Update notebook: test gpu, etc

3 files changed, 276 insertions, 374 deletions

diff --git a/notebooks/00-test-gpu.ipynb b/notebooks/00-test-gpu.ipynb
new file mode 100644
index 0000000..7ea06ae
--- /dev/null
+++ b/notebooks/00-test-gpu.ipynb
@@ -0,0 +1,276 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "8468e45a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "import torch"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "d1bc956b",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "True"
+      ]
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.cuda.is_available()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "652cfb26",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "1"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.cuda.device_count()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "0fc5e328",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.cuda.current_device()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "4ed93c82",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<torch.cuda.device at 0x7fbf7bf7b9a0>"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.cuda.device(0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "c03841ce",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'NVIDIA GeForce RTX 2070'"
+      ]
+     },
+     "execution_count": 6,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.cuda.get_device_name(0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "2fba9680-0c73-4a62-b24d-ceea79874717",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'11.7'"
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.version.cuda"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "7e62de10",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "t = torch.randn((2, 1, 3, 64))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "fc221b4e",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "torch.Size([2, 1, 3, 64])"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "t.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "ab80d75e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import torch.nn.functional as F"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "id": "bfe1fc90",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "torch.Size([2, 1, 7, 128])"
+      ]
+     },
+     "execution_count": 21,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "F.interpolate(t, scale_factor=[2.5, 2], mode=\"nearest\").shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "f69fa1fc",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "f = torch.nn.Conv2d(1, 1, 3, 1, padding=1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "id": "cd6df204",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "torch.Size([2, 1, 64, 64])"
+      ]
+     },
+     "execution_count": 20,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "f(torch.randn(2, 1, 64, 64)).shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "41693566",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.9.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/notebooks/03-look-at-iam-lines.ipynb b/notebooks/03-look-at-iam-lines.ipynb
index 9e9b24c..049a8b6 100644
--- a/notebooks/03-look-at-iam-lines.ipynb
+++ b/notebooks/03-look-at-iam-lines.ipynb
@@ -134,39 +134,6 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "(tensor([[[0., 0., 0.,  ..., 0., 0., 0.],\n",
-       "          [0., 0., 0.,  ..., 0., 0., 0.],\n",
-       "          [0., 0., 0.,  ..., 0., 0., 0.],\n",
-       "          ...,\n",
-       "          [0., 0., 0.,  ..., 0., 0., 0.],\n",
-       "          [0., 0., 0.,  ..., 0., 0., 0.],\n",
-       "          [0., 0., 0.,  ..., 0., 0., 0.]]]),\n",
-       " tensor([ 1, 32, 27, 14, 33, 16, 21, 22, 27, 20, 40, 14, 33, 40, 33, 21, 18, 40,\n",
-       "         29, 31, 28, 29, 40, 21, 28, 25, 17, 22, 27, 20, 40, 33, 21, 18, 40, 15,\n",
-       "         14, 31, 20, 18, 45, 32,  2,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,\n",
-       "          3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,\n",
-       "          3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3]))"
-      ]
-     },
-     "execution_count": 22,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "datamodule.data_val[16]"
-   ]
-  },
-  {
-   "cell_type": "code",
    "execution_count": 8,
    "metadata": {},
    "outputs": [
@@ -190,112 +157,6 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "['<b>',\n",
-       " '<s>',\n",
-       " '<e>',\n",
-       " '<p>',\n",
-       " '0',\n",
-       " '1',\n",
-       " '2',\n",
-       " '3',\n",
-       " '4',\n",
-       " '5',\n",
-       " '6',\n",
-       " '7',\n",
-       " '8',\n",
-       " '9',\n",
-       " 'a',\n",
-       " 'b',\n",
-       " 'c',\n",
-       " 'd',\n",
-       " 'e',\n",
-       " 'f',\n",
-       " 'g',\n",
-       " 'h',\n",
-       " 'i',\n",
-       " 'j',\n",
-       " 'k',\n",
-       " 'l',\n",
-       " 'm',\n",
-       " 'n',\n",
-       " 'o',\n",
-       " 'p',\n",
-       " 'q',\n",
-       " 'r',\n",
-       " 's',\n",
-       " 't',\n",
-       " 'u',\n",
-       " 'v',\n",
-       " 'w',\n",
-       " 'x',\n",
-       " 'y',\n",
-       " 'z',\n",
-       " ' ',\n",
-       " '!',\n",
-       " '\"',\n",
-       " '#',\n",
-       " '&',\n",
-       " \"'\",\n",
-       " '(',\n",
-       " ')',\n",
-       " '*',\n",
-       " '+',\n",
-       " ',',\n",
-       " '-',\n",
-       " '.',\n",
-       " '/',\n",
-       " ':',\n",
-       " ';',\n",
-       " '?',\n",
-       " '\\n']"
-      ]
-     },
-     "execution_count": 15,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "datamodule.tokenizer.mapping"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "tensor([ 1, 34, 27, 22, 28, 27, 40, 36, 28, 34, 25, 17, 40, 15, 18, 40, 29, 31,\n",
-       "        18, 29, 14, 31, 18, 17, 40, 33, 28, 40, 31, 18, 14, 16, 21,  2,  3,  3,\n",
-       "         3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,\n",
-       "         3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,\n",
-       "         3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  3])"
-      ]
-     },
-     "execution_count": 11,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "x =dataset[0][1]\n",
-    "x["
-   ]
-  },
-  {
-   "cell_type": "code",
    "execution_count": 10,
    "metadata": {},
    "outputs": [
diff --git a/notebooks/Untitled1.ipynb b/notebooks/Untitled1.ipynb
index a2d6168..92b35c9 100644
--- a/notebooks/Untitled1.ipynb
+++ b/notebooks/Untitled1.ipynb
@@ -430,241 +430,6 @@
     "plt.figure(figsize=(40, 20))\n",
     "plt.imshow(xxx, cmap='gray')"
    ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 79,
-   "id": "c7c30e67-0cd7-4c23-adcc-56c86450bd37",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "torch.device"
-      ]
-     },
-     "execution_count": 79,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "type(t.device)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 86,
-   "id": "a6f270cc-20a2-4aae-8006-cb956eeed44c",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [],
-   "source": [
-    "mv =-torch.finfo(t.dtype).max"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 93,
-   "id": "390d8a9d-2002-456f-93f9-b4e01b550024",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [],
-   "source": [
-    "a = torch.tensor([1., 1., 2., 2.])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 94,
-   "id": "55efcc9d-9f61-46fb-8417-0a3443332b93",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [],
-   "source": [
-    "b = torch.tensor([1., 1., 2., 2.]) != 2."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 95,
-   "id": "a0629f46-06b7-42dd-9fd7-d7d9da95faf6",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "tensor([-3.4028e+38, -3.4028e+38,  2.0000e+00,  2.0000e+00])"
-      ]
-     },
-     "execution_count": 95,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "a.masked_fill_(b, mv)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 96,
-   "id": "516339e8-445a-4459-8fec-f028e3201bce",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "tensor([ True,  True, False, False])"
-      ]
-     },
-     "execution_count": 96,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "b\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "id": "c0e733d8-c17d-46f5-b484-9c74e46d7308",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [],
-   "source": [
-    "from typing import Union\n",
-    "\n",
-    "import torch\n",
-    "\n",
-    "\n",
-    "def first_appearance(x: torch.Tensor, element: Union[int, float], dim: int = 1) -> torch.Tensor:\n",
-    "    \"\"\"Return indices of first appearance of element in x, collapsing along dim.\n",
-    "\n",
-    "    Based on https://discuss.pytorch.org/t/first-nonzero-index/24769/9\n",
-    "\n",
-    "    Parameters\n",
-    "    ----------\n",
-    "    x\n",
-    "        One or two-dimensional Tensor to search for element.\n",
-    "    element\n",
-    "        Item to search for inside x.\n",
-    "    dim\n",
-    "        Dimension of Tensor to collapse over.\n",
-    "\n",
-    "    Returns\n",
-    "    -------\n",
-    "    torch.Tensor\n",
-    "        Indices where element occurs in x. If element is not found,\n",
-    "        return length of x along dim. One dimension smaller than x.\n",
-    "\n",
-    "    Raises\n",
-    "    ------\n",
-    "    ValueError\n",
-    "        if x is not a 1 or 2 dimensional Tensor\n",
-    "\n",
-    "    Examples\n",
-    "    --------\n",
-    "    >>> first_appearance(torch.tensor([[1, 2, 3], [2, 3, 3], [1, 1, 1], [3, 1, 1]]), 3)\n",
-    "    tensor([2, 1, 3, 0])\n",
-    "    >>> first_appearance(torch.tensor([1, 2, 3]), 1, dim=0)\n",
-    "    tensor(0)\n",
-    "    \"\"\"\n",
-    "    if x.dim() > 2 or x.dim() == 0:\n",
-    "        raise ValueError(f\"only 1 or 2 dimensional Tensors allowed, got Tensor with dim {x.dim()}\")\n",
-    "    matches = x == element\n",
-    "    first_appearance_mask = (matches.cumsum(dim) == 1) & matches\n",
-    "    does_match, match_index = first_appearance_mask.max(dim)\n",
-    "    first_inds = torch.where(does_match, match_index, x.shape[dim])\n",
-    "    return first_inds"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "id": "26ff2314-b2df-408f-b83a-f5fc903145da",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "tensor([2, 3])"
-      ]
-     },
-     "execution_count": 2,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "first_appearance(torch.tensor([[1, 1, 3], [1, 1, 1]]), 3)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "id": "e8c9dd16-4917-40bc-8504-084035882ced",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "tensor([ True, False])"
-      ]
-     },
-     "execution_count": 5,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "torch.any(torch.isin(torch.tensor([[1, 1, 3], [1, 1, 1]]), 3), 1)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "id": "30eb1926-53d2-431a-b7c1-f95919887b84",
-   "metadata": {
-    "tags": []
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "tensor([2, 0])"
-      ]
-     },
-     "execution_count": 6,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "torch.tensor([[1, 1, 3], [1, 1, 1]]).argmax(dim=1)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "a02c61b3-c84f-4778-90d0-fe6aafa12ccc",
-   "metadata": {},
-   "outputs": [],
-   "source": []
   }
  ],
  "metadata": {