feature: support loading sdxl compvis weights (#449)

docs/changelog.md

@@ -1,7 +1,27 @@
 
 ## ChangeLog
 
-**14.0.0**
+**14.1.0**
+- 🎉 feature: make video generation smooth by adding frame interpolation
+- feature: SDXL weights in the compvis format can now be used
+- feature: allow video generation at any size specified by user
+- feature: video generations output in "bounce" format
+- feature: choose video output format: mp4, webp, or gif
+- feature: fix random seed handling in video generation
+- docs: auto-publish docs on push to master
+- build: remove imageio dependency
+- build: vendorize facexlib so we don't install its unneeded dependencies
+
+
+**14.0.4**
+- docs: add a documentation website at https://brycedrennan.github.io/imaginAIry/
+- build: remove fairscale dependency
+- fix: video generation was broken
+
+**14.0.3**
+- fix: several critical bugs with package
+- tests: add a wheel smoketest to detect these issues in the future
+
 **14.0.0**
 - 🎉 video generation using [Stable Video Diffusion](https://github.com/Stability-AI/generative-models)
   - add `--videogen` to any image generation to create a short video from the generated image

imaginairy/enhancers/video_interpolation/rife/RIFE_HDv3.py

@@ -6,7 +6,7 @@ from .IFNet_HDv3 import IFNet
 class Model:
     def __init__(self):
         self.flownet = IFNet()
-        self.version: float
+        self.version = None
 
     def eval(self):
         self.flownet.eval()

imaginairy/schema.py

@@ -760,6 +760,12 @@ class ImaginePrompt(BaseModel, protected_namespaces=()):
     def height(self) -> int:
         return self.size[1]
 
+    @property
+    def aspect_ratio(self) -> str:
+        from imaginairy.utils.img_utils import aspect_ratio
+
+        return aspect_ratio(width=self.width, height=self.height)
+
     @property
     def should_use_inpainting(self) -> bool:
         return bool(self.outpaint or self.mask_image or self.mask_prompt)
@@ -787,7 +793,7 @@ class ImaginePrompt(BaseModel, protected_namespaces=()):
             "    "
             f"negative-prompt:{neg_prompt}\n"
             "    "
-            f"size:{self.width}x{self.height}px "
+            f"size:{self.width}x{self.height}px-({self.aspect_ratio}) "
             f"seed:{self.seed} "
             f"prompt-strength:{self.prompt_strength} "
             f"steps:{self.steps} solver-type:{self.solver_type} "

imaginairy/utils/img_utils.py

@@ -258,3 +258,26 @@ def calc_scale_to_fit_within(height: int, width: int, max_size) -> float:
     height_ratio = max_height / height
 
     return min(width_ratio, height_ratio)
+
+
+def aspect_ratio(width, height):
+    """
+    Calculate the aspect ratio of a given width and height.
+
+    Args:
+    width (int): The width dimension.
+    height (int): The height dimension.
+
+    Returns:
+    str: The aspect ratio in the format 'X:Y'.
+    """
+    from math import gcd
+
+    # Calculate the greatest common divisor
+    divisor = gcd(width, height)
+
+    # Calculate the aspect ratio
+    x = width // divisor
+    y = height // divisor
+
+    return f"{x}:{y}"

imaginairy/utils/log_utils.py

@@ -5,6 +5,9 @@ import logging.config
 import re
 import time
 import warnings
+from contextlib import contextmanager
+from functools import lru_cache
+from logging import Logger
 from typing import Callable
 
 import torch.cuda
@@ -57,6 +60,73 @@ def increment_step():
     _CURRENT_LOGGING_CONTEXT.step_count += 1
 
 
+@contextmanager
+def timed_log_method(logger, level, msg, *args, hide_below_ms=0, **kwargs):
+    start_time = time.perf_counter()
+    try:
+        yield
+    finally:
+        end_time = time.perf_counter()
+        elapsed_ms = (end_time - start_time) * 1000
+        if elapsed_ms < hide_below_ms:
+            return
+        full_msg = f"{msg} (in {elapsed_ms:.1f}ms)"
+        logger.log(level, full_msg, *args, **kwargs, stacklevel=3)
+
+
+@lru_cache
+def add_timed_methods_to_logger():
+    """Monkey patches the default python logger to have timed logs"""
+
+    def create_timed_method(level):
+        def timed_method(self, msg, *args, hide_below_ms=0, **kwargs):
+            return timed_log_method(
+                self, level, msg, *args, hide_below_ms=hide_below_ms, **kwargs
+            )
+
+        return timed_method
+
+    logging.Logger.timed_debug = create_timed_method(logging.DEBUG)
+    logging.Logger.timed_info = create_timed_method(logging.INFO)
+    logging.Logger.timed_warning = create_timed_method(logging.WARNING)
+    logging.Logger.timed_error = create_timed_method(logging.ERROR)
+    logging.Logger.timed_critical = create_timed_method(logging.CRITICAL)
+
+
+add_timed_methods_to_logger()
+
+
+class TimedLogger(Logger):
+    def timed_debug(self, msg, *args, hide_below_ms=0, **kwargs):
+        return timed_log_method(
+            self, logging.DEBUG, msg, *args, hide_below_ms=hide_below_ms, **kwargs
+        )
+
+    def timed_info(self, msg, *args, hide_below_ms=0, **kwargs):
+        return timed_log_method(
+            self, logging.INFO, msg, *args, hide_below_ms=hide_below_ms, **kwargs
+        )
+
+    def timed_warning(self, msg, *args, hide_below_ms=0, **kwargs):
+        return timed_log_method(
+            self, logging.WARNING, msg, *args, hide_below_ms=hide_below_ms, **kwargs
+        )
+
+    def timed_error(self, msg, *args, hide_below_ms=0, **kwargs):
+        return timed_log_method(
+            self, logging.ERROR, msg, *args, hide_below_ms=hide_below_ms, **kwargs
+        )
+
+    def timed_critical(self, msg, *args, hide_below_ms=0, **kwargs):
+        return timed_log_method(
+            self, logging.CRITICAL, msg, *args, hide_below_ms=hide_below_ms, **kwargs
+        )
+
+
+def getLogger(name) -> TimedLogger:
+    return logging.getLogger(name)  # type: ignore
+
+
 class TimingContext:
     """Tracks time and memory usage of a block of code"""
 

imaginairy/utils/model_manager.py

@@ -38,6 +38,12 @@ from imaginairy.vendored.refiners.foundationals.latent_diffusion.model import (
     LatentDiffusionModel,
 )
 from imaginairy.weight_management import translators
+from imaginairy.weight_management.translators import (
+    DoubleTextEncoderTranslator,
+    diffusers_autoencoder_kl_to_refiners_translator,
+    diffusers_unet_sdxl_to_refiners_translator,
+    load_weight_map,
+)
 
 logger = logging.getLogger(__name__)
 
@@ -108,36 +114,6 @@ def load_model_from_config(config, weights_location, half_mode=False):
     return model
 
 
-def load_model_from_config_old(
-    config, weights_location, control_weights_locations=None, half_mode=False
-):
-    model = instantiate_from_config(config.model)
-    base_model_dict = load_state_dict(weights_location, half_mode=half_mode)
-    model.init_from_state_dict(base_model_dict)
-
-    control_weights_locations = control_weights_locations or []
-    controlnets = []
-    for control_weights_location in control_weights_locations:
-        controlnet_state_dict = load_state_dict(
-            control_weights_location, half_mode=half_mode
-        )
-        controlnet_state_dict = {
-            k.replace("control_model.", ""): v for k, v in controlnet_state_dict.items()
-        }
-        controlnet = instantiate_from_config(model.control_stage_config)
-        controlnet.load_state_dict(controlnet_state_dict)
-        controlnet.to(get_device())
-        controlnets.append(controlnet)
-        model.set_control_models(controlnets)
-
-    if half_mode:
-        model = model.half()
-
-    model.to(get_device())
-    model.eval()
-    return model
-
-
 def add_controlnet(base_state_dict, controlnet_state_dict):
     """Merges a base sd15 model with a controlnet model."""
     for key in controlnet_state_dict:
@@ -286,7 +262,7 @@ def _get_diffusion_model_refiners(
 
     architecture = iconfig.MODEL_ARCHITECTURE_LOOKUP[architecture_alias]
     if architecture.primary_alias in ("sd15", "sd15inpaint"):
-        sd = _get_sd15_diffusion_model_refiners(
+        sd = load_sd15_pipeline(
             weights_location=weights_location,
             for_inpainting=for_inpainting,
             device=device,
@@ -301,10 +277,11 @@ def _get_diffusion_model_refiners(
     MOST_RECENTLY_LOADED_MODEL = sd
 
     msg = (
-        f"sd dtype:{sd.dtype} device:{sd.device}\n"
-        f"sd.unet dtype:{sd.unet.dtype} device:{sd.unet.device}\n"
-        f"sd.lda dtype:{sd.lda.dtype} device:{sd.lda.device}\n"
-        f"sd.clip_text_encoder dtype:{sd.clip_text_encoder.dtype} device:{sd.clip_text_encoder.device}\n"
+        "Pipeline loaded "
+        f"sd[dtype:{sd.dtype} device:{sd.device}] "
+        f"sd.unet[dtype:{sd.unet.dtype} device:{sd.unet.device}] "
+        f"sd.lda[dtype:{sd.lda.dtype} device:{sd.lda.device}]"
+        f"sd.clip_text_encoder[dtype:{sd.clip_text_encoder.dtype} device:{sd.clip_text_encoder.device}]"
     )
     logger.debug(msg)
 
@@ -312,7 +289,7 @@ def _get_diffusion_model_refiners(
 
 
 # new
-def _get_sd15_diffusion_model_refiners(
+def load_sd15_pipeline(
     weights_location: str,
     for_inpainting: bool = False,
     device=None,
@@ -756,7 +733,9 @@ def load_sd15_diffusers_weights(base_url: str, device=None):
     return vae_weights, unet_weights, text_encoder_weights
 
 
-def load_sdxl_diffusers_weights(base_url: str, device=None, dtype=torch.float16):
+def load_sdxl_pipeline_from_diffusers_weights(
+    base_url: str, device=None, dtype=torch.float16
+):
     from imaginairy.utils import get_device
 
     device = device or get_device()
@@ -817,13 +796,47 @@ def load_sdxl_diffusers_weights(base_url: str, device=None, dtype=torch.float16)
     return sd
 
 
-def load_sdxl_pipeline(base_url, device=None):
-    logger.info(f"Loading SDXL weights from {base_url}")
+def load_sdxl_pipeline_from_compvis_weights(
+    base_url: str, device=None, dtype=torch.float16
+):
+    from imaginairy.utils import get_device
+
     device = device or get_device()
-    sd = load_sdxl_diffusers_weights(base_url, device=device)
+    unet_weights, vae_weights, text_encoder_weights = load_sdxl_compvis_weights(
+        base_url
+    )
+    lda = SDXLAutoencoderSliced(device="cpu", dtype=dtype)
+    lda.load_state_dict(vae_weights, assign=True)
+    del vae_weights
+
+    unet = SDXLUNet(device="cpu", dtype=dtype, in_channels=4)
+    unet.load_state_dict(unet_weights, assign=True)
+    del unet_weights
+
+    text_encoder = DoubleTextEncoder(device="cpu", dtype=torch.float32)
+    text_encoder.load_state_dict(text_encoder_weights, assign=True)
+    del text_encoder_weights
+    lda = lda.to(device=device, dtype=torch.float32)
+    unet = unet.to(device=device)
+    text_encoder = text_encoder.to(device=device)
+    sd = StableDiffusion_XL(
+        device=device, dtype=None, lda=lda, unet=unet, clip_text_encoder=text_encoder
+    )
+
     return sd
 
 
+def load_sdxl_pipeline(base_url, device=None):
+    device = device or get_device()
+
+    with logger.timed_info(f"Loaded SDXL pipeline from {base_url}"):
+        if is_diffusers_repo_url(base_url):
+            sd = load_sdxl_pipeline_from_diffusers_weights(base_url, device=device)
+        else:
+            sd = load_sdxl_pipeline_from_compvis_weights(base_url, device=device)
+        return sd
+
+
 def open_weights(filepath, device=None):
     from imaginairy.utils import get_device
 
@@ -940,3 +953,76 @@ def load_stable_diffusion_compvis_weights(weights_url):
     )
 
     return vae_state_dict, unet_state_dict, text_encoder_state_dict
+
+
+def load_sdxl_compvis_weights(url):
+    from safetensors import safe_open
+
+    weights_path = get_cached_url_path(url)
+    state_dict = {}
+    unet_state_dict = {}
+    vae_state_dict = {}
+    text_encoder_1_state_dict = {}
+    text_encoder_2_state_dict = {}
+    with safe_open(weights_path, framework="pt") as f:
+        for key in f.keys():  # noqa
+            if key.startswith("model.diffusion_model."):
+                unet_state_dict[key] = f.get_tensor(key)
+            elif key.startswith("first_stage_model"):
+                vae_state_dict[key] = f.get_tensor(key)
+            elif key.startswith("conditioner.embedders.0."):
+                text_encoder_1_state_dict[key] = f.get_tensor(key)
+            elif key.startswith("conditioner.embedders.1."):
+                text_encoder_2_state_dict[key] = f.get_tensor(key)
+            else:
+                state_dict[key] = f.get_tensor(key)
+                logger.warning(f"Unused key {key}")
+
+    unet_weightmap = load_weight_map("Compvis-UNet-SDXL-to-Diffusers")
+    vae_weightmap = load_weight_map("Compvis-Autoencoder-SDXL-to-Diffusers")
+    text_encoder_1_weightmap = load_weight_map("Compvis-TextEncoder-SDXL-to-Diffusers")
+    text_encoder_2_weightmap = load_weight_map(
+        "Compvis-OpenClipTextEncoder-SDXL-to-Diffusers"
+    )
+
+    diffusers_unet_state_dict = unet_weightmap.translate_weights(unet_state_dict)
+
+    refiners_unet_state_dict = (
+        diffusers_unet_sdxl_to_refiners_translator().translate_weights(
+            diffusers_unet_state_dict
+        )
+    )
+
+    diffusers_vae_state_dict = vae_weightmap.translate_weights(vae_state_dict)
+
+    refiners_vae_state_dict = (
+        diffusers_autoencoder_kl_to_refiners_translator().translate_weights(
+            diffusers_vae_state_dict
+        )
+    )
+
+    diffusers_text_encoder_1_state_dict = text_encoder_1_weightmap.translate_weights(
+        text_encoder_1_state_dict
+    )
+
+    for key in list(text_encoder_2_state_dict.keys()):
+        if key.endswith((".in_proj_bias", ".in_proj_weight")):
+            value = text_encoder_2_state_dict[key]
+            q, k, v = value.chunk(3, dim=0)
+            text_encoder_2_state_dict[f"{key}.0"] = q
+            text_encoder_2_state_dict[f"{key}.1"] = k
+            text_encoder_2_state_dict[f"{key}.2"] = v
+            del text_encoder_2_state_dict[key]
+
+    diffusers_text_encoder_2_state_dict = text_encoder_2_weightmap.translate_weights(
+        text_encoder_2_state_dict
+    )
+
+    refiners_text_encoder_weights = DoubleTextEncoderTranslator().translate_weights(
+        diffusers_text_encoder_1_state_dict, diffusers_text_encoder_2_state_dict
+    )
+    return (
+        refiners_unet_state_dict,
+        refiners_vae_state_dict,
+        refiners_text_encoder_weights,
+    )

imaginairy/weight_management/translation.py

@@ -1,4 +1,3 @@
-import logging
 import re
 from dataclasses import asdict, dataclass, field
 from typing import Dict
@@ -7,7 +6,9 @@ import torch
 from safetensors import safe_open
 from torch import device as Device
 
-logger = logging.getLogger(__name__)
+from imaginairy.utils.log_utils import getLogger
+
+logger = getLogger(__name__)
 
 
 TensorDict = Dict[str, torch.Tensor]
@@ -21,7 +22,7 @@ class WeightTranslationMap:
     source_aliases: dict[str, str] = field(default_factory=dict)
     reshapes: dict[str, tuple[int, ...]] = field(default_factory=dict)
 
-    def load_and_translate_weights(
+    def load_untranslated_weights(
         self, source_path: str, device: Device | str = "cpu"
     ) -> TensorDict:
         extension = source_path.split(".")[-1]
@@ -34,7 +35,12 @@ class WeightTranslationMap:
         else:
             msg = f"Unsupported extension {extension}"
             raise ValueError(msg)
+        return source_weights
 
+    def load_and_translate_weights(
+        self, source_path: str, device: Device | str = "cpu"
+    ) -> TensorDict:
+        source_weights = self.load_untranslated_weights(source_path, device=device)
         return self.translate_weights(source_weights)
 
     def translate_weights(self, source_weights: TensorDict) -> TensorDict:
@@ -90,19 +96,21 @@ def check_nan_path(path: str, device):
 
 
 def translate_weights(
-    source_weights: TensorDict, weight_map: WeightTranslationMap
+    source_weights: TensorDict, weight_map: WeightTranslationMap, nan_check=False
 ) -> TensorDict:
     new_state_dict: TensorDict = {}
     # check source weights for nan
-    for k, v in source_weights.items():
-        nan_count = torch.sum(torch.isnan(v)).item()
-        if nan_count:
-            msg = (
-                f"Found {nan_count} nan values in {k} of source state dict."
-                " This could indicate the source weights are corrupted and "
-                "need to be re-downloaded. "
-            )
-            logger.warning(msg)
+    if nan_check:
+        with logger.timed_debug("Checking for nans"):
+            for k, v in source_weights.items():
+                nan_count = torch.sum(torch.isnan(v)).item()
+                if nan_count:
+                    msg = (
+                        f"Found {nan_count} nan values in {k} of source state dict."
+                        " This could indicate the source weights are corrupted and "
+                        "need to be re-downloaded. "
+                    )
+                    logger.warning(msg)
 
     # print(f"Translating {len(source_weights)} weights")
     # print(f"Using {len(weight_map.name_map)} name mappings")
@@ -110,79 +118,85 @@ def translate_weights(
 
     source_weights = flatten_dict(source_weights)
 
-    for source_key in list(source_weights.keys()):
-        source_key = weight_map.source_aliases.get(source_key, source_key)
-        try:
-            target_key = weight_map.name_map[source_key]
-            # print(f"Found {source_prefix} -> {target_prefix}")
-        except KeyError:
-            continue
-        if target_key is None:
-            # mapped to None means we ignore it
-            source_weights.pop(source_key)
-        else:
-            # print(f"Adding {target_key}")
-            new_state_dict[target_key] = source_weights.pop(source_key)
+    with logger.timed_debug("Translating exact match keys", hide_below_ms=50):
+        for source_key in list(source_weights.keys()):
+            source_key = weight_map.source_aliases.get(source_key, source_key)
+            try:
+                target_key = weight_map.name_map[source_key]
+                # print(f"Found {source_prefix} -> {target_prefix}")
+            except KeyError:
+                continue
+            if target_key is None:
+                # mapped to None means we ignore it
+                source_weights.pop(source_key)
+                continue
 
-    for source_key in list(source_weights.keys()):
-        try:
-            source_prefix, suffix = source_key.rsplit(sep=".", maxsplit=1)
-        except ValueError:
-            # no dots
-            continue
-        # print(f"Checking {source_prefix} {suffix}")
-
-        source_prefix = weight_map.source_aliases.get(source_prefix, source_prefix)
-        try:
-            target_prefix = weight_map.name_map[source_prefix]
-            # print(f"Found {source_prefix} -> {target_prefix}")
-        except KeyError:
-            continue
-        if target_prefix is None:
-            # mapped to None means we ignore it
-            source_weights.pop(source_key)
-            continue
-        else:
-            target_key = ".".join([target_prefix, suffix])
-            # print(f"Adding {target_key}")
             new_state_dict[target_key] = source_weights.pop(source_key)
 
-    for source_key in list(source_weights.keys()):
-        try:
-            source_prefix, suffix = source_key.rsplit(sep=".", maxsplit=1)
-        except ValueError:
-            # no dots
-            continue
-        for pattern, replace_pattern in weight_map.regex_map.items():
-            match = re.match(pattern, source_prefix)
-            if match:
-                match_data = match.groupdict()
-                new_k = render_fstring(replace_pattern, match_data)
-                new_k = ".".join([new_k, suffix])
-                new_state_dict[new_k] = source_weights.pop(source_key)
+    with logger.timed_debug("Translating prefix matched keys", hide_below_ms=50):
+        for source_key in list(source_weights.keys()):
+            try:
+                source_prefix, suffix = source_key.rsplit(sep=".", maxsplit=1)
+            except ValueError:
+                # no dots
+                continue
+            # print(f"Checking {source_prefix} {suffix}")
+
+            source_prefix = weight_map.source_aliases.get(source_prefix, source_prefix)
+            try:
+                target_prefix = weight_map.name_map[source_prefix]
+                # print(f"Found {source_prefix} -> {target_prefix}")
+            except KeyError:
+                continue
+            if target_prefix is None:
+                # mapped to None means we ignore it
+                source_weights.pop(source_key)
+                continue
+            else:
+                target_key = ".".join([target_prefix, suffix])
+                # print(f"Adding {target_key}")
+                new_state_dict[target_key] = source_weights.pop(source_key)
+
+    with logger.timed_debug("Translating regex matched keys", hide_below_ms=50):
+        for source_key in list(source_weights.keys()):
+            try:
+                source_prefix, suffix = source_key.rsplit(sep=".", maxsplit=1)
+            except ValueError:
+                # no dots
+                continue
+            for pattern, replace_pattern in weight_map.regex_map.items():
+                match = re.match(pattern, source_prefix)
+                if match:
+                    match_data = match.groupdict()
+                    new_k = render_fstring(replace_pattern, match_data)
+                    new_k = ".".join([new_k, suffix])
+                    new_state_dict[new_k] = source_weights.pop(source_key)
 
     if source_weights:
         msg = f"Unmapped keys: {list(source_weights.keys())}"
         logger.info(msg)
         for k in source_weights:
             if isinstance(source_weights[k], torch.Tensor):
-                print(f"  {k}: {source_weights[k].shape}")
+                logger.info(f"  Unmapped key '{k}': {source_weights[k].shape}")
             else:
-                print(f"  {k}: {repr(source_weights[k])[:100]}")
+                logger.info(f"  Unmapped key '{k}': {repr(source_weights[k])[:100]}")
 
-    if weight_map.reshapes:
-        for key, new_shape in weight_map.reshapes.items():
-            if key in new_state_dict:
-                new_state_dict[key] = new_state_dict[key].reshape(new_shape)
+    with logger.timed_debug("reshaping weights", hide_below_ms=50):
+        if weight_map.reshapes:
+            for key, new_shape in weight_map.reshapes.items():
+                if key in new_state_dict:
+                    new_state_dict[key] = new_state_dict[key].reshape(new_shape)
 
     # check for nan values
-    for k in list(new_state_dict.keys()):
-        v = new_state_dict[k]
-        nan_count = torch.sum(torch.isnan(v)).item()
-        if nan_count:
-            logger.warning(
-                f"Found {nan_count} nan values in {k} of converted state dict."
-            )
+    if nan_check:
+        with logger.timed_debug("Checking for nans", hide_below_ms=50):
+            for k in list(new_state_dict.keys()):
+                v = new_state_dict[k]
+                nan_count = torch.sum(torch.isnan(v)).item()
+                if nan_count:
+                    logger.warning(
+                        f"Found {nan_count} nan values in {k} of converted state dict."
+                    )
 
     return new_state_dict
 

imaginairy/weight_management/translators.py

@@ -10,79 +10,91 @@ weight_map_folder = os.path.join(_current_folder, "weight_maps")
 
 
 @lru_cache
-def _weight_map(map_name: str) -> WeightTranslationMap:
+def load_weight_map(map_name: str) -> WeightTranslationMap:
     map_path = os.path.join(weight_map_folder, f"{map_name}.weightmap.json")
     return WeightTranslationMap.load(map_path)
 
 
 def transformers_text_encoder_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Transformers-ClipTextEncoder")
+    return load_weight_map("Transformers-ClipTextEncoder")
 
 
 def transformers_image_encoder_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Transformers-ClipImageEncoder-SD21")
+    return load_weight_map("Transformers-ClipImageEncoder-SD21")
 
 
 def diffusers_autoencoder_kl_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-AutoencoderKL-SD")
+    return load_weight_map("Diffusers-AutoencoderKL-SD")
 
 
 def diffusers_unet_sd15_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-UNet-SD15")
+    return load_weight_map("Diffusers-UNet-SD15")
 
 
 def diffusers_unet_sdxl_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-UNet-SDXL")
+    return load_weight_map("Diffusers-UNet-SDXL")
 
 
 def informative_drawings_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("InformativeDrawings")
+    return load_weight_map("InformativeDrawings")
 
 
 def diffusers_controlnet_sd15_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-Controlnet-SD15")
+    return load_weight_map("Diffusers-Controlnet-SD15")
 
 
 def diffusers_ip_adapter_sd15_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-IPAdapter-SD15")
+    return load_weight_map("Diffusers-IPAdapter-SD15")
 
 
 def diffusers_ip_adapter_sdxl_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-IPAdapter-SDXL")
+    return load_weight_map("Diffusers-IPAdapter-SDXL")
 
 
 def diffusers_ip_adapter_plus_sd15_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-IPAdapterPlus-SD15")
+    return load_weight_map("Diffusers-IPAdapterPlus-SD15")
 
 
 def diffusers_ip_adapter_plus_sdxl_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-IPAdapterPlus-SDXL")
+    return load_weight_map("Diffusers-IPAdapterPlus-SDXL")
 
 
 def diffusers_t2i_adapter_sd15_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-T2IAdapter-SD15")
+    return load_weight_map("Diffusers-T2IAdapter-SD15")
 
 
 def diffusers_t2i_adapter_sdxl_to_refiners_translator() -> WeightTranslationMap:
-    return _weight_map("Diffusers-T2IAdapter-SDXL")
+    return load_weight_map("Diffusers-T2IAdapter-SDXL")
 
 
 class DoubleTextEncoderTranslator:
     def __init__(self):
         self.translator = transformers_text_encoder_to_refiners_translator()
 
-    def load_and_translate_weights(
+    def load_untranslated_weights(
         self,
         text_encoder_l_weights_path: str,
         text_encoder_g_weights_path: str,
         device: Device | str = "cpu",
-    ) -> TensorDict:
-        text_encoder_l_weights = self.translator.load_and_translate_weights(
+    ) -> tuple[TensorDict, TensorDict]:
+        text_encoder_l_weights = self.translator.load_untranslated_weights(
             text_encoder_l_weights_path, device=device
         )
-        text_encoder_g_weights = self.translator.load_and_translate_weights(
+        text_encoder_g_weights = self.translator.load_untranslated_weights(
             text_encoder_g_weights_path, device=device
         )
+        return text_encoder_l_weights, text_encoder_g_weights
+
+    def load_and_translate_weights(
+        self,
+        text_encoder_l_weights_path: str,
+        text_encoder_g_weights_path: str,
+        device: Device | str = "cpu",
+    ) -> TensorDict:
+        text_encoder_l_weights, text_encoder_g_weights = self.load_untranslated_weights(
+            text_encoder_l_weights_path, text_encoder_g_weights_path, device=device
+        )
+
         return self.translate_weights(text_encoder_l_weights, text_encoder_g_weights)
 
     def translate_weights(
@@ -90,6 +102,13 @@ class DoubleTextEncoderTranslator:
     ) -> TensorDict:
         new_sd: TensorDict = {}
 
+        text_encoder_l_weights = self.translator.translate_weights(
+            text_encoder_l_weights
+        )
+        text_encoder_g_weights = self.translator.translate_weights(
+            text_encoder_g_weights
+        )
+
         for k in list(text_encoder_l_weights.keys()):
             if k.startswith("TransformerLayer_12"):
                 text_encoder_l_weights.pop(k)

imaginairy/weight_management/weight_maps/Compvis-Autoencoder-SDXL-to-Diffusers.weightmap.json

@@ -0,0 +1,256 @@
+{
+  "name_map": {
+    "first_stage_model.decoder.conv_in.bias": "decoder.conv_in.bias",
+    "first_stage_model.decoder.conv_in.weight": "decoder.conv_in.weight",
+    "first_stage_model.decoder.norm_out.bias": "decoder.conv_norm_out.bias",
+    "first_stage_model.decoder.norm_out.weight": "decoder.conv_norm_out.weight",
+    "first_stage_model.decoder.conv_out.bias": "decoder.conv_out.bias",
+    "first_stage_model.decoder.conv_out.weight": "decoder.conv_out.weight",
+    "first_stage_model.decoder.mid.attn_1.norm.bias": "decoder.mid_block.attentions.0.group_norm.bias",
+    "first_stage_model.decoder.mid.attn_1.norm.weight": "decoder.mid_block.attentions.0.group_norm.weight",
+    "first_stage_model.decoder.mid.attn_1.k.bias": "decoder.mid_block.attentions.0.to_k.bias",
+    "first_stage_model.decoder.mid.attn_1.k.weight": "decoder.mid_block.attentions.0.to_k.weight",
+    "first_stage_model.decoder.mid.attn_1.proj_out.bias": "decoder.mid_block.attentions.0.to_out.0.bias",
+    "first_stage_model.decoder.mid.attn_1.proj_out.weight": "decoder.mid_block.attentions.0.to_out.0.weight",
+    "first_stage_model.decoder.mid.attn_1.q.bias": "decoder.mid_block.attentions.0.to_q.bias",
+    "first_stage_model.decoder.mid.attn_1.q.weight": "decoder.mid_block.attentions.0.to_q.weight",
+    "first_stage_model.decoder.mid.attn_1.v.bias": "decoder.mid_block.attentions.0.to_v.bias",
+    "first_stage_model.decoder.mid.attn_1.v.weight": "decoder.mid_block.attentions.0.to_v.weight",
+    "first_stage_model.decoder.mid.block_1.conv1.bias": "decoder.mid_block.resnets.0.conv1.bias",
+    "first_stage_model.decoder.mid.block_1.conv1.weight": "decoder.mid_block.resnets.0.conv1.weight",
+    "first_stage_model.decoder.mid.block_1.conv2.bias": "decoder.mid_block.resnets.0.conv2.bias",
+    "first_stage_model.decoder.mid.block_1.conv2.weight": "decoder.mid_block.resnets.0.conv2.weight",
+    "first_stage_model.decoder.mid.block_1.norm1.bias": "decoder.mid_block.resnets.0.norm1.bias",
+    "first_stage_model.decoder.mid.block_1.norm1.weight": "decoder.mid_block.resnets.0.norm1.weight",
+    "first_stage_model.decoder.mid.block_1.norm2.bias": "decoder.mid_block.resnets.0.norm2.bias",
+    "first_stage_model.decoder.mid.block_1.norm2.weight": "decoder.mid_block.resnets.0.norm2.weight",
+    "first_stage_model.decoder.mid.block_2.conv1.bias": "decoder.mid_block.resnets.1.conv1.bias",
+    "first_stage_model.decoder.mid.block_2.conv1.weight": "decoder.mid_block.resnets.1.conv1.weight",
+    "first_stage_model.decoder.mid.block_2.conv2.bias": "decoder.mid_block.resnets.1.conv2.bias",
+    "first_stage_model.decoder.mid.block_2.conv2.weight": "decoder.mid_block.resnets.1.conv2.weight",
+    "first_stage_model.decoder.mid.block_2.norm1.bias": "decoder.mid_block.resnets.1.norm1.bias",
+    "first_stage_model.decoder.mid.block_2.norm1.weight": "decoder.mid_block.resnets.1.norm1.weight",
+    "first_stage_model.decoder.mid.block_2.norm2.bias": "decoder.mid_block.resnets.1.norm2.bias",
+    "first_stage_model.decoder.mid.block_2.norm2.weight": "decoder.mid_block.resnets.1.norm2.weight",
+    "first_stage_model.decoder.up.3.block.0.conv1.bias": "decoder.up_blocks.0.resnets.0.conv1.bias",
+    "first_stage_model.decoder.up.3.block.0.conv1.weight": "decoder.up_blocks.0.resnets.0.conv1.weight",
+    "first_stage_model.decoder.up.3.block.0.conv2.bias": "decoder.up_blocks.0.resnets.0.conv2.bias",
+    "first_stage_model.decoder.up.3.block.0.conv2.weight": "decoder.up_blocks.0.resnets.0.conv2.weight",
+    "first_stage_model.decoder.up.3.block.0.norm1.bias": "decoder.up_blocks.0.resnets.0.norm1.bias",
+    "first_stage_model.decoder.up.3.block.0.norm1.weight": "decoder.up_blocks.0.resnets.0.norm1.weight",
+    "first_stage_model.decoder.up.3.block.0.norm2.bias": "decoder.up_blocks.0.resnets.0.norm2.bias",
+    "first_stage_model.decoder.up.3.block.0.norm2.weight": "decoder.up_blocks.0.resnets.0.norm2.weight",
+    "first_stage_model.decoder.up.3.block.1.conv1.bias": "decoder.up_blocks.0.resnets.1.conv1.bias",
+    "first_stage_model.decoder.up.3.block.1.conv1.weight": "decoder.up_blocks.0.resnets.1.conv1.weight",
+    "first_stage_model.decoder.up.3.block.1.conv2.bias": "decoder.up_blocks.0.resnets.1.conv2.bias",
+    "first_stage_model.decoder.up.3.block.1.conv2.weight": "decoder.up_blocks.0.resnets.1.conv2.weight",
+    "first_stage_model.decoder.up.3.block.1.norm1.bias": "decoder.up_blocks.0.resnets.1.norm1.bias",
+    "first_stage_model.decoder.up.3.block.1.norm1.weight": "decoder.up_blocks.0.resnets.1.norm1.weight",
+    "first_stage_model.decoder.up.3.block.1.norm2.bias": "decoder.up_blocks.0.resnets.1.norm2.bias",
+    "first_stage_model.decoder.up.3.block.1.norm2.weight": "decoder.up_blocks.0.resnets.1.norm2.weight",
+    "first_stage_model.decoder.up.3.block.2.conv1.bias": "decoder.up_blocks.0.resnets.2.conv1.bias",
+    "first_stage_model.decoder.up.3.block.2.conv1.weight": "decoder.up_blocks.0.resnets.2.conv1.weight",
+    "first_stage_model.decoder.up.3.block.2.conv2.bias": "decoder.up_blocks.0.resnets.2.conv2.bias",
+    "first_stage_model.decoder.up.3.block.2.conv2.weight": "decoder.up_blocks.0.resnets.2.conv2.weight",
+    "first_stage_model.decoder.up.3.block.2.norm1.bias": "decoder.up_blocks.0.resnets.2.norm1.bias",
+    "first_stage_model.decoder.up.3.block.2.norm1.weight": "decoder.up_blocks.0.resnets.2.norm1.weight",
+    "first_stage_model.decoder.up.3.block.2.norm2.bias": "decoder.up_blocks.0.resnets.2.norm2.bias",
+    "first_stage_model.decoder.up.3.block.2.norm2.weight": "decoder.up_blocks.0.resnets.2.norm2.weight",
+    "first_stage_model.decoder.up.3.upsample.conv.bias": "decoder.up_blocks.0.upsamplers.0.conv.bias",
+    "first_stage_model.decoder.up.3.upsample.conv.weight": "decoder.up_blocks.0.upsamplers.0.conv.weight",
+    "first_stage_model.decoder.up.2.block.0.conv1.bias": "decoder.up_blocks.1.resnets.0.conv1.bias",
+    "first_stage_model.decoder.up.2.block.0.conv1.weight": "decoder.up_blocks.1.resnets.0.conv1.weight",
+    "first_stage_model.decoder.up.2.block.0.conv2.bias": "decoder.up_blocks.1.resnets.0.conv2.bias",
+    "first_stage_model.decoder.up.2.block.0.conv2.weight": "decoder.up_blocks.1.resnets.0.conv2.weight",
+    "first_stage_model.decoder.up.2.block.0.norm1.bias": "decoder.up_blocks.1.resnets.0.norm1.bias",
+    "first_stage_model.decoder.up.2.block.0.norm1.weight": "decoder.up_blocks.1.resnets.0.norm1.weight",
+    "first_stage_model.decoder.up.2.block.0.norm2.bias": "decoder.up_blocks.1.resnets.0.norm2.bias",
+    "first_stage_model.decoder.up.2.block.0.norm2.weight": "decoder.up_blocks.1.resnets.0.norm2.weight",
+    "first_stage_model.decoder.up.2.block.1.conv1.bias": "decoder.up_blocks.1.resnets.1.conv1.bias",
+    "first_stage_model.decoder.up.2.block.1.conv1.weight": "decoder.up_blocks.1.resnets.1.conv1.weight",
+    "first_stage_model.decoder.up.2.block.1.conv2.bias": "decoder.up_blocks.1.resnets.1.conv2.bias",
+    "first_stage_model.decoder.up.2.block.1.conv2.weight": "decoder.up_blocks.1.resnets.1.conv2.weight",
+    "first_stage_model.decoder.up.2.block.1.norm1.bias": "decoder.up_blocks.1.resnets.1.norm1.bias",
+    "first_stage_model.decoder.up.2.block.1.norm1.weight": "decoder.up_blocks.1.resnets.1.norm1.weight",
+    "first_stage_model.decoder.up.2.block.1.norm2.bias": "decoder.up_blocks.1.resnets.1.norm2.bias",
+    "first_stage_model.decoder.up.2.block.1.norm2.weight": "decoder.up_blocks.1.resnets.1.norm2.weight",
+    "first_stage_model.decoder.up.2.block.2.conv1.bias": "decoder.up_blocks.1.resnets.2.conv1.bias",
+    "first_stage_model.decoder.up.2.block.2.conv1.weight": "decoder.up_blocks.1.resnets.2.conv1.weight",
+    "first_stage_model.decoder.up.2.block.2.conv2.bias": "decoder.up_blocks.1.resnets.2.conv2.bias",
+    "first_stage_model.decoder.up.2.block.2.conv2.weight": "decoder.up_blocks.1.resnets.2.conv2.weight",
+    "first_stage_model.decoder.up.2.block.2.norm1.bias": "decoder.up_blocks.1.resnets.2.norm1.bias",
+    "first_stage_model.decoder.up.2.block.2.norm1.weight": "decoder.up_blocks.1.resnets.2.norm1.weight",
+    "first_stage_model.decoder.up.2.block.2.norm2.bias": "decoder.up_blocks.1.resnets.2.norm2.bias",
+    "first_stage_model.decoder.up.2.block.2.norm2.weight": "decoder.up_blocks.1.resnets.2.norm2.weight",
+    "first_stage_model.decoder.up.2.upsample.conv.bias": "decoder.up_blocks.1.upsamplers.0.conv.bias",
+    "first_stage_model.decoder.up.2.upsample.conv.weight": "decoder.up_blocks.1.upsamplers.0.conv.weight",
+    "first_stage_model.decoder.up.1.block.0.conv1.bias": "decoder.up_blocks.2.resnets.0.conv1.bias",
+    "first_stage_model.decoder.up.1.block.0.conv1.weight": "decoder.up_blocks.2.resnets.0.conv1.weight",
+    "first_stage_model.decoder.up.1.block.0.conv2.bias": "decoder.up_blocks.2.resnets.0.conv2.bias",
+    "first_stage_model.decoder.up.1.block.0.conv2.weight": "decoder.up_blocks.2.resnets.0.conv2.weight",
+    "first_stage_model.decoder.up.1.block.0.nin_shortcut.bias": "decoder.up_blocks.2.resnets.0.conv_shortcut.bias",
+    "first_stage_model.decoder.up.1.block.0.nin_shortcut.weight": "decoder.up_blocks.2.resnets.0.conv_shortcut.weight",
+    "first_stage_model.decoder.up.1.block.0.norm1.bias": "decoder.up_blocks.2.resnets.0.norm1.bias",
+    "first_stage_model.decoder.up.1.block.0.norm1.weight": "decoder.up_blocks.2.resnets.0.norm1.weight",
+    "first_stage_model.decoder.up.1.block.0.norm2.bias": "decoder.up_blocks.2.resnets.0.norm2.bias",
+    "first_stage_model.decoder.up.1.block.0.norm2.weight": "decoder.up_blocks.2.resnets.0.norm2.weight",
+    "first_stage_model.decoder.up.1.block.1.conv1.bias": "decoder.up_blocks.2.resnets.1.conv1.bias",
+    "first_stage_model.decoder.up.1.block.1.conv1.weight": "decoder.up_blocks.2.resnets.1.conv1.weight",
+    "first_stage_model.decoder.up.1.block.1.conv2.bias": "decoder.up_blocks.2.resnets.1.conv2.bias",
+    "first_stage_model.decoder.up.1.block.1.conv2.weight": "decoder.up_blocks.2.resnets.1.conv2.weight",
+    "first_stage_model.decoder.up.1.block.1.norm1.bias": "decoder.up_blocks.2.resnets.1.norm1.bias",
+    "first_stage_model.decoder.up.1.block.1.norm1.weight": "decoder.up_blocks.2.resnets.1.norm1.weight",
+    "first_stage_model.decoder.up.1.block.1.norm2.bias": "decoder.up_blocks.2.resnets.1.norm2.bias",
+    "first_stage_model.decoder.up.1.block.1.norm2.weight": "decoder.up_blocks.2.resnets.1.norm2.weight",
+    "first_stage_model.decoder.up.1.block.2.conv1.bias": "decoder.up_blocks.2.resnets.2.conv1.bias",
+    "first_stage_model.decoder.up.1.block.2.conv1.weight": "decoder.up_blocks.2.resnets.2.conv1.weight",
+    "first_stage_model.decoder.up.1.block.2.conv2.bias": "decoder.up_blocks.2.resnets.2.conv2.bias",
+    "first_stage_model.decoder.up.1.block.2.conv2.weight": "decoder.up_blocks.2.resnets.2.conv2.weight",
+    "first_stage_model.decoder.up.1.block.2.norm1.bias": "decoder.up_blocks.2.resnets.2.norm1.bias",
+    "first_stage_model.decoder.up.1.block.2.norm1.weight": "decoder.up_blocks.2.resnets.2.norm1.weight",
+    "first_stage_model.decoder.up.1.block.2.norm2.bias": "decoder.up_blocks.2.resnets.2.norm2.bias",
+    "first_stage_model.decoder.up.1.block.2.norm2.weight": "decoder.up_blocks.2.resnets.2.norm2.weight",
+    "first_stage_model.decoder.up.1.upsample.conv.bias": "decoder.up_blocks.2.upsamplers.0.conv.bias",
+    "first_stage_model.decoder.up.1.upsample.conv.weight": "decoder.up_blocks.2.upsamplers.0.conv.weight",
+    "first_stage_model.decoder.up.0.block.0.conv1.bias": "decoder.up_blocks.3.resnets.0.conv1.bias",
+    "first_stage_model.decoder.up.0.block.0.conv1.weight": "decoder.up_blocks.3.resnets.0.conv1.weight",
+    "first_stage_model.decoder.up.0.block.0.conv2.bias": "decoder.up_blocks.3.resnets.0.conv2.bias",
+    "first_stage_model.decoder.up.0.block.0.conv2.weight": "decoder.up_blocks.3.resnets.0.conv2.weight",
+    "first_stage_model.decoder.up.0.block.0.nin_shortcut.bias": "decoder.up_blocks.3.resnets.0.conv_shortcut.bias",
+    "first_stage_model.decoder.up.0.block.0.nin_shortcut.weight": "decoder.up_blocks.3.resnets.0.conv_shortcut.weight",
+    "first_stage_model.decoder.up.0.block.0.norm1.bias": "decoder.up_blocks.3.resnets.0.norm1.bias",
+    "first_stage_model.decoder.up.0.block.0.norm1.weight": "decoder.up_blocks.3.resnets.0.norm1.weight",
+    "first_stage_model.decoder.up.0.block.0.norm2.bias": "decoder.up_blocks.3.resnets.0.norm2.bias",
+    "first_stage_model.decoder.up.0.block.0.norm2.weight": "decoder.up_blocks.3.resnets.0.norm2.weight",
+    "first_stage_model.decoder.up.0.block.1.conv1.bias": "decoder.up_blocks.3.resnets.1.conv1.bias",
+    "first_stage_model.decoder.up.0.block.1.conv1.weight": "decoder.up_blocks.3.resnets.1.conv1.weight",
+    "first_stage_model.decoder.up.0.block.1.conv2.bias": "decoder.up_blocks.3.resnets.1.conv2.bias",
+    "first_stage_model.decoder.up.0.block.1.conv2.weight": "decoder.up_blocks.3.resnets.1.conv2.weight",
+    "first_stage_model.decoder.up.0.block.1.norm1.bias": "decoder.up_blocks.3.resnets.1.norm1.bias",
+    "first_stage_model.decoder.up.0.block.1.norm1.weight": "decoder.up_blocks.3.resnets.1.norm1.weight",
+    "first_stage_model.decoder.up.0.block.1.norm2.bias": "decoder.up_blocks.3.resnets.1.norm2.bias",
+    "first_stage_model.decoder.up.0.block.1.norm2.weight": "decoder.up_blocks.3.resnets.1.norm2.weight",
+    "first_stage_model.decoder.up.0.block.2.conv1.bias": "decoder.up_blocks.3.resnets.2.conv1.bias",
+    "first_stage_model.decoder.up.0.block.2.conv1.weight": "decoder.up_blocks.3.resnets.2.conv1.weight",
+    "first_stage_model.decoder.up.0.block.2.conv2.bias": "decoder.up_blocks.3.resnets.2.conv2.bias",
+    "first_stage_model.decoder.up.0.block.2.conv2.weight": "decoder.up_blocks.3.resnets.2.conv2.weight",
+    "first_stage_model.decoder.up.0.block.2.norm1.bias": "decoder.up_blocks.3.resnets.2.norm1.bias",
+    "first_stage_model.decoder.up.0.block.2.norm1.weight": "decoder.up_blocks.3.resnets.2.norm1.weight",
+    "first_stage_model.decoder.up.0.block.2.norm2.bias": "decoder.up_blocks.3.resnets.2.norm2.bias",
+    "first_stage_model.decoder.up.0.block.2.norm2.weight": "decoder.up_blocks.3.resnets.2.norm2.weight",
+    "first_stage_model.encoder.conv_in.bias": "encoder.conv_in.bias",
+    "first_stage_model.encoder.conv_in.weight": "encoder.conv_in.weight",
+    "first_stage_model.encoder.norm_out.bias": "encoder.conv_norm_out.bias",
+    "first_stage_model.encoder.norm_out.weight": "encoder.conv_norm_out.weight",
+    "first_stage_model.encoder.conv_out.bias": "encoder.conv_out.bias",
+    "first_stage_model.encoder.conv_out.weight": "encoder.conv_out.weight",
+    "first_stage_model.encoder.down.0.downsample.conv.bias": "encoder.down_blocks.0.downsamplers.0.conv.bias",
+    "first_stage_model.encoder.down.0.downsample.conv.weight": "encoder.down_blocks.0.downsamplers.0.conv.weight",
+    "first_stage_model.encoder.down.0.block.0.conv1.bias": "encoder.down_blocks.0.resnets.0.conv1.bias",
+    "first_stage_model.encoder.down.0.block.0.conv1.weight": "encoder.down_blocks.0.resnets.0.conv1.weight",
+    "first_stage_model.encoder.down.0.block.0.conv2.bias": "encoder.down_blocks.0.resnets.0.conv2.bias",
+    "first_stage_model.encoder.down.0.block.0.conv2.weight": "encoder.down_blocks.0.resnets.0.conv2.weight",
+    "first_stage_model.encoder.down.0.block.0.norm1.bias": "encoder.down_blocks.0.resnets.0.norm1.bias",
+    "first_stage_model.encoder.down.0.block.0.norm1.weight": "encoder.down_blocks.0.resnets.0.norm1.weight",
+    "first_stage_model.encoder.down.0.block.0.norm2.bias": "encoder.down_blocks.0.resnets.0.norm2.bias",
+    "first_stage_model.encoder.down.0.block.0.norm2.weight": "encoder.down_blocks.0.resnets.0.norm2.weight",
+    "first_stage_model.encoder.down.0.block.1.conv1.bias": "encoder.down_blocks.0.resnets.1.conv1.bias",
+    "first_stage_model.encoder.down.0.block.1.conv1.weight": "encoder.down_blocks.0.resnets.1.conv1.weight",
+    "first_stage_model.encoder.down.0.block.1.conv2.bias": "encoder.down_blocks.0.resnets.1.conv2.bias",
+    "first_stage_model.encoder.down.0.block.1.conv2.weight": "encoder.down_blocks.0.resnets.1.conv2.weight",
+    "first_stage_model.encoder.down.0.block.1.norm1.bias": "encoder.down_blocks.0.resnets.1.norm1.bias",
+    "first_stage_model.encoder.down.0.block.1.norm1.weight": "encoder.down_blocks.0.resnets.1.norm1.weight",
+    "first_stage_model.encoder.down.0.block.1.norm2.bias": "encoder.down_blocks.0.resnets.1.norm2.bias",
+    "first_stage_model.encoder.down.0.block.1.norm2.weight": "encoder.down_blocks.0.resnets.1.norm2.weight",
+    "first_stage_model.encoder.down.1.downsample.conv.bias": "encoder.down_blocks.1.downsamplers.0.conv.bias",
+    "first_stage_model.encoder.down.1.downsample.conv.weight": "encoder.down_blocks.1.downsamplers.0.conv.weight",
+    "first_stage_model.encoder.down.1.block.0.conv1.bias": "encoder.down_blocks.1.resnets.0.conv1.bias",
+    "first_stage_model.encoder.down.1.block.0.conv1.weight": "encoder.down_blocks.1.resnets.0.conv1.weight",
+    "first_stage_model.encoder.down.1.block.0.conv2.bias": "encoder.down_blocks.1.resnets.0.conv2.bias",
+    "first_stage_model.encoder.down.1.block.0.conv2.weight": "encoder.down_blocks.1.resnets.0.conv2.weight",
+    "first_stage_model.encoder.down.1.block.0.nin_shortcut.bias": "encoder.down_blocks.1.resnets.0.conv_shortcut.bias",
+    "first_stage_model.encoder.down.1.block.0.nin_shortcut.weight": "encoder.down_blocks.1.resnets.0.conv_shortcut.weight",
+    "first_stage_model.encoder.down.1.block.0.norm1.bias": "encoder.down_blocks.1.resnets.0.norm1.bias",
+    "first_stage_model.encoder.down.1.block.0.norm1.weight": "encoder.down_blocks.1.resnets.0.norm1.weight",
+    "first_stage_model.encoder.down.1.block.0.norm2.bias": "encoder.down_blocks.1.resnets.0.norm2.bias",
+    "first_stage_model.encoder.down.1.block.0.norm2.weight": "encoder.down_blocks.1.resnets.0.norm2.weight",
+    "first_stage_model.encoder.down.1.block.1.conv1.bias": "encoder.down_blocks.1.resnets.1.conv1.bias",
+    "first_stage_model.encoder.down.1.block.1.conv1.weight": "encoder.down_blocks.1.resnets.1.conv1.weight",
+    "first_stage_model.encoder.down.1.block.1.conv2.bias": "encoder.down_blocks.1.resnets.1.conv2.bias",
+    "first_stage_model.encoder.down.1.block.1.conv2.weight": "encoder.down_blocks.1.resnets.1.conv2.weight",
+    "first_stage_model.encoder.down.1.block.1.norm1.bias": "encoder.down_blocks.1.resnets.1.norm1.bias",
+    "first_stage_model.encoder.down.1.block.1.norm1.weight": "encoder.down_blocks.1.resnets.1.norm1.weight",
+    "first_stage_model.encoder.down.1.block.1.norm2.bias": "encoder.down_blocks.1.resnets.1.norm2.bias",
+    "first_stage_model.encoder.down.1.block.1.norm2.weight": "encoder.down_blocks.1.resnets.1.norm2.weight",
+    "first_stage_model.encoder.down.2.downsample.conv.bias": "encoder.down_blocks.2.downsamplers.0.conv.bias",
+    "first_stage_model.encoder.down.2.downsample.conv.weight": "encoder.down_blocks.2.downsamplers.0.conv.weight",
+    "first_stage_model.encoder.down.2.block.0.conv1.bias": "encoder.down_blocks.2.resnets.0.conv1.bias",
+    "first_stage_model.encoder.down.2.block.0.conv1.weight": "encoder.down_blocks.2.resnets.0.conv1.weight",
+    "first_stage_model.encoder.down.2.block.0.conv2.bias": "encoder.down_blocks.2.resnets.0.conv2.bias",
+    "first_stage_model.encoder.down.2.block.0.conv2.weight": "encoder.down_blocks.2.resnets.0.conv2.weight",
+    "first_stage_model.encoder.down.2.block.0.nin_shortcut.bias": "encoder.down_blocks.2.resnets.0.conv_shortcut.bias",
+    "first_stage_model.encoder.down.2.block.0.nin_shortcut.weight": "encoder.down_blocks.2.resnets.0.conv_shortcut.weight",
+    "first_stage_model.encoder.down.2.block.0.norm1.bias": "encoder.down_blocks.2.resnets.0.norm1.bias",
+    "first_stage_model.encoder.down.2.block.0.norm1.weight": "encoder.down_blocks.2.resnets.0.norm1.weight",
+    "first_stage_model.encoder.down.2.block.0.norm2.bias": "encoder.down_blocks.2.resnets.0.norm2.bias",
+    "first_stage_model.encoder.down.2.block.0.norm2.weight": "encoder.down_blocks.2.resnets.0.norm2.weight",
+    "first_stage_model.encoder.down.2.block.1.conv1.bias": "encoder.down_blocks.2.resnets.1.conv1.bias",
+    "first_stage_model.encoder.down.2.block.1.conv1.weight": "encoder.down_blocks.2.resnets.1.conv1.weight",
+    "first_stage_model.encoder.down.2.block.1.conv2.bias": "encoder.down_blocks.2.resnets.1.conv2.bias",
+    "first_stage_model.encoder.down.2.block.1.conv2.weight": "encoder.down_blocks.2.resnets.1.conv2.weight",
+    "first_stage_model.encoder.down.2.block.1.norm1.bias": "encoder.down_blocks.2.resnets.1.norm1.bias",
+    "first_stage_model.encoder.down.2.block.1.norm1.weight": "encoder.down_blocks.2.resnets.1.norm1.weight",
+    "first_stage_model.encoder.down.2.block.1.norm2.bias": "encoder.down_blocks.2.resnets.1.norm2.bias",
+    "first_stage_model.encoder.down.2.block.1.norm2.weight": "encoder.down_blocks.2.resnets.1.norm2.weight",
+    "first_stage_model.encoder.down.3.block.0.conv1.bias": "encoder.down_blocks.3.resnets.0.conv1.bias",
+    "first_stage_model.encoder.down.3.block.0.conv1.weight": "encoder.down_blocks.3.resnets.0.conv1.weight",
+    "first_stage_model.encoder.down.3.block.0.conv2.bias": "encoder.down_blocks.3.resnets.0.conv2.bias",
+    "first_stage_model.encoder.down.3.block.0.conv2.weight": "encoder.down_blocks.3.resnets.0.conv2.weight",
+    "first_stage_model.encoder.down.3.block.0.norm1.bias": "encoder.down_blocks.3.resnets.0.norm1.bias",
+    "first_stage_model.encoder.down.3.block.0.norm1.weight": "encoder.down_blocks.3.resnets.0.norm1.weight",
+    "first_stage_model.encoder.down.3.block.0.norm2.bias": "encoder.down_blocks.3.resnets.0.norm2.bias",
+    "first_stage_model.encoder.down.3.block.0.norm2.weight": "encoder.down_blocks.3.resnets.0.norm2.weight",
+    "first_stage_model.encoder.down.3.block.1.conv1.bias": "encoder.down_blocks.3.resnets.1.conv1.bias",
+    "first_stage_model.encoder.down.3.block.1.conv1.weight": "encoder.down_blocks.3.resnets.1.conv1.weight",
+    "first_stage_model.encoder.down.3.block.1.conv2.bias": "encoder.down_blocks.3.resnets.1.conv2.bias",
+    "first_stage_model.encoder.down.3.block.1.conv2.weight": "encoder.down_blocks.3.resnets.1.conv2.weight",
+    "first_stage_model.encoder.down.3.block.1.norm1.bias": "encoder.down_blocks.3.resnets.1.norm1.bias",
+    "first_stage_model.encoder.down.3.block.1.norm1.weight": "encoder.down_blocks.3.resnets.1.norm1.weight",
+    "first_stage_model.encoder.down.3.block.1.norm2.bias": "encoder.down_blocks.3.resnets.1.norm2.bias",
+    "first_stage_model.encoder.down.3.block.1.norm2.weight": "encoder.down_blocks.3.resnets.1.norm2.weight",
+    "first_stage_model.encoder.mid.attn_1.norm.bias": "encoder.mid_block.attentions.0.group_norm.bias",
+    "first_stage_model.encoder.mid.attn_1.norm.weight": "encoder.mid_block.attentions.0.group_norm.weight",
+    "first_stage_model.encoder.mid.attn_1.k.bias": "encoder.mid_block.attentions.0.to_k.bias",
+    "first_stage_model.encoder.mid.attn_1.k.weight": "encoder.mid_block.attentions.0.to_k.weight",
+    "first_stage_model.encoder.mid.attn_1.proj_out.bias": "encoder.mid_block.attentions.0.to_out.0.bias",
+    "first_stage_model.encoder.mid.attn_1.proj_out.weight": "encoder.mid_block.attentions.0.to_out.0.weight",
+    "first_stage_model.encoder.mid.attn_1.q.bias": "encoder.mid_block.attentions.0.to_q.bias",
+    "first_stage_model.encoder.mid.attn_1.q.weight": "encoder.mid_block.attentions.0.to_q.weight",
+    "first_stage_model.encoder.mid.attn_1.v.bias": "encoder.mid_block.attentions.0.to_v.bias",
+    "first_stage_model.encoder.mid.attn_1.v.weight": "encoder.mid_block.attentions.0.to_v.weight",
+    "first_stage_model.encoder.mid.block_1.conv1.bias": "encoder.mid_block.resnets.0.conv1.bias",
+    "first_stage_model.encoder.mid.block_1.conv1.weight": "encoder.mid_block.resnets.0.conv1.weight",
+    "first_stage_model.encoder.mid.block_1.conv2.bias": "encoder.mid_block.resnets.0.conv2.bias",
+    "first_stage_model.encoder.mid.block_1.conv2.weight": "encoder.mid_block.resnets.0.conv2.weight",
+    "first_stage_model.encoder.mid.block_1.norm1.bias": "encoder.mid_block.resnets.0.norm1.bias",
+    "first_stage_model.encoder.mid.block_1.norm1.weight": "encoder.mid_block.resnets.0.norm1.weight",
+    "first_stage_model.encoder.mid.block_1.norm2.bias": "encoder.mid_block.resnets.0.norm2.bias",
+    "first_stage_model.encoder.mid.block_1.norm2.weight": "encoder.mid_block.resnets.0.norm2.weight",
+    "first_stage_model.encoder.mid.block_2.conv1.bias": "encoder.mid_block.resnets.1.conv1.bias",
+    "first_stage_model.encoder.mid.block_2.conv1.weight": "encoder.mid_block.resnets.1.conv1.weight",
+    "first_stage_model.encoder.mid.block_2.conv2.bias": "encoder.mid_block.resnets.1.conv2.bias",
+    "first_stage_model.encoder.mid.block_2.conv2.weight": "encoder.mid_block.resnets.1.conv2.weight",
+    "first_stage_model.encoder.mid.block_2.norm1.bias": "encoder.mid_block.resnets.1.norm1.bias",
+    "first_stage_model.encoder.mid.block_2.norm1.weight": "encoder.mid_block.resnets.1.norm1.weight",
+    "first_stage_model.encoder.mid.block_2.norm2.bias": "encoder.mid_block.resnets.1.norm2.bias",
+    "first_stage_model.encoder.mid.block_2.norm2.weight": "encoder.mid_block.resnets.1.norm2.weight",
+    "first_stage_model.post_quant_conv.bias": "post_quant_conv.bias",
+    "first_stage_model.post_quant_conv.weight": "post_quant_conv.weight",
+    "first_stage_model.quant_conv.bias": "quant_conv.bias",
+    "first_stage_model.quant_conv.weight": "quant_conv.weight"
+  },
+  "regex_map": {},
+  "ignore_prefixes": [],
+  "source_aliases": {},
+  "reshapes": {}
+}

imaginairy/weight_management/weight_maps/Compvis-OpenClipTextEncoder-SDXL-to-Diffusers.weightmap.json

@@ -0,0 +1,527 @@
+{
+  "name_map": {
+    "conditioner.embedders.1.model.positional_embedding": "text_model.embeddings.position_embedding.weight",
+    "conditioner.embedders.1.model.token_embedding.weight": "text_model.embeddings.token_embedding.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.ln_1.bias": "text_model.encoder.layers.0.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.ln_1.weight": "text_model.encoder.layers.0.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.ln_2.bias": "text_model.encoder.layers.0.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.ln_2.weight": "text_model.encoder.layers.0.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.mlp.c_fc.bias": "text_model.encoder.layers.0.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.mlp.c_fc.weight": "text_model.encoder.layers.0.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.mlp.c_proj.bias": "text_model.encoder.layers.0.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.mlp.c_proj.weight": "text_model.encoder.layers.0.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.out_proj.bias": "text_model.encoder.layers.0.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.out_proj.weight": "text_model.encoder.layers.0.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.ln_1.bias": "text_model.encoder.layers.1.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.ln_1.weight": "text_model.encoder.layers.1.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.ln_2.bias": "text_model.encoder.layers.1.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.ln_2.weight": "text_model.encoder.layers.1.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.mlp.c_fc.bias": "text_model.encoder.layers.1.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.mlp.c_fc.weight": "text_model.encoder.layers.1.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.mlp.c_proj.bias": "text_model.encoder.layers.1.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.mlp.c_proj.weight": "text_model.encoder.layers.1.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.out_proj.bias": "text_model.encoder.layers.1.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.out_proj.weight": "text_model.encoder.layers.1.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.ln_1.bias": "text_model.encoder.layers.10.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.ln_1.weight": "text_model.encoder.layers.10.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.ln_2.bias": "text_model.encoder.layers.10.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.ln_2.weight": "text_model.encoder.layers.10.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.mlp.c_fc.bias": "text_model.encoder.layers.10.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.mlp.c_fc.weight": "text_model.encoder.layers.10.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.mlp.c_proj.bias": "text_model.encoder.layers.10.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.mlp.c_proj.weight": "text_model.encoder.layers.10.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.out_proj.bias": "text_model.encoder.layers.10.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.out_proj.weight": "text_model.encoder.layers.10.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.ln_1.bias": "text_model.encoder.layers.11.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.ln_1.weight": "text_model.encoder.layers.11.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.ln_2.bias": "text_model.encoder.layers.11.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.ln_2.weight": "text_model.encoder.layers.11.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.mlp.c_fc.bias": "text_model.encoder.layers.11.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.mlp.c_fc.weight": "text_model.encoder.layers.11.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.mlp.c_proj.bias": "text_model.encoder.layers.11.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.mlp.c_proj.weight": "text_model.encoder.layers.11.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.out_proj.bias": "text_model.encoder.layers.11.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.out_proj.weight": "text_model.encoder.layers.11.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.ln_1.bias": "text_model.encoder.layers.12.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.ln_1.weight": "text_model.encoder.layers.12.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.ln_2.bias": "text_model.encoder.layers.12.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.ln_2.weight": "text_model.encoder.layers.12.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.mlp.c_fc.bias": "text_model.encoder.layers.12.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.mlp.c_fc.weight": "text_model.encoder.layers.12.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.mlp.c_proj.bias": "text_model.encoder.layers.12.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.mlp.c_proj.weight": "text_model.encoder.layers.12.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.out_proj.bias": "text_model.encoder.layers.12.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.out_proj.weight": "text_model.encoder.layers.12.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.ln_1.bias": "text_model.encoder.layers.13.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.ln_1.weight": "text_model.encoder.layers.13.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.ln_2.bias": "text_model.encoder.layers.13.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.ln_2.weight": "text_model.encoder.layers.13.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.mlp.c_fc.bias": "text_model.encoder.layers.13.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.mlp.c_fc.weight": "text_model.encoder.layers.13.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.mlp.c_proj.bias": "text_model.encoder.layers.13.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.mlp.c_proj.weight": "text_model.encoder.layers.13.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.out_proj.bias": "text_model.encoder.layers.13.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.out_proj.weight": "text_model.encoder.layers.13.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.ln_1.bias": "text_model.encoder.layers.14.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.ln_1.weight": "text_model.encoder.layers.14.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.ln_2.bias": "text_model.encoder.layers.14.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.ln_2.weight": "text_model.encoder.layers.14.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.mlp.c_fc.bias": "text_model.encoder.layers.14.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.mlp.c_fc.weight": "text_model.encoder.layers.14.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.mlp.c_proj.bias": "text_model.encoder.layers.14.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.mlp.c_proj.weight": "text_model.encoder.layers.14.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.out_proj.bias": "text_model.encoder.layers.14.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.out_proj.weight": "text_model.encoder.layers.14.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.ln_1.bias": "text_model.encoder.layers.15.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.ln_1.weight": "text_model.encoder.layers.15.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.ln_2.bias": "text_model.encoder.layers.15.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.ln_2.weight": "text_model.encoder.layers.15.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.mlp.c_fc.bias": "text_model.encoder.layers.15.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.mlp.c_fc.weight": "text_model.encoder.layers.15.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.mlp.c_proj.bias": "text_model.encoder.layers.15.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.mlp.c_proj.weight": "text_model.encoder.layers.15.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.out_proj.bias": "text_model.encoder.layers.15.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.out_proj.weight": "text_model.encoder.layers.15.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.ln_1.bias": "text_model.encoder.layers.16.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.ln_1.weight": "text_model.encoder.layers.16.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.ln_2.bias": "text_model.encoder.layers.16.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.ln_2.weight": "text_model.encoder.layers.16.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.mlp.c_fc.bias": "text_model.encoder.layers.16.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.mlp.c_fc.weight": "text_model.encoder.layers.16.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.mlp.c_proj.bias": "text_model.encoder.layers.16.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.mlp.c_proj.weight": "text_model.encoder.layers.16.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.out_proj.bias": "text_model.encoder.layers.16.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.out_proj.weight": "text_model.encoder.layers.16.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.ln_1.bias": "text_model.encoder.layers.17.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.ln_1.weight": "text_model.encoder.layers.17.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.ln_2.bias": "text_model.encoder.layers.17.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.ln_2.weight": "text_model.encoder.layers.17.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.mlp.c_fc.bias": "text_model.encoder.layers.17.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.mlp.c_fc.weight": "text_model.encoder.layers.17.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.mlp.c_proj.bias": "text_model.encoder.layers.17.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.mlp.c_proj.weight": "text_model.encoder.layers.17.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.out_proj.bias": "text_model.encoder.layers.17.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.out_proj.weight": "text_model.encoder.layers.17.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.ln_1.bias": "text_model.encoder.layers.18.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.ln_1.weight": "text_model.encoder.layers.18.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.ln_2.bias": "text_model.encoder.layers.18.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.ln_2.weight": "text_model.encoder.layers.18.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.mlp.c_fc.bias": "text_model.encoder.layers.18.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.mlp.c_fc.weight": "text_model.encoder.layers.18.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.mlp.c_proj.bias": "text_model.encoder.layers.18.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.mlp.c_proj.weight": "text_model.encoder.layers.18.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.out_proj.bias": "text_model.encoder.layers.18.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.out_proj.weight": "text_model.encoder.layers.18.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.ln_1.bias": "text_model.encoder.layers.19.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.ln_1.weight": "text_model.encoder.layers.19.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.ln_2.bias": "text_model.encoder.layers.19.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.ln_2.weight": "text_model.encoder.layers.19.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.mlp.c_fc.bias": "text_model.encoder.layers.19.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.mlp.c_fc.weight": "text_model.encoder.layers.19.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.mlp.c_proj.bias": "text_model.encoder.layers.19.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.mlp.c_proj.weight": "text_model.encoder.layers.19.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.out_proj.bias": "text_model.encoder.layers.19.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.out_proj.weight": "text_model.encoder.layers.19.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.ln_1.bias": "text_model.encoder.layers.2.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.ln_1.weight": "text_model.encoder.layers.2.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.ln_2.bias": "text_model.encoder.layers.2.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.ln_2.weight": "text_model.encoder.layers.2.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.mlp.c_fc.bias": "text_model.encoder.layers.2.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.mlp.c_fc.weight": "text_model.encoder.layers.2.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.mlp.c_proj.bias": "text_model.encoder.layers.2.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.mlp.c_proj.weight": "text_model.encoder.layers.2.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.out_proj.bias": "text_model.encoder.layers.2.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.out_proj.weight": "text_model.encoder.layers.2.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.ln_1.bias": "text_model.encoder.layers.20.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.ln_1.weight": "text_model.encoder.layers.20.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.ln_2.bias": "text_model.encoder.layers.20.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.ln_2.weight": "text_model.encoder.layers.20.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.mlp.c_fc.bias": "text_model.encoder.layers.20.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.mlp.c_fc.weight": "text_model.encoder.layers.20.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.mlp.c_proj.bias": "text_model.encoder.layers.20.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.mlp.c_proj.weight": "text_model.encoder.layers.20.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.out_proj.bias": "text_model.encoder.layers.20.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.out_proj.weight": "text_model.encoder.layers.20.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.ln_1.bias": "text_model.encoder.layers.21.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.ln_1.weight": "text_model.encoder.layers.21.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.ln_2.bias": "text_model.encoder.layers.21.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.ln_2.weight": "text_model.encoder.layers.21.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.mlp.c_fc.bias": "text_model.encoder.layers.21.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.mlp.c_fc.weight": "text_model.encoder.layers.21.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.mlp.c_proj.bias": "text_model.encoder.layers.21.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.mlp.c_proj.weight": "text_model.encoder.layers.21.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.out_proj.bias": "text_model.encoder.layers.21.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.out_proj.weight": "text_model.encoder.layers.21.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.ln_1.bias": "text_model.encoder.layers.22.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.ln_1.weight": "text_model.encoder.layers.22.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.ln_2.bias": "text_model.encoder.layers.22.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.ln_2.weight": "text_model.encoder.layers.22.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.mlp.c_fc.bias": "text_model.encoder.layers.22.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.mlp.c_fc.weight": "text_model.encoder.layers.22.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.mlp.c_proj.bias": "text_model.encoder.layers.22.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.mlp.c_proj.weight": "text_model.encoder.layers.22.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.out_proj.bias": "text_model.encoder.layers.22.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.out_proj.weight": "text_model.encoder.layers.22.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.ln_1.bias": "text_model.encoder.layers.23.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.ln_1.weight": "text_model.encoder.layers.23.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.ln_2.bias": "text_model.encoder.layers.23.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.ln_2.weight": "text_model.encoder.layers.23.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.mlp.c_fc.bias": "text_model.encoder.layers.23.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.mlp.c_fc.weight": "text_model.encoder.layers.23.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.mlp.c_proj.bias": "text_model.encoder.layers.23.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.mlp.c_proj.weight": "text_model.encoder.layers.23.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.out_proj.bias": "text_model.encoder.layers.23.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.out_proj.weight": "text_model.encoder.layers.23.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.ln_1.bias": "text_model.encoder.layers.24.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.ln_1.weight": "text_model.encoder.layers.24.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.ln_2.bias": "text_model.encoder.layers.24.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.ln_2.weight": "text_model.encoder.layers.24.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.mlp.c_fc.bias": "text_model.encoder.layers.24.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.mlp.c_fc.weight": "text_model.encoder.layers.24.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.mlp.c_proj.bias": "text_model.encoder.layers.24.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.mlp.c_proj.weight": "text_model.encoder.layers.24.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.out_proj.bias": "text_model.encoder.layers.24.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.out_proj.weight": "text_model.encoder.layers.24.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.ln_1.bias": "text_model.encoder.layers.25.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.ln_1.weight": "text_model.encoder.layers.25.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.ln_2.bias": "text_model.encoder.layers.25.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.ln_2.weight": "text_model.encoder.layers.25.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.mlp.c_fc.bias": "text_model.encoder.layers.25.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.mlp.c_fc.weight": "text_model.encoder.layers.25.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.mlp.c_proj.bias": "text_model.encoder.layers.25.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.mlp.c_proj.weight": "text_model.encoder.layers.25.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.out_proj.bias": "text_model.encoder.layers.25.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.out_proj.weight": "text_model.encoder.layers.25.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.ln_1.bias": "text_model.encoder.layers.26.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.ln_1.weight": "text_model.encoder.layers.26.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.ln_2.bias": "text_model.encoder.layers.26.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.ln_2.weight": "text_model.encoder.layers.26.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.mlp.c_fc.bias": "text_model.encoder.layers.26.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.mlp.c_fc.weight": "text_model.encoder.layers.26.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.mlp.c_proj.bias": "text_model.encoder.layers.26.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.mlp.c_proj.weight": "text_model.encoder.layers.26.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.out_proj.bias": "text_model.encoder.layers.26.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.out_proj.weight": "text_model.encoder.layers.26.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.ln_1.bias": "text_model.encoder.layers.27.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.ln_1.weight": "text_model.encoder.layers.27.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.ln_2.bias": "text_model.encoder.layers.27.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.ln_2.weight": "text_model.encoder.layers.27.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.mlp.c_fc.bias": "text_model.encoder.layers.27.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.mlp.c_fc.weight": "text_model.encoder.layers.27.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.mlp.c_proj.bias": "text_model.encoder.layers.27.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.mlp.c_proj.weight": "text_model.encoder.layers.27.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.out_proj.bias": "text_model.encoder.layers.27.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.out_proj.weight": "text_model.encoder.layers.27.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.ln_1.bias": "text_model.encoder.layers.28.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.ln_1.weight": "text_model.encoder.layers.28.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.ln_2.bias": "text_model.encoder.layers.28.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.ln_2.weight": "text_model.encoder.layers.28.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.mlp.c_fc.bias": "text_model.encoder.layers.28.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.mlp.c_fc.weight": "text_model.encoder.layers.28.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.mlp.c_proj.bias": "text_model.encoder.layers.28.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.mlp.c_proj.weight": "text_model.encoder.layers.28.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.out_proj.bias": "text_model.encoder.layers.28.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.out_proj.weight": "text_model.encoder.layers.28.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.ln_1.bias": "text_model.encoder.layers.29.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.ln_1.weight": "text_model.encoder.layers.29.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.ln_2.bias": "text_model.encoder.layers.29.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.ln_2.weight": "text_model.encoder.layers.29.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.mlp.c_fc.bias": "text_model.encoder.layers.29.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.mlp.c_fc.weight": "text_model.encoder.layers.29.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.mlp.c_proj.bias": "text_model.encoder.layers.29.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.mlp.c_proj.weight": "text_model.encoder.layers.29.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.out_proj.bias": "text_model.encoder.layers.29.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.out_proj.weight": "text_model.encoder.layers.29.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.ln_1.bias": "text_model.encoder.layers.3.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.ln_1.weight": "text_model.encoder.layers.3.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.ln_2.bias": "text_model.encoder.layers.3.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.ln_2.weight": "text_model.encoder.layers.3.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.mlp.c_fc.bias": "text_model.encoder.layers.3.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.mlp.c_fc.weight": "text_model.encoder.layers.3.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.mlp.c_proj.bias": "text_model.encoder.layers.3.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.mlp.c_proj.weight": "text_model.encoder.layers.3.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.out_proj.bias": "text_model.encoder.layers.3.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.out_proj.weight": "text_model.encoder.layers.3.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.ln_1.bias": "text_model.encoder.layers.30.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.ln_1.weight": "text_model.encoder.layers.30.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.ln_2.bias": "text_model.encoder.layers.30.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.ln_2.weight": "text_model.encoder.layers.30.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.mlp.c_fc.bias": "text_model.encoder.layers.30.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.mlp.c_fc.weight": "text_model.encoder.layers.30.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.mlp.c_proj.bias": "text_model.encoder.layers.30.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.mlp.c_proj.weight": "text_model.encoder.layers.30.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.out_proj.bias": "text_model.encoder.layers.30.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.out_proj.weight": "text_model.encoder.layers.30.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.ln_1.bias": "text_model.encoder.layers.31.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.ln_1.weight": "text_model.encoder.layers.31.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.ln_2.bias": "text_model.encoder.layers.31.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.ln_2.weight": "text_model.encoder.layers.31.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.mlp.c_fc.bias": "text_model.encoder.layers.31.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.mlp.c_fc.weight": "text_model.encoder.layers.31.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.mlp.c_proj.bias": "text_model.encoder.layers.31.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.mlp.c_proj.weight": "text_model.encoder.layers.31.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.out_proj.bias": "text_model.encoder.layers.31.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.out_proj.weight": "text_model.encoder.layers.31.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.ln_1.bias": "text_model.encoder.layers.4.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.ln_1.weight": "text_model.encoder.layers.4.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.ln_2.bias": "text_model.encoder.layers.4.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.ln_2.weight": "text_model.encoder.layers.4.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.mlp.c_fc.bias": "text_model.encoder.layers.4.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.mlp.c_fc.weight": "text_model.encoder.layers.4.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.mlp.c_proj.bias": "text_model.encoder.layers.4.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.mlp.c_proj.weight": "text_model.encoder.layers.4.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.out_proj.bias": "text_model.encoder.layers.4.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.out_proj.weight": "text_model.encoder.layers.4.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.ln_1.bias": "text_model.encoder.layers.5.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.ln_1.weight": "text_model.encoder.layers.5.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.ln_2.bias": "text_model.encoder.layers.5.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.ln_2.weight": "text_model.encoder.layers.5.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.mlp.c_fc.bias": "text_model.encoder.layers.5.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.mlp.c_fc.weight": "text_model.encoder.layers.5.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.mlp.c_proj.bias": "text_model.encoder.layers.5.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.mlp.c_proj.weight": "text_model.encoder.layers.5.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.out_proj.bias": "text_model.encoder.layers.5.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.out_proj.weight": "text_model.encoder.layers.5.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.ln_1.bias": "text_model.encoder.layers.6.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.ln_1.weight": "text_model.encoder.layers.6.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.ln_2.bias": "text_model.encoder.layers.6.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.ln_2.weight": "text_model.encoder.layers.6.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.mlp.c_fc.bias": "text_model.encoder.layers.6.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.mlp.c_fc.weight": "text_model.encoder.layers.6.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.mlp.c_proj.bias": "text_model.encoder.layers.6.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.mlp.c_proj.weight": "text_model.encoder.layers.6.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.out_proj.bias": "text_model.encoder.layers.6.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.out_proj.weight": "text_model.encoder.layers.6.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.ln_1.bias": "text_model.encoder.layers.7.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.ln_1.weight": "text_model.encoder.layers.7.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.ln_2.bias": "text_model.encoder.layers.7.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.ln_2.weight": "text_model.encoder.layers.7.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.mlp.c_fc.bias": "text_model.encoder.layers.7.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.mlp.c_fc.weight": "text_model.encoder.layers.7.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.mlp.c_proj.bias": "text_model.encoder.layers.7.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.mlp.c_proj.weight": "text_model.encoder.layers.7.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.out_proj.bias": "text_model.encoder.layers.7.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.out_proj.weight": "text_model.encoder.layers.7.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.ln_1.bias": "text_model.encoder.layers.8.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.ln_1.weight": "text_model.encoder.layers.8.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.ln_2.bias": "text_model.encoder.layers.8.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.ln_2.weight": "text_model.encoder.layers.8.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.mlp.c_fc.bias": "text_model.encoder.layers.8.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.mlp.c_fc.weight": "text_model.encoder.layers.8.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.mlp.c_proj.bias": "text_model.encoder.layers.8.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.mlp.c_proj.weight": "text_model.encoder.layers.8.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.out_proj.bias": "text_model.encoder.layers.8.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.out_proj.weight": "text_model.encoder.layers.8.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.ln_1.bias": "text_model.encoder.layers.9.layer_norm1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.ln_1.weight": "text_model.encoder.layers.9.layer_norm1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.ln_2.bias": "text_model.encoder.layers.9.layer_norm2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.ln_2.weight": "text_model.encoder.layers.9.layer_norm2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.mlp.c_fc.bias": "text_model.encoder.layers.9.mlp.fc1.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.mlp.c_fc.weight": "text_model.encoder.layers.9.mlp.fc1.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.mlp.c_proj.bias": "text_model.encoder.layers.9.mlp.fc2.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.mlp.c_proj.weight": "text_model.encoder.layers.9.mlp.fc2.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.out_proj.bias": "text_model.encoder.layers.9.self_attn.out_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.out_proj.weight": "text_model.encoder.layers.9.self_attn.out_proj.weight",
+    "conditioner.embedders.1.model.ln_final.bias": "text_model.final_layer_norm.bias",
+    "conditioner.embedders.1.model.ln_final.weight": "text_model.final_layer_norm.weight",
+    "conditioner.embedders.1.model.text_projection.weight": "text_projection.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.in_proj_weight.0": "text_model.encoder.layers.0.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.in_proj_weight.1": "text_model.encoder.layers.0.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.in_proj_weight.2": "text_model.encoder.layers.0.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.in_proj_weight.0": "text_model.encoder.layers.1.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.in_proj_weight.1": "text_model.encoder.layers.1.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.in_proj_weight.2": "text_model.encoder.layers.1.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.in_proj_weight.0": "text_model.encoder.layers.10.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.in_proj_weight.1": "text_model.encoder.layers.10.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.in_proj_weight.2": "text_model.encoder.layers.10.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.in_proj_weight.0": "text_model.encoder.layers.11.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.in_proj_weight.1": "text_model.encoder.layers.11.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.in_proj_weight.2": "text_model.encoder.layers.11.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.in_proj_weight.0": "text_model.encoder.layers.12.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.in_proj_weight.1": "text_model.encoder.layers.12.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.in_proj_weight.2": "text_model.encoder.layers.12.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.in_proj_weight.0": "text_model.encoder.layers.13.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.in_proj_weight.1": "text_model.encoder.layers.13.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.in_proj_weight.2": "text_model.encoder.layers.13.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.in_proj_weight.0": "text_model.encoder.layers.14.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.in_proj_weight.1": "text_model.encoder.layers.14.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.in_proj_weight.2": "text_model.encoder.layers.14.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.in_proj_weight.0": "text_model.encoder.layers.15.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.in_proj_weight.1": "text_model.encoder.layers.15.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.in_proj_weight.2": "text_model.encoder.layers.15.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.in_proj_weight.0": "text_model.encoder.layers.16.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.in_proj_weight.1": "text_model.encoder.layers.16.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.in_proj_weight.2": "text_model.encoder.layers.16.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.in_proj_weight.0": "text_model.encoder.layers.17.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.in_proj_weight.1": "text_model.encoder.layers.17.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.in_proj_weight.2": "text_model.encoder.layers.17.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.in_proj_weight.0": "text_model.encoder.layers.18.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.in_proj_weight.1": "text_model.encoder.layers.18.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.in_proj_weight.2": "text_model.encoder.layers.18.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.in_proj_weight.0": "text_model.encoder.layers.19.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.in_proj_weight.1": "text_model.encoder.layers.19.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.in_proj_weight.2": "text_model.encoder.layers.19.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.in_proj_weight.0": "text_model.encoder.layers.2.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.in_proj_weight.1": "text_model.encoder.layers.2.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.in_proj_weight.2": "text_model.encoder.layers.2.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.in_proj_weight.0": "text_model.encoder.layers.20.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.in_proj_weight.1": "text_model.encoder.layers.20.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.in_proj_weight.2": "text_model.encoder.layers.20.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.in_proj_weight.0": "text_model.encoder.layers.21.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.in_proj_weight.1": "text_model.encoder.layers.21.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.in_proj_weight.2": "text_model.encoder.layers.21.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.in_proj_weight.0": "text_model.encoder.layers.22.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.in_proj_weight.1": "text_model.encoder.layers.22.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.in_proj_weight.2": "text_model.encoder.layers.22.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.in_proj_weight.0": "text_model.encoder.layers.23.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.in_proj_weight.1": "text_model.encoder.layers.23.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.in_proj_weight.2": "text_model.encoder.layers.23.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.in_proj_weight.0": "text_model.encoder.layers.24.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.in_proj_weight.1": "text_model.encoder.layers.24.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.in_proj_weight.2": "text_model.encoder.layers.24.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.in_proj_weight.0": "text_model.encoder.layers.25.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.in_proj_weight.1": "text_model.encoder.layers.25.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.in_proj_weight.2": "text_model.encoder.layers.25.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.in_proj_weight.0": "text_model.encoder.layers.26.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.in_proj_weight.1": "text_model.encoder.layers.26.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.in_proj_weight.2": "text_model.encoder.layers.26.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.in_proj_weight.0": "text_model.encoder.layers.27.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.in_proj_weight.1": "text_model.encoder.layers.27.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.in_proj_weight.2": "text_model.encoder.layers.27.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.in_proj_weight.0": "text_model.encoder.layers.28.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.in_proj_weight.1": "text_model.encoder.layers.28.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.in_proj_weight.2": "text_model.encoder.layers.28.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.in_proj_weight.0": "text_model.encoder.layers.29.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.in_proj_weight.1": "text_model.encoder.layers.29.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.in_proj_weight.2": "text_model.encoder.layers.29.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.in_proj_weight.0": "text_model.encoder.layers.3.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.in_proj_weight.1": "text_model.encoder.layers.3.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.in_proj_weight.2": "text_model.encoder.layers.3.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.in_proj_weight.0": "text_model.encoder.layers.30.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.in_proj_weight.1": "text_model.encoder.layers.30.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.in_proj_weight.2": "text_model.encoder.layers.30.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.in_proj_weight.0": "text_model.encoder.layers.31.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.in_proj_weight.1": "text_model.encoder.layers.31.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.in_proj_weight.2": "text_model.encoder.layers.31.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.in_proj_weight.0": "text_model.encoder.layers.4.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.in_proj_weight.1": "text_model.encoder.layers.4.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.in_proj_weight.2": "text_model.encoder.layers.4.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.in_proj_weight.0": "text_model.encoder.layers.5.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.in_proj_weight.1": "text_model.encoder.layers.5.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.in_proj_weight.2": "text_model.encoder.layers.5.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.in_proj_weight.0": "text_model.encoder.layers.6.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.in_proj_weight.1": "text_model.encoder.layers.6.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.in_proj_weight.2": "text_model.encoder.layers.6.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.in_proj_weight.0": "text_model.encoder.layers.7.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.in_proj_weight.1": "text_model.encoder.layers.7.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.in_proj_weight.2": "text_model.encoder.layers.7.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.in_proj_weight.0": "text_model.encoder.layers.8.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.in_proj_weight.1": "text_model.encoder.layers.8.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.in_proj_weight.2": "text_model.encoder.layers.8.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.in_proj_weight.0": "text_model.encoder.layers.9.self_attn.q_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.in_proj_weight.1": "text_model.encoder.layers.9.self_attn.k_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.in_proj_weight.2": "text_model.encoder.layers.9.self_attn.v_proj.weight",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.in_proj_bias.0": "text_model.encoder.layers.0.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.in_proj_bias.1": "text_model.encoder.layers.0.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.0.attn.in_proj_bias.2": "text_model.encoder.layers.0.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.in_proj_bias.0": "text_model.encoder.layers.1.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.in_proj_bias.1": "text_model.encoder.layers.1.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.1.attn.in_proj_bias.2": "text_model.encoder.layers.1.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.in_proj_bias.0": "text_model.encoder.layers.10.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.in_proj_bias.1": "text_model.encoder.layers.10.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.10.attn.in_proj_bias.2": "text_model.encoder.layers.10.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.in_proj_bias.0": "text_model.encoder.layers.11.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.in_proj_bias.1": "text_model.encoder.layers.11.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.11.attn.in_proj_bias.2": "text_model.encoder.layers.11.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.in_proj_bias.0": "text_model.encoder.layers.12.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.in_proj_bias.1": "text_model.encoder.layers.12.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.12.attn.in_proj_bias.2": "text_model.encoder.layers.12.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.in_proj_bias.0": "text_model.encoder.layers.13.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.in_proj_bias.1": "text_model.encoder.layers.13.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.13.attn.in_proj_bias.2": "text_model.encoder.layers.13.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.in_proj_bias.0": "text_model.encoder.layers.14.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.in_proj_bias.1": "text_model.encoder.layers.14.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.14.attn.in_proj_bias.2": "text_model.encoder.layers.14.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.in_proj_bias.0": "text_model.encoder.layers.15.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.in_proj_bias.1": "text_model.encoder.layers.15.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.15.attn.in_proj_bias.2": "text_model.encoder.layers.15.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.in_proj_bias.0": "text_model.encoder.layers.16.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.in_proj_bias.1": "text_model.encoder.layers.16.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.16.attn.in_proj_bias.2": "text_model.encoder.layers.16.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.in_proj_bias.0": "text_model.encoder.layers.17.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.in_proj_bias.1": "text_model.encoder.layers.17.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.17.attn.in_proj_bias.2": "text_model.encoder.layers.17.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.in_proj_bias.0": "text_model.encoder.layers.18.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.in_proj_bias.1": "text_model.encoder.layers.18.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.18.attn.in_proj_bias.2": "text_model.encoder.layers.18.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.in_proj_bias.0": "text_model.encoder.layers.19.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.in_proj_bias.1": "text_model.encoder.layers.19.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.19.attn.in_proj_bias.2": "text_model.encoder.layers.19.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.in_proj_bias.0": "text_model.encoder.layers.2.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.in_proj_bias.1": "text_model.encoder.layers.2.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.2.attn.in_proj_bias.2": "text_model.encoder.layers.2.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.in_proj_bias.0": "text_model.encoder.layers.20.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.in_proj_bias.1": "text_model.encoder.layers.20.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.20.attn.in_proj_bias.2": "text_model.encoder.layers.20.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.in_proj_bias.0": "text_model.encoder.layers.21.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.in_proj_bias.1": "text_model.encoder.layers.21.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.21.attn.in_proj_bias.2": "text_model.encoder.layers.21.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.in_proj_bias.0": "text_model.encoder.layers.22.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.in_proj_bias.1": "text_model.encoder.layers.22.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.22.attn.in_proj_bias.2": "text_model.encoder.layers.22.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.in_proj_bias.0": "text_model.encoder.layers.23.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.in_proj_bias.1": "text_model.encoder.layers.23.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.23.attn.in_proj_bias.2": "text_model.encoder.layers.23.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.in_proj_bias.0": "text_model.encoder.layers.24.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.in_proj_bias.1": "text_model.encoder.layers.24.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.24.attn.in_proj_bias.2": "text_model.encoder.layers.24.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.in_proj_bias.0": "text_model.encoder.layers.25.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.in_proj_bias.1": "text_model.encoder.layers.25.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.25.attn.in_proj_bias.2": "text_model.encoder.layers.25.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.in_proj_bias.0": "text_model.encoder.layers.26.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.in_proj_bias.1": "text_model.encoder.layers.26.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.26.attn.in_proj_bias.2": "text_model.encoder.layers.26.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.in_proj_bias.0": "text_model.encoder.layers.27.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.in_proj_bias.1": "text_model.encoder.layers.27.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.27.attn.in_proj_bias.2": "text_model.encoder.layers.27.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.in_proj_bias.0": "text_model.encoder.layers.28.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.in_proj_bias.1": "text_model.encoder.layers.28.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.28.attn.in_proj_bias.2": "text_model.encoder.layers.28.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.in_proj_bias.0": "text_model.encoder.layers.29.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.in_proj_bias.1": "text_model.encoder.layers.29.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.29.attn.in_proj_bias.2": "text_model.encoder.layers.29.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.in_proj_bias.0": "text_model.encoder.layers.3.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.in_proj_bias.1": "text_model.encoder.layers.3.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.3.attn.in_proj_bias.2": "text_model.encoder.layers.3.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.in_proj_bias.0": "text_model.encoder.layers.30.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.in_proj_bias.1": "text_model.encoder.layers.30.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.30.attn.in_proj_bias.2": "text_model.encoder.layers.30.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.in_proj_bias.0": "text_model.encoder.layers.31.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.in_proj_bias.1": "text_model.encoder.layers.31.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.31.attn.in_proj_bias.2": "text_model.encoder.layers.31.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.in_proj_bias.0": "text_model.encoder.layers.4.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.in_proj_bias.1": "text_model.encoder.layers.4.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.4.attn.in_proj_bias.2": "text_model.encoder.layers.4.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.in_proj_bias.0": "text_model.encoder.layers.5.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.in_proj_bias.1": "text_model.encoder.layers.5.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.5.attn.in_proj_bias.2": "text_model.encoder.layers.5.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.in_proj_bias.0": "text_model.encoder.layers.6.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.in_proj_bias.1": "text_model.encoder.layers.6.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.6.attn.in_proj_bias.2": "text_model.encoder.layers.6.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.in_proj_bias.0": "text_model.encoder.layers.7.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.in_proj_bias.1": "text_model.encoder.layers.7.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.7.attn.in_proj_bias.2": "text_model.encoder.layers.7.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.in_proj_bias.0": "text_model.encoder.layers.8.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.in_proj_bias.1": "text_model.encoder.layers.8.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.8.attn.in_proj_bias.2": "text_model.encoder.layers.8.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.in_proj_bias.0": "text_model.encoder.layers.9.self_attn.q_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.in_proj_bias.1": "text_model.encoder.layers.9.self_attn.k_proj.bias",
+    "conditioner.embedders.1.model.transformer.resblocks.9.attn.in_proj_bias.2": "text_model.encoder.layers.9.self_attn.v_proj.bias",
+    "conditioner.embedders.1.model.text_projection": "text_projection.weight",
+    "conditioner.embedders.1.model.logit_scale": null
+  },
+  "regex_map": {},
+  "ignore_prefixes": [],
+  "source_aliases": {},
+  "reshapes": {}
+}

imaginairy/weight_management/weight_maps/Compvis-TextEncoder-SDXL-to-Diffusers.weightmap.json

@@ -0,0 +1,206 @@
+{
+  "name_map": {
+    "conditioner.embedders.0.transformer.text_model.embeddings.position_embedding.weight": "text_model.embeddings.position_embedding.weight",
+    "conditioner.embedders.0.transformer.text_model.embeddings.token_embedding.weight": "text_model.embeddings.token_embedding.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.layer_norm1.bias": "text_model.encoder.layers.0.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.layer_norm1.weight": "text_model.encoder.layers.0.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.layer_norm2.bias": "text_model.encoder.layers.0.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.layer_norm2.weight": "text_model.encoder.layers.0.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.mlp.fc1.bias": "text_model.encoder.layers.0.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.mlp.fc1.weight": "text_model.encoder.layers.0.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.mlp.fc2.bias": "text_model.encoder.layers.0.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.mlp.fc2.weight": "text_model.encoder.layers.0.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.k_proj.bias": "text_model.encoder.layers.0.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.k_proj.weight": "text_model.encoder.layers.0.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.out_proj.bias": "text_model.encoder.layers.0.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.out_proj.weight": "text_model.encoder.layers.0.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.q_proj.bias": "text_model.encoder.layers.0.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.q_proj.weight": "text_model.encoder.layers.0.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.v_proj.bias": "text_model.encoder.layers.0.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.0.self_attn.v_proj.weight": "text_model.encoder.layers.0.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.layer_norm1.bias": "text_model.encoder.layers.1.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.layer_norm1.weight": "text_model.encoder.layers.1.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.layer_norm2.bias": "text_model.encoder.layers.1.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.layer_norm2.weight": "text_model.encoder.layers.1.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.mlp.fc1.bias": "text_model.encoder.layers.1.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.mlp.fc1.weight": "text_model.encoder.layers.1.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.mlp.fc2.bias": "text_model.encoder.layers.1.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.mlp.fc2.weight": "text_model.encoder.layers.1.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.k_proj.bias": "text_model.encoder.layers.1.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.k_proj.weight": "text_model.encoder.layers.1.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.out_proj.bias": "text_model.encoder.layers.1.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.out_proj.weight": "text_model.encoder.layers.1.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.q_proj.bias": "text_model.encoder.layers.1.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.q_proj.weight": "text_model.encoder.layers.1.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.v_proj.bias": "text_model.encoder.layers.1.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.1.self_attn.v_proj.weight": "text_model.encoder.layers.1.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.layer_norm1.bias": "text_model.encoder.layers.10.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.layer_norm1.weight": "text_model.encoder.layers.10.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.layer_norm2.bias": "text_model.encoder.layers.10.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.layer_norm2.weight": "text_model.encoder.layers.10.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.mlp.fc1.bias": "text_model.encoder.layers.10.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.mlp.fc1.weight": "text_model.encoder.layers.10.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.mlp.fc2.bias": "text_model.encoder.layers.10.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.mlp.fc2.weight": "text_model.encoder.layers.10.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.k_proj.bias": "text_model.encoder.layers.10.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.k_proj.weight": "text_model.encoder.layers.10.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.out_proj.bias": "text_model.encoder.layers.10.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.out_proj.weight": "text_model.encoder.layers.10.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.q_proj.bias": "text_model.encoder.layers.10.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.q_proj.weight": "text_model.encoder.layers.10.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.v_proj.bias": "text_model.encoder.layers.10.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.10.self_attn.v_proj.weight": "text_model.encoder.layers.10.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.layer_norm1.bias": "text_model.encoder.layers.11.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.layer_norm1.weight": "text_model.encoder.layers.11.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.layer_norm2.bias": "text_model.encoder.layers.11.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.layer_norm2.weight": "text_model.encoder.layers.11.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.mlp.fc1.bias": "text_model.encoder.layers.11.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.mlp.fc1.weight": "text_model.encoder.layers.11.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.mlp.fc2.bias": "text_model.encoder.layers.11.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.mlp.fc2.weight": "text_model.encoder.layers.11.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.k_proj.bias": "text_model.encoder.layers.11.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.k_proj.weight": "text_model.encoder.layers.11.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.out_proj.bias": "text_model.encoder.layers.11.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.out_proj.weight": "text_model.encoder.layers.11.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.q_proj.bias": "text_model.encoder.layers.11.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.q_proj.weight": "text_model.encoder.layers.11.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.v_proj.bias": "text_model.encoder.layers.11.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.11.self_attn.v_proj.weight": "text_model.encoder.layers.11.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.layer_norm1.bias": "text_model.encoder.layers.2.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.layer_norm1.weight": "text_model.encoder.layers.2.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.layer_norm2.bias": "text_model.encoder.layers.2.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.layer_norm2.weight": "text_model.encoder.layers.2.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.mlp.fc1.bias": "text_model.encoder.layers.2.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.mlp.fc1.weight": "text_model.encoder.layers.2.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.mlp.fc2.bias": "text_model.encoder.layers.2.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.mlp.fc2.weight": "text_model.encoder.layers.2.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.k_proj.bias": "text_model.encoder.layers.2.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.k_proj.weight": "text_model.encoder.layers.2.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.out_proj.bias": "text_model.encoder.layers.2.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.out_proj.weight": "text_model.encoder.layers.2.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.q_proj.bias": "text_model.encoder.layers.2.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.q_proj.weight": "text_model.encoder.layers.2.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.v_proj.bias": "text_model.encoder.layers.2.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.2.self_attn.v_proj.weight": "text_model.encoder.layers.2.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.layer_norm1.bias": "text_model.encoder.layers.3.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.layer_norm1.weight": "text_model.encoder.layers.3.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.layer_norm2.bias": "text_model.encoder.layers.3.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.layer_norm2.weight": "text_model.encoder.layers.3.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.mlp.fc1.bias": "text_model.encoder.layers.3.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.mlp.fc1.weight": "text_model.encoder.layers.3.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.mlp.fc2.bias": "text_model.encoder.layers.3.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.mlp.fc2.weight": "text_model.encoder.layers.3.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.k_proj.bias": "text_model.encoder.layers.3.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.k_proj.weight": "text_model.encoder.layers.3.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.out_proj.bias": "text_model.encoder.layers.3.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.out_proj.weight": "text_model.encoder.layers.3.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.q_proj.bias": "text_model.encoder.layers.3.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.q_proj.weight": "text_model.encoder.layers.3.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.v_proj.bias": "text_model.encoder.layers.3.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.3.self_attn.v_proj.weight": "text_model.encoder.layers.3.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.layer_norm1.bias": "text_model.encoder.layers.4.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.layer_norm1.weight": "text_model.encoder.layers.4.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.layer_norm2.bias": "text_model.encoder.layers.4.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.layer_norm2.weight": "text_model.encoder.layers.4.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.mlp.fc1.bias": "text_model.encoder.layers.4.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.mlp.fc1.weight": "text_model.encoder.layers.4.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.mlp.fc2.bias": "text_model.encoder.layers.4.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.mlp.fc2.weight": "text_model.encoder.layers.4.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.k_proj.bias": "text_model.encoder.layers.4.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.k_proj.weight": "text_model.encoder.layers.4.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.out_proj.bias": "text_model.encoder.layers.4.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.out_proj.weight": "text_model.encoder.layers.4.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.q_proj.bias": "text_model.encoder.layers.4.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.q_proj.weight": "text_model.encoder.layers.4.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.v_proj.bias": "text_model.encoder.layers.4.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.4.self_attn.v_proj.weight": "text_model.encoder.layers.4.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.layer_norm1.bias": "text_model.encoder.layers.5.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.layer_norm1.weight": "text_model.encoder.layers.5.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.layer_norm2.bias": "text_model.encoder.layers.5.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.layer_norm2.weight": "text_model.encoder.layers.5.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.mlp.fc1.bias": "text_model.encoder.layers.5.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.mlp.fc1.weight": "text_model.encoder.layers.5.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.mlp.fc2.bias": "text_model.encoder.layers.5.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.mlp.fc2.weight": "text_model.encoder.layers.5.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.k_proj.bias": "text_model.encoder.layers.5.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.k_proj.weight": "text_model.encoder.layers.5.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.out_proj.bias": "text_model.encoder.layers.5.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.out_proj.weight": "text_model.encoder.layers.5.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.q_proj.bias": "text_model.encoder.layers.5.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.q_proj.weight": "text_model.encoder.layers.5.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.v_proj.bias": "text_model.encoder.layers.5.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.5.self_attn.v_proj.weight": "text_model.encoder.layers.5.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.layer_norm1.bias": "text_model.encoder.layers.6.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.layer_norm1.weight": "text_model.encoder.layers.6.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.layer_norm2.bias": "text_model.encoder.layers.6.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.layer_norm2.weight": "text_model.encoder.layers.6.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.mlp.fc1.bias": "text_model.encoder.layers.6.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.mlp.fc1.weight": "text_model.encoder.layers.6.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.mlp.fc2.bias": "text_model.encoder.layers.6.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.mlp.fc2.weight": "text_model.encoder.layers.6.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.k_proj.bias": "text_model.encoder.layers.6.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.k_proj.weight": "text_model.encoder.layers.6.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.out_proj.bias": "text_model.encoder.layers.6.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.out_proj.weight": "text_model.encoder.layers.6.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.q_proj.bias": "text_model.encoder.layers.6.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.q_proj.weight": "text_model.encoder.layers.6.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.v_proj.bias": "text_model.encoder.layers.6.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.6.self_attn.v_proj.weight": "text_model.encoder.layers.6.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.layer_norm1.bias": "text_model.encoder.layers.7.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.layer_norm1.weight": "text_model.encoder.layers.7.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.layer_norm2.bias": "text_model.encoder.layers.7.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.layer_norm2.weight": "text_model.encoder.layers.7.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.mlp.fc1.bias": "text_model.encoder.layers.7.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.mlp.fc1.weight": "text_model.encoder.layers.7.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.mlp.fc2.bias": "text_model.encoder.layers.7.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.mlp.fc2.weight": "text_model.encoder.layers.7.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.k_proj.bias": "text_model.encoder.layers.7.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.k_proj.weight": "text_model.encoder.layers.7.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.out_proj.bias": "text_model.encoder.layers.7.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.out_proj.weight": "text_model.encoder.layers.7.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.q_proj.bias": "text_model.encoder.layers.7.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.q_proj.weight": "text_model.encoder.layers.7.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.v_proj.bias": "text_model.encoder.layers.7.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.7.self_attn.v_proj.weight": "text_model.encoder.layers.7.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.layer_norm1.bias": "text_model.encoder.layers.8.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.layer_norm1.weight": "text_model.encoder.layers.8.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.layer_norm2.bias": "text_model.encoder.layers.8.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.layer_norm2.weight": "text_model.encoder.layers.8.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.mlp.fc1.bias": "text_model.encoder.layers.8.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.mlp.fc1.weight": "text_model.encoder.layers.8.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.mlp.fc2.bias": "text_model.encoder.layers.8.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.mlp.fc2.weight": "text_model.encoder.layers.8.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.k_proj.bias": "text_model.encoder.layers.8.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.k_proj.weight": "text_model.encoder.layers.8.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.out_proj.bias": "text_model.encoder.layers.8.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.out_proj.weight": "text_model.encoder.layers.8.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.q_proj.bias": "text_model.encoder.layers.8.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.q_proj.weight": "text_model.encoder.layers.8.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.v_proj.bias": "text_model.encoder.layers.8.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.8.self_attn.v_proj.weight": "text_model.encoder.layers.8.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.layer_norm1.bias": "text_model.encoder.layers.9.layer_norm1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.layer_norm1.weight": "text_model.encoder.layers.9.layer_norm1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.layer_norm2.bias": "text_model.encoder.layers.9.layer_norm2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.layer_norm2.weight": "text_model.encoder.layers.9.layer_norm2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.mlp.fc1.bias": "text_model.encoder.layers.9.mlp.fc1.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.mlp.fc1.weight": "text_model.encoder.layers.9.mlp.fc1.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.mlp.fc2.bias": "text_model.encoder.layers.9.mlp.fc2.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.mlp.fc2.weight": "text_model.encoder.layers.9.mlp.fc2.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.k_proj.bias": "text_model.encoder.layers.9.self_attn.k_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.k_proj.weight": "text_model.encoder.layers.9.self_attn.k_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.out_proj.bias": "text_model.encoder.layers.9.self_attn.out_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.out_proj.weight": "text_model.encoder.layers.9.self_attn.out_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.q_proj.bias": "text_model.encoder.layers.9.self_attn.q_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.q_proj.weight": "text_model.encoder.layers.9.self_attn.q_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.v_proj.bias": "text_model.encoder.layers.9.self_attn.v_proj.bias",
+    "conditioner.embedders.0.transformer.text_model.encoder.layers.9.self_attn.v_proj.weight": "text_model.encoder.layers.9.self_attn.v_proj.weight",
+    "conditioner.embedders.0.transformer.text_model.final_layer_norm.bias": "text_model.final_layer_norm.bias",
+    "conditioner.embedders.0.transformer.text_model.final_layer_norm.weight": "text_model.final_layer_norm.weight",
+    "conditioner.embedders.0.text_projection": "text_projection.weight",
+    "conditioner.embedders.0.transformer.text_model.embeddings.position_ids": null
+  },
+  "regex_map": {},
+  "ignore_prefixes": [],
+  "source_aliases": {},
+  "reshapes": {}
+}

tox.ini

@@ -16,4 +16,8 @@ warn_unused_ignores = False
 
 [mypy-imaginairy.vendored.*]
 follow_imports = skip
+ignore_errors = True
+
+
+[mypy-logging.*]
 ignore_errors = True