tall.', 'is_correct': False }] With `mode='train'`, it would return only the first example, since it uses the correct label. With `mode='fewshot_eval'`, it would return both examples in a single batch. Args: ds: a tf.data.Dataset to preprocess. inputs_formats: A string or a list of strings to format with feature values to produce 'inputs'. Feature keys should be surrounded by curly braces to be replaced. targets_formats: A string or a list of strings to format with feature values to produce 'targets', one for each possible class value. Feature keys should be surrounded by curly braces to be replaced. mode: A string, one of 'train', 'eval', or 'fewshot_train') 'train' produces only the correct example(s) based on the label value(s). 'eval' produces an example for every possible class value, sequentially. 'fewshot_eval': produces an example for every possible class value, batched together for each input example. label_key: A string, the feature key for the integer label value(s). weight_key: A string, the feature key for the float example weight. Returns: A tf.data.Dataset containing 'idx', inputs', 'targets', and 'is_correct'.
53
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `parse_tsv` function. Write a Python function `def parse_tsv(line, field_names=None, field_delim='\t', field_columns=None)` to solve the following problem: Splits TSV lines into dict examples mapping field name to string value. Args: line: an example containing a comma/tab-delimited string. field_names: a list of strings, the ordered names of the TSV fields. Defaults to "inputs" and "targets". field_delim: a string, the delimiter to split on e.g. ',' for csv. field_columns: a list of column indices for each field. Defaults to consecutive numbering of the provided `field_names`. Returns: A feature dict mapping field name to string value. Here is the function: def parse_tsv(line, field_names=None, field_delim='\t', field_columns=None): """Splits TSV lines into dict examples mapping field name to string value. Args: line: an example containing a comma/tab-delimited string. field_names: a list of strings, the ordered names of the TSV fields. Defaults to "inputs" and "targets". field_delim: a string, the delimiter to split on e.g. ',' for csv. field_columns: a list of column indices for each field. Defaults to consecutive numbering of the provided `field_names`. Returns: A feature dict mapping field name to string value. """ field_names = field_names or ['inputs', 'targets'] field_columns = field_columns or list(range(len(field_names))) return dict( zip(field_names, tf.io.decode_csv( line, record_defaults=[''] * len(field_names), field_delim=field_delim, use_quote_delim=False, select_cols=field_columns)))
Splits TSV lines into dict examples mapping field name to string value. Args: line: an example containing a comma/tab-delimited string. field_names: a list of strings, the ordered names of the TSV fields. Defaults to "inputs" and "targets". field_delim: a string, the delimiter to split on e.g. ',' for csv. field_columns: a list of column indices for each field. Defaults to consecutive numbering of the provided `field_names`. Returns: A feature dict mapping field name to string value.
54
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `preprocess_tsv` function. Write a Python function `def preprocess_tsv(line, field_delim='\t', num_fields=2, inputs_format='{0}', targets_format='{1}', field_names=None, use_quote_delim=False)` to solve the following problem: r"""Parse tab-delimited strings into inputs and targets. This function takes a tf.data.Dataset of strings, each of which contains tab-delimited fields. The function returns a tf.data.Dataset of feature dictionaries of the form {"inputs": string, "targets": string}. inputs_format contains a template string and field numbers or names used to produce the "inputs" string. targets_format contains a template string and field numbers or names used to produce the "targets" string. Example (field numbers): The input dataset contains the lines: "6,7,42" "2,9,18" preprocess_tsv(dataset, field_delim=',', inputs_format='numerator: {2} denominator: {1}', targets_format='quotient: {0}' would produce a dataset containing the dictionaries: {"inputs": "numerator: 42 denominator: 7", "targets": "quotient: 6"} {"inputs": "numerator: 18 denominator: 9", "targets": "quotient: 2"} Example (field names): The input dataset contains the lines: "6,7,42" "2,9,18" preprocess_tsv(dataset, field_delim=',', field_names=['quot', 'denom', 'numer'], inputs_format='numerator: {numer} denominator: {denom}', targets_format='quotient: {quot}' would produce a dataset containing the dictionaries: {"inputs": "numerator: 42 denominator: 7", "targets": "quotient: 6"} {"inputs": "numerator: 18 denominator: 9", "targets": "quotient: 2"} Args: line: an example containing comma/tab-delimited string. field_delim: a string, the delimiter to split on e.g. ',' for csv. num_fields: an integer inputs_format: a string, the desired output format with placeholders for field values. targets_format: a string, the desired output format with placeholders for field values. field_names: a list of strings, the ordered names of the TSV fields. defaults to None (i.e. use field number in *_format) use_quote_delim: If false, treats double quotation marks as regular characters inside of the string fields (ignoring RFC 4180, Section 2, Bullet 5). Returns: A feature dict with 'inputs' and 'targets' features. Here is the function: def preprocess_tsv(line, field_delim='\t', num_fields=2, inputs_format='{0}', targets_format='{1}', field_names=None, use_quote_delim=False): r"""Parse tab-delimited strings into inputs and targets. This function takes a tf.data.Dataset of strings, each of which contains tab-delimited fields. The function returns a tf.data.Dataset of feature dictionaries of the form {"inputs": string, "targets": string}. inputs_format contains a template string and field numbers or names used to produce the "inputs" string. targets_format contains a template string and field numbers or names used to produce the "targets" string. Example (field numbers): The input dataset contains the lines: "6,7,42" "2,9,18" preprocess_tsv(dataset, field_delim=',', inputs_format='numerator: {2} denominator: {1}', targets_format='quotient: {0}' would produce a dataset containing the dictionaries: {"inputs": "numerator: 42 denominator: 7", "targets": "quotient: 6"} {"inputs": "numerator: 18 denominator: 9", "targets": "quotient: 2"} Example (field names): The input dataset contains the lines: "6,7,42" "2,9,18" preprocess_tsv(dataset, field_delim=',', field_names=['quot', 'denom', 'numer'], inputs_format='numerator: {numer} denominator: {denom}', targets_format='quotient: {quot}' would produce a dataset containing the dictionaries: {"inputs": "numerator: 42 denominator: 7", "targets": "quotient: 6"} {"inputs": "numerator: 18 denominator: 9", "targets": "quotient: 2"} Args: line: an example containing comma/tab-delimited string. field_delim: a string, the delimiter to split on e.g. ',' for csv. num_fields: an integer inputs_format: a string, the desired output format with placeholders for field values. targets_format: a string, the desired output format with placeholders for field values. field_names: a list of strings, the ordered names of the TSV fields. defaults to None (i.e. use field number in *_format) use_quote_delim: If false, treats double quotation marks as regular characters inside of the string fields (ignoring RFC 4180, Section 2, Bullet 5). Returns: A feature dict with 'inputs' and 'targets' features. """ def _format_part_with_field_numbers(part, field_values): found = re.findall(r'{(\d+)}', part) if found: return field_values[int(found[0])] else: return part def _format_part_with_field_names(part, field_names, field_values): field_names_re = '|'.join(['{{({})}}'.format(x) for x in field_names]) found = re.findall(field_names_re, part) if found: pos = field_names.index(''.join(found[0])) return field_values[int(pos)] else: return part def _format(format_string, field_names, field_values): if field_names is None: parts = [ _format_part_with_field_numbers(p, field_values) for p in re.split(r'({\d+})', format_string) ] else: field_names_re = '(' + '|'.join(['{{{}}}'.format(x) for x in field_names ]) + ')' parts = [ _format_part_with_field_names(p, field_names, field_values) for p in re.split(field_names_re, format_string) ] return tf.strings.join(parts) field_values = tf.io.decode_csv( line, record_defaults=[''] * (num_fields if field_names is None else len(field_names)), field_delim=field_delim, use_quote_delim=use_quote_delim) return { 'inputs': _format(inputs_format, field_names, field_values), 'targets': _format(targets_format, field_names, field_values) }
r"""Parse tab-delimited strings into inputs and targets. This function takes a tf.data.Dataset of strings, each of which contains tab-delimited fields. The function returns a tf.data.Dataset of feature dictionaries of the form {"inputs": string, "targets": string}. inputs_format contains a template string and field numbers or names used to produce the "inputs" string. targets_format contains a template string and field numbers or names used to produce the "targets" string. Example (field numbers): The input dataset contains the lines: "6,7,42" "2,9,18" preprocess_tsv(dataset, field_delim=',', inputs_format='numerator: {2} denominator: {1}', targets_format='quotient: {0}' would produce a dataset containing the dictionaries: {"inputs": "numerator: 42 denominator: 7", "targets": "quotient: 6"} {"inputs": "numerator: 18 denominator: 9", "targets": "quotient: 2"} Example (field names): The input dataset contains the lines: "6,7,42" "2,9,18" preprocess_tsv(dataset, field_delim=',', field_names=['quot', 'denom', 'numer'], inputs_format='numerator: {numer} denominator: {denom}', targets_format='quotient: {quot}' would produce a dataset containing the dictionaries: {"inputs": "numerator: 42 denominator: 7", "targets": "quotient: 6"} {"inputs": "numerator: 18 denominator: 9", "targets": "quotient: 2"} Args: line: an example containing comma/tab-delimited string. field_delim: a string, the delimiter to split on e.g. ',' for csv. num_fields: an integer inputs_format: a string, the desired output format with placeholders for field values. targets_format: a string, the desired output format with placeholders for field values. field_names: a list of strings, the ordered names of the TSV fields. defaults to None (i.e. use field number in *_format) use_quote_delim: If false, treats double quotation marks as regular characters inside of the string fields (ignoring RFC 4180, Section 2, Bullet 5). Returns: A feature dict with 'inputs' and 'targets' features.
55
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def select_random_chunk(dataset: tf.data.Dataset, output_features: Mapping[str, seqio.Feature], max_length: Optional[int] = None, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[ Sequence[str]] = None, sequence_length: Optional[Mapping[str, int]] = None, uniform_random_start: bool = False, min_length: Optional[int] = None, **unused_kwargs) -> tf.data.Dataset: """SeqIO wrapper for single_example_select_random_chunk().""" def _my_fn(x, seed): return single_example_select_random_chunk( x, seed, output_features=output_features, max_length=max_length, feature_key=feature_key, additional_feature_keys=additional_feature_keys, passthrough_feature_keys=passthrough_feature_keys, sequence_length=sequence_length, uniform_random_start=uniform_random_start, min_length=min_length) # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) return _my_fn(dataset) def reduce_concat_tokens(dataset, feature_key='targets', batch_size=128, **unused_kwargs): """Token-preprocessor to concatenate multiple unrelated documents. If we want to generate examples of exactly the right length, (to avoid wasting space on padding), then we use this function, folowed by split_tokens. Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. feature_key: an string batch_size: an integer - how many documents to concatenate into one Returns: a dataset """ dataset = dataset.map( lambda x: {feature_key: x[feature_key]}, num_parallel_calls=AUTOTUNE) dataset = dataset.padded_batch(batch_size, padded_shapes={feature_key: [-1]}) def _my_fn(x): tokens = tf.reshape(x[feature_key], [-1]) # strip padding tokens = tf.boolean_mask(tokens, tf.cast(tokens, tf.bool)) return {feature_key: tokens} return dataset.map(_my_fn, num_parallel_calls=AUTOTUNE) def split_tokens(dataset: tf.data.Dataset, min_tokens_per_segment: Optional[int] = None, max_tokens_per_segment: int = gin.REQUIRED, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[Sequence[str]] = None, **unused_kwargs) -> tf.data.Dataset: """Split examples into multiple examples each. The intended use case is to break up long examples for use in unsupervised transfer-learning. This function is generally preceded by select_random_chunk. If min_tokens_per_segment is provided, the segment length is chosen randomly per document from a log-uniform distribution. If min_tokens_per_segment is None, then the segment length is max_tokens_per_segment (except for a possibly shorter last segment in each document). Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. min_tokens_per_segment: an optional integer max_tokens_per_segment: an integer, the maximum number of tokens in each segment. Only the final segment may be shorter. feature_key: a string, the feature to split additional_feature_keys: Additional features to split. The same chunk size will be used, so they should be the same size as feature_key. passthrough_feature_keys: Features to pass through without any splitting. Returns: a dataset """ if passthrough_feature_keys: split_keys = set([feature_key] + (additional_feature_keys or [])) overlap_keys = split_keys & set(passthrough_feature_keys) if overlap_keys: raise ValueError( f'split keys {overlap_keys} also included in passthrough keys') def _split_tokens(x, seed): """Split one token sequence into multiple sequences.""" tokens = x[feature_key] n_tokens = tf.shape(tokens)[0] if min_tokens_per_segment is None: length = max_tokens_per_segment else: # pick a length - log-uniformly distributed length = tf.cast( tf.exp( tf.random.stateless_uniform( [], minval=math.log(min_tokens_per_segment), maxval=math.log(max_tokens_per_segment), seed=seed ) ), tf.int32) # Pad to a multiple of length, then use tf.reshape to split up the tokens # into num_segments segments each of the given length. num_segments = tf.cast( tf.math.ceil( tf.cast(n_tokens, tf.float32) / tf.cast(length, tf.float32)) , tf.int32) padding = num_segments * length - tf.shape(tokens)[0] feature_keys_to_split = [feature_key] orig_lengths = {} outputs = {} if additional_feature_keys is not None: feature_keys_to_split.extend(additional_feature_keys) for k in feature_keys_to_split: with tf.control_dependencies([ tf.assert_equal( tf.shape(tokens)[0], tf.shape(x[k])[0], message=(f'Additional feature {k} is not the same size as ' f'{feature_key} along axis 0 in split_tokens().') ) ]): shape = tf.shape(x[k])[1:] shape_list = x[k].shape[1:] padded = tf.pad( x[k], tf.concat([[[0, padding]], tf.zeros([len(shape_list), 2], dtype=tf.int32)], axis=0)) orig_lengths[k] = tf.concat( [tf.repeat(length, num_segments - 1), [length - padding]], axis=0) outputs[k] = tf.reshape( padded, tf.concat([[-1, length], shape], axis=0)) # To avoid memory issues, don't just replicate the passthrough features # for every segment; use tf.data to do it so the copies don't get # instantiated all at once. outputs_ds = tf.data.Dataset.from_tensor_slices(outputs) orig_lengths_ds = tf.data.Dataset.from_tensor_slices(orig_lengths) if passthrough_feature_keys: passthrough = {k: v for k, v in x.items() if k in passthrough_feature_keys} passthrough_ds = tf.data.Dataset.from_tensors(passthrough).repeat( tf.cast(num_segments, tf.int64)) return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds, passthrough_ds)) else: return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds)) def _strip_padding_and_merge_passthrough( inputs, orig_lengths, passthrough=None): output = {} for k, v in inputs.items(): output[k] = v[:orig_lengths[k]] if passthrough: for k, v in passthrough.items(): output[k] = passthrough[k] return output # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) dataset = _split_tokens(dataset).flat_map(lambda z: z) dataset = dataset.map( _strip_padding_and_merge_passthrough, num_parallel_calls=AUTOTUNE) return dataset def random_spans_helper(inputs_length=gin.REQUIRED, noise_density=gin.REQUIRED, mean_noise_span_length=gin.REQUIRED, extra_tokens_per_span_inputs=gin.REQUIRED, extra_tokens_per_span_targets=gin.REQUIRED, verbose=False): """Training parameters to avoid padding with random_spans_noise_mask. When training a model with random_spans_noise_mask, we would like to set the other training hyperparmeters in a way that avoids padding. This function helps us compute these hyperparameters. We assume that each noise span in the input is replaced by extra_tokens_per_span_inputs sentinel tokens, and each non-noise span in the targets is replaced by extra_tokens_per_span_targets sentinel tokens. This function tells us the required number of tokens in the raw example (for split_tokens()) as well as the length of the encoded targets. Note that this function assumes the inputs and targets will have EOS appended and includes that in the reported length. Args: inputs_length: an integer - desired length of the tokenized inputs sequence noise_density: a float mean_noise_span_length: a float extra_tokens_per_span_inputs: an integer extra_tokens_per_span_targets: an integer verbose: a bool indicating whether to log sequence lengths Returns: tokens_length: length of original text in tokens targets_length: an integer - length in tokens of encoded targets sequence """ def _tokens_length_to_inputs_length_targets_length(tokens_length): num_noise_tokens = int(round(tokens_length * noise_density)) num_nonnoise_tokens = tokens_length - num_noise_tokens num_noise_spans = int(round(num_noise_tokens / mean_noise_span_length)) # inputs contain all nonnoise tokens, sentinels for all noise spans # and one EOS token. return ( num_nonnoise_tokens + num_noise_spans * extra_tokens_per_span_inputs + 1, num_noise_tokens + num_noise_spans * extra_tokens_per_span_targets + 1) tokens_length = inputs_length - 1 while (_tokens_length_to_inputs_length_targets_length(tokens_length + 1)[0] <= inputs_length): tokens_length += 1 inputs_length, targets_length = ( _tokens_length_to_inputs_length_targets_length(tokens_length)) # minor hack to get the targets length to be equal to inputs length # which is more likely to have been set to a nice round number. if noise_density == 0.5 and targets_length > inputs_length: tokens_length -= 1 targets_length -= 1 if verbose: logging.info( 'tokens_length=%s inputs_length=%s targets_length=%s ' 'noise_density=%s mean_noise_span_length=%s ', tokens_length, inputs_length, targets_length, noise_density, mean_noise_span_length) return tokens_length, targets_length def denoise(dataset, output_features, noise_density=gin.REQUIRED, noise_mask_fn=gin.REQUIRED, inputs_fn=gin.REQUIRED, targets_fn=None, passthrough_feature_keys: Optional[Sequence[str]] = None, input_feature_key='inputs', **unused_kwargs): """SeqIO wrapper for single_example_denoise().""" def my_fn(features, seed): return single_example_denoise( features, seed, output_features=output_features, noise_density=noise_density, noise_mask_fn=noise_mask_fn, inputs_fn=inputs_fn, targets_fn=targets_fn, passthrough_feature_keys=passthrough_feature_keys, input_feature_key=input_feature_key) return my_fn(dataset) def random_spans_noise_mask(length, noise_density, seeds, mean_noise_span_length=3.0, random_roll=False): """Noise mask consisting of random spans of noise tokens. The number of noise tokens and the number of noise spans and non-noise spans are determined deterministically as follows: num_noise_tokens = round(length * noise_density) num_nonnoise_spans = num_noise_spans = round( num_noise_tokens / mean_noise_span_length) Spans alternate between non-noise and noise, beginning with non-noise. Subject to the above restrictions, all masks are equally likely. Args: length: an int32 scalar (length of the incoming token sequence) noise_density: a float - approximate density of output mask seeds: an int32 Tensor, shaped (2, 2) mean_noise_span_length: a number random_roll: bool, whether to roll the mask by a random integer offset in [0, length). Set random_roll to True to get a more uniform distribution of masked positions. Specifically, when random_roll is False (default) and a single span is enough to satisfy the noise density requirement, this fuction masks only the last few positions. Returns: a boolean tensor with shape [length] """ if noise_density == 0.0: return tf.zeros(length, tf.bool) orig_length = length # increase length to avoid degeneracy length = tf.maximum(length, 2) def to_int(x): return tf.cast(x, tf.int32) def to_float(x): return tf.cast(x, tf.float32) num_noise_tokens = to_int(tf.round(to_float(length) * noise_density)) # avoid degeneracy by ensuring positive numbers of noise and nonnoise tokens. num_noise_tokens = tf.minimum(tf.maximum(num_noise_tokens, 1), length - 1) num_noise_spans = to_int( tf.round(to_float(num_noise_tokens) / mean_noise_span_length)) # avoid degeneracy by ensuring positive number of noise spans num_noise_spans = tf.maximum(num_noise_spans, 1) num_nonnoise_tokens = length - num_noise_tokens # pick the lengths of the noise spans and the non-noise spans def _random_segmentation(num_items, num_segments, seed): """Partition a sequence of items randomly into non-empty segments. Args: num_items: an integer scalar > 0 num_segments: an integer scalar in [1, num_items] seed: an integer seed Returns: a Tensor with shape [num_segments] containing positive integers that add up to num_items """ first_in_segment = tf.pad( seqio.stateless_shuffle( to_int(tf.range(num_items - 1) < num_segments - 1), seed), [[1, 0]]) segment_id = tf.cumsum(first_in_segment) segment_length = tf.math.segment_sum(tf.ones_like(segment_id), segment_id) return segment_length noise_span_lengths = _random_segmentation( num_noise_tokens, num_noise_spans, seeds[0]) nonnoise_span_lengths = _random_segmentation( num_nonnoise_tokens, num_noise_spans, seeds[1]) interleaved_span_lengths = tf.reshape( tf.stack([nonnoise_span_lengths, noise_span_lengths], axis=1), [num_noise_spans * 2]) span_starts = tf.cumsum(interleaved_span_lengths)[:-1] span_start_indicator = tf.math.unsorted_segment_sum( tf.ones_like(span_starts), span_starts, length) span_num = tf.cumsum(span_start_indicator) is_noise = tf.equal(span_num % 2, 1) mask = is_noise[:orig_length] if random_roll: roll_seed = (seeds[0][0]+seeds[1][1], seeds[0][1]-seeds[1][0]) # new seed. # Roll the mask by a random offset e.g. for offset=2: [1,2,3,4] => [3,4,1,2] offset = tf.random.stateless_uniform( [1], seed=roll_seed, dtype=tf.int32, minval=0, maxval=length)[0] mask = tf.roll(mask, shift=offset, axis=0) return mask def noise_span_to_unique_sentinel(tokens, noise_mask, vocabulary, seeds): """Replace each run of consecutive noise tokens with a different sentinel. The idea here is to be able to align the dropped spans in the inputs with the markers in the targets. We want to generate training examples like "We hold X to be Y that" -> "X these truths Y self evident Z" Sentinels assigned in decreasing order within the sequence starting at vocabulary.size - 1. That is, we appropriate the last tokens in the vocabulary for additional use as sentinels. TODO(noam): we may want to try enlarging the vocabulary and leaving room for the sentinels instead. However, this requires enlarging the embedding tables in the model, so that is a bigger change. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an unused int32 Tensor Returns: a Tensor with the same shape and dtype as tokens """ del seeds prev_token_is_noise = tf.pad(noise_mask[:-1], [[1, 0]]) first_noise_tokens = tf.logical_and( noise_mask, tf.logical_not(prev_token_is_noise)) subsequent_noise_tokens = tf.logical_and(noise_mask, prev_token_is_noise) sentinel = sentinel_id(vocabulary) + 1 - tf.cumsum( tf.cast(first_noise_tokens, tokens.dtype)) tokens = tf.where(first_noise_tokens, sentinel, tokens) return tf.boolean_mask(tokens, tf.logical_not(subsequent_noise_tokens)) def nonnoise_span_to_unique_sentinel(tokens, noise_mask, vocabulary, seeds): return noise_span_to_unique_sentinel( tokens, tf.logical_not(noise_mask), vocabulary, seeds) The provided code snippet includes necessary dependencies for implementing the `span_corruption` function. Write a Python function `def span_corruption(dataset, sequence_length, output_features, mean_noise_span_length=3.0, noise_density=0.15, input_feature_key='inputs', merge_examples_to_reduce_padding=True, reserved_for_packing=None, passthrough_feature_keys: Optional[Sequence[str]] = None)` to solve the following problem: Final pretraining objective used in Raffel et al., 2019. Args: dataset: A tf.data.Dataset with dictionaries containing the key `input_feature_key`. sequence_length: dict mapping of feature key to int length for that feature. output_features: mapping of keys to features. mean_noise_span_length: the mean number of tokens per masked span per example. noise_density: what fraction of the tokens to mask. input_feature_key: which feature to use from the dataset as the input text tokens. merge_examples_to_reduce_padding: if True, combines multiple input examples to reduce padding. reserved_for_packing: if specified, reduces the desired inputs length by the specified amount to enable multiple examples to be packed together downstream. passthrough_feature_keys: a sequence of feature names that should be passed through to the output of this preprocessor. eg: ["tokens"]. Only supported if `merge_examples_to_reduce_padding` is set to False. Returns: a dataset Here is the function: def span_corruption(dataset, sequence_length, output_features, mean_noise_span_length=3.0, noise_density=0.15, input_feature_key='inputs', merge_examples_to_reduce_padding=True, reserved_for_packing=None, passthrough_feature_keys: Optional[Sequence[str]] = None): """Final pretraining objective used in Raffel et al., 2019. Args: dataset: A tf.data.Dataset with dictionaries containing the key `input_feature_key`. sequence_length: dict mapping of feature key to int length for that feature. output_features: mapping of keys to features. mean_noise_span_length: the mean number of tokens per masked span per example. noise_density: what fraction of the tokens to mask. input_feature_key: which feature to use from the dataset as the input text tokens. merge_examples_to_reduce_padding: if True, combines multiple input examples to reduce padding. reserved_for_packing: if specified, reduces the desired inputs length by the specified amount to enable multiple examples to be packed together downstream. passthrough_feature_keys: a sequence of feature names that should be passed through to the output of this preprocessor. eg: ["tokens"]. Only supported if `merge_examples_to_reduce_padding` is set to False. Returns: a dataset """ inputs_length = sequence_length[input_feature_key] if reserved_for_packing: inputs_length -= reserved_for_packing input_length, targets_length = random_spans_helper( extra_tokens_per_span_inputs=1, extra_tokens_per_span_targets=1, inputs_length=inputs_length, mean_noise_span_length=mean_noise_span_length, noise_density=noise_density) if sequence_length['targets'] < targets_length: raise ValueError( f'Expected targets length for span corruption ({targets_length}) is ' f'greater than configured targets length ' f"({sequence_length['targets']})") ds = dataset ds = select_random_chunk( ds, output_features=output_features, feature_key='targets', max_length=65536, passthrough_feature_keys=passthrough_feature_keys) if merge_examples_to_reduce_padding: if passthrough_feature_keys: raise ValueError('passthrough_feature_keys not supported with ' 'merge_examples_to_reduce_padding=True. ' f'Got: {passthrough_feature_keys}') ds = reduce_concat_tokens(ds, feature_key='targets', batch_size=128) ds = split_tokens( ds, feature_key='targets', min_tokens_per_segment=None, max_tokens_per_segment=input_length, passthrough_feature_keys=passthrough_feature_keys) ds = denoise( ds, output_features, inputs_fn=noise_span_to_unique_sentinel, targets_fn=nonnoise_span_to_unique_sentinel, noise_density=noise_density, noise_mask_fn=functools.partial( random_spans_noise_mask, mean_noise_span_length=mean_noise_span_length), input_feature_key=input_feature_key, passthrough_feature_keys=passthrough_feature_keys) return ds
Final pretraining objective used in Raffel et al., 2019. Args: dataset: A tf.data.Dataset with dictionaries containing the key `input_feature_key`. sequence_length: dict mapping of feature key to int length for that feature. output_features: mapping of keys to features. mean_noise_span_length: the mean number of tokens per masked span per example. noise_density: what fraction of the tokens to mask. input_feature_key: which feature to use from the dataset as the input text tokens. merge_examples_to_reduce_padding: if True, combines multiple input examples to reduce padding. reserved_for_packing: if specified, reduces the desired inputs length by the specified amount to enable multiple examples to be packed together downstream. passthrough_feature_keys: a sequence of feature names that should be passed through to the output of this preprocessor. eg: ["tokens"]. Only supported if `merge_examples_to_reduce_padding` is set to False. Returns: a dataset
56
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def select_random_chunk(dataset: tf.data.Dataset, output_features: Mapping[str, seqio.Feature], max_length: Optional[int] = None, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[ Sequence[str]] = None, sequence_length: Optional[Mapping[str, int]] = None, uniform_random_start: bool = False, min_length: Optional[int] = None, **unused_kwargs) -> tf.data.Dataset: """SeqIO wrapper for single_example_select_random_chunk().""" def _my_fn(x, seed): return single_example_select_random_chunk( x, seed, output_features=output_features, max_length=max_length, feature_key=feature_key, additional_feature_keys=additional_feature_keys, passthrough_feature_keys=passthrough_feature_keys, sequence_length=sequence_length, uniform_random_start=uniform_random_start, min_length=min_length) # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) return _my_fn(dataset) def reduce_concat_tokens(dataset, feature_key='targets', batch_size=128, **unused_kwargs): """Token-preprocessor to concatenate multiple unrelated documents. If we want to generate examples of exactly the right length, (to avoid wasting space on padding), then we use this function, folowed by split_tokens. Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. feature_key: an string batch_size: an integer - how many documents to concatenate into one Returns: a dataset """ dataset = dataset.map( lambda x: {feature_key: x[feature_key]}, num_parallel_calls=AUTOTUNE) dataset = dataset.padded_batch(batch_size, padded_shapes={feature_key: [-1]}) def _my_fn(x): tokens = tf.reshape(x[feature_key], [-1]) # strip padding tokens = tf.boolean_mask(tokens, tf.cast(tokens, tf.bool)) return {feature_key: tokens} return dataset.map(_my_fn, num_parallel_calls=AUTOTUNE) def split_tokens_to_inputs_length(dataset, sequence_length, output_features, **kwargs): max_tokens = sequence_length['inputs'] if output_features['inputs'].add_eos: # Leave room to insert an EOS token. max_tokens -= 1 return split_tokens(dataset, max_tokens_per_segment=max_tokens, **kwargs) def denoise(dataset, output_features, noise_density=gin.REQUIRED, noise_mask_fn=gin.REQUIRED, inputs_fn=gin.REQUIRED, targets_fn=None, passthrough_feature_keys: Optional[Sequence[str]] = None, input_feature_key='inputs', **unused_kwargs): """SeqIO wrapper for single_example_denoise().""" def my_fn(features, seed): return single_example_denoise( features, seed, output_features=output_features, noise_density=noise_density, noise_mask_fn=noise_mask_fn, inputs_fn=inputs_fn, targets_fn=targets_fn, passthrough_feature_keys=passthrough_feature_keys, input_feature_key=input_feature_key) return my_fn(dataset) def iid_noise_mask(length, noise_density, seeds): """Independent and identically distributed token noise. Args: length: an int32 scalar. noise_density: a float - approximate density of output mask. seeds: an int32 Tensor, shaped (1, 2), the random seed. Returns: a boolean tensor with shape [length]. """ return tf.random.stateless_uniform([length], seed=seeds[0]) < noise_density def noise_span_to_unique_sentinel(tokens, noise_mask, vocabulary, seeds): """Replace each run of consecutive noise tokens with a different sentinel. The idea here is to be able to align the dropped spans in the inputs with the markers in the targets. We want to generate training examples like "We hold X to be Y that" -> "X these truths Y self evident Z" Sentinels assigned in decreasing order within the sequence starting at vocabulary.size - 1. That is, we appropriate the last tokens in the vocabulary for additional use as sentinels. TODO(noam): we may want to try enlarging the vocabulary and leaving room for the sentinels instead. However, this requires enlarging the embedding tables in the model, so that is a bigger change. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an unused int32 Tensor Returns: a Tensor with the same shape and dtype as tokens """ del seeds prev_token_is_noise = tf.pad(noise_mask[:-1], [[1, 0]]) first_noise_tokens = tf.logical_and( noise_mask, tf.logical_not(prev_token_is_noise)) subsequent_noise_tokens = tf.logical_and(noise_mask, prev_token_is_noise) sentinel = sentinel_id(vocabulary) + 1 - tf.cumsum( tf.cast(first_noise_tokens, tokens.dtype)) tokens = tf.where(first_noise_tokens, sentinel, tokens) return tf.boolean_mask(tokens, tf.logical_not(subsequent_noise_tokens)) def nonnoise_span_to_unique_sentinel(tokens, noise_mask, vocabulary, seeds): return noise_span_to_unique_sentinel( tokens, tf.logical_not(noise_mask), vocabulary, seeds) The provided code snippet includes necessary dependencies for implementing the `iid_denoising` function. Write a Python function `def iid_denoising(dataset, sequence_length, output_features)` to solve the following problem: Baseline pretraining objective used in Raffel et al., 2019. Here is the function: def iid_denoising(dataset, sequence_length, output_features): """Baseline pretraining objective used in Raffel et al., 2019.""" ds = dataset ds = select_random_chunk(ds, output_features=output_features, feature_key='targets', max_length=65536) ds = reduce_concat_tokens(ds, feature_key='targets', batch_size=128) ds = split_tokens_to_inputs_length(ds, output_features=output_features, sequence_length=sequence_length) ds = denoise( ds, output_features, inputs_fn=noise_span_to_unique_sentinel, targets_fn=nonnoise_span_to_unique_sentinel, noise_density=0.15, noise_mask_fn=iid_noise_mask ) return ds
Baseline pretraining objective used in Raffel et al., 2019.
57
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def select_random_chunk(dataset: tf.data.Dataset, output_features: Mapping[str, seqio.Feature], max_length: Optional[int] = None, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[ Sequence[str]] = None, sequence_length: Optional[Mapping[str, int]] = None, uniform_random_start: bool = False, min_length: Optional[int] = None, **unused_kwargs) -> tf.data.Dataset: """SeqIO wrapper for single_example_select_random_chunk().""" def _my_fn(x, seed): return single_example_select_random_chunk( x, seed, output_features=output_features, max_length=max_length, feature_key=feature_key, additional_feature_keys=additional_feature_keys, passthrough_feature_keys=passthrough_feature_keys, sequence_length=sequence_length, uniform_random_start=uniform_random_start, min_length=min_length) # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) return _my_fn(dataset) def split_tokens_to_inputs_length(dataset, sequence_length, output_features, **kwargs): max_tokens = sequence_length['inputs'] if output_features['inputs'].add_eos: # Leave room to insert an EOS token. max_tokens -= 1 return split_tokens(dataset, max_tokens_per_segment=max_tokens, **kwargs) def denoise(dataset, output_features, noise_density=gin.REQUIRED, noise_mask_fn=gin.REQUIRED, inputs_fn=gin.REQUIRED, targets_fn=None, passthrough_feature_keys: Optional[Sequence[str]] = None, input_feature_key='inputs', **unused_kwargs): """SeqIO wrapper for single_example_denoise().""" def my_fn(features, seed): return single_example_denoise( features, seed, output_features=output_features, noise_density=noise_density, noise_mask_fn=noise_mask_fn, inputs_fn=inputs_fn, targets_fn=targets_fn, passthrough_feature_keys=passthrough_feature_keys, input_feature_key=input_feature_key) return my_fn(dataset) def random_prefix_noise_mask(length, noise_density, seeds): """First part of the sequence is noise (for prefix_lm). The length of the prefix is chosen uniformly between [1, length) noise_density must be 0.5. Args: length: an int32 scalar. noise_density: a float - must not exceed 0.5. seeds: an int32 Tensor, shaped (1, 2), the random seed. Returns: a boolean tensor with shape [length]. """ if noise_density > 0.5: raise NotImplementedError( 'noise density must not exceed 0.5 for random_prefix_noise_mask') max_input_tokens = length - 1 min_input_tokens = tf.minimum( max_input_tokens, tf.maximum( 1, tf.cast( tf.math.round((1 - 2 * noise_density) * tf.cast(max_input_tokens, tf.float32)), tf.int32))) num_input_tokens = tf.random.stateless_uniform( [], minval=min_input_tokens, maxval=max_input_tokens + 1, dtype=tf.int32, seed=seeds[0]) return tf.range(length, dtype=tf.int32) < num_input_tokens def drop_noise_tokens(tokens, noise_mask, vocabulary, seeds): """Drop noise tokens without inserting a sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an unused int32 Tensor Returns: a Tensor with the same shape and dtype as tokens """ del vocabulary del seeds return tf.boolean_mask(tokens, tf.logical_not(noise_mask)) def drop_nonnoise_tokens(tokens, noise_mask, vocabulary, seeds): """Drop non-noise tokens without inserting a sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an unused int32 Tensor Returns: a Tensor with the same shape and dtype as tokens """ del vocabulary del seeds return tf.boolean_mask(tokens, noise_mask) The provided code snippet includes necessary dependencies for implementing the `prefix_lm` function. Write a Python function `def prefix_lm(dataset, sequence_length, output_features, noise_density=0.5)` to solve the following problem: Prefix language modeling objective used in Raffel et al. 2019. Here is the function: def prefix_lm(dataset, sequence_length, output_features, noise_density=0.5): """Prefix language modeling objective used in Raffel et al. 2019.""" ds = dataset ds = select_random_chunk(ds, output_features=output_features, feature_key='targets', max_length=65536) ds = split_tokens_to_inputs_length(ds, output_features=output_features, sequence_length=sequence_length) ds = denoise( ds, output_features, inputs_fn=drop_nonnoise_tokens, targets_fn=drop_noise_tokens, noise_density=noise_density, noise_mask_fn=random_prefix_noise_mask, ) return ds
Prefix language modeling objective used in Raffel et al. 2019.
58
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def select_random_chunk(dataset: tf.data.Dataset, output_features: Mapping[str, seqio.Feature], max_length: Optional[int] = None, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[ Sequence[str]] = None, sequence_length: Optional[Mapping[str, int]] = None, uniform_random_start: bool = False, min_length: Optional[int] = None, **unused_kwargs) -> tf.data.Dataset: """SeqIO wrapper for single_example_select_random_chunk().""" def _my_fn(x, seed): return single_example_select_random_chunk( x, seed, output_features=output_features, max_length=max_length, feature_key=feature_key, additional_feature_keys=additional_feature_keys, passthrough_feature_keys=passthrough_feature_keys, sequence_length=sequence_length, uniform_random_start=uniform_random_start, min_length=min_length) # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) return _my_fn(dataset) def reduce_concat_tokens(dataset, feature_key='targets', batch_size=128, **unused_kwargs): """Token-preprocessor to concatenate multiple unrelated documents. If we want to generate examples of exactly the right length, (to avoid wasting space on padding), then we use this function, folowed by split_tokens. Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. feature_key: an string batch_size: an integer - how many documents to concatenate into one Returns: a dataset """ dataset = dataset.map( lambda x: {feature_key: x[feature_key]}, num_parallel_calls=AUTOTUNE) dataset = dataset.padded_batch(batch_size, padded_shapes={feature_key: [-1]}) def _my_fn(x): tokens = tf.reshape(x[feature_key], [-1]) # strip padding tokens = tf.boolean_mask(tokens, tf.cast(tokens, tf.bool)) return {feature_key: tokens} return dataset.map(_my_fn, num_parallel_calls=AUTOTUNE) def split_tokens(dataset: tf.data.Dataset, min_tokens_per_segment: Optional[int] = None, max_tokens_per_segment: int = gin.REQUIRED, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[Sequence[str]] = None, **unused_kwargs) -> tf.data.Dataset: """Split examples into multiple examples each. The intended use case is to break up long examples for use in unsupervised transfer-learning. This function is generally preceded by select_random_chunk. If min_tokens_per_segment is provided, the segment length is chosen randomly per document from a log-uniform distribution. If min_tokens_per_segment is None, then the segment length is max_tokens_per_segment (except for a possibly shorter last segment in each document). Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. min_tokens_per_segment: an optional integer max_tokens_per_segment: an integer, the maximum number of tokens in each segment. Only the final segment may be shorter. feature_key: a string, the feature to split additional_feature_keys: Additional features to split. The same chunk size will be used, so they should be the same size as feature_key. passthrough_feature_keys: Features to pass through without any splitting. Returns: a dataset """ if passthrough_feature_keys: split_keys = set([feature_key] + (additional_feature_keys or [])) overlap_keys = split_keys & set(passthrough_feature_keys) if overlap_keys: raise ValueError( f'split keys {overlap_keys} also included in passthrough keys') def _split_tokens(x, seed): """Split one token sequence into multiple sequences.""" tokens = x[feature_key] n_tokens = tf.shape(tokens)[0] if min_tokens_per_segment is None: length = max_tokens_per_segment else: # pick a length - log-uniformly distributed length = tf.cast( tf.exp( tf.random.stateless_uniform( [], minval=math.log(min_tokens_per_segment), maxval=math.log(max_tokens_per_segment), seed=seed ) ), tf.int32) # Pad to a multiple of length, then use tf.reshape to split up the tokens # into num_segments segments each of the given length. num_segments = tf.cast( tf.math.ceil( tf.cast(n_tokens, tf.float32) / tf.cast(length, tf.float32)) , tf.int32) padding = num_segments * length - tf.shape(tokens)[0] feature_keys_to_split = [feature_key] orig_lengths = {} outputs = {} if additional_feature_keys is not None: feature_keys_to_split.extend(additional_feature_keys) for k in feature_keys_to_split: with tf.control_dependencies([ tf.assert_equal( tf.shape(tokens)[0], tf.shape(x[k])[0], message=(f'Additional feature {k} is not the same size as ' f'{feature_key} along axis 0 in split_tokens().') ) ]): shape = tf.shape(x[k])[1:] shape_list = x[k].shape[1:] padded = tf.pad( x[k], tf.concat([[[0, padding]], tf.zeros([len(shape_list), 2], dtype=tf.int32)], axis=0)) orig_lengths[k] = tf.concat( [tf.repeat(length, num_segments - 1), [length - padding]], axis=0) outputs[k] = tf.reshape( padded, tf.concat([[-1, length], shape], axis=0)) # To avoid memory issues, don't just replicate the passthrough features # for every segment; use tf.data to do it so the copies don't get # instantiated all at once. outputs_ds = tf.data.Dataset.from_tensor_slices(outputs) orig_lengths_ds = tf.data.Dataset.from_tensor_slices(orig_lengths) if passthrough_feature_keys: passthrough = {k: v for k, v in x.items() if k in passthrough_feature_keys} passthrough_ds = tf.data.Dataset.from_tensors(passthrough).repeat( tf.cast(num_segments, tf.int64)) return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds, passthrough_ds)) else: return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds)) def _strip_padding_and_merge_passthrough( inputs, orig_lengths, passthrough=None): output = {} for k, v in inputs.items(): output[k] = v[:orig_lengths[k]] if passthrough: for k, v in passthrough.items(): output[k] = passthrough[k] return output # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) dataset = _split_tokens(dataset).flat_map(lambda z: z) dataset = dataset.map( _strip_padding_and_merge_passthrough, num_parallel_calls=AUTOTUNE) return dataset The provided code snippet includes necessary dependencies for implementing the `full_lm` function. Write a Python function `def full_lm(dataset, sequence_length, output_features)` to solve the following problem: Full language modeling objective with EOS only at document boundaries. Here is the function: def full_lm(dataset, sequence_length, output_features): """Full language modeling objective with EOS only at document boundaries.""" ds = dataset ds = select_random_chunk(ds, output_features=output_features, feature_key='targets', max_length=65536) ds = seqio.preprocessors.append_eos(ds, output_features) ds = reduce_concat_tokens(ds, feature_key='targets', batch_size=128) # Don't use `split_tokens_to_targets_length` since we've alrady added EOS. ds = split_tokens(ds, max_tokens_per_segment=sequence_length['targets']) return ds
Full language modeling objective with EOS only at document boundaries.
59
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `trim_tokens_at_front` function. Write a Python function `def trim_tokens_at_front(x, sequence_length, keys_to_trim=None, **unused_kwargs)` to solve the following problem: Token-preprocessor to trim sequence at the beginning. Args: x: an example with dictionaries containing keys_to_trim. sequence_length: a dict of ints. keys_to_trim: a list of feature keys. Returns: A preprocessed example. Here is the function: def trim_tokens_at_front(x, sequence_length, keys_to_trim=None, **unused_kwargs): """Token-preprocessor to trim sequence at the beginning. Args: x: an example with dictionaries containing keys_to_trim. sequence_length: a dict of ints. keys_to_trim: a list of feature keys. Returns: A preprocessed example. """ for key in (keys_to_trim or sequence_length.keys()): if key in x: # trim tokens, leaving room for EOS which gets added later x[key] = x[key][-(sequence_length[key] - 1):] return x
Token-preprocessor to trim sequence at the beginning. Args: x: an example with dictionaries containing keys_to_trim. sequence_length: a dict of ints. keys_to_trim: a list of feature keys. Returns: A preprocessed example.
60
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `trivia_qa_truncate_inputs` function. Write a Python function `def trivia_qa_truncate_inputs(dataset, output_features, sequence_length)` to solve the following problem: Token preprocessor for the trivia QA dataset to truncate inputs. This function takes a dataset containing "targets" and "inputs". It searches for the "targets" in the "inputs" and truncates the "inputs" to `sequence_length` while ensuring that the "targets" are present in the "inputs". The function will randomly select a subset of "inputs". If "targets" are not found in the "inputs", then the example is is dropped from the dataset. E.g. Input dataset { "inputs": [0, 3, 5, 7, 9, 11, 13, 15, 17, 18] "targets": [5, 7, 9] } Output dataset (assuming sequence_length['inputs'] = 4) { "inputs": [3, 5, 7, 9] "targets": [5, 7, 9] } or { "inputs": [5, 7, 9, 11] "targets": [5, 7, 9] } Args: dataset: a tf.data.Dataset with dictionaries containing the "inputs" and "targets". output_features: unused by this function. sequence_length: a dict, with keys as "inputs" and "targets" indicating the maximum number of tokens in each of the sequences. Returns: a dataset Here is the function: def trivia_qa_truncate_inputs(dataset, output_features, sequence_length): """Token preprocessor for the trivia QA dataset to truncate inputs. This function takes a dataset containing "targets" and "inputs". It searches for the "targets" in the "inputs" and truncates the "inputs" to `sequence_length` while ensuring that the "targets" are present in the "inputs". The function will randomly select a subset of "inputs". If "targets" are not found in the "inputs", then the example is is dropped from the dataset. E.g. Input dataset { "inputs": [0, 3, 5, 7, 9, 11, 13, 15, 17, 18] "targets": [5, 7, 9] } Output dataset (assuming sequence_length['inputs'] = 4) { "inputs": [3, 5, 7, 9] "targets": [5, 7, 9] } or { "inputs": [5, 7, 9, 11] "targets": [5, 7, 9] } Args: dataset: a tf.data.Dataset with dictionaries containing the "inputs" and "targets". output_features: unused by this function. sequence_length: a dict, with keys as "inputs" and "targets" indicating the maximum number of tokens in each of the sequences. Returns: a dataset """ del output_features @seqio.map_over_dataset(num_seeds=1) def my_fn(features, seed): """Function to map original dataset to the new dataset.""" inputs = features['inputs'] targets = features['targets'] ans_len = tf.shape(targets)[0] max_input_tokens = sequence_length['inputs'] def truncate_inputs(): """Helper function to truncate the inputs.""" def answer_in_context(context, answer): """Helper function that checks if the answer is present in the context. Args: context: Tensor, tokenized representation of the context answer: Tensor, tokenized representation of the answer Returns: result: boolean, indicates if the answer was present in the context. pos_mask: boolean mask, a mask for every possible start position of the answer in the context. Indicates whether the answer starts at the particular position. """ conv_inp = tf.reshape(tf.cast(context, tf.float32), [1, -1, 1]) ans_len = tf.shape(answer)[0] filters = tf.eye(ans_len, dtype=tf.float32) # Assume context len is N and answer len is M. # Use a convolution to create a matrix of (N-M) x M elements where # each row of the matrix is a sequence of len M. This matrix contains # all possible contiguous sequences of length M from the context. # Every row of this matrix is compared with the answer to check if the # answer exists in the context. strided = tf.nn.conv1d(conv_inp, tf.reshape(filters, [ans_len, 1, ans_len]), 1, 'VALID') strided = tf.cast(strided[0], answer.dtype) pos_mask = tf.reduce_all( tf.equal(strided, tf.reshape(answer, [1, -1])), 1) result = tf.reduce_any(pos_mask) return result, pos_mask def slice_inputs(inputs, answer_len, pos_mask, seed=None): """Helper function to slice inputs while keeping the answer.""" ans_start_pos = tf.cast(tf.where(pos_mask)[0][0], tf.int32) inputs_len = tf.shape(inputs)[0] start_range_min = tf.maximum( 0, ans_start_pos - (max_input_tokens - answer_len)) start_range_max = tf.minimum(ans_start_pos, inputs_len - max_input_tokens) + 1 start_pos = tf.random.stateless_uniform( [], minval=start_range_min, maxval=start_range_max, dtype=tf.int32, seed=seed) return inputs[start_pos:start_pos + max_input_tokens] result, pos_mask = answer_in_context(inputs, targets) if result: return slice_inputs(inputs, ans_len, pos_mask, seed=seed) else: return tf.constant([], dtype=inputs.dtype) if tf.greater(tf.shape(inputs)[0], max_input_tokens): inputs = truncate_inputs() return {'inputs': inputs, 'targets': features['targets']} dataset = my_fn(dataset) return dataset.filter(lambda x: tf.size(x['inputs']) > 0)
Token preprocessor for the trivia QA dataset to truncate inputs. This function takes a dataset containing "targets" and "inputs". It searches for the "targets" in the "inputs" and truncates the "inputs" to `sequence_length` while ensuring that the "targets" are present in the "inputs". The function will randomly select a subset of "inputs". If "targets" are not found in the "inputs", then the example is is dropped from the dataset. E.g. Input dataset { "inputs": [0, 3, 5, 7, 9, 11, 13, 15, 17, 18] "targets": [5, 7, 9] } Output dataset (assuming sequence_length['inputs'] = 4) { "inputs": [3, 5, 7, 9] "targets": [5, 7, 9] } or { "inputs": [5, 7, 9, 11] "targets": [5, 7, 9] } Args: dataset: a tf.data.Dataset with dictionaries containing the "inputs" and "targets". output_features: unused by this function. sequence_length: a dict, with keys as "inputs" and "targets" indicating the maximum number of tokens in each of the sequences. Returns: a dataset
61
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `unsupervised` function. Write a Python function `def unsupervised(dataset, preprocessors=None, output_features=None, sequence_length=None)` to solve the following problem: Configure this to point at unsupervised preprocessors. This function creates an extra level of indirection in case we want different unsupervised pretraining functions in the future which do not fit into the denoise() framework. This function should be used as a post-cache preprocessing function. Args: dataset: A tf.data.Dataset to process. preprocessors: a list of token-preprocessor functions. These functions should take unused kwargs if output_features or sequence_length is not used. output_features: dict(str, Feature), output features of the Task to be passed to the model. sequence_length: dict mapping feature key to int length for that feature. Returns: A preprocessed tf.data.Dataset. Here is the function: def unsupervised(dataset, preprocessors=None, output_features=None, sequence_length=None): """Configure this to point at unsupervised preprocessors. This function creates an extra level of indirection in case we want different unsupervised pretraining functions in the future which do not fit into the denoise() framework. This function should be used as a post-cache preprocessing function. Args: dataset: A tf.data.Dataset to process. preprocessors: a list of token-preprocessor functions. These functions should take unused kwargs if output_features or sequence_length is not used. output_features: dict(str, Feature), output features of the Task to be passed to the model. sequence_length: dict mapping feature key to int length for that feature. Returns: A preprocessed tf.data.Dataset. """ if preprocessors is None: logging.warning( 'unsupervised preprocessor got preprocessors=None; no preprocessing ' 'will be applied.' ) return dataset kwargs = {} if output_features: kwargs['output_features'] = output_features if sequence_length: kwargs['sequence_length'] = sequence_length for p in preprocessors: dataset = p(dataset, **kwargs) return dataset
Configure this to point at unsupervised preprocessors. This function creates an extra level of indirection in case we want different unsupervised pretraining functions in the future which do not fit into the denoise() framework. This function should be used as a post-cache preprocessing function. Args: dataset: A tf.data.Dataset to process. preprocessors: a list of token-preprocessor functions. These functions should take unused kwargs if output_features or sequence_length is not used. output_features: dict(str, Feature), output features of the Task to be passed to the model. sequence_length: dict mapping feature key to int length for that feature. Returns: A preprocessed tf.data.Dataset.
62
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def split_tokens(dataset: tf.data.Dataset, min_tokens_per_segment: Optional[int] = None, max_tokens_per_segment: int = gin.REQUIRED, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[Sequence[str]] = None, **unused_kwargs) -> tf.data.Dataset: def split_tokens_to_targets_length(dataset, sequence_length, output_features, **kwargs): max_tokens = sequence_length['targets'] if output_features['targets'].add_eos: # Leave room to insert an EOS token. max_tokens -= 1 return split_tokens(dataset, max_tokens_per_segment=max_tokens, **kwargs)
null
63
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def split_tokens(dataset: tf.data.Dataset, min_tokens_per_segment: Optional[int] = None, max_tokens_per_segment: int = gin.REQUIRED, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[Sequence[str]] = None, **unused_kwargs) -> tf.data.Dataset: """Split examples into multiple examples each. The intended use case is to break up long examples for use in unsupervised transfer-learning. This function is generally preceded by select_random_chunk. If min_tokens_per_segment is provided, the segment length is chosen randomly per document from a log-uniform distribution. If min_tokens_per_segment is None, then the segment length is max_tokens_per_segment (except for a possibly shorter last segment in each document). Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. min_tokens_per_segment: an optional integer max_tokens_per_segment: an integer, the maximum number of tokens in each segment. Only the final segment may be shorter. feature_key: a string, the feature to split additional_feature_keys: Additional features to split. The same chunk size will be used, so they should be the same size as feature_key. passthrough_feature_keys: Features to pass through without any splitting. Returns: a dataset """ if passthrough_feature_keys: split_keys = set([feature_key] + (additional_feature_keys or [])) overlap_keys = split_keys & set(passthrough_feature_keys) if overlap_keys: raise ValueError( f'split keys {overlap_keys} also included in passthrough keys') def _split_tokens(x, seed): """Split one token sequence into multiple sequences.""" tokens = x[feature_key] n_tokens = tf.shape(tokens)[0] if min_tokens_per_segment is None: length = max_tokens_per_segment else: # pick a length - log-uniformly distributed length = tf.cast( tf.exp( tf.random.stateless_uniform( [], minval=math.log(min_tokens_per_segment), maxval=math.log(max_tokens_per_segment), seed=seed ) ), tf.int32) # Pad to a multiple of length, then use tf.reshape to split up the tokens # into num_segments segments each of the given length. num_segments = tf.cast( tf.math.ceil( tf.cast(n_tokens, tf.float32) / tf.cast(length, tf.float32)) , tf.int32) padding = num_segments * length - tf.shape(tokens)[0] feature_keys_to_split = [feature_key] orig_lengths = {} outputs = {} if additional_feature_keys is not None: feature_keys_to_split.extend(additional_feature_keys) for k in feature_keys_to_split: with tf.control_dependencies([ tf.assert_equal( tf.shape(tokens)[0], tf.shape(x[k])[0], message=(f'Additional feature {k} is not the same size as ' f'{feature_key} along axis 0 in split_tokens().') ) ]): shape = tf.shape(x[k])[1:] shape_list = x[k].shape[1:] padded = tf.pad( x[k], tf.concat([[[0, padding]], tf.zeros([len(shape_list), 2], dtype=tf.int32)], axis=0)) orig_lengths[k] = tf.concat( [tf.repeat(length, num_segments - 1), [length - padding]], axis=0) outputs[k] = tf.reshape( padded, tf.concat([[-1, length], shape], axis=0)) # To avoid memory issues, don't just replicate the passthrough features # for every segment; use tf.data to do it so the copies don't get # instantiated all at once. outputs_ds = tf.data.Dataset.from_tensor_slices(outputs) orig_lengths_ds = tf.data.Dataset.from_tensor_slices(orig_lengths) if passthrough_feature_keys: passthrough = {k: v for k, v in x.items() if k in passthrough_feature_keys} passthrough_ds = tf.data.Dataset.from_tensors(passthrough).repeat( tf.cast(num_segments, tf.int64)) return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds, passthrough_ds)) else: return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds)) def _strip_padding_and_merge_passthrough( inputs, orig_lengths, passthrough=None): output = {} for k, v in inputs.items(): output[k] = v[:orig_lengths[k]] if passthrough: for k, v in passthrough.items(): output[k] = passthrough[k] return output # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) dataset = _split_tokens(dataset).flat_map(lambda z: z) dataset = dataset.map( _strip_padding_and_merge_passthrough, num_parallel_calls=AUTOTUNE) return dataset def split_tokens_to_random_length(dataset, sequence_length, output_features, **kwargs): max_tokens = sequence_length['inputs'] if output_features['inputs'].add_eos: # Leave room to insert an EOS token. max_tokens -= 1 return split_tokens(dataset, min_tokens_per_segment=8, max_tokens_per_segment=max_tokens, **kwargs)
null
64
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `concatenate_and_split_to_fixed_length` function. Write a Python function `def concatenate_and_split_to_fixed_length(dataset, sequence_length, output_features, feature_key='targets', **unused_kwargs)` to solve the following problem: Concatenate tokens across examples, then split to fixed-size chunks. Chunk length is determined by sequence_length[feature_key]. Args: dataset: a tf.data.Dataset sequence_length: a dict of ints. output_features: a dict mapping feature name to t5.data.Feature. feature_key: a string Returns: a tf.data.Dataset Here is the function: def concatenate_and_split_to_fixed_length(dataset, sequence_length, output_features, feature_key='targets', **unused_kwargs): """Concatenate tokens across examples, then split to fixed-size chunks. Chunk length is determined by sequence_length[feature_key]. Args: dataset: a tf.data.Dataset sequence_length: a dict of ints. output_features: a dict mapping feature name to t5.data.Feature. feature_key: a string Returns: a tf.data.Dataset """ dataset = dataset.map(lambda x: {feature_key: x[feature_key]}) max_tokens = sequence_length[feature_key] if output_features[feature_key].add_eos: # Leave room to insert an EOS token. max_tokens -= 1 return dataset.unbatch().batch(max_tokens)
Concatenate tokens across examples, then split to fixed-size chunks. Chunk length is determined by sequence_length[feature_key]. Args: dataset: a tf.data.Dataset sequence_length: a dict of ints. output_features: a dict mapping feature name to t5.data.Feature. feature_key: a string Returns: a tf.data.Dataset
65
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `filter_by_string_length` function. Write a Python function `def filter_by_string_length(dataset, feature_key='targets', min_length=1, max_length=1000000, **unused_kwargs)` to solve the following problem: Filter examples by string length. Args: dataset: a tf.data.Dataset (not tokenized) feature_key: a string min_length: an integer max_length: an integer Returns: a tf.data.Dataset Here is the function: def filter_by_string_length(dataset, feature_key='targets', min_length=1, max_length=1000000, **unused_kwargs): """Filter examples by string length. Args: dataset: a tf.data.Dataset (not tokenized) feature_key: a string min_length: an integer max_length: an integer Returns: a tf.data.Dataset """ def my_fn(x): l = tf.strings.length(x[feature_key]) return tf.logical_and(tf.greater_equal(l, min_length), tf.less_equal(l, max_length)) return dataset.filter(my_fn)
Filter examples by string length. Args: dataset: a tf.data.Dataset (not tokenized) feature_key: a string min_length: an integer max_length: an integer Returns: a tf.data.Dataset
66
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def random_spans_helper(inputs_length=gin.REQUIRED, noise_density=gin.REQUIRED, mean_noise_span_length=gin.REQUIRED, extra_tokens_per_span_inputs=gin.REQUIRED, extra_tokens_per_span_targets=gin.REQUIRED, verbose=False): """Training parameters to avoid padding with random_spans_noise_mask. When training a model with random_spans_noise_mask, we would like to set the other training hyperparmeters in a way that avoids padding. This function helps us compute these hyperparameters. We assume that each noise span in the input is replaced by extra_tokens_per_span_inputs sentinel tokens, and each non-noise span in the targets is replaced by extra_tokens_per_span_targets sentinel tokens. This function tells us the required number of tokens in the raw example (for split_tokens()) as well as the length of the encoded targets. Note that this function assumes the inputs and targets will have EOS appended and includes that in the reported length. Args: inputs_length: an integer - desired length of the tokenized inputs sequence noise_density: a float mean_noise_span_length: a float extra_tokens_per_span_inputs: an integer extra_tokens_per_span_targets: an integer verbose: a bool indicating whether to log sequence lengths Returns: tokens_length: length of original text in tokens targets_length: an integer - length in tokens of encoded targets sequence """ def _tokens_length_to_inputs_length_targets_length(tokens_length): num_noise_tokens = int(round(tokens_length * noise_density)) num_nonnoise_tokens = tokens_length - num_noise_tokens num_noise_spans = int(round(num_noise_tokens / mean_noise_span_length)) # inputs contain all nonnoise tokens, sentinels for all noise spans # and one EOS token. return ( num_nonnoise_tokens + num_noise_spans * extra_tokens_per_span_inputs + 1, num_noise_tokens + num_noise_spans * extra_tokens_per_span_targets + 1) tokens_length = inputs_length - 1 while (_tokens_length_to_inputs_length_targets_length(tokens_length + 1)[0] <= inputs_length): tokens_length += 1 inputs_length, targets_length = ( _tokens_length_to_inputs_length_targets_length(tokens_length)) # minor hack to get the targets length to be equal to inputs length # which is more likely to have been set to a nice round number. if noise_density == 0.5 and targets_length > inputs_length: tokens_length -= 1 targets_length -= 1 if verbose: logging.info( 'tokens_length=%s inputs_length=%s targets_length=%s ' 'noise_density=%s mean_noise_span_length=%s ', tokens_length, inputs_length, targets_length, noise_density, mean_noise_span_length) return tokens_length, targets_length The provided code snippet includes necessary dependencies for implementing the `random_spans_tokens_length` function. Write a Python function `def random_spans_tokens_length()` to solve the following problem: Helper for gin-configuring split_tokens with random_spans_noise_mask. Here is the function: def random_spans_tokens_length(): """Helper for gin-configuring split_tokens with random_spans_noise_mask.""" return random_spans_helper()[0]
Helper for gin-configuring split_tokens with random_spans_noise_mask.
67
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def random_spans_helper(inputs_length=gin.REQUIRED, noise_density=gin.REQUIRED, mean_noise_span_length=gin.REQUIRED, extra_tokens_per_span_inputs=gin.REQUIRED, extra_tokens_per_span_targets=gin.REQUIRED, verbose=False): """Training parameters to avoid padding with random_spans_noise_mask. When training a model with random_spans_noise_mask, we would like to set the other training hyperparmeters in a way that avoids padding. This function helps us compute these hyperparameters. We assume that each noise span in the input is replaced by extra_tokens_per_span_inputs sentinel tokens, and each non-noise span in the targets is replaced by extra_tokens_per_span_targets sentinel tokens. This function tells us the required number of tokens in the raw example (for split_tokens()) as well as the length of the encoded targets. Note that this function assumes the inputs and targets will have EOS appended and includes that in the reported length. Args: inputs_length: an integer - desired length of the tokenized inputs sequence noise_density: a float mean_noise_span_length: a float extra_tokens_per_span_inputs: an integer extra_tokens_per_span_targets: an integer verbose: a bool indicating whether to log sequence lengths Returns: tokens_length: length of original text in tokens targets_length: an integer - length in tokens of encoded targets sequence """ def _tokens_length_to_inputs_length_targets_length(tokens_length): num_noise_tokens = int(round(tokens_length * noise_density)) num_nonnoise_tokens = tokens_length - num_noise_tokens num_noise_spans = int(round(num_noise_tokens / mean_noise_span_length)) # inputs contain all nonnoise tokens, sentinels for all noise spans # and one EOS token. return ( num_nonnoise_tokens + num_noise_spans * extra_tokens_per_span_inputs + 1, num_noise_tokens + num_noise_spans * extra_tokens_per_span_targets + 1) tokens_length = inputs_length - 1 while (_tokens_length_to_inputs_length_targets_length(tokens_length + 1)[0] <= inputs_length): tokens_length += 1 inputs_length, targets_length = ( _tokens_length_to_inputs_length_targets_length(tokens_length)) # minor hack to get the targets length to be equal to inputs length # which is more likely to have been set to a nice round number. if noise_density == 0.5 and targets_length > inputs_length: tokens_length -= 1 targets_length -= 1 if verbose: logging.info( 'tokens_length=%s inputs_length=%s targets_length=%s ' 'noise_density=%s mean_noise_span_length=%s ', tokens_length, inputs_length, targets_length, noise_density, mean_noise_span_length) return tokens_length, targets_length The provided code snippet includes necessary dependencies for implementing the `random_spans_targets_length` function. Write a Python function `def random_spans_targets_length()` to solve the following problem: Helper for gin-configuring the targets sequence length. Here is the function: def random_spans_targets_length(): """Helper for gin-configuring the targets sequence length.""" return random_spans_helper()[1]
Helper for gin-configuring the targets sequence length.
68
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `regular_noise_mask` function. Write a Python function `def regular_noise_mask(length, noise_density, seeds, min_span_length=1, max_span_length=5)` to solve the following problem: Noise mask consisting of equally spaced spans of equal length. The span length and the offset are chosen randomly per-example. The beginning and end of the sequence may be part of shorter spans of noise. For example, if noise_density=0.25 and a span length of 2 is chosen, then the output might be: [T F F F F F F T T F F F F F F T T F F F F F F T T F F] Args: length: an int32 scalar. noise_density: a float - approximate density of output mask. seeds: an int32 Tensor, shaped (2, 2), the random seeds. min_span_length: an integer. max_span_length: an integer. Returns: a boolean tensor with shape [length]. Here is the function: def regular_noise_mask(length, noise_density, seeds, min_span_length=1, max_span_length=5): """Noise mask consisting of equally spaced spans of equal length. The span length and the offset are chosen randomly per-example. The beginning and end of the sequence may be part of shorter spans of noise. For example, if noise_density=0.25 and a span length of 2 is chosen, then the output might be: [T F F F F F F T T F F F F F F T T F F F F F F T T F F] Args: length: an int32 scalar. noise_density: a float - approximate density of output mask. seeds: an int32 Tensor, shaped (2, 2), the random seeds. min_span_length: an integer. max_span_length: an integer. Returns: a boolean tensor with shape [length]. """ span_length = tf.random.stateless_uniform( [], minval=min_span_length, maxval=max_span_length + 1, dtype=tf.int32, seed=seeds[0]) period = tf.cast( tf.round(tf.cast(span_length, tf.float32) / noise_density), tf.int32) offset = tf.random.stateless_uniform( [], maxval=period, dtype=tf.int32, seed=seeds[1]) return (tf.range(length, dtype=tf.int32) + offset) % period < span_length
Noise mask consisting of equally spaced spans of equal length. The span length and the offset are chosen randomly per-example. The beginning and end of the sequence may be part of shorter spans of noise. For example, if noise_density=0.25 and a span length of 2 is chosen, then the output might be: [T F F F F F F T T F F F F F F T T F F F F F F T T F F] Args: length: an int32 scalar. noise_density: a float - approximate density of output mask. seeds: an int32 Tensor, shaped (2, 2), the random seeds. min_span_length: an integer. max_span_length: an integer. Returns: a boolean tensor with shape [length].
69
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def noise_span_to_sentinel(tokens, noise_mask, vocabulary, seeds): """Replace each run of consecutive noise tokens with a single sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an unused int32 Tensor Returns: a Tensor with the same shape and dtype as tokens """ del seeds tokens = tf.where(noise_mask, tf.cast(sentinel_id(vocabulary), tokens.dtype), tokens) prev_token_is_noise = tf.pad(noise_mask[:-1], [[1, 0]]) subsequent_noise_tokens = tf.logical_and(noise_mask, prev_token_is_noise) return tf.boolean_mask(tokens, tf.logical_not(subsequent_noise_tokens)) def nonnoise_span_to_sentinel(tokens, noise_mask, vocabulary, seeds): return noise_span_to_sentinel( tokens, tf.logical_not(noise_mask), vocabulary, seeds)
null
70
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `permute_noise_tokens` function. Write a Python function `def permute_noise_tokens(tokens, noise_mask, vocabulary, seeds)` to solve the following problem: Permute the noise tokens, keeping the non-noise tokens where they are. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an int32 Tensor, sized (1, 2) Returns: a Tensor with the same shape and dtype as tokens Here is the function: def permute_noise_tokens(tokens, noise_mask, vocabulary, seeds): """Permute the noise tokens, keeping the non-noise tokens where they are. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an int32 Tensor, sized (1, 2) Returns: a Tensor with the same shape and dtype as tokens """ del vocabulary masked_only = tf.boolean_mask(tokens, noise_mask) permuted = seqio.stateless_shuffle(masked_only, seeds[0]) # pad to avoid errors when it has size 0 permuted = tf.pad(permuted, [[0, 1]]) indices = tf.cumsum(tf.cast(noise_mask, tf.int32), exclusive=True) return tf.where(noise_mask, tf.gather(permuted, indices), tokens)
Permute the noise tokens, keeping the non-noise tokens where they are. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an int32 Tensor, sized (1, 2) Returns: a Tensor with the same shape and dtype as tokens
71
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `noise_token_to_gathered_token` function. Write a Python function `def noise_token_to_gathered_token(tokens, noise_mask, vocabulary, seeds)` to solve the following problem: Replace each noise token with a random token from the sequence. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an int32 Tensor, sized (1, 2) Returns: a Tensor with the same shape and dtype as tokens Here is the function: def noise_token_to_gathered_token(tokens, noise_mask, vocabulary, seeds): """Replace each noise token with a random token from the sequence. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an int32 Tensor, sized (1, 2) Returns: a Tensor with the same shape and dtype as tokens """ del vocabulary indices = tf.random.stateless_uniform( shape=tf.shape(tokens), maxval=tf.size(tokens), dtype=tf.int32, seed=seeds[0]) return tf.where(noise_mask, tf.gather(tokens, indices), tokens)
Replace each noise token with a random token from the sequence. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: an unused vocabulary.Vocabulary seeds: an int32 Tensor, sized (1, 2) Returns: a Tensor with the same shape and dtype as tokens
72
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def noise_token_to_sentinel(tokens, noise_mask, vocabulary, seeds): """Replace each noise token with the given sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an unused int32 Tensor Returns: a Tensor with the same shape and dtype as tokens """ del seeds return tf.where(noise_mask, tf.cast(sentinel_id(vocabulary), tokens.dtype), tokens) def noise_token_to_random_token( tokens, noise_mask, vocabulary, seeds, num_reserved_tokens=3): """Replace each noise token with a random token from the vocabulary. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an int32 Tensor, shaped (1, 2) num_reserved_tokens: an integer Returns: a Tensor with the same shape and dtype as tokens """ return tf.where(noise_mask, tf.random.stateless_uniform( tf.shape(tokens), minval=num_reserved_tokens, maxval=vocabulary.vocab_size, dtype=tokens.dtype, seed=seeds[0]), tokens) The provided code snippet includes necessary dependencies for implementing the `noise_token_to_random_token_or_sentinel` function. Write a Python function `def noise_token_to_random_token_or_sentinel( tokens, noise_mask, vocabulary, seeds, random_prob=0.1)` to solve the following problem: Replace each noise token with a random token or a sentinel. For each masked token, with probability random_prob, we replace it by a random token from the vocabulary. Otherwise, we replace it with a sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an int32 Tensor, shaped (2, 2). random_prob: a float Returns: a Tensor with the same shape and dtype as tokens Here is the function: def noise_token_to_random_token_or_sentinel( tokens, noise_mask, vocabulary, seeds, random_prob=0.1): """Replace each noise token with a random token or a sentinel. For each masked token, with probability random_prob, we replace it by a random token from the vocabulary. Otherwise, we replace it with a sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an int32 Tensor, shaped (2, 2). random_prob: a float Returns: a Tensor with the same shape and dtype as tokens """ use_random = ( tf.random.stateless_uniform(tf.shape(tokens), seed=seeds[0]) < random_prob) return tf.where( use_random, noise_token_to_random_token( tokens, noise_mask, vocabulary, seeds=seeds[1:]), noise_token_to_sentinel( tokens, noise_mask, vocabulary, seeds=()))
Replace each noise token with a random token or a sentinel. For each masked token, with probability random_prob, we replace it by a random token from the vocabulary. Otherwise, we replace it with a sentinel. Args: tokens: a 1d integer Tensor noise_mask: a boolean Tensor with the same shape as tokens vocabulary: a vocabulary.Vocabulary seeds: an int32 Tensor, shaped (2, 2). random_prob: a float Returns: a Tensor with the same shape and dtype as tokens
73
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf def select_random_chunk(dataset: tf.data.Dataset, output_features: Mapping[str, seqio.Feature], max_length: Optional[int] = None, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[ Sequence[str]] = None, sequence_length: Optional[Mapping[str, int]] = None, uniform_random_start: bool = False, min_length: Optional[int] = None, **unused_kwargs) -> tf.data.Dataset: """SeqIO wrapper for single_example_select_random_chunk().""" def _my_fn(x, seed): return single_example_select_random_chunk( x, seed, output_features=output_features, max_length=max_length, feature_key=feature_key, additional_feature_keys=additional_feature_keys, passthrough_feature_keys=passthrough_feature_keys, sequence_length=sequence_length, uniform_random_start=uniform_random_start, min_length=min_length) # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) return _my_fn(dataset) def reduce_concat_tokens(dataset, feature_key='targets', batch_size=128, **unused_kwargs): """Token-preprocessor to concatenate multiple unrelated documents. If we want to generate examples of exactly the right length, (to avoid wasting space on padding), then we use this function, folowed by split_tokens. Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. feature_key: an string batch_size: an integer - how many documents to concatenate into one Returns: a dataset """ dataset = dataset.map( lambda x: {feature_key: x[feature_key]}, num_parallel_calls=AUTOTUNE) dataset = dataset.padded_batch(batch_size, padded_shapes={feature_key: [-1]}) def _my_fn(x): tokens = tf.reshape(x[feature_key], [-1]) # strip padding tokens = tf.boolean_mask(tokens, tf.cast(tokens, tf.bool)) return {feature_key: tokens} return dataset.map(_my_fn, num_parallel_calls=AUTOTUNE) def split_tokens(dataset: tf.data.Dataset, min_tokens_per_segment: Optional[int] = None, max_tokens_per_segment: int = gin.REQUIRED, feature_key: str = 'targets', additional_feature_keys: Optional[Sequence[str]] = None, passthrough_feature_keys: Optional[Sequence[str]] = None, **unused_kwargs) -> tf.data.Dataset: """Split examples into multiple examples each. The intended use case is to break up long examples for use in unsupervised transfer-learning. This function is generally preceded by select_random_chunk. If min_tokens_per_segment is provided, the segment length is chosen randomly per document from a log-uniform distribution. If min_tokens_per_segment is None, then the segment length is max_tokens_per_segment (except for a possibly shorter last segment in each document). Args: dataset: a tf.data.Dataset with dictionaries containing the key feature_key. min_tokens_per_segment: an optional integer max_tokens_per_segment: an integer, the maximum number of tokens in each segment. Only the final segment may be shorter. feature_key: a string, the feature to split additional_feature_keys: Additional features to split. The same chunk size will be used, so they should be the same size as feature_key. passthrough_feature_keys: Features to pass through without any splitting. Returns: a dataset """ if passthrough_feature_keys: split_keys = set([feature_key] + (additional_feature_keys or [])) overlap_keys = split_keys & set(passthrough_feature_keys) if overlap_keys: raise ValueError( f'split keys {overlap_keys} also included in passthrough keys') def _split_tokens(x, seed): """Split one token sequence into multiple sequences.""" tokens = x[feature_key] n_tokens = tf.shape(tokens)[0] if min_tokens_per_segment is None: length = max_tokens_per_segment else: # pick a length - log-uniformly distributed length = tf.cast( tf.exp( tf.random.stateless_uniform( [], minval=math.log(min_tokens_per_segment), maxval=math.log(max_tokens_per_segment), seed=seed ) ), tf.int32) # Pad to a multiple of length, then use tf.reshape to split up the tokens # into num_segments segments each of the given length. num_segments = tf.cast( tf.math.ceil( tf.cast(n_tokens, tf.float32) / tf.cast(length, tf.float32)) , tf.int32) padding = num_segments * length - tf.shape(tokens)[0] feature_keys_to_split = [feature_key] orig_lengths = {} outputs = {} if additional_feature_keys is not None: feature_keys_to_split.extend(additional_feature_keys) for k in feature_keys_to_split: with tf.control_dependencies([ tf.assert_equal( tf.shape(tokens)[0], tf.shape(x[k])[0], message=(f'Additional feature {k} is not the same size as ' f'{feature_key} along axis 0 in split_tokens().') ) ]): shape = tf.shape(x[k])[1:] shape_list = x[k].shape[1:] padded = tf.pad( x[k], tf.concat([[[0, padding]], tf.zeros([len(shape_list), 2], dtype=tf.int32)], axis=0)) orig_lengths[k] = tf.concat( [tf.repeat(length, num_segments - 1), [length - padding]], axis=0) outputs[k] = tf.reshape( padded, tf.concat([[-1, length], shape], axis=0)) # To avoid memory issues, don't just replicate the passthrough features # for every segment; use tf.data to do it so the copies don't get # instantiated all at once. outputs_ds = tf.data.Dataset.from_tensor_slices(outputs) orig_lengths_ds = tf.data.Dataset.from_tensor_slices(orig_lengths) if passthrough_feature_keys: passthrough = {k: v for k, v in x.items() if k in passthrough_feature_keys} passthrough_ds = tf.data.Dataset.from_tensors(passthrough).repeat( tf.cast(num_segments, tf.int64)) return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds, passthrough_ds)) else: return tf.data.Dataset.zip((outputs_ds, orig_lengths_ds)) def _strip_padding_and_merge_passthrough( inputs, orig_lengths, passthrough=None): output = {} for k, v in inputs.items(): output[k] = v[:orig_lengths[k]] if passthrough: for k, v in passthrough.items(): output[k] = passthrough[k] return output # Filter empty examples. dataset = dataset.filter(lambda x: tf.not_equal(tf.size(x[feature_key]), 0)) dataset = _split_tokens(dataset).flat_map(lambda z: z) dataset = dataset.map( _strip_padding_and_merge_passthrough, num_parallel_calls=AUTOTUNE) return dataset def trim_and_pad_dataset(dataset, sequence_length): """A wrapper to use `seqio.utils.trim_and_pad_dataset` as a preprocessor.""" return seqio.utils.trim_and_pad_dataset( dataset, feature_lengths=sequence_length) The provided code snippet includes necessary dependencies for implementing the `targets_for_prefix_lm_objective` function. Write a Python function `def targets_for_prefix_lm_objective(dataset, sequence_length, output_features)` to solve the following problem: Prepares targets to be used for prefix LM objective. Here is the function: def targets_for_prefix_lm_objective(dataset, sequence_length, output_features): """Prepares targets to be used for prefix LM objective.""" dataset = select_random_chunk( dataset, output_features, max_length=65536, feature_key='targets') dataset = seqio.preprocessors.append_eos(dataset, output_features) dataset = reduce_concat_tokens(dataset, batch_size=128) dataset = split_tokens( dataset, max_tokens_per_segment=sequence_length['targets']) dataset = trim_and_pad_dataset(dataset, sequence_length) return dataset
Prepares targets to be used for prefix LM objective.
74
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `pack_prefix_lm_encoder_decoder` function. Write a Python function `def pack_prefix_lm_encoder_decoder(ds, sequence_length, pad_id=0)` to solve the following problem: Pack two examples into one with the prefix LM objective. Here is the function: def pack_prefix_lm_encoder_decoder(ds, sequence_length, pad_id=0): """Pack two examples into one with the prefix LM objective.""" packed_length = next(iter(sequence_length.values())) assert packed_length % 2 == 0 assert all(l == packed_length for l in sequence_length.values()) @seqio.utils.map_over_dataset(num_seeds=1) def pack_examples(example_pair, seed): split_point = tf.random.stateless_uniform((), minval=1, maxval=packed_length, seed=seed, dtype=tf.int32) inputs = tf.concat([ example_pair['targets'][0][:split_point], example_pair['targets'][1][:packed_length - split_point] ], axis=0) inputs = tf.reshape(inputs, (packed_length,)) targets = tf.concat([ example_pair['targets'][0][split_point:], example_pair['targets'][1][packed_length - split_point:] ], axis=0) targets = tf.reshape(targets, (packed_length,)) encoder_segment_ids = tf.cast( tf.range(packed_length) >= split_point, tf.int32) + 1 decoder_segment_ids = tf.cast( tf.range(packed_length) >= (packed_length - split_point), tf.int32) + 1 decoder_input_tokens = seqio.utils.make_autoregressive_inputs( targets, sequence_id=decoder_segment_ids) encoder_positions = tf.concat( [tf.range(split_point), tf.range(packed_length - split_point)], axis=0) encoder_positions = tf.reshape(encoder_positions, (packed_length,)) decoder_positions = tf.concat( [tf.range(packed_length - split_point), tf.range(split_point)], axis=0) decoder_positions = tf.reshape(decoder_positions, (packed_length,)) decoder_loss_weights = tf.cast( tf.not_equal(targets, pad_id), dtype=tf.int32) return { 'encoder_input_tokens': inputs, 'decoder_target_tokens': targets, 'decoder_input_tokens': decoder_input_tokens, 'encoder_segment_ids': encoder_segment_ids, 'encoder_positions': encoder_positions, 'decoder_segment_ids': decoder_segment_ids, 'decoder_positions': decoder_positions, 'decoder_loss_weights': decoder_loss_weights, } # Note that the batch requires the lengths to be the same. return pack_examples(ds.batch(2))
Pack two examples into one with the prefix LM objective.
75
import collections import functools import math import re from typing import Any, Callable, Mapping, Optional, Protocol, Sequence, Union import uuid from absl import logging import babel import gin import seqio import tensorflow.compat.v2 as tf The provided code snippet includes necessary dependencies for implementing the `pack_prefix_lm_decoder_only` function. Write a Python function `def pack_prefix_lm_decoder_only(ds, sequence_length, loss_on_targets_only=True, pad_id=0)` to solve the following problem: Randomly split the tokens for the prefix LM objective. Here is the function: def pack_prefix_lm_decoder_only(ds, sequence_length, loss_on_targets_only=True, pad_id=0): """Randomly split the tokens for the prefix LM objective.""" packed_length = next(iter(sequence_length.values())) assert packed_length % 2 == 0 assert all(l == packed_length for l in sequence_length.values()) @seqio.utils.map_over_dataset(num_seeds=1) def pack_examples(example, seed): split_point = tf.random.stateless_uniform((), minval=1, maxval=packed_length, seed=seed, dtype=tf.int32) decoder_target_tokens = example['targets'] decoder_input_tokens = seqio.utils.make_autoregressive_inputs( decoder_target_tokens) if loss_on_targets_only: decoder_loss_weights = tf.cast( tf.range(packed_length) >= split_point, tf.int32) else: decoder_loss_weights = tf.ones((packed_length,), dtype=tf.int32) padding_mask = tf.cast( tf.not_equal(decoder_target_tokens, pad_id), dtype=tf.int32) decoder_loss_weights *= padding_mask decoder_causal_attention = tf.cast( tf.range(packed_length) <= split_point, tf.int32) return { 'decoder_target_tokens': decoder_target_tokens, 'decoder_input_tokens': decoder_input_tokens, 'decoder_loss_weights': decoder_loss_weights, 'decoder_causal_attention': decoder_causal_attention, } return pack_examples(ds)
Randomly split the tokens for the prefix LM objective.
76
import collections import os from absl import logging import numpy as np import pandas as pd import tensorflow.compat.v1 as tf import tensorflow_datasets as tfds Event = collections.namedtuple("event", ["step", "value"]) The provided code snippet includes necessary dependencies for implementing the `parse_events_files` function. Write a Python function `def parse_events_files(tb_summary_dir, seqio_summaries=False)` to solve the following problem: Parse all TensorBoard events files in tb_summary_dir. Args: tb_summary_dir: str, path to look for events files in. seqio_summaries: boolean, whether event summaries are generated by SeqIO Evaluator. Returns: A dict, where each key is a TensorBoard tag and each value is a list of Event tuples with step and value attributes. Here is the function: def parse_events_files(tb_summary_dir, seqio_summaries=False): """Parse all TensorBoard events files in tb_summary_dir. Args: tb_summary_dir: str, path to look for events files in. seqio_summaries: boolean, whether event summaries are generated by SeqIO Evaluator. Returns: A dict, where each key is a TensorBoard tag and each value is a list of Event tuples with step and value attributes. """ events = collections.defaultdict(list) for events_file in tf.io.gfile.glob(os.path.join(tb_summary_dir, "events.*")): try: serialized_events = list( tfds.as_numpy(tf.data.TFRecordDataset(events_file)))[1:] for idx, e in enumerate(tf.train.summary_iterator(events_file)): for v in e.summary.value: if seqio_summaries: event = tf.compat.v1.Event.FromString( serialized_events[idx-1]).summary.value[0] # Need to check if event has a tensor or scalar since we need to # handle both cases. if event.HasField("tensor"): metric_value = tf.make_ndarray(event.tensor) else: metric_value = event.simple_value else: metric_value = v.simple_value events[v.tag].append(Event(e.step, metric_value)) except tf.errors.DataLossError: logging.info("Skipping %s due to truncated record.", events_file) return events
Parse all TensorBoard events files in tb_summary_dir. Args: tb_summary_dir: str, path to look for events files in. seqio_summaries: boolean, whether event summaries are generated by SeqIO Evaluator. Returns: A dict, where each key is a TensorBoard tag and each value is a list of Event tuples with step and value attributes.
77
import collections import os from absl import logging import numpy as np import pandas as pd import tensorflow.compat.v1 as tf import tensorflow_datasets as tfds The provided code snippet includes necessary dependencies for implementing the `get_eval_metric_values` function. Write a Python function `def get_eval_metric_values(events, task_name=None)` to solve the following problem: Filter TensorBoard events to only include those for eval metrics. Args: events: dict of list of (step, value) tuples where keys are tags. task_name: string, if not provided, then the function will look for the task name in the events tags. Returns: Dict where key is task_name/metric_name and value is (step, value) tuple. Here is the function: def get_eval_metric_values(events, task_name=None): """Filter TensorBoard events to only include those for eval metrics. Args: events: dict of list of (step, value) tuples where keys are tags. task_name: string, if not provided, then the function will look for the task name in the events tags. Returns: Dict where key is task_name/metric_name and value is (step, value) tuple. """ eval_values = {} for tag, event_values in events.items(): if tag.startswith("eval"): if task_name: _, metric_name = tag.split("/") else: _, task_name_from_tag, metric_name = tag.split("/") eval_task_name = task_name if task_name else task_name_from_tag eval_values["{}/{}".format(eval_task_name, metric_name)] = event_values return eval_values
Filter TensorBoard events to only include those for eval metrics. Args: events: dict of list of (step, value) tuples where keys are tags. task_name: string, if not provided, then the function will look for the task name in the events tags. Returns: Dict where key is task_name/metric_name and value is (step, value) tuple.
78
import collections import os from absl import logging import numpy as np import pandas as pd import tensorflow.compat.v1 as tf import tensorflow_datasets as tfds METRIC_NAMES = collections.OrderedDict([ ("glue_average", Metric("Average GLUE Score")), ("glue_cola_v002/matthews_corrcoef", Metric("CoLA")), ("glue_sst2_v002/accuracy", Metric("SST-2")), ("glue_mrpc_v002/f1", Metric("MRPC (F1)", "MRPC")), ("glue_mrpc_v002/accuracy", Metric("MRPC (accuracy)", "MRPC")), ("glue_stsb_v002/pearson_corrcoef", Metric("STSB (Pearson)", "STSB")), ("glue_stsb_v002/spearman_corrcoef", Metric("STSB (Spearman)", "STSB")), ("glue_qqp_v002/f1", Metric("QQP (F1)", "QQP")), ("glue_qqp_v002/accuracy", Metric("QQP (accuracy)", "QQP")), ("glue_mnli_matched_v002/accuracy", Metric("MNLIm", "MNLI")), ("glue_mnli_mismatched_v002/accuracy", Metric("MNLImm", "MNLI")), ("glue_qnli_v002/accuracy", Metric("QNLI")), ("glue_rte_v002/accuracy", Metric("GLUE RTE")), ("cnn_dailymail_v002/rouge1", Metric("CNN/DM (ROUGE-1)", "CNN/DM")), ("cnn_dailymail_v002/rouge2", Metric("CNN/DM (ROUGE-2)", "CNN/DM")), ("cnn_dailymail_v002/rougeL", Metric("CNN/DM (ROUGE-L)", "CNN/DM")), ("cnn_dailymail_v002/rougeLsum", Metric("CNN/DM (ROUGE-L)", "CNN/DM")), ("squad_v010_allanswers/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010_allanswers/f1", Metric("SQuAD (F1)", "SQuAD")), ("squad_v010_allanswers_span/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010_allanswers_span/f1", Metric("SQuAD (F1)", "SQuAD")), ("squad_v010/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010/f1", Metric("SQuAD (F1)", "SQuAD")), ("super_glue_average", Metric("Average SuperGLUE Score")), ("super_glue_boolq_v102/accuracy", Metric("BoolQ (accuracy)")), ("super_glue_cb_v102/mean_3class_f1", Metric("CB (F1)", "CB")), ("super_glue_cb_v102/accuracy", Metric("CB (accuracy)", "CB")), ("super_glue_copa_v102/accuracy", Metric("CoPA")), ("super_glue_multirc_v102/f1", Metric("MultiRC (F1)", "MultiRC")), ("super_glue_multirc_v102/exact_match", Metric("MultiRC (EM)", "MultiRC")), ("super_glue_record_v102/f1", Metric("ReCoRD (F1)", "ReCoRD")), ("super_glue_record_v102/em", Metric("ReCoRD (EM)", "ReCoRD")), ("super_glue_rte_v102/accuracy", Metric("SuperGLUE RTE")), ("super_glue_wic_v102/accuracy", Metric("WiC")), ("super_glue_wsc_v102_simple_eval/accuracy", Metric("WSC")), ("dpr_v001_simple/accuracy", Metric("DPR")), ("wmt_t2t_ende_v003/bleu", Metric("WMT T2T En-De")), ("wmt14_ende_v003/bleu", Metric("WMT14 En-De")), ("wmt15_enfr_v003/bleu", Metric("WMT15 En-Fr")), ("wmt16_enro_v003/bleu", Metric("WMT16 En-Ro")), ]) def sort_columns(df, metric_names=None): metric_names = metric_names or METRIC_NAMES column_order = list(collections.OrderedDict.fromkeys( [m.name for m in metric_names.values() if m.name in df.columns] )) return df.reindex(columns=column_order)
null
79
import collections import os from absl import logging import numpy as np import pandas as pd import tensorflow.compat.v1 as tf import tensorflow_datasets as tfds METRIC_NAMES = collections.OrderedDict([ ("glue_average", Metric("Average GLUE Score")), ("glue_cola_v002/matthews_corrcoef", Metric("CoLA")), ("glue_sst2_v002/accuracy", Metric("SST-2")), ("glue_mrpc_v002/f1", Metric("MRPC (F1)", "MRPC")), ("glue_mrpc_v002/accuracy", Metric("MRPC (accuracy)", "MRPC")), ("glue_stsb_v002/pearson_corrcoef", Metric("STSB (Pearson)", "STSB")), ("glue_stsb_v002/spearman_corrcoef", Metric("STSB (Spearman)", "STSB")), ("glue_qqp_v002/f1", Metric("QQP (F1)", "QQP")), ("glue_qqp_v002/accuracy", Metric("QQP (accuracy)", "QQP")), ("glue_mnli_matched_v002/accuracy", Metric("MNLIm", "MNLI")), ("glue_mnli_mismatched_v002/accuracy", Metric("MNLImm", "MNLI")), ("glue_qnli_v002/accuracy", Metric("QNLI")), ("glue_rte_v002/accuracy", Metric("GLUE RTE")), ("cnn_dailymail_v002/rouge1", Metric("CNN/DM (ROUGE-1)", "CNN/DM")), ("cnn_dailymail_v002/rouge2", Metric("CNN/DM (ROUGE-2)", "CNN/DM")), ("cnn_dailymail_v002/rougeL", Metric("CNN/DM (ROUGE-L)", "CNN/DM")), ("cnn_dailymail_v002/rougeLsum", Metric("CNN/DM (ROUGE-L)", "CNN/DM")), ("squad_v010_allanswers/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010_allanswers/f1", Metric("SQuAD (F1)", "SQuAD")), ("squad_v010_allanswers_span/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010_allanswers_span/f1", Metric("SQuAD (F1)", "SQuAD")), ("squad_v010/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010/f1", Metric("SQuAD (F1)", "SQuAD")), ("super_glue_average", Metric("Average SuperGLUE Score")), ("super_glue_boolq_v102/accuracy", Metric("BoolQ (accuracy)")), ("super_glue_cb_v102/mean_3class_f1", Metric("CB (F1)", "CB")), ("super_glue_cb_v102/accuracy", Metric("CB (accuracy)", "CB")), ("super_glue_copa_v102/accuracy", Metric("CoPA")), ("super_glue_multirc_v102/f1", Metric("MultiRC (F1)", "MultiRC")), ("super_glue_multirc_v102/exact_match", Metric("MultiRC (EM)", "MultiRC")), ("super_glue_record_v102/f1", Metric("ReCoRD (F1)", "ReCoRD")), ("super_glue_record_v102/em", Metric("ReCoRD (EM)", "ReCoRD")), ("super_glue_rte_v102/accuracy", Metric("SuperGLUE RTE")), ("super_glue_wic_v102/accuracy", Metric("WiC")), ("super_glue_wsc_v102_simple_eval/accuracy", Metric("WSC")), ("dpr_v001_simple/accuracy", Metric("DPR")), ("wmt_t2t_ende_v003/bleu", Metric("WMT T2T En-De")), ("wmt14_ende_v003/bleu", Metric("WMT14 En-De")), ("wmt15_enfr_v003/bleu", Metric("WMT15 En-Fr")), ("wmt16_enro_v003/bleu", Metric("WMT16 En-Ro")), ]) The provided code snippet includes necessary dependencies for implementing the `compute_avg_glue` function. Write a Python function `def compute_avg_glue(df, metric_names=None)` to solve the following problem: Compute average GLUE and SuperGLUE scores from a DataFrame. Will only compute a given average score if all of the metrics for that benchmark appear as columns in the DataFrame. Args: df: pandas.DataFrame, columns should be metric names. metric_names: dict mapping tensorboard tag to metric name. Returns: A pandas.DataFrame which has GLUE and SuperGLUE averages calculated. Here is the function: def compute_avg_glue(df, metric_names=None): """Compute average GLUE and SuperGLUE scores from a DataFrame. Will only compute a given average score if all of the metrics for that benchmark appear as columns in the DataFrame. Args: df: pandas.DataFrame, columns should be metric names. metric_names: dict mapping tensorboard tag to metric name. Returns: A pandas.DataFrame which has GLUE and SuperGLUE averages calculated. """ # Use METRIC_NAMES defined at the top as default metric_names = metric_names or METRIC_NAMES all_glue_tags = { k for k in metric_names.keys() if "glue" in k and "average" not in k } superglue_tags = {k for k in all_glue_tags if "super" in k} glue_tags = all_glue_tags - superglue_tags average_keys = ["Average GLUE Score", "Average SuperGLUE Score"] for average_key, tags in zip(average_keys, [glue_tags, superglue_tags]): # Only compute average if all metric names appear as columns in the DF if {metric_names[t].name for t in tags}.issubset(set(df.columns)): # Compute average over each metric group group_to_metrics = collections.defaultdict(set) for tag in tags: metric = metric_names[tag] group_to_metrics[metric.group].add(metric.name) accum = None for metrics in group_to_metrics.values(): group_avg = np.mean([df[k] for k in metrics], axis=0) accum = group_avg if accum is None else accum + group_avg # Compute average across all groups average = accum/len(group_to_metrics) df[average_key] = average return df
Compute average GLUE and SuperGLUE scores from a DataFrame. Will only compute a given average score if all of the metrics for that benchmark appear as columns in the DataFrame. Args: df: pandas.DataFrame, columns should be metric names. metric_names: dict mapping tensorboard tag to metric name. Returns: A pandas.DataFrame which has GLUE and SuperGLUE averages calculated.
80
import collections import os from absl import logging import numpy as np import pandas as pd import tensorflow.compat.v1 as tf import tensorflow_datasets as tfds class Metric(object): def __init__(self, name, group=None): self.name = name self.group = group or name METRIC_NAMES = collections.OrderedDict([ ("glue_average", Metric("Average GLUE Score")), ("glue_cola_v002/matthews_corrcoef", Metric("CoLA")), ("glue_sst2_v002/accuracy", Metric("SST-2")), ("glue_mrpc_v002/f1", Metric("MRPC (F1)", "MRPC")), ("glue_mrpc_v002/accuracy", Metric("MRPC (accuracy)", "MRPC")), ("glue_stsb_v002/pearson_corrcoef", Metric("STSB (Pearson)", "STSB")), ("glue_stsb_v002/spearman_corrcoef", Metric("STSB (Spearman)", "STSB")), ("glue_qqp_v002/f1", Metric("QQP (F1)", "QQP")), ("glue_qqp_v002/accuracy", Metric("QQP (accuracy)", "QQP")), ("glue_mnli_matched_v002/accuracy", Metric("MNLIm", "MNLI")), ("glue_mnli_mismatched_v002/accuracy", Metric("MNLImm", "MNLI")), ("glue_qnli_v002/accuracy", Metric("QNLI")), ("glue_rte_v002/accuracy", Metric("GLUE RTE")), ("cnn_dailymail_v002/rouge1", Metric("CNN/DM (ROUGE-1)", "CNN/DM")), ("cnn_dailymail_v002/rouge2", Metric("CNN/DM (ROUGE-2)", "CNN/DM")), ("cnn_dailymail_v002/rougeL", Metric("CNN/DM (ROUGE-L)", "CNN/DM")), ("cnn_dailymail_v002/rougeLsum", Metric("CNN/DM (ROUGE-L)", "CNN/DM")), ("squad_v010_allanswers/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010_allanswers/f1", Metric("SQuAD (F1)", "SQuAD")), ("squad_v010_allanswers_span/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010_allanswers_span/f1", Metric("SQuAD (F1)", "SQuAD")), ("squad_v010/em", Metric("SQuAD (EM)", "SQuAD")), ("squad_v010/f1", Metric("SQuAD (F1)", "SQuAD")), ("super_glue_average", Metric("Average SuperGLUE Score")), ("super_glue_boolq_v102/accuracy", Metric("BoolQ (accuracy)")), ("super_glue_cb_v102/mean_3class_f1", Metric("CB (F1)", "CB")), ("super_glue_cb_v102/accuracy", Metric("CB (accuracy)", "CB")), ("super_glue_copa_v102/accuracy", Metric("CoPA")), ("super_glue_multirc_v102/f1", Metric("MultiRC (F1)", "MultiRC")), ("super_glue_multirc_v102/exact_match", Metric("MultiRC (EM)", "MultiRC")), ("super_glue_record_v102/f1", Metric("ReCoRD (F1)", "ReCoRD")), ("super_glue_record_v102/em", Metric("ReCoRD (EM)", "ReCoRD")), ("super_glue_rte_v102/accuracy", Metric("SuperGLUE RTE")), ("super_glue_wic_v102/accuracy", Metric("WiC")), ("super_glue_wsc_v102_simple_eval/accuracy", Metric("WSC")), ("dpr_v001_simple/accuracy", Metric("DPR")), ("wmt_t2t_ende_v003/bleu", Metric("WMT T2T En-De")), ("wmt14_ende_v003/bleu", Metric("WMT14 En-De")), ("wmt15_enfr_v003/bleu", Metric("WMT15 En-Fr")), ("wmt16_enro_v003/bleu", Metric("WMT16 En-Ro")), ]) The provided code snippet includes necessary dependencies for implementing the `scores_to_df` function. Write a Python function `def scores_to_df(scores, metric_names=None)` to solve the following problem: Convert `scores` into a pandas DataFrame. Here is the function: def scores_to_df(scores, metric_names=None): """Convert `scores` into a pandas DataFrame.""" # Use METRIC_NAMES defined at the top as default metric_names = metric_names or METRIC_NAMES for tag in scores.keys(): if tag not in metric_names: metric_names[tag] = Metric(tag) logging.warning( "TensorBoard tag %s not found in metric_names. " "Using tag as metric name.", tag) # Sort the tags in scores according to metric_names order sorted_tags = sorted( scores.keys(), key=lambda x: list(metric_names.keys()).index(x) ) columns = [metric_names[t].name for t in sorted_tags] # Convert scores to dict with the format # {step_number: {tag1: value, tag2: value, ...}} step_scores = collections.defaultdict( lambda: collections.OrderedDict([(t, np.nan) for t in sorted_tags]) ) for tag in sorted_tags: for step, value in scores[tag]: # If a job gets evicted and restarts from a prior checkpoint, it's # possible that a single step has more than one eval result. In that case, # we pick the max value across all the eval results. if step_scores[step][tag]: step_scores[step][tag] = max(value, step_scores[step][tag]) else: step_scores[step][tag] = value sorted_items = sorted(list(step_scores.items())) data = [list(r.values()) for _, r in sorted_items] index = [s for s, _ in sorted_items] df = pd.DataFrame(data, index, columns) df.index.name = "step" return df
Convert `scores` into a pandas DataFrame.
81
import collections import os from absl import logging import numpy as np import pandas as pd import tensorflow.compat.v1 as tf import tensorflow_datasets as tfds def metric_group_max(df, metric_names=None): """Find the step which achieves the highest mean value for a group of metrics.""" # Use METRIC_NAMES defined at the top as default metric_names = metric_names or METRIC_NAMES group_to_metrics = collections.defaultdict(set) for metric in metric_names.values(): group_to_metrics[metric.group].add(metric.name) group_df = pd.DataFrame() for group, metrics in group_to_metrics.items(): if not all(m in df for m in metrics): continue group_df[group] = df[list(metrics)].mean(axis=1) # Need to replace nan with large negative value for idxmax group_max_step = group_df.fillna(-1e9).idxmax(axis=0) metric_max = pd.Series() metric_max_step = pd.Series() for group_name, max_step in group_max_step.items(): for metric in group_to_metrics[group_name]: metric_max[metric] = df[metric][max_step] metric_max_step[metric] = max_step metric_max = metric_max.reindex(df.columns) metric_max_step = metric_max_step.reindex(df.columns) return metric_max, metric_max_step The provided code snippet includes necessary dependencies for implementing the `log_csv` function. Write a Python function `def log_csv(df, metric_names=None, output_file=None)` to solve the following problem: Log scores to be copy/pasted into a spreadsheet. Here is the function: def log_csv(df, metric_names=None, output_file=None): """Log scores to be copy/pasted into a spreadsheet.""" logging.info(",".join(df.columns)) metric_max, metric_max_step = metric_group_max(df, metric_names) max_row = "max," + ",".join("{:.3f}".format(m) for m in metric_max) logging.info(max_row) idx_row = "step," + ",".join("{:d}".format(i) for i in metric_max_step) logging.info(idx_row) if output_file is not None: with tf.io.gfile.GFile(output_file, "w") as f: csv_string = df.to_csv(float_format="%.3f") f.write(csv_string + max_row + "\n" + idx_row)
Log scores to be copy/pasted into a spreadsheet.
82
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `bleu` function. Write a Python function `def bleu(targets, predictions, tokenizer="intl")` to solve the following problem: Computes BLEU score. Args: targets: list of strings or list of list of strings if multiple references are present. predictions: list of strings tokenizer: tokenizer option for corpus_bleu Returns: bleu_score across all targets and predictions Here is the function: def bleu(targets, predictions, tokenizer="intl"): """Computes BLEU score. Args: targets: list of strings or list of list of strings if multiple references are present. predictions: list of strings tokenizer: tokenizer option for corpus_bleu Returns: bleu_score across all targets and predictions """ if isinstance(targets[0], list): targets = [[x for x in target] for target in targets] else: # Need to wrap targets in another list for corpus_bleu. targets = [targets] bleu_score = sacrebleu.corpus_bleu(predictions, targets, smooth_method="exp", smooth_value=0.0, force=False, lowercase=False, tokenize=tokenizer, use_effective_order=False) return {"bleu": bleu_score.score}
Computes BLEU score. Args: targets: list of strings or list of list of strings if multiple references are present. predictions: list of strings tokenizer: tokenizer option for corpus_bleu Returns: bleu_score across all targets and predictions
83
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def _prepare_summary_rouge(summary): # Make sure the summary is not bytes-type # Add newlines between sentences so that rougeLsum is computed correctly. summary = summary.replace(" . ", " .\n") return summary The provided code snippet includes necessary dependencies for implementing the `rouge` function. Write a Python function `def rouge( targets, predictions, score_keys=("rouge1", "rouge2", "rougeLsum"), verbose=True, **kwargs, )` to solve the following problem: Computes rouge score nondeterministically using the bootstrap. Args: targets: list of strings. predictions: list of strings. score_keys: list of strings with the keys to compute. verbose: whether to enable additional logging. **kwargs: additional keyword arguments for RougeScorer. Returns: dict with score_key: rouge score across all targets and predictions Here is the function: def rouge( targets, predictions, score_keys=("rouge1", "rouge2", "rougeLsum"), verbose=True, **kwargs, ): """Computes rouge score nondeterministically using the bootstrap. Args: targets: list of strings. predictions: list of strings. score_keys: list of strings with the keys to compute. verbose: whether to enable additional logging. **kwargs: additional keyword arguments for RougeScorer. Returns: dict with score_key: rouge score across all targets and predictions """ scorer = rouge_scorer.RougeScorer(rouge_types=score_keys, **kwargs) aggregator = scoring.BootstrapAggregator() for prediction, target in zip(predictions, targets): target = _prepare_summary_rouge(target) prediction = _prepare_summary_rouge(prediction) aggregator.add_scores(scorer.score(target=target, prediction=prediction)) result = aggregator.aggregate() if verbose: for key in score_keys: logging.info( "%s = %.2f, 95%% confidence [%.2f, %.2f]", key, result[key].mid.fmeasure*100, result[key].low.fmeasure*100, result[key].high.fmeasure*100, ) return {key: result[key].mid.fmeasure*100 for key in score_keys}
Computes rouge score nondeterministically using the bootstrap. Args: targets: list of strings. predictions: list of strings. score_keys: list of strings with the keys to compute. verbose: whether to enable additional logging. **kwargs: additional keyword arguments for RougeScorer. Returns: dict with score_key: rouge score across all targets and predictions
84
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def _prepare_summary_rouge(summary): # Make sure the summary is not bytes-type # Add newlines between sentences so that rougeLsum is computed correctly. summary = summary.replace(" . ", " .\n") return summary The provided code snippet includes necessary dependencies for implementing the `rouge_mean` function. Write a Python function `def rouge_mean( targets, predictions, score_keys=("rouge1", "rouge2", "rougeLsum"), **kwargs, )` to solve the following problem: Computes rouge score deterministically (no bootstrap). Args: targets: list of strings predictions: list of strings score_keys: list of strings with the keys to compute **kwargs: additional keyword arguments for RougeScorer. Returns: dict with score_key: rouge score across all targets and predictions Here is the function: def rouge_mean( targets, predictions, score_keys=("rouge1", "rouge2", "rougeLsum"), **kwargs, ): """Computes rouge score deterministically (no bootstrap). Args: targets: list of strings predictions: list of strings score_keys: list of strings with the keys to compute **kwargs: additional keyword arguments for RougeScorer. Returns: dict with score_key: rouge score across all targets and predictions """ scorer = rouge_scorer.RougeScorer(rouge_types=score_keys, **kwargs) count = 0 sum_scores = collections.defaultdict(float) for prediction, target in zip(predictions, targets): target = _prepare_summary_rouge(target) prediction = _prepare_summary_rouge(prediction) scores = scorer.score(target=target, prediction=prediction) count += 1 for k, v in scores.items(): sum_scores[k] += v.fmeasure if count == 0: raise ValueError("Predictions and targets must both have nonzero length") result = {k: v / count for k, v in sum_scores.items()} return {key: result[key] * 100 for key in score_keys}
Computes rouge score deterministically (no bootstrap). Args: targets: list of strings predictions: list of strings score_keys: list of strings with the keys to compute **kwargs: additional keyword arguments for RougeScorer. Returns: dict with score_key: rouge score across all targets and predictions
85
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def squad(targets, predictions): """Computes SQuAD metrics, maximizing over answers per question. Args: targets: list of lists of strings predictions: list of strings Returns: dict with score_key: squad score across all targets and predictions """ targets = [[qa_utils.normalize_squad(t) for t in u] for u in targets] predictions = [qa_utils.normalize_squad(p) for p in predictions] return qa_utils.qa_metrics(targets, predictions) The provided code snippet includes necessary dependencies for implementing the `span_squad` function. Write a Python function `def span_squad(targets, predictions)` to solve the following problem: Computes SQuAD metrics for span prediction tasks. Uses qa metric function to compute EM and F1 score. Args: targets: list of dict of answers (list of strings) and context (string) predictions: list of strings, each string is contains the space tokenized ids in the format: "start: 3 end: 6" Returns: dict with score_key: squad score across all targets and predictions Here is the function: def span_squad(targets, predictions): """Computes SQuAD metrics for span prediction tasks. Uses qa metric function to compute EM and F1 score. Args: targets: list of dict of answers (list of strings) and context (string) predictions: list of strings, each string is contains the space tokenized ids in the format: "start: 3 end: 6" Returns: dict with score_key: squad score across all targets and predictions """ assert len(targets) == len(predictions) def space_tok(s): return re.sub(r"\W", " ", s).split() def get_answer_text_from_context(context, answer_tokens): """Find the answer in the context given the answer tokens.""" # In the initial training iterations, the model can output garbage. # Returning an empty string in such cases. if len(answer_tokens) < 4: return "" # Model sometimes predicts words instead of numbers in the answer. Return # an empty string in that case. try: start_index = int(answer_tokens[1]) end_index = int(answer_tokens[3]) except ValueError: return "" return " ".join(context[start_index:end_index+1]) contexts = [space_tok(t["context"]) for t in targets] answers = [t["answers"] for t in targets] predictions = [space_tok(p) for p in predictions] final_predictions = [ get_answer_text_from_context(c, p) for c, p in zip(contexts, predictions) ] return squad(answers, final_predictions)
Computes SQuAD metrics for span prediction tasks. Uses qa metric function to compute EM and F1 score. Args: targets: list of dict of answers (list of strings) and context (string) predictions: list of strings, each string is contains the space tokenized ids in the format: "start: 3 end: 6" Returns: dict with score_key: squad score across all targets and predictions
86
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `trivia_qa` function. Write a Python function `def trivia_qa(targets, predictions)` to solve the following problem: Computes TriviaQA metrics, maximizing over answers per question. Args: targets: list of lists of strings predictions: list of strings Returns: dict with score_key: squad score across all targets and predictions Here is the function: def trivia_qa(targets, predictions): """Computes TriviaQA metrics, maximizing over answers per question. Args: targets: list of lists of strings predictions: list of strings Returns: dict with score_key: squad score across all targets and predictions """ targets = [[qa_utils.normalize_trivia_qa(t) for t in u] for u in targets] predictions = [qa_utils.normalize_trivia_qa(p) for p in predictions] return qa_utils.qa_metrics(targets, predictions)
Computes TriviaQA metrics, maximizing over answers per question. Args: targets: list of lists of strings predictions: list of strings Returns: dict with score_key: squad score across all targets and predictions
87
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def accuracy(targets, predictions): return {"accuracy": 100*sklearn.metrics.accuracy_score(targets, predictions)}
null
88
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `sequence_accuracy` function. Write a Python function `def sequence_accuracy(targets, predictions)` to solve the following problem: Computes per-sequence accuracy. For each example, returns 1.0 if the target sequence EXACTLY matches the predicted sequence. Else, 0.0. Args: targets: list of strings predictions: list of strings Returns: float. Average sequence-level accuracy. Here is the function: def sequence_accuracy(targets, predictions): """Computes per-sequence accuracy. For each example, returns 1.0 if the target sequence EXACTLY matches the predicted sequence. Else, 0.0. Args: targets: list of strings predictions: list of strings Returns: float. Average sequence-level accuracy. """ assert len(targets) == len(predictions) seq_acc = 100 * np.mean([p == t for p, t in zip(predictions, targets)]) return {"sequence_accuracy": seq_acc}
Computes per-sequence accuracy. For each example, returns 1.0 if the target sequence EXACTLY matches the predicted sequence. Else, 0.0. Args: targets: list of strings predictions: list of strings Returns: float. Average sequence-level accuracy.
89
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `pearson_corrcoef` function. Write a Python function `def pearson_corrcoef(targets, predictions)` to solve the following problem: Pearson correlation coefficient. Here is the function: def pearson_corrcoef(targets, predictions): """Pearson correlation coefficient.""" return {"pearson_corrcoef": 100 * scipy.stats.pearsonr(targets, predictions)[0]}
Pearson correlation coefficient.
90
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `spearman_corrcoef` function. Write a Python function `def spearman_corrcoef(targets, predictions)` to solve the following problem: Spearman correlation coefficient. Here is the function: def spearman_corrcoef(targets, predictions): """Spearman correlation coefficient.""" return {"spearman_corrcoef": 100 * scipy.stats.spearmanr(targets, predictions)[0]}
Spearman correlation coefficient.
91
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `all_match` function. Write a Python function `def all_match(targets, predictions)` to solve the following problem: Computes whether all targets match all predictions exactly. Here is the function: def all_match(targets, predictions): """Computes whether all targets match all predictions exactly.""" return {"exact_match": 100 * float(np.array_equal(targets, predictions))}
Computes whether all targets match all predictions exactly.
92
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `deduplicate_metric` function. Write a Python function `def deduplicate_metric(metric_fn, group_key: str = "group", value_key: str = "value")` to solve the following problem: Returns a metric that only considers one example per group. Useful for things like ReCoRD where inputs may be replicated during training to handle multiple labels, but where at eval we only want a single copy of each example. Args: metric_fn: function, the metric to compute on the unique examples. group_key: the key for the grouping value in the target dictionary. value_key: the key for the value in the dictionaries. Returns: A metric function that deduplicated based on the grouping key before returning a metric. Here is the function: def deduplicate_metric(metric_fn, group_key: str = "group", value_key: str = "value"): """Returns a metric that only considers one example per group. Useful for things like ReCoRD where inputs may be replicated during training to handle multiple labels, but where at eval we only want a single copy of each example. Args: metric_fn: function, the metric to compute on the unique examples. group_key: the key for the grouping value in the target dictionary. value_key: the key for the value in the dictionaries. Returns: A metric function that deduplicated based on the grouping key before returning a metric. """ def _deduplicated_metric(targets, predictions): """Deduplicate targets and predictions and pass that to the metric fn.""" processed_groups = set() deduplicated_targets = [] deduplicated_predictions = [] for targ, pred in zip(targets, predictions): group = targ[group_key] if group in processed_groups: continue processed_groups.add(group) deduplicated_targets.append(targ[value_key]) deduplicated_predictions.append(pred[value_key]) return metric_fn(deduplicated_targets, deduplicated_predictions) return _deduplicated_metric
Returns a metric that only considers one example per group. Useful for things like ReCoRD where inputs may be replicated during training to handle multiple labels, but where at eval we only want a single copy of each example. Args: metric_fn: function, the metric to compute on the unique examples. group_key: the key for the grouping value in the target dictionary. value_key: the key for the value in the dictionaries. Returns: A metric function that deduplicated based on the grouping key before returning a metric.
93
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `mean_group_metric` function. Write a Python function `def mean_group_metric(metric_fn, group_key="group", value_key="value", return_subgroup_scores=False)` to solve the following problem: Returns a metric that averages `metric_fn` on sub-groups of results. The sub-groups are defined by aggregating results (targets and predictions) by accessing the feature specified by `group_key` in the target dicts. **WARNING**: Using this function can produce unreliable results if you do not pass in full groups. For example, if you evaluate over a random subsample of a validation set and do not retain all of the examples in each group, you may get results which aren't directly comparable to using the full validation set. Args: metric_fn: function, the metric to compute on the subgroups. group_key: string, the key for the grouping value in the target dictionary. value_key: string, the key for the value in the dictionaries. return_subgroup_scores: If true, include the scores for each sub-group. Here is the function: def mean_group_metric(metric_fn, group_key="group", value_key="value", return_subgroup_scores=False): """Returns a metric that averages `metric_fn` on sub-groups of results. The sub-groups are defined by aggregating results (targets and predictions) by accessing the feature specified by `group_key` in the target dicts. **WARNING**: Using this function can produce unreliable results if you do not pass in full groups. For example, if you evaluate over a random subsample of a validation set and do not retain all of the examples in each group, you may get results which aren't directly comparable to using the full validation set. Args: metric_fn: function, the metric to compute on the subgroups. group_key: string, the key for the grouping value in the target dictionary. value_key: string, the key for the value in the dictionaries. return_subgroup_scores: If true, include the scores for each sub-group. """ def my_metric(targets, predictions): """Computes mean of `metric_fn` over subgroups of results.""" grouped_values = collections.defaultdict(lambda: ([], [])) for targ, pred in zip(targets, predictions): g = targ[group_key] grouped_values[g][0].append(targ[value_key]) grouped_values[g][1].append(pred[value_key]) group_scores = collections.defaultdict(list) for group, (targets, predictions) in grouped_values.items(): for metric, score in metric_fn(targets, predictions).items(): group_scores[metric].append(score) if return_subgroup_scores: group_scores["%s-%s" % (group, metric)].append(score) return {metric: np.mean(scores) for metric, scores in group_scores.items()} return my_metric
Returns a metric that averages `metric_fn` on sub-groups of results. The sub-groups are defined by aggregating results (targets and predictions) by accessing the feature specified by `group_key` in the target dicts. **WARNING**: Using this function can produce unreliable results if you do not pass in full groups. For example, if you evaluate over a random subsample of a validation set and do not retain all of the examples in each group, you may get results which aren't directly comparable to using the full validation set. Args: metric_fn: function, the metric to compute on the subgroups. group_key: string, the key for the grouping value in the target dictionary. value_key: string, the key for the value in the dictionaries. return_subgroup_scores: If true, include the scores for each sub-group.
94
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def f1_score_with_invalid(targets, predictions): """Compute F1 score, but any prediction != 0 or 1 is counted as incorrect. Args: targets: np.ndarray of targets, either 0 or 1 predictions: np.ndarray of predictions, any integer value Returns: F1 score, where any prediction != 0 or 1 is counted as wrong. """ targets, predictions = np.asarray(targets), np.asarray(predictions) # Get indices of invalid predictions invalid_idx_mask = np.logical_and(predictions != 0, predictions != 1) # For any prediction != 0 or 1, set it to the opposite of what the target is predictions[invalid_idx_mask] = 1 - targets[invalid_idx_mask] return {"f1": 100 * sklearn.metrics.f1_score(targets, predictions)} The provided code snippet includes necessary dependencies for implementing the `multirc_f1_over_all_answers` function. Write a Python function `def multirc_f1_over_all_answers(targets, predictions)` to solve the following problem: Special metric for MultiRC which computes F1 score over all examples. This is necessary because the targets/predictions for MultiRC are dicts and the f1_score_with_invalid expects a list of True/False labels, not dicts. As a result we just need to key in the "value" for each of the example dicts before feeding into f1_score_with_invalid. Args: targets: list of dicts, where each dict has a "value" key. predictions: list of dicts, where each dict has a "value" key. Returns: F1 score over values, where any prediction != 0 or 1 is counted as wrong. Here is the function: def multirc_f1_over_all_answers(targets, predictions): """Special metric for MultiRC which computes F1 score over all examples. This is necessary because the targets/predictions for MultiRC are dicts and the f1_score_with_invalid expects a list of True/False labels, not dicts. As a result we just need to key in the "value" for each of the example dicts before feeding into f1_score_with_invalid. Args: targets: list of dicts, where each dict has a "value" key. predictions: list of dicts, where each dict has a "value" key. Returns: F1 score over values, where any prediction != 0 or 1 is counted as wrong. """ return f1_score_with_invalid( [t["value"] for t in targets], [p["value"] for p in predictions] )
Special metric for MultiRC which computes F1 score over all examples. This is necessary because the targets/predictions for MultiRC are dicts and the f1_score_with_invalid expects a list of True/False labels, not dicts. As a result we just need to key in the "value" for each of the example dicts before feeding into f1_score_with_invalid. Args: targets: list of dicts, where each dict has a "value" key. predictions: list of dicts, where each dict has a "value" key. Returns: F1 score over values, where any prediction != 0 or 1 is counted as wrong.
95
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def auc(targets, predictions, targets_threshold=None): """Compute Area Under the ROC and PR curves. ROC - Receiver Operating Characteristic PR - Precision and Recall Args: targets: np.ndarray of targets, either 0 or 1, or continuous values. predictions: np.ndarray of predictions, any value. targets_threshold: float, if target values are continuous values, this threshold binarizes them. Returns: A dictionary with AUC-ROC and AUC-PR scores. """ if targets_threshold is not None: targets = np.array(targets) targets = np.where(targets < targets_threshold, np.zeros_like(targets, dtype=np.int32), np.ones_like(targets, dtype=np.int32)) return { "auc-roc": sklearn.metrics.roc_auc_score(targets, predictions), "auc-pr": sklearn.metrics.average_precision_score(targets, predictions), } The provided code snippet includes necessary dependencies for implementing the `score_auc` function. Write a Python function `def score_auc(targets, scores, targets_threshold=None)` to solve the following problem: Compute Area Under the ROC and PR curves. ROC - Receiver Operating Characteristic PR - Precision and Recall Args: targets: np.ndarray of targets, either 0 or 1, or continuous values. scores: np.ndarray of scores, any value. targets_threshold: float, if target values are continuous values, this threshold binarizes them. Returns: A dictionary with AUC-ROC and AUC-PR scores. Here is the function: def score_auc(targets, scores, targets_threshold=None): """Compute Area Under the ROC and PR curves. ROC - Receiver Operating Characteristic PR - Precision and Recall Args: targets: np.ndarray of targets, either 0 or 1, or continuous values. scores: np.ndarray of scores, any value. targets_threshold: float, if target values are continuous values, this threshold binarizes them. Returns: A dictionary with AUC-ROC and AUC-PR scores. """ return auc( targets=targets, predictions=scores, targets_threshold=targets_threshold)
Compute Area Under the ROC and PR curves. ROC - Receiver Operating Characteristic PR - Precision and Recall Args: targets: np.ndarray of targets, either 0 or 1, or continuous values. scores: np.ndarray of scores, any value. targets_threshold: float, if target values are continuous values, this threshold binarizes them. Returns: A dictionary with AUC-ROC and AUC-PR scores.
96
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def mean_multiclass_f1(num_classes, **metric_fn_kwargs): """Computes the unweighted average of the F1 per class.""" return sklearn_metrics_wrapper( "fbeta_score", metric_dict_str="mean_%dclass_f1" % num_classes, metric_post_process_fn=lambda x: 100 * x, beta=1, labels=range(num_classes), average="macro", **metric_fn_kwargs) The provided code snippet includes necessary dependencies for implementing the `rank_classification` function. Write a Python function `def rank_classification( targets: Sequence[Tuple[Sequence[int], bool, float, int]], scores: Sequence[float], num_classes: Optional[int] = None, normalize_by_target_length: bool = False, idx_len: int = 2, ) -> Dict[str, Union[float, int]]` to solve the following problem: Computes standard metrics classification based on log likelihood ranking. This metric is intended to be used along with the `rank_classification` preprocessor and postprocessor. Each example is scored (by log likelihood) for every possible label, and the label with the best score is selected as the prediction. In the case of multiple labels, a prediction matching any will be considered correct. For problems with two labels, AUC-pr and AUC-roc retrieval metrics will be reported for the positive class, which is assumed to have an 'idx' of 1. If more labels are present, only accuracy and F-1 will be reported. Args: targets: list of tuples, the 'idx', 'is_correct', 'weight' fields, and length of target tokens from ground truth examples. scores: list of float, a flat list of log likelihood scores for each possible label for each example. num_classes: int or None, the number of possible classes for the label or None if the number of classes vary. normalize_by_target_length: bool, if True the scores are normalized by the target token lengths. idx_len: int, The number of elems in the idx field in the targets. This is generally 2 (input_id, target_id). Returns: Accuracy, f1, and AUC scores. Raises: ValueError: if `targets` is not a sequence of 4-tuples. Here is the function: def rank_classification( targets: Sequence[Tuple[Sequence[int], bool, float, int]], scores: Sequence[float], num_classes: Optional[int] = None, normalize_by_target_length: bool = False, idx_len: int = 2, ) -> Dict[str, Union[float, int]]: """Computes standard metrics classification based on log likelihood ranking. This metric is intended to be used along with the `rank_classification` preprocessor and postprocessor. Each example is scored (by log likelihood) for every possible label, and the label with the best score is selected as the prediction. In the case of multiple labels, a prediction matching any will be considered correct. For problems with two labels, AUC-pr and AUC-roc retrieval metrics will be reported for the positive class, which is assumed to have an 'idx' of 1. If more labels are present, only accuracy and F-1 will be reported. Args: targets: list of tuples, the 'idx', 'is_correct', 'weight' fields, and length of target tokens from ground truth examples. scores: list of float, a flat list of log likelihood scores for each possible label for each example. num_classes: int or None, the number of possible classes for the label or None if the number of classes vary. normalize_by_target_length: bool, if True the scores are normalized by the target token lengths. idx_len: int, The number of elems in the idx field in the targets. This is generally 2 (input_id, target_id). Returns: Accuracy, f1, and AUC scores. Raises: ValueError: if `targets` is not a sequence of 4-tuples. """ assert len(targets) == len(scores) if len(targets[0]) != 4: raise ValueError( f"`targets` should contain 4 elements but has {len(targets[0])}.") normalized_scores = [] if normalize_by_target_length: for target, score in zip(targets, scores): _, _, _, target_length = target score = score / target_length normalized_scores.append(score) scores = normalized_scores idx_0 = targets[0][0] if not hasattr(idx_0, "__len__") or len(idx_0) != idx_len: raise ValueError("The first element of `targets` ('idx') should be " f"{idx_len}-dimensional. Got {idx_0}.") # Sort by 'idx' since the function relies on this assumption. # ((idx, is_correct, weight, target_length), score) get_idx = lambda x: x[0][0] targets, scores = zip(*sorted(zip(targets, scores), key=get_idx)) def all_unique(indices): seen = set() for idx in indices: if idx in seen: return False seen.add(idx) return True indices = (t[0] for t in targets) if not all_unique(indices): err_msg = ( "rank_classification metric function received targets list with" " non-unique indices. There's no way to distinguish the items, so the" " metric can't be computed. Most likely, this is caused by using SeqIO" " rank_classification preprocessor (or a similar one) in a multi-host" " experiment. To fix this, either use Pathways, or make the ids random" " with big enough range to make them unique." ) raise ValueError(err_msg) if not num_classes: # Assuming variable classes. Can only compute accuracy. num_correct = 0 total = 0 # (((input idx, output idx), is_correct, weight, target_length), score) get_grp = lambda x: x[0][0][0] for _, grp in itertools.groupby(zip(targets, scores), get_grp): exs, log_likelihoods = zip(*grp) prediction = np.argmax(log_likelihoods) weights = exs[prediction][2] num_correct += exs[prediction][1] * weights total += weights return {"accuracy": 100 * num_correct / total} assert len(targets) % num_classes == 0, f"{len(targets)} % {num_classes} != 0" labels_indicator = np.array([is_correct for _, is_correct, _, _ in targets ]).reshape((-1, num_classes)) weights = np.array([weight for _, _, weight, _ in targets]).reshape( (-1, num_classes))[:, 0] log_likelihoods = np.array(scores, np.float32).reshape((-1, num_classes)) predictions = log_likelihoods.argmax(-1) if np.any(labels_indicator.sum(axis=-1) > 1): # multiple-answer case logging.info( "Multiple labels detected. Predictions matching any label will be " "considered correct.") num_examples = len(labels_indicator) return { "accuracy": (100 * np.average( labels_indicator[np.arange(num_examples), predictions], weights=weights)) } predictions_indicator = np.eye(num_classes)[predictions] def exp_normalize(x): b = x.max(-1)[:, np.newaxis] y = np.exp(x - b) return y / y.sum(-1)[:, np.newaxis] probs = exp_normalize(log_likelihoods) metrics = { "accuracy": 100 * sklearn.metrics.accuracy_score( labels_indicator, predictions_indicator, sample_weight=weights), } if num_classes > 2: metrics.update( mean_multiclass_f1(num_classes, sample_weight=weights)(labels_indicator, predictions_indicator)) logging.warning("AUC-pr and AUC-roc are not supported when num_classes > 2") else: metrics.update({ "f1": 100 * sklearn.metrics.f1_score( labels_indicator.argmax(-1), predictions, sample_weight=weights) }) labels_indicator = labels_indicator[:, 1] probs = probs[:, 1] metrics.update({ "auc-roc": 100 * sklearn.metrics.roc_auc_score( labels_indicator, probs, multi_class="ovr", sample_weight=weights, average="macro"), "auc-pr": 100 * sklearn.metrics.average_precision_score( labels_indicator, probs, sample_weight=weights, average="macro"), }) return metrics
Computes standard metrics classification based on log likelihood ranking. This metric is intended to be used along with the `rank_classification` preprocessor and postprocessor. Each example is scored (by log likelihood) for every possible label, and the label with the best score is selected as the prediction. In the case of multiple labels, a prediction matching any will be considered correct. For problems with two labels, AUC-pr and AUC-roc retrieval metrics will be reported for the positive class, which is assumed to have an 'idx' of 1. If more labels are present, only accuracy and F-1 will be reported. Args: targets: list of tuples, the 'idx', 'is_correct', 'weight' fields, and length of target tokens from ground truth examples. scores: list of float, a flat list of log likelihood scores for each possible label for each example. num_classes: int or None, the number of possible classes for the label or None if the number of classes vary. normalize_by_target_length: bool, if True the scores are normalized by the target token lengths. idx_len: int, The number of elems in the idx field in the targets. This is generally 2 (input_id, target_id). Returns: Accuracy, f1, and AUC scores. Raises: ValueError: if `targets` is not a sequence of 4-tuples.
97
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring def _coqa_tokenize(inp: str) -> Sequence[str]: """Normalize English text and tokenize into words based on spaces. Adapted from official evaluation tokenization at https://stanfordnlp.github.io/coqa/. Args: inp: string. Returns: Tokenization of normalized text as List[str] """ def remove_articles(text): regex = re.compile(r"\b(a|an|the)\b", re.UNICODE) return re.sub(regex, " ", text) def normalize_whitespace(text): return " ".join(text.split()) def remove_punc(text): exclude = set(string.punctuation) return "".join(ch for ch in text if ch not in exclude) return normalize_whitespace(remove_articles(remove_punc(inp.lower()))).split() def _sequence_f1(target_tokens: Sequence[str], prediction_tokens: Sequence[str]) -> float: """Given target and prediction tokens, return token-wise F1 score.""" if not (target_tokens or prediction_tokens): return int(target_tokens == prediction_tokens) common_token_counts = ( collections.Counter(target_tokens) & collections.Counter(prediction_tokens)) sum_common = sum(common_token_counts.values()) if sum_common == 0: return 0 precision = 1.0 * sum_common / len(prediction_tokens) recall = 1.0 * sum_common / len(target_tokens) f1 = (2 * precision * recall) / (precision + recall) return f1 The provided code snippet includes necessary dependencies for implementing the `coqa_f1` function. Write a Python function `def coqa_f1( targets: Sequence[Sequence[str]], predictions: Sequence[str] ) -> Mapping[str, float]` to solve the following problem: Return mean sequence F1 score over all QA turns. Here is the function: def coqa_f1( targets: Sequence[Sequence[str]], predictions: Sequence[str] ) -> Mapping[str, float]: """Return mean sequence F1 score over all QA turns.""" f1s = [] for (target, p) in zip(targets, predictions): assert isinstance(target, Sequence) prediction_tokens = _coqa_tokenize(p) example_f1s = [ _sequence_f1(_coqa_tokenize(t), prediction_tokens) for t in target ] f1s.append(max(example_f1s)) return {"f1": np.mean(np.array(f1s)) * 100}
Return mean sequence F1 score over all QA turns.
98
import collections import itertools import re import string from typing import Any, Dict, Mapping, Optional, Sequence, Tuple, Union from absl import logging import editdistance import flax import jax.numpy as jnp import numpy as np import sacrebleu import scipy.stats import seqio import sklearn.metrics from t5.evaluation import qa_utils import tensorflow.compat.v2 as tf from rouge_score import rouge_scorer from rouge_score import scoring The provided code snippet includes necessary dependencies for implementing the `edit_distance` function. Write a Python function `def edit_distance(targets, predictions, lower=True)` to solve the following problem: Word-level edit distance between targets and predictions. Here is the function: def edit_distance(targets, predictions, lower=True): """Word-level edit distance between targets and predictions.""" edit_distances = [] for pred, target in zip(predictions, targets): if lower: pred = pred.lower() target = target.lower() # For simplicity, use regex-based tokenization that treats each # contiguous chunk of characters matched by \w as a word. pred = re.split("[^\\w]", pred) target = re.split("[^\\w]", target) edit_distances.append(editdistance.distance(pred, target)) return {"min_edit": min(edit_distances), "max_edit": max(edit_distances), "mean_edit": np.mean(edit_distances), "median_edit": np.median(edit_distances), "sum_edit": sum(edit_distances)}
Word-level edit distance between targets and predictions.
99
import os import re from absl import app from absl import flags from absl import logging import numpy as np import tensorflow.compat.v1 as tf def average_tensors(tensors): result = tensors[0] for t in tensors[1:]: result += t return result / len(tensors)
null