github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-09-09 04:50:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

[misc] update pre-commit and run all files (#4752)

Browse Source 
* [misc] update pre-commit

* [misc] run pre-commit

* [misc] remove useless configuration files

* [misc] ignore cuda for clang-format
This commit is contained in:
Hongxin Liu
2023-09-19 14:20:26 +08:00
committed by GitHub
parent 3c6b831c26
commit 079bf3cb26
1268 changed files with 50037 additions and 38444 deletions
Download Patch File Download Diff File Expand all files Collapse all files

101
tests/kit/model_zoo/transformers/bloom.py
View File

@@ -25,7 +25,7 @@ def data_gen_for_lm():
# LM data gen
# the `labels` of LM is the token of the output, cause no padding, use `input_ids` as `labels`
data = data_gen()
data['labels'] = data['input_ids'].clone()
data["labels"] = data["input_ids"].clone()
return data
@@ -33,14 +33,14 @@ def data_gen_for_token_classification():
# token classification data gen
# `labels` is the type not the token id for token classification, 0 or 1
data = data_gen()
data['labels'] = torch.tensor([[0, 0, 0, 0, 0, 0, 0, 0]], dtype=torch.int64)
data["labels"] = torch.tensor([[0, 0, 0, 0, 0, 0, 0, 0]], dtype=torch.int64)
return data
def data_gen_for_sequence_classification():
# sequence classification data gen
data = data_gen()
data['labels'] = torch.tensor([0], dtype=torch.int64)
data["labels"] = torch.tensor([0], dtype=torch.int64)
return data
@@ -54,62 +54,69 @@ def data_gen_for_question_answering():
input_ids = torch.tensor(
[[57647, 1620, 23967, 620, 107373, 34, 91514, 620, 107373, 1620, 267, 35378, 48946, 18161, 48946, 18161]],
dtype=torch.int64)
dtype=torch.int64,
)
attention_mask = torch.tensor([[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]], dtype=torch.int64)
start_positions = torch.tensor([1], dtype=torch.int64)
end_positions = torch.tensor([10], dtype=torch.int64)
return dict(input_ids=input_ids,
attention_mask=attention_mask,
start_positions=start_positions,
end_positions=end_positions)
return dict(
input_ids=input_ids, attention_mask=attention_mask, start_positions=start_positions, end_positions=end_positions
)
# define output transform function
output_transform_fn = lambda x: x
# define loss function
loss_fn_for_bloom_model = lambda x: torch.nn.functional.mse_loss(x.last_hidden_state,
torch.ones_like(x.last_hidden_state))
loss_fn_for_bloom_model = lambda x: torch.nn.functional.mse_loss(
x.last_hidden_state, torch.ones_like(x.last_hidden_state)
)
loss_fn_for_causal_lm = lambda x: x.loss
loss_fn_for_classification = lambda x: x.loss
loss_fn_for_question_answering = lambda x: x.loss
config = transformers.BloomConfig(n_layer=2,
n_head=4,
vocab_size=250880,
hidden_dropout=0,
attention_dropout=0,
hidden_size=64,
pad_token_id=50256)
config = transformers.BloomConfig(
n_layer=2, n_head=4, vocab_size=250880, hidden_dropout=0, attention_dropout=0, hidden_size=64, pad_token_id=50256
)
# register the following models
model_zoo.register(name='transformers_bloom',
model_fn=lambda: transformers.BloomModel(config),
data_gen_fn=data_gen,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_bloom_model,
model_attribute=ModelAttribute(has_control_flow=True))
model_zoo.register(name='transformers_bloom_for_causal_lm',
model_fn=lambda: transformers.BloomForCausalLM(config),
data_gen_fn=data_gen_for_lm,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_causal_lm,
model_attribute=ModelAttribute(has_control_flow=True))
model_zoo.register(name='transformers_bloom_for_sequence_classification',
model_fn=lambda: transformers.BloomForSequenceClassification(config),
data_gen_fn=data_gen_for_sequence_classification,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_classification,
model_attribute=ModelAttribute(has_control_flow=True))
model_zoo.register(name='transformers_bloom_for_token_classification',
model_fn=lambda: transformers.BloomForTokenClassification(config),
data_gen_fn=data_gen_for_token_classification,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_classification,
model_attribute=ModelAttribute(has_control_flow=True))
model_zoo.register(name='transformers_bloom_for_question_answering',
model_fn=lambda: transformers.BloomForQuestionAnswering(config),
data_gen_fn=data_gen_for_question_answering,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_question_answering,
model_attribute=ModelAttribute(has_control_flow=True))
model_zoo.register(
name="transformers_bloom",
model_fn=lambda: transformers.BloomModel(config),
data_gen_fn=data_gen,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_bloom_model,
model_attribute=ModelAttribute(has_control_flow=True),
)
model_zoo.register(
name="transformers_bloom_for_causal_lm",
model_fn=lambda: transformers.BloomForCausalLM(config),
data_gen_fn=data_gen_for_lm,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_causal_lm,
model_attribute=ModelAttribute(has_control_flow=True),
)
model_zoo.register(
name="transformers_bloom_for_sequence_classification",
model_fn=lambda: transformers.BloomForSequenceClassification(config),
data_gen_fn=data_gen_for_sequence_classification,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_classification,
model_attribute=ModelAttribute(has_control_flow=True),
)
model_zoo.register(
name="transformers_bloom_for_token_classification",
model_fn=lambda: transformers.BloomForTokenClassification(config),
data_gen_fn=data_gen_for_token_classification,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_classification,
model_attribute=ModelAttribute(has_control_flow=True),
)
model_zoo.register(
name="transformers_bloom_for_question_answering",
model_fn=lambda: transformers.BloomForQuestionAnswering(config),
data_gen_fn=data_gen_for_question_answering,
output_transform_fn=output_transform_fn,
loss_fn=loss_fn_for_question_answering,
model_attribute=ModelAttribute(has_control_flow=True),
)