File size: 15,920 Bytes
00d591a
1312362
6e62cd2
915f4ae
 
00d591a
915f4ae
 
 
 
 
00d591a
915f4ae
 
 
00d591a
618d32d
 
00d591a
 
 
915f4ae
00d591a
 
 
915f4ae
618d32d
00d591a
 
618d32d
 
 
00d591a
618d32d
 
 
 
 
 
 
 
 
 
 
 
 
 
00d591a
 
915f4ae
 
 
 
 
 
 
 
 
618d32d
 
 
 
 
 
915f4ae
618d32d
 
 
 
 
 
915f4ae
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
618d32d
 
 
 
 
 
 
 
 
 
 
00d591a
618d32d
 
 
00d591a
915f4ae
618d32d
 
915f4ae
618d32d
 
 
00d591a
915f4ae
 
 
 
 
618d32d
 
 
 
 
 
 
 
 
 
 
 
915f4ae
 
 
 
 
 
 
 
 
 
 
 
 
00d591a
915f4ae
 
00d591a
618d32d
 
 
00d591a
 
 
 
 
 
 
618d32d
00d591a
 
 
 
 
915f4ae
 
 
 
 
 
 
 
 
 
618d32d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
00d591a
618d32d
 
 
915f4ae
00d591a
 
915f4ae
618d32d
 
915f4ae
 
00d591a
618d32d
 
 
915f4ae
 
618d32d
 
 
 
 
 
 
 
 
915f4ae
 
 
618d32d
 
915f4ae
 
 
 
 
 
 
 
618d32d
 
 
 
 
 
 
 
 
 
 
 
915f4ae
00d591a
618d32d
00d591a
 
 
618d32d
00d591a
 
618d32d
 
 
 
 
915f4ae
00d591a
 
 
915f4ae
618d32d
915f4ae
00d591a
 
 
 
618d32d
 
 
 
 
 
 
915f4ae
00d591a
915f4ae
618d32d
 
 
00d591a
 
618d32d
 
 
00d591a
 
 
 
 
 
 
 
 
 
915f4ae
 
 
00d591a
 
 
 
915f4ae
 
00d591a
618d32d
915f4ae
618d32d
 
 
 
915f4ae
 
00d591a
618d32d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
915f4ae
 
 
618d32d
 
1312362
e11ace5
915f4ae
618d32d
 
 
 
 
 
 
 
 
 
915f4ae
 
00d591a
618d32d
 
9e77c17
915f4ae
00d591a
 
 
8eea048
 
 
00d591a
915f4ae
00d591a
915f4ae
618d32d
 
 
915f4ae
00d591a
618d32d
00d591a
618d32d
 
 
 
 
 
00d591a
 
 
 
 
618d32d
 
 
 
 
 
 
 
 
 
00d591a
 
 
915f4ae
 
00d591a
915f4ae
00d591a
6e62cd2
e11ace5
618d32d
 
 
 
 
 
 
 
e11ace5
00d591a
e11ace5
915f4ae
618d32d
 
 
 
 
 
 
9f48eda
00d591a
9f48eda
618d32d
 
 
 
 
9f48eda
915f4ae
b03fd98
618d32d
 
 
 
 
915f4ae
 
618d32d
 
 
 
 
 
 
 
 
 
00d591a
618d32d
 
 
 
 
915f4ae
618d32d
00d591a
915f4ae
00d591a
618d32d
00d591a
618d32d
915f4ae
00d591a
915f4ae
618d32d
 
 
 
 
 
 
 
 
 
 
 
 
915f4ae
 
 
 
00d591a
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
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
import os
import spaces
import time
import gradio as gr
import torch
from torch import Tensor, nn
from PIL import Image
from torchvision import transforms
from dataclasses import dataclass
import math
from typing import Callable
import random
from tqdm import tqdm
import bitsandbytes as bnb
from bitsandbytes.nn.modules import Params4bit, QuantState
from transformers import (
    MarianTokenizer,
    MarianMTModel,
    CLIPTextModel, CLIPTokenizer,
    T5EncoderModel, T5Tokenizer
)
from diffusers import AutoencoderKL
from huggingface_hub import hf_hub_download
from safetensors.torch import load_file
from einops import rearrange, repeat

# 1) ์žฅ์น˜ ์„ค์ •
torch_device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# 2) ๋ฒˆ์—ญ ๋ชจ๋ธ์„ CPU์—์„œ, ๋ฐ˜๋“œ์‹œ PyTorch ์ฒดํฌํฌ์ธํŠธ๋กœ ๋กœ๋“œ
trans_tokenizer = MarianTokenizer.from_pretrained(
    "Helsinki-NLP/opus-mt-ko-en"
)
trans_model = MarianMTModel.from_pretrained(
    "Helsinki-NLP/opus-mt-ko-en",
    from_tf=True,         # TF ์ฒดํฌํฌ์ธํŠธ๋ผ๋„ PyTorch ๋กœ๋“œ
    torch_dtype=torch.float32,
).to(torch.device("cpu"))

def translate_ko_to_en(text: str, max_length: int = 512) -> str:
    """ํ•œ๊ธ€ โ†’ ์˜์–ด ๋ฒˆ์—ญ (CPU)"""
    batch = trans_tokenizer([text], return_tensors="pt", padding=True)
    # ๋ชจ๋ธ์€ CPU์— ์žˆ์œผ๋ฏ€๋กœ .to("cpu") ํ•ด์ค„ ํ•„์š” ์—†์Œ
    gen = trans_model.generate(
        **batch, max_length=max_length
    )
    return trans_tokenizer.batch_decode(gen, skip_special_tokens=True)[0]

# ---------------- Encoders ----------------

class HFEmbedder(nn.Module):
    def __init__(self, version: str, max_length: int, **hf_kwargs):
        super().__init__()
        self.is_clip = version.startswith("openai")
        self.max_length = max_length
        self.output_key = "pooler_output" if self.is_clip else "last_hidden_state"

        if self.is_clip:
            self.tokenizer: CLIPTokenizer = CLIPTokenizer.from_pretrained(
                version, max_length=max_length
            )
            self.hf_module: CLIPTextModel = CLIPTextModel.from_pretrained(
                version, **hf_kwargs
            )
        else:
            self.tokenizer: T5Tokenizer = T5Tokenizer.from_pretrained(
                version, max_length=max_length
            )
            self.hf_module: T5EncoderModel = T5EncoderModel.from_pretrained(
                version, **hf_kwargs
            )

        self.hf_module = self.hf_module.eval().requires_grad_(False)

    def forward(self, text: list[str]) -> Tensor:
        batch_encoding = self.tokenizer(
            text,
            truncation=True,
            max_length=self.max_length,
            padding="max_length",
            return_tensors="pt",
        )
        outputs = self.hf_module(
            input_ids=batch_encoding["input_ids"].to(self.hf_module.device),
            attention_mask=None,
            output_hidden_states=False,
        )
        return outputs[self.output_key]

# T5, CLIP, VAE ๋ชจ๋‘ GPU/CPU(device)๋กœ ์ด๋™
t5   = HFEmbedder(
    "DeepFloyd/t5-v1_1-xxl",
    max_length=512,
    torch_dtype=torch.bfloat16
).to(torch_device)
clip = HFEmbedder(
    "openai/clip-vit-large-patch14",
    max_length=77,
    torch_dtype=torch.bfloat16
).to(torch_device)
ae   = AutoencoderKL.from_pretrained(
    "black-forest-labs/FLUX.1-dev",
    subfolder="vae",
    torch_dtype=torch.bfloat16
).to(torch_device)

# ---------------- NF4 ์ง€์› ์ฝ”๋“œ ----------------

def functional_linear_4bits(x, weight, bias):
    out = bnb.matmul_4bit(
        x, weight.t(), bias=bias, quant_state=weight.quant_state
    )
    return out.to(x)

def copy_quant_state(state: QuantState, device: torch.device = None) -> QuantState:
    if state is None:
        return None
    device = device or state.absmax.device
    state2 = (
        QuantState(
            absmax=state.state2.absmax.to(device),
            shape=state.state2.shape,
            code=state.state2.code.to(device),
            blocksize=state.state2.blocksize,
            quant_type=state.state2.quant_type,
            dtype=state.state2.dtype,
        )
        if state.nested
        else None
    )
    return QuantState(
        absmax=state.absmax.to(device),
        shape=state.shape,
        code=state.code.to(device),
        blocksize=state.blocksize,
        quant_type=state.quant_type,
        dtype=state.dtype,
        offset=state.offset.to(device) if state.nested else None,
        state2=state2,
    )

class ForgeParams4bit(Params4bit):
    def to(self, *args, **kwargs):
        device, dtype, non_blocking, _ = torch._C._nn._parse_to(*args, **kwargs)
        if device is not None and device.type == "cuda" and not self.bnb_quantized:
            return self._quantize(device)
        new = ForgeParams4bit(
            torch.nn.Parameter.to(
                self, device=device, dtype=dtype, non_blocking=non_blocking
            ),
            requires_grad=self.requires_grad,
            quant_state=copy_quant_state(self.quant_state, device),
            compress_statistics=False,
            blocksize=self.blocksize,
            quant_type=self.quant_type,
            quant_storage=self.quant_storage,
            bnb_quantized=self.bnb_quantized,
            module=self.module,
        )
        self.module.quant_state = new.quant_state
        self.data = new.data
        self.quant_state = new.quant_state
        return new

class ForgeLoader4Bit(torch.nn.Module):
    def __init__(self, *, device, dtype, quant_type, **kwargs):
        super().__init__()
        self.dummy = torch.nn.Parameter(torch.empty(1, device=device, dtype=dtype))
        self.weight = None
        self.quant_state = None
        self.bias = None
        self.quant_type = quant_type

    def _load_from_state_dict(
        self,
        state_dict,
        prefix,
        local_metadata,
        strict,
        missing_keys,
        unexpected_keys,
        error_msgs,
    ):
        qs_keys = {
            k[len(prefix + "weight.") :]
            for k in state_dict
            if k.startswith(prefix + "weight.")
        }
        if any("bitsandbytes" in k for k in qs_keys):
            qs = {
                k: state_dict[prefix + "weight." + k] for k in qs_keys
            }
            self.weight = ForgeParams4bit.from_prequantized(
                data=state_dict[prefix + "weight"],
                quantized_stats=qs,
                requires_grad=False,
                device=torch.device("cuda"),
                module=self,
            )
            self.quant_state = self.weight.quant_state
            if prefix + "bias" in state_dict:
                self.bias = torch.nn.Parameter(
                    state_dict[prefix + "bias"].to(self.dummy)
                )
            del self.dummy
        else:
            super()._load_from_state_dict(
                state_dict,
                prefix,
                local_metadata,
                strict,
                missing_keys,
                unexpected_keys,
                error_msgs,
            )

class Linear(ForgeLoader4Bit):
    def __init__(self, *args, device=None, dtype=None, **kwargs):
        super().__init__(device=device, dtype=dtype, quant_type="nf4")

    def forward(self, x):
        self.weight.quant_state = self.quant_state
        if self.bias is not None and self.bias.dtype != x.dtype:
            self.bias.data = self.bias.data.to(x.dtype)
        return functional_linear_4bits(x, self.weight, self.bias)

nn.Linear = Linear

# ---------------- Flux ๋ชจ๋ธ ์ •์˜ (์›๋ณธ ๊ทธ๋Œ€๋กœ) ----------------

def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor) -> Tensor:
    # ... (์ƒ๋žต ์—†์ด ์›๋ณธ ์ฝ”๋“œ ๊ทธ๋Œ€๋กœ)
    q, k = apply_rope(q, k, pe)
    x = torch.nn.functional.scaled_dot_product_attention(q, k, v)
    x = x.permute(0, 2, 1, 3).reshape(x.size(0), x.size(2), -1)
    return x

# apply_rope, rope, EmbedND, timestep_embedding, MLPEmbedder, RMSNorm, QKNorm,
# SelfAttention, Modulation, DoubleStreamBlock, SingleStreamBlock,
# LastLayer, FluxParams, Flux ํด๋ž˜์Šค๊นŒ์ง€ ์ „๋ถ€ ์›๋ณธ๊ณผ ๋™์ผํ•˜๊ฒŒ ํฌํ•จํ•˜์„ธ์š”.

# ---------------- ๋ชจ๋ธ ๋กœ๋“œ ----------------

sd = load_file(
    hf_hub_download(
        repo_id="lllyasviel/flux1-dev-bnb-nf4",
        filename="flux1-dev-bnb-nf4-v2.safetensors",
    )
)
sd = {
    k.replace("model.diffusion_model.", ""): v
    for k, v in sd.items()
    if "model.diffusion_model" in k
}

model = Flux().to(torch_device, dtype=torch.bfloat16)
model.load_state_dict(sd)
model_zero_init = False

# ---------------- ์œ ํ‹ธ๋ฆฌํ‹ฐ ํ•จ์ˆ˜ ----------------

def get_image(image) -> torch.Tensor | None:
    if image is None:
        return None
    image = Image.fromarray(image).convert("RGB")
    tfm = transforms.Compose(
        [
            transforms.ToTensor(),
            transforms.Lambda(lambda x: 2.0 * x - 1.0),
        ]
    )
    return tfm(image)[None, ...]

def prepare(t5, clip, img, prompt):
    bs, c, h, w = img.shape
    img = rearrange(
        img, "b c (h ph) (w pw) -> b (h w) (c ph pw)", ph=2, pw=2
    )
    if bs == 1 and isinstance(prompt, list):
        img = repeat(img, "1 ... -> bs ...", bs=len(prompt))
    img_ids = torch.zeros(h // 2, w // 2, 3, device=img.device)
    img_ids[..., 1] = torch.arange(h // 2, device=img.device)[:, None]
    img_ids[..., 2] = torch.arange(w // 2, device=img.device)[None, :]
    img_ids = repeat(img_ids, "h w c -> b (h w) c", b=img.shape[0])

    txt = t5([prompt] if isinstance(prompt, str) else prompt)
    if txt.shape[0] == 1 and img.shape[0] > 1:
        txt = repeat(txt, "1 ... -> bs ...", bs=img.shape[0])
    txt_ids = torch.zeros(txt.size(0), txt.size(1), 3, device=img.device)

    vec = clip([prompt] if isinstance(prompt, str) else prompt)
    if vec.shape[0] == 1 and img.shape[0] > 1:
        vec = repeat(vec, "1 ... -> bs ...", bs=img.shape[0])

    return {
        "img": img,
        "img_ids": img_ids,
        "txt": txt,
        "txt_ids": txt_ids,
        "vec": vec,
    }

def get_schedule(num_steps, image_seq_len, base_shift=0.5, max_shift=1.15, shift=True):
    timesteps = torch.linspace(1, 0, num_steps + 1)
    if shift:
        mu = ((max_shift - base_shift) / (4096 - 256)) * image_seq_len + (
            base_shift - (256 * (max_shift - base_shift) / (4096 - 256))
        )
        timesteps = timesteps.exp().div((1 / timesteps - 1) ** 1 + mu)
    return timesteps.tolist()

def denoise(model, img, img_ids, txt, txt_ids, vec, timesteps, guidance):
    guidance_vec = torch.full(
        (img.size(0),), guidance, device=img.device, dtype=img.dtype
    )
    for t_curr, t_prev in tqdm(
        zip(timesteps[:-1], timesteps[1:]), total=len(timesteps) - 1
    ):
        t_vec = torch.full(
            (img.size(0),), t_curr, device=img.device, dtype=img.dtype
        )
        pred = model(
            img=img,
            img_ids=img_ids,
            txt=txt,
            txt_ids=txt_ids,
            y=vec,
            timesteps=t_vec,
            guidance=guidance_vec,
        )
        img = img + (t_prev - t_curr) * pred
    return img

# ---------------- Gradio ๋ฐ๋ชจ ----------------

@spaces.GPU
@torch.no_grad()
def generate_image(
    prompt,
    width,
    height,
    guidance,
    inference_steps,
    seed,
    do_img2img,
    init_image,
    image2image_strength,
    resize_img,
    progress=gr.Progress(track_tqdm=True),
):
    # ํ•œ๊ธ€ ๊ฐ์ง€ ์‹œ CPU ๋ฒˆ์—ญ๊ธฐ ์‚ฌ์šฉ
    if any("\u3131" <= c <= "\u318E" or "\uAC00" <= c <= "\uD7A3" for c in prompt):
        prompt = translate_ko_to_en(prompt)

    if seed == 0:
        seed = random.randint(1, 1_000_000)

    global model_zero_init, model
    if not model_zero_init:
        model = model.to(torch_device)
        model_zero_init = True

    if do_img2img and init_image is not None:
        init_img = get_image(init_image)
        if resize_img:
            init_img = torch.nn.functional.interpolate(
                init_img, (height, width)
            )
        else:
            h0, w0 = init_img.shape[-2:]
            init_img = init_img[..., : 16 * (h0 // 16), : 16 * (w0 // 16)]
            height, width = init_img.shape[-2:]
        init_img = ae.encode(
            init_img.to(torch_device).to(torch.bfloat16)
        ).latent_dist.sample()
        init_img = (
            init_img - ae.config.shift_factor
        ) * ae.config.scaling_factor
    else:
        init_img = None

    generator = torch.Generator(device=str(torch_device)).manual_seed(seed)
    x = torch.randn(
        1,
        16,
        2 * math.ceil(height / 16),
        2 * math.ceil(width / 16),
        device=torch_device,
        dtype=torch.bfloat16,
        generator=generator,
    )
    timesteps = get_schedule(
        inference_steps, (x.shape[-1] * x.shape[-2]) // 4, shift=True
    )
    if do_img2img and init_img is not None:
        t_idx = int((1 - image2image_strength) * inference_steps)
        t = timesteps[t_idx]
        timesteps = timesteps[t_idx:]
        x = t * x + (1 - t) * init_img.to(x.dtype)

    inp = prepare(t5, clip, x, prompt)
    x = denoise(model, **inp, timesteps=timesteps, guidance=guidance)

    x = rearrange(
        x[:, inp["txt"].shape[1] :, ...].float(),
        "b (h w) (c ph pw) -> b c (h ph) (w pw)",
        h=math.ceil(height / 16),
        w=math.ceil(width / 16),
        ph=2,
        pw=2,
    )
    with torch.autocast(device_type=torch_device.type, dtype=torch.bfloat16):
        x = (x / ae.config.scaling_factor) + ae.config.shift_factor
        x = ae.decode(x).sample

    x = x.clamp(-1, 1)
    img = Image.fromarray(
        (127.5 * (rearrange(x[0], "c h w -> h w c") + 1.0))
        .cpu()
        .byte()
        .numpy()
    )

    return img, seed

css = """
footer {
    visibility: hidden;
}
"""

def create_demo():
    with gr.Blocks(theme="Yntec/HaleyCH_Theme_Orange", css=css) as demo:
        gr.Markdown(
            "# News! Multilingual version "
            "[https://huggingface.co/spaces/ginigen/FLUXllama-Multilingual]"
            "(https://huggingface.co/spaces/ginigen/FLUXllama-Multilingual)"
        )
        with gr.Row():
            with gr.Column():
                prompt = gr.Textbox(
                    label="Prompt(ํ•œ๊ธ€ ๊ฐ€๋Šฅ)",
                    value="A cute and fluffy golden retriever puppy sitting upright...",
                )
                width = gr.Slider(128, 2048, 64, label="Width", value=768)
                height = gr.Slider(128, 2048, 64, label="Height", value=768)
                guidance = gr.Slider(1.0, 5.0, 0.1, label="Guidance", value=3.5)
                steps = gr.Slider(1, 30, 1, label="Inference steps", value=30)
                seed = gr.Number(label="Seed", precision=0)
                do_i2i = gr.Checkbox(label="Image to Image", value=False)
                init_img = gr.Image(label="Input Image", visible=False)
                strength = gr.Slider(
                    0.0, 1.0, 0.01, label="Noising strength", value=0.8, visible=False
                )
                resize = gr.Checkbox(label="Resize image", value=True, visible=False)
                btn = gr.Button("Generate")
            with gr.Column():
                out_img = gr.Image(label="Generated Image")
                out_seed = gr.Text(label="Used Seed")

        do_i2i.change(
            fn=lambda x: [gr.update(visible=x)] * 3,
            inputs=[do_i2i],
            outputs=[init_img, strength, resize],
        )
        btn.click(
            fn=generate_image,
            inputs=[
                prompt,
                width,
                height,
                guidance,
                steps,
                seed,
                do_i2i,
                init_img,
                strength,
                resize,
            ],
            outputs=[out_img, out_seed],
        )
    return demo

if __name__ == "__main__":
    create_demo().launch()