Skip to content

Updated KV_pytorch and ORT inference for VLMs to incorporate image_idx #557

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
+32 −11
Open

Updated KV_pytorch and ORT inference for VLMs to incorporate image_idx #557

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 7 additions & 4 deletions QEfficient/utils/generate_inputs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -249,7 +249,7 @@ def prepare_pytorch_inputs(self):

num_hidden_layers = txt_cfg.num_hidden_layers
num_key_value_heads = txt_cfg.num_key_value_heads
head_dim = txt_cfg.hidden_size // txt_cfg.num_attention_heads
head_dim = getattr(txt_cfg, "head_dim", txt_cfg.hidden_size // txt_cfg.num_attention_heads)
if hasattr(txt_cfg, "cross_attention_layers"):
cross_attention_layers = txt_cfg.cross_attention_layers

Expand Down Expand Up @@ -287,7 +287,7 @@ def prepare_vlm_ort_inputs(self):
txt_cfg = self.config.llm_config
num_hidden_layers = txt_cfg.num_hidden_layers
num_key_value_heads = txt_cfg.num_key_value_heads
head_dim = txt_cfg.hidden_size // txt_cfg.num_attention_heads
head_dim = getattr(txt_cfg, "head_dim", txt_cfg.hidden_size // txt_cfg.num_attention_heads)
if hasattr(txt_cfg, "cross_attention_layers"):
cross_attention_layers = txt_cfg.cross_attention_layers
vis_cfg = self.config.vision_config
Expand All @@ -298,6 +298,7 @@ def prepare_vlm_ort_inputs(self):
if "attention_mask" in inputs.keys():
inputs["position_ids"] = inputs.pop("attention_mask").cumsum(1) - 1
inputs["past_key_values"] = []
inputs["image_idx"] = np.array([[0]])

vision_inputs = {
k: v for k, v in inputs.items() if k in {"pixel_values", "aspect_ratio_ids", "aspect_ratio_mask"}
Expand Down Expand Up @@ -349,6 +350,7 @@ def update_vlm_ort_outputs(self, ort_outputs):
outputs["image_features_RetainedState"] = (
ort_outputs["image_features_RetainedState"] if "image_features_RetainedState" in ort_outputs else None
)
outputs["image_idx"] = ort_outputs["image_idx_output"]
return outputs

def update_vlm_ort_inputs(self, inputs, ort_outputs):
Expand Down Expand Up @@ -414,7 +416,7 @@ def prepare_pytorch_inputs(self):

num_hidden_layers = txt_cfg.num_hidden_layers
num_key_value_heads = txt_cfg.num_key_value_heads
head_dim = txt_cfg.hidden_size // txt_cfg.num_attention_heads
head_dim = getattr(txt_cfg, "head_dim", txt_cfg.hidden_size // txt_cfg.num_attention_heads)

inputs["position_ids"] = inputs.pop("attention_mask").cumsum(1) - 1
inputs["past_key_values"] = []
Expand All @@ -435,7 +437,7 @@ def prepare_vlm_ort_inputs(self):
txt_cfg = self.config.llm_config
num_hidden_layers = txt_cfg.num_hidden_layers
num_key_value_heads = txt_cfg.num_key_value_heads
head_dim = txt_cfg.hidden_size // txt_cfg.num_attention_heads
head_dim = getattr(txt_cfg, "head_dim", txt_cfg.hidden_size // txt_cfg.num_attention_heads)

question = "<image>\n" + self.prompt
pixel_values = self.processor.load_image(self.image, max_num=12)
Expand All @@ -449,6 +451,7 @@ def prepare_vlm_ort_inputs(self):
if "attention_mask" in inputs.keys():
inputs["position_ids"] = inputs.pop("attention_mask").cumsum(1) - 1
inputs["past_key_values"] = []
inputs["image_idx"] = np.array([[0]])

vision_inputs = {
k: v for k, v in inputs.items() if k in {"pixel_values", "aspect_ratio_ids", "aspect_ratio_mask"}
Expand Down
32 changes: 25 additions & 7 deletions QEfficient/utils/run_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -129,7 +129,6 @@ def run_kv_model_on_pytorch(self, model):

generated_ids = []
inputs = self.input_handler.prepare_pytorch_inputs()

pt_outputs = model(**inputs)
for _ in range(1, self.gen_len):
generated_ids.append(pt_outputs["logits"].argmax(-1).reshape(-1, 1))
Expand Down Expand Up @@ -291,9 +290,11 @@ def run_vlm_kv_model_on_pytorch(self, model):
generation_len = self.gen_len
generated_ids = torch.full((self.batch_size, generation_len), self.processor.tokenizer.pad_token_id)
inputs = self.input_handler_vlm.prepare_pytorch_inputs()
inputs["image_idx"] = torch.tensor([[0]])

outputs = model(**inputs)
inputs["input_ids"] = outputs[0].argmax(2)
inputs["image_idx"] = outputs[2]
if "cross_attention_mask" in inputs:
bs, _, num_images, img_tiles = inputs["cross_attention_mask"].shape
inputs["cross_attention_mask"] = torch.ones((bs, 1, num_images, img_tiles), dtype=torch.int64)
Expand All @@ -308,6 +309,7 @@ def run_vlm_kv_model_on_pytorch(self, model):
for num_token in range(1, self.gen_len):
outputs = model(**inputs)
inputs["input_ids"] = outputs[0].argmax(2)
inputs["image_idx"] = outputs[2]
inputs["position_ids"] += 1
streamer.put(inputs["input_ids"])
generated_ids[:, num_token] = inputs["input_ids"].squeeze(1)
Expand Down Expand Up @@ -363,15 +365,23 @@ def run_vlm_kv_model_on_ort(self, model_path):

added_initializers, decoder_session = self.setup_ort_session(decoder_path)
generated_ids = []
finished_sequences = lang_inputs["input_ids"] == self.processor.tokenizer.eos_token_id

ort_outputs = self.run_ort_session(lang_inputs, session=decoder_session)
ort_outputs = self.input_handler_vlm.update_vlm_ort_outputs(ort_outputs)
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
lang_inputs = self.input_handler_vlm.update_vlm_ort_inputs(lang_inputs, ort_outputs)

for _ in range(1, self.gen_len):
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
lang_inputs = self.input_handler_vlm.update_vlm_ort_inputs(lang_inputs, ort_outputs)
finished_sequences |= lang_inputs["input_ids"] == self.processor.tokenizer.eos_token_id
if finished_sequences.all():
break

ort_outputs = self.run_ort_session(lang_inputs, decoder_session)
ort_outputs = self.input_handler_vlm.update_vlm_ort_outputs(ort_outputs)
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
lang_inputs = self.input_handler_vlm.update_vlm_ort_inputs(lang_inputs, ort_outputs)

generated_ids = np.concatenate(generated_ids, axis=1)
predicted_string = self.processor.tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
print("ORT KV_OFFLOAD Session Outputs:")
Expand All @@ -383,14 +393,22 @@ def run_vlm_kv_model_on_ort(self, model_path):
added_initializers, session = self.setup_ort_session(model_path)
generated_ids = []
inputs = {**vision_inputs, **lang_inputs}
finished_sequences = inputs["input_ids"] == self.processor.tokenizer.eos_token_id

ort_outputs = self.run_ort_session(inputs, session=session)
ort_outputs = self.input_handler_vlm.update_vlm_ort_outputs(ort_outputs)
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
inputs = self.input_handler_vlm.update_vlm_ort_inputs(inputs, ort_outputs)

for _ in range(1, self.gen_len):
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
inputs = self.input_handler_vlm.update_vlm_ort_inputs(inputs, ort_outputs)
finished_sequences |= inputs["input_ids"] == self.processor.tokenizer.eos_token_id
if finished_sequences.all():
break
ort_outputs = self.run_ort_session(inputs, session)
ort_outputs = self.input_handler_vlm.update_vlm_ort_outputs(ort_outputs)
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
generated_ids.append(ort_outputs["logits"].argmax(-1).reshape(-1, 1))
inputs = self.input_handler_vlm.update_vlm_ort_inputs(inputs, ort_outputs)

generated_ids = np.concatenate(generated_ids, axis=1)
predicted_string = self.processor.tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
print("ORT Session Outputs:")
Expand Down
Loading