Spaces:
Runtime error
Runtime error
File size: 7,413 Bytes
eb0678a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 |
# This file may have been modified by Flash-VStream Authors (Flash-VStream Modifications”). All Flash-VStream Modifications are Copyright 2024 Flash-VStream Authors.
# ------------------------------------------------------------------------
# Based on https://github.com/haotian-liu/LLaVA. Below is the original copyright:
# Copyright 2023 Haotian Liu
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import os
import warnings
import shutil
from transformers import AutoTokenizer, AutoModelForCausalLM, AutoConfig, BitsAndBytesConfig
import torch
from flash_vstream.model import VStreamLlamaForCausalLM, VStreamConfig
from flash_vstream.constants import DEFAULT_IMAGE_PATCH_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN
def load_pretrained_model(model_path, model_base, model_name, load_8bit=False, load_4bit=False, device_map="auto", device="cuda", **kwargs):
kwargs = {"device_map": device_map, **kwargs}
if device != "cuda":
kwargs['device_map'] = {"": device}
if load_8bit:
kwargs['load_in_8bit'] = True
elif load_4bit:
kwargs['load_in_4bit'] = True
kwargs['quantization_config'] = BitsAndBytesConfig(
load_in_4bit=True,
bnb_4bit_compute_dtype=torch.float16,
bnb_4bit_use_double_quant=True,
bnb_4bit_quant_type='nf4'
)
else:
kwargs['torch_dtype'] = torch.float16
if 'vstream' in model_name.lower():
# Load LLaMA-VStream model
if 'lora' in model_name.lower() and model_base is None:
warnings.warn('There is `lora` in model name but no `model_base` is provided. If you are loading a LoRA model, please provide the `model_base` argument. Detailed instruction: https://github.com/haotian-liu/LLaVA#launch-a-model-worker-lora-weights-unmerged.')
if 'lora' in model_name.lower() and model_base is not None:
lora_cfg_pretrained = AutoConfig.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_base, use_fast=False)
print('(LoRA) Loading LLaMA-VStream from base model...')
model = VStreamLlamaForCausalLM.from_pretrained(model_base, low_cpu_mem_usage=True, config=lora_cfg_pretrained, **kwargs)
token_num, tokem_dim = model.lm_head.out_features, model.lm_head.in_features
if model.lm_head.weight.shape[0] != token_num:
model.lm_head.weight = torch.nn.Parameter(torch.empty(token_num, tokem_dim, device=model.device, dtype=model.dtype))
model.model.embed_tokens.weight = torch.nn.Parameter(torch.empty(token_num, tokem_dim, device=model.device, dtype=model.dtype))
print('(LoRA) Loading additional LLaMA-VStream weights...')
if os.path.exists(os.path.join(model_path, 'non_lora_trainables.bin')):
non_lora_trainables = torch.load(os.path.join(model_path, 'non_lora_trainables.bin'), map_location='cpu')
else:
# this is probably from HF Hub
from huggingface_hub import hf_hub_download
def load_from_hf(repo_id, filename, subfolder=None):
cache_file = hf_hub_download(
repo_id=repo_id,
filename=filename,
subfolder=subfolder)
return torch.load(cache_file, map_location='cpu')
non_lora_trainables = load_from_hf(model_path, 'non_lora_trainables.bin')
non_lora_trainables = {(k[11:] if k.startswith('base_model.') else k): v for k, v in non_lora_trainables.items()}
if any(k.startswith('model.model.') for k in non_lora_trainables):
non_lora_trainables = {(k[6:] if k.startswith('model.') else k): v for k, v in non_lora_trainables.items()}
model.load_state_dict(non_lora_trainables, strict=False)
from peft import PeftModel
print('Loading LoRA weights...')
model = PeftModel.from_pretrained(model, model_path)
print('Merging LoRA weights...')
model = model.merge_and_unload()
print('Model is loaded...')
elif model_base is not None:
# this may be mm projector only
print('Loading LLaMA-VStream from base model...')
tokenizer = AutoTokenizer.from_pretrained(model_base, use_fast=False)
cfg_pretrained = AutoConfig.from_pretrained(model_path)
model = VStreamLlamaForCausalLM.from_pretrained(model_base, low_cpu_mem_usage=True, config=cfg_pretrained, **kwargs)
mm_projector_weights = torch.load(os.path.join(model_path, 'mm_projector.bin'), map_location='cpu')
mm_projector_weights = {k: v.to(torch.float16) for k, v in mm_projector_weights.items()}
model.load_state_dict(mm_projector_weights, strict=False)
else:
tokenizer = AutoTokenizer.from_pretrained(model_path, use_fast=False)
model = VStreamLlamaForCausalLM.from_pretrained(model_path, low_cpu_mem_usage=True, **kwargs)
else:
# Load language model
if model_base is not None:
# PEFT model
from peft import PeftModel
tokenizer = AutoTokenizer.from_pretrained(model_base, use_fast=False)
model = AutoModelForCausalLM.from_pretrained(model_base, low_cpu_mem_usage=True, **kwargs)
print(f"Loading LoRA weights from {model_path}")
model = PeftModel.from_pretrained(model, model_path)
print(f"Merging weights")
model = model.merge_and_unload()
print('Convert to FP16...')
model.to(torch.float16)
else:
use_fast = False
tokenizer = AutoTokenizer.from_pretrained(model_path, use_fast=False)
model = AutoModelForCausalLM.from_pretrained(model_path, low_cpu_mem_usage=True, **kwargs)
image_processor = None
if 'vstream' in model_name.lower():
mm_use_im_start_end = getattr(model.config, "mm_use_im_start_end", False)
mm_use_im_patch_token = getattr(model.config, "mm_use_im_patch_token", True)
if mm_use_im_patch_token:
tokenizer.add_tokens([DEFAULT_IMAGE_PATCH_TOKEN], special_tokens=True)
if mm_use_im_start_end:
tokenizer.add_tokens([DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN], special_tokens=True)
model.resize_token_embeddings(len(tokenizer))
vision_tower = model.get_vision_tower()
if not vision_tower.is_loaded:
vision_tower.load_model()
vision_tower.to(device=device, dtype=torch.float16)
image_processor = vision_tower.image_processor
if hasattr(model.config, "max_sequence_length"):
context_len = model.config.max_sequence_length
else:
context_len = 2048
return tokenizer, model, image_processor, context_len
|