{
    "info": {
        "author": "Xiang Gao",
        "author_email": "qasdfgtyuiop@gmail.com",
        "bugtrack_url": null,
        "classifiers": [],
        "description": "# TorchSnooper\n\nStatus:\n\n![PyPI](https://img.shields.io/pypi/v/TorchSnooper.svg)\n![PyPI - Downloads](https://img.shields.io/pypi/dm/TorchSnooper.svg)\n\nChecks:\n\n[![Build Status](https://zasdfgbnm.visualstudio.com/TorchSnooper/_apis/build/status/flake8?branchName=master)](https://zasdfgbnm.visualstudio.com/TorchSnooper/_build/latest?definitionId=12&branchName=master)\n[![Build Status](https://zasdfgbnm.visualstudio.com/TorchSnooper/_apis/build/status/test?branchName=master)](https://zasdfgbnm.visualstudio.com/TorchSnooper/_build/latest?definitionId=13&branchName=master)\n[![Build Status](https://zasdfgbnm.visualstudio.com/TorchSnooper/_apis/build/status/deploy-test-pypi?branchName=master)](https://zasdfgbnm.visualstudio.com/TorchSnooper/_build/latest?definitionId=19&branchName=master)\n\nDeploy (only succeed on tagged commits):\n\n[![Build Status](https://zasdfgbnm.visualstudio.com/TorchSnooper/_apis/build/status/deploy-pypi?branchName=master)](https://zasdfgbnm.visualstudio.com/TorchSnooper/_build/latest?definitionId=14&branchName=master)\n\nDo you want to look at the shape/dtype/etc. of every step of you model, but tired of manually writing prints?\n\nAre you bothered by errors like `RuntimeError: Expected object of scalar type Double but got scalar type Float`, and want to quickly figure out the problem?\n\nTorchSnooper is a [PySnooper](https://github.com/cool-RR/PySnooper) extension that helps you debugging these errors.\n\nTo use TorchSnooper, you just use it like using PySnooper. Remember to replace the `pysnooper.snoop` with `torchsnooper.snoop` in your code.\n\nTo install:\n\n```\npip install torchsnooper\n```\n\nTorchSnooper also support [snoop](https://github.com/alexmojaki/snoop). To use TorchSnooper with snoop, simply execute:\n```python\ntorchsnooper.register_snoop()\n```\nor\n```python\ntorchsnooper.register_snoop(verbose=True)\n```\nat the beginning, and use snoop normally.\n\n# Example 1: Monitoring device and dtype\n\nWe're writing a simple function:\n\n```python\ndef myfunc(mask, x):\n    y = torch.zeros(6)\n    y.masked_scatter_(mask, x)\n    return y\n```\n\nand use it like below\n\n```python\nmask = torch.tensor([0, 1, 0, 1, 1, 0], device='cuda')\nsource = torch.tensor([1.0, 2.0, 3.0], device='cuda')\ny = myfunc(mask, source)\n```\n\nThe above code seems to be correct, but unfortunately, we are getting the following error:\n\n```\nRuntimeError: Expected object of backend CPU but got backend CUDA for argument #2 'mask'\n```\n\nWhat is the problem? Let's snoop it! Decorate our function with `torchsnooper.snoop()`:\n\n```python\nimport torch\nimport torchsnooper\n\n@torchsnooper.snoop()\ndef myfunc(mask, x):\n    y = torch.zeros(6)\n    y.masked_scatter_(mask, x)\n    return y\n\nmask = torch.tensor([0, 1, 0, 1, 1, 0], device='cuda')\nsource = torch.tensor([1.0, 2.0, 3.0], device='cuda')\ny = myfunc(mask, source)\n```\n\nRun our script, and we will see:\n\n```\nStarting var:.. mask = tensor<(6,), int64, cuda:0>\nStarting var:.. x = tensor<(3,), float32, cuda:0>\n21:41:42.941668 call         5 def myfunc(mask, x):\n21:41:42.941834 line         6     y = torch.zeros(6)\nNew var:....... y = tensor<(6,), float32, cpu>\n21:41:42.943443 line         7     y.masked_scatter_(mask, x)\n21:41:42.944404 exception    7     y.masked_scatter_(mask, x)\n```\n\nNow pay attention to the devices of tensors, we notice\n```\nNew var:....... y = tensor<(6,), float32, cpu>\n```\n\nNow, it's clear that, the problem is because `y` is a tensor on CPU, that is,\nwe forget to specify the device on `y = torch.zeros(6)`. Changing it to\n`y = torch.zeros(6, device='cuda')`, this problem is solved.\n\nBut when running the script again we are getting another error:\n\n```\nRuntimeError: Expected object of scalar type Byte but got scalar type Long for argument #2 'mask'\n```\n\nLook at the trace above again, pay attention to the dtype of variables, we notice\n\n```\nStarting var:.. mask = tensor<(6,), int64, cuda:0>\n```\n\nOK, the problem is that, we didn't make the `mask` in the input a byte tensor. Changing the line into\n```\nmask = torch.tensor([0, 1, 0, 1, 1, 0], device='cuda', dtype=torch.uint8)\n```\nProblem solved.\n\n# Example 2: Monitoring shape\n\nWe are building a linear model\n\n```python\nclass Model(torch.nn.Module):\n\n    def __init__(self):\n        super().__init__()\n        self.layer = torch.nn.Linear(2, 1)\n\n    def forward(self, x):\n        return self.layer(x)\n```\n\nand we want to fit `y = x1 + 2 * x2 + 3`, so we create a dataset:\n\n```python\nx = torch.tensor([[0.0, 0.0], [0.0, 1.0], [1.0, 0.0], [1.0, 1.0]])\ny = torch.tensor([3.0, 5.0, 4.0, 6.0])\n```\n\nWe train our model on this dataset using SGD optimizer:\n\n```python\nmodel = Model()\noptimizer = torch.optim.SGD(model.parameters(), lr=0.1)\nfor _ in range(10):\n    optimizer.zero_grad()\n    pred = model(x)\n    squared_diff = (y - pred) ** 2\n    loss = squared_diff.mean()\n    print(loss.item())\n    loss.backward()\n    optimizer.step()\n```\n\nBut unfortunately, the loss does not go down to a low enough number.\n\nWhat's wrong? Let's snoop it! Putting the training loop inside snoop:\n\n```python\nwith torchsnooper.snoop():\n    for _ in range(100):\n        optimizer.zero_grad()\n        pred = model(x)\n        squared_diff = (y - pred) ** 2\n        loss = squared_diff.mean()\n        print(loss.item())\n        loss.backward()\n        optimizer.step()\n```\n\nPart of the trace looks like:\n\n```\nNew var:....... x = tensor<(4, 2), float32, cpu>\nNew var:....... y = tensor<(4,), float32, cpu>\nNew var:....... model = Model(  (layer): Linear(in_features=2, out_features=1, bias=True))\nNew var:....... optimizer = SGD (Parameter Group 0    dampening: 0    lr: 0....omentum: 0    nesterov: False    weight_decay: 0)\n22:27:01.024233 line        21     for _ in range(100):\nNew var:....... _ = 0\n22:27:01.024439 line        22         optimizer.zero_grad()\n22:27:01.024574 line        23         pred = model(x)\nNew var:....... pred = tensor<(4, 1), float32, cpu, grad>\n22:27:01.026442 line        24         squared_diff = (y - pred) ** 2\nNew var:....... squared_diff = tensor<(4, 4), float32, cpu, grad>\n22:27:01.027369 line        25         loss = squared_diff.mean()\nNew var:....... loss = tensor<(), float32, cpu, grad>\n22:27:01.027616 line        26         print(loss.item())\n22:27:01.027793 line        27         loss.backward()\n22:27:01.050189 line        28         optimizer.step()\n```\n\nWe notice that, `y` has shape `(4,)`, but `pred` has shape `(4, 1)`. As a result, `squared_diff` has shape `(4, 4)` due to broadcasting!\n\nThis is not the expected behavior, let's fix it: `pred = model(x).squeeze()`, now everything looks good:\n\n```\nNew var:....... x = tensor<(4, 2), float32, cpu>\nNew var:....... y = tensor<(4,), float32, cpu>\nNew var:....... model = Model(  (layer): Linear(in_features=2, out_features=1, bias=True))\nNew var:....... optimizer = SGD (Parameter Group 0    dampening: 0    lr: 0....omentum: 0    nesterov: False    weight_decay: 0)\n22:28:19.778089 line        21     for _ in range(100):\nNew var:....... _ = 0\n22:28:19.778293 line        22         optimizer.zero_grad()\n22:28:19.778436 line        23         pred = model(x).squeeze()\nNew var:....... pred = tensor<(4,), float32, cpu, grad>\n22:28:19.780250 line        24         squared_diff = (y - pred) ** 2\nNew var:....... squared_diff = tensor<(4,), float32, cpu, grad>\n22:28:19.781099 line        25         loss = squared_diff.mean()\nNew var:....... loss = tensor<(), float32, cpu, grad>\n22:28:19.781361 line        26         print(loss.item())\n22:28:19.781537 line        27         loss.backward()\n22:28:19.798983 line        28         optimizer.step()\n```\n\nAnd the final model converge to the desired values.\n\n# Example 3: Monitoring nan and inf\n\nLet's say we have a model that output the likelihood of something. For this example, we will just use a mock:\n\n```python\nclass MockModel(torch.nn.Module):\n\n    def __init__(self):\n        super(MockModel, self).__init__()\n        self.unused = torch.nn.Linear(6, 4)\n\n    def forward(self, x):\n        return torch.tensor([0.0, 0.25, 0.9, 0.75]) + self.unused(x) * 0.0\n\nmodel = MockModel()\n```\n\nDuring training, we want to minimize the negative log likelihood, we have code:\n\n```python\nfor epoch in range(100):\n    batch_input = torch.randn(6, 6)\n    likelihood = model(batch_input)\n    log_likelihood = likelihood.log()\n    target = -log_likelihood.mean()\n    print(target.item())\n\n    optimizer.zero_grad()\n    target.backward()\n    optimizer.step()\n```\n\nUnfortunately, we first get `inf` then `nan` for our target during training. What's wrong? Let's snoop it:\n\n```python\nwith torchsnooper.snoop():\n    for epoch in range(100):\n        batch_input = torch.randn(6, 6)\n        likelihood = model(batch_input)\n        log_likelihood = likelihood.log()\n        target = -log_likelihood.mean()\n        print(target.item())\n\n        optimizer.zero_grad()\n        target.backward()\n        optimizer.step()\n```\n\nWe will see the part of the output of the snoop looks like:\n\n```\n19:30:20.928316 line        18     for epoch in range(100):\nNew var:....... epoch = 0\n19:30:20.928575 line        19         batch_input = torch.randn(6, 6)\nNew var:....... batch_input = tensor<(6, 6), float32, cpu>\n19:30:20.929671 line        20         likelihood = model(batch_input)\nNew var:....... likelihood = tensor<(6, 4), float32, cpu, grad>\n19:30:20.930284 line        21         log_likelihood = likelihood.log()\nNew var:....... log_likelihood = tensor<(6, 4), float32, cpu, grad, has_inf>\n19:30:20.930672 line        22         target = -log_likelihood.mean()\nNew var:....... target = tensor<(), float32, cpu, grad, has_inf>\n19:30:20.931136 line        23         print(target.item())\n19:30:20.931508 line        25         optimizer.zero_grad()\n19:30:20.931871 line        26         target.backward()\ninf\n19:30:20.960028 line        27         optimizer.step()\n19:30:20.960673 line        18     for epoch in range(100):\nModified var:.. epoch = 1\n19:30:20.961043 line        19         batch_input = torch.randn(6, 6)\n19:30:20.961423 line        20         likelihood = model(batch_input)\nModified var:.. likelihood = tensor<(6, 4), float32, cpu, grad, has_nan>\n19:30:20.961910 line        21         log_likelihood = likelihood.log()\nModified var:.. log_likelihood = tensor<(6, 4), float32, cpu, grad, has_nan>\n19:30:20.962302 line        22         target = -log_likelihood.mean()\nModified var:.. target = tensor<(), float32, cpu, grad, has_nan>\n19:30:20.962715 line        23         print(target.item())\n19:30:20.963089 line        25         optimizer.zero_grad()\n19:30:20.963464 line        26         target.backward()\n19:30:20.964051 line        27         optimizer.step()\n```\n\nReading the output, we find that, at the first epoch (`epoch = 0`), the `log_likelihood` has `has_inf` flag.\nThe `has_inf` flag means, your tensor contains `inf` in its value. The same flag appears for `target`.\nAnd at the second epoch, starting from `likelihood`, tensors all have a `has_nan` flag.\n\nFrom our experience with deep learning, we would guess this is because the first epoch has `inf`, which causes\nthe gradient to be `nan`, and when parameters are updated, these `nan` propagate to parameters and causing all\nfuture steps to have `nan` result.\n\nTaking a deeper look, we figure out that the `likelihood` contains 0 in it, which leads to `log(0) = -inf`. Changing\nthe line\n```python\nlog_likelihood = likelihood.log()\n```\ninto\n```python\nlog_likelihood = likelihood.clamp(min=1e-8).log()\n```\nProblem solved.\n\n\n",
        "description_content_type": "text/markdown",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/zasdfgbnm/TorchSnooper",
        "keywords": "",
        "license": "",
        "maintainer": "",
        "maintainer_email": "",
        "name": "TorchSnooper",
        "package_url": "https://pypi.org/project/TorchSnooper/",
        "platform": "",
        "project_url": "https://pypi.org/project/TorchSnooper/",
        "project_urls": {
            "Homepage": "https://github.com/zasdfgbnm/TorchSnooper"
        },
        "release_url": "https://pypi.org/project/TorchSnooper/0.7.1/",
        "requires_dist": [
            "pysnooper (>=0.1.0)",
            "numpy"
        ],
        "requires_python": "",
        "summary": "Debug PyTorch code using PySnooper.",
        "version": "0.7.1"
    },
    "last_serial": 5801420,
    "releases": {
        "0.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "48474abdfb938be6bda97554d4d78c8f",
                    "sha256": "ea10791acc776b67e46916f169628565782a72a67aa97ee41f351262d27a2215"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "48474abdfb938be6bda97554d4d78c8f",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 2714,
                "upload_time": "2019-05-29T01:13:22",
                "url": "https://files.pythonhosted.org/packages/52/1a/f1bfff0832922eb66155eaa0b1a17e0d66565237a84ec3fa95c305068c68/TorchSnooper-0.1-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "39ca12fbd2c348260211605b1c878162",
                    "sha256": "6ff62a860e87c2b6aa908b8d9b058ef705e143712411e1799179bfb64ea1c1ce"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.1.tar.gz",
                "has_sig": false,
                "md5_digest": "39ca12fbd2c348260211605b1c878162",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 5483,
                "upload_time": "2019-05-29T01:14:00",
                "url": "https://files.pythonhosted.org/packages/18/ad/4ae99c9becaeddf55e5d9e48e85a42618c9b7235dd2ed7b18583bb7fdd96/TorchSnooper-0.1.tar.gz"
            }
        ],
        "0.1.3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "1c0daf04aef95fcd413622af5a44710d",
                    "sha256": "abb50d9d2948fd31eee6fca2b9152067831da91a465db7f5ebaa63aaf208883e"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.1.3.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "1c0daf04aef95fcd413622af5a44710d",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 4819,
                "upload_time": "2019-05-29T03:30:44",
                "url": "https://files.pythonhosted.org/packages/e5/e3/29b0e47c68cde3b8b7dc6dea0cff8cc3abd4a0776ce779f919b3c433ce90/TorchSnooper-0.1.3.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "30de607365ad183025aeb5cade8fc331",
                    "sha256": "071070ac0e6d2b24a04fb4452ba249dea1e8ef2f44bf73d4fff5200d55af0934"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.1.3-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "30de607365ad183025aeb5cade8fc331",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 2689,
                "upload_time": "2019-05-29T03:30:43",
                "url": "https://files.pythonhosted.org/packages/3d/14/26394429e9de04f316946ba4a6dc464744e7f700c5be0e26ef7b6145e46f/TorchSnooper-0.1.3-py3-none-any.whl"
            }
        ],
        "0.2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "6c9e7ac76eb38ee9eb116a629d72cfa8",
                    "sha256": "ec0f466a2e39feb7f4e159f95476acb1c753a6dacd61934a04032e047cad2c75"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.2.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "6c9e7ac76eb38ee9eb116a629d72cfa8",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 4836,
                "upload_time": "2019-05-30T06:32:46",
                "url": "https://files.pythonhosted.org/packages/8f/1e/a4d9c3739940b56b22c9ab1ce87c62d0a3281677771159edc311202c5134/TorchSnooper-0.2.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "534935f659f130f6b0778210dbe3b29e",
                    "sha256": "b8431c6e8b95da33fce2ca0f76603a2c5102e69c17760f5235e917311afb9e43"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.2-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "534935f659f130f6b0778210dbe3b29e",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 3490,
                "upload_time": "2019-05-30T06:32:45",
                "url": "https://files.pythonhosted.org/packages/e9/5c/302069d01b531162f0be29d258461c007392863b745a6bd57d3063062c0f/TorchSnooper-0.2-py3-none-any.whl"
            }
        ],
        "0.3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "42fe5867b8a7cb2b1511799e5c926432",
                    "sha256": "3ffb5c4507f24118de1a3ba51ae065652499fa65fbd0d55e47f73bd71f3d5453"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.3.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "42fe5867b8a7cb2b1511799e5c926432",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 4372,
                "upload_time": "2019-06-28T18:55:48",
                "url": "https://files.pythonhosted.org/packages/b0/51/af8d268c6cc0fbe14268cec35c1642ce5191b98a2e06cde2b73764c1d04d/TorchSnooper-0.3.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "0f11c877afcb2302dee19e5a88a06ad2",
                    "sha256": "0b1a945c6fb860f2165fd8f33694187e3ff34b2a676218674a6884d3f5f057ae"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.3-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "0f11c877afcb2302dee19e5a88a06ad2",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 3387,
                "upload_time": "2019-06-28T18:55:46",
                "url": "https://files.pythonhosted.org/packages/34/21/7aad0aa419b14a5b8ecd7c087876ae7d082fc53aa0afe780be781157e837/TorchSnooper-0.3-py3-none-any.whl"
            }
        ],
        "0.4": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "66477e95b6ec5cbb226906c043ee6701",
                    "sha256": "41059eaae11e620e648f4546fc998ff0713b72739ace381d364c148f8cd63d7f"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.4.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "66477e95b6ec5cbb226906c043ee6701",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 4611,
                "upload_time": "2019-07-01T23:08:17",
                "url": "https://files.pythonhosted.org/packages/c1/41/8546e55d19f02a9c709439170b1078201746e93251a5f5aa2e1f5f70f733/TorchSnooper-0.4.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "f965de504b352a6c371c8ad418ae1554",
                    "sha256": "3d876854140f04b337ea855da22d311aab03fbbe5437902839b9cf03adbfaefe"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.4-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "f965de504b352a6c371c8ad418ae1554",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 3475,
                "upload_time": "2019-07-01T23:08:15",
                "url": "https://files.pythonhosted.org/packages/cd/6d/7f14a6b3521e294e36aae70422f9c39b1001d81553214fe8a36d6ec319b0/TorchSnooper-0.4-py3-none-any.whl"
            }
        ],
        "0.4.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "596f855ea38106c26b8847e330121d26",
                    "sha256": "ca20917a03d2dca23c4474d28a635981ec5917d80d0d34d5e4b72327ab8d40f5"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.4.1.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "596f855ea38106c26b8847e330121d26",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 6817,
                "upload_time": "2019-07-02T00:07:10",
                "url": "https://files.pythonhosted.org/packages/b2/b5/754db71878deeb8bb591b3becaa19ae19df86a4e3efdaf26c684302373a3/TorchSnooper-0.4.1.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "47950d50797b57227e74ab677b8a5fdc",
                    "sha256": "696e54b09cbc95d07b7c040f206fb0812308a37528c0dd77c1edb5cdd4c12480"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.4.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "47950d50797b57227e74ab677b8a5fdc",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 5627,
                "upload_time": "2019-07-02T00:07:08",
                "url": "https://files.pythonhosted.org/packages/e1/62/bd535c0b1533be9e573cfd753a4b89c2e643c6716871798d44425d234885/TorchSnooper-0.4.1-py3-none-any.whl"
            }
        ],
        "0.5": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "cfb6a28d1c926b85315eb9be7b9852d7",
                    "sha256": "5d67cb3a4abda7bf5758d77ff4a3d7c5ecf951bf6aa03a64d321d351f8b6f05e"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.5.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "cfb6a28d1c926b85315eb9be7b9852d7",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 8084,
                "upload_time": "2019-07-05T23:49:30",
                "url": "https://files.pythonhosted.org/packages/fd/5e/5b348b3d35bf8cfa8ffac6c32b933900a00d6e4e07f62bdb598386f681b7/TorchSnooper-0.5.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "4f65a531a6ba536cb1c2b012f13edf37",
                    "sha256": "2f4261d222843084ec1757cd2bee80e29fd741d7b631e480bfa57350ccfc7b77"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.5-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "4f65a531a6ba536cb1c2b012f13edf37",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 6662,
                "upload_time": "2019-07-05T23:49:28",
                "url": "https://files.pythonhosted.org/packages/df/bb/147923a1274155034189a3a622ac99781009a15928101b7f47afd1892cdb/TorchSnooper-0.5-py3-none-any.whl"
            }
        ],
        "0.6": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "bbeb04f359669e5af2d12e73180de7e9",
                    "sha256": "1fe1468ffac899b4b53392d6a374d796745e986917f8a943dc445491b54d9e30"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.6.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "bbeb04f359669e5af2d12e73180de7e9",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 8925,
                "upload_time": "2019-07-10T01:57:31",
                "url": "https://files.pythonhosted.org/packages/e9/ac/5c3a244284d6db22d3a84929d4b7e9dd9f6ee48d0b9d705a3dfcdd712c21/TorchSnooper-0.6.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "8a3d63b832d63bde76ef33bd215aa385",
                    "sha256": "e136cb90420ce39cbd6527524046582286633cc1a270447e420a82ac0a2aa8a2"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.6-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "8a3d63b832d63bde76ef33bd215aa385",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 7004,
                "upload_time": "2019-07-10T01:57:30",
                "url": "https://files.pythonhosted.org/packages/79/98/4eaf43200a6a4ea31eae1163527db3cf0d2587f449a5cefa81931b4dcede/TorchSnooper-0.6-py3-none-any.whl"
            }
        ],
        "0.6.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "034522fa84e34fbd13b6af3e276bd28b",
                    "sha256": "48a4363745454ad051e228d1cfd6339fe1be9e9655d9601789846f36fa0bf516"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.6.1.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "034522fa84e34fbd13b6af3e276bd28b",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 8974,
                "upload_time": "2019-07-10T05:28:11",
                "url": "https://files.pythonhosted.org/packages/ed/9f/ce053efdb9ccb6890d7d87b2257ba10a72d7d19cae3e341f8704104a6d92/TorchSnooper-0.6.1.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "08957137dc3060f247e290549a3c183c",
                    "sha256": "161036f9190fcff7ccc2aac8816eec8d5d3a0c29e5e612ce4a6a86bf8f59fd11"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.6.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "08957137dc3060f247e290549a3c183c",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 7031,
                "upload_time": "2019-07-10T05:28:10",
                "url": "https://files.pythonhosted.org/packages/de/54/ea99bdc44a773cafc8cbd29c9c4d5fe98d84244cf5849e1a1aadb48aec14/TorchSnooper-0.6.1-py3-none-any.whl"
            }
        ],
        "0.7": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "2c5e2bad1a21fd5a6022295ddb4ca7fb",
                    "sha256": "a967b627b40f20019cd1e1517ed5aac7bb3154d591401e1976531d7d40987e86"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.7.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "2c5e2bad1a21fd5a6022295ddb4ca7fb",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 9210,
                "upload_time": "2019-08-08T18:58:40",
                "url": "https://files.pythonhosted.org/packages/56/ef/6904da7e8b7bb1d985b7c2aeb103a5763bd1f9d64416bc6d17d500ae3050/TorchSnooper-0.7.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "37993306713a0bb55fa92b89a72c2157",
                    "sha256": "beb31bae9fa65c923c48f2de877a489dba1e10bcbc65d0e8f18af5e4c898c5e0"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.7-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "37993306713a0bb55fa92b89a72c2157",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 7142,
                "upload_time": "2019-08-08T18:58:38",
                "url": "https://files.pythonhosted.org/packages/7b/67/1e30f88a689d423f46fb1752bc26001678bad506cc915bbea735ecdb962d/TorchSnooper-0.7-py3-none-any.whl"
            }
        ],
        "0.7.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "ac0020fcd65fe90195121c5a161f9744",
                    "sha256": "45290ee3e963e21961f188beab76f24a5137932bc085314715f1fe21ef9aa09d"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.7.1.linux-x86_64.tar.gz",
                "has_sig": false,
                "md5_digest": "ac0020fcd65fe90195121c5a161f9744",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 9214,
                "upload_time": "2019-09-09T02:58:37",
                "url": "https://files.pythonhosted.org/packages/fa/1b/429a26cee50ef7bd31bcdc8ecf5585ec6402dbd9eb64f3b8b8f24a853b25/TorchSnooper-0.7.1.linux-x86_64.tar.gz"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "52570036546acaaf55f3053fc7d4caae",
                    "sha256": "f362d3156e2130f7e46abfff2824321eff9270dab1e48a3fbbed1c86a8f97e76"
                },
                "downloads": -1,
                "filename": "TorchSnooper-0.7.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "52570036546acaaf55f3053fc7d4caae",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 7171,
                "upload_time": "2019-09-09T02:58:36",
                "url": "https://files.pythonhosted.org/packages/33/52/91ad263db398cf1fb5689c562bab82d35184501dc7ad62805d535fd27fa7/TorchSnooper-0.7.1-py3-none-any.whl"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "ac0020fcd65fe90195121c5a161f9744",
                "sha256": "45290ee3e963e21961f188beab76f24a5137932bc085314715f1fe21ef9aa09d"
            },
            "downloads": -1,
            "filename": "TorchSnooper-0.7.1.linux-x86_64.tar.gz",
            "has_sig": false,
            "md5_digest": "ac0020fcd65fe90195121c5a161f9744",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 9214,
            "upload_time": "2019-09-09T02:58:37",
            "url": "https://files.pythonhosted.org/packages/fa/1b/429a26cee50ef7bd31bcdc8ecf5585ec6402dbd9eb64f3b8b8f24a853b25/TorchSnooper-0.7.1.linux-x86_64.tar.gz"
        },
        {
            "comment_text": "",
            "digests": {
                "md5": "52570036546acaaf55f3053fc7d4caae",
                "sha256": "f362d3156e2130f7e46abfff2824321eff9270dab1e48a3fbbed1c86a8f97e76"
            },
            "downloads": -1,
            "filename": "TorchSnooper-0.7.1-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "52570036546acaaf55f3053fc7d4caae",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": null,
            "size": 7171,
            "upload_time": "2019-09-09T02:58:36",
            "url": "https://files.pythonhosted.org/packages/33/52/91ad263db398cf1fb5689c562bab82d35184501dc7ad62805d535fd27fa7/TorchSnooper-0.7.1-py3-none-any.whl"
        }
    ]
}