From d80ac2fd2df4e7fb8a28acfa512bb11472b5cc99 Mon Sep 17 00:00:00 2001
From: liugz18 <57401541+liugz18@users.noreply.github.com>
Date: 星期四, 18 七月 2024 21:34:55 +0800
Subject: [PATCH] Rename 'res' in line 514 to avoid with naming conflict with line 365
---
funasr/train_utils/load_pretrained_model.py | 159 ++++++++++++++++++++++------------------------------
1 files changed, 67 insertions(+), 92 deletions(-)
diff --git a/funasr/train_utils/load_pretrained_model.py b/funasr/train_utils/load_pretrained_model.py
index a6596a0..8ed613c 100644
--- a/funasr/train_utils/load_pretrained_model.py
+++ b/funasr/train_utils/load_pretrained_model.py
@@ -7,121 +7,96 @@
import torch
import torch.nn
import torch.optim
-
-
-def filter_state_dict(
- dst_state: Dict[str, Union[float, torch.Tensor]],
- src_state: Dict[str, Union[float, torch.Tensor]],
-):
- """Filter name, size mismatch instances between dicts.
-
- Args:
- dst_state: reference state dict for filtering
- src_state: target state dict for filtering
-
- """
- match_state = {}
- for key, value in src_state.items():
- if key in dst_state and (dst_state[key].size() == src_state[key].size()):
- match_state[key] = value
- else:
- if key not in dst_state:
- logging.warning(
- f"Filter out {key} from pretrained dict"
- + " because of name not found in target dict"
- )
- else:
- logging.warning(
- f"Filter out {key} from pretrained dict"
- + " because of size mismatch"
- + f"({dst_state[key].size()}-{src_state[key].size()})"
- )
- return match_state
+import pdb
def load_pretrained_model(
- init_param: str,
+ path: str,
model: torch.nn.Module,
- ignore_init_mismatch: bool,
+ ignore_init_mismatch: bool = True,
map_location: str = "cpu",
oss_bucket=None,
+ scope_map=[],
+ excludes=None,
+ **kwargs,
):
"""Load a model state and set it to the model.
Args:
- init_param: <file_path>:<src_key>:<dst_key>:<exclude_Keys>
+ init_param: <file_path>:<src_key>:<dst_key>:<exclude_Keys>
Examples:
- >>> load_pretrained_model("somewhere/model.pb", model)
- >>> load_pretrained_model("somewhere/model.pb:decoder:decoder", model)
- >>> load_pretrained_model("somewhere/model.pb:decoder:decoder:", model)
- >>> load_pretrained_model(
- ... "somewhere/model.pb:decoder:decoder:decoder.embed", model
- ... )
- >>> load_pretrained_model("somewhere/decoder.pb::decoder", model)
+
"""
- sps = init_param.split(":", 4)
- if len(sps) == 4:
- path, src_key, dst_key, excludes = sps
- elif len(sps) == 3:
- path, src_key, dst_key = sps
- excludes = None
- elif len(sps) == 2:
- path, src_key = sps
- dst_key, excludes = None, None
- else:
- (path,) = sps
- src_key, dst_key, excludes = None, None, None
- if src_key == "":
- src_key = None
- if dst_key == "":
- dst_key = None
- if dst_key is None:
- obj = model
- else:
+ obj = model
+ dst_state = obj.state_dict()
- def get_attr(obj: Any, key: str):
- """Get an nested attribute.
-
- >>> class A(torch.nn.Module):
- ... def __init__(self):
- ... super().__init__()
- ... self.linear = torch.nn.Linear(10, 10)
- >>> a = A()
- >>> assert A.linear.weight is get_attr(A, 'linear.weight')
-
- """
- if key.strip() == "":
- return obj
- for k in key.split("."):
- obj = getattr(obj, k)
- return obj
-
- obj = get_attr(model, dst_key)
+ logging.info(f"ckpt: {path}")
if oss_bucket is None:
src_state = torch.load(path, map_location=map_location)
else:
buffer = BytesIO(oss_bucket.get_object(path).read())
src_state = torch.load(buffer, map_location=map_location)
+
+ src_state = src_state["state_dict"] if "state_dict" in src_state else src_state
+ src_state = src_state["model_state_dict"] if "model_state_dict" in src_state else src_state
src_state = src_state["model"] if "model" in src_state else src_state
+
+ if isinstance(scope_map, str):
+ scope_map = scope_map.split(",")
+ scope_map += ["module.", "None"]
+ logging.info(f"scope_map: {scope_map}")
+
if excludes is not None:
- for e in excludes.split(","):
- src_state = {k: v for k, v in src_state.items() if not k.startswith(e)}
+ if isinstance(excludes, str):
+ excludes = excludes.split(",")
- if src_key is not None:
- src_state = {
- k[len(src_key) + 1 :]: v
- for k, v in src_state.items()
- if k.startswith(src_key)
- }
+ logging.info(f"excludes: {excludes}")
- dst_state = obj.state_dict()
- if ignore_init_mismatch:
- src_state = filter_state_dict(dst_state, src_state)
+ for k in dst_state.keys():
+ excludes_flag = False
+ if excludes is not None:
+ for k_ex in excludes:
+ if k.startswith(k_ex):
+ logging.info(f"key: {k} matching: {k_ex}, excluded")
+ excludes_flag = True
+ break
+ if excludes_flag:
+ continue
- logging.debug("Loaded src_state keys: {}".format(src_state.keys()))
- logging.debug("Loaded dst_state keys: {}".format(dst_state.keys()))
- dst_state.update(src_state)
- obj.load_state_dict(dst_state)
+ k_src = k
+
+ if scope_map is not None:
+ src_prefix = ""
+ dst_prefix = ""
+ for i in range(0, len(scope_map), 2):
+ src_prefix = scope_map[i] if scope_map[i].lower() != "none" else ""
+ dst_prefix = scope_map[i + 1] if scope_map[i + 1].lower() != "none" else ""
+
+ if dst_prefix == "" and (src_prefix + k) in src_state.keys():
+ k_src = src_prefix + k
+ if not k_src.startswith("module."):
+ logging.info(f"init param, map: {k} from {k_src} in ckpt")
+ elif (
+ k.startswith(dst_prefix)
+ and k.replace(dst_prefix, src_prefix, 1) in src_state.keys()
+ ):
+ k_src = k.replace(dst_prefix, src_prefix, 1)
+ if not k_src.startswith("module."):
+ logging.info(f"init param, map: {k} from {k_src} in ckpt")
+
+ if k_src in src_state.keys():
+ if ignore_init_mismatch and dst_state[k].shape != src_state[k_src].shape:
+ logging.info(
+ f"ignore_init_mismatch:{ignore_init_mismatch}, dst: {k, dst_state[k].shape}, src: {k_src, src_state[k_src].shape}"
+ )
+ else:
+ dst_state[k] = src_state[k_src]
+
+ else:
+ print(f"Warning, miss key in ckpt: {k}, {path}")
+
+ flag = obj.load_state_dict(dst_state, strict=True)
+ logging.info(f"Loading ckpt: {path}, status: {flag}")
--
Gitblit v1.9.1