{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/dscilab_dungvo/workspace/vlm_clone/VLMEvalKit_old\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "%cd ../VLMEvalKit_old\n",
    "from vlmeval.config import supported_VLM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "InternLM2ForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly overwritten. However, it doesn't directly inherit from `GenerationMixin`. From 👉v4.50👈 onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.\n",
      "  - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes\n",
      "  - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).\n",
      "  - If you are not the owner of the model architecture class, please contact the model code owner to update it.\n",
      "Loading checkpoint shards: 100%|██████████████| 13/13 [00:08<00:00,  1.52it/s]\n",
      "Some weights of the model checkpoint at OpenGVLab/InternVL2_5-26B-AWQ were not used when initializing InternVLChatModel: ['language_model.model.layers.0.attention.wo.qweight', 'language_model.model.layers.0.attention.wo.qzeros', 'language_model.model.layers.0.attention.wo.scales', 'language_model.model.layers.0.attention.wqkv.qweight', 'language_model.model.layers.0.attention.wqkv.qzeros', 'language_model.model.layers.0.attention.wqkv.scales', 'language_model.model.layers.0.feed_forward.w1.qweight', 'language_model.model.layers.0.feed_forward.w1.qzeros', 'language_model.model.layers.0.feed_forward.w1.scales', 'language_model.model.layers.0.feed_forward.w2.qweight', 'language_model.model.layers.0.feed_forward.w2.qzeros', 'language_model.model.layers.0.feed_forward.w2.scales', 'language_model.model.layers.0.feed_forward.w3.qweight', 'language_model.model.layers.0.feed_forward.w3.qzeros', 'language_model.model.layers.0.feed_forward.w3.scales', 'language_model.model.layers.1.attention.wo.qweight', 'language_model.model.layers.1.attention.wo.qzeros', 'language_model.model.layers.1.attention.wo.scales', 'language_model.model.layers.1.attention.wqkv.qweight', 'language_model.model.layers.1.attention.wqkv.qzeros', 'language_model.model.layers.1.attention.wqkv.scales', 'language_model.model.layers.1.feed_forward.w1.qweight', 'language_model.model.layers.1.feed_forward.w1.qzeros', 'language_model.model.layers.1.feed_forward.w1.scales', 'language_model.model.layers.1.feed_forward.w2.qweight', 'language_model.model.layers.1.feed_forward.w2.qzeros', 'language_model.model.layers.1.feed_forward.w2.scales', 'language_model.model.layers.1.feed_forward.w3.qweight', 'language_model.model.layers.1.feed_forward.w3.qzeros', 'language_model.model.layers.1.feed_forward.w3.scales', 'language_model.model.layers.10.attention.wo.qweight', 'language_model.model.layers.10.attention.wo.qzeros', 'language_model.model.layers.10.attention.wo.scales', 'language_model.model.layers.10.attention.wqkv.qweight', 'language_model.model.layers.10.attention.wqkv.qzeros', 'language_model.model.layers.10.attention.wqkv.scales', 'language_model.model.layers.10.feed_forward.w1.qweight', 'language_model.model.layers.10.feed_forward.w1.qzeros', 'language_model.model.layers.10.feed_forward.w1.scales', 'language_model.model.layers.10.feed_forward.w2.qweight', 'language_model.model.layers.10.feed_forward.w2.qzeros', 'language_model.model.layers.10.feed_forward.w2.scales', 'language_model.model.layers.10.feed_forward.w3.qweight', 'language_model.model.layers.10.feed_forward.w3.qzeros', 'language_model.model.layers.10.feed_forward.w3.scales', 'language_model.model.layers.11.attention.wo.qweight', 'language_model.model.layers.11.attention.wo.qzeros', 'language_model.model.layers.11.attention.wo.scales', 'language_model.model.layers.11.attention.wqkv.qweight', 'language_model.model.layers.11.attention.wqkv.qzeros', 'language_model.model.layers.11.attention.wqkv.scales', 'language_model.model.layers.11.feed_forward.w1.qweight', 'language_model.model.layers.11.feed_forward.w1.qzeros', 'language_model.model.layers.11.feed_forward.w1.scales', 'language_model.model.layers.11.feed_forward.w2.qweight', 'language_model.model.layers.11.feed_forward.w2.qzeros', 'language_model.model.layers.11.feed_forward.w2.scales', 'language_model.model.layers.11.feed_forward.w3.qweight', 'language_model.model.layers.11.feed_forward.w3.qzeros', 'language_model.model.layers.11.feed_forward.w3.scales', 'language_model.model.layers.12.attention.wo.qweight', 'language_model.model.layers.12.attention.wo.qzeros', 'language_model.model.layers.12.attention.wo.scales', 'language_model.model.layers.12.attention.wqkv.qweight', 'language_model.model.layers.12.attention.wqkv.qzeros', 'language_model.model.layers.12.attention.wqkv.scales', 'language_model.model.layers.12.feed_forward.w1.qweight', 'language_model.model.layers.12.feed_forward.w1.qzeros', 'language_model.model.layers.12.feed_forward.w1.scales', 'language_model.model.layers.12.feed_forward.w2.qweight', 'language_model.model.layers.12.feed_forward.w2.qzeros', 'language_model.model.layers.12.feed_forward.w2.scales', 'language_model.model.layers.12.feed_forward.w3.qweight', 'language_model.model.layers.12.feed_forward.w3.qzeros', 'language_model.model.layers.12.feed_forward.w3.scales', 'language_model.model.layers.13.attention.wo.qweight', 'language_model.model.layers.13.attention.wo.qzeros', 'language_model.model.layers.13.attention.wo.scales', 'language_model.model.layers.13.attention.wqkv.qweight', 'language_model.model.layers.13.attention.wqkv.qzeros', 'language_model.model.layers.13.attention.wqkv.scales', 'language_model.model.layers.13.feed_forward.w1.qweight', 'language_model.model.layers.13.feed_forward.w1.qzeros', 'language_model.model.layers.13.feed_forward.w1.scales', 'language_model.model.layers.13.feed_forward.w2.qweight', 'language_model.model.layers.13.feed_forward.w2.qzeros', 'language_model.model.layers.13.feed_forward.w2.scales', 'language_model.model.layers.13.feed_forward.w3.qweight', 'language_model.model.layers.13.feed_forward.w3.qzeros', 'language_model.model.layers.13.feed_forward.w3.scales', 'language_model.model.layers.14.attention.wo.qweight', 'language_model.model.layers.14.attention.wo.qzeros', 'language_model.model.layers.14.attention.wo.scales', 'language_model.model.layers.14.attention.wqkv.qweight', 'language_model.model.layers.14.attention.wqkv.qzeros', 'language_model.model.layers.14.attention.wqkv.scales', 'language_model.model.layers.14.feed_forward.w1.qweight', 'language_model.model.layers.14.feed_forward.w1.qzeros', 'language_model.model.layers.14.feed_forward.w1.scales', 'language_model.model.layers.14.feed_forward.w2.qweight', 'language_model.model.layers.14.feed_forward.w2.qzeros', 'language_model.model.layers.14.feed_forward.w2.scales', 'language_model.model.layers.14.feed_forward.w3.qweight', 'language_model.model.layers.14.feed_forward.w3.qzeros', 'language_model.model.layers.14.feed_forward.w3.scales', 'language_model.model.layers.15.attention.wo.qweight', 'language_model.model.layers.15.attention.wo.qzeros', 'language_model.model.layers.15.attention.wo.scales', 'language_model.model.layers.15.attention.wqkv.qweight', 'language_model.model.layers.15.attention.wqkv.qzeros', 'language_model.model.layers.15.attention.wqkv.scales', 'language_model.model.layers.15.feed_forward.w1.qweight', 'language_model.model.layers.15.feed_forward.w1.qzeros', 'language_model.model.layers.15.feed_forward.w1.scales', 'language_model.model.layers.15.feed_forward.w2.qweight', 'language_model.model.layers.15.feed_forward.w2.qzeros', 'language_model.model.layers.15.feed_forward.w2.scales', 'language_model.model.layers.15.feed_forward.w3.qweight', 'language_model.model.layers.15.feed_forward.w3.qzeros', 'language_model.model.layers.15.feed_forward.w3.scales', 'language_model.model.layers.16.attention.wo.qweight', 'language_model.model.layers.16.attention.wo.qzeros', 'language_model.model.layers.16.attention.wo.scales', 'language_model.model.layers.16.attention.wqkv.qweight', 'language_model.model.layers.16.attention.wqkv.qzeros', 'language_model.model.layers.16.attention.wqkv.scales', 'language_model.model.layers.16.feed_forward.w1.qweight', 'language_model.model.layers.16.feed_forward.w1.qzeros', 'language_model.model.layers.16.feed_forward.w1.scales', 'language_model.model.layers.16.feed_forward.w2.qweight', 'language_model.model.layers.16.feed_forward.w2.qzeros', 'language_model.model.layers.16.feed_forward.w2.scales', 'language_model.model.layers.16.feed_forward.w3.qweight', 'language_model.model.layers.16.feed_forward.w3.qzeros', 'language_model.model.layers.16.feed_forward.w3.scales', 'language_model.model.layers.17.attention.wo.qweight', 'language_model.model.layers.17.attention.wo.qzeros', 'language_model.model.layers.17.attention.wo.scales', 'language_model.model.layers.17.attention.wqkv.qweight', 'language_model.model.layers.17.attention.wqkv.qzeros', 'language_model.model.layers.17.attention.wqkv.scales', 'language_model.model.layers.17.feed_forward.w1.qweight', 'language_model.model.layers.17.feed_forward.w1.qzeros', 'language_model.model.layers.17.feed_forward.w1.scales', 'language_model.model.layers.17.feed_forward.w2.qweight', 'language_model.model.layers.17.feed_forward.w2.qzeros', 'language_model.model.layers.17.feed_forward.w2.scales', 'language_model.model.layers.17.feed_forward.w3.qweight', 'language_model.model.layers.17.feed_forward.w3.qzeros', 'language_model.model.layers.17.feed_forward.w3.scales', 'language_model.model.layers.18.attention.wo.qweight', 'language_model.model.layers.18.attention.wo.qzeros', 'language_model.model.layers.18.attention.wo.scales', 'language_model.model.layers.18.attention.wqkv.qweight', 'language_model.model.layers.18.attention.wqkv.qzeros', 'language_model.model.layers.18.attention.wqkv.scales', 'language_model.model.layers.18.feed_forward.w1.qweight', 'language_model.model.layers.18.feed_forward.w1.qzeros', 'language_model.model.layers.18.feed_forward.w1.scales', 'language_model.model.layers.18.feed_forward.w2.qweight', 'language_model.model.layers.18.feed_forward.w2.qzeros', 'language_model.model.layers.18.feed_forward.w2.scales', 'language_model.model.layers.18.feed_forward.w3.qweight', 'language_model.model.layers.18.feed_forward.w3.qzeros', 'language_model.model.layers.18.feed_forward.w3.scales', 'language_model.model.layers.19.attention.wo.qweight', 'language_model.model.layers.19.attention.wo.qzeros', 'language_model.model.layers.19.attention.wo.scales', 'language_model.model.layers.19.attention.wqkv.qweight', 'language_model.model.layers.19.attention.wqkv.qzeros', 'language_model.model.layers.19.attention.wqkv.scales', 'language_model.model.layers.19.feed_forward.w1.qweight', 'language_model.model.layers.19.feed_forward.w1.qzeros', 'language_model.model.layers.19.feed_forward.w1.scales', 'language_model.model.layers.19.feed_forward.w2.qweight', 'language_model.model.layers.19.feed_forward.w2.qzeros', 'language_model.model.layers.19.feed_forward.w2.scales', 'language_model.model.layers.19.feed_forward.w3.qweight', 'language_model.model.layers.19.feed_forward.w3.qzeros', 'language_model.model.layers.19.feed_forward.w3.scales', 'language_model.model.layers.2.attention.wo.qweight', 'language_model.model.layers.2.attention.wo.qzeros', 'language_model.model.layers.2.attention.wo.scales', 'language_model.model.layers.2.attention.wqkv.qweight', 'language_model.model.layers.2.attention.wqkv.qzeros', 'language_model.model.layers.2.attention.wqkv.scales', 'language_model.model.layers.2.feed_forward.w1.qweight', 'language_model.model.layers.2.feed_forward.w1.qzeros', 'language_model.model.layers.2.feed_forward.w1.scales', 'language_model.model.layers.2.feed_forward.w2.qweight', 'language_model.model.layers.2.feed_forward.w2.qzeros', 'language_model.model.layers.2.feed_forward.w2.scales', 'language_model.model.layers.2.feed_forward.w3.qweight', 'language_model.model.layers.2.feed_forward.w3.qzeros', 'language_model.model.layers.2.feed_forward.w3.scales', 'language_model.model.layers.20.attention.wo.qweight', 'language_model.model.layers.20.attention.wo.qzeros', 'language_model.model.layers.20.attention.wo.scales', 'language_model.model.layers.20.attention.wqkv.qweight', 'language_model.model.layers.20.attention.wqkv.qzeros', 'language_model.model.layers.20.attention.wqkv.scales', 'language_model.model.layers.20.feed_forward.w1.qweight', 'language_model.model.layers.20.feed_forward.w1.qzeros', 'language_model.model.layers.20.feed_forward.w1.scales', 'language_model.model.layers.20.feed_forward.w2.qweight', 'language_model.model.layers.20.feed_forward.w2.qzeros', 'language_model.model.layers.20.feed_forward.w2.scales', 'language_model.model.layers.20.feed_forward.w3.qweight', 'language_model.model.layers.20.feed_forward.w3.qzeros', 'language_model.model.layers.20.feed_forward.w3.scales', 'language_model.model.layers.21.attention.wo.qweight', 'language_model.model.layers.21.attention.wo.qzeros', 'language_model.model.layers.21.attention.wo.scales', 'language_model.model.layers.21.attention.wqkv.qweight', 'language_model.model.layers.21.attention.wqkv.qzeros', 'language_model.model.layers.21.attention.wqkv.scales', 'language_model.model.layers.21.feed_forward.w1.qweight', 'language_model.model.layers.21.feed_forward.w1.qzeros', 'language_model.model.layers.21.feed_forward.w1.scales', 'language_model.model.layers.21.feed_forward.w2.qweight', 'language_model.model.layers.21.feed_forward.w2.qzeros', 'language_model.model.layers.21.feed_forward.w2.scales', 'language_model.model.layers.21.feed_forward.w3.qweight', 'language_model.model.layers.21.feed_forward.w3.qzeros', 'language_model.model.layers.21.feed_forward.w3.scales', 'language_model.model.layers.22.attention.wo.qweight', 'language_model.model.layers.22.attention.wo.qzeros', 'language_model.model.layers.22.attention.wo.scales', 'language_model.model.layers.22.attention.wqkv.qweight', 'language_model.model.layers.22.attention.wqkv.qzeros', 'language_model.model.layers.22.attention.wqkv.scales', 'language_model.model.layers.22.feed_forward.w1.qweight', 'language_model.model.layers.22.feed_forward.w1.qzeros', 'language_model.model.layers.22.feed_forward.w1.scales', 'language_model.model.layers.22.feed_forward.w2.qweight', 'language_model.model.layers.22.feed_forward.w2.qzeros', 'language_model.model.layers.22.feed_forward.w2.scales', 'language_model.model.layers.22.feed_forward.w3.qweight', 'language_model.model.layers.22.feed_forward.w3.qzeros', 'language_model.model.layers.22.feed_forward.w3.scales', 'language_model.model.layers.23.attention.wo.qweight', 'language_model.model.layers.23.attention.wo.qzeros', 'language_model.model.layers.23.attention.wo.scales', 'language_model.model.layers.23.attention.wqkv.qweight', 'language_model.model.layers.23.attention.wqkv.qzeros', 'language_model.model.layers.23.attention.wqkv.scales', 'language_model.model.layers.23.feed_forward.w1.qweight', 'language_model.model.layers.23.feed_forward.w1.qzeros', 'language_model.model.layers.23.feed_forward.w1.scales', 'language_model.model.layers.23.feed_forward.w2.qweight', 'language_model.model.layers.23.feed_forward.w2.qzeros', 'language_model.model.layers.23.feed_forward.w2.scales', 'language_model.model.layers.23.feed_forward.w3.qweight', 'language_model.model.layers.23.feed_forward.w3.qzeros', 'language_model.model.layers.23.feed_forward.w3.scales', 'language_model.model.layers.24.attention.wo.qweight', 'language_model.model.layers.24.attention.wo.qzeros', 'language_model.model.layers.24.attention.wo.scales', 'language_model.model.layers.24.attention.wqkv.qweight', 'language_model.model.layers.24.attention.wqkv.qzeros', 'language_model.model.layers.24.attention.wqkv.scales', 'language_model.model.layers.24.feed_forward.w1.qweight', 'language_model.model.layers.24.feed_forward.w1.qzeros', 'language_model.model.layers.24.feed_forward.w1.scales', 'language_model.model.layers.24.feed_forward.w2.qweight', 'language_model.model.layers.24.feed_forward.w2.qzeros', 'language_model.model.layers.24.feed_forward.w2.scales', 'language_model.model.layers.24.feed_forward.w3.qweight', 'language_model.model.layers.24.feed_forward.w3.qzeros', 'language_model.model.layers.24.feed_forward.w3.scales', 'language_model.model.layers.25.attention.wo.qweight', 'language_model.model.layers.25.attention.wo.qzeros', 'language_model.model.layers.25.attention.wo.scales', 'language_model.model.layers.25.attention.wqkv.qweight', 'language_model.model.layers.25.attention.wqkv.qzeros', 'language_model.model.layers.25.attention.wqkv.scales', 'language_model.model.layers.25.feed_forward.w1.qweight', 'language_model.model.layers.25.feed_forward.w1.qzeros', 'language_model.model.layers.25.feed_forward.w1.scales', 'language_model.model.layers.25.feed_forward.w2.qweight', 'language_model.model.layers.25.feed_forward.w2.qzeros', 'language_model.model.layers.25.feed_forward.w2.scales', 'language_model.model.layers.25.feed_forward.w3.qweight', 'language_model.model.layers.25.feed_forward.w3.qzeros', 'language_model.model.layers.25.feed_forward.w3.scales', 'language_model.model.layers.26.attention.wo.qweight', 'language_model.model.layers.26.attention.wo.qzeros', 'language_model.model.layers.26.attention.wo.scales', 'language_model.model.layers.26.attention.wqkv.qweight', 'language_model.model.layers.26.attention.wqkv.qzeros', 'language_model.model.layers.26.attention.wqkv.scales', 'language_model.model.layers.26.feed_forward.w1.qweight', 'language_model.model.layers.26.feed_forward.w1.qzeros', 'language_model.model.layers.26.feed_forward.w1.scales', 'language_model.model.layers.26.feed_forward.w2.qweight', 'language_model.model.layers.26.feed_forward.w2.qzeros', 'language_model.model.layers.26.feed_forward.w2.scales', 'language_model.model.layers.26.feed_forward.w3.qweight', 'language_model.model.layers.26.feed_forward.w3.qzeros', 'language_model.model.layers.26.feed_forward.w3.scales', 'language_model.model.layers.27.attention.wo.qweight', 'language_model.model.layers.27.attention.wo.qzeros', 'language_model.model.layers.27.attention.wo.scales', 'language_model.model.layers.27.attention.wqkv.qweight', 'language_model.model.layers.27.attention.wqkv.qzeros', 'language_model.model.layers.27.attention.wqkv.scales', 'language_model.model.layers.27.feed_forward.w1.qweight', 'language_model.model.layers.27.feed_forward.w1.qzeros', 'language_model.model.layers.27.feed_forward.w1.scales', 'language_model.model.layers.27.feed_forward.w2.qweight', 'language_model.model.layers.27.feed_forward.w2.qzeros', 'language_model.model.layers.27.feed_forward.w2.scales', 'language_model.model.layers.27.feed_forward.w3.qweight', 'language_model.model.layers.27.feed_forward.w3.qzeros', 'language_model.model.layers.27.feed_forward.w3.scales', 'language_model.model.layers.28.attention.wo.qweight', 'language_model.model.layers.28.attention.wo.qzeros', 'language_model.model.layers.28.attention.wo.scales', 'language_model.model.layers.28.attention.wqkv.qweight', 'language_model.model.layers.28.attention.wqkv.qzeros', 'language_model.model.layers.28.attention.wqkv.scales', 'language_model.model.layers.28.feed_forward.w1.qweight', 'language_model.model.layers.28.feed_forward.w1.qzeros', 'language_model.model.layers.28.feed_forward.w1.scales', 'language_model.model.layers.28.feed_forward.w2.qweight', 'language_model.model.layers.28.feed_forward.w2.qzeros', 'language_model.model.layers.28.feed_forward.w2.scales', 'language_model.model.layers.28.feed_forward.w3.qweight', 'language_model.model.layers.28.feed_forward.w3.qzeros', 'language_model.model.layers.28.feed_forward.w3.scales', 'language_model.model.layers.29.attention.wo.qweight', 'language_model.model.layers.29.attention.wo.qzeros', 'language_model.model.layers.29.attention.wo.scales', 'language_model.model.layers.29.attention.wqkv.qweight', 'language_model.model.layers.29.attention.wqkv.qzeros', 'language_model.model.layers.29.attention.wqkv.scales', 'language_model.model.layers.29.feed_forward.w1.qweight', 'language_model.model.layers.29.feed_forward.w1.qzeros', 'language_model.model.layers.29.feed_forward.w1.scales', 'language_model.model.layers.29.feed_forward.w2.qweight', 'language_model.model.layers.29.feed_forward.w2.qzeros', 'language_model.model.layers.29.feed_forward.w2.scales', 'language_model.model.layers.29.feed_forward.w3.qweight', 'language_model.model.layers.29.feed_forward.w3.qzeros', 'language_model.model.layers.29.feed_forward.w3.scales', 'language_model.model.layers.3.attention.wo.qweight', 'language_model.model.layers.3.attention.wo.qzeros', 'language_model.model.layers.3.attention.wo.scales', 'language_model.model.layers.3.attention.wqkv.qweight', 'language_model.model.layers.3.attention.wqkv.qzeros', 'language_model.model.layers.3.attention.wqkv.scales', 'language_model.model.layers.3.feed_forward.w1.qweight', 'language_model.model.layers.3.feed_forward.w1.qzeros', 'language_model.model.layers.3.feed_forward.w1.scales', 'language_model.model.layers.3.feed_forward.w2.qweight', 'language_model.model.layers.3.feed_forward.w2.qzeros', 'language_model.model.layers.3.feed_forward.w2.scales', 'language_model.model.layers.3.feed_forward.w3.qweight', 'language_model.model.layers.3.feed_forward.w3.qzeros', 'language_model.model.layers.3.feed_forward.w3.scales', 'language_model.model.layers.30.attention.wo.qweight', 'language_model.model.layers.30.attention.wo.qzeros', 'language_model.model.layers.30.attention.wo.scales', 'language_model.model.layers.30.attention.wqkv.qweight', 'language_model.model.layers.30.attention.wqkv.qzeros', 'language_model.model.layers.30.attention.wqkv.scales', 'language_model.model.layers.30.feed_forward.w1.qweight', 'language_model.model.layers.30.feed_forward.w1.qzeros', 'language_model.model.layers.30.feed_forward.w1.scales', 'language_model.model.layers.30.feed_forward.w2.qweight', 'language_model.model.layers.30.feed_forward.w2.qzeros', 'language_model.model.layers.30.feed_forward.w2.scales', 'language_model.model.layers.30.feed_forward.w3.qweight', 'language_model.model.layers.30.feed_forward.w3.qzeros', 'language_model.model.layers.30.feed_forward.w3.scales', 'language_model.model.layers.31.attention.wo.qweight', 'language_model.model.layers.31.attention.wo.qzeros', 'language_model.model.layers.31.attention.wo.scales', 'language_model.model.layers.31.attention.wqkv.qweight', 'language_model.model.layers.31.attention.wqkv.qzeros', 'language_model.model.layers.31.attention.wqkv.scales', 'language_model.model.layers.31.feed_forward.w1.qweight', 'language_model.model.layers.31.feed_forward.w1.qzeros', 'language_model.model.layers.31.feed_forward.w1.scales', 'language_model.model.layers.31.feed_forward.w2.qweight', 'language_model.model.layers.31.feed_forward.w2.qzeros', 'language_model.model.layers.31.feed_forward.w2.scales', 'language_model.model.layers.31.feed_forward.w3.qweight', 'language_model.model.layers.31.feed_forward.w3.qzeros', 'language_model.model.layers.31.feed_forward.w3.scales', 'language_model.model.layers.32.attention.wo.qweight', 'language_model.model.layers.32.attention.wo.qzeros', 'language_model.model.layers.32.attention.wo.scales', 'language_model.model.layers.32.attention.wqkv.qweight', 'language_model.model.layers.32.attention.wqkv.qzeros', 'language_model.model.layers.32.attention.wqkv.scales', 'language_model.model.layers.32.feed_forward.w1.qweight', 'language_model.model.layers.32.feed_forward.w1.qzeros', 'language_model.model.layers.32.feed_forward.w1.scales', 'language_model.model.layers.32.feed_forward.w2.qweight', 'language_model.model.layers.32.feed_forward.w2.qzeros', 'language_model.model.layers.32.feed_forward.w2.scales', 'language_model.model.layers.32.feed_forward.w3.qweight', 'language_model.model.layers.32.feed_forward.w3.qzeros', 'language_model.model.layers.32.feed_forward.w3.scales', 'language_model.model.layers.33.attention.wo.qweight', 'language_model.model.layers.33.attention.wo.qzeros', 'language_model.model.layers.33.attention.wo.scales', 'language_model.model.layers.33.attention.wqkv.qweight', 'language_model.model.layers.33.attention.wqkv.qzeros', 'language_model.model.layers.33.attention.wqkv.scales', 'language_model.model.layers.33.feed_forward.w1.qweight', 'language_model.model.layers.33.feed_forward.w1.qzeros', 'language_model.model.layers.33.feed_forward.w1.scales', 'language_model.model.layers.33.feed_forward.w2.qweight', 'language_model.model.layers.33.feed_forward.w2.qzeros', 'language_model.model.layers.33.feed_forward.w2.scales', 'language_model.model.layers.33.feed_forward.w3.qweight', 'language_model.model.layers.33.feed_forward.w3.qzeros', 'language_model.model.layers.33.feed_forward.w3.scales', 'language_model.model.layers.34.attention.wo.qweight', 'language_model.model.layers.34.attention.wo.qzeros', 'language_model.model.layers.34.attention.wo.scales', 'language_model.model.layers.34.attention.wqkv.qweight', 'language_model.model.layers.34.attention.wqkv.qzeros', 'language_model.model.layers.34.attention.wqkv.scales', 'language_model.model.layers.34.feed_forward.w1.qweight', 'language_model.model.layers.34.feed_forward.w1.qzeros', 'language_model.model.layers.34.feed_forward.w1.scales', 'language_model.model.layers.34.feed_forward.w2.qweight', 'language_model.model.layers.34.feed_forward.w2.qzeros', 'language_model.model.layers.34.feed_forward.w2.scales', 'language_model.model.layers.34.feed_forward.w3.qweight', 'language_model.model.layers.34.feed_forward.w3.qzeros', 'language_model.model.layers.34.feed_forward.w3.scales', 'language_model.model.layers.35.attention.wo.qweight', 'language_model.model.layers.35.attention.wo.qzeros', 'language_model.model.layers.35.attention.wo.scales', 'language_model.model.layers.35.attention.wqkv.qweight', 'language_model.model.layers.35.attention.wqkv.qzeros', 'language_model.model.layers.35.attention.wqkv.scales', 'language_model.model.layers.35.feed_forward.w1.qweight', 'language_model.model.layers.35.feed_forward.w1.qzeros', 'language_model.model.layers.35.feed_forward.w1.scales', 'language_model.model.layers.35.feed_forward.w2.qweight', 'language_model.model.layers.35.feed_forward.w2.qzeros', 'language_model.model.layers.35.feed_forward.w2.scales', 'language_model.model.layers.35.feed_forward.w3.qweight', 'language_model.model.layers.35.feed_forward.w3.qzeros', 'language_model.model.layers.35.feed_forward.w3.scales', 'language_model.model.layers.36.attention.wo.qweight', 'language_model.model.layers.36.attention.wo.qzeros', 'language_model.model.layers.36.attention.wo.scales', 'language_model.model.layers.36.attention.wqkv.qweight', 'language_model.model.layers.36.attention.wqkv.qzeros', 'language_model.model.layers.36.attention.wqkv.scales', 'language_model.model.layers.36.feed_forward.w1.qweight', 'language_model.model.layers.36.feed_forward.w1.qzeros', 'language_model.model.layers.36.feed_forward.w1.scales', 'language_model.model.layers.36.feed_forward.w2.qweight', 'language_model.model.layers.36.feed_forward.w2.qzeros', 'language_model.model.layers.36.feed_forward.w2.scales', 'language_model.model.layers.36.feed_forward.w3.qweight', 'language_model.model.layers.36.feed_forward.w3.qzeros', 'language_model.model.layers.36.feed_forward.w3.scales', 'language_model.model.layers.37.attention.wo.qweight', 'language_model.model.layers.37.attention.wo.qzeros', 'language_model.model.layers.37.attention.wo.scales', 'language_model.model.layers.37.attention.wqkv.qweight', 'language_model.model.layers.37.attention.wqkv.qzeros', 'language_model.model.layers.37.attention.wqkv.scales', 'language_model.model.layers.37.feed_forward.w1.qweight', 'language_model.model.layers.37.feed_forward.w1.qzeros', 'language_model.model.layers.37.feed_forward.w1.scales', 'language_model.model.layers.37.feed_forward.w2.qweight', 'language_model.model.layers.37.feed_forward.w2.qzeros', 'language_model.model.layers.37.feed_forward.w2.scales', 'language_model.model.layers.37.feed_forward.w3.qweight', 'language_model.model.layers.37.feed_forward.w3.qzeros', 'language_model.model.layers.37.feed_forward.w3.scales', 'language_model.model.layers.38.attention.wo.qweight', 'language_model.model.layers.38.attention.wo.qzeros', 'language_model.model.layers.38.attention.wo.scales', 'language_model.model.layers.38.attention.wqkv.qweight', 'language_model.model.layers.38.attention.wqkv.qzeros', 'language_model.model.layers.38.attention.wqkv.scales', 'language_model.model.layers.38.feed_forward.w1.qweight', 'language_model.model.layers.38.feed_forward.w1.qzeros', 'language_model.model.layers.38.feed_forward.w1.scales', 'language_model.model.layers.38.feed_forward.w2.qweight', 'language_model.model.layers.38.feed_forward.w2.qzeros', 'language_model.model.layers.38.feed_forward.w2.scales', 'language_model.model.layers.38.feed_forward.w3.qweight', 'language_model.model.layers.38.feed_forward.w3.qzeros', 'language_model.model.layers.38.feed_forward.w3.scales', 'language_model.model.layers.39.attention.wo.qweight', 'language_model.model.layers.39.attention.wo.qzeros', 'language_model.model.layers.39.attention.wo.scales', 'language_model.model.layers.39.attention.wqkv.qweight', 'language_model.model.layers.39.attention.wqkv.qzeros', 'language_model.model.layers.39.attention.wqkv.scales', 'language_model.model.layers.39.feed_forward.w1.qweight', 'language_model.model.layers.39.feed_forward.w1.qzeros', 'language_model.model.layers.39.feed_forward.w1.scales', 'language_model.model.layers.39.feed_forward.w2.qweight', 'language_model.model.layers.39.feed_forward.w2.qzeros', 'language_model.model.layers.39.feed_forward.w2.scales', 'language_model.model.layers.39.feed_forward.w3.qweight', 'language_model.model.layers.39.feed_forward.w3.qzeros', 'language_model.model.layers.39.feed_forward.w3.scales', 'language_model.model.layers.4.attention.wo.qweight', 'language_model.model.layers.4.attention.wo.qzeros', 'language_model.model.layers.4.attention.wo.scales', 'language_model.model.layers.4.attention.wqkv.qweight', 'language_model.model.layers.4.attention.wqkv.qzeros', 'language_model.model.layers.4.attention.wqkv.scales', 'language_model.model.layers.4.feed_forward.w1.qweight', 'language_model.model.layers.4.feed_forward.w1.qzeros', 'language_model.model.layers.4.feed_forward.w1.scales', 'language_model.model.layers.4.feed_forward.w2.qweight', 'language_model.model.layers.4.feed_forward.w2.qzeros', 'language_model.model.layers.4.feed_forward.w2.scales', 'language_model.model.layers.4.feed_forward.w3.qweight', 'language_model.model.layers.4.feed_forward.w3.qzeros', 'language_model.model.layers.4.feed_forward.w3.scales', 'language_model.model.layers.40.attention.wo.qweight', 'language_model.model.layers.40.attention.wo.qzeros', 'language_model.model.layers.40.attention.wo.scales', 'language_model.model.layers.40.attention.wqkv.qweight', 'language_model.model.layers.40.attention.wqkv.qzeros', 'language_model.model.layers.40.attention.wqkv.scales', 'language_model.model.layers.40.feed_forward.w1.qweight', 'language_model.model.layers.40.feed_forward.w1.qzeros', 'language_model.model.layers.40.feed_forward.w1.scales', 'language_model.model.layers.40.feed_forward.w2.qweight', 'language_model.model.layers.40.feed_forward.w2.qzeros', 'language_model.model.layers.40.feed_forward.w2.scales', 'language_model.model.layers.40.feed_forward.w3.qweight', 'language_model.model.layers.40.feed_forward.w3.qzeros', 'language_model.model.layers.40.feed_forward.w3.scales', 'language_model.model.layers.41.attention.wo.qweight', 'language_model.model.layers.41.attention.wo.qzeros', 'language_model.model.layers.41.attention.wo.scales', 'language_model.model.layers.41.attention.wqkv.qweight', 'language_model.model.layers.41.attention.wqkv.qzeros', 'language_model.model.layers.41.attention.wqkv.scales', 'language_model.model.layers.41.feed_forward.w1.qweight', 'language_model.model.layers.41.feed_forward.w1.qzeros', 'language_model.model.layers.41.feed_forward.w1.scales', 'language_model.model.layers.41.feed_forward.w2.qweight', 'language_model.model.layers.41.feed_forward.w2.qzeros', 'language_model.model.layers.41.feed_forward.w2.scales', 'language_model.model.layers.41.feed_forward.w3.qweight', 'language_model.model.layers.41.feed_forward.w3.qzeros', 'language_model.model.layers.41.feed_forward.w3.scales', 'language_model.model.layers.42.attention.wo.qweight', 'language_model.model.layers.42.attention.wo.qzeros', 'language_model.model.layers.42.attention.wo.scales', 'language_model.model.layers.42.attention.wqkv.qweight', 'language_model.model.layers.42.attention.wqkv.qzeros', 'language_model.model.layers.42.attention.wqkv.scales', 'language_model.model.layers.42.feed_forward.w1.qweight', 'language_model.model.layers.42.feed_forward.w1.qzeros', 'language_model.model.layers.42.feed_forward.w1.scales', 'language_model.model.layers.42.feed_forward.w2.qweight', 'language_model.model.layers.42.feed_forward.w2.qzeros', 'language_model.model.layers.42.feed_forward.w2.scales', 'language_model.model.layers.42.feed_forward.w3.qweight', 'language_model.model.layers.42.feed_forward.w3.qzeros', 'language_model.model.layers.42.feed_forward.w3.scales', 'language_model.model.layers.43.attention.wo.qweight', 'language_model.model.layers.43.attention.wo.qzeros', 'language_model.model.layers.43.attention.wo.scales', 'language_model.model.layers.43.attention.wqkv.qweight', 'language_model.model.layers.43.attention.wqkv.qzeros', 'language_model.model.layers.43.attention.wqkv.scales', 'language_model.model.layers.43.feed_forward.w1.qweight', 'language_model.model.layers.43.feed_forward.w1.qzeros', 'language_model.model.layers.43.feed_forward.w1.scales', 'language_model.model.layers.43.feed_forward.w2.qweight', 'language_model.model.layers.43.feed_forward.w2.qzeros', 'language_model.model.layers.43.feed_forward.w2.scales', 'language_model.model.layers.43.feed_forward.w3.qweight', 'language_model.model.layers.43.feed_forward.w3.qzeros', 'language_model.model.layers.43.feed_forward.w3.scales', 'language_model.model.layers.44.attention.wo.qweight', 'language_model.model.layers.44.attention.wo.qzeros', 'language_model.model.layers.44.attention.wo.scales', 'language_model.model.layers.44.attention.wqkv.qweight', 'language_model.model.layers.44.attention.wqkv.qzeros', 'language_model.model.layers.44.attention.wqkv.scales', 'language_model.model.layers.44.feed_forward.w1.qweight', 'language_model.model.layers.44.feed_forward.w1.qzeros', 'language_model.model.layers.44.feed_forward.w1.scales', 'language_model.model.layers.44.feed_forward.w2.qweight', 'language_model.model.layers.44.feed_forward.w2.qzeros', 'language_model.model.layers.44.feed_forward.w2.scales', 'language_model.model.layers.44.feed_forward.w3.qweight', 'language_model.model.layers.44.feed_forward.w3.qzeros', 'language_model.model.layers.44.feed_forward.w3.scales', 'language_model.model.layers.45.attention.wo.qweight', 'language_model.model.layers.45.attention.wo.qzeros', 'language_model.model.layers.45.attention.wo.scales', 'language_model.model.layers.45.attention.wqkv.qweight', 'language_model.model.layers.45.attention.wqkv.qzeros', 'language_model.model.layers.45.attention.wqkv.scales', 'language_model.model.layers.45.feed_forward.w1.qweight', 'language_model.model.layers.45.feed_forward.w1.qzeros', 'language_model.model.layers.45.feed_forward.w1.scales', 'language_model.model.layers.45.feed_forward.w2.qweight', 'language_model.model.layers.45.feed_forward.w2.qzeros', 'language_model.model.layers.45.feed_forward.w2.scales', 'language_model.model.layers.45.feed_forward.w3.qweight', 'language_model.model.layers.45.feed_forward.w3.qzeros', 'language_model.model.layers.45.feed_forward.w3.scales', 'language_model.model.layers.46.attention.wo.qweight', 'language_model.model.layers.46.attention.wo.qzeros', 'language_model.model.layers.46.attention.wo.scales', 'language_model.model.layers.46.attention.wqkv.qweight', 'language_model.model.layers.46.attention.wqkv.qzeros', 'language_model.model.layers.46.attention.wqkv.scales', 'language_model.model.layers.46.feed_forward.w1.qweight', 'language_model.model.layers.46.feed_forward.w1.qzeros', 'language_model.model.layers.46.feed_forward.w1.scales', 'language_model.model.layers.46.feed_forward.w2.qweight', 'language_model.model.layers.46.feed_forward.w2.qzeros', 'language_model.model.layers.46.feed_forward.w2.scales', 'language_model.model.layers.46.feed_forward.w3.qweight', 'language_model.model.layers.46.feed_forward.w3.qzeros', 'language_model.model.layers.46.feed_forward.w3.scales', 'language_model.model.layers.47.attention.wo.qweight', 'language_model.model.layers.47.attention.wo.qzeros', 'language_model.model.layers.47.attention.wo.scales', 'language_model.model.layers.47.attention.wqkv.qweight', 'language_model.model.layers.47.attention.wqkv.qzeros', 'language_model.model.layers.47.attention.wqkv.scales', 'language_model.model.layers.47.feed_forward.w1.qweight', 'language_model.model.layers.47.feed_forward.w1.qzeros', 'language_model.model.layers.47.feed_forward.w1.scales', 'language_model.model.layers.47.feed_forward.w2.qweight', 'language_model.model.layers.47.feed_forward.w2.qzeros', 'language_model.model.layers.47.feed_forward.w2.scales', 'language_model.model.layers.47.feed_forward.w3.qweight', 'language_model.model.layers.47.feed_forward.w3.qzeros', 'language_model.model.layers.47.feed_forward.w3.scales', 'language_model.model.layers.5.attention.wo.qweight', 'language_model.model.layers.5.attention.wo.qzeros', 'language_model.model.layers.5.attention.wo.scales', 'language_model.model.layers.5.attention.wqkv.qweight', 'language_model.model.layers.5.attention.wqkv.qzeros', 'language_model.model.layers.5.attention.wqkv.scales', 'language_model.model.layers.5.feed_forward.w1.qweight', 'language_model.model.layers.5.feed_forward.w1.qzeros', 'language_model.model.layers.5.feed_forward.w1.scales', 'language_model.model.layers.5.feed_forward.w2.qweight', 'language_model.model.layers.5.feed_forward.w2.qzeros', 'language_model.model.layers.5.feed_forward.w2.scales', 'language_model.model.layers.5.feed_forward.w3.qweight', 'language_model.model.layers.5.feed_forward.w3.qzeros', 'language_model.model.layers.5.feed_forward.w3.scales', 'language_model.model.layers.6.attention.wo.qweight', 'language_model.model.layers.6.attention.wo.qzeros', 'language_model.model.layers.6.attention.wo.scales', 'language_model.model.layers.6.attention.wqkv.qweight', 'language_model.model.layers.6.attention.wqkv.qzeros', 'language_model.model.layers.6.attention.wqkv.scales', 'language_model.model.layers.6.feed_forward.w1.qweight', 'language_model.model.layers.6.feed_forward.w1.qzeros', 'language_model.model.layers.6.feed_forward.w1.scales', 'language_model.model.layers.6.feed_forward.w2.qweight', 'language_model.model.layers.6.feed_forward.w2.qzeros', 'language_model.model.layers.6.feed_forward.w2.scales', 'language_model.model.layers.6.feed_forward.w3.qweight', 'language_model.model.layers.6.feed_forward.w3.qzeros', 'language_model.model.layers.6.feed_forward.w3.scales', 'language_model.model.layers.7.attention.wo.qweight', 'language_model.model.layers.7.attention.wo.qzeros', 'language_model.model.layers.7.attention.wo.scales', 'language_model.model.layers.7.attention.wqkv.qweight', 'language_model.model.layers.7.attention.wqkv.qzeros', 'language_model.model.layers.7.attention.wqkv.scales', 'language_model.model.layers.7.feed_forward.w1.qweight', 'language_model.model.layers.7.feed_forward.w1.qzeros', 'language_model.model.layers.7.feed_forward.w1.scales', 'language_model.model.layers.7.feed_forward.w2.qweight', 'language_model.model.layers.7.feed_forward.w2.qzeros', 'language_model.model.layers.7.feed_forward.w2.scales', 'language_model.model.layers.7.feed_forward.w3.qweight', 'language_model.model.layers.7.feed_forward.w3.qzeros', 'language_model.model.layers.7.feed_forward.w3.scales', 'language_model.model.layers.8.attention.wo.qweight', 'language_model.model.layers.8.attention.wo.qzeros', 'language_model.model.layers.8.attention.wo.scales', 'language_model.model.layers.8.attention.wqkv.qweight', 'language_model.model.layers.8.attention.wqkv.qzeros', 'language_model.model.layers.8.attention.wqkv.scales', 'language_model.model.layers.8.feed_forward.w1.qweight', 'language_model.model.layers.8.feed_forward.w1.qzeros', 'language_model.model.layers.8.feed_forward.w1.scales', 'language_model.model.layers.8.feed_forward.w2.qweight', 'language_model.model.layers.8.feed_forward.w2.qzeros', 'language_model.model.layers.8.feed_forward.w2.scales', 'language_model.model.layers.8.feed_forward.w3.qweight', 'language_model.model.layers.8.feed_forward.w3.qzeros', 'language_model.model.layers.8.feed_forward.w3.scales', 'language_model.model.layers.9.attention.wo.qweight', 'language_model.model.layers.9.attention.wo.qzeros', 'language_model.model.layers.9.attention.wo.scales', 'language_model.model.layers.9.attention.wqkv.qweight', 'language_model.model.layers.9.attention.wqkv.qzeros', 'language_model.model.layers.9.attention.wqkv.scales', 'language_model.model.layers.9.feed_forward.w1.qweight', 'language_model.model.layers.9.feed_forward.w1.qzeros', 'language_model.model.layers.9.feed_forward.w1.scales', 'language_model.model.layers.9.feed_forward.w2.qweight', 'language_model.model.layers.9.feed_forward.w2.qzeros', 'language_model.model.layers.9.feed_forward.w2.scales', 'language_model.model.layers.9.feed_forward.w3.qweight', 'language_model.model.layers.9.feed_forward.w3.qzeros', 'language_model.model.layers.9.feed_forward.w3.scales']\n",
      "- This IS expected if you are initializing InternVLChatModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
      "- This IS NOT expected if you are initializing InternVLChatModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "Some weights of InternVLChatModel were not initialized from the model checkpoint at OpenGVLab/InternVL2_5-26B-AWQ and are newly initialized: ['language_model.model.layers.0.attention.wo.weight', 'language_model.model.layers.0.attention.wqkv.weight', 'language_model.model.layers.0.feed_forward.w1.weight', 'language_model.model.layers.0.feed_forward.w2.weight', 'language_model.model.layers.0.feed_forward.w3.weight', 'language_model.model.layers.1.attention.wo.weight', 'language_model.model.layers.1.attention.wqkv.weight', 'language_model.model.layers.1.feed_forward.w1.weight', 'language_model.model.layers.1.feed_forward.w2.weight', 'language_model.model.layers.1.feed_forward.w3.weight', 'language_model.model.layers.10.attention.wo.weight', 'language_model.model.layers.10.attention.wqkv.weight', 'language_model.model.layers.10.feed_forward.w1.weight', 'language_model.model.layers.10.feed_forward.w2.weight', 'language_model.model.layers.10.feed_forward.w3.weight', 'language_model.model.layers.11.attention.wo.weight', 'language_model.model.layers.11.attention.wqkv.weight', 'language_model.model.layers.11.feed_forward.w1.weight', 'language_model.model.layers.11.feed_forward.w2.weight', 'language_model.model.layers.11.feed_forward.w3.weight', 'language_model.model.layers.12.attention.wo.weight', 'language_model.model.layers.12.attention.wqkv.weight', 'language_model.model.layers.12.feed_forward.w1.weight', 'language_model.model.layers.12.feed_forward.w2.weight', 'language_model.model.layers.12.feed_forward.w3.weight', 'language_model.model.layers.13.attention.wo.weight', 'language_model.model.layers.13.attention.wqkv.weight', 'language_model.model.layers.13.feed_forward.w1.weight', 'language_model.model.layers.13.feed_forward.w2.weight', 'language_model.model.layers.13.feed_forward.w3.weight', 'language_model.model.layers.14.attention.wo.weight', 'language_model.model.layers.14.attention.wqkv.weight', 'language_model.model.layers.14.feed_forward.w1.weight', 'language_model.model.layers.14.feed_forward.w2.weight', 'language_model.model.layers.14.feed_forward.w3.weight', 'language_model.model.layers.15.attention.wo.weight', 'language_model.model.layers.15.attention.wqkv.weight', 'language_model.model.layers.15.feed_forward.w1.weight', 'language_model.model.layers.15.feed_forward.w2.weight', 'language_model.model.layers.15.feed_forward.w3.weight', 'language_model.model.layers.16.attention.wo.weight', 'language_model.model.layers.16.attention.wqkv.weight', 'language_model.model.layers.16.feed_forward.w1.weight', 'language_model.model.layers.16.feed_forward.w2.weight', 'language_model.model.layers.16.feed_forward.w3.weight', 'language_model.model.layers.17.attention.wo.weight', 'language_model.model.layers.17.attention.wqkv.weight', 'language_model.model.layers.17.feed_forward.w1.weight', 'language_model.model.layers.17.feed_forward.w2.weight', 'language_model.model.layers.17.feed_forward.w3.weight', 'language_model.model.layers.18.attention.wo.weight', 'language_model.model.layers.18.attention.wqkv.weight', 'language_model.model.layers.18.feed_forward.w1.weight', 'language_model.model.layers.18.feed_forward.w2.weight', 'language_model.model.layers.18.feed_forward.w3.weight', 'language_model.model.layers.19.attention.wo.weight', 'language_model.model.layers.19.attention.wqkv.weight', 'language_model.model.layers.19.feed_forward.w1.weight', 'language_model.model.layers.19.feed_forward.w2.weight', 'language_model.model.layers.19.feed_forward.w3.weight', 'language_model.model.layers.2.attention.wo.weight', 'language_model.model.layers.2.attention.wqkv.weight', 'language_model.model.layers.2.feed_forward.w1.weight', 'language_model.model.layers.2.feed_forward.w2.weight', 'language_model.model.layers.2.feed_forward.w3.weight', 'language_model.model.layers.20.attention.wo.weight', 'language_model.model.layers.20.attention.wqkv.weight', 'language_model.model.layers.20.feed_forward.w1.weight', 'language_model.model.layers.20.feed_forward.w2.weight', 'language_model.model.layers.20.feed_forward.w3.weight', 'language_model.model.layers.21.attention.wo.weight', 'language_model.model.layers.21.attention.wqkv.weight', 'language_model.model.layers.21.feed_forward.w1.weight', 'language_model.model.layers.21.feed_forward.w2.weight', 'language_model.model.layers.21.feed_forward.w3.weight', 'language_model.model.layers.22.attention.wo.weight', 'language_model.model.layers.22.attention.wqkv.weight', 'language_model.model.layers.22.feed_forward.w1.weight', 'language_model.model.layers.22.feed_forward.w2.weight', 'language_model.model.layers.22.feed_forward.w3.weight', 'language_model.model.layers.23.attention.wo.weight', 'language_model.model.layers.23.attention.wqkv.weight', 'language_model.model.layers.23.feed_forward.w1.weight', 'language_model.model.layers.23.feed_forward.w2.weight', 'language_model.model.layers.23.feed_forward.w3.weight', 'language_model.model.layers.24.attention.wo.weight', 'language_model.model.layers.24.attention.wqkv.weight', 'language_model.model.layers.24.feed_forward.w1.weight', 'language_model.model.layers.24.feed_forward.w2.weight', 'language_model.model.layers.24.feed_forward.w3.weight', 'language_model.model.layers.25.attention.wo.weight', 'language_model.model.layers.25.attention.wqkv.weight', 'language_model.model.layers.25.feed_forward.w1.weight', 'language_model.model.layers.25.feed_forward.w2.weight', 'language_model.model.layers.25.feed_forward.w3.weight', 'language_model.model.layers.26.attention.wo.weight', 'language_model.model.layers.26.attention.wqkv.weight', 'language_model.model.layers.26.feed_forward.w1.weight', 'language_model.model.layers.26.feed_forward.w2.weight', 'language_model.model.layers.26.feed_forward.w3.weight', 'language_model.model.layers.27.attention.wo.weight', 'language_model.model.layers.27.attention.wqkv.weight', 'language_model.model.layers.27.feed_forward.w1.weight', 'language_model.model.layers.27.feed_forward.w2.weight', 'language_model.model.layers.27.feed_forward.w3.weight', 'language_model.model.layers.28.attention.wo.weight', 'language_model.model.layers.28.attention.wqkv.weight', 'language_model.model.layers.28.feed_forward.w1.weight', 'language_model.model.layers.28.feed_forward.w2.weight', 'language_model.model.layers.28.feed_forward.w3.weight', 'language_model.model.layers.29.attention.wo.weight', 'language_model.model.layers.29.attention.wqkv.weight', 'language_model.model.layers.29.feed_forward.w1.weight', 'language_model.model.layers.29.feed_forward.w2.weight', 'language_model.model.layers.29.feed_forward.w3.weight', 'language_model.model.layers.3.attention.wo.weight', 'language_model.model.layers.3.attention.wqkv.weight', 'language_model.model.layers.3.feed_forward.w1.weight', 'language_model.model.layers.3.feed_forward.w2.weight', 'language_model.model.layers.3.feed_forward.w3.weight', 'language_model.model.layers.30.attention.wo.weight', 'language_model.model.layers.30.attention.wqkv.weight', 'language_model.model.layers.30.feed_forward.w1.weight', 'language_model.model.layers.30.feed_forward.w2.weight', 'language_model.model.layers.30.feed_forward.w3.weight', 'language_model.model.layers.31.attention.wo.weight', 'language_model.model.layers.31.attention.wqkv.weight', 'language_model.model.layers.31.feed_forward.w1.weight', 'language_model.model.layers.31.feed_forward.w2.weight', 'language_model.model.layers.31.feed_forward.w3.weight', 'language_model.model.layers.32.attention.wo.weight', 'language_model.model.layers.32.attention.wqkv.weight', 'language_model.model.layers.32.feed_forward.w1.weight', 'language_model.model.layers.32.feed_forward.w2.weight', 'language_model.model.layers.32.feed_forward.w3.weight', 'language_model.model.layers.33.attention.wo.weight', 'language_model.model.layers.33.attention.wqkv.weight', 'language_model.model.layers.33.feed_forward.w1.weight', 'language_model.model.layers.33.feed_forward.w2.weight', 'language_model.model.layers.33.feed_forward.w3.weight', 'language_model.model.layers.34.attention.wo.weight', 'language_model.model.layers.34.attention.wqkv.weight', 'language_model.model.layers.34.feed_forward.w1.weight', 'language_model.model.layers.34.feed_forward.w2.weight', 'language_model.model.layers.34.feed_forward.w3.weight', 'language_model.model.layers.35.attention.wo.weight', 'language_model.model.layers.35.attention.wqkv.weight', 'language_model.model.layers.35.feed_forward.w1.weight', 'language_model.model.layers.35.feed_forward.w2.weight', 'language_model.model.layers.35.feed_forward.w3.weight', 'language_model.model.layers.36.attention.wo.weight', 'language_model.model.layers.36.attention.wqkv.weight', 'language_model.model.layers.36.feed_forward.w1.weight', 'language_model.model.layers.36.feed_forward.w2.weight', 'language_model.model.layers.36.feed_forward.w3.weight', 'language_model.model.layers.37.attention.wo.weight', 'language_model.model.layers.37.attention.wqkv.weight', 'language_model.model.layers.37.feed_forward.w1.weight', 'language_model.model.layers.37.feed_forward.w2.weight', 'language_model.model.layers.37.feed_forward.w3.weight', 'language_model.model.layers.38.attention.wo.weight', 'language_model.model.layers.38.attention.wqkv.weight', 'language_model.model.layers.38.feed_forward.w1.weight', 'language_model.model.layers.38.feed_forward.w2.weight', 'language_model.model.layers.38.feed_forward.w3.weight', 'language_model.model.layers.39.attention.wo.weight', 'language_model.model.layers.39.attention.wqkv.weight', 'language_model.model.layers.39.feed_forward.w1.weight', 'language_model.model.layers.39.feed_forward.w2.weight', 'language_model.model.layers.39.feed_forward.w3.weight', 'language_model.model.layers.4.attention.wo.weight', 'language_model.model.layers.4.attention.wqkv.weight', 'language_model.model.layers.4.feed_forward.w1.weight', 'language_model.model.layers.4.feed_forward.w2.weight', 'language_model.model.layers.4.feed_forward.w3.weight', 'language_model.model.layers.40.attention.wo.weight', 'language_model.model.layers.40.attention.wqkv.weight', 'language_model.model.layers.40.feed_forward.w1.weight', 'language_model.model.layers.40.feed_forward.w2.weight', 'language_model.model.layers.40.feed_forward.w3.weight', 'language_model.model.layers.41.attention.wo.weight', 'language_model.model.layers.41.attention.wqkv.weight', 'language_model.model.layers.41.feed_forward.w1.weight', 'language_model.model.layers.41.feed_forward.w2.weight', 'language_model.model.layers.41.feed_forward.w3.weight', 'language_model.model.layers.42.attention.wo.weight', 'language_model.model.layers.42.attention.wqkv.weight', 'language_model.model.layers.42.feed_forward.w1.weight', 'language_model.model.layers.42.feed_forward.w2.weight', 'language_model.model.layers.42.feed_forward.w3.weight', 'language_model.model.layers.43.attention.wo.weight', 'language_model.model.layers.43.attention.wqkv.weight', 'language_model.model.layers.43.feed_forward.w1.weight', 'language_model.model.layers.43.feed_forward.w2.weight', 'language_model.model.layers.43.feed_forward.w3.weight', 'language_model.model.layers.44.attention.wo.weight', 'language_model.model.layers.44.attention.wqkv.weight', 'language_model.model.layers.44.feed_forward.w1.weight', 'language_model.model.layers.44.feed_forward.w2.weight', 'language_model.model.layers.44.feed_forward.w3.weight', 'language_model.model.layers.45.attention.wo.weight', 'language_model.model.layers.45.attention.wqkv.weight', 'language_model.model.layers.45.feed_forward.w1.weight', 'language_model.model.layers.45.feed_forward.w2.weight', 'language_model.model.layers.45.feed_forward.w3.weight', 'language_model.model.layers.46.attention.wo.weight', 'language_model.model.layers.46.attention.wqkv.weight', 'language_model.model.layers.46.feed_forward.w1.weight', 'language_model.model.layers.46.feed_forward.w2.weight', 'language_model.model.layers.46.feed_forward.w3.weight', 'language_model.model.layers.47.attention.wo.weight', 'language_model.model.layers.47.attention.wqkv.weight', 'language_model.model.layers.47.feed_forward.w1.weight', 'language_model.model.layers.47.feed_forward.w2.weight', 'language_model.model.layers.47.feed_forward.w3.weight', 'language_model.model.layers.5.attention.wo.weight', 'language_model.model.layers.5.attention.wqkv.weight', 'language_model.model.layers.5.feed_forward.w1.weight', 'language_model.model.layers.5.feed_forward.w2.weight', 'language_model.model.layers.5.feed_forward.w3.weight', 'language_model.model.layers.6.attention.wo.weight', 'language_model.model.layers.6.attention.wqkv.weight', 'language_model.model.layers.6.feed_forward.w1.weight', 'language_model.model.layers.6.feed_forward.w2.weight', 'language_model.model.layers.6.feed_forward.w3.weight', 'language_model.model.layers.7.attention.wo.weight', 'language_model.model.layers.7.attention.wqkv.weight', 'language_model.model.layers.7.feed_forward.w1.weight', 'language_model.model.layers.7.feed_forward.w2.weight', 'language_model.model.layers.7.feed_forward.w3.weight', 'language_model.model.layers.8.attention.wo.weight', 'language_model.model.layers.8.attention.wqkv.weight', 'language_model.model.layers.8.feed_forward.w1.weight', 'language_model.model.layers.8.feed_forward.w2.weight', 'language_model.model.layers.8.feed_forward.w3.weight', 'language_model.model.layers.9.attention.wo.weight', 'language_model.model.layers.9.attention.wqkv.weight', 'language_model.model.layers.9.feed_forward.w1.weight', 'language_model.model.layers.9.feed_forward.w2.weight', 'language_model.model.layers.9.feed_forward.w3.weight']\n",
      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
     ]
    },
    {
     "ename": "OutOfMemoryError",
     "evalue": "CUDA out of memory. Tried to allocate 192.00 MiB. GPU 0 has a total capacity of 39.39 GiB of which 186.81 MiB is free. Process 1779400 has 39.19 GiB memory in use. Of the allocated memory 38.27 GiB is allocated by PyTorch, and 530.99 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mOutOfMemoryError\u001b[0m                          Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[3], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[43msupported_VLM\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mInternVL2_5-26B-AWQ\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/vlm_clone/VLMEvalKit_old/vlmeval/vlm/internvl_chat.py:293\u001b[0m, in \u001b[0;36mInternVLChat.__init__\u001b[0;34m(self, model_path, load_in_8bit, cot_prompt, version, **kwargs)\u001b[0m\n\u001b[1;32m    286\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmodel \u001b[38;5;241m=\u001b[39m AutoModel\u001b[38;5;241m.\u001b[39mfrom_pretrained(\n\u001b[1;32m    287\u001b[0m         model_path,\n\u001b[1;32m    288\u001b[0m         torch_dtype\u001b[38;5;241m=\u001b[39mtorch\u001b[38;5;241m.\u001b[39mbfloat16,\n\u001b[1;32m    289\u001b[0m         load_in_8bit\u001b[38;5;241m=\u001b[39mload_in_8bit,\n\u001b[1;32m    290\u001b[0m         trust_remote_code\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mTrue\u001b[39;00m,\n\u001b[1;32m    291\u001b[0m     )\u001b[38;5;241m.\u001b[39meval()\n\u001b[1;32m    292\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m load_in_8bit:\n\u001b[0;32m--> 293\u001b[0m         \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmodel \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mcuda\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m    295\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mimage_size \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mmodel\u001b[38;5;241m.\u001b[39mconfig\u001b[38;5;241m.\u001b[39mvision_config\u001b[38;5;241m.\u001b[39mimage_size\n\u001b[1;32m    296\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mversion \u001b[38;5;241m=\u001b[39m version\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/transformers/modeling_utils.py:3164\u001b[0m, in \u001b[0;36mPreTrainedModel.to\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   3159\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m dtype_present_in_args:\n\u001b[1;32m   3160\u001b[0m         \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\n\u001b[1;32m   3161\u001b[0m             \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mYou cannot cast a GPTQ model in a new `dtype`. Make sure to load the model using `from_pretrained` using the desired\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m   3162\u001b[0m             \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m `dtype` by passing the correct `torch_dtype` argument.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m   3163\u001b[0m         )\n\u001b[0;32m-> 3164\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43msuper\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/torch/nn/modules/module.py:1340\u001b[0m, in \u001b[0;36mModule.to\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1337\u001b[0m         \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m   1338\u001b[0m             \u001b[38;5;28;01mraise\u001b[39;00m\n\u001b[0;32m-> 1340\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_apply\u001b[49m\u001b[43m(\u001b[49m\u001b[43mconvert\u001b[49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/torch/nn/modules/module.py:900\u001b[0m, in \u001b[0;36mModule._apply\u001b[0;34m(self, fn, recurse)\u001b[0m\n\u001b[1;32m    898\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m recurse:\n\u001b[1;32m    899\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m module \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mchildren():\n\u001b[0;32m--> 900\u001b[0m         \u001b[43mmodule\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_apply\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfn\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    902\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mcompute_should_use_set_data\u001b[39m(tensor, tensor_applied):\n\u001b[1;32m    903\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m torch\u001b[38;5;241m.\u001b[39m_has_compatible_shallow_copy_type(tensor, tensor_applied):\n\u001b[1;32m    904\u001b[0m         \u001b[38;5;66;03m# If the new tensor has compatible tensor type as the existing tensor,\u001b[39;00m\n\u001b[1;32m    905\u001b[0m         \u001b[38;5;66;03m# the current behavior is to change the tensor in-place using `.data =`,\u001b[39;00m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    910\u001b[0m         \u001b[38;5;66;03m# global flag to let the user control whether they want the future\u001b[39;00m\n\u001b[1;32m    911\u001b[0m         \u001b[38;5;66;03m# behavior of overwriting the existing tensor or not.\u001b[39;00m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/torch/nn/modules/module.py:900\u001b[0m, in \u001b[0;36mModule._apply\u001b[0;34m(self, fn, recurse)\u001b[0m\n\u001b[1;32m    898\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m recurse:\n\u001b[1;32m    899\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m module \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mchildren():\n\u001b[0;32m--> 900\u001b[0m         \u001b[43mmodule\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_apply\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfn\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    902\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mcompute_should_use_set_data\u001b[39m(tensor, tensor_applied):\n\u001b[1;32m    903\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m torch\u001b[38;5;241m.\u001b[39m_has_compatible_shallow_copy_type(tensor, tensor_applied):\n\u001b[1;32m    904\u001b[0m         \u001b[38;5;66;03m# If the new tensor has compatible tensor type as the existing tensor,\u001b[39;00m\n\u001b[1;32m    905\u001b[0m         \u001b[38;5;66;03m# the current behavior is to change the tensor in-place using `.data =`,\u001b[39;00m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    910\u001b[0m         \u001b[38;5;66;03m# global flag to let the user control whether they want the future\u001b[39;00m\n\u001b[1;32m    911\u001b[0m         \u001b[38;5;66;03m# behavior of overwriting the existing tensor or not.\u001b[39;00m\n",
      "    \u001b[0;31m[... skipping similar frames: Module._apply at line 900 (3 times)]\u001b[0m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/torch/nn/modules/module.py:900\u001b[0m, in \u001b[0;36mModule._apply\u001b[0;34m(self, fn, recurse)\u001b[0m\n\u001b[1;32m    898\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m recurse:\n\u001b[1;32m    899\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m module \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mchildren():\n\u001b[0;32m--> 900\u001b[0m         \u001b[43mmodule\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_apply\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfn\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    902\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mcompute_should_use_set_data\u001b[39m(tensor, tensor_applied):\n\u001b[1;32m    903\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m torch\u001b[38;5;241m.\u001b[39m_has_compatible_shallow_copy_type(tensor, tensor_applied):\n\u001b[1;32m    904\u001b[0m         \u001b[38;5;66;03m# If the new tensor has compatible tensor type as the existing tensor,\u001b[39;00m\n\u001b[1;32m    905\u001b[0m         \u001b[38;5;66;03m# the current behavior is to change the tensor in-place using `.data =`,\u001b[39;00m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    910\u001b[0m         \u001b[38;5;66;03m# global flag to let the user control whether they want the future\u001b[39;00m\n\u001b[1;32m    911\u001b[0m         \u001b[38;5;66;03m# behavior of overwriting the existing tensor or not.\u001b[39;00m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/torch/nn/modules/module.py:927\u001b[0m, in \u001b[0;36mModule._apply\u001b[0;34m(self, fn, recurse)\u001b[0m\n\u001b[1;32m    923\u001b[0m \u001b[38;5;66;03m# Tensors stored in modules are graph leaves, and we don't want to\u001b[39;00m\n\u001b[1;32m    924\u001b[0m \u001b[38;5;66;03m# track autograd history of `param_applied`, so we have to use\u001b[39;00m\n\u001b[1;32m    925\u001b[0m \u001b[38;5;66;03m# `with torch.no_grad():`\u001b[39;00m\n\u001b[1;32m    926\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m torch\u001b[38;5;241m.\u001b[39mno_grad():\n\u001b[0;32m--> 927\u001b[0m     param_applied \u001b[38;5;241m=\u001b[39m \u001b[43mfn\u001b[49m\u001b[43m(\u001b[49m\u001b[43mparam\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    928\u001b[0m p_should_use_set_data \u001b[38;5;241m=\u001b[39m compute_should_use_set_data(param, param_applied)\n\u001b[1;32m    930\u001b[0m \u001b[38;5;66;03m# subclasses may have multiple child tensors so we need to use swap_tensors\u001b[39;00m\n",
      "File \u001b[0;32m/dscilab_dungvo/workspace/conda/envs/vlmeval/lib/python3.10/site-packages/torch/nn/modules/module.py:1326\u001b[0m, in \u001b[0;36mModule.to.<locals>.convert\u001b[0;34m(t)\u001b[0m\n\u001b[1;32m   1319\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m convert_to_format \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m t\u001b[38;5;241m.\u001b[39mdim() \u001b[38;5;129;01min\u001b[39;00m (\u001b[38;5;241m4\u001b[39m, \u001b[38;5;241m5\u001b[39m):\n\u001b[1;32m   1320\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m t\u001b[38;5;241m.\u001b[39mto(\n\u001b[1;32m   1321\u001b[0m             device,\n\u001b[1;32m   1322\u001b[0m             dtype \u001b[38;5;28;01mif\u001b[39;00m t\u001b[38;5;241m.\u001b[39mis_floating_point() \u001b[38;5;129;01mor\u001b[39;00m t\u001b[38;5;241m.\u001b[39mis_complex() \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m   1323\u001b[0m             non_blocking,\n\u001b[1;32m   1324\u001b[0m             memory_format\u001b[38;5;241m=\u001b[39mconvert_to_format,\n\u001b[1;32m   1325\u001b[0m         )\n\u001b[0;32m-> 1326\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mt\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mto\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   1327\u001b[0m \u001b[43m        \u001b[49m\u001b[43mdevice\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1328\u001b[0m \u001b[43m        \u001b[49m\u001b[43mdtype\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mif\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mis_floating_point\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01mor\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mt\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mis_complex\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01melse\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mNone\u001b[39;49;00m\u001b[43m,\u001b[49m\n\u001b[1;32m   1329\u001b[0m \u001b[43m        \u001b[49m\u001b[43mnon_blocking\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   1330\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1331\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mNotImplementedError\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[1;32m   1332\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mstr\u001b[39m(e) \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mCannot copy out of meta tensor; no data!\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n",
      "\u001b[0;31mOutOfMemoryError\u001b[0m: CUDA out of memory. Tried to allocate 192.00 MiB. GPU 0 has a total capacity of 39.39 GiB of which 186.81 MiB is free. Process 1779400 has 39.19 GiB memory in use. Of the allocated memory 38.27 GiB is allocated by PyTorch, and 530.99 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
     ]
    }
   ],
   "source": [
    "model = supported_VLM['InternVL2_5-4B-AWQ']()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}