ref
stringlengths 4
432
| prompt
stringlengths 1.74k
34.3k
|
|---|---|
class BlipImageBaseProcessor(BaseProcessor): | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: DLYuanGod/TinyGPT-V
# Path: minigpt4/common/registry.py
class Registry:
def register_builder(cls, name):
def wrap(builder_cls):
def register_task(cls, name):
def wrap(task_cls):
def register_model(cls, name):
def wrap(model_cls):
def register_processor(cls, name):
def wrap(processor_cls):
def register_lr_scheduler(cls, name):
def wrap(lr_sched_cls):
def register_runner(cls, name):
def wrap(runner_cls):
def register_path(cls, name, path):
def register(cls, name, obj):
def get_builder_class(cls, name):
def get_model_class(cls, name):
def get_task_class(cls, name):
def get_processor_class(cls, name):
def get_lr_scheduler_class(cls, name):
def get_runner_class(cls, name):
def list_runners(cls):
def list_models(cls):
def list_tasks(cls):
def list_processors(cls):
def list_lr_schedulers(cls):
def list_datasets(cls):
def get_path(cls, name):
def get(cls, name, default=None, no_warning=False):
def unregister(cls, name):
# Path: minigpt4/processors/base_processor.py
class BaseProcessor:
def __init__(self):
self.transform = lambda x: x
return
def __call__(self, item):
return self.transform(item)
@classmethod
def from_config(cls, cfg=None):
return cls()
def build(self, **kwargs):
cfg = OmegaConf.create(kwargs)
return self.from_config(cfg)
# Path: minigpt4/processors/randaugment.py
class RandomAugment(object):
def __init__(self, N=2, M=10, isPIL=False, augs=[]):
self.N = N
self.M = M
self.isPIL = isPIL
if augs:
self.augs = augs
else:
self.augs = list(arg_dict.keys())
def get_random_ops(self):
sampled_ops = np.random.choice(self.augs, self.N)
return [(op, 0.5, self.M) for op in sampled_ops]
def __call__(self, img):
if self.isPIL:
img = np.array(img)
ops = self.get_random_ops()
for name, prob, level in ops:
if np.random.random() > prob:
continue
args = arg_dict[name](level)
img = func_dict[name](img, *args)
return img
# Path: minigpt4/processors/blip_processors.py
import re
from minigpt4.common.registry import registry
from minigpt4.processors.base_processor import BaseProcessor
from minigpt4.processors.randaugment import RandomAugment
from omegaconf import OmegaConf
from torchvision import transforms
from torchvision.transforms.functional import InterpolationMode
"""
Copyright (c) 2022, salesforce.com, inc.
All rights reserved.
SPDX-License-Identifier: BSD-3-Clause
For full license text, see the LICENSE_Lavis file in the repo root or https://opensource.org/licenses/BSD-3-Clause
"""
|
rs = tool.runffmpeg(params) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: jianchang512/vocal-separate
# Path: vocal/cfg.py
LANG = "en" if locale.getdefaultlocale()[0].split('_')[0].lower() != 'zh' else "zh"
ROOT_DIR = os.getcwd()
MODEL_DIR = os.path.join(ROOT_DIR, 'pretrained_models')
STATIC_DIR = os.path.join(ROOT_DIR, 'static')
TMP_DIR = os.path.join(STATIC_DIR, 'tmp')
FILES_DIR = os.path.join(STATIC_DIR, 'files')
# Path: vocal/tool.py
def runffmpeg(arg):
def checkupdate():
def openweb(web_address):
# Path: vocal/cfg.py
ROOT_DIR = os.getcwd()
# Path: start.py
import logging
import threading
import sys
import os
import subprocess
from flask import Flask, request, render_template, jsonify, send_from_directory
from gevent.pywsgi import WSGIServer, WSGIHandler,LoggingLogAdapter
from logging.handlers import RotatingFileHandler
from vocal import cfg, tool
from vocal.cfg import ROOT_DIR
from spleeter.separator import Separator
class CustomRequestHandler(WSGIHandler):
def log_request(self):
pass
# 禁用 Werkzeug 默认的日志处理器
log = logging.getLogger('werkzeug')
log.handlers[:] = []
log.setLevel(logging.WARNING)
app = Flask(__name__, static_folder=os.path.join(ROOT_DIR, 'static'), static_url_path='/static',
template_folder=os.path.join(ROOT_DIR, 'templates'))
root_log = logging.getLogger() # Flask的根日志记录器
root_log.handlers = []
root_log.setLevel(logging.WARNING)
# 配置日志
app.logger.setLevel(logging.WARNING) # 设置日志级别为 INFO
# 创建 RotatingFileHandler 对象，设置写入的文件路径和大小限制
file_handler = RotatingFileHandler(os.path.join(ROOT_DIR, 'vocal.log'), maxBytes=1024 * 1024, backupCount=5)
# 创建日志的格式
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
# 设置文件处理器的级别和格式
file_handler.setLevel(logging.WARNING)
file_handler.setFormatter(formatter)
# 将文件处理器添加到日志记录器中
app.logger.addHandler(file_handler)
@app.route('/static/<path:filename>')
def static_files(filename):
return send_from_directory(app.config['STATIC_FOLDER'], filename)
@app.route('/')
def index():
return render_template("index.html",cuda=cfg.cuda, language=cfg.LANG,root_dir=ROOT_DIR.replace('\\', '/'))
# 上传音频
@app.route('/upload', methods=['POST'])
def upload():
try:
# 获取上传的文件
audio_file = request.files['audio']
# 如果是mp4
noextname, ext = os.path.splitext(audio_file.filename)
ext = ext.lower()
# 如果是视频，先分离
wav_file = os.path.join(cfg.TMP_DIR, f'{noextname}.wav')
if os.path.exists(wav_file) and os.path.getsize(wav_file) > 0:
return jsonify({'code': 0, 'msg': cfg.transobj['lang1'], "data": os.path.basename(wav_file)})
msg=""
if ext in ['.mp4', '.mov', '.avi', '.mkv', '.mpeg', '.mp3', '.flac']:
video_file = os.path.join(cfg.TMP_DIR, f'{noextname}{ext}')
audio_file.save(video_file)
params = [
"-i",
video_file,
]
if ext not in ['.mp3', '.flac']:
params.append('-vn')
params.append(wav_file)
|
self.layers = _get_clones(decoder_layer, num_layers) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: ali-vilab/dreamtalk
# Path: core/networks/transformer.py
def _get_activation_fn(activation):
"""Return an activation function given a string"""
if activation == "relu":
return F.relu
if activation == "gelu":
return F.gelu
if activation == "glu":
return F.glu
raise RuntimeError(F"activation should be relu/gelu, not {activation}.")
# Path: core/networks/transformer.py
def _get_clones(module, N):
return nn.ModuleList([copy.deepcopy(module) for i in range(N)])
# Path: core/networks/dynamic_linear.py
class DynamicLinear(nn.Module):
def __init__(self, in_planes, out_planes, cond_planes, bias=True, K=4, temperature=30, ratio=4, init_weight=True):
super().__init__()
self.dynamic_conv = DynamicConv(
in_planes,
out_planes,
cond_planes,
kernel_size=1,
stride=1,
padding=0,
bias=bias,
K=K,
ratio=ratio,
temperature=temperature,
init_weight=init_weight,
)
def forward(self, x, cond):
"""
Args:
x (_type_): (L, B, C_in)
cond (_type_): (B, C_style)
Returns:
_type_: (L, B, C_out)
"""
x = x.permute(1, 2, 0).unsqueeze(-1)
out = self.dynamic_conv(x, cond)
# (B, C_out, L, 1)
out = out.squeeze().permute(2, 0, 1)
return out
# Path: core/networks/dynamic_fc_decoder.py
import torch.nn as nn
import torch
from core.networks.transformer import _get_activation_fn, _get_clones
from core.networks.dynamic_linear import DynamicLinear
class DynamicFCDecoderLayer(nn.Module):
def __init__(
self,
d_model,
nhead,
d_style,
dynamic_K,
dynamic_ratio,
dim_feedforward=2048,
dropout=0.1,
activation="relu",
normalize_before=False,
):
super().__init__()
self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
self.multihead_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout)
# Implementation of Feedforward model
# self.linear1 = nn.Linear(d_model, dim_feedforward)
self.linear1 = DynamicLinear(d_model, dim_feedforward, d_style, K=dynamic_K, ratio=dynamic_ratio)
self.dropout = nn.Dropout(dropout)
self.linear2 = nn.Linear(dim_feedforward, d_model)
# self.linear2 = DynamicLinear(dim_feedforward, d_model, d_style, K=dynamic_K, ratio=dynamic_ratio)
self.norm1 = nn.LayerNorm(d_model)
self.norm2 = nn.LayerNorm(d_model)
self.norm3 = nn.LayerNorm(d_model)
self.dropout1 = nn.Dropout(dropout)
self.dropout2 = nn.Dropout(dropout)
self.dropout3 = nn.Dropout(dropout)
self.activation = _get_activation_fn(activation)
self.normalize_before = normalize_before
def with_pos_embed(self, tensor, pos):
return tensor if pos is None else tensor + pos
def forward_post(
self,
tgt,
memory,
style,
tgt_mask=None,
memory_mask=None,
tgt_key_padding_mask=None,
memory_key_padding_mask=None,
pos=None,
query_pos=None,
):
# q = k = self.with_pos_embed(tgt, query_pos)
tgt2 = self.self_attn(tgt, tgt, value=tgt, attn_mask=tgt_mask, key_padding_mask=tgt_key_padding_mask)[0]
tgt = tgt + self.dropout1(tgt2)
tgt = self.norm1(tgt)
tgt2 = self.multihead_attn(
query=tgt, key=memory, value=memory, attn_mask=memory_mask, key_padding_mask=memory_key_padding_mask
)[0]
tgt = tgt + self.dropout2(tgt2)
tgt = self.norm2(tgt)
# tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt, style))), style)
tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt, style))))
tgt = tgt + self.dropout3(tgt2)
tgt = self.norm3(tgt)
return tgt
# def forward_pre(
# self,
# tgt,
# memory,
# tgt_mask=None,
# memory_mask=None,
# tgt_key_padding_mask=None,
# memory_key_padding_mask=None,
# pos=None,
# query_pos=None,
# ):
# tgt2 = self.norm1(tgt)
# # q = k = self.with_pos_embed(tgt2, query_pos)
# tgt2 = self.self_attn(tgt2, tgt2, value=tgt2, attn_mask=tgt_mask, key_padding_mask=tgt_key_padding_mask)[0]
# tgt = tgt + self.dropout1(tgt2)
# tgt2 = self.norm2(tgt)
# tgt2 = self.multihead_attn(
# query=tgt2, key=memory, value=memory, attn_mask=memory_mask, key_padding_mask=memory_key_padding_mask
# )[0]
# tgt = tgt + self.dropout2(tgt2)
# tgt2 = self.norm3(tgt)
# tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt2))))
# tgt = tgt + self.dropout3(tgt2)
# return tgt
def forward(
self,
tgt,
memory,
style,
tgt_mask=None,
memory_mask=None,
tgt_key_padding_mask=None,
memory_key_padding_mask=None,
pos=None,
query_pos=None,
):
if self.normalize_before:
raise NotImplementedError
# return self.forward_pre(
# tgt, memory, tgt_mask, memory_mask, tgt_key_padding_mask, memory_key_padding_mask, pos, query_pos
# )
return self.forward_post(
tgt, memory, style, tgt_mask, memory_mask, tgt_key_padding_mask, memory_key_padding_mask, pos, query_pos
)
class DynamicFCDecoder(nn.Module):
def __init__(self, decoder_layer, num_layers, norm=None, return_intermediate=False):
super().__init__()
|
linear_cls = LoRACompatibleLinear if not USE_PEFT_BACKEND else nn.Linear | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: jiawei-ren/dreamgaussian4d
# Path: diffusers/src/diffusers/utils/constants.py
USE_PEFT_BACKEND = _required_peft_version and _required_transformers_version
# Path: diffusers/src/diffusers/models/lora.py
class LoRACompatibleLinear(nn.Linear):
"""
A Linear layer that can be used with LoRA.
"""
def __init__(self, *args, lora_layer: Optional[LoRALinearLayer] = None, **kwargs):
super().__init__(*args, **kwargs)
self.lora_layer = lora_layer
def set_lora_layer(self, lora_layer: Optional[LoRALinearLayer]):
self.lora_layer = lora_layer
def _fuse_lora(self, lora_scale: float = 1.0, safe_fusing: bool = False):
if self.lora_layer is None:
return
dtype, device = self.weight.data.dtype, self.weight.data.device
w_orig = self.weight.data.float()
w_up = self.lora_layer.up.weight.data.float()
w_down = self.lora_layer.down.weight.data.float()
if self.lora_layer.network_alpha is not None:
w_up = w_up * self.lora_layer.network_alpha / self.lora_layer.rank
fused_weight = w_orig + (lora_scale * torch.bmm(w_up[None, :], w_down[None, :])[0])
if safe_fusing and torch.isnan(fused_weight).any().item():
raise ValueError(
"This LoRA weight seems to be broken. "
f"Encountered NaN values when trying to fuse LoRA weights for {self}."
"LoRA weights will not be fused."
)
self.weight.data = fused_weight.to(device=device, dtype=dtype)
# we can drop the lora layer now
self.lora_layer = None
# offload the up and down matrices to CPU to not blow the memory
self.w_up = w_up.cpu()
self.w_down = w_down.cpu()
self._lora_scale = lora_scale
def _unfuse_lora(self):
if not (getattr(self, "w_up", None) is not None and getattr(self, "w_down", None) is not None):
return
fused_weight = self.weight.data
dtype, device = fused_weight.dtype, fused_weight.device
w_up = self.w_up.to(device=device).float()
w_down = self.w_down.to(device).float()
unfused_weight = fused_weight.float() - (self._lora_scale * torch.bmm(w_up[None, :], w_down[None, :])[0])
self.weight.data = unfused_weight.to(device=device, dtype=dtype)
self.w_up = None
self.w_down = None
def forward(self, hidden_states: torch.Tensor, scale: float = 1.0) -> torch.Tensor:
if self.lora_layer is None:
out = super().forward(hidden_states)
return out
else:
out = super().forward(hidden_states) + (scale * self.lora_layer(hidden_states))
return out
# Path: diffusers/src/diffusers/models/activations.py
import torch
import torch.nn.functional as F
from torch import nn
from ..utils import USE_PEFT_BACKEND
from .lora import LoRACompatibleLinear
# coding=utf-8
# Copyright 2023 HuggingFace Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
ACTIVATION_FUNCTIONS = {
"swish": nn.SiLU(),
"silu": nn.SiLU(),
"mish": nn.Mish(),
"gelu": nn.GELU(),
"relu": nn.ReLU(),
}
def get_activation(act_fn: str) -> nn.Module:
"""Helper function to get activation function from string.
Args:
act_fn (str): Name of activation function.
Returns:
nn.Module: Activation function.
"""
act_fn = act_fn.lower()
if act_fn in ACTIVATION_FUNCTIONS:
return ACTIVATION_FUNCTIONS[act_fn]
else:
raise ValueError(f"Unsupported activation function: {act_fn}")
class GELU(nn.Module):
r"""
GELU activation function with tanh approximation support with `approximate="tanh"`.
Parameters:
dim_in (`int`): The number of channels in the input.
dim_out (`int`): The number of channels in the output.
approximate (`str`, *optional*, defaults to `"none"`): If `"tanh"`, use tanh approximation.
"""
def __init__(self, dim_in: int, dim_out: int, approximate: str = "none"):
super().__init__()
self.proj = nn.Linear(dim_in, dim_out)
self.approximate = approximate
def gelu(self, gate: torch.Tensor) -> torch.Tensor:
if gate.device.type != "mps":
return F.gelu(gate, approximate=self.approximate)
# mps: gelu is not implemented for float16
return F.gelu(gate.to(dtype=torch.float32), approximate=self.approximate).to(dtype=gate.dtype)
def forward(self, hidden_states):
hidden_states = self.proj(hidden_states)
hidden_states = self.gelu(hidden_states)
return hidden_states
class GEGLU(nn.Module):
r"""
A [variant](https://arxiv.org/abs/2002.05202) of the gated linear unit activation function.
Parameters:
dim_in (`int`): The number of channels in the input.
dim_out (`int`): The number of channels in the output.
"""
def __init__(self, dim_in: int, dim_out: int):
super().__init__()
|
if (cur_input_ids == IMAGE_TOKEN_INDEX).sum() == 0: | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Meituan-AutoML/MobileVLM
# Path: mobilevlm/model/vision_encoder.py
def build_vision_tower(model_cfg, **kwargs):
vision_tower = getattr(model_cfg, 'mm_vision_tower', getattr(model_cfg, 'vision_tower', None))
is_absolute_path_exists = os.path.exists(vision_tower)
if is_absolute_path_exists or vision_tower.startswith("openai") or vision_tower.startswith("laion"):
vision_tower_type = getattr(model_cfg, 'vision_tower_type', None)
if vision_tower_type == "clip":
return CLIPVisionTower(vision_tower, args=model_cfg, **kwargs)
raise ValueError(f'Unknown vision tower: {vision_tower}')
# Path: mobilevlm/model/vision_projector.py
def build_vision_projector(config, delay_load=False, **kwargs):
projector_type = getattr(config, 'mm_projector_type', 'linear')
if projector_type == 'linear':
return nn.Linear(config.mm_hidden_size, config.hidden_size)
elif projector_type.startswith('mlp'):
mlp_gelu_match = re.match(r'^mlp(\d+)x_gelu$', projector_type)
if mlp_gelu_match:
mlp_depth = int(mlp_gelu_match.group(1))
modules = [nn.Linear(config.mm_hidden_size, config.hidden_size)]
for _ in range(1, mlp_depth):
modules.append(nn.GELU())
modules.append(nn.Linear(config.hidden_size, config.hidden_size))
return nn.Sequential(*modules)
elif projector_type.startswith('ldpnet'):
return LDPNetProjector(config)
raise ValueError(f'Unknown projector type: {projector_type}')
# Path: mobilevlm/constants.py
IGNORE_INDEX = -100
# Path: mobilevlm/constants.py
IMAGE_TOKEN_INDEX = -200
# Path: mobilevlm/constants.py
DEFAULT_IMAGE_PATCH_TOKEN = "<im_patch>"
# Path: mobilevlm/constants.py
DEFAULT_IM_START_TOKEN = "<im_start>"
# Path: mobilevlm/constants.py
DEFAULT_IM_END_TOKEN = "<im_end>"
# Path: mobilevlm/model/mobilevlm.py
import torch
import torch.nn as nn
from abc import ABC, abstractmethod
from transformers import AutoTokenizer, BitsAndBytesConfig
from mobilevlm.model.vision_encoder import build_vision_tower
from mobilevlm.model.vision_projector import build_vision_projector
from mobilevlm.constants import IGNORE_INDEX, IMAGE_TOKEN_INDEX, \
DEFAULT_IMAGE_PATCH_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN
from mobilevlm.model.mobilellama import MobileLlamaForCausalLM
class MobileVLMMetaModel:
def __init__(self, config):
super(MobileVLMMetaModel, self).__init__(config)
if hasattr(config, "mm_vision_tower"):
self.vision_tower = build_vision_tower(config, delay_load=False)
self.mm_projector = build_vision_projector(config)
def get_vision_tower(self):
vision_tower = getattr(self, 'vision_tower', None)
if type(vision_tower) is list:
vision_tower = vision_tower[0]
return vision_tower
def initialize_vision_modules(self, model_args, fsdp=None):
mm_vision_select_layer = model_args.mm_vision_select_layer
mm_vision_select_feature = model_args.mm_vision_select_feature
pretrain_mm_mlp_adapter = model_args.pretrain_mm_mlp_adapter
self.config.mm_vision_tower = model_args.vision_tower
self.config.use_mm_proj = True
self.config.mm_projector_type = getattr(model_args, 'mm_projector_type', 'linear')
self.config.mm_vision_select_layer = mm_vision_select_layer
self.config.mm_vision_select_feature = mm_vision_select_feature
# Build VisionTower
vision_tower = build_vision_tower(model_args)
if fsdp is not None and len(fsdp) > 0:
self.vision_tower = [vision_tower]
else:
self.vision_tower = vision_tower
self.config.mm_hidden_size = vision_tower.hidden_size
# Build Vision-Projector
self.mm_projector = build_vision_projector(self.config)
if pretrain_mm_mlp_adapter is not None:
mm_projector_weights = torch.load(pretrain_mm_mlp_adapter, map_location='cpu')
def get_w(weights, keyword):
return {k.split(keyword + '.')[1]: v for k, v in weights.items() if keyword in k}
self.mm_projector.load_state_dict(get_w(mm_projector_weights, 'mm_projector'))
class MobileVLMMetaForCausalLM(ABC):
@abstractmethod
def get_model(self):
pass
def get_vision_tower(self):
return self.get_model().get_vision_tower()
def encode_images(self, images):
image_features = self.get_model().get_vision_tower()(images)
image_features = self.get_model().mm_projector(image_features)
return image_features
def prepare_inputs_labels_for_multimodal(
self, input_ids, attention_mask, past_key_values, labels, images
):
vision_tower = self.get_vision_tower()
if vision_tower is None or images is None or input_ids.shape[1] == 1:
if past_key_values is not None and vision_tower is not None and images is not None and input_ids.shape[1] == 1:
attention_mask = torch.ones((attention_mask.shape[0], past_key_values[-1][-1].shape[-2] + 1), dtype=attention_mask.dtype, device=attention_mask.device)
return input_ids, attention_mask, past_key_values, None, labels
if type(images) is list or images.ndim == 5:
concat_images = torch.cat([image for image in images], dim=0)
image_features = self.encode_images(concat_images)
split_sizes = [image.shape[0] for image in images]
image_features = torch.split(image_features, split_sizes, dim=0)
image_features = [x.flatten(0, 1) for x in image_features]
else:
image_features = self.encode_images(images)
new_input_embeds = []
new_labels = [] if labels is not None else None
cur_image_idx = 0
for batch_idx, cur_input_ids in enumerate(input_ids):
|
configs = (lora_config, llama_adapter_config, prefix_config) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: kinggongzilla/ai-clone-whatsapp
# Path: configs/datasets.py
class custom_dataset:
# Path: configs/peft.py
class lora_config:
r: int=8
lora_alpha: int=32
target_modules: List[str] = field(default_factory=lambda: ["q_proj", "v_proj"])
bias= "none"
task_type: str= "CAUSAL_LM"
lora_dropout: float=0.05
inference_mode: bool = False
# Path: configs/peft.py
class llama_adapter_config:
adapter_len: int= 10
adapter_layers: int= 30
task_type: str= "CAUSAL_LM"
# Path: configs/peft.py
class prefix_config:
num_virtual_tokens: int=30
task_type: str= "CAUSAL_LM"
# Path: configs/training.py
class train_config:
whatsapp_username: str="" # your own whatsapp user name as it is in the chat .txt files
model_name: str="mistralai/Mistral-7B-Instruct-v0.2"
enable_fsdp: bool=False
low_cpu_fsdp: bool=False
run_validation: bool=False
batch_size_training: int=1
batching_strategy: str="packing" #alternative: padding
context_length: int=4096
gradient_accumulation_steps: int=1
gradient_clipping: bool = False
gradient_clipping_threshold: float = 1.0
num_epochs: int=1
num_workers_dataloader: int=1
lr: float=1e-4
weight_decay: float=0.0
gamma: float= 0.85
seed: int=42
use_fp16: bool=True
mixed_precision: bool=True
val_batch_size: int=1
dataset = "custom_dataset"
data_dir: str = "data/preprocessing/processed_chats"
peft_method: str = "lora" # None , llama_adapter, prefix
use_peft: bool=True
output_dir: str = "checkpoints"
freeze_layers: bool = False
num_freeze_layers: int = 1
quantization: bool = True
one_gpu: bool = False
save_model: bool = True
dist_checkpoint_root_folder: str="PATH/to/save/FSDP/model" # will be used if using FSDP
dist_checkpoint_folder: str="fine-tuned" # will be used if using FSDP
save_optimizer: bool=False # will be used if using FSDP
use_fast_kernels: bool = False # Enable using SDPA from PyTroch Accelerated Transformers, make use Flash Attention and Xformer memory-efficient kernels
# Path: data/sampler.py
class LengthBasedBatchSampler(torch.utils.data.BatchSampler):
def __init__(self, data_source, batch_size: int, drop_last: bool, shuffle: bool=True) -> None:
if isinstance(next(iter(data_source)), dict):
first_key = next(iter(next(iter(data_source)).keys()))
self.lengths = [len(d[first_key]) for d in data_source]
else:
self.lengths = [len(d) for d in data_source]
self.batch_size = batch_size
self.drop_last = drop_last
self.shuffle = shuffle
def __iter__(self):
ids = np.argsort(self.lengths)
if self.drop_last:
ids = ids[:len(ids) // self.batch_size * self.batch_size]
batches = [ids[i:i+self.batch_size] for i in range(0, len(ids), self.batch_size)]
if self.shuffle:
random.shuffle(batches)
for b in batches:
yield b
def __len__(self):
if self.drop_last:
return len(self.lengths) // self.batch_size
else:
return len(self.lengths) // self.batch_size + (len(self.lengths) % self.batch_size > 0)
# Path: data/sampler.py
class DistributedLengthBasedBatchSampler(torch.utils.data.BatchSampler):
def __init__(self, data_source, batch_size: int, num_replicas: int, rank: int, shuffle: bool = True, seed: int = 0) -> None:
random.seed(seed)
self.batch_sampler = LengthBasedBatchSampler(
data_source, batch_size=batch_size, drop_last=True, shuffle=shuffle
)
self.num_replicas = num_replicas
self.rank = rank
def __iter__(self):
max_length = len(self.batch_sampler) // self.num_replicas * self.num_replicas
return islice(self.batch_sampler, self.rank, max_length, self.num_replicas)
def __len__(self):
return len(self.batch_sampler) // self.num_replicas
# Path: utils/dataset_utils.py
DATASET_PREPROC = {
"custom_dataset": get_custom_dataset,
}
# Path: utils/config_utils.py
import inspect
import torch.distributed as dist
from dataclasses import asdict
from torch.utils.data import DistributedSampler
from peft import (
LoraConfig,
AdaptionPromptConfig,
PrefixTuningConfig,
)
from transformers import default_data_collator
from transformers.data import DataCollatorForSeq2Seq
from configs import datasets, lora_config, llama_adapter_config, prefix_config, train_config
from data.sampler import LengthBasedBatchSampler, DistributedLengthBasedBatchSampler
from utils.dataset_utils import DATASET_PREPROC
# Copyright (c) Meta Platforms, Inc. and affiliates.
# This software may be used and distributed according to the terms of the Llama 2 Community License Agreement.
def update_config(config, **kwargs):
if isinstance(config, (tuple, list)):
for c in config:
update_config(c, **kwargs)
else:
for k, v in kwargs.items():
if hasattr(config, k):
setattr(config, k, v)
elif "." in k:
# allow --some_config.some_param=True
config_name, param_name = k.split(".")
if type(config).__name__ == config_name:
if hasattr(config, param_name):
setattr(config, param_name, v)
else:
# In case of specialized config we can warm user
print(f"Warning: {config_name} does not accept parameter: {k}")
elif isinstance(config, train_config):
print(f"Warning: unknown parameter {k}")
def generate_peft_config(train_config, kwargs):
|
cost_giou = -generalized_box_iou(box_cxcywh_to_xyxy(bz_boxes), box_cxcywh_to_xyxy(bz_gtboxs)) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: FoundationVision/UniRef
# Path: projects/UniRef/uniref/util/box_ops.py
def box_cxcywh_to_xyxy(x):
# print('box:\n', x)
x_c, y_c, w, h = x.unbind(-1)
b = [(x_c - 0.5 * w), (y_c - 0.5 * h),
(x_c + 0.5 * w), (y_c + 0.5 * h)]
return torch.stack(b, dim=-1)
# Path: projects/UniRef/uniref/util/box_ops.py
def generalized_box_iou(boxes1, boxes2):
"""
Generalized IoU from https://giou.stanford.edu/
The boxes should be in [x0, y0, x1, y1] format
Returns a [N, M] pairwise matrix, where N = len(boxes1)
and M = len(boxes2)
"""
# degenerate boxes gives inf / nan results
# so do an early check
assert (boxes1[:, 2:] >= boxes1[:, :2]).all()
assert (boxes2[:, 2:] >= boxes2[:, :2]).all()
iou, union = box_iou(boxes1, boxes2)
lt = torch.min(boxes1[:, None, :2], boxes2[:, :2])
rb = torch.max(boxes1[:, None, 2:], boxes2[:, 2:])
wh = (rb - lt).clamp(min=0) # [N,M,2]
area = wh[:, :, 0] * wh[:, :, 1]
return iou - (area - union) / (area+1e-7)
# Path: projects/UniRef/uniref/models/deformable_detr/matcher.py
import torch
import torch.nn.functional as F
import torchvision.ops as ops
from scipy.optimize import linear_sum_assignment
from torch import nn
from ...util.box_ops import box_cxcywh_to_xyxy, generalized_box_iou
# ------------------------------------------------------------------------
# Deformable DETR
# Copyright (c) 2020 SenseTime. All Rights Reserved.
# Licensed under the Apache License, Version 2.0 [see LICENSE for details]
# ------------------------------------------------------------------------
# Modified from DETR (https://github.com/facebookresearch/detr)
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
# ------------------------------------------------------------------------
"""
Modules to compute the matching cost and solve the corresponding LSAP.
"""
class HungarianMatcher(nn.Module):
"""This class computes an assignment between the targets and the predictions of the network
For efficiency reasons, the targets don't include the no_object. Because of this, in general,
there are more predictions than targets. In this case, we do a 1-to-1 matching of the best predictions,
while the others are un-matched (and thus treated as non-objects).
"""
def __init__(self,
cost_class: float = 1,
cost_bbox: float = 1,
cost_giou: float = 1):
"""Creates the matcher
Params:
cost_class: This is the relative weight of the classification error in the matching cost
cost_bbox: This is the relative weight of the L1 error of the bounding box coordinates in the matching cost
cost_giou: This is the relative weight of the giou loss of the bounding box in the matching cost
"""
super().__init__()
self.cost_class = cost_class
self.cost_bbox = cost_bbox
self.cost_giou = cost_giou
assert cost_class != 0 or cost_bbox != 0 or cost_giou != 0, "all costs cant be 0"
def forward_ota(self, outputs, targets):
""" simOTA for detr
"""
with torch.no_grad():
bs, num_queries = outputs["pred_logits"].shape[:2]
out_prob = outputs["pred_logits"].sigmoid()
out_bbox = outputs["pred_boxes"] # 跳过frame 维度
indices = []
matched_ids = []
for batch_idx in range(bs):
bz_boxes = out_bbox[batch_idx] #[300,4]
bz_out_prob = out_prob[batch_idx]
bz_tgt_ids = targets[batch_idx]["labels"]
num_insts = len(bz_tgt_ids)
bz_gtboxs = targets[batch_idx]['boxes'].reshape(num_insts,4) #[num_gt, 4]
fg_mask, is_in_boxes_and_center = \
self.get_in_boxes_info(bz_boxes,bz_gtboxs,expanded_strides=32)
pair_wise_ious = ops.box_iou(box_cxcywh_to_xyxy(bz_boxes), box_cxcywh_to_xyxy(bz_gtboxs))
# pair_wise_ious_loss = -torch.log(pair_wise_ious + 1e-8)
# Compute the classification cost.
alpha = 0.25
gamma = 2.0
neg_cost_class = (1 - alpha) * (bz_out_prob ** gamma) * (-(1 - bz_out_prob + 1e-8).log())
pos_cost_class = alpha * ((1 - bz_out_prob) ** gamma) * (-(bz_out_prob + 1e-8).log())
cost_class = pos_cost_class[:, bz_tgt_ids] - neg_cost_class[:, bz_tgt_ids]
|
self.encoding = get_encoding(3, self.cfg.dir_encoding_config) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: xhuangcv/humannorm
# Path: threestudio/models/materials/base.py
class BaseMaterial(BaseModule):
@dataclass
class Config(BaseModule.Config):
pass
cfg: Config
requires_normal: bool = False
requires_tangent: bool = False
def configure(self):
pass
def forward(self, *args, **kwargs) -> Float[Tensor, "*B 3"]:
raise NotImplementedError
def export(self, *args, **kwargs) -> Dict[str, Any]:
return {}
# Path: threestudio/models/networks.py
def get_encoding(n_input_dims: int, config) -> nn.Module:
# input suppose to be range [0, 1]
encoding: nn.Module
if config.otype == "ProgressiveBandFrequency":
encoding = ProgressiveBandFrequency(n_input_dims, config_to_primitive(config))
elif config.otype == "ProgressiveBandHashGrid":
encoding = ProgressiveBandHashGrid(n_input_dims, config_to_primitive(config))
else:
encoding = TCNNEncoding(n_input_dims, config_to_primitive(config))
encoding = CompositeEncoding(
encoding,
include_xyz=config.get("include_xyz", False),
xyz_scale=2.0,
xyz_offset=-1.0,
) # FIXME: hard coded
return encoding
# Path: threestudio/models/networks.py
def get_mlp(n_input_dims, n_output_dims, config) -> nn.Module:
network: nn.Module
if config.otype == "VanillaMLP":
network = VanillaMLP(n_input_dims, n_output_dims, config_to_primitive(config))
elif config.otype == "SphereInitVanillaMLP":
network = SphereInitVanillaMLP(
n_input_dims, n_output_dims, config_to_primitive(config)
)
else:
assert (
config.get("sphere_init", False) is False
), "sphere_init=True only supported by VanillaMLP"
network = TCNNNetwork(n_input_dims, n_output_dims, config_to_primitive(config))
return network
# Path: threestudio/utils/ops.py
def dot(x, y):
return torch.sum(x * y, -1, keepdim=True)
# Path: threestudio/utils/ops.py
def get_activation(name) -> Callable:
if name is None:
return lambda x: x
name = name.lower()
if name == "none":
return lambda x: x
elif name == "lin2srgb":
return lambda x: torch.where(
x > 0.0031308,
torch.pow(torch.clamp(x, min=0.0031308), 1.0 / 2.4) * 1.055 - 0.055,
12.92 * x,
).clamp(0.0, 1.0)
elif name == "exp":
return lambda x: torch.exp(x)
elif name == "shifted_exp":
return lambda x: torch.exp(x - 1.0)
elif name == "trunc_exp":
return trunc_exp
elif name == "shifted_trunc_exp":
return lambda x: trunc_exp(x - 1.0)
elif name == "sigmoid":
return lambda x: torch.sigmoid(x)
elif name == "tanh":
return lambda x: torch.tanh(x)
elif name == "shifted_softplus":
return lambda x: F.softplus(x - 1.0)
elif name == "scale_-11_01":
return lambda x: x * 0.5 + 0.5
else:
try:
return getattr(F, name)
except AttributeError:
raise ValueError(f"Unknown activation function: {name}")
# Path: threestudio/models/materials/neural_radiance_material.py
import random
import torch
import torch.nn as nn
import torch.nn.functional as F
import threestudio
from dataclasses import dataclass, field
from threestudio.models.materials.base import BaseMaterial
from threestudio.models.networks import get_encoding, get_mlp
from threestudio.utils.ops import dot, get_activation
from threestudio.utils.typing import *
@threestudio.register("neural-radiance-material")
class NeuralRadianceMaterial(BaseMaterial):
@dataclass
class Config(BaseMaterial.Config):
input_feature_dims: int = 8
color_activation: str = "sigmoid"
dir_encoding_config: dict = field(
default_factory=lambda: {"otype": "SphericalHarmonics", "degree": 3}
)
mlp_network_config: dict = field(
default_factory=lambda: {
"otype": "FullyFusedMLP",
"activation": "ReLU",
"n_neurons": 16,
"n_hidden_layers": 2,
}
)
cfg: Config
def configure(self) -> None:
|
rs = tool.runffmpeg(params) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: jianchang512/stt
# Path: stslib/cfg.py
LANG = "en" if locale.getdefaultlocale()[0].split('_')[0].lower() != 'zh' else "zh"
ROOT_DIR = os.getcwd()
MODEL_DIR = os.path.join(ROOT_DIR, 'models')
STATIC_DIR = os.path.join(ROOT_DIR, 'static')
TMP_DIR = os.path.join(STATIC_DIR, 'tmp')
# Path: stslib/tool.py
def runffmpeg(arg):
def checkupdate():
def openweb(web_address):
def ms_to_time_string(*, ms=0, seconds=None):
# Path: stslib/cfg.py
ROOT_DIR = os.getcwd()
# Path: start.py
import logging
import re
import threading
import sys
import torch
import os
from flask import Flask, request, render_template, jsonify, send_from_directory
from gevent.pywsgi import WSGIServer, WSGIHandler, LoggingLogAdapter
from logging.handlers import RotatingFileHandler
from stslib import cfg, tool
from stslib.cfg import ROOT_DIR
from faster_whisper import WhisperModel
device = "cuda" if torch.cuda.is_available() else "cpu"
class CustomRequestHandler(WSGIHandler):
def log_request(self):
pass
# 配置日志
# 禁用 Werkzeug 默认的日志处理器
log = logging.getLogger('werkzeug')
log.handlers[:] = []
log.setLevel(logging.WARNING)
app = Flask(__name__, static_folder=os.path.join(ROOT_DIR, 'static'), static_url_path='/static',
template_folder=os.path.join(ROOT_DIR, 'templates'))
root_log = logging.getLogger() # Flask的根日志记录器
root_log.handlers = []
root_log.setLevel(logging.WARNING)
# 配置日志
app.logger.setLevel(logging.WARNING) # 设置日志级别为 INFO
# 创建 RotatingFileHandler 对象，设置写入的文件路径和大小限制
file_handler = RotatingFileHandler(os.path.join(ROOT_DIR, 'sts.log'), maxBytes=1024 * 1024, backupCount=5)
# 创建日志的格式
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
# 设置文件处理器的级别和格式
file_handler.setLevel(logging.WARNING)
file_handler.setFormatter(formatter)
# 将文件处理器添加到日志记录器中
app.logger.addHandler(file_handler)
@app.route('/static/<path:filename>')
def static_files(filename):
return send_from_directory(app.config['STATIC_FOLDER'], filename)
@app.route('/')
def index():
return render_template("index.html",
cuda=cfg.cuda,
lang_code=cfg.lang_code,
language=cfg.LANG,
root_dir=ROOT_DIR.replace('\\', '/'))
# 上传音频
@app.route('/upload', methods=['POST'])
def upload():
try:
# 获取上传的文件
audio_file = request.files['audio']
# 如果是mp4
noextname, ext = os.path.splitext(audio_file.filename)
ext = ext.lower()
# 如果是视频，先分离
wav_file = os.path.join(cfg.TMP_DIR, f'{noextname}.wav')
if os.path.exists(wav_file) and os.path.getsize(wav_file) > 0:
return jsonify({'code': 0, 'msg': cfg.transobj['lang1'], "data": os.path.basename(wav_file)})
msg = ""
if ext in ['.mp4', '.mov', '.avi', '.mkv', '.mpeg', '.mp3', '.flac']:
video_file = os.path.join(cfg.TMP_DIR, f'{noextname}{ext}')
audio_file.save(video_file)
params = [
"-i",
video_file,
]
if ext not in ['.mp3', '.flac']:
params.append('-vn')
params.append(wav_file)
|
self.similar_filter = SimilarImageFilter() | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: jesenzhang/ComfyUI_StreamDiffusion
# Path: streamdiffusion/image_filter.py
class SimilarImageFilter:
def __init__(self, threshold: float = 0.98, max_skip_frame: float = 10) -> None:
self.threshold = threshold
self.prev_tensor = None
self.cos = torch.nn.CosineSimilarity(dim=0, eps=1e-6)
self.max_skip_frame = max_skip_frame
self.skip_count = 0
def __call__(self, x: torch.Tensor) -> Optional[torch.Tensor]:
if self.prev_tensor is None:
self.prev_tensor = x.detach().clone()
return x
else:
cos_sim = self.cos(self.prev_tensor.reshape(-1), x.reshape(-1)).item()
sample = random.uniform(0, 1)
if self.threshold >= 1:
skip_prob = 0
else:
skip_prob = max(0, 1 - (1 - cos_sim) / (1 - self.threshold))
# not skip frame
if skip_prob < sample:
self.prev_tensor = x.detach().clone()
return x
# skip frame
else:
if self.skip_count > self.max_skip_frame:
self.skip_count = 0
self.prev_tensor = x.detach().clone()
return x
else:
self.skip_count += 1
return None
def set_threshold(self, threshold: float) -> None:
self.threshold = threshold
def set_max_skip_frame(self, max_skip_frame: float) -> None:
self.max_skip_frame = max_skip_frame
# Path: streamdiffusion/image_utils.py
def postprocess_image(
image: torch.Tensor,
output_type: str = "pil",
do_denormalize: Optional[List[bool]] = None,
) -> Union[torch.Tensor, np.ndarray, PIL.Image.Image]:
if not isinstance(image, torch.Tensor):
raise ValueError(
f"Input for postprocessing is in incorrect format: {type(image)}. We only support pytorch tensor"
)
if output_type == "latent":
return image
do_normalize_flg = True
if do_denormalize is None:
do_denormalize = [do_normalize_flg] * image.shape[0]
image = torch.stack(
[
denormalize(image[i]) if do_denormalize[i] else image[i]
for i in range(image.shape[0])
]
)
if output_type == "pt":
return image
image = pt_to_numpy(image)
if output_type == "np":
return image
if output_type == "pil":
return numpy_to_pil(image)
# Path: streamdiffusion/pipeline.py
import time
import numpy as np
import PIL.Image
import torch
from typing import List, Optional, Union, Any, Dict, Tuple, Literal
from diffusers import LCMScheduler, StableDiffusionPipeline
from diffusers.image_processor import VaeImageProcessor
from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_img2img import (
retrieve_latents,
)
from .image_filter import SimilarImageFilter
from .image_utils import postprocess_image
class StreamDiffusion:
def __init__(
self,
pipe: StableDiffusionPipeline,
t_index_list: List[int],
torch_dtype: torch.dtype = torch.float16,
width: int = 512,
height: int = 512,
do_add_noise: bool = True,
use_denoising_batch: bool = True,
frame_buffer_size: int = 1,
cfg_type: Literal["none", "full", "self", "initialize"] = "self",
) -> None:
self.device = pipe.device
self.dtype = torch_dtype
self.generator = None
self.height = height
self.width = width
self.latent_height = int(height // pipe.vae_scale_factor)
self.latent_width = int(width // pipe.vae_scale_factor)
self.frame_bff_size = frame_buffer_size
self.denoising_steps_num = len(t_index_list)
self.cfg_type = cfg_type
if use_denoising_batch:
self.batch_size = self.denoising_steps_num * frame_buffer_size
if self.cfg_type == "initialize":
self.trt_unet_batch_size = (
self.denoising_steps_num + 1
) * self.frame_bff_size
elif self.cfg_type == "full":
self.trt_unet_batch_size = (
2 * self.denoising_steps_num * self.frame_bff_size
)
else:
self.trt_unet_batch_size = self.denoising_steps_num * frame_buffer_size
else:
self.trt_unet_batch_size = self.frame_bff_size
self.batch_size = frame_buffer_size
self.t_list = t_index_list
self.do_add_noise = do_add_noise
self.use_denoising_batch = use_denoising_batch
self.similar_image_filter = False
|
for key, value in _state_dict(model).items(): | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: neobundy/MLX-Stable-Diffusion-WebUI
# Path: stable_diffusion/config.py
class DiffuserModelPathConfig:
class BaseConfig:
class AutoencoderConfig(BaseConfig):
class CLIPTextModelConfig(BaseConfig):
class UNetConfig(BaseConfig):
class DiffusionConfig(BaseConfig):
def __init__(self, model_path: str = "./diffuser_models"):
def unet_config(self):
def unet(self):
def scheduler(self):
def text_encoder_config(self):
def text_encoder(self):
def vae_config(self):
def vae(self):
def diffusion_config(self):
def tokenizer_vocab(self):
def tokenizer_merges(self):
def __getitem__(self, key):
def __setitem__(self, key, value):
# Path: stable_diffusion/model_io.py
_DEBUG = False
def _debug_print(*args, **kwargs):
def _from_numpy(x):
def map_unet_weights(key, value):
def map_clip_text_encoder_weights(key, value):
def map_vae_weights(key, value):
def _flatten(params):
def _load_safetensor_weights(mapper, model, weight_file, float16: bool = False):
def _check_key(key: str, part: str):
def load_unet(key: str = _DEFAULT_MODEL, float16: bool = False):
def load_text_encoder(key: str = _DEFAULT_MODEL, float16: bool = False):
def load_autoencoder(key: str = _DEFAULT_MODEL, float16: bool = False):
def load_diffusion_config(key: str = _DEFAULT_MODEL):
def load_tokenizer(key: str = _DEFAULT_MODEL):
def load_unet_local(weights_path: str, config_path: str, float16: bool = False):
def load_text_encoder_local(weights_path: str, config_path: str, float16: bool = False):
def load_autoencoder_local(weights_path: str, config_path: str, float16: bool = False):
def load_diffusion_config_local(config_path:str):
def load_tokenizer_local(vocab_path: str, merges_path: str):
def load_diffuser_model(diffuser_model_path: str, float16: bool = False):
# Path: utils.py
def _state_dict(model):
"""Return the model's state_dict as a dictionary."""
state_dict = {}
for name, param in model.parameters().items():
state_dict[name] = param
return state_dict
# Path: utils.py
def get_state_dict_from_safetensor(checkpoint_path: str):
"""Return the state_dict from the checkpoint."""
state_dict = {}
with safetensor_open(checkpoint_path, framework="numpy") as f:
# Access the data in the file
for key in f.keys():
tensor = f.get_tensor(key)
state_dict[key] = tensor
return state_dict
# Path: model_inspector.py
from stable_diffusion.config import PathConfig
from stable_diffusion.model_io import preload_models_from_safetensor_weights
from utils import _state_dict
from utils import get_state_dict_from_safetensor
INSPECTION_FILE = "model_inspection.txt"
NUM_ITEMS = 100
MODEL_FILE = "./models/v2-1_512-ema-pruned.safetensors"
MODEL_FILE1 = "./unet/diffusion_pytorch_model_test.safetensors"
MODEL_FILE2 = "./unet/xxmix9realistic_v40.safetensors"
# Recreate the inspection file at every execution of the script
with open(INSPECTION_FILE, 'w') as f:
pass
def write_to_file(*args, **kwargs):
"""Write the text to the inspection file."""
# Convert the arguments to a string
message = ' '.join(map(str, args))
# Print the message to the console
print(message, **kwargs)
# Open the log file in append mode and write the message
with open(INSPECTION_FILE, 'a') as f:
f.write(message + '\n')
def inspect_model(path_config: PathConfig, keys_only=True):
"""Inspect the contents of the models."""
# Load the models using the provided config and weights paths
unet_model = load_unet_local(path_config.unet_config, MODEL_FILE)
text_encoder_model = load_text_encoder_local(MODEL_FILE)
autoencoder_model = load_autoencoder_local(MODEL_FILE)
diffusion_config = load_diffusion_config_local(path_config.diffusion_config)
tokenizer = load_tokenizer_local(path_config.tokenizer_vocab, path_config.tokenizer_merges)
# Convert the models' state_dict to a dictionary and iterate over it
for model_name, model in zip(["unet", "text_encoder", "autoencoder"], [unet_model, text_encoder_model, autoencoder_model]):
write_to_file("-" * 50)
write_to_file(f"Model: {model_name}")
write_to_file("-" * 50)
|
.where(meme_source.c.type == MemeSourceType.TELEGRAM) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: ffmemes/ff-backend
# Path: src/database.py
DATABASE_URL = str(settings.DATABASE_URL)
async def fetch_one(select_query: Select | Insert | Update) -> dict[str, Any] | None:
async def fetch_all(select_query: Select | Insert | Update) -> list[dict[str, Any]]:
async def execute(select_query: Insert | Update) -> CursorResult:
# Path: src/storage/parsers/schemas.py
class TgChannelPostParsingResult(CustomModel):
post_id: int
url: str
content: str | None = None # post text
media: list[dict] | None = None
views: int
date: datetime
mentions: list[str] | None = None # mentioned usernames
hashtags: list[str] | None = None
forwarded: dict | None = None
forwarded_url: str | None = None # url to forwarded post
link_preview: dict | None = None
out_links: list[str] | None = None
# Path: src/storage/parsers/schemas.py
class VkGroupPostParsingResult(CustomModel):
post_id: str
url: str
content: str | None = None # post text
media: list[str]
date: datetime
views: int
likes: int
reposts: int
comments: int
# Path: src/storage/constants.py
class MemeSourceType(str, Enum):
TELEGRAM = "telegram"
VK = "vk"
REDDIT = "reddit"
INSTAGRAM = "instagram"
TWITTER = "twitter"
TIKTOK = "tiktok"
USER_UPLOAD = "user upload"
# Path: src/storage/constants.py
class MemeSourceStatus(str, Enum):
IN_MODERATION = "in_moderation"
PARSING_ENABLED = "parsing_enabled"
PARSING_DISABLED = "parsing_disabled"
# Path: src/storage/constants.py
class MemeType(str, Enum):
IMAGE = "image"
ANIMATION = "animation"
VIDEO = "video"
# Path: src/storage/constants.py
class MemeStatus(str, Enum):
CREATED = "created"
OK = "ok"
DUPLICATE = "duplicate"
AD = "ad"
BROKEN_CONTENT_LINK = "broken_content_link"
# TODO: more statuses?
# IN_MODERATION = "in_moderation"
# Path: src/storage/constants.py
MEME_RAW_TELEGRAM_MEME_SOURCE_POST_UNIQUE_CONSTRAINT = "meme_raw_telegram_meme_source_id_post_id_key"
# Path: src/storage/constants.py
MEME_RAW_VK_MEME_SOURCE_POST_UNIQUE_CONSTRAINT = "meme_raw_vk_meme_source_id_post_id_key"
# Path: src/storage/service.py
from typing import Any
from datetime import datetime
from sqlalchemy import select, nulls_first, text
from sqlalchemy.dialects.postgresql import insert
from src.database import (
language,
meme,
meme_source,
meme_raw_telegram,
meme_raw_vk,
execute, fetch_one, fetch_all,
)
from src.storage.parsers.schemas import TgChannelPostParsingResult, VkGroupPostParsingResult
from src.storage.constants import (
MemeSourceType,
MemeSourceStatus,
MemeType,
MemeStatus,
MEME_RAW_TELEGRAM_MEME_SOURCE_POST_UNIQUE_CONSTRAINT,
MEME_RAW_VK_MEME_SOURCE_POST_UNIQUE_CONSTRAINT,
)
async def insert_parsed_posts_from_telegram(
meme_source_id: int,
telegram_posts: list[TgChannelPostParsingResult],
) -> None:
posts = [
post.model_dump() | {"meme_source_id": meme_source_id}
for post in telegram_posts
]
insert_statement = insert(meme_raw_telegram).values(posts)
insert_posts_query = insert_statement.on_conflict_do_update(
constraint=MEME_RAW_TELEGRAM_MEME_SOURCE_POST_UNIQUE_CONSTRAINT,
set_={
"media": insert_statement.excluded.media,
"views": insert_statement.excluded.views,
"updated_at": datetime.utcnow(),
},
)
await execute(insert_posts_query)
async def insert_parsed_posts_from_vk(
meme_source_id: int,
vk_posts: list[VkGroupPostParsingResult],
) -> None:
posts = [
post.model_dump() | {"meme_source_id": meme_source_id}
for post in vk_posts
]
insert_statement = insert(meme_raw_vk).values(posts)
insert_posts_query = insert_statement.on_conflict_do_update(
constraint=MEME_RAW_VK_MEME_SOURCE_POST_UNIQUE_CONSTRAINT,
set_={
"media": insert_statement.excluded.media,
"views": insert_statement.excluded.views,
"likes": insert_statement.excluded.likes,
"reposts": insert_statement.excluded.reposts,
"comments": insert_statement.excluded.comments,
"updated_at": datetime.utcnow(),
},
)
await execute(insert_posts_query)
async def get_telegram_sources_to_parse(limit=10) -> list[dict[str, Any]]:
select_query = (
select(meme_source)
|
self.target = encode_prompts(tokenizer, text_encoder, [target_prompt]) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Con6924/SPM
# Path: src/misc/clip_templates.py
# Path: src/engine/train_util.py
def encode_prompts(
tokenizer: CLIPTokenizer,
text_encoder: CLIPTokenizer,
prompts: list[str],
return_tokens: bool = False,
):
text_tokens = text_tokenize(tokenizer, prompts)
text_embeddings = text_encode(text_encoder, text_tokens)
if return_tokens:
return text_embeddings, torch.unique(text_tokens, dim=1)
return text_embeddings
# Path: src/configs/prompt.py
from typing import Literal, Optional, Union
from pathlib import Path
from pydantic import BaseModel, root_validator
from transformers import CLIPTextModel, CLIPTokenizer
from src.misc.clip_templates import imagenet_templates
from src.engine.train_util import encode_prompts
import yaml
import pandas as pd
import random
import torch
class PromptEmbedsXL:
text_embeds: torch.FloatTensor
pooled_embeds: torch.FloatTensor
def __init__(self, embeds) -> None:
self.text_embeds, self.pooled_embeds = embeds
PROMPT_EMBEDDING = Union[torch.FloatTensor, PromptEmbedsXL]
class PromptEmbedsCache:
prompts: dict[str, PROMPT_EMBEDDING] = {}
def __setitem__(self, __name: str, __value: PROMPT_EMBEDDING) -> None:
self.prompts[__name] = __value
def __getitem__(self, __name: str) -> Optional[PROMPT_EMBEDDING]:
if __name in self.prompts:
return self.prompts[__name]
else:
return None
class PromptSettings(BaseModel): # yaml
target: str
positive: str = None # if None, target will be used
unconditional: str = "" # default is ""
neutral: str = None # if None, unconditional will be used
action: ACTION_TYPES = "erase" # default is "erase"
guidance_scale: float = 1.0 # default is 1.0
resolution: int = 512 # default is 512
dynamic_resolution: bool = False # default is False
batch_size: int = 1 # default is 1
dynamic_crops: bool = False # default is False. only used when model is XL
use_template: bool = False # default is False
la_strength: float = 1000.0
sampling_batch_size: int = 4
seed: int = None
case_number: int = 0
@root_validator(pre=True)
def fill_prompts(cls, values):
keys = values.keys()
if "target" not in keys:
raise ValueError("target must be specified")
if "positive" not in keys:
values["positive"] = values["target"]
if "unconditional" not in keys:
values["unconditional"] = ""
if "neutral" not in keys:
values["neutral"] = values["unconditional"]
return values
class PromptEmbedsPair:
target: PROMPT_EMBEDDING # the concept that do not want to generate
positive: PROMPT_EMBEDDING # generate the concept
unconditional: PROMPT_EMBEDDING # uncondition (default should be empty)
neutral: PROMPT_EMBEDDING # base condition (default should be empty)
use_template: bool = False # use clip template or not
guidance_scale: float
resolution: int
dynamic_resolution: bool
batch_size: int
dynamic_crops: bool
loss_fn: torch.nn.Module
action: ACTION_TYPES
def __init__(
self,
loss_fn: torch.nn.Module,
target: PROMPT_EMBEDDING,
positive: PROMPT_EMBEDDING,
unconditional: PROMPT_EMBEDDING,
neutral: PROMPT_EMBEDDING,
settings: PromptSettings,
) -> None:
self.loss_fn = loss_fn
self.target = target
self.positive = positive
self.unconditional = unconditional
self.neutral = neutral
self.settings = settings
self.use_template = settings.use_template
self.guidance_scale = settings.guidance_scale
self.resolution = settings.resolution
self.dynamic_resolution = settings.dynamic_resolution
self.batch_size = settings.batch_size
self.dynamic_crops = settings.dynamic_crops
self.action = settings.action
self.la_strength = settings.la_strength
self.sampling_batch_size = settings.sampling_batch_size
def _prepare_embeddings(
self,
cache: PromptEmbedsCache,
tokenizer: CLIPTokenizer,
text_encoder: CLIPTextModel,
):
"""
Prepare embeddings for training. When use_template is True, the embeddings will be
format using a template, and then be processed by the model.
"""
if not self.use_template:
return
template = random.choice(imagenet_templates)
target_prompt = template.format(self.settings.target)
if cache[target_prompt]:
self.target = cache[target_prompt]
else:
|
input_seqs = read_fasta_file(fasta_file) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: dakpinaroglu/Frame2seq
# Path: frame2seq/utils/residue_constants.py
def load_stereo_chemical_props() -> Tuple[Mapping[str, List[Bond]],
def make_bond_key(atom1_name, atom2_name):
def sequence_to_onehot(
sequence: str,
mapping: Mapping[str, int],
) -> np.ndarray:
def _make_standard_atom_mask() -> np.ndarray:
def _make_rigid_transformation_4x4(ex, ey, translation):
AA_TO_ID = {
'A': 0,
'C': 1,
'D': 2,
'E': 3,
'F': 4,
'G': 5,
'H': 6,
'I': 7,
'K': 8,
'L': 9,
'M': 10,
'N': 11,
'P': 12,
'Q': 13,
'R': 14,
'S': 15,
'T': 16,
'V': 17,
'W': 18,
'Y': 19,
'X': 20,
}
ID_TO_AA = {
0: 'A',
1: 'C',
2: 'D',
3: 'E',
4: 'F',
5: 'G',
6: 'H',
7: 'I',
8: 'K',
9: 'L',
10: 'M',
11: 'N',
12: 'P',
13: 'Q',
14: 'R',
15: 'S',
16: 'T',
17: 'V',
18: 'W',
19: 'Y',
20: 'X',
}
STANDARD_ATOM_MASK = _make_standard_atom_mask()
# Path: frame2seq/utils/util.py
def get_neg_pll(probs, seq):
seq_probs = torch.gather(probs, 1, seq.unsqueeze(-1)).squeeze(-1)
neg_pll = -1 * torch.log(seq_probs)
avg_neg_pll = neg_pll.sum().item() / len(neg_pll)
return neg_pll, avg_neg_pll
# Path: frame2seq/utils/util.py
def read_fasta_file(fasta_file):
"""
Read a fasta file and return a list of sequences.
"""
with open(fasta_file, 'r') as f:
lines = f.readlines()
sequences = []
for line in lines:
if line[0] == '>':
sequences.append(lines[lines.index(line) + 1].strip())
return sequences
# Path: frame2seq/utils/pdb2input.py
def get_inference_inputs(pdb_file, chain_id):
atom_positions, aatype, seq_mask = get_parsed_inputs(pdb_file, chain_id)
seq_mask = seq_mask.unsqueeze(0)
aatype = torch.from_numpy(aatype)
aatype = aatype.unsqueeze(0)
X = atom_positions
X = X.unsqueeze(0)
return seq_mask, aatype, X
# Path: frame2seq/utils/pred2output.py
def output_csv(preds, csv_dir):
"""
Given average negative pseudo-log-likelihoods, write to a csv file.
"""
df = pd.DataFrame(columns=[
'PDBID', 'Chain ID', 'Sample Number', 'Scored sequence',
'Average negative pseudo-log-likelihood', 'Temperature'
],
data=preds)
df.to_csv(f"{csv_dir}/scores.csv", index=False)
# Path: frame2seq/utils/pred2output.py
def output_indiv_csv(scores, csv_dir):
"""
Given per-residue negative pseudo-log-likelihoods, write to a csv file.
"""
pdbid = scores['pdbid']
chain = scores['chain']
sample = scores['sample']
res_idx = scores['res_idx']
neg_pll = scores['neg_pll']
df = pd.DataFrame(
list(zip(res_idx, neg_pll)),
columns=['Residue index', 'Negative pseudo-log-likelihood'])
df.to_csv(f"{csv_dir}/{pdbid}_{chain}_seq{sample}.csv", index=False)
# Path: frame2seq/utils/score.py
import os
import torch
from tqdm import tqdm
from frame2seq.utils import residue_constants
from frame2seq.utils.util import get_neg_pll, read_fasta_file
from frame2seq.utils.pdb2input import get_inference_inputs
from frame2seq.utils.pred2output import output_csv, output_indiv_csv
def score(self, pdb_file, chain_id, fasta_file, save_indiv_neg_pll):
temperature = 1.0
seq_mask, aatype, X = get_inference_inputs(pdb_file, chain_id)
seq_mask = seq_mask.to(self.device)
aatype = aatype.to(self.device)
X = X.to(self.device)
str_form = [residue_constants.ID_TO_AA[int(i)] for i in aatype[0]]
input_aatype_onehot = residue_constants.sequence_to_onehot(
sequence=str_form,
mapping=residue_constants.AA_TO_ID,
)
input_aatype_onehot = torch.from_numpy(input_aatype_onehot).float()
input_aatype_onehot = input_aatype_onehot.unsqueeze(0)
input_aatype_onehot = input_aatype_onehot.to(self.device)
input_aatype_onehot = torch.zeros_like(input_aatype_onehot)
input_aatype_onehot[:, :,
20] = 1 # all positions are masked (set to unknown)
scores, preds = {}, []
with torch.no_grad():
pred_seq1 = self.models[0].forward(X, seq_mask, input_aatype_onehot)
pred_seq2 = self.models[1].forward(X, seq_mask, input_aatype_onehot)
pred_seq3 = self.models[2].forward(X, seq_mask, input_aatype_onehot)
pred_seq = (pred_seq1 + pred_seq2 + pred_seq3) / 3 # ensemble
pred_seq = pred_seq / temperature
pred_seq = torch.nn.functional.softmax(pred_seq, dim=-1)
pred_seq = pred_seq[seq_mask]
if fasta_file is not None:
|
self.push_screen(Main()) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: davep/oshit
# Path: oshit/app/data/config.py
@lru_cache(maxsize=None)
def load_configuration() -> Configuration:
"""Load the configuration.
Returns:
The configuration.
Note:
As a side-effect, if the configuration doesn't exist a default one
will be saved to storage.
This function is designed so that it's safe and low-cost to
repeatedly call it. The configuration is cached and will only be
loaded from storage when necessary.
"""
source = configuration_file()
return (
Configuration(**loads(source.read_text(encoding="utf-8")))
if source.exists()
else save_configuration(Configuration())
)
# Path: oshit/app/data/config.py
def save_configuration(configuration: Configuration) -> Configuration:
"""Save the given configuration.
Args:
The configuration to store.
Returns:
The configuration.
"""
load_configuration.cache_clear()
configuration_file().write_text(
dumps(asdict(configuration), indent=4), encoding="utf-8"
)
return load_configuration()
# Path: oshit/app/screens/main.py
class Main(Screen[None]):
"""The main screen of the application."""
CONTEXT_HELP = """
## Application keys
| Key | Description |
| - | - |
| <kbd>F1</kbd> | This help screen. |
| <kbd>F2</kbd> | Toggle compact/relaxed display. |
| <kbd>F3</kbd> | Toggle dark/light mode. |
| <kbd>F12</kbd> | Quit the application. |
| <kbd>t</kbd> | View the top stories. |
| <kbd>n</kbd> | View the new stories. |
| <kbd>b</kbd> | View the best stories. |
| <kbd>a</kbd> | View the AskHN stories. |
| <kbd>s</kbd> | View the ShowHN stories. |
| <kbd>j</kbd> | View the jobs. |
"""
CSS = """
TabbedContent, LoadingIndicator {
background: $panel;
}
"""
TITLE = f"Orange Site Hit v{__version__}"
BINDINGS = [
Binding("f1", "help", "Help"),
Binding("f2", "compact", "Compact/Relaxed"),
Binding("f3", "toggle_dark"),
Binding("f12", "quit", "Quit"),
Binding("t", "go('top')"),
Binding("n", "go('new')"),
Binding("b", "go('best')"),
Binding("a", "go('ask')"),
Binding("s", "go('show')"),
Binding("j", "go('jobs')"),
Binding("down, enter", "pane"),
]
def __init__(self) -> None:
"""Initialise the screen."""
super().__init__()
config = load_configuration()
self._hn = HN(
max_concurrency=config.maximum_concurrency,
timeout=config.connection_timeout,
)
"""The HackerNews client object."""
def compose(self) -> ComposeResult:
"""Compose the main screen's layout."""
yield Header()
with HackerNews():
yield Items("top", "t", self._hn.top_stories)
yield Items("new", "n", self._hn.new_stories)
yield Items("best", "b", self._hn.best_stories)
yield Items("ask", "a", self._hn.latest_ask_stories)
yield Items("show", "s", self._hn.latest_show_stories)
yield Items("jobs", "j", self._hn.latest_job_stories)
yield Footer()
def _refresh_subtitle(self) -> None:
"""Refresh the subtitle of the screen."""
self.sub_title = self.query_one(HackerNews).description
def on_mount(self) -> None:
"""Configure things once the DOM is ready."""
self.set_interval(0.95, self._refresh_subtitle)
def action_help(self) -> None:
"""Show the help screen."""
self.app.push_screen(Help(self))
def action_go(self, items: str) -> None:
"""Go to the given list of items.
Args:
items: The name of the list of items to go to.
"""
self.query_one(HackerNews).active = items
self.query_one(HackerNews).focus_active_pane()
def action_compact(self) -> None:
"""Toggle the compact display."""
news = self.query_one(HackerNews)
news.compact = not news.compact
@on(ShowUser)
def show_user(self, event: ShowUser) -> None:
"""Handle a request to show the details of a user."""
self.app.push_screen(UserDetails(self._hn, event.user))
@on(ShowComments)
def show_comments(self, event: ShowComments) -> None:
"""Handle a request to show the comments for an article."""
self.app.push_screen(Comments(self._hn, event.article))
# Path: oshit/app/oshit.py
from textual.app import App
from .data import load_configuration, save_configuration
from .screens import Main
"""The main application class."""
##############################################################################
# Textual imports.
##############################################################################
# Local imports.
##############################################################################
class OSHit(App[None]):
"""The Orange Site Hit application."""
ENABLE_COMMAND_PALETTE = False
def __init__(self) -> None:
"""Initialise the application."""
super().__init__()
self.dark = load_configuration().dark_mode
def on_mount(self) -> None:
"""Get things going once the app is up and running."""
|
self.retrieval_memory = RetrievalMemory(persistent_db_path, embedding_model_name, collection_name) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Maximilian-Winter/llama-cpp-agent
# Path: src/llama_cpp_agent/function_calling.py
class LlamaCppFunctionTool:
def __init__(self, pydantic_model: Type[BaseModel], has_markdown_code_block=False, has_triple_quoted_string=False,
**additional_parameters):
self.model = pydantic_model
self.look_for_field_string = has_markdown_code_block or has_triple_quoted_string
self.has_markdown_code_block = has_markdown_code_block
self.has_triple_quoted_string = has_triple_quoted_string
self.additional_parameters = additional_parameters if additional_parameters else {}
def __call__(self, *args, **kwargs):
return self.model(**kwargs)
# Path: src/llama_cpp_agent/agent_memory/core_memory_manager.py
class CoreMemoryManager:
def __init__(self, core_memory: dict):
self.core_memory = core_memory
def add_to_core_memory(self, key: str, child_key: str, value) -> str:
"""
Adds or updates an entry in the core memory.
"""
if key not in self.core_memory:
self.core_memory[key] = {}
self.core_memory[key][child_key] = value
return f"Core memory updated. Key: {key}, Child Key: {child_key}"
def replace_in_core_memory(self, key: str, child_key: str, new_value) -> str:
"""
Replaces an existing entry in the core memory.
"""
if key in self.core_memory and child_key in self.core_memory[key]:
self.core_memory[key][child_key] = new_value
return f"Core memory replaced. Key: {key}, Child Key: {child_key}"
else:
return "Key or child key not found in core memory."
def remove_from_core_memory(self, key: str, child_key: str) -> str:
"""
Removes a specific field from a core memory entry.
"""
if key in self.core_memory and child_key in self.core_memory[key]:
del self.core_memory[key][child_key]
return f"Core memory entry removed. Key: {key}, Child Key: {child_key}"
else:
return "Key or child key not found in core memory."
def build_core_memory_context(self):
output = json.dumps(self.core_memory, indent=4)
context = f"# Core-Memory:\n{output if output != '{}' else 'Empty'}"
return context
def load(self, file_path):
with open(file_path, 'r', encoding='utf-8') as file:
self.core_memory = json.load(file)
def save(self, file_path):
with open(file_path, 'w', encoding='utf-8') as file:
json.dump(self.core_memory, file, indent=4)
# Path: src/llama_cpp_agent/agent_memory/retrieval_memory_manager.py
class RetrievalMemoryManager:
def __init__(self, retrieval_memory: RetrievalMemory):
def add_memory_to_retrieval(self, description: str, importance: float = 1.0) -> str:
def retrieve_memories(self, query: str, max_results: int = 5) -> str:
# Path: src/llama_cpp_agent/agent_memory/memory_tools.py
from pydantic import BaseModel, Field
from ..function_calling import LlamaCppFunctionTool
from .core_memory_manager import CoreMemoryManager
from .retrieval_memory_manager import RetrievalMemoryManager, RetrievalMemory
class AddCoreMemory(BaseModel):
"""
Add a new entry to the core memory.
"""
key: str = Field(..., description="The key identifier for the core memory entry.")
field: str = Field(..., description="A secondary key or field within the core memory entry.")
value: str = Field(..., description="The value or data to be stored in the specified core memory entry.")
def run(self, core_memory_manager: CoreMemoryManager):
return core_memory_manager.add_to_core_memory(self.key, self.field, self.value)
# Replace Core Memory Model
class ReplaceCoreMemory(BaseModel):
"""
Replace an entry in the core memory.
"""
key: str = Field(..., description="The key identifier for the core memory entry.")
field: str = Field(..., description="The specific field within the core memory entry to be replaced.")
new_value: str = Field(...,
description="The new value to replace the existing data in the specified core memory field.")
def run(self, core_memory_manager: CoreMemoryManager):
return core_memory_manager.replace_in_core_memory(self.key, self.field, self.value)
class RemoveCoreMemory(BaseModel):
"""
Remove an entry in the core memory.
"""
key: str = Field(..., description="The key identifier for the core memory entry to be removed.")
field: str = Field(..., description="The specific field within the core memory entry to be removed.")
def run(self, core_memory_manager: CoreMemoryManager):
return core_memory_manager.remove_from_core_memory(self.key, self.field)
class RetrieveMemories(BaseModel):
"""
Retrieve memories from the retrieval memory based on a query.
"""
query: str = Field(..., description="The query to be used to retrieve memories from the retrieval memory.")
def run(self, retrieval_memory_manager: RetrievalMemoryManager):
return retrieval_memory_manager.retrieve_memories(self.query)
class AddRetrievalMemory(BaseModel):
"""
Add memory to the retrieval memory.
"""
memory: str = Field(..., description="The memory to be added to the retrieval memory.")
importance: float = Field(..., description="The importance of the memory to be added to the retrieval memory.")
def run(self, retrieval_memory_manager: RetrievalMemoryManager):
return retrieval_memory_manager.add_memory_to_retrieval(self.memory, self.importance)
class AgentRetrievalMemory:
def __init__(self, persistent_db_path="./retrieval_memory", embedding_model_name="all-MiniLM-L6-v2",
collection_name="retrieval_memory_collection"):
|
"dot": Dot, | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: tedivm/paracelsus
# Path: paracelsus/transformers/dot.py
class Dot:
comment_format: str = "dot"
metadata: MetaData
graph: pydot.Dot
def __init__(self, metaclass: MetaData) -> None:
self.metadata = metaclass
self.graph = pydot.Dot("database", graph_type="graph")
for table in self.metadata.tables.values():
node = pydot.Node(name=table.name)
node.set_label(self._table_label(table))
node.set_shape("none")
node.set_margin("0")
self.graph.add_node(node)
for column in table.columns:
for foreign_key in column.foreign_keys:
key_parts = foreign_key.target_fullname.split(".")
left_table = key_parts[0]
left_column = key_parts[1]
edge = pydot.Edge(left_table, table.name)
edge.set_label(column.name)
edge.set_dir("both")
edge.set_arrowhead("none")
if not column.unique:
edge.set_arrowhead("crow")
l_column = self.metadata.tables[left_table].columns[left_column]
edge.set_arrowtail("none")
if not l_column.unique and not l_column.primary_key:
edge.set_arrowtail("crow")
self.graph.add_edge(edge)
def _table_label(self, table: Table) -> str:
column_output = ""
columns = sorted(table.columns, key=utils.column_sort_key)
for column in columns:
attributes = set([])
if column.primary_key:
attributes.add("Primary Key")
if len(column.foreign_keys) > 0:
attributes.add("Foreign Key")
if column.unique:
attributes.add("Unique")
column_output += f' <tr><td align="left">{column.type}</td><td align="left">{column.name}</td><td>{", ".join(sorted(attributes))}</td></tr>\n'
return f"""<
<table border="0" cellborder="1" cellspacing="0" cellpadding="4">
<tr><td colspan="3" bgcolor="lightblue"><b>{table.name}</b></td></tr>
{column_output.rstrip()}
</table>
>"""
def __str__(self) -> str:
return self.graph.to_string()
# Path: paracelsus/transformers/mermaid.py
class Mermaid:
comment_format: str = "mermaid"
metadata: MetaData
def __init__(self, metaclass: MetaData) -> None:
self.metadata = metaclass
def _table(self, table: Table) -> str:
output = f"\t{table.name}"
output += " {\n"
columns = sorted(table.columns, key=utils.column_sort_key)
for column in columns:
output += self._column(column)
output += "\t}\n\n"
return output
def _column(self, column: Column) -> str:
column_str = f"{column.type} {column.name}"
if column.primary_key:
if len(column.foreign_keys) > 0:
column_str += " PK,FK"
else:
column_str += " PK"
elif len(column.foreign_keys) > 0:
column_str += " FK"
options = []
if column.nullable:
options.append("nullable")
if column.unique:
options.append("unique")
if column.index:
options.append("indexed")
if len(options) > 0:
column_str += f' "{",".join(options)}"'
return f"\t\t{column_str}\n"
def _relationships(self, column: Column) -> str:
output = ""
column_name = column.name
right_table = column.table.name
if column.unique:
right_operand = "o|"
else:
right_operand = "o{"
for foreign_key in column.foreign_keys:
key_parts = foreign_key.target_fullname.split(".")
left_table = key_parts[0]
left_column = key_parts[1]
left_operand = ""
lcolumn = self.metadata.tables[left_table].columns[left_column]
if lcolumn.unique or lcolumn.primary_key:
left_operand = "||"
else:
left_operand = "}o"
output += f"\t{left_table} {left_operand}--{right_operand} {right_table} : {column_name}\n"
return output
def __str__(self) -> str:
output = "erDiagram\n"
for table in self.metadata.tables.values():
output += self._table(table)
for table in self.metadata.tables.values():
for column in table.columns.values():
if len(column.foreign_keys) > 0:
output += self._relationships(column)
return output
# Path: paracelsus/cli.py
import importlib
import re
import sys
import typer
from enum import Enum
from pathlib import Path
from typing import List
from typing_extensions import Annotated
from .transformers.dot import Dot
from .transformers.mermaid import Mermaid
from . import _version
app = typer.Typer()
transformers = {
"mmd": Mermaid,
"mermaid": Mermaid,
|
send_message(update.from_id, "😫 You are not allowed to use this bot.") | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: winniesi/tg-gemini-bot
# Path: api/auth.py
def is_authorized(from_id: int, user_name: str) -> bool:
if str(user_name) in ALLOWED_USERS:
return True
return False
# Path: api/context.py
class ChatManager:
"""setting up a basic conversation storage manager"""
def __init__(self):
self.chats: Dict[str, ChatConversation] = {}
def _new_chat(self, username: str) -> ChatConversation:
chat = ChatConversation()
self.chats[username] = chat
return chat
def get_chat(self, username: str) -> ChatConversation:
if self.chats.get(username) is None:
return self._new_chat(username)
return self.chats[username]
# Path: api/context.py
class ImageChatManger:
def __init__(self, prompt, file_id: str) -> None:
self.prompt = prompt
self.file_id = file_id
def tel_photo_url(self) -> str:
"""process telegram photo url"""
r_file_id = requests.get(
f"https://api.telegram.org/bot{BOT_TOKEN}/getFile?file_id={self.file_id}"
)
file_path = r_file_id.json().get("result").get("file_path")
download_url = f"https://api.telegram.org/file/bot{BOT_TOKEN}/{file_path}"
return download_url
def photo_bytes(self) -> BytesIO:
"""get photo bytes"""
photo_url = self.tel_photo_url()
response = requests.get(photo_url)
photo_bytes = BytesIO(response.content)
return photo_bytes
def send_image(self) -> str:
response = generate_text_with_image(self.prompt, self.photo_bytes())
return response
# Path: api/telegram.py
class Update:
def __init__(self, update: Dict) -> None:
self.update = update
self.from_id = update["message"]["from"]["id"]
self.type = self._type()
self.text = self._text()
self.photo_caption = self._photo_caption()
self.file_id = self._file_id()
self.user_name = update["message"]["from"]["username"]
def _type(self):
if "text" in self.update["message"]:
return "text"
elif "photo" in self.update["message"]:
return "photo"
else:
return ""
def _photo_caption(self):
if self.type == "photo":
return self.update["message"].get("caption", "describe the photo")
return ""
def _text(self):
if self.type == "text":
return self.update["message"]["text"]
return ""
def _file_id(self):
if self.type == "photo":
return self.update["message"]["photo"][0]["file_id"]
return ""
# Path: api/telegram.py
def send_message(chat_id, text):
"""send text message"""
payload = {
"chat_id": chat_id,
"text": escape(text),
"parse_mode": "MarkdownV2",
}
r = requests.post(f"{TELEGRAM_API}/sendMessage", data=payload)
print(f"Sent message: {text} to {chat_id}")
return r
# Path: api/handle.py
from .auth import is_authorized
from .context import ChatManager, ImageChatManger
from .telegram import Update, send_message
"""
All the chat that comes through the Telegram bot gets passed to the
handle_message function. This function checks out if the user has the
green light to chat with the bot. Once that's sorted, it figures out if
the user sent words or an image and deals with it accordingly.
For text messages, it fires up the ChatManager class that keeps track of
the back-and-forth with that user.
As for images, in Gemini pro, they're context-free, so you can handle
them pretty straight-up without much fuss.
"""
chat_manager = ChatManager()
def handle_message(update_data):
update = Update(update_data)
authorized = is_authorized(update.from_id, update.user_name)
if not authorized:
|
raise error.flowFileException('Flow file does not exist.') | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: usail-hkust/LLMTSCS
# Path: utils/utils.py
def oneline_wrapper(dic_agent_conf, dic_traffic_env_conf, dic_path, roadnet, trafficflow):
results_table = []
all_rewards = []
all_queue_len = []
all_travel_time = []
for i in range(1):
dic_path["PATH_TO_MODEL"] = (dic_path["PATH_TO_MODEL"].split(".")[0] + ".json" +
time.strftime('%m_%d_%H_%M_%S', time.localtime(time.time())))
dic_path["PATH_TO_WORK_DIRECTORY"] = (dic_path["PATH_TO_WORK_DIRECTORY"].split(".")[0] + ".json" +
time.strftime('%m_%d_%H_%M_%S', time.localtime(time.time())))
oneline = OneLine(dic_agent_conf=dic_agent_conf,
dic_traffic_env_conf=merge(config.dic_traffic_env_conf, dic_traffic_env_conf),
dic_path=merge(config.DIC_PATH, dic_path),
roadnet=roadnet,
trafficflow=trafficflow
)
round_results = oneline.train(round=i)
results_table.append([round_results['test_reward_over'], round_results['test_avg_queue_len_over'],
round_results['test_avg_travel_time_over']])
all_rewards.append(round_results['test_reward_over'])
all_queue_len.append(round_results['test_avg_queue_len_over'])
all_travel_time.append(round_results['test_avg_travel_time_over'])
# delete junk
cmd_delete_model = 'rm -rf <dir>'.replace("<dir>", dic_path["PATH_TO_MODEL"])
cmd_delete_work = 'find <dir> -type f ! -name "state_action.json" -exec rm -rf {} \;'.replace("<dir>", dic_path["PATH_TO_WORK_DIRECTORY"])
os.system(cmd_delete_model)
os.system(cmd_delete_work)
results_table.append([np.average(all_rewards), np.average(all_queue_len), np.average(all_travel_time)])
results_table.append([np.std(all_rewards), np.std(all_queue_len), np.std(all_travel_time)])
table_logger = wandb.init(
project=dic_traffic_env_conf['PROJECT_NAME'],
group=f"{dic_traffic_env_conf['MODEL_NAME']}-{roadnet}-{trafficflow}-{len(dic_agent_conf['FIXED_TIME'])}_Phases",
name="exp_results",
config=merge(merge(dic_agent_conf, dic_path), dic_traffic_env_conf),
)
columns = ["reward", "avg_queue_len", "avg_travel_time"]
logger_table = wandb.Table(columns=columns, data=results_table)
table_logger.log({"results": logger_table})
wandb.finish()
return
# Path: utils/error.py
class flowFileException(Exception):
def __init__(self, message):
def __str__(self):
# Path: run_advanced_maxpressure.py
from utils.utils import oneline_wrapper
from utils import error
from multiprocessing import Process
import os
import time
import argparse
def parse_args():
parser = argparse.ArgumentParser()
parser.add_argument("--memo", type=str, default='AdvancedMaxPressure')
parser.add_argument("--model", type=str, default="AdvancedMaxPressure")
parser.add_argument("--proj_name", type=str, default="chatgpt-TSCS")
parser.add_argument("--eightphase", action="store_true", default=False)
parser.add_argument("--multi_process", action="store_true", default=True)
parser.add_argument("--workers", type=int, default=1)
parser.add_argument("--dataset", type=str, default="template")
parser.add_argument("--traffic_file", type=str, default="flow_main_stream.json")
return parser.parse_args()
def main(in_args):
traffic_file_list = []
if in_args.dataset == 'jinan':
count = 3600
road_net = "3_4"
traffic_file_list = ["anon_3_4_jinan_real.json", "anon_3_4_jinan_real_2000.json", "anon_3_4_jinan_real_2500.json"]
template = "Jinan"
elif in_args.dataset == 'hangzhou':
count = 3600
road_net = "4_4"
traffic_file_list = ["anon_4_4_hangzhou_real.json", "anon_4_4_hangzhou_real_5816.json"]
template = "Hangzhou"
elif in_args.dataset == 'newyork_16x3':
count = 3600
road_net = "16_3"
traffic_file_list = ["anon_16_3_newyork_real.json"]
template = "NewYork"
elif in_args.dataset == 'newyork_28x7':
count = 3600
road_net = "28_7"
traffic_file_list = ["anon_28_7_newyork_real_double.json", "anon_28_7_newyork_real_triple.json"]
template = "NewYork"
elif in_args.dataset == 'template':
count = 3600
road_net = "1_1"
traffic_file_list = ["flow_main_stream.json"]
template = "template"
# flow_file error
try:
if in_args.traffic_file not in traffic_file_list:
|
def validate(self, data: ndarray) -> Optional[InferredField]: | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: ohadmata/shmessy
# Path: src/shmessy/schema.py
class InferredField(BaseModel):
inferred_type: Optional[str] = None
inferred_pattern: Optional[Any] = None
# Path: src/shmessy/schema.py
class ValidatorTypes(str, Enum):
NUMERIC = "NUMERIC"
STRING = "STRING"
# Path: src/shmessy/types/base.py
class BaseType(ABC):
weight: int = 0
validator_types: Tuple[ValidatorTypes]
@abstractmethod
def validate(self, data: ndarray) -> Optional[InferredField]:
pass
@abstractmethod
def fix(self, column: Series, inferred_field: InferredField) -> Series:
pass
def is_validator_type_valid(self, dtype: Type) -> bool:
for possible_validator_type in self.validator_types:
if self._check_single_validator_type(dtype, possible_validator_type):
return True
return False
@staticmethod
def _check_single_validator_type(
dtype: Type, possible_validator_type: ValidatorTypes
) -> bool:
if possible_validator_type == ValidatorTypes.NUMERIC and not issubdtype(
dtype, number
):
return False
if possible_validator_type == ValidatorTypes.STRING and not (
issubdtype(dtype, object_) or issubdtype(dtype, str_)
):
return False
return True
@property
def name(self) -> str:
return str(self.__class__.__name__.replace("Type", ""))
# Path: src/shmessy/types/unix_timestamp.py
import logging
import math
from datetime import datetime
from enum import Enum
from typing import Optional
from numpy import ndarray
from pandas import Series, to_datetime
from ..schema import InferredField, ValidatorTypes
from .base import BaseType
logger = logging.getLogger(__name__)
class TimestampResolution(str, Enum):
SECONDS = "s"
MILLISECONDS = "ms"
NANOSECONDS = "ns"
class UnixTimestampType(BaseType):
weight = 4
validator_types = (ValidatorTypes.NUMERIC,)
min_valid_year: int = 1980
max_valid_year: int = 2100
@staticmethod
def _unix_timestamp_resolution(value: float) -> TimestampResolution:
number_of_digits = len(str(int(value)))
if number_of_digits == 10:
return TimestampResolution.SECONDS
if number_of_digits == 13:
return TimestampResolution.MILLISECONDS
if number_of_digits == 16:
return TimestampResolution.NANOSECONDS
@staticmethod
def _fix_input_resolution(
value: float, selected_resolution: TimestampResolution
) -> float:
if selected_resolution == TimestampResolution.SECONDS:
return value
if selected_resolution == TimestampResolution.MILLISECONDS:
return value / 1000
if selected_resolution == TimestampResolution.NANOSECONDS:
return value / 1000 / 1000
|
bought_token_curr_price = get_price(desired_token_address)
| You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: kokiez/solana-sniper
# Path: birdeye.py
def get_price(token_address):
url = f"https://api.dexscreener.com/latest/dex/tokens/{token_address}"
exclude = ['EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v', 'Es9vMFrzaCERmJfrF4H2FYD4KCoNkY11McCe8BenwNYB']
response = requests.get(url).json()
if token_address not in exclude:
for pair in response['pairs']:
if pair['quoteToken']['address'] == 'So11111111111111111111111111111111111111112':
return float(pair['priceUsd'])
else:
return response['pairs'][0]['priceUsd']
return None
# Path: birdeye.py
def getSymbol(token):
# usdc and usdt
exclude = ['EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v', 'Es9vMFrzaCERmJfrF4H2FYD4KCoNkY11McCe8BenwNYB']
if token not in exclude:
url = f"https://api.dexscreener.com/latest/dex/tokens/{token}"
Token_Symbol = ""
Sol_symbol=""
try:
response = requests.get(url)
# Check if the request was successful (status code 200)
if response.status_code == 200:
resp = response.json()
print("Response:",resp['pairs'][0]['baseToken']['symbol'])
for pair in resp['pairs']:
quoteToken = pair['quoteToken']['symbol']
if quoteToken == 'SOL':
Token_Symbol = pair['baseToken']['symbol']
Sol_symbol = quoteToken
return Token_Symbol, Sol_symbol
else:
print(f"[getSymbol] Request failed with status code {response.status_code}")
except requests.exceptions.RequestException as e:
print(f"[getSymbol] error occurred: {e}")
except:
a = 1
return Token_Symbol, Sol_symbol
else:
if token == 'EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v':
return "USDC", "SOL"
elif token == 'EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v':
return "USDT", "SOL"
# Path: webhook.py
def sendWebhook(title_type_info, description):
global error_webhook
global webhook_url
title = ""
title_type = title_type_info.split("|")
if title_type[0] == "msg":
title = title_type[1]
color = colors["Green"]
webhook(title, color, description, webhook_url)
elif title_type[0] == "msg_b":
title = title_type[1]
color = colors["DarkAqua"]
webhook(title, color, description, webhook_url)
elif title_type[0] == "msg_s":
title = title_type[1]
color = colors["DarkAqua"]
webhook(title, color, description, webhook_url)
elif title_type[0] == "i_s": #invest or slippage was changed etc
title = title_type[1]
color = colors["DarkPurple"]
webhook(title, color, description, webhook_url)
elif title_type[0] == "e": #error
title = title_type[1]
color = colors["DarkRed"]
webhook(title, color, description, error_webhook)
elif title_type[0] == "a": #alert
title = title_type[1]
color = colors["LuminousVividPink"]
webhook(title, color, description, webhook_url)
elif title_type[0] == "w": #wallet info
title = title_type[1]
color = colors["Gold"]
webhook(title, color, description, webhook_url)
# Path: monitor_price_strategy.py
import time
from birdeye import get_price, getSymbol
from webhook import sendWebhook
"""If you have ton of trades then best to use Simulate Transaction and modify this part of code to your needs"""
"""
Only Take Profit
"""
def limit_order(bought_token_price,desired_token_address, take_profit_ratio, execution_time, txB):
token_symbol, SOl_Symbol = getSymbol(desired_token_address)
# CALCULATE SELL LIMIT
sell_limit_token_price = bought_token_price * take_profit_ratio
print("-" * 79)
print(f"| {'Bought Price':<12} | {'Sell Limit':<12} | {'Tx Buy':<50} |")
print("-" * 79)
print(f"|{bought_token_price:.12f} | {sell_limit_token_price:.12f} {txB:<50} |")
print("-" * 79)
sendWebhook(f"msg_b|BUY INFO {token_symbol}",f"Bought Price: {bought_token_price:.12f}\n**Sell Limit: {sell_limit_token_price:.15f}**\nTotal Buy Execution time: {execution_time} seconds\nBuy TXN: https://solscan.io/tx/{txB} |")
# LOOP = CHECK IF PRICE >= SELL LIMIT | checks price every 5 seconds
priceLow = True
# while priceLow and isTimePassed(time_limit) == False:
while priceLow:
# Check if time limit has been passed for the token bought or not
|
self.LayerNorm = LayerNormTorchAlike(config.hidden_size, eps=config.layer_norm_eps, correction=True) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: enochyearn/MLX_RoBERTa
# Path: custom/nn/layers/normalization.py
class LayerNormBasselCorrected(Module):
r"""Applies layer normalization [1] on the inputs with Bessel's Correction used by default like PyTorch.
Computes
.. math::
y = \frac{x - E[x]}{\sqrt{Var[x]} + \epsilon} \gamma + \beta,
where :math:`\gamma` and :math:`\beta` are learned per feature dimension
parameters initialized at 1 and 0 respectively.
Var[x] would by default apply Bessel's Correction.
[1]: https://arxiv.org/abs/1607.06450
Args:
dims (int): The feature dimension of the input to normalize over
eps (float): A small additive constant for numerical stability
affine (bool): If True learn an affine transform to apply after the
normalization
correction (bool):
"""
def __init__(self, dims: int, eps: float = 1e-5, affine: bool = True, correction: bool = True):
super().__init__()
if affine:
self.bias = mx.zeros((dims,))
self.weight = mx.ones((dims,))
self.eps = eps
self.dims = dims
self.correction = correction
def _extra_repr(self):
return f"{self.dims}, eps={self.eps}, affine={'weight' in self}"
def __call__(self, x):
means = mx.mean(x, axis=-1, keepdims=True)
var = mx.var(x, axis=-1, keepdims=True, ddof=int(self.correction))
x = (x - means) * mx.rsqrt(var + self.eps)
return (self.weight * x + self.bias) if "weight" in self else x
# Path: custom/nn/layers/normalization.py
class LayerNormTorchAlike(Module):
r"""Applies layer normalization [1] on the inputs in PyTorch's style.
MLX's official LayerNorm has a different behavior with PyTorch's.
Computes
.. math::
y = \frac{x - E[x]}{\sqrt{Var[x]} + \epsilon} \gamma + \beta,
where :math:`\gamma` and :math:`\beta` are learned per feature dimension
parameters initialized at 1 and 0 respectively.
Var[x] would by default apply Bessel's Correction.
[1]: https://arxiv.org/abs/1607.06450
Args:
dims (int): The feature dimension of the input to normalize over
eps (float): A small additive constant for numerical stability
affine (bool): If True learn an affine transform to apply after the
normalization
correction (bool):
"""
def __init__(self, dims: int, eps: float = 1e-5, affine: bool = True, correction: bool = True):
super().__init__()
if affine:
self.bias = mx.zeros((dims,))
self.weight = mx.ones((dims,))
self.eps = eps
self.dims = dims
self.correction = correction
def _extra_repr(self):
return f"{self.dims}, eps={self.eps}, affine={'weight' in self}"
def __call__(self, x):
# Calculate the mean of all elements;
# i.e. the means for each element $\mathbb{E}[X]$
mean = x.mean(axis=-1, keepdims=True)
# Calculate the squared mean of all elements;
# i.e. the means for each element $\mathbb{E}[X^2]$
mean_x2 = (x ** 2).mean(axis=-1, keepdims=True)
# Variance of all element $Var[X] = \mathbb{E}[X^2] - \mathbb{E}[X]^2$
var = mean_x2 - mean ** 2
# Normalize $$\hat{X} = \frac{X - \mathbb{E}[X]}{\sqrt{Var[X] + \epsilon}}$$
x_norm = (x - mean) / mx.sqrt(var + self.eps)
# Scale and shift $$\text{LN}(x) = \gamma \hat{X} + \beta$$
x_norm = self.weight * x_norm + self.bias
return x_norm
# Path: mlx_roberta.py
import argparse
import time
import mlx.core as mx
import mlx.nn as nn
import numpy as np
import math
from mlx.utils import tree_unflatten
from collections import OrderedDict
from custom.nn.layers.normalization import LayerNormBasselCorrected, LayerNormTorchAlike
from transformers import RobertaTokenizer
from dataclasses import dataclass
# utils
@dataclass
class ModelConfig:
intermediate_size: int = 3072
hidden_size: int = 768
no_heads: int = 12
hidden_layers: int = 12
vocab_size: int = 50265
attention_probs_dropout_prob: float = 0.1
hidden_dropout_prob: float = 0.1
layer_norm_eps: float = 1e-5
max_position_embeddings: int = 514
# QA model's parameters
num_labels: int = 2
type_vocab_size: int = 2
pad_token_id: int = 1
chunk_size_feed_forward: int = 0
model_configs = {
"deepset/roberta-base-squad2": ModelConfig(),
"roberta-base": ModelConfig(),
}
model_types = {
"deepset/roberta-base-squad2": "qa",
"roberta-base": "base",
}
class RobertaEmbeddings(nn.Module):
def __init__(self, config):
super().__init__()
self.word_embeddings = nn.Embedding(config.vocab_size, config.hidden_size)
self.position_embeddings = nn.Embedding(config.max_position_embeddings, config.hidden_size)
self.token_type_embeddings = nn.Embedding(config.type_vocab_size, config.hidden_size)
|
@app.get("/tables", response_model=RList[Table]) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: zy7y/dfs-generate
# Path: entity.py
class CodeGen(BaseVo):
name: str
code: str
@field_serializer("code")
def serialize_code(self, code: str, _info):
_code = black.format_str(code, mode=black.FileMode())
return isort.code(_code)
# Path: entity.py
class Conf(SQLModel, table=True):
__tablename__ = "dfs_conf"
id: int = Field(None, primary_key=True)
db_uri: str = Field(..., description="数据库连接")
@classmethod
def get_db_uri_last_new(cls):
"""获取最新的db_url"""
with Session(engine) as session:
query = select(cls).order_by(cls.id.desc())
latest_conf = session.exec(query).first()
if latest_conf:
return latest_conf.db_uri
else:
return None
@classmethod
def create(cls, uri) -> "Conf":
with Session(engine) as session:
obj = cls(db_uri=uri)
session.add(obj)
session.commit()
session.refresh(obj)
return obj
@classmethod
def get_last_uri_with_metadata(cls):
uri = cls.get_db_uri_last_new()
return uri, get_metadata_by_db_uri(uri)
# Path: entity.py
class DBConf(SQLModel):
user: str
password: str
port: int
host: str
db: str
def get_db_uri(self):
return f"mysql+pymysql://{self.user}:{self.password}@{self.host}:{self.port}/{self.db}"
def get_metadata(self):
return get_metadata_by_db_uri(self.get_db_uri())
# Path: entity.py
class R(BaseModel, Generic[T]):
code: int = 20000
msg: str = "ok"
data: Optional[T] = None
@classmethod
def success(cls, **kwargs):
return cls(**kwargs)
@classmethod
def error(cls, msg):
return cls(code=40000, msg=msg)
# Path: entity.py
class RList(R[T]):
data: List[T] = Field(default_factory=list)
# Path: entity.py
class Table(BaseVo):
table_name: str
table_comment: Optional[str] = None
# Path: generate/main.py
def generate_code(table: Table, uri: str):
return [
{"name": "model.py", "code": GenerateEntity(table).render()},
{"name": "router.py", "code": render_router(table.name)},
{"name": "main.py", "code": render_main(table.name)},
{"name": "db.py", "code": render_db(uri)},
]
# Path: main.py
from fastapi import FastAPI, Query
from fastapi.requests import Request
from fastapi.responses import FileResponse
from fastapi.staticfiles import StaticFiles
from entity import CodeGen, Conf, DBConf, R, RList, Table
from generate.main import generate_code
import uvicorn
app = FastAPI(
title="dfs-generate", description="FastAPI SQLModel 逆向生成代码", docs_url=None
)
app.mount("/static", StaticFiles(directory="static"), name="static")
@app.get("/", include_in_schema=False)
def index():
return FileResponse("static/index.html")
|
v_search = VectorSearchEngine(item_embedding) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: CrawlScript/Torch-MGDCF
# Path: torch_mgdcf/metrics/ranking.py
def ndcg_score(reference, hypothesis):
"""
Normalized Discounted Cumulative Gain (nDCG)
Normalized version of DCG:
nDCG = DCG(hypothesis)/DCG(reference)
Parameters:
reference - a gold standard (perfect) ordering Ex: [5,4,3,2,1]
hypothesis - a proposed ordering Ex: [5,2,2,3,1]
Returns:
ndcg_score - normalized score
"""
return dcg_score(hypothesis)/dcg_score(reference)
# Path: torch_mgdcf/metrics/ranking.py
def precision_score(reference, hypothesis):
result = np.sum(hypothesis, dtype=np.float32)/len(hypothesis)
return result
# Path: torch_mgdcf/metrics/ranking.py
def recall_score(reference, hypothesis):
result = np.sum(hypothesis, dtype=np.float32) / len(reference)
return result
# Path: torch_mgdcf/vector_search/vector_search.py
class VectorSearchEngine(object):
def __init__(self, vectors):
super().__init__()
if isinstance(vectors, torch.Tensor):
self.vectors = vectors.detach().cpu().numpy()
else:
self.vectors = np.array(vectors)
self.dim = self.vectors.shape[1]
self.index = faiss.IndexFlatIP(self.dim)
self.index.add(self.vectors)
def search(self, query_vectors, k=10):
query_vectors = np.asarray(query_vectors)
topK_distances, topK_indices = self.index.search(query_vectors, k)
return topK_distances, topK_indices
# Path: torch_mgdcf/evaluation/ranking.py
from tqdm import tqdm
from torch_mgdcf.metrics.ranking import ndcg_score, precision_score, recall_score
from torch_mgdcf.vector_search.vector_search import VectorSearchEngine
import numpy as np
import torch
# coding=utf-8
# The code is from our another project GRecX: https://github.com/maenzhier/grecx_datasets
def score(ground_truth, pred_items, k_list, metrics):
pred_match = [1 if item in ground_truth else 0 for item in pred_items]
max_k = k_list[-1]
if len(ground_truth) > max_k:
ndcg_gold = [1] * max_k
else:
ndcg_gold = [1] * len(ground_truth) + [0] * (max_k - len(ground_truth))
res_score = []
for metric in metrics:
if metric == "ndcg":
score_func = ndcg_score
elif metric == "precision":
score_func = precision_score
elif metric == "recall":
score_func = recall_score
else:
raise Exception("Not Found Metric : {}".format(metric))
for k in k_list:
if metric == "ndcg":
res_score.append(score_func(ndcg_gold[:k], pred_match[:k]))
else:
res_score.append(score_func(ground_truth, pred_match[:k]))
return res_score
def evaluate_mean_global_metrics(user_items_dict, user_mask_items_dict,
user_embedding, item_embedding,
k_list=[10, 20], metrics=["ndcg"]):
|
@MODELS.register_module() | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: KyanChen/TTP
# Path: mmseg/utils/typing_utils.py
# Path: opencd/registry.py
MODELS = Registry('model', parent=MMENGINE_MODELS, locations=['opencd.models'])
# Path: opencd/models/data_preprocessor.py
from numbers import Number
from typing import Any, Dict, List, Optional, Sequence, Union
from mmengine.model import BaseDataPreprocessor
from mmseg.utils import SampleList
from opencd.registry import MODELS
import numpy as np
import torch
import torch.nn.functional as F
# Copyright (c) Open-CD. All rights reserved.
def stack_batch(inputs: List[torch.Tensor],
data_samples: Optional[SampleList] = None,
size: Optional[tuple] = None,
size_divisor: Optional[int] = None,
pad_val: Union[int, float] = 0,
seg_pad_val: Union[int, float] = 255) -> torch.Tensor:
"""Stack multiple inputs to form a batch and pad the images and gt_sem_segs
to the max shape use the right bottom padding mode.
Args:
inputs (List[Tensor]): The input multiple tensors. each is a
CHW 3D-tensor.
data_samples (list[:obj:`SegDataSample`]): The list of data samples.
It usually includes information such as `gt_sem_seg`.
size (tuple, optional): Fixed padding size.
size_divisor (int, optional): The divisor of padded size.
pad_val (int, float): The padding value. Defaults to 0
seg_pad_val (int, float): The padding value. Defaults to 255
Returns:
Tensor: The 4D-tensor.
List[:obj:`SegDataSample`]: After the padding of the gt_seg_map.
"""
assert isinstance(inputs, list), \
f'Expected input type to be list, but got {type(inputs)}'
assert len({tensor.ndim for tensor in inputs}) == 1, \
f'Expected the dimensions of all inputs must be the same, ' \
f'but got {[tensor.ndim for tensor in inputs]}'
assert inputs[0].ndim == 3, f'Expected tensor dimension to be 3, ' \
f'but got {inputs[0].ndim}'
assert len({tensor.shape[0] for tensor in inputs}) == 1, \
f'Expected the channels of all inputs must be the same, ' \
f'but got {[tensor.shape[0] for tensor in inputs]}'
# only one of size and size_divisor should be valid
assert (size is not None) ^ (size_divisor is not None), \
'only one of size and size_divisor should be valid'
padded_inputs = []
padded_samples = []
inputs_sizes = [(img.shape[-2], img.shape[-1]) for img in inputs]
max_size = np.stack(inputs_sizes).max(0)
if size_divisor is not None and size_divisor > 1:
# the last two dims are H,W, both subject to divisibility requirement
max_size = (max_size +
(size_divisor - 1)) // size_divisor * size_divisor
for i in range(len(inputs)):
tensor = inputs[i]
if size is not None:
width = max(size[-1] - tensor.shape[-1], 0)
height = max(size[-2] - tensor.shape[-2], 0)
# (padding_left, padding_right, padding_top, padding_bottom)
padding_size = (0, width, 0, height)
elif size_divisor is not None:
width = max(max_size[-1] - tensor.shape[-1], 0)
height = max(max_size[-2] - tensor.shape[-2], 0)
padding_size = (0, width, 0, height)
else:
padding_size = [0, 0, 0, 0]
# pad img
pad_img = F.pad(tensor, padding_size, value=pad_val)
padded_inputs.append(pad_img)
# pad gt_sem_seg
if data_samples is not None:
data_sample = data_samples[i]
gt_sem_seg = data_sample.gt_sem_seg.data
del data_sample.gt_sem_seg.data
data_sample.gt_sem_seg.data = F.pad(
gt_sem_seg, padding_size, value=seg_pad_val)
if 'gt_edge_map' in data_sample:
gt_edge_map = data_sample.gt_edge_map.data
del data_sample.gt_edge_map.data
data_sample.gt_edge_map.data = F.pad(
gt_edge_map, padding_size, value=seg_pad_val)
if 'gt_seg_map_from' in data_sample:
gt_seg_map_from = data_sample.gt_seg_map_from.data
del data_sample.gt_seg_map_from.data
data_sample.gt_seg_map_from.data = F.pad(
gt_seg_map_from, padding_size, value=seg_pad_val)
if 'gt_seg_map_to' in data_sample:
gt_seg_map_to = data_sample.gt_seg_map_to.data
del data_sample.gt_seg_map_to.data
data_sample.gt_seg_map_to.data = F.pad(
gt_seg_map_to, padding_size, value=seg_pad_val)
data_sample.set_metainfo({
'img_shape': tensor.shape[-2:],
'pad_shape': data_sample.gt_sem_seg.shape,
'padding_size': padding_size
})
padded_samples.append(data_sample)
else:
padded_samples.append(
dict(
img_padding_size=padding_size,
pad_shape=pad_img.shape[-2:]))
return torch.stack(padded_inputs, dim=0), padded_samples
|
await free(str(phone_number).replace("+", ""))
| You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: N0rz3/Phunter
# Path: lib/free_lookup.py
async def free(phone_number):
r = await Request("https://free-lookup.net/{}".format(phone_number), headers={'user-agent': random.choice(agent)}).get()
html_body = BeautifulSoup(r.text, "html.parser")
list_info = html_body.findChild("ul", class_="report-summary__list").findAll("div")
info_dict = {
k.text.strip(): info.text.strip() if info.text.strip() else "Not found"
for _, (k, info) in enumerate(zip(list_info[::2], list_info[1::2]))
}
print(f"\n [{GREEN}>{WHITE}] Free-lookup")
for key, value in info_dict.items():
if value != "Not found":
print(f" ├── {key}: {value}")
else:
continue
# Path: lib/spam.py
async def spamcalls(p_n):
print(f"\n [{GREEN}>{WHITE}] Spamcalls")
url = f"https://spamcalls.net/en/number/{p_n}"
r = await Request(url, headers={'user-agent': random.choice(user_agent)}).get()
if r.status_code == 200:
print(f" └── {RED}!{WHITE} Spammer")
else:
print(f" └── {GREEN}>{WHITE} Not spammer")
# Path: lib/lookup.py
import phonenumbers
import json
from phonenumbers import carrier
from .reputation import *
from .free_lookup import free
from .spam import spamcalls
from lib.text import *
async def lookup(phone_number):
print()
parsed = phonenumbers.parse(phone_number)
operator = carrier.name_for_number(parsed, "fr")
line = phonenumbers.number_type(parsed)
if line == phonenumbers.PhoneNumberType.FIXED_LINE:
ligne = f" [{GREEN}>{WHITE}] Line type: Fixed"
elif line == phonenumbers.PhoneNumberType.MOBILE:
ligne = f" [{GREEN}>{WHITE}] Line type: Mobile"
else:
ligne = " [-] Line not found"
possible = phonenumbers.is_possible_number(parsed)
valid = phonenumbers.is_valid_number(parsed)
with open("lib/country.json", "r") as file:
read = json.load(file)
d = 0
countrys = []
for country, code in read.items():
d += 1
if phone_number.startswith(code):
countrys.append(country)
if d == 153:
break
else:
continue
else:
continue
print(f"{WHITE}📞 Phone number: {BLUE}{phone_number}{WHITE}")
if possible == True:
pos = {"possible": "✔️"}
else:
pos = {"possible": "❌"}
if valid == True:
val = {"valid": "✔️"}
else:
val = {"valid": "❌"}
print(f" [{GREEN}>{WHITE}] Possible: {pos['possible']}")
print(f" [{GREEN}>{WHITE}] Valid: {val['valid']}")
print()
if operator != "":
print(f" [{GREEN}>{WHITE}] Operator: {operator}")
else:
print(f" [-] Not Operator")
try:
print(f" [{GREEN}>{WHITE}] Possible location: " + str(countrys).replace("[", "").replace("]", "").replace("'", ""))
except:
print(f" [-] Not location")
print(ligne)
await reputation(phone_number)
|
coordinator = hass.data[DOMAIN][DATA_COORDINATORS][COORDINATOR_CHARGESESSIONS] | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: dan-r/HomeAssistant-Ohme
# Path: custom_components/ohme/const.py
DOMAIN = "ohme"
# Path: custom_components/ohme/const.py
DATA_COORDINATORS = "coordinators"
# Path: custom_components/ohme/const.py
COORDINATOR_CHARGESESSIONS = 0
# Path: custom_components/ohme/const.py
COORDINATOR_ADVANCED = 3
# Path: custom_components/ohme/const.py
DATA_CLIENT = "client"
# Path: custom_components/ohme/coordinator.py
class OhmeChargeSessionsCoordinator(DataUpdateCoordinator):
"""Coordinator to pull main charge state and power/current draw."""
def __init__(self, hass):
"""Initialise coordinator."""
super().__init__(
hass,
_LOGGER,
name="Ohme Charge Sessions",
update_interval=timedelta(seconds=30),
)
self._client = hass.data[DOMAIN][DATA_CLIENT]
async def _async_update_data(self):
"""Fetch data from API endpoint."""
try:
return await self._client.async_get_charge_sessions()
except BaseException:
raise UpdateFailed("Error communicating with API")
# Path: custom_components/ohme/coordinator.py
class OhmeAdvancedSettingsCoordinator(DataUpdateCoordinator):
"""Coordinator to pull CT clamp reading."""
def __init__(self, hass):
"""Initialise coordinator."""
super().__init__(
hass,
_LOGGER,
name="Ohme Advanced Settings",
update_interval=timedelta(minutes=1),
)
self._client = hass.data[DOMAIN][DATA_CLIENT]
async def _async_update_data(self):
"""Fetch data from API endpoint."""
try:
return await self._client.async_get_advanced_settings()
except BaseException:
raise UpdateFailed("Error communicating with API")
# Path: custom_components/ohme/utils.py
def charge_graph_in_slot(charge_start, points, skip_format=False):
"""Are we currently in a charge slot?"""
now = int(time())
data = points if skip_format else _format_charge_graph(charge_start, points)
# Loop through every value, skipping the last
for idx in range(0, len(data) - 1):
# This is our current point
if data[idx]["t"] < now and data[idx + 1]["t"] > now:
# If the delta line we are on is steeper than 10,
# we are in a charge slot.
if data[idx + 1]["y"] - data[idx]["y"] > 10:
return True
break
return False
# Path: custom_components/ohme/binary_sensor.py
import logging
from homeassistant.components.binary_sensor import (
BinarySensorDeviceClass,
BinarySensorEntity
)
from homeassistant.helpers.update_coordinator import CoordinatorEntity
from homeassistant.core import HomeAssistant, callback
from homeassistant.helpers.entity import generate_entity_id
from homeassistant.util.dt import (utcnow)
from .const import DOMAIN, DATA_COORDINATORS, COORDINATOR_CHARGESESSIONS, COORDINATOR_ADVANCED, DATA_CLIENT
from .coordinator import OhmeChargeSessionsCoordinator, OhmeAdvancedSettingsCoordinator
from .utils import charge_graph_in_slot
"""Platform for sensor integration."""
from __future__ import annotations
_LOGGER = logging.getLogger(__name__)
async def async_setup_entry(
hass: core.HomeAssistant,
config_entry: config_entries.ConfigEntry,
async_add_entities,
):
"""Setup sensors and configure coordinator."""
client = hass.data[DOMAIN][DATA_CLIENT]
|
) -> Union[IPResponse, None]: | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Almas-Ali/SpyIP
# Path: spyip/exceptions.py
class TooManyRequests(Exception):
pass
# Path: spyip/exceptions.py
class ConnectionTimeout(Exception):
pass
# Path: spyip/exceptions.py
class StatusError(Exception):
pass
# Path: spyip/models.py
class IPResponse(BaseModel):
"""
Example response from API:
{
"status": "success",
"continent": "Asia",
"continentCode": "AS",
"country": "India",
"countryCode": "IN",
"region": "DL",
"regionName": "National Capital Territory of Delhi",
"city": "New Delhi",
"district": "",
"zip": "110001",
"lat": 28.6139,
"lon": 77.209,
"timezone": "Asia/Kolkata",
"offset": 19800,
"currency": "INR",
"isp": "Google LLC",
"org": "Google LLC",
"as": "AS15169 Google LLC",
"asname": "GOOGLE",
"mobile": false,
"proxy": false,
"hosting": true,
"query": "142.250.193.206",
}
"""
status: str = Field(..., description='Status of the request.')
continent: str = Field(..., description='Continent name.')
continentCode: str = Field(..., description='Continent code.')
country: str = Field(..., description='Country name.')
countryCode: str = Field(..., description='Country code.')
region: str = Field(..., description='Region code.')
regionName: str = Field(..., description='Region name.')
city: str = Field(..., description='City name.')
district: str = Field(..., description='District name.')
zip_: str = Field(..., description='Zip code.')
lat: float = Field(..., description='Latitude.')
lon: float = Field(..., description='Longitude.')
timezone: str = Field(..., description='Timezone.')
offset: int = Field(..., description='Offset.')
currency: str = Field(..., description='Currency.')
isp: str = Field(..., description='ISP name.')
org: str = Field(..., description='Organization name.')
as_: str = Field(..., description='AS number and name.')
asname: str = Field(..., description='AS name.')
mobile: bool = Field(..., description='Mobile status.')
proxy: bool = Field(..., description='Proxy status.')
hosting: bool = Field(..., description='Hosting status.')
query: str = Field(..., description='IP address.')
class Config:
def alias_generator(x):
return x.replace('_', '')
populate_by_name = True
# fields = { # Alias for reserved keywords
# "as_": "as",
# "zip_": "zip",
# }
@field_validator('status')
def check_status(cls, v):
if v != 'success':
raise ValueError('Invalid IP address.')
return v
def json(self, **kwargs) -> str:
return self.model_dump_json(**kwargs)
# Path: spyip/models.py
class DNSResponse(BaseModel):
"""
Example response from API:
"dns": {
"ip": "74.125.73.83",
"geo": "United States - Google"
}
"""
ip: str = Field(..., description='IP address.')
geo: str = Field(..., description='Geo location.')
def json(self, **kwargs) -> str:
return self.model_dump_json(**kwargs)
# Path: spyip/backend.py
from typing import List, Union
from .exceptions import (
TooManyRequests,
ConnectionTimeout,
StatusError,
)
from .models import (
IPResponse,
DNSResponse,
)
import asyncio
import random
import string
import httpx
def get_random_string(length: int = 32) -> str:
"""Generate a random string of fixed length."""
letters = string.ascii_lowercase + string.digits
return ''.join(random.sample(letters, length))
# API endpoints for IP address lookup
trace_me_url = 'http://ip-api.com/json/'
trace_ip_url = 'http://ip-api.com/json/%(query)s'
trace_dns_url = f'http://{get_random_string(32)}.edns.ip-api.com/json/'
trace_ip_batch_url = 'http://ip-api.com/batch'
headers = {
'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8',
'Accept-Encoding': 'gzip, deflate',
'Accept-Language': 'en-US,en;q=0.5',
'Connection': 'keep-alive',
'Upgrade-Insecure-Requests': '1',
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:89.0) Gecko/20100101 Firefox/89.0',
}
def trace_me(
timeout: int = 5,
lang: str = 'en',
|
process_task(fake_job, fake_pipeline, fake_executor, fake_path, parallel_exec=True) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: leopedroso45/Stable-Diffusion-ImageGen
# Path: sevsd/process_task.py
def check_cuda_and_clear_cache():
r"""
Clears the CUDA cache if available, otherwise performs garbage collection.
This function is called to manage memory usage, particularly when working with large models or multiple image generations.
"""
if torch.cuda.is_available():
torch.cuda.empty_cache()
else:
gc.collect()
# Path: sevsd/process_task.py
def process_task(job, pipeline, executor, path, parallel_exec=True):
r"""
Processes a single image generation job using the specified pipeline and execution parameters.
This function handles the generation of one or more images based on a given job description. It supports both parallel and sequential execution modes. Generated images are saved to the specified path.
Parameters:
job (dict): A dictionary containing details for the image generation task. It includes 'prompt' and optionally 'negative_prompt'.
pipeline (callable): The Stable Diffusion pipeline callable used for generating images.
executor (dict): A dictionary containing execution parameters such as 'num_of_exec', 'cfg_scale', and 'inference_steps'.
path (str): The directory path where generated images will be saved.
parallel_exec (bool, optional): If True, generates all specified images in parallel. Defaults to True.
The function saves each generated image with a unique timestamp in the specified path and prints the save location. In case of any exceptions, they are caught and printed.
Example:
job = {
"prompt": "A scenic landscape",
"negative_prompt": "blurred image, black and white, watermarked image"
}
executor = {
"num_of_exec": 2,
"cfg_scale": 7,
"inference_steps": 50
}
pipeline = setup_pipeline("CompVis/stable-diffusion-v1-4")
process_task(job, pipeline, executor, "./generated-images", parallel_exec=False)
Note:
This function also handles CUDA cache clearing and garbage collection for memory management.
"""
def call_generate_image():
images = generate_image(job, pipeline, executor, parallel_exec)
if images is not None:
for image in images:
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S%f")
image_path = f"{path}/generated_image_{timestamp}.png"
image.save(image_path)
print(f"[sevsd] - image saved at {image_path}")
else:
print("[sevsd] - image generation failed due to memory constraints.")
check_cuda_and_clear_cache()
try:
path = check_os_path(path)
if job is not None:
if parallel_exec is not True:
num_images = executor.get("num_of_exec", 1)
for _ in range(num_images):
call_generate_image()
else:
call_generate_image()
except Exception as e:
print(f"[sevsd] - exception: {e}")
finally:
check_cuda_and_clear_cache()
# Path: sevsd/process_task.py
def check_os_path(path):
r"""
Checks if the given path exists, and if not, creates the necessary directories.
This function ensures that the output path for saving images is available.
Parameters:
path (str): The directory path to check and create if necessary.
Returns:
str: The verified or created directory path.
"""
if not os.path.exists(path):
os.makedirs(path)
print(f"[sevsd] - created path: {path}")
return path
# Path: tests/test_process_task.py
import unittest
import sys
from unittest.mock import patch, MagicMock
from sevsd.process_task import check_cuda_and_clear_cache, process_task, check_os_path
sys.path.append('../')
class TestProcessTask(unittest.TestCase):
@patch('sevsd.process_task.generate_image')
def test_process_task(self, mock_generate_image):
mock_image = MagicMock()
mock_image.save = MagicMock()
mock_generate_image.return_value = [mock_image]
fake_job = {"prompt": "prompt", "details": (None, 50, 1, 7.5)}
fake_pipeline = MagicMock()
fake_executor = {"num_of_exec": 1, "cfg_scale": 7}
fake_path = "test_path"
|
hooks.append(Hook(t, lambda grad: grad.T)) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Emperor-WS/PyEmber
# Path: ember/autograd/hook.py
class Hook:
"""
Hook class for attaching gradient functions to tensors.
Hooks allow users to attach custom gradient functions to tensors for
monitoring or modifying gradients during backpropagation.
Attributes:
- tensor (Tensor): The target tensor.
- grad_fn (callable): The gradient function to be applied to the tensor.
Methods:
- __init__(self, tensor, grad_fn): Constructor for Hook class.
- __repr__(self): String representation of the Hook instance.
"""
__slots__ = 'tensor', 'grad_fn'
def __init__(self, tensor, grad_fn):
"""
Constructor for the Hook class.
Args:
- tensor (Tensor): The target tensor.
- grad_fn (callable): The gradient function to be applied to the tensor.
"""
self.tensor = tensor
self.grad_fn = grad_fn
def __repr__(self):
"""
String representation of the Hook instance.
Returns:
- str: A string containing information about the tensor and its associated gradient function.
"""
# Extract the class name from the qualified name of the gradient function
grad_name = self.grad_fn.__qualname__.split('.')[0]
return f"Hook(tensor_id={self.tensor.id}, grad_fn={grad_name.upper()})"
# Path: ember/autograd/_utils.py
def numpy_unpad(x, pad_width):
"""
Remove padding from an array.
Args:
- x (numpy.ndarray): Input array.
- pad_width (tuple of ints): Amount of padding on each dimension.
Returns:
- numpy.ndarray: Unpadded array.
"""
slices = []
for pad in pad_width:
end = None if pad[1] == 0 else -pad[1]
slices.append(slice(pad[0], end ))
return x[tuple(slices)]
# Path: ember/autograd/_utils.py
def inv_permutation(permutation):
"""
Compute the inverse of a permutation.
Args:
- permutation (list): List representing a permutation.
Returns:
- list: Inverse permutation.
"""
inverse = [0] * len(permutation)
for original_idx, permuted_idx in enumerate(permutation):
inverse[permuted_idx] = original_idx
return inverse
# Path: ember/autograd/numeric.py
import numpy as np
import ember
from .hook import Hook
from ._utils import numpy_unpad, inv_permutation
def _T(t):
"""
Transpose operation on the input tensor.
Args:
- t: Input tensor.
Returns:
- Tensor: Resultant tensor with the transpose operation applied.
"""
t = ember.to_tensor(t) # Convert the input tensor to a Tensor
data = t.data.T # Transpose operation
requires_grad = t.requires_grad # Set requires_grad based on input tensor
hooks = []
# Register a hook for gradient computation if the input tensor requires it
if requires_grad:
|
labels_dict = read_yaml(parsed_args.params)["labels_mapping"] | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Hassi34/iot-device-identification
# Path: src/utils/common.py
def read_yaml(path_to_yaml: str) -> dict:
with open(path_to_yaml) as yaml_file:
content = yaml.safe_load(yaml_file)
return content
# Path: src/utils/sys_logging.py
def get_logger(logs_filepath: str):
logger.add(
logs_filepath,
format="{time} | {level} | {name}.{module}:{line} | {message}",
level="DEBUG",
rotation="10 KB",
retention="10 days",
compression="zip",
colorize=True,
enqueue=True,
catch=True,
encoding="utf-8",
)
return logger
# Path: src/utils/common.py
def write_dict_to_yaml(dict_input: dict, yaml_file_path: str):
try:
current_file_data = read_yaml(yaml_file_path)
current_file_data.update(dict_input)
with open(yaml_file_path, "w") as f:
yaml.dump(current_file_data, f)
except (FileNotFoundError , AttributeError):
with open(yaml_file_path, "w") as f:
yaml.dump(dict_input, f)
# Path: src/utils/data_ops.py
def gzip_np_arr(np_array: np.ndarray, filepath: str):
with gzip.GzipFile(filepath, "w") as f:
np.save(file=f, arr=np_array)
# Path: src/utils/data_ops.py
def get_fitted_pipeline(df, columns, KNN_IMPUTER_NEIGHBORS: int = 3):
ct = ColumnTransformer(
transformers=[("input_features", "passthrough", columns)], remainder="drop"
)
imputer = KNNImputer(n_neighbors=KNN_IMPUTER_NEIGHBORS)
scaler = StandardScaler()
pipeline = Pipeline(
steps=[("select_columns", ct), ("imputer", imputer), ("scaler", scaler)]
)
return pipeline.fit(df)
# Path: src/stage_03_preprocess_data.py
import argparse
import joblib
import pandas as pd
from src.utils.common import read_yaml
from src.utils.sys_logging import get_logger
from sklearn.preprocessing import LabelEncoder
from src.utils.common import write_dict_to_yaml
from src.utils.data_ops import gzip_np_arr
from sklearn.model_selection import train_test_split
from src.utils.data_ops import get_fitted_pipeline
from pathlib import Path
STAGE = "Preprocess Data"
def preprocess_data():
complete_df = pd.read_parquet(RAW_DATA_FILE_PATH)
logger.info(
f'The raw data file has been loaded from "{RAW_DATA_FILE_PATH}" with the shape "{complete_df.shape}"'
)
duplicate_rows = complete_df.duplicated().sum()
if duplicate_rows > 0:
logger.warning(
f"Found {duplicate_rows} duplicate rows, removing duplicate rows..."
)
complete_df = complete_df.drop_duplicates(keep="first")
X = complete_df.drop([TARGET_COLUMN_NAME], axis=1)
y = complete_df[TARGET_COLUMN_NAME]
feature_cols = params["input_features_schema"]
feature_cols = list(feature_cols.keys())
logger.info(f"Read {len(feature_cols)} feature columns from params")
data_processing_pipeline = get_fitted_pipeline(
X, feature_cols, KNN_IMPUTER_NEIGHBORS=KNN_IMPUTER_NEIGHBORS
)
Path(DATA_PREPROCESSING_PIPELINE_FILE_PATH).parent.absolute().mkdir(parents=True, exist_ok=True)
joblib.dump(data_processing_pipeline, DATA_PREPROCESSING_PIPELINE_FILE_PATH, compress=1)
logger.info(f"Saved the preprocessing pipeline to {DATA_PREPROCESSING_PIPELINE_FILE_PATH}")
data_processing_pipeline = joblib.load(DATA_PREPROCESSING_PIPELINE_FILE_PATH)
data_processing_pipeline
data_processing_pipeline = joblib.load(DATA_PREPROCESSING_PIPELINE_FILE_PATH)
logger.info(
f'Loaded sklearn data preprocessing pipeline from "{DATA_PREPROCESSING_PIPELINE_FILE_PATH}"'
)
X_transformed = data_processing_pipeline.transform(X)
logger.info(f'Dataframe shape after transformation is "{X_transformed.shape}"')
le = LabelEncoder()
le.fit(y)
labels_mapping_dict = {"labels_mapping": ""}
le_dict = dict(zip(le.transform(le.classes_), le.classes_))
le_dict = {int(k): v for k, v in le_dict.items()}
labels_mapping_dict["labels_mapping"] = le_dict
logger.info(f"Label encoding map has the dictionary: {le_dict}")
write_dict_to_yaml(labels_mapping_dict, parsed_args.params)
logger.info(f'Updated the label encoding map in the file at "{parsed_args.params}"')
|
text = "".join(text2sep_kata(text)[0]) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: see2023/Bert-VITS2-ext
# Path: config.py
class Resample_config:
class Preprocess_text_config:
class Bert_gen_config:
class Emo_gen_config:
class Train_ms_config:
class Webui_config:
class Server_config:
class Translate_config:
class Config:
def __init__(self, in_dir: str, out_dir: str, sampling_rate: int = 44100):
def from_dict(cls, dataset_path: str, data: Dict[str, any]):
def __init__(
self,
transcription_path: str,
cleaned_path: str,
train_path: str,
val_path: str,
config_path: str,
val_per_lang: int = 5,
max_val_total: int = 10000,
clean: bool = True,
):
def from_dict(cls, dataset_path: str, data: Dict[str, any]):
def __init__(
self,
config_path: str,
num_processes: int = 2,
device: str = "cuda",
use_multi_device: bool = False,
):
def from_dict(cls, dataset_path: str, data: Dict[str, any]):
def __init__(
self,
config_path: str,
num_processes: int = 2,
device: str = "cuda",
use_multi_device: bool = False,
):
def from_dict(cls, dataset_path: str, data: Dict[str, any]):
def __init__(
self,
config_path: str,
env: Dict[str, any],
base: Dict[str, any],
model: str,
num_workers: int,
spec_cache: bool,
keep_ckpts: int,
):
def from_dict(cls, dataset_path: str, data: Dict[str, any]):
def __init__(
self,
device: str,
model: str,
v_model: str,
config_path: str,
language_identification_library: str,
port: int = 7860,
share: bool = False,
debug: bool = False,
):
def from_dict(cls, dataset_path: str, data: Dict[str, any]):
def __init__(
self, models: List[Dict[str, any]], port: int = 5000, device: str = "cuda"
):
def from_dict(cls, data: Dict[str, any]):
def __init__(self, app_key: str, secret_key: str):
def from_dict(cls, data: Dict[str, any]):
def __init__(self, config_path: str):
# Path: text/japanese.py
def text2sep_kata(text: str) -> (list, list):
parsed = pyopenjtalk.run_frontend(text)
res = []
sep = []
for parts in parsed:
word, yomi = replace_punctuation(parts["string"]), parts["pron"].replace(
"’", ""
)
if yomi:
if re.match(_MARKS, yomi):
if len(word) > 1:
word = [replace_punctuation(i) for i in list(word)]
yomi = word
res += yomi
sep += word
continue
elif word not in rep_map.keys() and word not in rep_map.values():
word = ","
yomi = word
res.append(yomi)
else:
if word in _SYMBOL_TOKENS:
res.append(word)
elif word in ("っ", "ッ"):
res.append("ッ")
elif word in _NO_YOMI_TOKENS:
pass
else:
res.append(word)
sep.append(word)
return sep, [hira2kata(i) for i in res], get_accent(parsed)
# Path: for_deploy/infer_utils.py
import sys
import torch
from transformers import (
AutoModelForMaskedLM,
AutoTokenizer,
DebertaV2Model,
DebertaV2Tokenizer,
ClapModel,
ClapProcessor,
)
from config import config
from text.japanese import text2sep_kata
class BertFeature:
def __init__(self, model_path, language="ZH"):
self.model_path = model_path
self.language = language
self.tokenizer = None
self.model = None
self.device = None
self._prepare()
def _get_device(self, device=config.bert_gen_config.device):
if (
sys.platform == "darwin"
and torch.backends.mps.is_available()
and device == "cpu"
):
device = "mps"
if not device:
device = "cuda"
return device
def _prepare(self):
self.device = self._get_device()
if self.language == "EN":
self.tokenizer = DebertaV2Tokenizer.from_pretrained(self.model_path)
self.model = DebertaV2Model.from_pretrained(self.model_path).to(self.device)
else:
self.tokenizer = AutoTokenizer.from_pretrained(self.model_path)
self.model = AutoModelForMaskedLM.from_pretrained(self.model_path).to(
self.device
)
self.model.eval()
def get_bert_feature(self, text, word2ph):
if self.language == "JP":
|
color = get_activation(self.cfg.color_activation)(features) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: chinhsuanwu/ifusion-threestudio
# Path: threestudio/models/materials/base.py
class BaseMaterial(BaseModule):
@dataclass
class Config(BaseModule.Config):
pass
cfg: Config
requires_normal: bool = False
requires_tangent: bool = False
def configure(self):
pass
def forward(self, *args, **kwargs) -> Float[Tensor, "*B 3"]:
raise NotImplementedError
def export(self, *args, **kwargs) -> Dict[str, Any]:
return {}
# Path: threestudio/models/networks.py
def get_encoding(n_input_dims: int, config) -> nn.Module:
# input suppose to be range [0, 1]
encoding: nn.Module
if config.otype == "ProgressiveBandFrequency":
encoding = ProgressiveBandFrequency(n_input_dims, config_to_primitive(config))
elif config.otype == "ProgressiveBandHashGrid":
encoding = ProgressiveBandHashGrid(n_input_dims, config_to_primitive(config))
elif config.otype == "HashGridSpatialTime":
encoding = TCNNEncodingSpatialTime(n_input_dims, config) # 4D-fy encoding
else:
encoding = TCNNEncoding(n_input_dims, config_to_primitive(config))
encoding = CompositeEncoding(
encoding,
include_xyz=config.get("include_xyz", False),
xyz_scale=2.0,
xyz_offset=-1.0,
) # FIXME: hard coded
return encoding
# Path: threestudio/models/networks.py
def get_mlp(n_input_dims, n_output_dims, config) -> nn.Module:
network: nn.Module
if config.otype == "VanillaMLP":
network = VanillaMLP(n_input_dims, n_output_dims, config_to_primitive(config))
elif config.otype == "SphereInitVanillaMLP":
network = SphereInitVanillaMLP(
n_input_dims, n_output_dims, config_to_primitive(config)
)
else:
assert (
config.get("sphere_init", False) is False
), "sphere_init=True only supported by VanillaMLP"
network = TCNNNetwork(n_input_dims, n_output_dims, config_to_primitive(config))
return network
# Path: threestudio/utils/ops.py
def dot(x, y):
return torch.sum(x * y, -1, keepdim=True)
# Path: threestudio/utils/ops.py
def get_activation(name) -> Callable:
if name is None:
return lambda x: x
name = name.lower()
if name == "none":
return lambda x: x
elif name == "lin2srgb":
return lambda x: torch.where(
x > 0.0031308,
torch.pow(torch.clamp(x, min=0.0031308), 1.0 / 2.4) * 1.055 - 0.055,
12.92 * x,
).clamp(0.0, 1.0)
elif name == "exp":
return lambda x: torch.exp(x)
elif name == "shifted_exp":
return lambda x: torch.exp(x - 1.0)
elif name == "trunc_exp":
return trunc_exp
elif name == "shifted_trunc_exp":
return lambda x: trunc_exp(x - 1.0)
elif name == "sigmoid":
return lambda x: torch.sigmoid(x)
elif name == "tanh":
return lambda x: torch.tanh(x)
elif name == "shifted_softplus":
return lambda x: F.softplus(x - 1.0)
elif name == "scale_-11_01":
return lambda x: x * 0.5 + 0.5
else:
try:
return getattr(F, name)
except AttributeError:
raise ValueError(f"Unknown activation function: {name}")
# Path: threestudio/models/materials/no_material.py
import random
import torch
import torch.nn as nn
import torch.nn.functional as F
import threestudio
from dataclasses import dataclass, field
from threestudio.models.materials.base import BaseMaterial
from threestudio.models.networks import get_encoding, get_mlp
from threestudio.utils.ops import dot, get_activation
from threestudio.utils.typing import *
@threestudio.register("no-material")
class NoMaterial(BaseMaterial):
@dataclass
class Config(BaseMaterial.Config):
n_output_dims: int = 3
color_activation: str = "sigmoid"
input_feature_dims: Optional[int] = None
mlp_network_config: Optional[dict] = None
requires_normal: bool = False
cfg: Config
def configure(self) -> None:
self.use_network = False
if (
self.cfg.input_feature_dims is not None
and self.cfg.mlp_network_config is not None
):
self.network = get_mlp(
self.cfg.input_feature_dims,
self.cfg.n_output_dims,
self.cfg.mlp_network_config,
)
self.use_network = True
self.requires_normal = self.cfg.requires_normal
def forward(
self, features: Float[Tensor, "B ... Nf"], **kwargs
) -> Float[Tensor, "B ... Nc"]:
if not self.use_network:
assert (
features.shape[-1] == self.cfg.n_output_dims
), f"Expected {self.cfg.n_output_dims} output dims, only got {features.shape[-1]} dims input."
|
origin_sync += f'{TEXT["bright_green"]}{glyph("ahead")} {ahead}{RESET}' | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: jasursadikov/mud
# Path: utils.py
TEXT = {
'white': '\033[37m',
'gray': '\033[90m',
'black': '\033[30m',
'red': '\033[31m',
'green': '\033[32m',
'yellow': '\033[33m',
'blue': '\033[34m',
'magenta': '\033[35m',
'cyan': '\033[36m',
'bright_white': '\033[97m',
'bright_red': '\033[91m',
'bright_green': '\033[92m',
'bright_yellow': '\033[93m',
'bright_blue': '\033[94m',
'bright_magenta': '\033[95m',
'bright_cyan': '\033[96m',
}
# Path: utils.py
BACK = {
'white': '\033[47m',
'medium_gray': '\033[100m',
'black': '\033[40m',
'red': '\033[41m',
'green': '\033[42m',
'yellow': '\033[43m',
'blue': '\033[44m',
'magenta': '\033[45m',
'cyan': '\033[46m',
'bright_white': '\033[107m',
'bright_red': '\033[101m',
'bright_green': '\033[102m',
'bright_yellow': '\033[103m',
'bright_blue': '\033[104m',
'bright_magenta': '\033[105m',
'bright_cyan': '\033[106m',
}
# Path: utils.py
RESET = '\033[0m'
# Path: utils.py
STYLES = {
'bold': '\033[1m',
'dim': '\033[2m',
'italic': '\033[3m',
'underline': '\033[4m',
'blink': '\033[5m',
}
# Path: utils.py
END_STYLES = {
'bold': '\033[22m',
'dim': '\033[22m',
'italic': '\033[23m',
'underline': '\033[24m',
'blink': '\033[25m',
}
# Path: utils.py
def glyph(key: str) -> str:
return GLYPHS[key][0] if settings.mud_settings['nerd_fonts'] else GLYPHS[key][1]
# Path: commands.py
import utils
import asyncio
import subprocess
from utils import TEXT, BACK, RESET, STYLES, END_STYLES, glyph
from typing import List, Dict
from collections import Counter
from prettytable import PrettyTable, PLAIN_COLUMNS
class Commands:
def __init__(self, repos):
self.repos = repos
self.label_color_cache = {}
self.current_color_index = 0
# `mud status` command implementation
def status(self, repos: Dict[str, List[str]]) -> None:
table = self._get_table()
for path, tags in repos.items():
formatted_path = self._get_formatted_path(path)
branch = self._get_branch_status(path)
author = self._get_authors_name(path)
commit = self._get_commit_message(path, 30)
colored_labels = self._get_formatted_labels(tags)
# Sync with origin status
ahead_behind_cmd = subprocess.run(['git', 'rev-list', '--left-right', '--count', 'HEAD...@{upstream}'],
text=True, cwd=path, capture_output=True)
stdout = ahead_behind_cmd.stdout.strip().split()
if len(stdout) >= 2:
ahead, behind = stdout[0], stdout[1]
origin_sync = ''
if ahead and ahead != '0':
|
self.mm_projector = build_vision_projector(config) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Q-MM/PureMM
# Path: model/multimodal_encoder/builder.py
def build_vision_tower(vision_tower_cfg, **kwargs):
vision_tower = getattr(vision_tower_cfg, 'mm_vision_tower', getattr(vision_tower_cfg, 'vision_tower', None))
is_absolute_path_exists = os.path.exists(vision_tower)
if is_absolute_path_exists or vision_tower.startswith("openai") or vision_tower.startswith("laion"):
return CLIPVisionTower(vision_tower, args=vision_tower_cfg, **kwargs)
raise ValueError(f'Unknown vision tower: {vision_tower}')
# Path: model/multimodal_projector/builder.py
def build_vision_projector(config, delay_load=False, **kwargs):
projector_type = getattr(config, 'mm_projector_type', 'linear')
if projector_type == 'linear':
return nn.Linear(config.mm_hidden_size, config.hidden_size)
mlp_gelu_match = re.match(r'^mlp(\d+)x_gelu$', projector_type)
if mlp_gelu_match:
mlp_depth = int(mlp_gelu_match.group(1))
modules = [nn.Linear(config.mm_hidden_size, config.hidden_size)]
for _ in range(1, mlp_depth):
modules.append(nn.GELU())
modules.append(nn.Linear(config.hidden_size, config.hidden_size))
return nn.Sequential(*modules)
larger_mlp_gelu_match = re.match(r'^larger_mlp(\d+)x_gelu$', projector_type)
if larger_mlp_gelu_match:
mlp_depth = int(mlp_gelu_match.group(1))
modules = [nn.Linear(config.mm_hidden_size, config.mm_hidden_size)]
for _ in range(1, mlp_depth-1):
modules.append(nn.GELU())
modules.append(nn.Linear(config.mm_hidden_size, config.mm_hidden_size))
modules.append(nn.Linear(config.mm_hidden_size, config.hidden_size))
return nn.Sequential(*modules)
if projector_type == 'identity':
return IdentityMap()
raise ValueError(f'Unknown projector type: {projector_type}')
# Path: model/PureMM_arch.py
from abc import ABC, abstractmethod
from .multimodal_encoder.builder import build_vision_tower
from .multimodal_projector.builder import build_vision_projector
import torch
import torch.nn as nn
# Copyright 2023 Haotian Liu
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
IGNORE_INDEX = -100
IMAGE_TOKEN_INDEX = -200
DEFAULT_IMAGE_TOKEN = "<image>"
DEFAULT_IMAGE_PATCH_TOKEN = "<im_patch>"
DEFAULT_IM_START_TOKEN = "<im_start>"
DEFAULT_IM_END_TOKEN = "<im_end>"
def rank0_print(rank, *args):
if rank == 0:
print(*args)
class PureMMMetaModel:
def __init__(self, config):
super(PureMMMetaModel, self).__init__(config)
if hasattr(config, "mm_vision_tower"):
self.vision_tower = build_vision_tower(config, delay_load=True)
# self.mm_projector = nn.Linear(config.mm_hidden_size, config.hidden_size)
|
return SpotifyApi(os.getenv("CLIENT_ID"), os.getenv("CLIENT_SECRET")) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: Ananya2001-an/spotify-py-sdk
# Path: spotify_py_sdk/spotify_api.py
class SpotifyApi:
"""Create an api instance and call the various endpoint methods.
:param client_id: Client_ID for your app
:type client_id: str
:param client_secret: Client_Secret for your app
:type client_secret: str
:param config: pass :class:`SdkConfig` instance, defaults to None
:type config: :class:`SdkConfig`, optional
"""
_root_url: str = "https://api.spotify.com/v1/"
def __init__(self, client_id: str, client_secret: str, config: Optional[SdkConfig] = None):
"""Constructor method
"""
self.access_token_manager: AccessTokenManager = AccessTokenManager(client_id, client_secret)
self.sdk_config: Optional[SdkConfig] = config
self.albums: Albums = Albums(self)
self.artists: Artists = Artists(self)
self.audiobooks: Audiobooks = Audiobooks(self)
self.browse: Browse = Browse(self)
self.chapters: Chapters = Chapters(self)
self.episodes: Episodes = Episodes(self)
self.recommendations: Recommendations = Recommendations(self)
self.markets: Markets = Markets(self)
# self.player: Player = Player(self) # need different auth strategy; yet to be implemented
self.playlists: Playlists = Playlists(self)
self.shows: Shows = Shows(self)
self.tracks: Tracks = Tracks(self)
self.users: Users = Users(self)
self.search: Search = Search(self)
# self.current_user: CurrentUser = CurrentUser(self) # need different auth strategy; yet to be implemented
@classmethod
def fetch_results(cls, url: str, opts: dict):
"""Fetch results by making a request to the given URL
"""
try:
result = requests.request(method=opts["method"], url=url, headers=opts["headers"], data=opts["body"])
return result.json()
except HTTPError as e:
raise f"Failed to fetch result! {e}"
def make_request(self, method: Literal["GET", "POST", "PUT", "DELETE"], url: str, body: Optional[any] = None,
content_type: Optional[str] = None):
"""Get access token and make necessary request call to the api endpoint
"""
try:
access_token = self.access_token_manager.get_access_token()
except HTTPError as e:
raise "Access Token not available! Authenticate again."
full_url = SpotifyApi._root_url + url
opts = {
"method": method,
"headers": {
"Authorization": f"Bearer {access_token}",
"Content-Type": content_type if content_type else "application/json"
},
"body": json.dumps(body) if body and type(body) is not str else body
}
try:
if self.sdk_config:
if self.sdk_config.before_request:
self.sdk_config.before_request(full_url, opts)
if self.sdk_config.fetch:
result = self.sdk_config.fetch(full_url, opts)
else:
result = SpotifyApi.fetch_results(full_url, opts)
if self.sdk_config.after_request:
self.sdk_config.after_request(full_url, opts, result)
return result
return SpotifyApi.fetch_results(full_url, opts)
except (HTTPError, ValueError, InterruptedError) as e:
raise e
# handled = self.sdk_config.error_handler.handleErrors(e)
# if not handled:
# raise Exception("Failed to make request! Try again.")
# Path: spotify_py_sdk/endpoints/recommendations.py
class RecommendationsRequestRequiredArguments:
def __init__(self, seed_artists: Optional[list[str]] = None, seed_genres: Optional[list[str]] = None, seed_tracks: Optional[list[str]] = None):
self.seed_artists = seed_artists
self.seed_genres = seed_genres
self.seed_tracks = seed_tracks
# Path: tests/endpoints/test_recommendations.py
import json
import pytest
import os
from spotify_py_sdk import SpotifyApi
from spotify_py_sdk.endpoints.recommendations import RecommendationsRequestRequiredArguments
from dotenv import load_dotenv
load_dotenv()
@pytest.fixture
def api():
|
return AsyncSeq2SeqTrainer if training_args.async_grad else Seq2SeqTrainer | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: kyleliang919/Optimizer-Zoo
# Path: optimizer_zoo/Trainer/async_trainer.py
class AsyncTrainer(Trainer):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.accelerator.sync_gradients = None
def training_step(self, model, inputs):
# make sure the gradient is not automatically synced
with model.no_sync():
model.train()
inputs = self._prepare_inputs(inputs)
if is_sagemaker_mp_enabled():
loss_mb = smp_forward_backward(model, inputs, self.args.gradient_accumulation_steps)
return loss_mb.reduce_mean().detach().to(self.args.device)
with self.compute_loss_context_manager():
loss = self.compute_loss(model, inputs)
if self.args.n_gpu > 1:
loss = loss.mean() # mean() to average on multi-gpu parallel training
if self.use_apex:
with amp.scale_loss(loss, self.optimizer) as scaled_loss:
scaled_loss.backward()
else:
self.accelerator.backward(loss)
return loss.detach() / self.args.gradient_accumulation_steps
# Path: optimizer_zoo/Trainer/async_trainer.py
class AsyncSFTTrainer(SFTTrainer):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
def training_step(self, model, inputs):
# make sure the gradient is not automatically synced
with model.no_sync():
model.train()
inputs = self._prepare_inputs(inputs)
if is_sagemaker_mp_enabled():
loss_mb = smp_forward_backward(model, inputs, self.args.gradient_accumulation_steps)
return loss_mb.reduce_mean().detach().to(self.args.device)
with self.compute_loss_context_manager():
loss = self.compute_loss(model, inputs)
if self.args.n_gpu > 1:
loss = loss.mean() # mean() to average on multi-gpu parallel training
if self.use_apex:
with amp.scale_loss(loss, self.optimizer) as scaled_loss:
scaled_loss.backward()
else:
self.accelerator.backward(loss)
return loss.detach() / self.args.gradient_accumulation_steps
# Path: optimizer_zoo/Trainer/async_trainer.py
class AsyncDPOTrainer(DPOTrainer):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
def training_step(self, model, inputs):
# make sure the gradient is not automatically synced
with model.no_sync():
model.train()
inputs = self._prepare_inputs(inputs)
if is_sagemaker_mp_enabled():
loss_mb = smp_forward_backward(model, inputs, self.args.gradient_accumulation_steps)
return loss_mb.reduce_mean().detach().to(self.args.device)
with self.compute_loss_context_manager():
loss = self.compute_loss(model, inputs)
if self.args.n_gpu > 1:
loss = loss.mean() # mean() to average on multi-gpu parallel training
if self.use_apex:
with amp.scale_loss(loss, self.optimizer) as scaled_loss:
scaled_loss.backward()
else:
self.accelerator.backward(loss)
return loss.detach() / self.args.gradient_accumulation_steps
# Path: optimizer_zoo/Trainer/async_trainer.py
class AsyncSeq2SeqTrainer(Seq2SeqTrainer):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.accelerator.sync_gradients = None
def training_step(self, model, inputs):
# make sure the gradient is not automatically synced
with model.no_sync():
model.train()
inputs = self._prepare_inputs(inputs)
if is_sagemaker_mp_enabled():
loss_mb = smp_forward_backward(model, inputs, self.args.gradient_accumulation_steps)
return loss_mb.reduce_mean().detach().to(self.args.device)
with self.compute_loss_context_manager():
loss = self.compute_loss(model, inputs)
if self.args.n_gpu > 1:
loss = loss.mean() # mean() to average on multi-gpu parallel training
if self.use_apex:
with amp.scale_loss(loss, self.optimizer) as scaled_loss:
scaled_loss.backward()
else:
self.accelerator.backward(loss)
return loss.detach() / self.args.gradient_accumulation_steps
# Path: optimizer_zoo/Trainer/utils.py
from transformers import Trainer, Seq2SeqTrainer
from trl import SFTTrainer, DPOTrainer
from .async_trainer import AsyncTrainer, AsyncSFTTrainer, AsyncDPOTrainer, AsyncSeq2SeqTrainer
def create_trainer(training_args):
if training_args.task == "pretraining":
return AsyncTrainer if training_args.async_grad else Trainer
elif training_args.task == "sft":
return AsyncSFTTrainer if training_args.async_grad else SFTTrainer
elif training_args.task == "dpo":
return AsyncDPOTrainer if training_args.async_grad else DPOTrainer
elif training_args.task == "seq2seq":
|
residual_block(channel_size=16), | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: giaminhgist/3D-DAM
# Path: lib/model/attention_block.py
class SpatialAttention3D(nn.Module):
def __init__(self, out_channel=64, kernel_size=3, stride=1, padding=1):
super(SpatialAttention3D, self).__init__()
self.conv = nn.Conv3d(2, out_channel,
kernel_size=kernel_size, stride=stride, padding=padding, bias=False)
self.sigmoid = nn.Sigmoid()
def forward(self, x):
residual = x
avg_out = torch.mean(x, dim=1, keepdim=True)
max_out, _ = torch.max(x, dim=1, keepdim=True)
x = torch.cat([avg_out, max_out], dim=1)
x = self.conv(x)
x = self.sigmoid(x)
out = x * residual
return out
# Path: lib/model/attention_block.py
class ChannelAttention3D(nn.Module):
def __init__(self, in_planes=64, ratio=8):
super(ChannelAttention3D, self).__init__()
self.avg_pool = nn.AdaptiveAvgPool3d(1)
self.max_pool = nn.AdaptiveMaxPool3d(1)
self.fc = nn.Sequential(nn.Conv3d(in_planes, in_planes // ratio, 1, bias=False),
nn.ReLU(),
nn.Conv3d(in_planes // ratio, in_planes, 1, bias=False))
self.sigmoid = nn.Sigmoid()
def forward(self, x):
residual = x
avg_out = self.fc(self.avg_pool(x))
max_out = self.fc(self.max_pool(x))
out = avg_out + max_out
return self.sigmoid(out) * residual
# Path: lib/model/attention_block.py
class residual_block(nn.Module):
def __init__(self, channel_size=64):
super(residual_block, self).__init__()
self.conv = nn.Conv3d(channel_size, channel_size, kernel_size=3, padding=1)
self.relu = nn.ReLU()
self.bn = nn.BatchNorm3d(channel_size)
def forward(self, x):
residual = x
y = self.conv(x)
y = self.bn(y)
y = self.relu(y)
out = y + residual
return out
# Path: lib/model/DuoAttention.py
import numpy as np
import torch
from torch import nn
from lib.model.attention_block import SpatialAttention3D, ChannelAttention3D, residual_block
class DAM(nn.Module):
def __init__(self, channels=64):
super(DAM, self).__init__()
self.sa = SpatialAttention3D(out_channel=channels)
self.ca = ChannelAttention3D(in_planes=channels)
def forward(self, x):
residual = x
out = self.ca(x)
out = self.sa(out)
out = out + residual
return out
class Duo_Attention(nn.Module):
def __init__(
self, input_size=(1, 169, 208, 179), num_classes=3, dropout=0
):
super().__init__()
self.conv = nn.Sequential(
nn.Conv3d(input_size[0], 8, 3, padding=1),
nn.BatchNorm3d(8),
nn.ReLU(),
# nn.MaxPool3d(2, 2),
nn.Conv3d(8, 16, 3, padding=1, stride=2),
nn.BatchNorm3d(16),
nn.ReLU(),
|
decrypted_message = decrypt_message(encrypted_input.encode(), key) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: itsluminous/EasyEncryption
# Path: core.py
def generate_key():
"""Generate a Fernet key."""
return Fernet.generate_key()
# Path: core.py
def encrypt_message(message, key):
"""Encrypt a message using the provided key."""
fernet = Fernet(key)
encrypted = fernet.encrypt(message.encode())
return encrypted
# Path: core.py
def decrypt_message(encrypted_message, key):
"""Decrypt an encrypted message using the provided key."""
fernet = Fernet(key)
decrypted = fernet.decrypt(encrypted_message).decode()
return decrypted
# Path: core.py
def encrypt_file(file_path, key):
"""Encrypt a file using the provided key."""
try:
with open(file_path, 'r', encoding='utf-8') as file:
content = file.read()
encrypted_content = encrypt_message(content, key)
with open(file_path + '.enc', 'wb') as encrypted_file:
encrypted_file.write(encrypted_content)
print(f"\nFile '{file_path}' encrypted successfully.")
except FileNotFoundError:
print("\nFile not found.")
# Path: core.py
def decrypt_file(file_path, key):
"""Decrypt an encrypted file using the provided key."""
try:
with open(file_path, 'rb', encoding='utf-8') as file:
encrypted_content = file.read()
decrypted_content = decrypt_message(encrypted_content, key)
decrypted_file_path = file_path[:-4] # Remove the '.enc' extension
with open(decrypted_file_path, 'w', encoding='utf-8') as decrypted_file:
decrypted_file.write(decrypted_content)
print(f"\nFile '{file_path}' decrypted successfully.")
except FileNotFoundError:
print("\nFile not found.")
except ValueError:
print("\nInvalid decryption key or file content.")
# Path: script.py
from core import generate_key, encrypt_message, decrypt_message, encrypt_file, decrypt_file
"""
Script providing a user interface for encryption and decryption operations.
"""
def generate_new_key():
"""
Generate a new encryption key.
Returns:
- bytes: New encryption key.
"""
key = generate_key()
print(f"\nGenerated Key: {key.decode()}")
return key
def enter_user_key():
"""
Prompt user to enter a key.
Returns:
- bytes: User-entered key.
"""
print("\nEnter the key:")
return input().encode()
def encrypt_user_message(key):
"""
Encrypt a user-entered message.
Parameters:
- key (bytes): Encryption key.
"""
if key is None:
print("\nPlease generate or enter a key first.")
else:
print("\nEnter a message to encrypt (press Enter twice to finish):")
lines = []
while True:
line = input()
if not line:
break
lines.append(line)
user_input = '\n'.join(lines)
encrypted_message = encrypt_message(user_input, key)
print(f"\nEncrypted message: {encrypted_message}")
def decrypt_user_message(key):
"""
Decrypt a user-entered message.
Parameters:
- key (bytes): Decryption key.
"""
if key is None:
print("\nPlease generate or enter a key first.")
else:
print("\nEnter the encrypted message (press Enter twice to finish):")
lines = []
while True:
line = input()
if not line:
break
lines.append(line)
encrypted_input = '\n'.join(lines)
|
response = await resource_not_found(obj, exc) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: gardenifi/server
# Path: app/main_app.py
INVALID_DATA = "Invalid data: Unable to process the provided data"
class GlobalVars:
class WifiData(BaseModel):
class ValveData(BaseModel):
class BleData(BaseModel):
def __init__(self):
def refresh_set(self):
def refresh_set(self, value):
async def index():
async def resource_not_found(request: Request, exc: HTTPException):
async def read_ble_data(page: int = None):
async def write_ble_data(data: BleData):
async def discover_wifi(chunked: int = None, page: int = None):
async def save_wifi(data: WifiData):
async def turn(data: ValveData):
async def check_mqtt():
def web_server():
def setup_gpio():
def parse_arguments():
def main():
# Path: app/main_app.py
@app.exception_handler(404)
async def resource_not_found(request: Request, exc: HTTPException):
"""Not found error."""
logger.error(f"Request: {request}")
return JSONResponse(status_code=404, content={"detail": str(exc.detail)})
# Path: tests/api/resource_not_found_test.py
import json
import pytest
from fastapi.testclient import TestClient
from fastapi import HTTPException, Request
from fastapi.responses import JSONResponse
from app.main_app import app
from app.main_app import resource_not_found
"""MIT License
Copyright (c) 2023, Marios Karagiannopoulos
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.
**Attribution Requirement:**
When using or distributing the software, an attribution to Marios Karagiannopoulos must be included.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.
"""
client = TestClient(app)
scope = {"type": "http", "http_version": "1.1", "method": "GET", "path": "/"}
@pytest.fixture(scope="function")
async def request_obj():
"""Request object creation fixture"""
return Request(scope)
class TestResourceNotFound:
"""
Test class for the 'resource_not_found' error handler function.
"""
@pytest.mark.asyncio
async def test_returns_json_response_with_status_code_404_and_detail_of_httpexception(self, obj=request_obj):
"""
Test for returning a JSONResponse object with status code 404 and the detail of the HTTPException passed as an argument.
"""
exc = HTTPException(status_code=404, detail="Not found")
|
Logmanager(args.log) | You will be given python files from a code repository, with the current file being shown last. Your task is to predict the next line of code in the current file.
NOTE: You should only predict the next line in the current file. Do not produce more than one line, and do not provide any explanation.
====REPOSITORY====
# Repo Name: xiaoye0x0/pfgo_tg_bot
# Path: utils/task/model.py
class Task(metaclass=SingletonMeta):
def __init__(self, args) -> None:
self.conf_file = args.config
self.bot_token: str = ""
self.pfgo_url: str = ""
self.username: str = ""
self.password: str = ""
self.hide: list = []
self.webhook_url = ""
self.webhook_port = ""
self.running_host = ""
self.running_port = 0
self._init_conf()
def _init_conf(self):
config = configparser.ConfigParser()
config.read(self.conf_file)
self.bot_token = config.get("bot", "token")
self.pfgo_url = config.get("pfgo", "url")
self.username = config.get("pfgo", "username")
self.password = config.get("pfgo", "password")
self.hide += config.get("pfgo", "hide").split(",")
self.webhook_url = config.get("webhook", "webhook_url")
self.webhook_port = config.get("webhook", "webhook_port")
self.running_host = config.get("webhook", "running_host")
self.running_port = int(config.get("webhook", "running_port"))
# Path: utils/log.py
class Logmanager(metaclass=SingletonMeta):
log_list = []
log_list_lock = threading.Lock()
path = "./"
def __init__(self, path: str) -> None:
Logmanager.path = path
@classmethod
def create_logger(cls, name=None):
if name is None:
name = "default"
logger = logging.getLogger(name)
if name not in cls.log_list:
with Logmanager.log_list_lock:
if name not in cls.log_list:
cls.log_list.append(name)
logger.setLevel(logging.INFO)
logfile = f"{Logmanager.path}/log.log"
fh = RotatingFileHandler(
logfile,
mode="a",
maxBytes=1024 * 1024 * 10,
backupCount=2,
encoding="utf-8",
)
formatter = logging.Formatter(
"[%(name)s] [%(asctime)s] [%(levelname)s] %(message)s",
"%Y%m%d-%H:%M:%S",
)
fh.setFormatter(formatter)
logger.addHandler(fh)
ch = logging.StreamHandler()
ch.setFormatter(formatter)
logger.addHandler(ch)
fh.close()
ch.close()
return logger
# Path: utils/task/set_args.py
import os
import argparse
from .model import Task
from ..log import Logmanager
def is_file_exists(file_path) -> bool:
r = os.path.exists(file_path)
if not r:
LOGGER.error(f"文件{file_path}不存在")
return r
def create_folder_if_not_exists(folder_path):
if not folder_path:
return
if not os.path.exists(folder_path):
os.makedirs(folder_path)
def parse_command_line_args():
"""
-c --config: 配置文件
--log: 日志存放位置
"""
parser = argparse.ArgumentParser(description="运行参数")
parser.add_argument("--config", "-c", type=str, default="./config.ini", help="配置文件")
parser.add_argument("--log", type=str, default="./", help="日志存放文件夹的位置,默认放到当前路径")
args = parser.parse_args()
# 初始化日志模块
global LOGGER
create_folder_if_not_exists(args.log)
|
End of preview. Expand
in Dataset Viewer.
README.md exists but content is empty.
Use the Edit dataset card button to edit it.
- Downloads last month
- 39