Mirrors/modelscope
1
0
Fork 0
mirror of https://github.com/modelscope/modelscope.git synced 2025-12-24 20:19:22 +01:00
Code Issues Packages Projects Releases Wiki Activity

Fix some words (#261)

Browse Source
This commit is contained in:
chenxujun
2023-04-10 10:16:59 +08:00
committed by GitHub
parent 6dd94ff2bc
commit 94eeaffbf4
13 changed files with 25 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
modelscope/cli/download.py
View File

@@ -7,7 +7,7 @@ from modelscope.hub.snapshot_download import snapshot_download
def subparser_func(args):
""" Fuction which will be called for a specific sub parser.
""" Function which will be called for a specific sub parser.
"""
return DownloadCMD(args)

2
modelscope/cli/modelcard.py
View File

@@ -18,7 +18,7 @@ template_path = os.path.join(curren_path, 'template')
def subparser_func(args):
""" Fuction which will be called for a specific sub parser.
""" Function which will be called for a specific sub parser.
"""
return ModelCardCMD(args)

2
modelscope/cli/pipeline.py
View File

@@ -13,7 +13,7 @@ template_path = os.path.join(curren_path, 'template')
def subparser_func(args):
""" Fuction which will be called for a specific sub parser.
""" Function which will be called for a specific sub parser.
"""
return PipelineCMD(args)

2
modelscope/cli/plugins.py
View File

@@ -9,7 +9,7 @@ plugins_manager = PluginsManager()
def subparser_func(args):
""" Fuction which will be called for a specific sub parser.
""" Function which will be called for a specific sub parser.
"""
return PluginsCMD(args)

4
modelscope/cli/template/template.tpl
View File

@@ -24,7 +24,7 @@ class ${model_name}(TorchModel):
def init_model(self, **kwargs):
"""Provide default implementation based on TorchModel and user can reimplement it.
include init model and load ckpt from the model_dir, maybe include preprocessor
if nothing to do, then return lambdx x: x
if nothing to do, then return lambda x: x
"""
return lambda x: x
@@ -41,7 +41,7 @@ class ${preprocessor_name}(Preprocessor):
def init_preprocessor(self, **kwarg):
""" Provide default implementation based on preprocess_cfg and user can reimplement it.
if nothing to do, then return lambdx x: x
if nothing to do, then return lambda x: x
"""
return lambda x: x

2
modelscope/exporters/nlp/model_for_token_classification_exporter.py
View File

@@ -89,7 +89,7 @@ class ModelForSequenceClassificationExporter(TorchModelExporter):
outputs_origin = list(numpify_tensor_nested(outputs_origin))
outputs_origin = [outputs_origin[0]
] # keeo `predictions`, drop other outputs
] # keep `predictions`, drop other outputs
np_dummy_inputs = numpify_tensor_nested(dummy_inputs)
np_dummy_inputs['label_mask'] = np_dummy_inputs['label_mask'].astype(

2
modelscope/models/audio/kws/farfield/fsmn_sele_v2.py
View File

@@ -18,7 +18,7 @@ class FSMNUnit(nn.Module):
Args:
dimlinear: input / output dimension
dimproj: fsmn input / output dimension
lorder: left ofder
lorder: left order
rorder: right order
"""
super(FSMNUnit, self).__init__()

2
modelscope/models/audio/kws/nearfield/fsmn.py
View File

@@ -435,7 +435,7 @@ class FSMN(nn.Module):
"""
Args:
input (torch.Tensor): Input tensor (B, T, D)
in_cache(torhc.Tensor): (B, D, C), C is the accumulated cache size
in_cache(torch.Tensor): (B, D, C), C is the accumulated cache size
"""
# print("FSMN forward!!!!")

8
modelscope/models/audio/kws/nearfield/model.py
View File

@@ -39,8 +39,8 @@ class FSMNDecorator(TorchModel):
model_dir (str): the model path.
cmvn_file (str): cmvn file
backbone (dict): params related to backbone
input_dim (int): input dimention of network
output_dim (int): output dimention of network
input_dim (int): input dimension of network
output_dim (int): output dimension of network
training (bool): training or inference mode
"""
super().__init__(model_dir, *args, **kwargs)
@@ -108,7 +108,7 @@ class FSMNDecorator(TorchModel):
class KWSModel(nn.Module):
"""Our model consists of four parts:
1. global_cmvn: Optional, (idim, idim)
2. preprocessing: feature dimention projection, (idim, hdim)
2. preprocessing: feature dimension projection, (idim, hdim)
3. backbone: backbone or feature extractor of the whole network, (hdim, hdim)
4. classifier: output layer or classifier of KWS model, (hdim, odim)
5. activation:
@@ -133,7 +133,7 @@ class KWSModel(nn.Module):
odim (int): output dimension of network
hdim (int): hidden dimension of network
global_cmvn (nn.Module): cmvn for input feature, (idim, idim)
preprocessing (nn.Module): feature dimention projection, (idim, hdim)
preprocessing (nn.Module): feature dimension projection, (idim, hdim)
backbone (nn.Module): backbone or feature extractor of the whole network, (hdim, hdim)
classifier (nn.Module): output layer or classifier of KWS model, (hdim, odim)
activation (nn.Module): nn.Identity for training, nn.Sigmoid for inference

2
modelscope/models/cv/action_recognition/temporal_patch_shift_transformer.py
View File

@@ -871,7 +871,7 @@ class SwinTransformer2D_TPS(nn.Module):
Args:
logger (logging.Logger): The logger used to print
debugging infomation.
debugging information.
"""
checkpoint = torch.load(self.pretrained, map_location='cpu')
state_dict = checkpoint['model']

2
modelscope/models/cv/body_3d_keypoints/cannonical_pose/body_3d_pose.py
View File

@@ -181,7 +181,7 @@ class BodyKeypointsDetection3D(TorchModel):
"camera_pose": Tensor, [1, NUM_FRAME, OUT_NUM_JOINTS, OUT_3D_FEATURE_DIM],
3D human pose keypoints in camera frame.
"camera_traj": Tensor, [1, NUM_FRAME, 1, 3],
root keypoints coordinates in camere frame.
root keypoints coordinates in camera frame.
"""
inputs_2d = input['inputs_2d']
pose2d_rr = input['pose2d_rr']

10
modelscope/models/cv/body_3d_keypoints/hdformer/directed_graph.py
View File

@@ -46,14 +46,14 @@ class DiGraph():
super().__init__()
self.num_nodes = len(skeleton.parents())
self.directed_edges_hop1 = [
(parrent, child)
for child, parrent in enumerate(skeleton.parents()) if parrent >= 0
(parent, child) for child, parent in enumerate(skeleton.parents())
if parent >= 0
]
self.directed_edges_hop2 = [(0, 1, 2), (0, 4, 5), (0, 7, 8), (1, 2, 3),
(4, 5, 6), (7, 8, 9),
(7, 8, 11), (7, 8, 14), (8, 9, 10),
(8, 11, 12), (8, 14, 15), (11, 12, 13),
(14, 15, 16)] # (parrent, child)
(14, 15, 16)] # (parent, child)
self.directed_edges_hop3 = [(0, 1, 2, 3), (0, 4, 5, 6), (0, 7, 8, 9),
(7, 8, 9, 10), (7, 8, 11, 12),
(7, 8, 14, 15), (8, 11, 12, 13),
@@ -112,8 +112,8 @@ class Graph():
# edge is a list of [child, parent] paris
self.num_node = len(skeleton.parents())
self_link = [(i, i) for i in range(self.num_node)]
neighbor_link = [(child, parrent)
for child, parrent in enumerate(skeleton.parents())]
neighbor_link = [(child, parent)
for child, parent in enumerate(skeleton.parents())]
self.self_link = self_link
self.neighbor_link = neighbor_link
self.edge = self_link + neighbor_link

11
modelscope/models/cv/cartoon/facelib/facer.py
View File

@@ -94,7 +94,7 @@ class FaceAna():
sorted_bboxes = [bboxes[x] for x in picked]
return np.array(sorted_bboxes)
def judge_boxs(self, previuous_bboxs, now_bboxs):
def judge_boxs(self, previous_bboxs, now_bboxs):
def iou(rec1, rec2):
@@ -116,17 +116,16 @@ class FaceAna():
return intersect / (sum_area - intersect)
if previuous_bboxs is None:
if previous_bboxs is None:
return now_bboxs
result = []
for i in range(now_bboxs.shape[0]):
contain = False
for j in range(previuous_bboxs.shape[0]):
if iou(now_bboxs[i], previuous_bboxs[j]) > self.iou_thres:
result.append(
self.smooth(now_bboxs[i], previuous_bboxs[j]))
for j in range(previous_bboxs.shape[0]):
if iou(now_bboxs[i], previous_bboxs[j]) > self.iou_thres:
result.append(self.smooth(now_bboxs[i], previous_bboxs[j]))
contain = True
break
if not contain: