xinference 0.14.4.post1__py3-none-any.whl → 0.15.1__py3-none-any.whl

xinference/thirdparty/fish_speech/fish_speech/models/vqgan/modules/wavenet.py

@@ -1,225 +0,0 @@
-import math
-from typing import Optional
-
-import torch
-import torch.nn.functional as F
-from torch import nn
-
-
-class Mish(nn.Module):
-    def forward(self, x):
-        return x * torch.tanh(F.softplus(x))
-
-
-class DiffusionEmbedding(nn.Module):
-    """Diffusion Step Embedding"""
-
-    def __init__(self, d_denoiser):
-        super(DiffusionEmbedding, self).__init__()
-        self.dim = d_denoiser
-
-    def forward(self, x):
-        device = x.device
-        half_dim = self.dim // 2
-        emb = math.log(10000) / (half_dim - 1)
-        emb = torch.exp(torch.arange(half_dim, device=device) * -emb)
-        emb = x[:, None] * emb[None, :]
-        emb = torch.cat((emb.sin(), emb.cos()), dim=-1)
-        return emb
-
-
-class LinearNorm(nn.Module):
-    """LinearNorm Projection"""
-
-    def __init__(self, in_features, out_features, bias=False):
-        super(LinearNorm, self).__init__()
-        self.linear = nn.Linear(in_features, out_features, bias)
-
-        nn.init.xavier_uniform_(self.linear.weight)
-        if bias:
-            nn.init.constant_(self.linear.bias, 0.0)
-
-    def forward(self, x):
-        x = self.linear(x)
-        return x
-
-
-class ConvNorm(nn.Module):
-    """1D Convolution"""
-
-    def __init__(
-        self,
-        in_channels,
-        out_channels,
-        kernel_size=1,
-        stride=1,
-        padding=None,
-        dilation=1,
-        bias=True,
-        w_init_gain="linear",
-    ):
-        super(ConvNorm, self).__init__()
-
-        if padding is None:
-            assert kernel_size % 2 == 1
-            padding = int(dilation * (kernel_size - 1) / 2)
-
-        self.conv = nn.Conv1d(
-            in_channels,
-            out_channels,
-            kernel_size=kernel_size,
-            stride=stride,
-            padding=padding,
-            dilation=dilation,
-            bias=bias,
-        )
-        nn.init.kaiming_normal_(self.conv.weight)
-
-    def forward(self, signal):
-        conv_signal = self.conv(signal)
-
-        return conv_signal
-
-
-class ResidualBlock(nn.Module):
-    """Residual Block"""
-
-    def __init__(
-        self,
-        residual_channels,
-        use_linear_bias=False,
-        dilation=1,
-        condition_channels=None,
-    ):
-        super(ResidualBlock, self).__init__()
-        self.conv_layer = ConvNorm(
-            residual_channels,
-            2 * residual_channels,
-            kernel_size=3,
-            stride=1,
-            padding=dilation,
-            dilation=dilation,
-        )
-
-        if condition_channels is not None:
-            self.diffusion_projection = LinearNorm(
-                residual_channels, residual_channels, use_linear_bias
-            )
-            self.condition_projection = ConvNorm(
-                condition_channels, 2 * residual_channels, kernel_size=1
-            )
-
-        self.output_projection = ConvNorm(
-            residual_channels, 2 * residual_channels, kernel_size=1
-        )
-
-    def forward(self, x, condition=None, diffusion_step=None):
-        y = x
-
-        if diffusion_step is not None:
-            diffusion_step = self.diffusion_projection(diffusion_step).unsqueeze(-1)
-            y = y + diffusion_step
-
-        y = self.conv_layer(y)
-
-        if condition is not None:
-            condition = self.condition_projection(condition)
-            y = y + condition
-
-        gate, filter = torch.chunk(y, 2, dim=1)
-        y = torch.sigmoid(gate) * torch.tanh(filter)
-
-        y = self.output_projection(y)
-        residual, skip = torch.chunk(y, 2, dim=1)
-
-        return (x + residual) / math.sqrt(2.0), skip
-
-
-class WaveNet(nn.Module):
-    def __init__(
-        self,
-        input_channels: Optional[int] = None,
-        output_channels: Optional[int] = None,
-        residual_channels: int = 512,
-        residual_layers: int = 20,
-        dilation_cycle: Optional[int] = 4,
-        is_diffusion: bool = False,
-        condition_channels: Optional[int] = None,
-    ):
-        super().__init__()
-
-        # Input projection
-        self.input_projection = None
-        if input_channels is not None and input_channels != residual_channels:
-            self.input_projection = ConvNorm(
-                input_channels, residual_channels, kernel_size=1
-            )
-
-        if input_channels is None:
-            input_channels = residual_channels
-
-        self.input_channels = input_channels
-
-        # Residual layers
-        self.residual_layers = nn.ModuleList(
-            [
-                ResidualBlock(
-                    residual_channels=residual_channels,
-                    use_linear_bias=False,
-                    dilation=2 ** (i % dilation_cycle) if dilation_cycle else 1,
-                    condition_channels=condition_channels,
-                )
-                for i in range(residual_layers)
-            ]
-        )
-
-        # Skip projection
-        self.skip_projection = ConvNorm(
-            residual_channels, residual_channels, kernel_size=1
-        )
-
-        # Output projection
-        self.output_projection = None
-        if output_channels is not None and output_channels != residual_channels:
-            self.output_projection = ConvNorm(
-                residual_channels, output_channels, kernel_size=1
-            )
-
-        if is_diffusion:
-            self.diffusion_embedding = DiffusionEmbedding(residual_channels)
-            self.mlp = nn.Sequential(
-                LinearNorm(residual_channels, residual_channels * 4, False),
-                Mish(),
-                LinearNorm(residual_channels * 4, residual_channels, False),
-            )
-
-        self.apply(self._init_weights)
-
-    def _init_weights(self, m):
-        if isinstance(m, (nn.Conv1d, nn.Linear)):
-            nn.init.trunc_normal_(m.weight, std=0.02)
-            if getattr(m, "bias", None) is not None:
-                nn.init.constant_(m.bias, 0)
-
-    def forward(self, x, t=None, condition=None):
-        if self.input_projection is not None:
-            x = self.input_projection(x)
-            x = F.silu(x)
-
-        if t is not None:
-            t = self.diffusion_embedding(t)
-            t = self.mlp(t)
-
-        skip = []
-        for layer in self.residual_layers:
-            x, skip_connection = layer(x, condition, t)
-            skip.append(skip_connection)
-
-        x = torch.sum(torch.stack(skip), dim=0) / math.sqrt(len(self.residual_layers))
-        x = self.skip_projection(x)
-
-        if self.output_projection is not None:
-            x = F.silu(x)
-            x = self.output_projection(x)
-
-        return x

xinference/thirdparty/fish_speech/tools/auto_rerank.py

@@ -1,159 +0,0 @@
-import os
-
-os.environ["MODELSCOPE_CACHE"] = ".cache/"
-
-import string
-import time
-from threading import Lock
-
-import librosa
-import numpy as np
-import opencc
-import torch
-from faster_whisper import WhisperModel
-
-t2s_converter = opencc.OpenCC("t2s")
-
-
-def load_model(*, device="cuda"):
-    model = WhisperModel(
-        "medium",
-        device=device,
-        compute_type="float16",
-        download_root="faster_whisper",
-    )
-    print("faster_whisper loaded!")
-    return model
-
-
-@torch.no_grad()
-def batch_asr_internal(model: WhisperModel, audios, sr):
-    resampled_audios = []
-    for audio in audios:
-
-        if isinstance(audio, np.ndarray):
-            audio = torch.from_numpy(audio).float()
-
-        if audio.dim() > 1:
-            audio = audio.squeeze()
-
-        assert audio.dim() == 1
-        audio_np = audio.numpy()
-        resampled_audio = librosa.resample(audio_np, orig_sr=sr, target_sr=16000)
-        resampled_audios.append(resampled_audio)
-
-    trans_results = []
-
-    for resampled_audio in resampled_audios:
-        segments, info = model.transcribe(
-            resampled_audio,
-            language=None,
-            beam_size=5,
-            initial_prompt="Punctuation is needed in any language.",
-        )
-        trans_results.append(list(segments))
-
-    results = []
-    for trans_res, audio in zip(trans_results, audios):
-
-        duration = len(audio) / sr * 1000
-        huge_gap = False
-        max_gap = 0.0
-
-        text = None
-        last_tr = None
-
-        for tr in trans_res:
-            delta = tr.text.strip()
-            if tr.id > 1:
-                max_gap = max(tr.start - last_tr.end, max_gap)
-                text += delta
-            else:
-                text = delta
-
-            last_tr = tr
-            if max_gap > 3.0:
-                huge_gap = True
-                break
-
-        sim_text = t2s_converter.convert(text)
-        results.append(
-            {
-                "text": sim_text,
-                "duration": duration,
-                "huge_gap": huge_gap,
-            }
-        )
-
-    return results
-
-
-global_lock = Lock()
-
-
-def batch_asr(model, audios, sr):
-    return batch_asr_internal(model, audios, sr)
-
-
-def is_chinese(text):
-    return True
-
-
-def calculate_wer(text1, text2, debug=False):
-    chars1 = remove_punctuation(text1)
-    chars2 = remove_punctuation(text2)
-
-    m, n = len(chars1), len(chars2)
-
-    if m > n:
-        chars1, chars2 = chars2, chars1
-        m, n = n, m
-
-    prev = list(range(m + 1))  # row 0 distance: [0, 1, 2, ...]
-    curr = [0] * (m + 1)
-
-    for j in range(1, n + 1):
-        curr[0] = j
-        for i in range(1, m + 1):
-            if chars1[i - 1] == chars2[j - 1]:
-                curr[i] = prev[i - 1]
-            else:
-                curr[i] = min(prev[i], curr[i - 1], prev[i - 1]) + 1
-        prev, curr = curr, prev
-
-    edits = prev[m]
-    tot = max(len(chars1), len(chars2))
-    wer = edits / tot
-
-    if debug:
-        print("            gt:   ", chars1)
-        print("          pred:   ", chars2)
-        print(" edits/tot = wer: ", edits, "/", tot, "=", wer)
-
-    return wer
-
-
-def remove_punctuation(text):
-    chinese_punctuation = (
-        " \n\t”“！？｡。＂＃＄％＆＇（）＊＋，－／：；＜＝＞＠［＼］＾＿｀｛｜｝～｟｠｢｣､、〃《》「」『』【】〔〕〖〗〘〙〚〛〜〝〞〟〰〾〿–—"
-        '‛""„‟…‧﹏'
-    )
-    all_punctuation = string.punctuation + chinese_punctuation
-    translator = str.maketrans("", "", all_punctuation)
-    text_without_punctuation = text.translate(translator)
-    return text_without_punctuation
-
-
-if __name__ == "__main__":
-    model = load_model()
-    audios = [
-        librosa.load("44100.wav", sr=44100)[0],
-        librosa.load("lengyue.wav", sr=44100)[0],
-    ]
-    print(np.array(audios[0]))
-    print(batch_asr(model, audios, 44100))
-
-    start_time = time.time()
-    for _ in range(10):
-        print(batch_asr(model, audios, 44100))
-    print("Time taken:", time.time() - start_time)

xinference/thirdparty/fish_speech/tools/gen_ref.py

@@ -1,36 +0,0 @@
-import json
-from pathlib import Path
-
-
-def scan_folder(base_path):
-    wav_lab_pairs = {}
-
-    base = Path(base_path)
-    for suf in ["wav", "lab"]:
-        for f in base.rglob(f"*.{suf}"):
-            relative_path = f.relative_to(base)
-            parts = relative_path.parts
-            print(parts)
-            if len(parts) >= 3:
-                character = parts[0]
-                emotion = parts[1]
-
-                if character not in wav_lab_pairs:
-                    wav_lab_pairs[character] = {}
-                if emotion not in wav_lab_pairs[character]:
-                    wav_lab_pairs[character][emotion] = []
-                wav_lab_pairs[character][emotion].append(str(f.name))
-
-    return wav_lab_pairs
-
-
-def save_to_json(data, output_file):
-    with open(output_file, "w", encoding="utf-8") as file:
-        json.dump(data, file, ensure_ascii=False, indent=2)
-
-
-base_path = "ref_data"
-out_ref_file = "ref_data.json"
-
-wav_lab_pairs = scan_folder(base_path)
-save_to_json(wav_lab_pairs, out_ref_file)

xinference/thirdparty/fish_speech/tools/merge_asr_files.py

@@ -1,55 +0,0 @@
-import os
-from pathlib import Path
-
-from pydub import AudioSegment
-from tqdm import tqdm
-
-from tools.file import AUDIO_EXTENSIONS, list_files
-
-
-def merge_and_delete_files(save_dir, original_files):
-    save_path = Path(save_dir)
-    audio_slice_files = list_files(
-        path=save_dir, extensions=AUDIO_EXTENSIONS.union([".lab"]), recursive=True
-    )
-    audio_files = {}
-    label_files = {}
-    for file_path in tqdm(audio_slice_files, desc="Merging audio files"):
-        rel_path = Path(file_path).relative_to(save_path)
-        (save_path / rel_path.parent).mkdir(parents=True, exist_ok=True)
-        if file_path.suffix == ".wav":
-            prefix = rel_path.parent / file_path.stem.rsplit("-", 1)[0]
-            if prefix == rel_path.parent / file_path.stem:
-                continue
-            audio = AudioSegment.from_wav(file_path)
-            if prefix in audio_files.keys():
-                audio_files[prefix] = audio_files[prefix] + audio
-            else:
-                audio_files[prefix] = audio
-
-        elif file_path.suffix == ".lab":
-            prefix = rel_path.parent / file_path.stem.rsplit("-", 1)[0]
-            if prefix == rel_path.parent / file_path.stem:
-                continue
-            with open(file_path, "r", encoding="utf-8") as f:
-                label = f.read()
-            if prefix in label_files.keys():
-                label_files[prefix] = label_files[prefix] + ", " + label
-            else:
-                label_files[prefix] = label
-
-    for prefix, audio in audio_files.items():
-        output_audio_path = save_path / f"{prefix}.wav"
-        audio.export(output_audio_path, format="wav")
-
-    for prefix, label in label_files.items():
-        output_label_path = save_path / f"{prefix}.lab"
-        with open(output_label_path, "w", encoding="utf-8") as f:
-            f.write(label)
-
-    for file_path in original_files:
-        os.remove(file_path)
-
-
-if __name__ == "__main__":
-    merge_and_delete_files("/made/by/spicysama/laziman", [__file__])

xinference/web/ui/build/static/css/main.4bafd904.css.map

@@ -1 +0,0 @@
-{"version":3,"file":"static/css/main.4bafd904.css","mappings":"AAAA,WAKE,cAAe,CAJf,aAMF,CACA,4BAFE,kBAAmB,CAFnB,YAAa,CAFb,iBAAkB,CAClB,WAaF,CARA,iBAGE,SAAU,CAGV,YAAa,CAJb,QAMF,CACA,WACE,YAAa,CACb,6BACF,CACA,eAEE,kBAAmB,CADnB,YAEF,CACA,YAGE,eAAgB,CADhB,mBAAoB,CADpB,iBAAkB,CAGlB,UACF,CACA,GAEE,oBAAqB,CACrB,2BAA4B,CAF5B,mBAAoB,CAMpB,cAAe,CAHf,eAAgB,CAIhB,cAAiB,CAHjB,sBAAuB,CACvB,qBAGF,CACA,eACE,UAAW,CACX,eAAgB,CAChB,cACF,CACA,kBAOE,kBAAmB,CALnB,WAAY,CAGZ,YAAa,CACb,6BAA8B,CAH9B,UAAW,CAFX,iBAAkB,CAGlB,WAIF,CACA,iBAGE,wBAA6B,CAD7B,cAAiB,CADjB,SAGF,CACA,YAME,wBAAqB,CAHrB,iBAAkB,CADlB,WAAY,CADZ,UAMF,CACA,iBAEE,UAAc,CADd,cAAe,CAEf,iBAAkB,CAClB,aAAc,CACd,iBACF,CACA,SAEE,WAAY,CAIZ,6BAA8B,CAH9B,SAAU,CAFV,iBAAkB,CAGlB,UAIF,CACA,mBAFE,kBAAmB,CAFnB,YASF,CALA,UAEE,qBAAsB,CAEtB,WACF,CACA,cAEE,eAAgB,CADhB,eAEF,CACA,SACE,eACF,CACA,WACE,cACF,CACA,WAGE,qBAAsB,CADtB,YAAa,CAEb,WAAY,CACZ,iBAAkB,CAJlB,YAKF,CACA,aAIE,UAAW,CAHX,YAAa,CACb,6BAA8B,CAC9B,qBAEF,CACA,wBACE,cAAe,CACf,eACF,CACA,SAEE,cAAe,CACf,eAAgB,CAEhB,sBAAuB,CADvB,kBAAmB,CAHnB,WAKF,CACA,UACE,WACF,CACA,OAKE,UAAW,CADX,cAAe,CAFf,QAAS,CADT,iBAAkB,CAElB,OAAQ,CAGR,iCAA6B,CAA7B,yBACF,CACA,cAEE,kBAAmB,CADnB,YAEF,CACA,aAEE,aAAuB,CADvB,iBAEF,CACA,YAIE,qBAAsB,CAEtB,iBAAkB,CADlB,UAAW,CAJX,YAAa,CACb,qBAAsB,CACtB,iBAIF,CACA,kBAGE,kBAAmB,CAFnB,YAAa,CACb,6BAA8B,CAE9B,kBACF,CACA,YACE,cAAe,CACf,eACF,CACA,UAEE,YAAa,CADb,WAEF,CACA,cAIE,qBAAsB,CACtB,iBAAkB,CAElB,UAAW,CALX,WAAY,CACZ,gBAAiB,CAGjB,WAAY,CALZ,UAOF,CACA,SACE,YAAa,CACb,mBAAoB,CACpB,eACF,CClLA,UAEE,UAAW,CACX,cAAe,CAFf,wBAGF,CAEA,gBACE,aACF,CCRA,SAIE,eAAgB,CAFhB,cAAe,CACf,cAAe,CAEf,aAAc,CACd,qBAAsB,CALtB,iBAAkB,CAMlB,8BACF,CAEA,SACE,cAAe,CACf,cAAe,CACf,eACF,CAEA,WAEE,kBAAmB,CAInB,UAAW,CALX,YAAa,CAEb,cAAe,CAEf,UAAW,CADX,QAGF,CAEA,MAGE,cAAe,CACf,gBAAiB,CAHjB,iBAAkB,CAClB,WAGF,CAEA,YACE,aACF,CAEA,OACE,wBACF,CAEA,SAEE,eAAgB,CADhB,iBAAkB,CAGlB,8BAAgC,CADhC,UAEF,CAEA,MAGE,eAAgB,CADhB,iCAA6B,CAA7B,yBAA6B,CAD7B,OAGF,CAEA,iBAGE,kBAAmB,CAFnB,YAAa,CACb,cAAe,CAEf,UACF,CAEA,gBAGE,kBAAmB,CAFnB,YAAa,CACb,6BAEF,CAEA,eAEE,eAAgB,CADhB,gBAEF,CAEA,UAIE,qBAAsB,CACtB,iBAAkB,CAElB,UAAW,CALX,wBAAyB,CACzB,gBAAiB,CAGjB,WAAY,CALZ,UAOF,CAEA,QACE,0BACF,CAEA,MAEE,qBAAsB,CAGtB,kBAAmB,CAFnB,kBAAmB,CAGnB,eAAgB,CAFhB,YAAa,CAHb,iBAMF,CAEA,uBACE,mCAA4B,CAA5B,2BACF,CAEA,WAUE,wBAAyB,CADzB,kBAAmB,CAJnB,WAAY,CAGZ,gBAAiB,CAPjB,iBAAkB,CAClB,UAAW,CAKX,iBAAkB,CAJlB,oBAAqB,CAGrB,kCAA2B,CAA3B,0BAA2B,CAK3B,8BAAgC,CAPhC,UAQF,CAEA,iBAEE,wBAAyB,CADzB,cAEF,CAEA,YAEE,UAAW,CADX,wBAEF","sources":["scenes/launch_model/styles/modelCardStyle.css","components/copyComponent/style.css","scenes/register_model/styles/registerModelStyle.css"],"sourcesContent":[".container {\n  display: block;\n  position: relative;\n  width: 300px;\n  height: 300px;\n  cursor: pointer;\n  border-radius: 20px;\n}\n.descriptionCard {\n  position: relative;\n  top: -1px;\n  left: -1px;\n  width: 300px;\n  height: 300px;\n  padding: 20px;\n  border-radius: 20px;\n}\n.cardTitle {\n  display: flex;\n  justify-content: space-between;\n}\n.iconButtonBox {\n  display: flex;\n  align-items: center;\n}\n.drawerCard {\n  position: relative;\n  padding: 20px 80px 0;\n  min-height: 100%;\n  width: 60vw;\n}\n.p {\n  display: -webkit-box;\n  -webkit-line-clamp: 4;\n  -webkit-box-orient: vertical;\n  overflow: hidden;\n  text-overflow: ellipsis;\n  word-break: break-word;\n  font-size: 14px;\n  padding: 0px 10px;\n}\n.formContainer {\n  height: 80%;\n  overflow: scroll;\n  padding: 0 10px;\n}\n.buttonsContainer {\n  position: absolute;\n  bottom: 50px;\n  left: 100px;\n  right: 100px;\n  display: flex;\n  justify-content: space-between;\n  align-items: center;\n}\n.buttonContainer {\n  width: 45%;\n  border-width: 0px;\n  background-color: transparent;\n}\n.buttonItem {\n  width: 100%;\n  padding: 5px;\n  border-radius: 4px;\n  border: 1px solid #e5e7eb;\n  border-width: 1px;\n  border-color: #e5e7eb;\n}\n.instructionText {\n  font-size: 12px;\n  color: #666666;\n  font-style: italic;\n  margin: 30px 0;\n  text-align: center;\n}\n.iconRow {\n  position: absolute;\n  bottom: 20px;\n  left: 20px;\n  right: 20px;\n  display: flex;\n  justify-content: space-between;\n  align-items: center;\n}\n.iconItem {\n  display: flex;\n  flex-direction: column;\n  align-items: center;\n  margin: 20px;\n}\n.boldIconText {\n  font-weight: bold;\n  font-size: 1.2em;\n}\n.muiIcon {\n  font-size: 1.5em;\n}\n.smallText {\n  font-size: 0.8em;\n}\n.dialogBox {\n  width: 1241px;\n  height: 607px;\n  background-color: #fff;\n  margin: 32px;\n  overflow-x: scroll;\n}\n.dialogTitle {\n  display: flex;\n  justify-content: space-between;\n  padding: 20px 20px 7px;\n  color: #000;\n}\n.dialogTitle-model_name {\n  font-size: 18px;\n  font-weight: 700;\n}\n.pathBox {\n  width: 160px;\n  cursor: pointer;\n  overflow: hidden;\n  white-space: nowrap;\n  text-overflow: ellipsis;\n}\n.pathBox2 {\n  width: 300px;\n}\n.empty {\n  position: absolute;\n  left: 50%;\n  top: 30%;\n  font-size: 20px;\n  color: #555;\n  transform: translate(-50%, 0);\n}\n.deleteDialog {\n  display: flex;\n  align-items: center;\n}\n.warningIcon {\n  margin-right: 10px;\n  color: rgb(237, 108, 2);\n}\n.jsonDialog {\n  display: flex;\n  flex-direction: column;\n  padding: 10px 30px;\n  background-color: #fff;\n  color: #000;\n  border-radius: 8px;\n}\n.jsonDialog-title {\n  display: flex;\n  justify-content: space-between;\n  align-items: center;\n  margin: 10px 0 20px 0;\n}\n.title-name {\n  font-size: 16px;\n  font-weight: 700;\n}\n.main-box {\n  width: 700px;\n  height: 500px;\n}\n.textarea-box {\n  width: 100%;\n  height: 100%;\n  padding: 5px 10px;\n  border: 1px solid #ddd;\n  border-radius: 5px;\n  resize: none;\n  color: #444;\n}\n.but-box {\n  display: flex;\n  justify-content: end;\n  margin-top: 20px;\n}\n",".copyText {\n  font-size: 14px !important;\n  color: #666;\n  cursor: pointer;\n}\n\n.copyText:hover {\n  color: #1976d2;\n}\n",".formBox {\n  position: relative;\n  max-width: 50vw;\n  min-width: 50vw;\n  max-height: 80vh;\n  overflow: auto;\n  padding: 40px 20px 0 0;\n  transition: all 0.4s ease-in-out;\n}\n\n.broaden {\n  max-width: 100%;\n  min-width: 100%;\n  padding-right: 0;\n}\n\n.show-json {\n  display: flex;\n  align-items: center;\n  position: fixed;\n  top: 90px;\n  right: 60px;\n  color: #444;\n}\n\n.icon {\n  position: absolute;\n  right: -40px;\n  cursor: pointer;\n  margin-left: 20px;\n}\n\n.icon:hover {\n  color: #1976d2;\n}\n\n.arrow {\n  font-size: 24px !important;\n}\n\n.jsonBox {\n  position: relative;\n  min-height: 80vh;\n  width: 100%;\n  transition: all 0.4s ease-in-out;\n}\n\n.hide {\n  width: 0;\n  transform: translate(30vw, 0);\n  overflow: hidden;\n}\n\n.checkboxWrapper {\n  display: flex;\n  flex-wrap: wrap;\n  align-items: center;\n  width: 100%;\n}\n\n.jsonBox-header {\n  display: flex;\n  justify-content: space-between;\n  align-items: center;\n}\n\n.jsonBox-title {\n  line-height: 40px;\n  font-weight: 700;\n}\n\n.textarea {\n  width: 100%;\n  height: calc(100% - 40px);\n  padding: 5px 10px;\n  border: 1px solid #ddd;\n  border-radius: 5px;\n  resize: none;\n  color: #444;\n}\n\n.addBtn {\n  margin-left: 20px !important;\n}\n\n.item {\n  position: relative;\n  background-color: #eee;\n  margin: 10px 50px 0;\n  padding: 20px;\n  border-radius: 10px;\n  overflow: hidden;\n}\n\n.item:hover .deleteBtn {\n  transform: translateX(-50px);\n}\n\n.deleteBtn {\n  position: absolute;\n  right: 20px;\n  top: calc(50% - 25px);\n  width: 50px;\n  height: 50px;\n  transform: translateX(80px);\n  text-align: center;\n  line-height: 70px;\n  border-radius: 25px;\n  background-color: #1976d2;\n  transition: all 0.3s ease-in-out;\n}\n\n.deleteBtn:hover {\n  cursor: pointer;\n  box-shadow: 0 0 10px #aaa;\n}\n\n.deleteIcon {\n  font-size: 28px !important;\n  color: #fff;\n}\n"],"names":[],"sourceRoot":""}