dsipts 1.1.11__py3-none-any.whl → 1.1.12__py3-none-any.whl

dsipts/data_structure/data_structure.py

@@ -683,10 +683,7 @@ class TimeSeries():
         #self.model.apply(weight_init_zeros)
 
         self.config = config
-        try:
-            self.model = torch.compile(self.model)
-        except:
-            beauty_string('Can not compile the model','block',self.verbose)
+
 
         beauty_string('Setting the model','block',self.verbose)
         beauty_string(model,'',self.verbose)
@@ -812,8 +809,17 @@ class TimeSeries():
             weight_exists = False
             beauty_string('I can not load a previous model','section',self.verbose)
 
+        self.model.to(torch.device("cuda:0" if torch.cuda.is_available() else "cpu"))
+        if self.model.can_be_compiled():
+            try:
+                self.model = torch.compile(self.model)
+                beauty_string('Model COMPILED','block',self.verbose)
+
+            except:
+                beauty_string('Can not compile the model','block',self.verbose)
+        else:
+            beauty_string('Model can not still be compiled, be patient','block',self.verbose)
 
-        
         
         if OLD_PL:
             trainer = pl.Trainer(default_root_dir=dirpath,
@@ -895,10 +901,19 @@ class TimeSeries():
             self.losses = pd.DataFrame()
 
         try:
+
             if OLD_PL:
-                self.model = self.model.load_from_checkpoint(self.checkpoint_file_last)
+                if isinstance(self.model, torch._dynamo.eval_frame.OptimizedModule):
+                    self.model = self.model._orig_mod 
+                    self.model.load_from_checkpoint(self.checkpoint_file_last)
+                else:
+                    self.model = self.model.load_from_checkpoint(self.checkpoint_file_last)
             else:
-                self.model = self.model.__class__.load_from_checkpoint(self.checkpoint_file_last)
+                if isinstance(self.model, torch._dynamo.eval_frame.OptimizedModule):
+                    mm = self.model._orig_mod 
+                    self.model = mm.__class__.load_from_checkpoint(self.checkpoint_file_last)
+                else:
+                    self.model = self.model.__class__.load_from_checkpoint(self.checkpoint_file_last)
 
         except Exception as _:
             beauty_string(f'There is a problem loading the weights on file MAYBE CHANGED HOW WEIGHTS ARE LOADED {self.checkpoint_file_last}','section',self.verbose)
@@ -1186,6 +1201,6 @@ class TimeSeries():
                 self.model = self.model.load_from_checkpoint(tmp_path,verbose=self.verbose,)
             else:
                 self.model = self.model.__class__.load_from_checkpoint(tmp_path,verbose=self.verbose,)
-
+            self.model.to(torch.device("cuda:0" if torch.cuda.is_available() else "cpu"))
         except Exception as e:
             beauty_string(f'There is a problem loading the weights on file {tmp_path} {e}','section',self.verbose)

dsipts/models/Autoformer.py

@@ -148,7 +148,8 @@ class Autoformer(Base):
             projection=nn.Linear(d_model, self.out_channels*self.mul, bias=True)
         )
         self.projection = nn.Linear(self.past_channels,self.out_channels*self.mul )
-
+    def can_be_compiled(self):
+        return True
     def forward(self, batch):
         
 

dsipts/models/CrossFormer.py

@@ -114,7 +114,8 @@ class CrossFormer(Base):
         
                                          
         
-     
+    def can_be_compiled(self):
+        return True
         
     def forward(self, batch):
 

dsipts/models/D3VAE.py

@@ -98,7 +98,8 @@ class D3VAE(Base):
         self.gamma = 0.01
         self.lambda1 = 1.0
 
-        
+    def can_be_compiled(self):
+        return False   
     def configure_optimizers(self):
         """
         Each model has optim_config and scheduler_config

dsipts/models/Diffusion.py

@@ -425,6 +425,9 @@ class Diffusion(Base):
         loss = self.compute_loss(batch,out)
         return loss
 
+    def can_be_compiled(self):
+        return False
+
     # function to concat embedded categorical variables
     def cat_categorical_vars(self, batch:dict):
         """Extracting categorical context about past and future

dsipts/models/DilatedConv.py

@@ -234,7 +234,8 @@ class DilatedConv(Base):
         self.return_additional_loss = True
         
 
-        
+    def can_be_compiled(self):
+        return True
 
     def forward(self, batch):
         """It is mandatory to implement this method

dsipts/models/DilatedConvED.py

@@ -228,7 +228,8 @@ class DilatedConvED(Base):
                                                 nn.BatchNorm1d(hidden_RNN) if use_bn else nn.Dropout(dropout_rate) ,
                                                 Permute() if use_bn else nn.Identity() ,
                                                 nn.Linear(hidden_RNN ,self.mul))
-        
+    def can_be_compiled(self):
+        return True  
 
 
     def forward(self, batch):

dsipts/models/Duet.py

@@ -136,7 +136,8 @@ class Duet(Base):
                                          activation(),
                                          nn.Linear(dim*2,self.out_channels*self.mul  ))
 
-
+    def can_be_compiled(self):
+        return False  
     def forward(self, batch:dict)-> float:
         # x: [Batch, Input length, Channel]
         x_enc = batch['x_num_past'].to(self.device)

dsipts/models/ITransformer.py

@@ -101,6 +101,9 @@ class ITransformer(Base):
         )
         self.projector = nn.Linear(d_model, self.future_steps*self.mul, bias=True)
 
+    def can_be_compiled(self):
+        return True  
+
     def forecast(self, x_enc, x_mark_enc, x_dec, x_mark_dec):
         if self.use_norm:
             # Normalization from Non-stationary Transformer

dsipts/models/Informer.py

@@ -124,7 +124,8 @@ class Informer(Base):
         
                 
         
-  
+    def can_be_compiled(self):
+        return True  
         
     def forward(self,batch): 
         #x_enc, x_mark_enc, x_dec, x_mark_dec,enc_self_mask=None, dec_self_mask=None, dec_enc_mask=None):

dsipts/models/LinearTS.py

@@ -143,7 +143,8 @@ class LinearTS(Base):
                                                     activation(),
                                                     nn.BatchNorm1d(hidden_size//8) if use_bn else nn.Dropout(dropout_rate) ,    
                                                     nn.Linear(hidden_size//8,self.future_steps*self.mul)))
-                               
+    def can_be_compiled(self):
+        return True                            
     def forward(self, batch):
       
         x =  batch['x_num_past'].to(self.device)

dsipts/models/PatchTST.py

@@ -133,6 +133,9 @@ class PatchTST(Base):
     
         #self.final_linear = nn.Sequential(nn.Linear(past_channels,past_channels//2),activation(),nn.Dropout(dropout_rate), nn.Linear(past_channels//2,out_channels)  )
     
+    def can_be_compiled(self):
+        return True  
+    
     def forward(self, batch):           # x: [Batch, Input length, Channel]
         
 

dsipts/models/RNN.py

@@ -148,7 +148,8 @@ class RNN(Base):
                                             activation(),
                                             MyBN(hidden_RNN//8) if use_bn else nn.Dropout(dropout_rate) ,
                                             nn.Linear(hidden_RNN//8,1)))
-
+    def can_be_compiled(self):
+        return True  
   
 
     def forward(self, batch):

dsipts/models/Samformer.py

@@ -85,7 +85,9 @@ class Samformer(Base):
                                          activation(),
                                          nn.Linear(dim*2,self.out_channels*self.mul  ))
 
-
+    def can_be_compiled(self):
+        return True  
+  
     def forward(self, batch:dict)-> float:
 
         x = batch['x_num_past'].to(self.device)

dsipts/models/Simple.py

@@ -67,7 +67,9 @@ class Simple(Base):
         self.linear = (nn.Sequential(nn.Linear(emb_past_out_channel*self.past_steps+emb_fut_out_channel*self.future_steps+self.past_steps*self.past_channels+self.future_channels*self.future_steps,hidden_size),
                                                     activation(),nn.Dropout(dropout_rate),
                                                     nn.Linear(hidden_size,self.out_channels*self.future_steps*self.mul)))
-                               
+    def can_be_compiled(self):
+        return True  
+                             
     def forward(self, batch):
       
         x =  batch['x_num_past'].to(self.device)

dsipts/models/TFT.py

@@ -111,6 +111,10 @@ class TFT(Base):
 
         self.outLinear = nn.Linear(d_model, self.out_channels*self.mul)
 
+    def can_be_compiled(self):
+        return False  
+  
+
     def forward(self, batch:dict) -> torch.Tensor:
         """Temporal Fusion Transformer
 

dsipts/models/TIDE.py

@@ -106,7 +106,10 @@ class TIDE(Base):
 
         # linear for Y lookback
         self.linear_target = nn.Linear(self.past_steps*self.out_channels, self.future_steps*self.out_channels*self.mul)
-
+    
+    def can_be_compiled(self):
+        return False  
+  
 
     def forward(self, batch:dict)-> float:
         """training process of the diffusion network

dsipts/models/TTM.py

@@ -12,7 +12,7 @@ except:
     from .base import Base
 
 
-from .ttm.utils import get_model, get_frequency_token, count_parameters
+from .ttm.utils import get_model, get_frequency_token, count_parameters, DEFAULT_FREQUENCY_MAPPING
 from ..data_structure.utils import beauty_string
 from .utils import  get_scope
 
@@ -43,6 +43,7 @@ class TTM(Base):
         super().__init__(**kwargs)
         self.save_hyperparameters(logger=False)
 
+        
 
         self.index_fut = list(exogenous_channel_indices_cont)
 
@@ -52,6 +53,14 @@ class TTM(Base):
             self.index_fut_cat = []
         self.freq = freq
 
+        base_freq_token = get_frequency_token(self.freq)  # e.g., shape [n_token] or scalar
+        # ensure it's a tensor of integer type
+        if not torch.is_tensor(base_freq_token):
+            base_freq_token = torch.tensor(base_freq_token)
+        base_freq_token = base_freq_token.long()
+        self.register_buffer("token", base_freq_token, persistent=True)
+                
+        
         self.model = get_model(
             model_path=model_path,
             context_length=self.past_steps,
@@ -98,14 +107,14 @@ class TTM(Base):
             input[:,:,i] = input[:, :, i] / (e-1)
         return input
 
-
+    def can_be_compiled(self):
+        return True  
+  
     def forward(self, batch):
-        x_enc = batch['x_num_past']
+        x_enc = batch['x_num_past'].to(self.device)
         original_indexes = batch['idx_target'][0].tolist()
 
 
-
-
         if 'x_cat_past' in batch.keys():
             x_mark_enc = batch['x_cat_past'].to(torch.float32).to(self.device)
             x_mark_enc = self._scaler_past(x_mark_enc)
@@ -113,7 +122,7 @@ class TTM(Base):
         else:
             past_values = x_enc
         
-        future_values = torch.zeros_like(past_values)
+        future_values = torch.zeros_like(past_values).to(self.device)
         future_values = future_values[:,:self.future_steps,:]
 
         if 'x_num_future' in batch.keys(): 
@@ -124,8 +133,12 @@ class TTM(Base):
             future_values[:,:,self.index_cat_fut] = x_mark_dec
         
 
-        #investigating!!
-        freq_token = get_frequency_token(self.freq).repeat(past_values.shape[0])
+        #investigating!! problem with dynamo!
+        #freq_token = get_frequency_token(self.freq).repeat(past_values.shape[0])
+
+        batch_size = past_values.shape[0]
+        freq_token = self.token.repeat(batch_size).long().to(self.device)
+
 
         res = self.model(
             past_values= past_values,
@@ -134,7 +147,7 @@ class TTM(Base):
             future_observed_mask = None,
             output_hidden_states =  False,
             return_dict = False,
-            freq_token= freq_token, ##investigating
+            freq_token= freq_token,#[0:past_values.shape[0]], ##investigating
             static_categorical_values = None
         )
 

dsipts/models/TimeXER.py

@@ -125,7 +125,9 @@ class TimeXER(Base):
 
         
 
-
+    def can_be_compiled(self):
+        return True  
+  
 
 
     def forward(self, batch:dict)-> float:

dsipts/models/base_v2.py

@@ -307,7 +307,7 @@ class Base(pl.LightningModule):
         self.train_epoch_count +=1
         return loss
 
-    
+
     def validation_step(self, batch, batch_idx):
         """
         pythotrch lightening stuff
@@ -320,15 +320,14 @@ class Base(pl.LightningModule):
         else:
             y_hat = self(batch)
             score = 0
-        if batch_idx==0:
+        #log_this_batch = (batch_idx == 0) and (self.count_epoch % int(max(self.trainer.max_epochs / 100,1)) == 1)
 
+        #if log_this_batch:
             #track the predictions! We can do better than this but maybe it is better to firstly update pytorch-lightening 
-
-            if self.count_epoch%int(max(self.trainer.max_epochs/100,1))==1:
-                self._val_outputs.append({
-                    "y": batch['y'].detach().cpu(),
-                    "y_hat": y_hat.detach().cpu()
-                })                
+        self._val_outputs=[{
+                "y": batch['y'].detach().cpu(),
+                "y_hat": y_hat.detach().cpu()
+            }]               
         self.validation_epoch_metrics+= (self.compute_loss(batch,y_hat)+score).detach()
         self.validation_epoch_count+=1
         return None

dsipts/models/duet/layers.py

@@ -219,7 +219,7 @@ class SparseDispatcher(object):
         # expand according to batch index so we can just split by _part_sizes
         inp_exp = inp[self._batch_index].squeeze(1)
         return torch.split(inp_exp, self._part_sizes, dim=0)
-
+ 
     def combine(self, expert_out, multiply_by_gates=True):
         """Sum together the expert output, weighted by the gates.
         The slice corresponding to a particular batch element `b` is computed
@@ -234,7 +234,9 @@ class SparseDispatcher(object):
           a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
         """
         # apply exp to expert outputs, so we are not longer in log space
+  
         stitched = torch.cat(expert_out, 0)
+
         if multiply_by_gates:
             # stitched = stitched.mul(self._nonzero_gates)
             stitched = torch.einsum("i...,ij->i...", stitched, self._nonzero_gates)
@@ -430,9 +432,11 @@ class Linear_extractor_cluster(nn.Module):
         expert_inputs = dispatcher.dispatch(x_norm)
 
         gates = dispatcher.expert_to_gates()
+
         expert_outputs = [
             self.experts[i](expert_inputs[i]) for i in range(self.num_experts)
         ]
+        #y = dispatcher.combine([e for e in expert_outputs if len(e)>0])
+        #with torch._dynamo.disable():
         y = dispatcher.combine(expert_outputs)
-
         return y, loss

{dsipts-1.1.11.dist-info → dsipts-1.1.12.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dsipts
-Version: 1.1.11
+Version: 1.1.12
 Summary: Unified library for timeseries modelling
 Author-email: Andrea Gobbi <agobbi@fbk.eu>
 Project-URL: Homepage, https://github.com/DSIP-FBK/DSIPTS

{dsipts-1.1.11.dist-info → dsipts-1.1.12.dist-info}/RECORD

@@ -3,33 +3,33 @@ dsipts/data_management/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3
 dsipts/data_management/monash.py,sha256=aZxq9FbIH6IsU8Lwou1hAokXjgOAK-wdl2VAeFg2k4M,13075
 dsipts/data_management/public_datasets.py,sha256=yXFzOZZ-X0ZG1DoqVU-zFmEGVMc2033YDQhRgYxY8ws,6793
 dsipts/data_structure/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-dsipts/data_structure/data_structure.py,sha256=uyGkc1eDjETpXb8rgMMbRUjG8i9Xiiu6vZc64xfTiew,59914
+dsipts/data_structure/data_structure.py,sha256=KVkjTVjc7NznJIou4LYGzMbzE7ye-K3ll65GEgn2qKg,60814
 dsipts/data_structure/modifiers.py,sha256=qlry9dfw8pEE0GrvgwROZJkJ6oPpUnjEHPIG5qIetss,7948
 dsipts/data_structure/utils.py,sha256=QwfKPZgSy6DIw5n6ztOdPJIAnzo4EnlMTgRbpiWnyko,6593
-dsipts/models/Autoformer.py,sha256=ddGT3L9T4gAXNJHx1TsuYZy7j63Anyr0rkqqXaOoSu4,8447
-dsipts/models/CrossFormer.py,sha256=iO64L3S01jxuWA9dmm8FsK1WRvBIXbZ0PQ2tZlEQg4w,6481
-dsipts/models/D3VAE.py,sha256=NstHIniNteBRrkfL7SJ3-bJEl3l3IIxoSxavRV3j16U,6857
-dsipts/models/Diffusion.py,sha256=pUujnrdeSSkj4jC1RORbcptt03KpuCsGVwg414o4LPg,40733
-dsipts/models/DilatedConv.py,sha256=_c0NvFuT3vbYmo9A8cQchGo1XVb0qOpzBprNEkkAgiE,14292
-dsipts/models/DilatedConvED.py,sha256=fXk1-EWiRC5J_VIepTjYKya_D02SlEAkyiJcCjhW_XU,14004
-dsipts/models/Duet.py,sha256=EharWHT_r7tEYIk7BkozVLPZ0xptE5mmQmeFGm3uBsA,7628
-dsipts/models/ITransformer.py,sha256=qMsk27PqpnakNY1YM_rbkj8MO6BaG06N3b6m30Oa0RQ,7256
-dsipts/models/Informer.py,sha256=ByJ00qGk12ONFF7NZWAACzxxRb5UXcu5wpkGMYX9Cq4,6920
-dsipts/models/LinearTS.py,sha256=B0-Sz4POwUyl-PN2ssSx8L-ZHgwrQQPcMmreyvSS47U,9104
-dsipts/models/PatchTST.py,sha256=Z7DM1Kw5Ym8Hh9ywj0j9RuFtKaz_yVZmKFIYafjceM8,9061
+dsipts/models/Autoformer.py,sha256=nUQvPC_qtajLT1AHdNJmF_P3ZL01j3spkZ4ubxdGF3g,8497
+dsipts/models/CrossFormer.py,sha256=ClW6H_hrtLJH0iqTC7q_ya_Bwc_Xu-0lpAN5w2DSUYk,6526
+dsipts/models/D3VAE.py,sha256=d1aY6kGjBSxZncN-KPWpdUGunu182ng2QFInGFrKYQM,6903
+dsipts/models/Diffusion.py,sha256=owst4IxA3hkEEIrn5K-zwAYWUzEhouiRPwM4nTLcyoE,40786
+dsipts/models/DilatedConv.py,sha256=TMDzd_cNgCZa6YusVVVGbTGGH3YlMz0IZZ9ZxRrJ3i4,14334
+dsipts/models/DilatedConvED.py,sha256=KwG83yHqoEx_Vmea69zTPsSP1-0GdOUrtXwvhNDuWj8,14048
+dsipts/models/Duet.py,sha256=m67PStuYE6vkFUFUofBrrLryx1ZUZropyVGcu_ygOx8,7681
+dsipts/models/ITransformer.py,sha256=2WXqqEvnWH2DqRQyXfGm4Eg4_q32GFy2XnNeoTl-KmY,7310
+dsipts/models/Informer.py,sha256=gxCdU2KkNhadyMujBA5A0eP6SPN4Q0IkEIogLYwvz5k,6970
+dsipts/models/LinearTS.py,sha256=vXaGpbbkfdpzpTEWZ1hs6QI6j3vDvevD3SyKQXo6Sdg,9151
+dsipts/models/PatchTST.py,sha256=1O09cPMg8USdkt5q6szTiz5dIY45kizsf6gt6vLKnQo,9119
 dsipts/models/Persistent.py,sha256=URwyaBb0M7zbPXSGMImtHlwC9XCy-OquFCwfWvn3P70,1249
-dsipts/models/RNN.py,sha256=GbH6QyrGhvQg-Hnt_0l3YSnhNHE0Hl0AWsZpdQUAzug,9633
-dsipts/models/Samformer.py,sha256=s61Hi1o9iuw-KgSBPfiE80oJcK1j2fUA6N9f5BJgKJc,5551
-dsipts/models/Simple.py,sha256=K82E88A62NhV_7U9Euu2cn3Q8P287HDR7eIy7VqgwbM,3909
-dsipts/models/TFT.py,sha256=JO2-AKIUag7bfm9Oeo4KmGfdYZJbzQBHPDqGVg0WUZI,13830
-dsipts/models/TIDE.py,sha256=i8qXac2gImEVgE2X6cNxqW5kuQP3rzWMlQNdgJbNmKM,13033
-dsipts/models/TTM.py,sha256=gc-8yzEtn8ZdRVvsZfZvz7iE-RgqpZc-JGmOCQr4U_0,5215
-dsipts/models/TimeXER.py,sha256=aCg0003LxYZzqZWyWugpbW_iOybcdHN4OH6_v77qp4o,7056
+dsipts/models/RNN.py,sha256=RnsRDAQ2z5-XNaJVZd6Q7z23WvPR2uLVdi7BNQyF7QE,9685
+dsipts/models/Samformer.py,sha256=Kt7B9ID3INtFDAVKIM1LTly5-UfKCaVZ9uxAJmYv6B4,5606
+dsipts/models/Simple.py,sha256=8wRSO-gh_Z6Sl8fYMV-RIXIL0RrO5u5dDtsaq-OsKg0,3960
+dsipts/models/TFT.py,sha256=JiI90ikfP8aaR_rtczu8CyGMNLTgml13aYQifgIC_yo,13888
+dsipts/models/TIDE.py,sha256=S1KlKqFOR3jJ9DDiTqeaKvya9hYBsNHBVqwJsYX3FLU,13094
+dsipts/models/TTM.py,sha256=lOOo5dR5nOmf37cND6C8ft8TVl0kzNeraIuABw7eI5g,5897
+dsipts/models/TimeXER.py,sha256=EkmlHfT2RegY6Ce6q8EUEV1a_WZ6SkYibnOZXqsyd_8,7111
 dsipts/models/VQVAEA.py,sha256=sNJi8UZh-10qEIKcZK3SzhlOFUUjvqjoglzeZBFaeZM,13789
 dsipts/models/VVA.py,sha256=BnPkJ0Nzue0oShSHZVRNlf5RvT0Iwtf9bx19vLB9Nn0,11939
 dsipts/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dsipts/models/base.py,sha256=Gqsycy8ZXGaIVx9vvmYRpBCqdUxGE4tvC5ltgxlpEYY,19640
-dsipts/models/base_v2.py,sha256=eraXo1IBEQmyW41f1dz3Q-i-61vZ2AS3tVz6_X8J0Pg,19886
+dsipts/models/base_v2.py,sha256=03cueZExRhkJyBVIHuUPB8sjsCd5Go1HJAR81CADg-c,19896
 dsipts/models/utils.py,sha256=kjTwyktNCFMpPUy6zoleBCSKlvMvK_Jkgyh2T1OXg3E,24497
 dsipts/models/autoformer/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dsipts/models/autoformer/layers.py,sha256=xHt8V1lKdD1cIvgxXdDbI_EqOz4zgOQ6LP8l7M1pAxM,13276
@@ -47,7 +47,7 @@ dsipts/models/d3vae/neural_operations.py,sha256=C70kUtQ0ox9MeXBdu4rPDqt022_hVtcN
 dsipts/models/d3vae/resnet.py,sha256=3bnlrEBM2DGiAJV8TeSv2tm27Gm-_P6hee41t8QQFL8,5520
 dsipts/models/d3vae/utils.py,sha256=fmUsE_67uwizjeR1_pDdsndyQddbqt27Lv31XBEn-gw,23798
 dsipts/models/duet/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-dsipts/models/duet/layers.py,sha256=ikMAKr4DzhGt9J1KuBBQzVYXMHZUZVsV29P6TVJCa_Y,18141
+dsipts/models/duet/layers.py,sha256=TTrhlfSwIXE_7gO9rsdKJD9Bdy3B_JJPCo8vYZJ8Fvg,18258
 dsipts/models/duet/masked.py,sha256=lkdAB5kwAgV7QfBSVP_QeDr_mB09Rz4302p-KwZpUV4,7111
 dsipts/models/informer/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dsipts/models/informer/attn.py,sha256=ghrQGfAqt-Z_7qU5D_aixobmwk6pBKMLAdaNfg-QZbo,6839
@@ -76,7 +76,7 @@ dsipts/models/vva/minigpt.py,sha256=bg0JddqSD322uxSGexen3nPXL_hGTsk3vNLR62d7-w8,
 dsipts/models/vva/vqvae.py,sha256=RzCQ_M9xBprp7_x20dSV3EQqlO0FjPUGWV-qdyKrQsM,19680
 dsipts/models/xlstm/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dsipts/models/xlstm/xLSTM.py,sha256=ZKZZmffmIq1Vb71CR4GSyM8viqVx-u0FChxhcNgHub8,10081
-dsipts-1.1.11.dist-info/METADATA,sha256=fbMTKqi7b_vlvtmVSp5XJdkFrEC9SFF3DG_fKy58k_8,24795
-dsipts-1.1.11.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-dsipts-1.1.11.dist-info/top_level.txt,sha256=i6o0rf5ScFwZK21E89dSKjVNjUBkrEQpn0-Vij43748,7
-dsipts-1.1.11.dist-info/RECORD,,
+dsipts-1.1.12.dist-info/METADATA,sha256=nxE2kAg9RvG5Py27sMNbQ-mUIu9mtZrDo2WocLpJdQ4,24795
+dsipts-1.1.12.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+dsipts-1.1.12.dist-info/top_level.txt,sha256=i6o0rf5ScFwZK21E89dSKjVNjUBkrEQpn0-Vij43748,7
+dsipts-1.1.12.dist-info/RECORD,,