File size: 12,463 Bytes
62c110b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# Copyright 2024 UC Berkeley Team and The HuggingFace Team. All rights reserved.
#
# 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.

# DISCLAIMER: This file is strongly influenced by https://github.com/ermongroup/ddim

from dataclasses import dataclass
from typing import Optional, Tuple, Union

import flax
import jax
import jax.numpy as jnp

from ..configuration_utils import ConfigMixin, register_to_config
from .scheduling_utils_flax import (
    CommonSchedulerState,
    FlaxKarrasDiffusionSchedulers,
    FlaxSchedulerMixin,
    FlaxSchedulerOutput,
    add_noise_common,
    get_velocity_common,
)


@flax.struct.dataclass
class DDPMSchedulerState:
    common: CommonSchedulerState

    # setable values
    init_noise_sigma: jnp.ndarray
    timesteps: jnp.ndarray
    num_inference_steps: Optional[int] = None

    @classmethod
    def create(cls, common: CommonSchedulerState, init_noise_sigma: jnp.ndarray, timesteps: jnp.ndarray):
        return cls(common=common, init_noise_sigma=init_noise_sigma, timesteps=timesteps)


@dataclass
class FlaxDDPMSchedulerOutput(FlaxSchedulerOutput):
    state: DDPMSchedulerState


class FlaxDDPMScheduler(FlaxSchedulerMixin, ConfigMixin):
    """
    Denoising diffusion probabilistic models (DDPMs) explores the connections between denoising score matching and
    Langevin dynamics sampling.

    [`~ConfigMixin`] takes care of storing all config attributes that are passed in the scheduler's `__init__`
    function, such as `num_train_timesteps`. They can be accessed via `scheduler.config.num_train_timesteps`.
    [`SchedulerMixin`] provides general loading and saving functionality via the [`SchedulerMixin.save_pretrained`] and
    [`~SchedulerMixin.from_pretrained`] functions.

    For more details, see the original paper: https://arxiv.org/abs/2006.11239

    Args:
        num_train_timesteps (`int`): number of diffusion steps used to train the model.
        beta_start (`float`): the starting `beta` value of inference.
        beta_end (`float`): the final `beta` value.
        beta_schedule (`str`):
            the beta schedule, a mapping from a beta range to a sequence of betas for stepping the model. Choose from
            `linear`, `scaled_linear`, or `squaredcos_cap_v2`.
        trained_betas (`np.ndarray`, optional):
            option to pass an array of betas directly to the constructor to bypass `beta_start`, `beta_end` etc.
        variance_type (`str`):
            options to clip the variance used when adding noise to the denoised sample. Choose from `fixed_small`,
            `fixed_small_log`, `fixed_large`, `fixed_large_log`, `learned` or `learned_range`.
        clip_sample (`bool`, default `True`):
            option to clip predicted sample between -1 and 1 for numerical stability.
        prediction_type (`str`, default `epsilon`):
            indicates whether the model predicts the noise (epsilon), or the samples. One of `epsilon`, `sample`.
            `v-prediction` is not supported for this scheduler.
        dtype (`jnp.dtype`, *optional*, defaults to `jnp.float32`):
            the `dtype` used for params and computation.
    """

    _compatibles = [e.name for e in FlaxKarrasDiffusionSchedulers]

    dtype: jnp.dtype

    @property
    def has_state(self):
        return True

    @register_to_config
    def __init__(
        self,
        num_train_timesteps: int = 1000,
        beta_start: float = 0.0001,
        beta_end: float = 0.02,
        beta_schedule: str = "linear",
        trained_betas: Optional[jnp.ndarray] = None,
        variance_type: str = "fixed_small",
        clip_sample: bool = True,
        prediction_type: str = "epsilon",
        dtype: jnp.dtype = jnp.float32,
    ):
        self.dtype = dtype

    def create_state(self, common: Optional[CommonSchedulerState] = None) -> DDPMSchedulerState:
        if common is None:
            common = CommonSchedulerState.create(self)

        # standard deviation of the initial noise distribution
        init_noise_sigma = jnp.array(1.0, dtype=self.dtype)

        timesteps = jnp.arange(0, self.config.num_train_timesteps).round()[::-1]

        return DDPMSchedulerState.create(
            common=common,
            init_noise_sigma=init_noise_sigma,
            timesteps=timesteps,
        )

    def scale_model_input(
        self, state: DDPMSchedulerState, sample: jnp.ndarray, timestep: Optional[int] = None
    ) -> jnp.ndarray:
        """
        Args:
            state (`PNDMSchedulerState`): the `FlaxPNDMScheduler` state data class instance.
            sample (`jnp.ndarray`): input sample
            timestep (`int`, optional): current timestep

        Returns:
            `jnp.ndarray`: scaled input sample
        """
        return sample

    def set_timesteps(
        self, state: DDPMSchedulerState, num_inference_steps: int, shape: Tuple = ()
    ) -> DDPMSchedulerState:
        """
        Sets the discrete timesteps used for the diffusion chain. Supporting function to be run before inference.

        Args:
            state (`DDIMSchedulerState`):
                the `FlaxDDPMScheduler` state data class instance.
            num_inference_steps (`int`):
                the number of diffusion steps used when generating samples with a pre-trained model.
        """

        step_ratio = self.config.num_train_timesteps // num_inference_steps
        # creates integer timesteps by multiplying by ratio
        # rounding to avoid issues when num_inference_step is power of 3
        timesteps = (jnp.arange(0, num_inference_steps) * step_ratio).round()[::-1]

        return state.replace(
            num_inference_steps=num_inference_steps,
            timesteps=timesteps,
        )

    def _get_variance(self, state: DDPMSchedulerState, t, predicted_variance=None, variance_type=None):
        alpha_prod_t = state.common.alphas_cumprod[t]
        alpha_prod_t_prev = jnp.where(t > 0, state.common.alphas_cumprod[t - 1], jnp.array(1.0, dtype=self.dtype))

        # For t > 0, compute predicted variance βt (see formula (6) and (7) from https://arxiv.org/pdf/2006.11239.pdf)
        # and sample from it to get previous sample
        # x_{t-1} ~ N(pred_prev_sample, variance) == add variance to pred_sample
        variance = (1 - alpha_prod_t_prev) / (1 - alpha_prod_t) * state.common.betas[t]

        if variance_type is None:
            variance_type = self.config.variance_type

        # hacks - were probably added for training stability
        if variance_type == "fixed_small":
            variance = jnp.clip(variance, a_min=1e-20)
        # for rl-diffuser https://arxiv.org/abs/2205.09991
        elif variance_type == "fixed_small_log":
            variance = jnp.log(jnp.clip(variance, a_min=1e-20))
        elif variance_type == "fixed_large":
            variance = state.common.betas[t]
        elif variance_type == "fixed_large_log":
            # Glide max_log
            variance = jnp.log(state.common.betas[t])
        elif variance_type == "learned":
            return predicted_variance
        elif variance_type == "learned_range":
            min_log = variance
            max_log = state.common.betas[t]
            frac = (predicted_variance + 1) / 2
            variance = frac * max_log + (1 - frac) * min_log

        return variance

    def step(
        self,
        state: DDPMSchedulerState,
        model_output: jnp.ndarray,
        timestep: int,
        sample: jnp.ndarray,
        key: Optional[jax.Array] = None,
        return_dict: bool = True,
    ) -> Union[FlaxDDPMSchedulerOutput, Tuple]:
        """
        Predict the sample at the previous timestep by reversing the SDE. Core function to propagate the diffusion
        process from the learned model outputs (most often the predicted noise).

        Args:
            state (`DDPMSchedulerState`): the `FlaxDDPMScheduler` state data class instance.
            model_output (`jnp.ndarray`): direct output from learned diffusion model.
            timestep (`int`): current discrete timestep in the diffusion chain.
            sample (`jnp.ndarray`):
                current instance of sample being created by diffusion process.
            key (`jax.Array`): a PRNG key.
            return_dict (`bool`): option for returning tuple rather than FlaxDDPMSchedulerOutput class

        Returns:
            [`FlaxDDPMSchedulerOutput`] or `tuple`: [`FlaxDDPMSchedulerOutput`] if `return_dict` is True, otherwise a
            `tuple`. When returning a tuple, the first element is the sample tensor.

        """
        t = timestep

        if key is None:
            key = jax.random.PRNGKey(0)

        if model_output.shape[1] == sample.shape[1] * 2 and self.config.variance_type in ["learned", "learned_range"]:
            model_output, predicted_variance = jnp.split(model_output, sample.shape[1], axis=1)
        else:
            predicted_variance = None

        # 1. compute alphas, betas
        alpha_prod_t = state.common.alphas_cumprod[t]
        alpha_prod_t_prev = jnp.where(t > 0, state.common.alphas_cumprod[t - 1], jnp.array(1.0, dtype=self.dtype))
        beta_prod_t = 1 - alpha_prod_t
        beta_prod_t_prev = 1 - alpha_prod_t_prev

        # 2. compute predicted original sample from predicted noise also called
        # "predicted x_0" of formula (15) from https://arxiv.org/pdf/2006.11239.pdf
        if self.config.prediction_type == "epsilon":
            pred_original_sample = (sample - beta_prod_t ** (0.5) * model_output) / alpha_prod_t ** (0.5)
        elif self.config.prediction_type == "sample":
            pred_original_sample = model_output
        elif self.config.prediction_type == "v_prediction":
            pred_original_sample = (alpha_prod_t**0.5) * sample - (beta_prod_t**0.5) * model_output
        else:
            raise ValueError(
                f"prediction_type given as {self.config.prediction_type} must be one of `epsilon`, `sample` "
                " for the FlaxDDPMScheduler."
            )

        # 3. Clip "predicted x_0"
        if self.config.clip_sample:
            pred_original_sample = jnp.clip(pred_original_sample, -1, 1)

        # 4. Compute coefficients for pred_original_sample x_0 and current sample x_t
        # See formula (7) from https://arxiv.org/pdf/2006.11239.pdf
        pred_original_sample_coeff = (alpha_prod_t_prev ** (0.5) * state.common.betas[t]) / beta_prod_t
        current_sample_coeff = state.common.alphas[t] ** (0.5) * beta_prod_t_prev / beta_prod_t

        # 5. Compute predicted previous sample µ_t
        # See formula (7) from https://arxiv.org/pdf/2006.11239.pdf
        pred_prev_sample = pred_original_sample_coeff * pred_original_sample + current_sample_coeff * sample

        # 6. Add noise
        def random_variance():
            split_key = jax.random.split(key, num=1)
            noise = jax.random.normal(split_key, shape=model_output.shape, dtype=self.dtype)
            return (self._get_variance(state, t, predicted_variance=predicted_variance) ** 0.5) * noise

        variance = jnp.where(t > 0, random_variance(), jnp.zeros(model_output.shape, dtype=self.dtype))

        pred_prev_sample = pred_prev_sample + variance

        if not return_dict:
            return (pred_prev_sample, state)

        return FlaxDDPMSchedulerOutput(prev_sample=pred_prev_sample, state=state)

    def add_noise(
        self,
        state: DDPMSchedulerState,
        original_samples: jnp.ndarray,
        noise: jnp.ndarray,
        timesteps: jnp.ndarray,
    ) -> jnp.ndarray:
        return add_noise_common(state.common, original_samples, noise, timesteps)

    def get_velocity(
        self,
        state: DDPMSchedulerState,
        sample: jnp.ndarray,
        noise: jnp.ndarray,
        timesteps: jnp.ndarray,
    ) -> jnp.ndarray:
        return get_velocity_common(state.common, sample, noise, timesteps)

    def __len__(self):
        return self.config.num_train_timesteps