fix compatibility issue for transformers 4.46+
Browse files
configuration_intern_vit.py
CHANGED
|
@@ -3,6 +3,7 @@
|
|
| 3 |
# Copyright (c) 2024 OpenGVLab
|
| 4 |
# Licensed under The MIT License [see LICENSE for details]
|
| 5 |
# --------------------------------------------------------
|
|
|
|
| 6 |
import os
|
| 7 |
from typing import Union
|
| 8 |
|
|
|
|
| 3 |
# Copyright (c) 2024 OpenGVLab
|
| 4 |
# Licensed under The MIT License [see LICENSE for details]
|
| 5 |
# --------------------------------------------------------
|
| 6 |
+
|
| 7 |
import os
|
| 8 |
from typing import Union
|
| 9 |
|
configuration_internvl_chat.py
CHANGED
|
@@ -47,12 +47,12 @@ class InternVLChatConfig(PretrainedConfig):
|
|
| 47 |
logger.info('llm_config is None. Initializing the LlamaConfig config with default values (`LlamaConfig`).')
|
| 48 |
|
| 49 |
self.vision_config = InternVisionConfig(**vision_config)
|
| 50 |
-
if llm_config['architectures'][0] == 'LlamaForCausalLM':
|
| 51 |
self.llm_config = LlamaConfig(**llm_config)
|
| 52 |
-
elif llm_config['architectures'][0] == 'InternLM2ForCausalLM':
|
| 53 |
self.llm_config = InternLM2Config(**llm_config)
|
| 54 |
else:
|
| 55 |
-
raise ValueError('Unsupported architecture: {}'.format(llm_config['architectures'][0]))
|
| 56 |
self.use_backbone_lora = use_backbone_lora
|
| 57 |
self.use_llm_lora = use_llm_lora
|
| 58 |
self.select_layer = select_layer
|
|
|
|
| 47 |
logger.info('llm_config is None. Initializing the LlamaConfig config with default values (`LlamaConfig`).')
|
| 48 |
|
| 49 |
self.vision_config = InternVisionConfig(**vision_config)
|
| 50 |
+
if llm_config.get(['architectures'])[0] == 'LlamaForCausalLM':
|
| 51 |
self.llm_config = LlamaConfig(**llm_config)
|
| 52 |
+
elif llm_config.get(['architectures'])[0] == 'InternLM2ForCausalLM':
|
| 53 |
self.llm_config = InternLM2Config(**llm_config)
|
| 54 |
else:
|
| 55 |
+
raise ValueError('Unsupported architecture: {}'.format(llm_config.get(['architectures'])[0]))
|
| 56 |
self.use_backbone_lora = use_backbone_lora
|
| 57 |
self.use_llm_lora = use_llm_lora
|
| 58 |
self.select_layer = select_layer
|
modeling_internvl_chat.py
CHANGED
|
@@ -3,6 +3,7 @@
|
|
| 3 |
# Copyright (c) 2024 OpenGVLab
|
| 4 |
# Licensed under The MIT License [see LICENSE for details]
|
| 5 |
# --------------------------------------------------------
|
|
|
|
| 6 |
import warnings
|
| 7 |
from typing import Any, List, Optional, Tuple, Union
|
| 8 |
|
|
@@ -236,7 +237,7 @@ class InternVLChatModel(PreTrainedModel):
|
|
| 236 |
model_inputs = tokenizer(queries, return_tensors='pt', padding=True)
|
| 237 |
input_ids = model_inputs['input_ids'].to(self.device)
|
| 238 |
attention_mask = model_inputs['attention_mask'].to(self.device)
|
| 239 |
-
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep)
|
| 240 |
generation_config['eos_token_id'] = eos_token_id
|
| 241 |
generation_output = self.generate(
|
| 242 |
pixel_values=pixel_values,
|
|
@@ -245,7 +246,7 @@ class InternVLChatModel(PreTrainedModel):
|
|
| 245 |
**generation_config
|
| 246 |
)
|
| 247 |
responses = tokenizer.batch_decode(generation_output, skip_special_tokens=True)
|
| 248 |
-
responses = [response.split(template.sep)[0].strip() for response in responses]
|
| 249 |
return responses
|
| 250 |
|
| 251 |
def chat(self, tokenizer, pixel_values, question, generation_config, history=None, return_history=False,
|
|
@@ -264,7 +265,7 @@ class InternVLChatModel(PreTrainedModel):
|
|
| 264 |
|
| 265 |
template = get_conv_template(self.template)
|
| 266 |
template.system_message = self.system_message
|
| 267 |
-
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep)
|
| 268 |
|
| 269 |
history = [] if history is None else history
|
| 270 |
for (old_question, old_answer) in history:
|
|
@@ -293,7 +294,7 @@ class InternVLChatModel(PreTrainedModel):
|
|
| 293 |
**generation_config
|
| 294 |
)
|
| 295 |
response = tokenizer.batch_decode(generation_output, skip_special_tokens=True)[0]
|
| 296 |
-
response = response.split(template.sep)[0].strip()
|
| 297 |
history.append((question, response))
|
| 298 |
if return_history:
|
| 299 |
return response, history
|
|
@@ -313,7 +314,6 @@ class InternVLChatModel(PreTrainedModel):
|
|
| 313 |
visual_features: Optional[torch.FloatTensor] = None,
|
| 314 |
generation_config: Optional[GenerationConfig] = None,
|
| 315 |
output_hidden_states: Optional[bool] = None,
|
| 316 |
-
return_dict: Optional[bool] = None,
|
| 317 |
**generate_kwargs,
|
| 318 |
) -> torch.LongTensor:
|
| 319 |
|
|
@@ -341,7 +341,6 @@ class InternVLChatModel(PreTrainedModel):
|
|
| 341 |
attention_mask=attention_mask,
|
| 342 |
generation_config=generation_config,
|
| 343 |
output_hidden_states=output_hidden_states,
|
| 344 |
-
return_dict=return_dict,
|
| 345 |
use_cache=True,
|
| 346 |
**generate_kwargs,
|
| 347 |
)
|
|
|
|
| 3 |
# Copyright (c) 2024 OpenGVLab
|
| 4 |
# Licensed under The MIT License [see LICENSE for details]
|
| 5 |
# --------------------------------------------------------
|
| 6 |
+
|
| 7 |
import warnings
|
| 8 |
from typing import Any, List, Optional, Tuple, Union
|
| 9 |
|
|
|
|
| 237 |
model_inputs = tokenizer(queries, return_tensors='pt', padding=True)
|
| 238 |
input_ids = model_inputs['input_ids'].to(self.device)
|
| 239 |
attention_mask = model_inputs['attention_mask'].to(self.device)
|
| 240 |
+
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep.strip())
|
| 241 |
generation_config['eos_token_id'] = eos_token_id
|
| 242 |
generation_output = self.generate(
|
| 243 |
pixel_values=pixel_values,
|
|
|
|
| 246 |
**generation_config
|
| 247 |
)
|
| 248 |
responses = tokenizer.batch_decode(generation_output, skip_special_tokens=True)
|
| 249 |
+
responses = [response.split(template.sep.strip())[0].strip() for response in responses]
|
| 250 |
return responses
|
| 251 |
|
| 252 |
def chat(self, tokenizer, pixel_values, question, generation_config, history=None, return_history=False,
|
|
|
|
| 265 |
|
| 266 |
template = get_conv_template(self.template)
|
| 267 |
template.system_message = self.system_message
|
| 268 |
+
eos_token_id = tokenizer.convert_tokens_to_ids(template.sep.strip())
|
| 269 |
|
| 270 |
history = [] if history is None else history
|
| 271 |
for (old_question, old_answer) in history:
|
|
|
|
| 294 |
**generation_config
|
| 295 |
)
|
| 296 |
response = tokenizer.batch_decode(generation_output, skip_special_tokens=True)[0]
|
| 297 |
+
response = response.split(template.sep.strip())[0].strip()
|
| 298 |
history.append((question, response))
|
| 299 |
if return_history:
|
| 300 |
return response, history
|
|
|
|
| 314 |
visual_features: Optional[torch.FloatTensor] = None,
|
| 315 |
generation_config: Optional[GenerationConfig] = None,
|
| 316 |
output_hidden_states: Optional[bool] = None,
|
|
|
|
| 317 |
**generate_kwargs,
|
| 318 |
) -> torch.LongTensor:
|
| 319 |
|
|
|
|
| 341 |
attention_mask=attention_mask,
|
| 342 |
generation_config=generation_config,
|
| 343 |
output_hidden_states=output_hidden_states,
|
|
|
|
| 344 |
use_cache=True,
|
| 345 |
**generate_kwargs,
|
| 346 |
)
|