Skip to content

BLOOM - modifying slow tests #17963

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

Closed
younesbelkada wants to merge 3 commits into from
Closed

BLOOM - modifying slow tests #17963

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
564314d
modifying tests
younesbelkada Jun 30, 2022
6982516
make quality
younesbelkada Jun 30, 2022
4d936dd
add correct revision for backward compt
younesbelkada Jul 4, 2022
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
52 changes: 15 additions & 37 deletions tests/models/bloom/test_modeling_bloom.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -378,26 +378,26 @@ def test_model_from_pretrained(self):
@require_torch_gpu
def test_simple_generation(self):
path_350m = "bigscience/bloom-350m"
model = BloomForCausalLM.from_pretrained(path_350m, torch_dtype="auto", use_cache=True).cuda()
model = BloomForCausalLM.from_pretrained(
path_350m, torch_dtype=torch.float32, use_cache=False, revision="gs555750"
Copy link
Contributor

@patrickvonplaten patrickvonplaten Jul 4, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why not use the most up-to-date model here?

).cuda()
model = model.eval()
tokenizer = BloomTokenizerFast.from_pretrained(path_350m)

input_sentence = "I enjoy walking with my cute dog"
EXPECTED_OUTPUT = (
"I enjoy walking with my cute dog, and I love to watch the kids play. I am a very active person, and I am"
" a very good listener. I am a very good person, and I am a very good person. I am a"
)
EXPECTED_OUTPUT = "I enjoy walking with my cute dog, and I love to watch the kids"

input_ids = tokenizer.encode(input_sentence, return_tensors="pt")
greedy_output = model.generate(input_ids.cuda(), max_length=50)

greedy_output = model.generate(input_ids.cuda(), max_length=15)
self.assertEqual(tokenizer.decode(greedy_output[0], skip_special_tokens=True), EXPECTED_OUTPUT)

@slow
@require_torch_gpu
def test_batch_generation(self):
path_350m = "bigscience/bloom-350m"
model = BloomForCausalLM.from_pretrained(path_350m, torch_dtype="auto", use_cache=True).cuda()
model = BloomForCausalLM.from_pretrained(
path_350m, torch_dtype="auto", use_cache=True, revision="gs555750"
Copy link
Contributor

@patrickvonplaten patrickvonplaten Jul 27, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why do we add a revision here?

Copy link
Contributor Author

@younesbelkada younesbelkada Jul 27, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think that we don't need it there as well you are right

).cuda()
model = model.eval()
tokenizer = BloomTokenizerFast.from_pretrained(path_350m, padding_side="left")

Expand All @@ -417,7 +417,9 @@ def test_batch_generation(self):
@require_torch_gpu
def test_batch_generation_padd(self):
path_350m = "bigscience/bloom-350m"
model = BloomForCausalLM.from_pretrained(path_350m, torch_dtype="auto", use_cache=True).cuda()
model = BloomForCausalLM.from_pretrained(
path_350m, torch_dtype=torch.float32, use_cache=True, revision="gs555750"
).cuda()
model = model.eval()
tokenizer = BloomTokenizerFast.from_pretrained(path_350m, padding_side="left")

Expand All @@ -428,7 +430,10 @@ def test_batch_generation_padd(self):
input_ids_without_pad = tokenizer.encode(input_sentence_without_pad, return_tensors="pt")

greedy_output = model.generate(
input_ids["input_ids"].cuda(), attention_mask=input_ids["attention_mask"], max_length=50, do_sample=False
input_ids["input_ids"].cuda(),
attention_mask=input_ids["attention_mask"].cuda(),
max_length=50,
do_sample=False,
)
greedy_output_without_pad = model.generate(input_ids_without_pad.cuda(), max_length=50, do_sample=False)

Expand All @@ -441,33 +446,6 @@ def test_batch_generation_padd(self):
tokenizer.decode(greedy_output_without_pad[0, :-3], skip_special_tokens=True),
)

@slow
def test_right_left_batched_input(self):
Copy link
Contributor

@patrickvonplaten patrickvonplaten Jul 27, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why would we delete this?

Copy link
Contributor Author

@younesbelkada younesbelkada Jul 27, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Because we always use padding_side=left for batched generation for autoregressive models so for me there is no point keeping this test

path_1b3 = "bigscience/bloom-1b3"
model = BloomForCausalLM.from_pretrained(path_1b3, use_cache=True)
model = model.eval()

tokenizer = BloomTokenizerFast.from_pretrained(path_1b3)
tokenizer.padding_side = "right"

inputs = ["Hello there", "Joe Biden is the president of the"]
inputs_right = tokenizer(inputs, return_tensors="pt", padding=True)

tokenizer.padding_side = "left"
inputs_left = tokenizer(inputs, return_tensors="pt", padding=True)

# test token values are different
self.assertNotEqual(inputs_right["input_ids"].tolist(), inputs_left["input_ids"].tolist())

# test reconstructions are the same
outputs_right = model.generate(**inputs_right, max_length=10, do_sample=False)
outputs_left = model.generate(**inputs_left, max_length=10, do_sample=False)

self.assertEqual(
tokenizer.decode(outputs_right[0], skip_special_tokens=True),
tokenizer.decode(outputs_left[0], skip_special_tokens=True),
)


@require_torch
class BloomEmbeddingTest(unittest.TestCase):
Expand Down