{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d06d32d3",
   "metadata": {},
   "source": [
    "# Import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "29b55fb3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "from PIL import Image\n",
    "from torchvision.io import read_image\n",
    "from transformers import AutoModel\n",
    "device = 'cuda'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a713178e",
   "metadata": {},
   "source": [
    "# Plot Utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3a7e5670",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_qualitative(image, sim, palette):\n",
    "    qualitative_plot = np.zeros((sim.shape[0], sim.shape[1], 3)).astype(np.uint8)\n",
    "\n",
    "    for j in list(np.unique(sim)):\n",
    "        qualitative_plot[sim == j] = np.array(palette[j])\n",
    "    plt.axis('off')\n",
    "    plt.imshow(image)\n",
    "    plt.imshow(qualitative_plot, alpha=0.6)\n",
    "    plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6e0c4fc",
   "metadata": {},
   "source": [
    "# Target Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "04851c21",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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",
      "image/png": "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",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=578x578>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "image = Image.open(\"assets/pikachu.png\").convert(\"RGB\")\n",
    "image"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0679fca6",
   "metadata": {},
   "source": [
    "# Model Loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "76d731a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using cache found in /raid/homes/lorenzo.bianchi/.cache/torch/hub/facebookresearch_dinov2_main\n",
      "/raid/homes/lorenzo.bianchi/.cache/torch/hub/facebookresearch_dinov2_main/dinov2/layers/swiglu_ffn.py:51: UserWarning: xFormers is not available (SwiGLU)\n",
      "  warnings.warn(\"xFormers is not available (SwiGLU)\")\n",
      "/raid/homes/lorenzo.bianchi/.cache/torch/hub/facebookresearch_dinov2_main/dinov2/layers/attention.py:33: UserWarning: xFormers is not available (Attention)\n",
      "  warnings.warn(\"xFormers is not available (Attention)\")\n",
      "/raid/homes/lorenzo.bianchi/.cache/torch/hub/facebookresearch_dinov2_main/dinov2/layers/block.py:40: UserWarning: xFormers is not available (Block)\n",
      "  warnings.warn(\"xFormers is not available (Block)\")\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for cls_token: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for pos_embed: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for register_tokens: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for mask_token: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for patch_embed.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for patch_embed.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.0.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.1.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.2.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.3.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.4.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.5.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.6.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.7.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.8.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.9.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.10.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.11.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.12.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.13.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.14.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.15.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.16.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.17.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.18.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.19.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.20.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.21.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.22.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.norm1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.norm1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.attn.qkv.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.attn.qkv.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.attn.proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.attn.proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.ls1.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.norm2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.norm2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.mlp.fc1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.mlp.fc1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.mlp.fc2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.mlp.fc2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for blocks.23.ls2.gamma: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for norm.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for norm.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for positional_embedding: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for text_projection: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for logit_scale: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.class_embedding: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.positional_embedding: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.proj: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.conv1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.ln_pre.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.ln_pre.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.0.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.1.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.2.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.3.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.4.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.5.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.6.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.7.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.8.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.9.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.10.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.transformer.resblocks.11.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.ln_post.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for visual.ln_post.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.0.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.1.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.2.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.3.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.4.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.5.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.6.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.7.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.8.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.9.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.10.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.attn.in_proj_weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.attn.in_proj_bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.attn.out_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.attn.out_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.ln_1.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.ln_1.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.mlp.c_fc.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.mlp.c_fc.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.mlp.c_proj.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.mlp.c_proj.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.ln_2.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for transformer.resblocks.11.ln_2.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for token_embedding.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for ln_final.weight: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "/raid/homes/lorenzo.bianchi/miniconda3/envs/talk2dino_hf/lib/python3.10/site-packages/torch/nn/modules/module.py:2400: UserWarning: for ln_final.bias: copying from a non-meta parameter in the checkpoint to a meta parameter in the current model, which is a no-op. (Did you mean to pass `assign=True` to assign items in the state dictionary to their corresponding key in the module instead of copying them in place?)\n",
      "  warnings.warn(\n",
      "Some weights of the model checkpoint at . were not used when initializing Talk2DINO: ['pamr.aff_m.kernel', 'pamr.aff_std.kernel', 'pamr.aff_x.kernel']\n",
      "- This IS expected if you are initializing Talk2DINO 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 Talk2DINO from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n"
     ]
    }
   ],
   "source": [
    "model = AutoModel.from_pretrained(\"lorebianchi98/Talk2DINO-ViTL\", trust_remote_code=True).to(device).eval()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3f7e38c",
   "metadata": {},
   "source": [
    "# Mapping Texts & Images in DINOv2 space "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "eca3dce5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([1, 1024]), torch.Size([1, 1369, 1024]))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with torch.no_grad():\n",
    "    text_embed = model.encode_text(\"a pikachu\")\n",
    "    image_embed = model.encode_image(image)\n",
    "\n",
    "text_embed.shape, image_embed.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "34e00da9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# normalize the features to perform cosine similarity\n",
    "text_embed = text_embed / text_embed.norm(dim=-1, keepdim=True)\n",
    "image_embed = image_embed / image_embed.norm(dim=-1, keepdim=True)\n",
    "\n",
    "similarity = (image_embed @ text_embed.T).squeeze(0, -1).cpu().numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "9154cfbc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "patches_x_row = similarity.shape[0]**0.5\n",
    "plt.figure(figsize=(10, 10))\n",
    "plt.axis('off')\n",
    "plt.imshow(similarity.reshape(int(patches_x_row), int(patches_x_row)), cmap='magma')\n",
    "plt.colorbar()\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df020e09",
   "metadata": {},
   "source": [
    "# Generate masks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "80559b97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with_background = False\n",
    "text = \"pikachu,traffic_sign,forest,road,cap\".replace(\"_\", \" \").split(\",\")\n",
    "input_file = \"assets/pikachu.png\"\n",
    "\n",
    "img = read_image(input_file).to(device).float().unsqueeze(0)\n",
    "palette = [\n",
    "    [255, 0, 0],\n",
    "    [255, 255, 0],\n",
    "    [0, 255, 0],\n",
    "    [0, 255, 255],\n",
    "    [0, 0, 255],\n",
    "    [128, 128, 128]\n",
    "]\n",
    "if len(text) > len(palette):\n",
    "    for _ in range(len(text) - len(palette)):\n",
    "        palette.append([np.random.randint(0, 255) for _ in range(3)])\n",
    "        \n",
    "if with_background:\n",
    "    palette.insert(0, [0, 0, 0])\n",
    "    model.with_bg_clean = True\n",
    "\n",
    "with torch.no_grad():\n",
    "    text_emb = model.build_dataset_class_tokens(\"sub_imagenet_template\", text)\n",
    "    text_emb = model.build_text_embedding(text_emb)\n",
    "    \n",
    "    mask, _ = model.generate_masks(img, img_metas=None, text_emb=text_emb, classnames=text, apply_pamr=True)\n",
    "    if with_background:\n",
    "        background = torch.ones_like(mask[:, :1]) * 0.55\n",
    "        mask = torch.cat([background, mask], dim=1)\n",
    "    \n",
    "    mask = mask.argmax(dim=1)\n",
    "    \n",
    "plot_qualitative(img.cpu()[0].permute(1,2,0).int().numpy(), mask.cpu()[0].numpy(), palette)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "talk2dino_hf",
   "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": "3.10.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}