This commit is contained in:
Jakub Pokrywka 2022-06-01 10:00:34 +02:00
parent 55f0bea16b
commit 7e8223b1f1

View File

@ -37,6 +37,16 @@
"https://www.aclweb.org/anthology/W03-0419.pdf"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"conda env export -n <env-name> > environment.yml\n",
" \n",
" \n",
"conda env create -f path/to/environment.yml"
]
},
{
"cell_type": "code",
"execution_count": 1,
@ -56,7 +66,7 @@
},
{
"cell_type": "code",
"execution_count": 1,
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
@ -86,7 +96,7 @@
},
{
"cell_type": "code",
"execution_count": 2,
"execution_count": 3,
"metadata": {
"scrolled": false
},
@ -101,7 +111,7 @@
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "60fb8337cb5b4ab28969b9e1d60a851c",
"model_id": "960f4cf0de594e48ad7a84740cf301a3",
"version_major": 2,
"version_minor": 0
},
@ -119,7 +129,7 @@
},
{
"cell_type": "code",
"execution_count": 3,
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
@ -134,7 +144,7 @@
},
{
"cell_type": "code",
"execution_count": 4,
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
@ -143,7 +153,7 @@
},
{
"cell_type": "code",
"execution_count": 5,
"execution_count": 6,
"metadata": {},
"outputs": [
{
@ -152,7 +162,7 @@
"21"
]
},
"execution_count": 5,
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
@ -163,7 +173,7 @@
},
{
"cell_type": "code",
"execution_count": 6,
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
@ -173,7 +183,7 @@
},
{
"cell_type": "code",
"execution_count": 7,
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
@ -183,7 +193,7 @@
},
{
"cell_type": "code",
"execution_count": 8,
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
@ -192,7 +202,7 @@
},
{
"cell_type": "code",
"execution_count": 9,
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
@ -201,7 +211,7 @@
},
{
"cell_type": "code",
"execution_count": 10,
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
@ -210,7 +220,7 @@
},
{
"cell_type": "code",
"execution_count": 11,
"execution_count": 12,
"metadata": {
"scrolled": true
},
@ -221,7 +231,7 @@
},
{
"cell_type": "code",
"execution_count": 12,
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
@ -230,7 +240,7 @@
},
{
"cell_type": "code",
"execution_count": 13,
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
@ -239,7 +249,7 @@
},
{
"cell_type": "code",
"execution_count": 14,
"execution_count": 15,
"metadata": {},
"outputs": [
{
@ -248,7 +258,7 @@
"tensor([ 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 3])"
]
},
"execution_count": 14,
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
@ -259,7 +269,7 @@
},
{
"cell_type": "code",
"execution_count": 15,
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
@ -307,7 +317,7 @@
},
{
"cell_type": "code",
"execution_count": 16,
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
@ -316,7 +326,7 @@
},
{
"cell_type": "code",
"execution_count": 17,
"execution_count": 18,
"metadata": {},
"outputs": [],
"source": [
@ -336,7 +346,7 @@
},
{
"cell_type": "code",
"execution_count": 18,
"execution_count": 19,
"metadata": {},
"outputs": [],
"source": [
@ -345,7 +355,7 @@
},
{
"cell_type": "code",
"execution_count": 19,
"execution_count": 20,
"metadata": {
"scrolled": true
},
@ -356,7 +366,7 @@
},
{
"cell_type": "code",
"execution_count": 20,
"execution_count": 21,
"metadata": {},
"outputs": [],
"source": [
@ -367,7 +377,7 @@
},
{
"cell_type": "code",
"execution_count": 21,
"execution_count": 22,
"metadata": {},
"outputs": [],
"source": [
@ -389,7 +399,7 @@
},
{
"cell_type": "code",
"execution_count": 22,
"execution_count": 23,
"metadata": {},
"outputs": [],
"source": [
@ -406,7 +416,7 @@
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "fc9748f7f63c47fea274592f4dba2c73",
"model_id": "caddabb06f894f529ba7143ce62b8c2e",
"version_major": 2,
"version_minor": 0
},
@ -418,32 +428,11 @@
"output_type": "display_data"
},
{
"name": "stdout",
"name": "stderr",
"output_type": "stream",
"text": [
"> \u001b[0;32m/tmp/ipykernel_306568/4048919537.py\u001b[0m(12)\u001b[0;36m<cell line: 1>\u001b[0;34m()\u001b[0m\n",
"\u001b[0;32m 10 \u001b[0;31m \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0mcrf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0memissions\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mtags\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0m\u001b[0;32m 11 \u001b[0;31m \u001b[0;32mimport\u001b[0m \u001b[0mpdb\u001b[0m\u001b[0;34m;\u001b[0m \u001b[0mpdb\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mset_trace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0m\u001b[0;32m---> 12 \u001b[0;31m \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0m\u001b[0;32m 13 \u001b[0;31m \u001b[0moptimizer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0m\u001b[0;32m 14 \u001b[0;31m\u001b[0;34m\u001b[0m\u001b[0m\n",
"\u001b[0m\n",
"ipdb> batch_tokens\n",
"tensor([ 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 3])\n",
"ipdb> tags.shape\n",
"torch.Size([11, 1])\n",
"ipdb> tags\n",
"tensor([[0],\n",
" [3],\n",
" [0],\n",
" [7],\n",
" [0],\n",
" [0],\n",
" [0],\n",
" [7],\n",
" [0],\n",
" [0],\n",
" [0]])\n"
"/home/kuba/anaconda3/envs/zajeciaei/lib/python3.10/site-packages/torchcrf/__init__.py:249: UserWarning: where received a uint8 condition tensor. This behavior is deprecated and will be removed in a future version of PyTorch. Use a boolean condition instead. (Triggered internally at /opt/conda/conda-bld/pytorch_1646755897462/work/aten/src/ATen/native/TensorCompare.cpp:333.)\n",
" score = torch.where(mask[i].unsqueeze(1), next_score, score)\n"
]
}
],
@ -458,7 +447,7 @@
"\n",
" optimizer.zero_grad()\n",
" loss = -crf(emissions,tags)\n",
" import pdb; pdb.set_trace()\n",
" #import pdb; pdb.set_trace()\n",
" loss.backward()\n",
" optimizer.step()\n",
" \n",
@ -469,213 +458,40 @@
},
{
"cell_type": "code",
"execution_count": 26,
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['T_destination',\n",
" '__annotations__',\n",
" '__call__',\n",
" '__class__',\n",
" '__delattr__',\n",
" '__dict__',\n",
" '__dir__',\n",
" '__doc__',\n",
" '__eq__',\n",
" '__format__',\n",
" '__ge__',\n",
" '__getattr__',\n",
" '__getattribute__',\n",
" '__gt__',\n",
" '__hash__',\n",
" '__init__',\n",
" '__init_subclass__',\n",
" '__le__',\n",
" '__lt__',\n",
" '__module__',\n",
" '__ne__',\n",
" '__new__',\n",
" '__reduce__',\n",
" '__reduce_ex__',\n",
" '__repr__',\n",
" '__setattr__',\n",
" '__setstate__',\n",
" '__sizeof__',\n",
" '__str__',\n",
" '__subclasshook__',\n",
" '__weakref__',\n",
" '_apply',\n",
" '_backward_hooks',\n",
" '_buffers',\n",
" '_call_impl',\n",
" '_compute_normalizer',\n",
" '_compute_score',\n",
" '_forward_hooks',\n",
" '_forward_pre_hooks',\n",
" '_get_backward_hooks',\n",
" '_get_name',\n",
" '_is_full_backward_hook',\n",
" '_load_from_state_dict',\n",
" '_load_state_dict_pre_hooks',\n",
" '_maybe_warn_non_full_backward_hook',\n",
" '_modules',\n",
" '_named_members',\n",
" '_non_persistent_buffers_set',\n",
" '_parameters',\n",
" '_register_load_state_dict_pre_hook',\n",
" '_register_state_dict_hook',\n",
" '_replicate_for_data_parallel',\n",
" '_save_to_state_dict',\n",
" '_slow_forward',\n",
" '_state_dict_hooks',\n",
" '_validate',\n",
" '_version',\n",
" '_viterbi_decode',\n",
" 'add_module',\n",
" 'apply',\n",
" 'batch_first',\n",
" 'bfloat16',\n",
" 'buffers',\n",
" 'children',\n",
" 'cpu',\n",
" 'cuda',\n",
" 'decode',\n",
" 'double',\n",
" 'dump_patches',\n",
" 'end_transitions',\n",
" 'eval',\n",
" 'extra_repr',\n",
" 'float',\n",
" 'forward',\n",
" 'get_buffer',\n",
" 'get_extra_state',\n",
" 'get_parameter',\n",
" 'get_submodule',\n",
" 'half',\n",
" 'load_state_dict',\n",
" 'modules',\n",
" 'named_buffers',\n",
" 'named_children',\n",
" 'named_modules',\n",
" 'named_parameters',\n",
" 'num_tags',\n",
" 'parameters',\n",
" 'register_backward_hook',\n",
" 'register_buffer',\n",
" 'register_forward_hook',\n",
" 'register_forward_pre_hook',\n",
" 'register_full_backward_hook',\n",
" 'register_module',\n",
" 'register_parameter',\n",
" 'requires_grad_',\n",
" 'reset_parameters',\n",
" 'set_extra_state',\n",
" 'share_memory',\n",
" 'start_transitions',\n",
" 'state_dict',\n",
" 'to',\n",
" 'to_empty',\n",
" 'train',\n",
" 'training',\n",
" 'transitions',\n",
" 'type',\n",
" 'xpu',\n",
" 'zero_grad']"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"dir(crf)\n"
]
},
{
"cell_type": "code",
"execution_count": 34,
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Parameter containing:\n",
"tensor([[ 0.1427, 0.0082, -0.0852, -0.0714, -0.0514, 0.0753, 0.0389, 0.0018,\n",
" -0.0806],\n",
" [-0.0809, -0.0508, 0.0520, -0.0619, 0.0181, -0.0729, -0.1430, -0.1055,\n",
" 0.0384],\n",
" [-0.0011, -0.1476, 0.0425, -0.0081, -0.1181, -0.0098, -0.0567, 0.0311,\n",
" -0.0696],\n",
" [-0.0443, -0.0741, 0.0463, -0.0967, -0.0403, -0.0243, 0.0098, -0.0063,\n",
" -0.0811],\n",
" [ 0.0632, -0.1175, -0.0992, 0.0198, 0.0310, -0.0059, 0.0191, -0.1303,\n",
" -0.1423],\n",
" [ 0.0029, 0.0296, 0.0152, -0.0418, -0.1068, -0.0920, -0.0380, 0.0461,\n",
" 0.0167],\n",
" [-0.1167, -0.0559, -0.0428, -0.0115, -0.1006, -0.1511, 0.0035, -0.0273,\n",
" -0.1201],\n",
" [-0.0378, 0.0481, -0.1474, -0.0154, 0.0347, -0.0392, -0.0755, -0.1227,\n",
" 0.0448],\n",
" [-0.0383, -0.0402, 0.0054, 0.0145, -0.1353, -0.0460, 0.0257, -0.0322,\n",
" 0.0023]], requires_grad=True)"
]
},
"execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"crf.transitions"
]
},
{
"cell_type": "code",
"execution_count": 31,
"execution_count": null,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"[Parameter containing:\n",
" tensor([-0.0432, -0.1150, -0.1045, -0.0779, -0.0858, 0.0287, -0.1437, -0.1446,\n",
" 0.0335], requires_grad=True),\n",
" Parameter containing:\n",
" tensor([ 0.0838, -0.0097, -0.1136, 0.0010, -0.1177, 0.0225, 0.0292, -0.0837,\n",
" -0.1063], requires_grad=True),\n",
" Parameter containing:\n",
" tensor([[ 0.1427, 0.0082, -0.0852, -0.0714, -0.0514, 0.0753, 0.0389, 0.0018,\n",
" -0.0806],\n",
" [-0.0809, -0.0508, 0.0520, -0.0619, 0.0181, -0.0729, -0.1430, -0.1055,\n",
" 0.0384],\n",
" [-0.0011, -0.1476, 0.0425, -0.0081, -0.1181, -0.0098, -0.0567, 0.0311,\n",
" -0.0696],\n",
" [-0.0443, -0.0741, 0.0463, -0.0967, -0.0403, -0.0243, 0.0098, -0.0063,\n",
" -0.0811],\n",
" [ 0.0632, -0.1175, -0.0992, 0.0198, 0.0310, -0.0059, 0.0191, -0.1303,\n",
" -0.1423],\n",
" [ 0.0029, 0.0296, 0.0152, -0.0418, -0.1068, -0.0920, -0.0380, 0.0461,\n",
" 0.0167],\n",
" [-0.1167, -0.0559, -0.0428, -0.0115, -0.1006, -0.1511, 0.0035, -0.0273,\n",
" -0.1201],\n",
" [-0.0378, 0.0481, -0.1474, -0.0154, 0.0347, -0.0392, -0.0755, -0.1227,\n",
" 0.0448],\n",
" [-0.0383, -0.0402, 0.0054, 0.0145, -0.1353, -0.0460, 0.0257, -0.0322,\n",
" 0.0023]], requires_grad=True)]"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"list(crf.parameters())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"eval_model(validation_tokens_ids, validation_labels)"
]
},
{
"cell_type": "code",
"execution_count": null,
@ -688,9 +504,7 @@
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"eval_model(validation_tokens_ids, validation_labels)"
]
"source": []
},
{
"cell_type": "code",