From 46f65b0016cd39c9ee5e71b512cb5595a358674c Mon Sep 17 00:00:00 2001 From: deep1 <> Date: Thu, 19 Oct 2023 15:11:23 +0800 Subject: [PATCH] no more deep :'( --- mjc_notes.md | 4 + ..._train_nanda_probe_w_counterfact_70%.ipynb | 1791 ++++++++++++++++- 2 files changed, 1784 insertions(+), 11 deletions(-) diff --git a/mjc_notes.md b/mjc_notes.md index a04daef..2596bfe 100644 --- a/mjc_notes.md +++ b/mjc_notes.md @@ -1785,3 +1785,7 @@ Ok so I got a center of mass intervention to the heads :) Now I just need to debug the dataloading script... nesxt week TODO: I need to split test train based on example_i, not row! +TODO: the plotted and metric acc don't line up + +maybe I need both head and mlp? +or just residual stream? diff --git a/notebooks/027_train_nanda_probe_w_counterfact_70%.ipynb b/notebooks/027_train_nanda_probe_w_counterfact_70%.ipynb index 225b793..f438fef 100644 --- a/notebooks/027_train_nanda_probe_w_counterfact_70%.ipynb +++ b/notebooks/027_train_nanda_probe_w_counterfact_70%.ipynb @@ -1085,9 +1085,34 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 55, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "text/plain": [ + "(array([ 3., 5., 10., 12., 53., 48., 13., 11., 6., 3.]),\n", + " array([-0.56379092, -0.4535192 , -0.34324747, -0.23297575, -0.12270403,\n", + " -0.01243231, 0.09783942, 0.20811114, 0.31838286, 0.42865458,\n", + " 0.5389263 ]),\n", + " )" + ] + }, + "execution_count": 55, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], "source": [] }, { @@ -1391,11 +1416,36 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 49, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "text/plain": [ + "array([[-1.17034912e-02, 4.29992676e-02, 1.35879517e-02, ...,\n", + " -3.09600830e-02, -4.75311279e-03, 4.70581055e-02],\n", + " [ 7.62939453e-02, 6.79492950e-04, -1.53686523e-01, ...,\n", + " -1.41113281e-01, -2.22534180e-01, -6.31103516e-02],\n", + " [-3.50646973e-02, 3.27392578e-01, -3.23181152e-02, ...,\n", + " -3.94287109e-01, -9.31549072e-03, -2.96142578e-01],\n", + " ...,\n", + " [ 5.80139160e-02, -8.45703125e-01, 4.98657227e-02, ...,\n", + " -3.53271484e-01, -4.83398438e-01, 3.98681641e-01],\n", + " [-1.47216797e-01, 1.85913086e-01, -9.24316406e-01, ...,\n", + " -3.83300781e-01, 1.16119385e-02, 2.29980469e-01],\n", + " [ 3.99169922e-01, 1.27075195e-01, -2.48535156e-01, ...,\n", + " -8.30566406e-01, -6.40625000e-01, 9.44213867e-02]], dtype=float32)" + ] + }, + "execution_count": 49, + "metadata": {}, + "output_type": "execute_result" + } + ], "source": [ - "# max_epochs= 10\n" + "# a = ds2['head_activation'][..., 0]\n", + "# b = ds2['head_activation'][..., 1]\n", + "dm.hs0[0]\n" ] }, { @@ -1403,6 +1453,41 @@ "execution_count": null, "metadata": {}, "outputs": [], + "source": [] + }, + { + "cell_type": "code", + "execution_count": 44, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "4 2\n", + "torch.Size([164, 12, 5120]) x\n", + "0\n", + "1\n" + ] + }, + { + "data": { + "text/plain": [ + "PLConvProbeLinear(\n", + " (probe): Sequential(\n", + " (0): BatchNorm1d(61440, eps=1e-05, momentum=0.1, affine=False, track_running_stats=True)\n", + " (1): Linear(in_features=61440, out_features=128, bias=True)\n", + " (2): ReLU()\n", + " (3): Linear(in_features=128, out_features=1, bias=True)\n", + " )\n", + ")" + ] + }, + "execution_count": 44, + "metadata": {}, + "output_type": "execute_result" + } + ], "source": [ "dl_train = dm.train_dataloader()\n", "dl_val = dm.val_dataloader()\n", @@ -1414,8 +1499,8 @@ "c_in = np.prod(x.shape[1:-1])\n", "net = PLConvProbeLinear(c_in=c_in, total_steps=max_epochs*len(dl_train), lr=lr, \n", " weight_decay=wd, \n", - " depth=2,\n", - " hs=128*4\n", + " depth=1,\n", + " hs=128\n", " # x_feats=x_feats\n", " )\n", "net\n" @@ -1423,9 +1508,1472 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 45, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Using bfloat16 Automatic Mixed Precision (AMP)\n", + "GPU available: True (cuda), used: True\n", + "TPU available: False, using: 0 TPU cores\n", + "IPU available: False, using: 0 IPUs\n", + "HPU available: False, using: 0 HPUs\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n", + "\n", + " | Name | Type | Params\n", + "-------------------------------------\n", + "0 | probe | Sequential | 7.9 M \n", + "-------------------------------------\n", + "7.9 M Trainable params\n", + "0 Non-trainable params\n", + "7.9 M Total params\n", + "31.458 Total estimated model params size (MB)\n" + ] + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "c6dd3a95148c4f6a8ff9b30b8a445145", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Sanity Checking: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "da557f0b3805480bab49c311f1db3675", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Training: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "ca68c829fbee4b9c8577f43d10c80ffd", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "79196746277042b2a13b122b65fadcb8", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "f15df65551ec4eeea9deff7441840e91", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "03d85b9bfa21443cbcc37fa216755128", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3d04b914a0264e5ca9e456fb2a5d9ce5", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a246e8bebed546de8c930177b7d90016", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "0cf1366124c44964b6f0998f70359596", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "793061533cd740898b9c04ecfda2bcc0", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "da031da0caf1464a815f493c7c058156", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "8933a8d9e8c2450b8a549ae9ff0faf91", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "12411458c0a54fae83596a93ec6f63fb", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3ed684fb42f640c29a26b2065bedd8c9", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "c670587c3eaf43be8deb54eb378e91c3", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "cea730dda9014e929d271baacbec0589", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "89a1d2802f0a4aed90d8ab99901ffb2a", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3e60bd754ce34d0993c28cac3810995b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3090805108124cdf8a35262dd261b131", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "30865d48cee444588eae65a3251b4d06", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "c9298ac6869f4d208c1b10cfb5c08403", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "43d838e9470c40ee95f30dcdb5a864af", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "29a250608d0d454babb4d66c39ecb943", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "0b30e6c4b35d46acb62295e64f57815c", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "235e9b9c86624938ad618c4dcd978906", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "e2c7b84f529b47b087515e37a1a85624", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "ce5431b4b7fd4059ae1c96e927404a1f", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "706b67912b3a4d9e8cf6ff74172bf390", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "5b70c1c01b894a90a80fd5da71300e3f", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "6b6e9d67d54e4b9490050d40cfecfddc", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a28d0b8318f143e4ba4496a74c38c0fd", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "cbc422ac92194fba9feedf85ed46585b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "50a40b78d61b420d88477db2557d344b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a3ea72896d0e49d7b23d4f331e2a27db", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "6083eb30887b497da094ac84a2ed0459", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "8a21cc6c21db429a874ddd4f8c49c5a4", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "88b2d20a24034426b7cde53a0433462a", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "070c24d50af749a89ac45a64708ec4b0", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "bf602404108c42e2a92fc361448e4c3b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "ed423e32b7e84430b9ab35105c19a522", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "082bcd9a9f8b4cbfad6888b20ad18bd8", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "2cb1e9aa17204258b9350a72dd8e5235", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a82715e5ffa942caaa3ab012bcfaacb5", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "c70c05f0322d43efad9b7a5416bd6b40", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "96f6e6d9d78a4539ae5dda5c48b1b0e4", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b1b63e254f9d4f74b882fed1743a4579", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "0005da41153f456e90579168c802363c", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "643fef2a14e84444a8f8753b772a758a", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b3d03eb6c77a45d9bb00987ce1ab6e99", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "6ffcfe112bef4df389f6a51840a86375", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "73f9625870d641f9a0b5520702d83a6b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "13f6aa90ba8e4b548bca815ea1c64f03", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "7ed012c6bc024505b0c84dc551db9087", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "87795d3e810744339f77669b67c0b1f7", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "8287e1a1b3314a739e934fd869c8cd3a", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "d38e803ffaa14e43bbc59c434c5a2164", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "bf6a1f19854a4a52b275ccd2376fb0ed", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3e5f4a9f9c2045999d6d624c873cfb7c", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a3aa1539236748b39d1ec73a1ec00cea", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "85954e7b9f0b4d87be6d24ea940af69f", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b0621581ad244184b8c0516910404015", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3043c2fcff314b54b7ba30c7493ce372", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "5a6081560a9544b19e90854aa59e70d9", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "5cfa7901489c48c1b595112310a33799", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "10948e3bde784e1c8361b1c412e6aabc", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "7394af6413bf46ef93a3c401e39cf6ba", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a089422574d04f7fad2a47c04770ca30", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "5acb42c9870d4debad9e782c946d5a6b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "3aff0d5825444368b0c447a93878a4d9", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b0e65916cd394cb18af2f5d87267d3bb", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "426a227dd3794cf490ce6c62fd61102c", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "c8e2657475d344aebae40b1869c74c43", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "c873683aaa03474fb3dddaf07ac8bf2e", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "bba01ff5f053464b9d92612af2d6733c", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "e87cc2491bf84a76848da35551062834", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "9a4d24b231a74ad1b875b968d23ef7fe", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "ecce325703da43b2b8403f0443960df2", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b1da3e84ce3842b090bf989e301a0550", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b83d4fdee35a4cf2824f110a001b1fc1", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "e657002582ae4f0586d76646893f3f33", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "f7621e06d9a24690b4949affdd48be0f", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "879282e5cb8747a4bb3625865d0ce8c3", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "6e670fbba496442d9e91ee4d4d4c5ee4", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "6ea90719333b48db859777ce6a14edde", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "6b0dff9e821d4573ad8ee4d0a60a12ce", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "afe19407694d4010b8dc8097c3d5f949", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "9d9704c417a54d6883789fe18bd748b6", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b1779e538fcf40cbb21728958f476b21", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "89fa645117d541219c420d4422b7360b", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "559fc5e9a1c74ef984340e607afec21f", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "51e88159cbfa48039edaf8b92359aa78", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "d39fd52907484b33a0d2732446bc53c1", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "8cc2134453fd4b868bdffc70270a3e44", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "cb7861c606134068b8e3bb8add87693c", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "813013fd6bd048c18d3219c36586c163", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "0e97275564024e6fb2c0902b6f0bc20f", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "f41fd5e4a9504e8d84e9a8255db33d57", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "43a4608fb9d64ba193e58527bcc4fd87", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "b49303ba92e4418ea03c671aeb47844e", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "25801e723abd4ec28d60eb435359e5a4", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "d48c4839f0114423971a814160aa5d66", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "5624ed0b474344e3a958d76d26be0f18", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Validation: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "`Trainer.fit` stopped: `max_epochs=100` reached.\n" + ] + } + ], "source": [ "trainer = pl.Trainer(precision=\"bf16-mixed\",\n", " gradient_clip_val=20,\n", @@ -1438,9 +2986,198 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 46, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n" + ] + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a0ff2f2d1f5d44f69be523672f37f478", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Testing: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/html": [ + "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
+       "┃   Runningstage.testing    ┃                           ┃                           ┃                           ┃\n",
+       "┃          metric           ┃       DataLoader 0        ┃       DataLoader 1        ┃       DataLoader 2        ┃\n",
+       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
+       "│         test/acc          │    0.5667275786399841     │    0.5620437860488892     │     0.543795645236969     │\n",
+       "│         test/loss         │     0.694634735584259     │    0.6583271026611328     │    0.7604641914367676     │\n",
+       "│          test/n           │           547.0           │           274.0           │           274.0           │\n",
+       "└───────────────────────────┴───────────────────────────┴───────────────────────────┴───────────────────────────┘\n",
+       "
\n" + ], + "text/plain": [ + "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓\n", + "┃\u001b[1m \u001b[0m\u001b[1m Runningstage.testing \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m┃\n", + "┃\u001b[1m \u001b[0m\u001b[1m metric \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 0 \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 1 \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m DataLoader 2 \u001b[0m\u001b[1m \u001b[0m┃\n", + "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩\n", + "│\u001b[36m \u001b[0m\u001b[36m test/acc \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.5667275786399841 \u001b[0m\u001b[35m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.5620437860488892 \u001b[0m\u001b[35m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.543795645236969 \u001b[0m\u001b[35m \u001b[0m│\n", + "│\u001b[36m \u001b[0m\u001b[36m test/loss \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.694634735584259 \u001b[0m\u001b[35m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.6583271026611328 \u001b[0m\u001b[35m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 0.7604641914367676 \u001b[0m\u001b[35m \u001b[0m│\n", + "│\u001b[36m \u001b[0m\u001b[36m test/n \u001b[0m\u001b[36m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 547.0 \u001b[0m\u001b[35m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 274.0 \u001b[0m\u001b[35m \u001b[0m│\u001b[35m \u001b[0m\u001b[35m 274.0 \u001b[0m\u001b[35m \u001b[0m│\n", + "└───────────────────────────┴───────────────────────────┴───────────────────────────┴───────────────────────────┘\n" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n" + ] + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "a0bdacbd4d134c09badaa5ac63465ba4", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Predicting: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n" + ] + }, + { + "data": { + "application/vnd.jupyter.widget-view+json": { + "model_id": "7ef9bedcef90449b9c0fe95f5c4ed190", + "version_major": 2, + "version_minor": 0 + }, + "text/plain": [ + "Predicting: 0it [00:00, ?it/s]" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "probe results on subsets of the data\n", + "acc=55.29%,\tn=548,\t[] \n", + "acc=54.31%,\tn=197,\t[instructed_to_lie==True] \n", + "acc=55.84%,\tn=351,\t[instructed_to_lie==False] \n", + "acc=56.24%,\tn=521,\t[llm_ans==label_true] \n", + "acc=54.50%,\tn=378,\t[llm_ans==label_instructed] \n", + "acc=37.04%,\tn=27,\t[instructed_to_lie==True & llm_ans==label_instructed] \n", + "acc=57.06%,\tn=170,\t[instructed_to_lie==True & llm_ans!=label_instructed] \n", + "probe accuracy for quadrants\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
llm gavediddidn't
instructed to
tell a truth0.56NaN
tell a lie0.370.57
\n", + "
" + ], + "text/plain": [ + "llm gave did didn't\n", + "instructed to \n", + "tell a truth 0.56 NaN\n", + "tell a lie 0.37 0.57" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "⭐PRIMARY METRIC⭐ acc=55.29% from probe\n", + "⭐SECONDARY METRIC⭐ acc_lie_lie=37.04% from probe\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], "source": [ "\n", "# look at hist\n", @@ -1464,6 +3201,38 @@ "rs['testval_metrics'] = rs['test']\n" ] }, + { + "cell_type": "code", + "execution_count": 53, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "epoch\n", + "0 0.681901\n", + "1 0.806216\n", + "2 0.745887\n", + "3 0.703839\n", + "4 0.522852\n", + " ... \n", + "95 0.848263\n", + "96 0.861060\n", + "97 0.861060\n", + "98 0.868373\n", + "99 0.895795\n", + "Name: train/acc, Length: 100, dtype: float64" + ] + }, + "execution_count": 53, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "df_hist['train/acc']\n" + ] + }, { "cell_type": "code", "execution_count": null,