Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix multitask prompt tuning other inits #8

Merged
merged 3 commits into from
Feb 9, 2024
Merged

Fix multitask prompt tuning other inits #8

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
81a33f3
mpt fix
mayank31398 Feb 8, 2024
a487da2
drop print statement
mayank31398 Feb 8, 2024
049f927
drop print statement
mayank31398 Feb 8, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 24 additions & 4 deletions tests/test_multitask_prompt_tuning.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -249,15 +249,34 @@ def test_bf16_inference(self) -> None:
mpt = mpt.to(self.torch_device)
_ = mpt.generate(input_ids=input_ids, task_ids=task_ids)

def test_generate_text_with_random_init(self) -> None:
model = LlamaForCausalLM(self._create_test_llama_config())

config = self._create_multitask_prompt_tuning_config()
config.prompt_tuning_init = MultitaskPromptTuningInit.RANDOM

model = get_peft_model(model, config)
model = model.to(self.torch_device)

input_ids = torch.LongTensor([[1, 1, 1], [2, 1, 2]]).to(self.torch_device)
attention_mask = torch.LongTensor([[1, 1, 1], [1, 0, 1]]).to(self.torch_device)
task_ids = torch.LongTensor([0]).to(self.torch_device)

# check if `generate` works
_ = model.generate(input_ids=input_ids, attention_mask=attention_mask, task_ids=task_ids)

with self.assertRaises(ValueError):
# check if `generate` raises an error if task_ids are not passed
_ = model.generate(input_ids, attention_mask=attention_mask)

@parameterized.expand(
[
MultitaskPromptTuningInit.TEXT,
MultitaskPromptTuningInit.AVERAGE_SOURCE_TASKS,
MultitaskPromptTuningInit.EXACT_SOURCE_TASK,
MultitaskPromptTuningInit.ONLY_SOURCE_SHARED,
],
)
def test_generate_text(self, prompt_tuning_init) -> None:
def test_generate_text_with_other_init(self, prompt_tuning_init) -> None:
with tempfile.TemporaryDirectory() as tmp_dirname:
model = LlamaForCausalLM(self._create_test_llama_config())
model = get_peft_model(model, self._create_multitask_prompt_tuning_config())
Expand All @@ -273,6 +292,7 @@ def test_generate_text(self, prompt_tuning_init) -> None:
prompt_tuning_init=prompt_tuning_init,
prompt_tuning_init_state_dict_path=os.path.join(tmp_dirname, WEIGHTS_NAME),
)
model = LlamaForCausalLM(self._create_test_llama_config())
model = get_peft_model(model, config)
model = model.to(self.torch_device)

Expand All @@ -283,6 +303,6 @@ def test_generate_text(self, prompt_tuning_init) -> None:
# check if `generate` works
_ = model.generate(input_ids=input_ids, attention_mask=attention_mask, task_ids=task_ids)

with self.assertRaises(TypeError):
# check if `generate` raises an error if no positional arguments are passed
with self.assertRaises(ValueError):
# check if `generate` raises an error if task_ids are not passed
_ = model.generate(input_ids, attention_mask=attention_mask)