Skip to content

Dev #194

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
helloyongyang wants to merge 2 commits into from
Closed

Dev #194

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
79586ab
update internvl2
helloyongyang Nov 17, 2024
09ff699
update
helloyongyang Nov 17, 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
1 change: 0 additions & 1 deletion llmc/data/dataset/base_dataset.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -255,7 +255,6 @@ def get_calib_dataset(self):
samples = self.get_calib_samples()
if self.calib_dataset_type in ['txt', 'img', 'img_txt']:
logger.info(f'len(samples) all : {len(samples)}')
assert len(samples) % int(os.environ['WORLD_SIZE']) == 0
samples = samples[int(os.environ['RANK'])::int(os.environ['WORLD_SIZE'])]
logger.info(f'len(samples) rank : {len(samples)}')
calib_samples = []
Expand Down
91 changes: 16 additions & 75 deletions llmc/models/internvl2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -129,21 +129,25 @@ def build_model(self):
self.vlm_model.img_context_token_id = tokenizer.convert_tokens_to_ids(IMG_CONTEXT_TOKEN)

def batch_process(self, img_qas):
if len(img_qas) == 1:
return self.single_process(img_qas[0])
tokenizer = AutoTokenizer.from_pretrained(self.model_path, trust_remote_code=True)
questions = []
pixel_values_list = []
num_patches_list = []
for idx in range(len(img_qas)):
img_path = img_qas[idx]['img']
pixel_values = load_image(img_path, max_num=12).to(
next(self.vlm_model.parameters()).dtype
)
pixel_values_list.append(pixel_values)
num_patches_list.append(pixel_values.size(0))
questions.append(f"<image>\n{img_qas[idx]['question']}")
pixel_values = torch.cat(pixel_values_list, dim=0)
_num_patches_i = []
if img_path is not None:
if not isinstance(img_path, list):
img_path = [img_path]
for img_idx in range(len(img_path)):
pixel_values = load_image(img_path[img_idx], max_num=12).to(
next(self.vlm_model.parameters()).dtype
)
pixel_values_list.append(pixel_values)
_num_patches_i.append(pixel_values.size(0))
num_patches_list.append(_num_patches_i)
questions.append(img_qas[idx]['question'])
pixel_values = torch.cat(pixel_values_list, dim=0) if len(pixel_values_list) > 0 else None
generation_config = dict()

IMG_CONTEXT_TOKEN = '<IMG_CONTEXT>'
Expand All @@ -163,12 +167,10 @@ def batch_process(self, img_qas):
template.append_message(template.roles[0], question)
template.append_message(template.roles[1], None)
query = template.get_prompt()
image_tokens = (IMG_START_TOKEN +
IMG_CONTEXT_TOKEN * self.vlm_model.num_image_token * num_patches +
IMG_END_TOKEN)
query = query.replace('<image>', image_tokens, 1)
for _num_patches_i in num_patches:
image_tokens = IMG_START_TOKEN + IMG_CONTEXT_TOKEN * self.vlm_model.num_image_token * _num_patches_i + IMG_END_TOKEN # noqa
query = query.replace('<image>', image_tokens, 1)
queries.append(query)

tokenizer.padding_side = 'left'
model_inputs = tokenizer(queries, return_tensors='pt', padding=True)
input_ids = model_inputs['input_ids']
Expand All @@ -183,64 +185,3 @@ def batch_process(self, img_qas):
**generation_config
}
return inputs

def single_process(self, img_qa):
tokenizer = AutoTokenizer.from_pretrained(self.model_path, trust_remote_code=True)
num_patches_list = None
pixel_values_list = []
if img_qa['img'] is not None:
if isinstance(img_qa['img'], list):
num_patches_list = []
for img_idx in range(len(img_qa['img'])):
pixel_values = load_image(img_qa['img'][img_idx], max_num=12).to(
next(self.vlm_model.parameters()).dtype
)
pixel_values_list.append(pixel_values)
num_patches_list.append(pixel_values.size(0))
pixel_values = torch.cat(pixel_values_list, dim=0)
else:
pixel_values = load_image(img_qa['img'], max_num=12).to(
next(self.vlm_model.parameters()).dtype
)
else:
pixel_values = None
question = img_qa['question']
if pixel_values is not None and '<image>' not in question:
question = '<image>\n' + question
if num_patches_list is None:
num_patches_list = [pixel_values.shape[0]] if pixel_values is not None else []
generation_config = dict()

IMG_CONTEXT_TOKEN = '<IMG_CONTEXT>'
IMG_START_TOKEN = '<img>'
IMG_END_TOKEN = '</img>'
try:
template = get_conv_template(self.vlm_model.template)
except Exception:
raise Exception(
'InternLM2 conversation.py not be found. '
'Please copy it from model path to llmc/models.'
)
template.system_message = self.vlm_model.system_message
template.append_message(template.roles[0], question)
template.append_message(template.roles[1], None)
query = template.get_prompt()
for num_patches in num_patches_list:
image_tokens = (IMG_START_TOKEN +
IMG_CONTEXT_TOKEN * self.vlm_model.num_image_token * num_patches +
IMG_END_TOKEN)
query = query.replace('<image>', image_tokens, 1)

model_inputs = tokenizer(query, return_tensors='pt')
input_ids = model_inputs['input_ids']
attention_mask = model_inputs['attention_mask']
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep)
generation_config['eos_token_id'] = eos_token_id

inputs = {
'pixel_values': pixel_values,
'input_ids': input_ids,
'attention_mask': attention_mask,
**generation_config
}
return inputs
Loading