fix(pt): fix precision (deepmodeling#4344)

Tried to implement the decorator as in deepmodeling#4343, but encountered JIT
errors.

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

## Release Notes

- **New Features**
- Enhanced precision handling across various descriptor classes and
methods, ensuring consistent tensor operations.
- Updated output formats in several classes to improve clarity and
usability.
- Introduced a new environment variable for stricter control over tensor
precision handling.
- Added a new parameter to the `DipoleFittingNet` class for excluding
specific types.

- **Bug Fixes**
- Removed conditions that skipped tests for "float32" data type,
allowing all tests to run consistently.

- **Documentation**
- Improved error messages for dimension mismatches and unsupported
parameters, enhancing user understanding.

- **Tests**
- Adjusted test parameters for consistency in handling `fparam` and
`aparam` across multiple test cases.
- Simplified tensor handling in tests by removing unnecessary type
conversions before compression.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

deepmd/pt/model/descriptor/dpa1.py

@@ -22,6 +22,7 @@
     env,
 )
 from deepmd.pt.utils.env import (
+    PRECISION_DICT,
     RESERVED_PRECISON_DICT,
 )
 from deepmd.pt.utils.tabulate import (
@@ -311,6 +312,7 @@ def __init__(
             use_tebd_bias=use_tebd_bias,
             type_map=type_map,
         )
+        self.prec = PRECISION_DICT[precision]
         self.tebd_dim = tebd_dim
         self.concat_output_tebd = concat_output_tebd
         self.trainable = trainable
@@ -678,6 +680,8 @@ def forward(
             The smooth switch function. shape: nf x nloc x nnei
 
         """
+        # cast the input to internal precsion
+        extended_coord = extended_coord.to(dtype=self.prec)
         del mapping
         nframes, nloc, nnei = nlist.shape
         nall = extended_coord.view(nframes, -1).shape[1] // 3
@@ -693,7 +697,13 @@ def forward(
         if self.concat_output_tebd:
             g1 = torch.cat([g1, g1_inp], dim=-1)
 
-        return g1, rot_mat, g2, h2, sw
+        return (
+            g1.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            rot_mat.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            g2.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION) if g2 is not None else None,
+            h2.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            sw.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+        )
 
     @classmethod
     def update_sel(

deepmd/pt/model/descriptor/dpa2.py

@@ -27,6 +27,9 @@
 from deepmd.pt.utils import (
     env,
 )
+from deepmd.pt.utils.env import (
+    PRECISION_DICT,
+)
 from deepmd.pt.utils.nlist import (
     build_multiple_neighbor_list,
     get_multiple_nlist_key,
@@ -268,6 +271,7 @@ def init_subclass_params(sub_data, sub_class):
         )
         self.concat_output_tebd = concat_output_tebd
         self.precision = precision
+        self.prec = PRECISION_DICT[self.precision]
         self.smooth = smooth
         self.exclude_types = exclude_types
         self.env_protection = env_protection
@@ -745,6 +749,9 @@ def forward(
             The smooth switch function. shape: nf x nloc x nnei
 
         """
+        # cast the input to internal precsion
+        extended_coord = extended_coord.to(dtype=self.prec)
+
         use_three_body = self.use_three_body
         nframes, nloc, nnei = nlist.shape
         nall = extended_coord.view(nframes, -1).shape[1] // 3
@@ -810,7 +817,13 @@ def forward(
         )
         if self.concat_output_tebd:
             g1 = torch.cat([g1, g1_inp], dim=-1)
-        return g1, rot_mat, g2, h2, sw
+        return (
+            g1.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            rot_mat.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            g2.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            h2.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            sw.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+        )
 
     @classmethod
     def update_sel(

deepmd/pt/model/descriptor/repformers.py

@@ -22,6 +22,9 @@
 from deepmd.pt.utils import (
     env,
 )
+from deepmd.pt.utils.env import (
+    PRECISION_DICT,
+)
 from deepmd.pt.utils.env_mat_stat import (
     EnvMatStatSe,
 )
@@ -237,6 +240,7 @@ def __init__(
         self.reinit_exclude(exclude_types)
         self.env_protection = env_protection
         self.precision = precision
+        self.prec = PRECISION_DICT[precision]
         self.trainable_ln = trainable_ln
         self.ln_eps = ln_eps
         self.epsilon = 1e-4
@@ -286,12 +290,8 @@ def __init__(
         self.layers = torch.nn.ModuleList(layers)
 
         wanted_shape = (self.ntypes, self.nnei, 4)
-        mean = torch.zeros(
-            wanted_shape, dtype=env.GLOBAL_PT_FLOAT_PRECISION, device=env.DEVICE
-        )
-        stddev = torch.ones(
-            wanted_shape, dtype=env.GLOBAL_PT_FLOAT_PRECISION, device=env.DEVICE
-        )
+        mean = torch.zeros(wanted_shape, dtype=self.prec, device=env.DEVICE)
+        stddev = torch.ones(wanted_shape, dtype=self.prec, device=env.DEVICE)
         self.register_buffer("mean", mean)
         self.register_buffer("stddev", stddev)
         self.stats = None

deepmd/pt/model/descriptor/se_a.py

@@ -118,6 +118,7 @@ def __init__(
         super().__init__()
         self.type_map = type_map
         self.compress = False
+        self.prec = PRECISION_DICT[precision]
         self.sea = DescrptBlockSeA(
             rcut,
             rcut_smth,
@@ -337,7 +338,18 @@ def forward(
             The smooth switch function.
 
         """
-        return self.sea.forward(nlist, coord_ext, atype_ext, None, mapping)
+        # cast the input to internal precsion
+        coord_ext = coord_ext.to(dtype=self.prec)
+        g1, rot_mat, g2, h2, sw = self.sea.forward(
+            nlist, coord_ext, atype_ext, None, mapping
+        )
+        return (
+            g1.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            rot_mat.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            None,
+            None,
+            sw.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+        )
 
     def set_stat_mean_and_stddev(
         self,
@@ -742,7 +754,6 @@ def forward(
         )
 
         dmatrix = dmatrix.view(-1, self.nnei, 4)
-        dmatrix = dmatrix.to(dtype=self.prec)
         nfnl = dmatrix.shape[0]
         # pre-allocate a shape to pass jit
         xyz_scatter = torch.zeros(
@@ -811,8 +822,8 @@ def forward(
         result = result.view(nf, nloc, self.filter_neuron[-1] * self.axis_neuron)
         rot_mat = rot_mat.view([nf, nloc] + list(rot_mat.shape[1:]))  # noqa:RUF005
         return (
-            result.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
-            rot_mat.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            result,
+            rot_mat,
             None,
             None,
             sw,

deepmd/pt/model/descriptor/se_atten.py

@@ -227,12 +227,8 @@ def __init__(
         )
 
         wanted_shape = (self.ntypes, self.nnei, 4)
-        mean = torch.zeros(
-            wanted_shape, dtype=env.GLOBAL_PT_FLOAT_PRECISION, device=env.DEVICE
-        )
-        stddev = torch.ones(
-            wanted_shape, dtype=env.GLOBAL_PT_FLOAT_PRECISION, device=env.DEVICE
-        )
+        mean = torch.zeros(wanted_shape, dtype=self.prec, device=env.DEVICE)
+        stddev = torch.ones(wanted_shape, dtype=self.prec, device=env.DEVICE)
         self.register_buffer("mean", mean)
         self.register_buffer("stddev", stddev)
         self.tebd_dim_input = self.tebd_dim if self.type_one_side else self.tebd_dim * 2
@@ -568,8 +564,6 @@ def forward(
                 # nfnl x nnei x ng
                 # gg = gg_s * gg_t + gg_s
                 gg_t = gg_t.reshape(-1, gg_t.size(-1))
-                # Convert all tensors to the required precision at once
-                ss, rr, gg_t = (t.to(self.prec) for t in (ss, rr, gg_t))
                 xyz_scatter = torch.ops.deepmd.tabulate_fusion_se_atten(
                     self.compress_data[0].contiguous(),
                     self.compress_info[0].cpu().contiguous(),

deepmd/pt/model/descriptor/se_r.py

@@ -456,6 +456,8 @@ def forward(
             The smooth switch function.
 
         """
+        # cast the input to internal precsion
+        coord_ext = coord_ext.to(dtype=self.prec)
         del mapping, comm_dict
         nf = nlist.shape[0]
         nloc = nlist.shape[1]
@@ -474,7 +476,6 @@ def forward(
 
         assert self.filter_layers is not None
         dmatrix = dmatrix.view(-1, self.nnei, 1)
-        dmatrix = dmatrix.to(dtype=self.prec)
         nfnl = dmatrix.shape[0]
         # pre-allocate a shape to pass jit
         xyz_scatter = torch.zeros(
@@ -519,7 +520,7 @@ def forward(
             None,
             None,
             None,
-            sw,
+            sw.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
         )
 
     def set_stat_mean_and_stddev(

deepmd/pt/model/descriptor/se_t.py

@@ -154,6 +154,7 @@ def __init__(
         super().__init__()
         self.type_map = type_map
         self.compress = False
+        self.prec = PRECISION_DICT[precision]
         self.seat = DescrptBlockSeT(
             rcut,
             rcut_smth,
@@ -373,7 +374,18 @@ def forward(
             The smooth switch function.
 
         """
-        return self.seat.forward(nlist, coord_ext, atype_ext, None, mapping)
+        # cast the input to internal precsion
+        coord_ext = coord_ext.to(dtype=self.prec)
+        g1, rot_mat, g2, h2, sw = self.seat.forward(
+            nlist, coord_ext, atype_ext, None, mapping
+        )
+        return (
+            g1.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            None,
+            None,
+            None,
+            sw.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+        )
 
     def set_stat_mean_and_stddev(
         self,
@@ -801,7 +813,6 @@ def forward(
             protection=self.env_protection,
         )
         dmatrix = dmatrix.view(-1, self.nnei, 4)
-        dmatrix = dmatrix.to(dtype=self.prec)
         nfnl = dmatrix.shape[0]
         # pre-allocate a shape to pass jit
         result = torch.zeros(
@@ -832,8 +843,6 @@ def forward(
                 env_ij = torch.einsum("ijm,ikm->ijk", rr_i, rr_j)
                 if self.compress:
                     ebd_env_ij = env_ij.view(-1, 1)
-                    ebd_env_ij = ebd_env_ij.to(dtype=self.prec)
-                    env_ij = env_ij.to(dtype=self.prec)
                     res_ij = torch.ops.deepmd.tabulate_fusion_se_t(
                         compress_data_ii.contiguous(),
                         compress_info_ii.cpu().contiguous(),
@@ -853,7 +862,7 @@ def forward(
         # xyz_scatter /= (self.nnei * self.nnei)
         result = result.view(nf, nloc, self.filter_neuron[-1])
         return (
-            result.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            result,
             None,
             None,
             None,

deepmd/pt/model/descriptor/se_t_tebd.py

@@ -161,6 +161,7 @@ def __init__(
             smooth=smooth,
             seed=child_seed(seed, 1),
         )
+        self.prec = PRECISION_DICT[precision]
         self.use_econf_tebd = use_econf_tebd
         self.type_map = type_map
         self.smooth = smooth
@@ -441,12 +442,14 @@ def forward(
             The smooth switch function. shape: nf x nloc x nnei
 
         """
+        # cast the input to internal precsion
+        extended_coord = extended_coord.to(dtype=self.prec)
         del mapping
         nframes, nloc, nnei = nlist.shape
         nall = extended_coord.view(nframes, -1).shape[1] // 3
         g1_ext = self.type_embedding(extended_atype)
         g1_inp = g1_ext[:, :nloc, :]
-        g1, g2, h2, rot_mat, sw = self.se_ttebd(
+        g1, _, _, _, sw = self.se_ttebd(
             nlist,
             extended_coord,
             extended_atype,
@@ -456,7 +459,13 @@ def forward(
         if self.concat_output_tebd:
             g1 = torch.cat([g1, g1_inp], dim=-1)
 
-        return g1, rot_mat, g2, h2, sw
+        return (
+            g1.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            None,
+            None,
+            None,
+            sw.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+        )
 
     @classmethod
     def update_sel(
@@ -540,12 +549,8 @@ def __init__(
         self.reinit_exclude(exclude_types)
 
         wanted_shape = (self.ntypes, self.nnei, 4)
-        mean = torch.zeros(
-            wanted_shape, dtype=env.GLOBAL_PT_FLOAT_PRECISION, device=env.DEVICE
-        )
-        stddev = torch.ones(
-            wanted_shape, dtype=env.GLOBAL_PT_FLOAT_PRECISION, device=env.DEVICE
-        )
+        mean = torch.zeros(wanted_shape, dtype=self.prec, device=env.DEVICE)
+        stddev = torch.ones(wanted_shape, dtype=self.prec, device=env.DEVICE)
         self.register_buffer("mean", mean)
         self.register_buffer("stddev", stddev)
         self.tebd_dim_input = self.tebd_dim * 2
@@ -849,7 +854,7 @@ def forward(
         # nf x nl x ng
         result = res_ij.view(nframes, nloc, self.filter_neuron[-1])
         return (
-            result.to(dtype=env.GLOBAL_PT_FLOAT_PRECISION),
+            result,
             None,
             None,
             None,

deepmd/pt/model/network/mlp.py

@@ -200,7 +200,8 @@ def forward(
             The output.
         """
         ori_prec = xx.dtype
-        xx = xx.to(self.prec)
+        if not env.DP_DTYPE_PROMOTION_STRICT:
+            xx = xx.to(self.prec)
         yy = (
             torch.matmul(xx, self.matrix) + self.bias
             if self.bias is not None
@@ -215,7 +216,8 @@ def forward(
                 yy += torch.concat([xx, xx], dim=-1)
             else:
                 yy = yy
-        yy = yy.to(ori_prec)
+        if not env.DP_DTYPE_PROMOTION_STRICT:
+            yy = yy.to(ori_prec)
         return yy
 
     def serialize(self) -> dict:

deepmd/pt/model/task/dipole.py

@@ -180,6 +180,8 @@ def forward(
     ):
         nframes, nloc, _ = descriptor.shape
         assert gr is not None, "Must provide the rotation matrix for dipole fitting."
+        # cast the input to internal precsion
+        gr = gr.to(self.prec)
         # (nframes, nloc, m1)
         out = self._forward_common(descriptor, atype, gr, g2, h2, fparam, aparam)[
             self.var_name