File size: 10,386 Bytes
3966f89 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 |
#!/usr/bin/env python3
import os
import glob
import time
import pandas as pd
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from diffusers import StableDiffusionPipeline
import fitz
import requests
from PIL import Image
import logging
import asyncio
import aiofiles
from io import BytesIO
from dataclasses import dataclass
from typing import Optional
import gradio as gr
logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s")
logger = logging.getLogger(__name__)
log_records = []
class LogCaptureHandler(logging.Handler):
def emit(self, record):
log_records.append(record)
logger.addHandler(LogCaptureHandler())
@dataclass
class ModelConfig:
name: str
base_model: str
size: str
domain: Optional[str] = None
model_type: str = "causal_lm"
@property
def model_path(self):
return f"models/{self.name}"
@dataclass
class DiffusionConfig:
name: str
base_model: str
size: str
domain: Optional[str] = None
@property
def model_path(self):
return f"diffusion_models/{self.name}"
class ModelBuilder:
def __init__(self):
self.config = None
self.model = None
self.tokenizer = None
def load_model(self, model_path: str, config: Optional[ModelConfig] = None):
self.model = AutoModelForCausalLM.from_pretrained(model_path)
self.tokenizer = AutoTokenizer.from_pretrained(model_path)
if self.tokenizer.pad_token is None:
self.tokenizer.pad_token = self.tokenizer.eos_token
if config:
self.config = config
self.model.to("cuda" if torch.cuda.is_available() else "cpu")
return self
def save_model(self, path: str):
os.makedirs(os.path.dirname(path), exist_ok=True)
self.model.save_pretrained(path)
self.tokenizer.save_pretrained(path)
class DiffusionBuilder:
def __init__(self):
self.config = None
self.pipeline = None
def load_model(self, model_path: str, config: Optional[DiffusionConfig] = None):
self.pipeline = StableDiffusionPipeline.from_pretrained(model_path, torch_dtype=torch.float32).to("cpu")
if config:
self.config = config
return self
def save_model(self, path: str):
os.makedirs(os.path.dirname(path), exist_ok=True)
self.pipeline.save_pretrained(path)
def generate(self, prompt: str):
return self.pipeline(prompt, num_inference_steps=20).images[0]
def generate_filename(sequence, ext="png"):
timestamp = time.strftime("%d%m%Y%H%M%S")
return f"{sequence}_{timestamp}.{ext}"
def get_gallery_files(file_types):
return sorted(list(set([f for ext in file_types for f in glob.glob(f"*.{ext}")]))) # Deduplicate files
async def process_image_gen(prompt, output_file, builder):
if builder and isinstance(builder, DiffusionBuilder) and builder.pipeline:
pipeline = builder.pipeline
else:
pipeline = StableDiffusionPipeline.from_pretrained("OFA-Sys/small-stable-diffusion-v0", torch_dtype=torch.float32).to("cpu")
gen_image = pipeline(prompt, num_inference_steps=20).images[0]
gen_image.save(output_file)
return gen_image
# Smart Uploader Functions
def upload_files(files, links_title, links_url, history, selected_files):
uploaded = {"images": [], "videos": [], "documents": [], "datasets": [], "links": []}
if files:
for file in files:
ext = file.name.split('.')[-1].lower()
output_path = f"uploaded_{int(time.time())}_{file.name}"
with open(output_path, "wb") as f:
f.write(file.read())
if ext in ["jpg", "png"]:
uploaded["images"].append(output_path)
elif ext == "mp4":
uploaded["videos"].append(output_path)
elif ext in ["md", "pdf", "docx"]:
uploaded["documents"].append(output_path)
elif ext in ["csv", "xlsx"]:
uploaded["datasets"].append(output_path)
history.append(f"Uploaded: {output_path}")
selected_files[output_path] = False # Default unchecked
if links_title and links_url:
links = list(zip(links_title.split('\n'), links_url.split('\n')))
for title, url in links:
if title and url:
link_entry = f"[{title}]({url})"
uploaded["links"].append(link_entry)
history.append(f"Added Link: {link_entry}")
selected_files[link_entry] = False
return uploaded, history, selected_files
def update_galleries(history, selected_files):
galleries = {
"images": get_gallery_files(["jpg", "png"]),
"videos": get_gallery_files(["mp4"]),
"documents": get_gallery_files(["md", "pdf", "docx"]),
"datasets": get_gallery_files(["csv", "xlsx"]),
"links": [f for f in selected_files.keys() if f.startswith('[') and '](' in f and f.endswith(')')]
}
gallery_outputs = {
"images": [(Image.open(f), os.path.basename(f)) for f in galleries["images"][:4]],
"videos": [(f, os.path.basename(f)) for f in galleries["videos"][:4]], # Video preview as file path
"documents": [(Image.frombytes("RGB", fitz.open(f)[0].get_pixmap(matrix=fitz.Matrix(0.5, 0.5)).size, fitz.open(f)[0].get_pixmap(matrix=fitz.Matrix(0.5, 0.5)).samples) if f.endswith('.pdf') else f, os.path.basename(f)) for f in galleries["documents"][:4]],
"datasets": [(f, os.path.basename(f)) for f in galleries["datasets"][:4]], # Text preview
"links": [(f, f.split(']')[0][1:]) for f in galleries["links"][:4]]
}
history.append(f"Updated galleries: {sum(len(g) for g in galleries.values())} files")
return gallery_outputs, history, selected_files
def toggle_selection(file_list, selected_files):
for file in file_list:
selected_files[file] = not selected_files.get(file, False)
return selected_files
def image_gen(prompt, builder, history, selected_files):
selected = [f for f, sel in selected_files.items() if sel and f.endswith(('.jpg', '.png'))]
if not selected:
return "No images selected", None, history, selected_files
output_file = generate_filename("gen_output", "png")
gen_image = asyncio.run(process_image_gen(prompt, output_file, builder))
history.append(f"Image Gen: {prompt} -> {output_file}")
selected_files[output_file] = True
return f"Image saved to {output_file}", gen_image, history, selected_files
# Gradio UI
with gr.Blocks(title="AI Vision & SFT Titans π") as demo:
gr.Markdown("# AI Vision & SFT Titans π")
history = gr.State(value=[])
builder = gr.State(value=None)
selected_files = gr.State(value={})
with gr.Row():
with gr.Column(scale=1):
gr.Markdown("## π File Tree")
with gr.Accordion("π³ Uploads", open=True):
with gr.Row():
gr.Markdown("### πΌοΈ Images (jpg/png)")
img_gallery = gr.Gallery(label="Images", columns=4, height="auto")
with gr.Row():
gr.Markdown("### π₯ Videos (mp4)")
vid_gallery = gr.Gallery(label="Videos", columns=4, height="auto")
with gr.Row():
gr.Markdown("### π Docs (md/pdf/docx)")
doc_gallery = gr.Gallery(label="Documents", columns=4, height="auto")
with gr.Row():
gr.Markdown("### π Data (csv/xlsx)")
data_gallery = gr.Gallery(label="Datasets", columns=4, height="auto")
with gr.Row():
gr.Markdown("### π Links")
link_gallery = gr.Gallery(label="Links", columns=4, height="auto")
gr.Markdown("## π History")
history_output = gr.Textbox(label="Log", lines=5, interactive=False)
with gr.Column(scale=3):
with gr.Row():
gr.Markdown("## π οΈ Toolbar")
upload_btn = gr.Button("π€ Upload")
select_btn = gr.Button("β
Select")
gen_btn = gr.Button("π¨ Generate")
with gr.Tabs():
with gr.TabItem("π€ Smart Upload"):
file_upload = gr.File(label="Upload Files", file_count="multiple", type="binary")
links_title = gr.Textbox(label="Link Titles (one per line)", lines=3)
links_url = gr.Textbox(label="Link URLs (one per line)", lines=3)
upload_status = gr.Textbox(label="Status")
with gr.TabItem("π Operations"):
prompt = gr.Textbox(label="Image Gen Prompt", value="Generate a neon version")
op_status = gr.Textbox(label="Status")
op_output = gr.Image(label="Output")
upload_btn.click(
upload_files,
inputs=[file_upload, links_title, links_url, history, selected_files],
outputs=[upload_status, history, selected_files]
).then(
update_galleries,
inputs=[history, selected_files],
outputs=[img_gallery, vid_gallery, doc_gallery, data_gallery, link_gallery, history, selected_files]
)
select_btn.click(
toggle_selection,
inputs=[gr.Dropdown(choices=list(selected_files.value.keys()), multiselect=True, label="Select Files"), selected_files],
outputs=[selected_files]
).then(
update_galleries,
inputs=[history, selected_files],
outputs=[img_gallery, vid_gallery, doc_gallery, data_gallery, link_gallery, history, selected_files]
)
gen_btn.click(
image_gen,
inputs=[prompt, builder, history, selected_files],
outputs=[op_status, op_output, history, selected_files]
).then(
update_galleries,
inputs=[history, selected_files],
outputs=[img_gallery, vid_gallery, doc_gallery, data_gallery, link_gallery, history, selected_files]
)
# Update history output
demo.load(lambda h: "\n".join(h[-5:]), inputs=[history], outputs=[history_output])
demo.launch() |