inference works but is hella slow

llms/mlx_lm/models/cache.py

@@ -342,7 +342,7 @@ class MambaCache(_BaseCache):
 
 class Mamba2Cache(_BaseCache):
     conv_states: Optional[mx.array] = None
-    ssm_states: Optional[mx.array] = None
+    ssm_state: Optional[mx.array] = None
 
     def __getitem__(self, idx: int) -> Optional[mx.array]:
         if idx == 0:

llms/mlx_lm/models/mamba2.py

@@ -103,89 +103,55 @@ class DepthWiseConv1d(nn.Module):
         assert in_channels == out_channels, "In and out channels must be same for depthwise convolution"
         assert self.groups == in_channels, "Groups must be equal to in_channels for depthwise convolution"
         
-        # Weight shape: (channels, 1, kernel_size) to match pretrained weights
         self.weight = mx.random.normal((in_channels, 1, kernel_size))
         self.bias = mx.zeros((out_channels,)) if bias else None
 
-    def __call__(self, x: mx.array, cache=None, cache_idx: int = 0) -> mx.array:
+    def __call__(self, x: mx.array, cache=None) -> mx.array:
         B, L, C = x.shape
         K = self.kernel_size
 
-        # Validate input dimensions
         assert C == self.in_channels, f"Input channels {C} doesn't match expected {self.in_channels}"
         
-        # Handle padding and caching
+        if cache is not None and 'conv_states' in cache:
-        if cache is not None:
+            conv_states = cache['conv_states']
-            conv_states = cache[cache_idx]
             if conv_states is not None:
-                # Validate cache shape
                 assert conv_states.shape[0] == B, "Cache batch size mismatch"
                 assert conv_states.shape[2] == C, "Cache channel count mismatch"
                 x = mx.concatenate([conv_states, x], axis=1)
-                L = x.shape[1]
-        else:
-            # Add left padding of size (kernel_size - 1)
-            pad_left = K - 1
-            x = mx.pad(x, [(0, 0), (pad_left, 0), (0, 0)])
-            L = x.shape[1]
-
-        # Pre-allocate output array if possible
-        outputs = []
-        
-        # Process each channel independently
-        for c in range(C):
-            # Extract and prepare channel data
-            x_c = x[:, :, c]  # Shape: [B, L]
-            x_c = mx.expand_dims(x_c, axis=1)  # Shape: [B, 1, L]
             
-            # Prepare filter weights
+        # Process each channel independently
-            w_c = self.weight[c]  # Get channel weights
+        outputs = []
-            # Ensure filter is 3D: [depth(1), in_channels(1), kernel_size]
+        for c in range(C):
+            x_c = x[:, :, c]
+            x_c = mx.expand_dims(x_c, axis=1)
+            
+            w_c = self.weight[c]
             if w_c.ndim == 2:
                 w_c = mx.expand_dims(w_c, axis=0)
             elif w_c.ndim == 1:
                 w_c = mx.expand_dims(mx.expand_dims(w_c, axis=0), axis=0)
             
-            # Handle inference mode (single token)
+            # Apply convolution
-            if L < K:
+            y_c = mx.conv_general(
-                pad_size = K - L
+                x_c,
-                x_c = mx.pad(x_c, [(0, 0), (0, 0), (pad_size, 0)])
+                w_c,
+                stride=1,
+                padding=0
+            )
             
-            # Apply 1D convolution
+            if self.bias is not None:
-            try:
+                y_c = y_c + self.bias[c]
-                y_c = mx.conv_general(
+            
-                    x_c,
+            outputs.append(mx.squeeze(y_c, axis=1))
-                    w_c,
-                    stride=1,
-                    padding=0  # Padding already handled
-                )
-                
-                if self.bias is not None:
-                    y_c = y_c + self.bias[c]
-                
-                # Remove singleton dimension and add to outputs
-                outputs.append(mx.squeeze(y_c, axis=1))
-                
-            except Exception as e:
-                raise RuntimeError(f"Convolution failed for channel {c}. Shapes: input={x_c.shape}, weight={w_c.shape}") from e
 
-        # Stack channel outputs along last dimension
+        y = mx.stack(outputs, axis=-1)
-        y = mx.stack(outputs, axis=-1)  # Shape: [B, L', C]
         
-        # Update cache if needed
+        # Update cache
         if cache is not None:
-            # Store last (kernel_size - 1) tokens or entire input if shorter
+            cache['conv_states'] = x[:, -K+1:, :] if x.shape[1] >= K else x
-            new_cache = x[:, -(K-1):, :] if L >= K else x
+            
-            cache[cache_idx] = new_cache
-
-            if new_cache.shape != cache[cache_idx].shape:
-                cache[cache_idx] = new_cache
-                print(f"Cache updated at index {cache_idx}")
-            else:
-                print(f"Skipping cache update at index {cache_idx}, shapes are identical.")
-
         return y
-    
+
 
 class Mamba2Block(nn.Module):
     def __init__(self, args: ModelArgs):
@@ -217,34 +183,22 @@ class Mamba2Block(nn.Module):
             self.out_proj.weight = self.out_proj.weight * layer_scale
 
     def __call__(self, x: mx.array, cache=None):
-        # if cache is not None and self.args.use_cache:
         if cache is not None:
             return self.step(x, cache)
 
-        # Calculate sizes
+        # Regular forward pass code remains the same...
         d_model = self.args.intermediate_size
         d_state = self.args.state_size
         n_heads = self.args.num_heads
         
-        # Compute A
         A = -mx.exp(self.A_log)
-        
-        # Project input
         zxbcdt = self.in_proj(x)
         
-        # Correct splits for z, xBC, dt
+        splits = [d_model, d_model + 2 * d_state, n_heads]
-        splits = [
-            d_model,                # z
-            d_model + 2 * d_state,  # xBC (delta, B, C concatenated)
-            n_heads                 # dt
-        ]
-        
-        # Split using cumulative indices
         z = zxbcdt[:, :, :splits[0]]
         xBC = zxbcdt[:, :, splits[0]:splits[0] + splits[1]]
         dt = zxbcdt[:, :, -splits[2]:]
 
-        # Process dt
         dt = mx.clip(
             nn.softplus(dt + self.dt_bias),
             self.args.time_step_min,
@@ -252,46 +206,26 @@ class Mamba2Block(nn.Module):
         )
         dt = mx.maximum(dt, self.args.time_step_floor)
 
-        # Process convolution
         xBC = silu(self.conv1d(xBC))
 
-        # Split convolved xBC into x, B, C
         x = xBC[:, :, :d_model]
         B = xBC[:, :, d_model:d_model + d_state]
         C = xBC[:, :, -d_state:]
 
-        # Reshape for SSM computation
         b, l, hp = x.shape
         h = self.args.num_heads
         p = hp // h
         x = mx.reshape(x, (b, l, h, p))
 
-        # Compute SSM
+        y, ssm_state = ssd(x * mx.expand_dims(dt, -1), A * dt, B, C, self.args.chunk_size)
-        y, ssm_state = ssd(
-            x * mx.expand_dims(dt, -1),
-            A * dt,
-            B,
-            C,
-            self.args.chunk_size
-        )
-
-        # Add skip connection
         y = y + x * mx.expand_dims(self.D, -1)
-        
-        # Reshape back
         y = mx.reshape(y, (b, l, h * p))
 
-        # Apply norm and projection
         y = self.norm(y + z)
         y = self.out_proj(y)
 
-        # Update cache if needed
-        if cache is not None and self.args.use_cache:
-            cache[1] = ssm_state
-
-        # Cast if needed
         if self.args.residual_in_fp32:
-            y.astype(mx.float32)
+            y = y.astype(mx.float32)
 
         return y
 
@@ -300,9 +234,17 @@ class Mamba2Block(nn.Module):
         seq_len = u.shape[1]
         outputs = []
 
-        # Initialize SSM state if needed
+        # Initialize cache if needed
-        if cache[1] is None:
+        if cache.conv_states is None:
-            cache[1] = mx.zeros((
+            conv_dim = self.args.intermediate_size + 2 * self.args.state_size
+            cache.conv_states = mx.zeros((
+                batch_size,
+                self.args.conv_kernel - 1,
+                conv_dim
+            ))
+            
+        if cache.ssm_state is None:
+            cache.ssm_state = mx.zeros((
                 batch_size,
                 self.args.num_heads,
                 self.args.head_dim,
@@ -310,26 +252,17 @@ class Mamba2Block(nn.Module):
             ))
 
         for pos in range(seq_len):
-            # Getting stuck here in last position, also cache from pos 0 is the same.
-            # Get single token
             u_t = u[:, pos:pos+1, :]
-
-            # Project input
             zxbcdt = self.in_proj(u_t)
             
-            # Calculate sizes
             d_model = self.args.intermediate_size
             d_state = self.args.state_size
             n_heads = self.args.num_heads
-            d_head = self.args.head_dim
             
-            # Split projected input
-            # conv_dim = d_model + 2 * d_state (this should match self.conv1d.in_channels)
             z = zxbcdt[:, :, :d_model]
-            xBC = zxbcdt[:, :, d_model:d_model + 2*d_state + d_model]  # Include the full conv dimension
+            xBC = zxbcdt[:, :, d_model:d_model + 2*d_state + d_model]
             dt = zxbcdt[:, :, -(n_heads):]
             
-            # Process dt
             dt = mx.reshape(dt, (batch_size, n_heads))
             dt = mx.clip(
                 nn.softplus(dt + self.dt_bias),
@@ -338,49 +271,43 @@ class Mamba2Block(nn.Module):
             )
             dt = mx.maximum(dt, self.args.time_step_floor)
 
-            # Process convolution with correct dimensions
+            # Create a temporary cache dictionary for the convolution
-            xBC = self.conv1d(xBC, cache=cache, cache_idx=0)
+            conv_cache = {'conv_states': cache.conv_states}
+            xBC = self.conv1d(xBC, cache=conv_cache)
+            cache.conv_states = conv_cache['conv_states']
+            
             xBC = silu(xBC)
 
-            # Split convolved xBC into x, B, C with correct dimensions
             x = xBC[:, :, :d_model]
             B = xBC[:, :, d_model:d_model + d_state]
             C = xBC[:, :, -d_state:]
             
-            # Reshape tensors for SSM computation
+            x = mx.reshape(x, (batch_size, 1, n_heads, self.args.head_dim))
-            x = mx.reshape(x, (batch_size, 1, n_heads, d_head))
+            x = mx.squeeze(x, axis=1)
-            x = mx.squeeze(x, axis=1)  # (batch, heads, dim)
+
-            
             B = mx.reshape(B, (batch_size, 1, d_state))
             B = mx.broadcast_to(B, (batch_size, n_heads, d_state))
-            B = mx.expand_dims(B, axis=2)  # (batch, heads, 1, state)
+            B = mx.expand_dims(B, axis=2)
-            
+
             C = mx.reshape(C, (batch_size, 1, d_state))
             C = mx.broadcast_to(C, (batch_size, n_heads, d_state))
-            C = mx.expand_dims(C, axis=3)  # (batch, heads, state, 1)
+            C = mx.expand_dims(C, axis=3)
 
-            # Compute SSM updates
             A = -mx.exp(self.A_log)
             dA = mx.exp(dt * mx.expand_dims(A, 0))
-            dA = mx.expand_dims(mx.expand_dims(dA, -1), -1)  # (batch, heads, 1, 1)
+            dA = mx.expand_dims(mx.expand_dims(dA, -1), -1)
 
-            # Update state with proper shapes
+            x = mx.expand_dims(x, axis=3)
-            x = mx.expand_dims(x, axis=3)  # (batch, heads, dim, 1)
+            dBx = mx.matmul(x, B)
-            dBx = mx.matmul(x, B)  # (batch, heads, dim, state)
             
-            ssm_state = cache[1]
+            cache.ssm_state = cache.ssm_state * dA + dBx
-            ssm_state = ssm_state * dA + dBx
-            cache[1] = ssm_state
 
-            # Compute output
+            y = mx.matmul(cache.ssm_state, C)
-            y = mx.matmul(ssm_state, C)  # (batch, heads, dim, 1)
+            y = mx.squeeze(y, axis=-1)
-            y = mx.squeeze(y, axis=-1)  # (batch, heads, dim)
             
-            # Add skip connection
             y = y + x[:, :, :, 0] * mx.expand_dims(self.D, -1)
             
-            # Reshape and process output
+            y = mx.reshape(y, (batch_size, 1, n_heads * self.args.head_dim))
-            y = mx.reshape(y, (batch_size, 1, n_heads * d_head))
             y = self.norm(y + z)
             y = self.out_proj(y)
 
@@ -440,7 +367,6 @@ class Model(nn.Module):
         else:
             logits = self.lm_head(x)
 
-        print('ouput')
         return logits
 
     def make_cache(self, batch_size=1):