fixed the AR+NAR dual model, the resp_emb has to be split up (classifier might too)

vall_e/models/ar.py

@@ -53,6 +53,10 @@ class AR(Base):
 			return self.config.interleave
 		return False
 
+	@property
+	def dual(self) -> bool:
+		return False
+
 	def _prune(self, l: Tensor):
 		indices = (l == self.stop_token).nonzero()
 		if len(indices) == 0:

vall_e/models/ar_nar.py

@@ -16,7 +16,7 @@ class AR_NAR(Base):
 
 	@property
 	def norm_type(self):
-		return "ln"
+		return "ln" if self.n_resp_levels == 1 else "adaln"
 
 	@property
 	def arch_type(self) -> str:
@@ -44,15 +44,15 @@ class AR_NAR(Base):
 
 	@property
 	def recurrent_chunk_size(self) -> int:
-		if cfg.mode == "training":
-			return 0
-		return cfg.inference.recurrent_chunk_size
+		return 0
 
 	@property
 	def interleave(self) -> bool:
-		if hasattr(self, "config") and self.config:
-			return self.config.interleave
 		return False
+	
+	@property
+	def dual(self) -> bool:
+		return True
 
 	def _prune(self, l: Tensor):
 		indices = (l == self.stop_token).nonzero()
@@ -60,18 +60,6 @@ class AR_NAR(Base):
 			return l
 		return l[: indices.min().item()]
 
-	def _interleave( self, codes ):
-		if not self.interleave:
-			return codes
-
-		return codes.flatten()
-
-	def _deinterleave( self, codes, length = 0 ):
-		if not self.interleave:
-			return codes
-
-		return torch.unflatten( codes[:codes.shape[0] // self.n_prom_levels * self.n_prom_levels], 0, ( codes.shape[0] // self.n_prom_levels, self.n_prom_levels ) )
-
 	@staticmethod
 	def _unsqueeze_list(x_list, axis=-1):
 		return [x.unsqueeze(dim=axis) for x in x_list]
@@ -243,7 +231,7 @@ def example_usage():
 
 	def train():
 		engine.train()
-		t = trange(5000)
+		t = trange(1000)
 		for i in t:
 			stats = {"step": i}
 			stats |= engine.traverse(text_list=text_list, proms_list=proms_list, resps_list=resps_list)

vall_e/models/base.py

@@ -126,6 +126,10 @@ class Base(nn.Module):
 	def interleave(self) -> bool:
 		return False
 
+	@property
+	def dual(self) -> bool:
+		return False
+
 	@property
 	def stop_token(self):
 		if not self.causal:
@@ -151,7 +155,7 @@ class Base(nn.Module):
 		n_heads: int = 8,
 		n_layers: int = 12,
 		p_dropout: float = 0.1,
-		
+
 		config = None, 
 	):
 		super().__init__()
@@ -169,7 +173,11 @@ class Base(nn.Module):
 
 		self.text_emb = Embedding(n_tokens, d_model)
 		self.proms_emb = MultiEmbedding(self.n_prom_levels, n_prom_tokens, d_model)
-		self.resps_emb = MultiEmbedding(self.n_resp_levels, n_resp_tokens, d_model)
+
+		if self.dual:
+			self.resps_emb = nn.ModuleList([MultiEmbedding(self.n_resp_levels, n_resp_tokens, d_model) for _ in range(2)])
+		else:
+			self.resps_emb = MultiEmbedding(self.n_resp_levels, n_resp_tokens, d_model)
 
 		self.sep = nn.Parameter(torch.randn(d_model))
 
@@ -254,12 +262,20 @@ class Base(nn.Module):
 
 		state: dict | None = None,
 	):
-		x_list = self._samplewise_merge_tensors(
-			self.text_emb(text_list),
-			self.proms_emb(proms_list),
-			self.resps_emb(resps_list),
-			sep=self.sep,
-		)
+		if self.dual:
+			x_list = self._samplewise_merge_tensors(
+				self.text_emb(text_list),
+				self.proms_emb(proms_list),
+				self.resps_emb[0 if quant_levels is None else 1](resps_list),
+				sep=self.sep,
+			)
+		else:
+			x_list = self._samplewise_merge_tensors(
+				self.text_emb(text_list),
+				self.proms_emb(proms_list),
+				self.resps_emb(resps_list),
+				sep=self.sep,
+			)
 
 		x, m = list_to_tensor(x_list)
 		

vall_e/models/nar.py

@@ -47,6 +47,10 @@ class NAR(Base):
 	def interleave(self) -> bool:
 		return False
 
+	@property
+	def dual(self) -> bool:
+		return False
+
 	def forward(
 		self,
 		text_list: list[Tensor],