repo stringclasses 147
values | number int64 1 172k | title stringlengths 2 476 | body stringlengths 0 5k | url stringlengths 39 70 | state stringclasses 2
values | labels listlengths 0 9 | created_at timestamp[ns, tz=UTC]date 2017-01-18 18:50:08 2026-01-06 07:33:18 | updated_at timestamp[ns, tz=UTC]date 2017-01-18 19:20:07 2026-01-06 08:03:39 | comments int64 0 58 ⌀ | user stringlengths 2 28 |
|---|---|---|---|---|---|---|---|---|---|---|
pytorch/pytorch | 20,271 | Official instructions for how to build libtorch don't have same structure as prebuilt binaries | On Slack, Geoffrey Yu asked:
> Are there instructions for building libtorch from source? I feel like I'm missing something since I've tried building with `tools/build_libtorch.py`. However the build output doesn't seem to have the same structure as the prebuilt libtorch that you can download on pytorch.org
@pjh5 ... | https://github.com/pytorch/pytorch/issues/20271 | closed | [
"high priority",
"module: binaries",
"module: build",
"module: docs",
"module: cpp",
"triaged"
] | 2019-05-08T13:02:51Z | 2019-05-30T19:52:28Z | null | ezyang |
huggingface/transformers | 591 | What is the use of [SEP]? | Hello. I know that [CLS] means the start of a sentence and [SEP] makes BERT know the second sentence has begun. [SEP] can’t stop one sentence from extracting information from another sentence. However, I have a question.
If I have 2 sentences, which are s1 and s2., and our fine-tuning task is the same. In one way, I a... | https://github.com/huggingface/transformers/issues/591 | closed | [] | 2019-05-07T04:12:16Z | 2019-05-21T10:51:31Z | null | RomanShen |
pytorch/pytorch | 20,090 | How to add dynamically allocated strings to Pickler? | The following code prints `111` and `111`, instead of `222` and `111`, because `222` is skipped [here](https://github.com/pytorch/pytorch/blob/master/torch/csrc/jit/pickler.cpp#L68). Is this by design as Pickler only works for statically allocated strings? Or is there a way to correctly add dynamically allocated string... | https://github.com/pytorch/pytorch/issues/20090 | closed | [
"oncall: jit",
"triaged"
] | 2019-05-03T04:43:08Z | 2019-05-17T21:45:41Z | null | mrshenli |
huggingface/neuralcoref | 157 | Performance? | Hi there,
Thanks for the nice package!
Are there any performance comparisons with other systems? (say, Lee et el'18: https://arxiv.org/pdf/1804.05392.pdf).
| https://github.com/huggingface/neuralcoref/issues/157 | closed | [
"question",
"perf / accuracy"
] | 2019-04-30T21:38:56Z | 2019-10-16T08:48:09Z | null | danyaljj |
pytorch/examples | 554 | Where is the hook? | On the tutorial, I see it says this is an example of hook. So where is the hook? | https://github.com/pytorch/examples/issues/554 | closed | [] | 2019-04-30T11:38:40Z | 2019-05-27T21:00:53Z | null | yanbixing |
pytorch/pytorch | 19,908 | c++/pytorch How to convert tensor to image array? | ## ❓ Questions and Help
### Please note that this issue tracker is not a help form and this issue will be closed.
I would like to convert a tensor to image array and use tensor.data<short>() method. But it doesn't work.
My function is showed below:
```
#include <torch/script.h> // One-stop header.
#include... | https://github.com/pytorch/pytorch/issues/19908 | closed | [] | 2019-04-29T07:49:29Z | 2019-04-29T09:58:24Z | null | JingLiRaysightmed |
pytorch/pytorch | 19,822 | How to use torch.tensor(n) in Python3 to adapt to ’ at::TensorImpl‘ | ## ❓ Questions and Help
Hi ,when nms_cpp compiled by cpp_extension ,it didnt work but work in pytorch0.4.0.:
## TypeError: gpu_nms(): incompatible function arguments. The following argument types are supported:
1. (arg0: at::TensorImpl, arg1: at::TensorImpl, arg2: at::TensorImpl, arg3:
float) -> int
##... | https://github.com/pytorch/pytorch/issues/19822 | closed | [] | 2019-04-27T08:40:06Z | 2019-04-28T07:42:41Z | null | liuchanfeng165 |
pytorch/pytorch | 19,744 | How to select cl.exe for a config of cpp_extension? | ## ❓ Questions and Help
Hi, I got this error and dont wanna change vs15 again because the compiler keeps in OS,if I use cl.exe of VS2015 not VS2017 to compile my cpp_extension by setup.py and how to modify setup.py
 and how to obtain the predictive probability value? | Hello everyone
I've been using tensorflow before, but I met torch when I added functionality to a tool. My ultimate goal was to get the classification probability. On a binary classification problem, I used model. Eval () (inputs). numpy () to get the prediction results.
like this
2.19903 -2.06323
2.22841 -2.090... | https://github.com/pytorch/pytorch/issues/19611 | closed | [
"triaged"
] | 2019-04-23T09:29:27Z | 2019-04-23T19:26:02Z | null | xujiameng |
pytorch/pytorch | 19,561 | How to do prediction/inference for a batch of images at a time with libtorch? | Anybody knows how to do prediction/inference for a batch of images at a time with libtorch/pytorch C++?
any reply would be appreciated, thank you!
| https://github.com/pytorch/pytorch/issues/19561 | closed | [] | 2019-04-22T08:40:04Z | 2019-04-22T20:59:27Z | null | asa008 |
pytorch/examples | 547 | where can I get the inference code for classification? | I have trained the resnet-18 model for classification on my own dataset with examples/imagenet/main.py. And now I want to infernece the images, but there is no inference code. | https://github.com/pytorch/examples/issues/547 | closed | [] | 2019-04-22T03:20:25Z | 2019-04-24T10:50:50Z | 2 | ShaneYS |
pytorch/pytorch | 19,453 | How to load PyTorch model with LSTM using C++ api | ## 🐛 Bug
<!-- A clear and concise description of what the bug is. -->
## To Reproduce
Steps to reproduce the behavior:
1. Establish a PyTorch model with LSTM module using python, and store the script module after using torch.jit.trace. Python code like this:
```python
class MyModule(nn.Module):
de... | https://github.com/pytorch/pytorch/issues/19453 | open | [
"module: cpp",
"triaged"
] | 2019-04-19T02:11:06Z | 2024-07-24T20:52:00Z | null | SixerWang |
pytorch/tutorials | 484 | different results | HI, i copied to code exactly and and ran with all the appropriate downloads and im not achieving the resutls stated with 40 epochs. the loss stays at around 2.2 with test accuracy of 100/837 .. around 11%..
is there something i need to change to get over 50% accuracy ? | https://github.com/pytorch/tutorials/issues/484 | closed | [] | 2019-04-18T16:04:34Z | 2021-06-16T17:49:27Z | 1 | taylerpauls |
pytorch/examples | 544 | the error when I run the example for the imagenet | When I tried to run the model for the example/imagenet, I encounter such error.So could you tell me how to solve the problem?
python /home/zrz/code/imagenet_dist/examples-master/imagenet/main.py -a resnet18 -/home/zrz/dataset/imagenet/imagenet2012/ILSVRC2012/raw-data/imagenet-data
=> creating model 'resnet18'
... | https://github.com/pytorch/examples/issues/544 | closed | [] | 2019-04-14T06:22:38Z | 2022-03-10T05:56:43Z | 4 | runzeer |
pytorch/examples | 543 | [Important BUG] non-consistent behavior between "final evaluation" and "eval on each epoch" for mnist example | It is a common sense that, during evaluation, the model is not trained by the dev dataset.
However, I noticed a strange different behavior between the two results:
(1) train 10 epoch, having final evaluate on test data
(2) train 10 epoch, having an evaluation after each training epoch on test data
## Prior know... | https://github.com/pytorch/examples/issues/543 | open | [
"help wanted",
"nlp"
] | 2019-04-12T06:09:13Z | 2022-03-10T06:03:31Z | 1 | Jacob-Ma |
pytorch/examples | 542 | non-deterministic behavior on PyTorch mnist example | I tried PyTorch `examples/mnist/main.py` example to check if it is deterministic.
Although I modified the code to set the seed on everything, it still gives quite different results on GPU.
Do you know how to make the code be deterministic? Thank you very much.
Below is the code I have run and the output.
```... | https://github.com/pytorch/examples/issues/542 | closed | [] | 2019-04-12T04:31:25Z | 2019-04-12T05:53:14Z | 1 | Jacob-Ma |
pytorch/tutorials | 476 | Example with torch.empty in What is PyTorch? is misleading | In this [example](https://github.com/pytorch/tutorials/blob/master/beginner_source/blitz/tensor_tutorial.py) the output of a `torch.empty` call looks the same result I would obtain with `torch.zeros`, while it should be filled with garbage values.
This might be misleading to beginners. | https://github.com/pytorch/tutorials/issues/476 | closed | [] | 2019-04-11T09:19:06Z | 2019-08-23T21:18:53Z | null | alexchapeaux |
pytorch/pytorch | 19,098 | [C++ front end] how to use clamp to clip gradients? | ## ❓ Questions and Help
hi, I wonder if this could clip the gradients:
for(int i=0; i<net.parameters().size(); i++)
{
net.parameters().at(i).grad() = torch::clamp(net.parameters().at(i).grad(), -GRADIENT_CLIP, GRADIENT_CLIP);
}
optimizer.step();
I found it doesn't seem to work, and I still got la... | https://github.com/pytorch/pytorch/issues/19098 | closed | [] | 2019-04-10T05:37:24Z | 2019-04-10T05:38:01Z | null | ZhuXingJune |
pytorch/pytorch | 19,012 | I have a piece of code that is written in LUA and I want to know what is the pytorch equivalent of the code.?How do I implement these lines in pytorch .. Can somebody help me with it? The code is mentioned in the comment | ## ❓ Questions and Help
### Please note that this issue tracker is not a help form and this issue will be closed.
We have a set of [listed resources available on the website](https://pytorch.org/resources). Our primary means of support is our discussion forum:
- [Discussion Forum](https://discuss.pytorch.org/)... | https://github.com/pytorch/pytorch/issues/19012 | closed | [] | 2019-04-08T10:15:59Z | 2019-04-08T10:30:32Z | null | AshishRMenon |
pytorch/pytorch | 18,951 | Completed code with bug report for hdf5 dataset. How to fix? | Hello all, I want to report the issue of pytorch with hdf5 loader. The full source code and bug are provided
The problem is that I want to call the `test_dataloader.py` in two terminals. The file is used to load the custom hdf5 dataset (`custom_h5_loader`). To generate h5 files, you may need first run the file `conve... | https://github.com/pytorch/pytorch/issues/18951 | closed | [] | 2019-04-05T14:50:55Z | 2019-04-06T12:36:34Z | null | John1231983 |
pytorch/pytorch | 18,872 | How to convert a cudnn.BLSTM model to nn.LSTM bidirectional model | ## ❓ Questions and Help
### Please note that this issue tracker is not a help form and this issue will be closed.
We have a set of [listed resources available on the website](https://pytorch.org/resources). Our primary means of support is our discussion forum:
- [Discussion Forum](https://discuss.pytorch.org/)... | https://github.com/pytorch/pytorch/issues/18872 | closed | [] | 2019-04-04T18:30:20Z | 2019-04-04T19:00:04Z | null | rafikg |
pytorch/pytorch | 18,837 | How to use libtorch api torch::nn::parallel::data_parallel train on multi-gpu | ## 📚 Documentation
<!-- A clear and concise description of what content in https://pytorch.org/docs is an issue. If this has to do with the general https://pytorch.org website, please file an issue at https://github.com/pytorch/pytorch.github.io/issues/new/choose instead. If this has to do with https://pytorch.org/... | https://github.com/pytorch/pytorch/issues/18837 | closed | [
"module: performance",
"oncall: distributed",
"module: multi-gpu",
"module: docs",
"module: cpp",
"module: nn",
"triaged"
] | 2019-04-04T02:48:41Z | 2020-06-25T16:48:51Z | null | DDFlyInCode |
pytorch/tutorials | 468 | auxilary net confusion Inception_v3 Vs. GoogLeNet in finetune script? | Hi, I followed the finetune tutorial (but using this script to train from scratch): for `inception` as there is only one `aux_logit` below snippet working fine.
```
elif model_name == "inception":
""" Inception v3
Be careful, expects (299,299) sized images and has auxiliary output
""... | https://github.com/pytorch/tutorials/issues/468 | closed | [] | 2019-04-03T15:43:05Z | 2019-04-07T10:52:28Z | 0 | rajasekharponakala |
pytorch/pytorch | 18,781 | TORCH_CUDA_ARCH_LIST=All should know what is possible | ## 🐛 Bug
When setting TORCH_CUDA_ARCH_LIST=All, I expect Torch to compile with all CUDA architectures available to my current version of CUDA. Instead, it attempted to build for cuda 2.0.
See error:
nvcc fatal : Unsupported gpu architecture 'compute_20'
## To Reproduce
Steps to reproduce the behavior:
... | https://github.com/pytorch/pytorch/issues/18781 | closed | [
"module: build",
"module: docs",
"module: cuda",
"module: molly-guard",
"triaged"
] | 2019-04-03T00:31:35Z | 2024-08-04T05:06:56Z | null | xsacha |
pytorch/tutorials | 463 | The input of a GRU is of shape (seq_len, batch, input_size). I wonder does seq_len means anything? | I noticed that in the documentation of pytorch GRU, the input shape should be (seq_len, batch, input_size), thus the input ought to be a sequence and the model will deal with the sequence inside itself. But in this notebook, the author passes a tensor only of length one in each iteration in function `train`. I mean if ... | https://github.com/pytorch/tutorials/issues/463 | closed | [] | 2019-04-02T09:32:10Z | 2019-08-23T21:49:49Z | 1 | CSUN1997 |
pytorch/text | 522 | how to set random seed for BucketIterator to guarantee that it produce the same itrator every time you run the code ? | https://github.com/pytorch/text/issues/522 | closed | [
"obsolete"
] | 2019-04-02T01:47:03Z | 2022-01-25T04:04:42Z | null | zide05 | |
pytorch/pytorch | 18,677 | How to compile/install caffe2 with cuda 9.0? | I'm building caffe2 on ubuntu 18.04 with CUDA 9.0? But when I run "python setup.py install" command, I have met issue about version of CUDA. It needs to CUDA 9.2 instead of 9.0 but i only want to build with 9.0.
How to pass it?
Thank you! | https://github.com/pytorch/pytorch/issues/18677 | open | [
"caffe2"
] | 2019-04-01T06:54:50Z | 2019-05-27T12:43:18Z | null | TuanHAnhVN |
pytorch/tutorials | 459 | Inappropriate example code of vector-Jacobian product in (AUTOGRAD: AUTOMATIC DIFFERENTIATION) | I want to check the vector-Jacobian product. But in the example code the x is randomly generated and the the function relation between x and y is not clear. So how can I check if I correctly understand the vector-Jacobian product ? Can you improve it ? | https://github.com/pytorch/tutorials/issues/459 | closed | [] | 2019-03-30T13:57:54Z | 2021-06-16T20:24:11Z | 2 | guixianjin |
pytorch/tutorials | 458 | how can i remove the layers in model. | I want to use the googlenet model to do finetuing, remove the fc layer, and then add other layers. | https://github.com/pytorch/tutorials/issues/458 | closed | [] | 2019-03-30T08:28:08Z | 2021-06-16T20:25:08Z | 1 | wangjue-wzq |
pytorch/examples | 534 | when i run the main.py in imagenet,i encounter a problem | 
| https://github.com/pytorch/examples/issues/534 | open | [
"help wanted"
] | 2019-03-27T02:09:15Z | 2022-03-10T06:03:41Z | 0 | Xavier-cvpr |
pytorch/examples | 531 | Running examples/world_language_model | Hello, I am trying to run 'examples/world_language_model'.
However, when I do 'python main.py --cuda' in the above directory.
It prints an error like this.
Does anyone know how to solve this problem? | https://github.com/pytorch/examples/issues/531 | closed | [] | 2019-03-21T11:51:02Z | 2020-03-25T00:54:09Z | 1 | ohcurrent |
huggingface/transformers | 370 | What is Synthetic Self-Training? | The current best performing model on[ SQuAD 2.0](https://rajpurkar.github.io/SQuAD-explorer/) is BERT + N-Gram Masking + Synthetic Self-Training (ensemble):
What is Synthetic Self-Training?
| https://github.com/huggingface/transformers/issues/370 | closed | [
"Discussion",
"wontfix"
] | 2019-03-12T20:40:50Z | 2019-07-13T20:58:32Z | null | hsm207 |
pytorch/examples | 524 | [super_resolution]How can I get 'model_epoch_500.pth' file? | When I run:
`python super_resolve.py --input_image dataset/BSDS300/images/test/16077.jpg --model model_epoch_500.pth --output_filename out.png` .
It outpus :
` [Errno 2] No such file or directory: 'model_epoch_500.pth'`
How can I get 'model_epoch_500.pth' file? | https://github.com/pytorch/examples/issues/524 | closed | [] | 2019-03-08T11:08:17Z | 2019-05-31T08:57:22Z | 0 | dyfloveslife |
pytorch/pytorch | 17,654 | I want to know how to use the select(int64_t dim, int64_t index) in at::Tensor?What is the definition of a parameter ? | ## ❓ Questions and Help
### Please note that this issue tracker is not a help form and this issue will be closed.
We have a set of [listed resources available on the website](https://pytorch.org/resources). Our primary means of support is our discussion forum:
- [Discussion Forum](https://discuss.pytorch.org/)... | https://github.com/pytorch/pytorch/issues/17654 | closed | [] | 2019-03-04T12:57:04Z | 2019-03-04T17:05:24Z | null | SongyiGao |
pytorch/tutorials | 441 | Add 'Open in Colab' Button to Tutorial Code | Is it possible to edit the downloadable Jupyter notebooks at the bottom of each 60-minute blitz section? In a utopian scenario there would be a button at the top of each file, allowing the user to 'Open in Colab'. If this button was present the user would fix the code so that each cell was ran with the output visible.
... | https://github.com/pytorch/tutorials/issues/441 | closed | [] | 2019-03-04T12:26:51Z | 2019-03-31T10:34:34Z | 2 | pynchmeister |
pytorch/examples | 518 | How to train from scratch on custom model | Hi,
Thank you very much for the code.
I am new to pytorch (have worked a lot with tensorflow), and have a question which is probably basic, but I can't find the answer.
If I want to train ImageNet on a model which doesn't appear in the code list (under "model names"), but I do have the model as pth.tar file, how a... | https://github.com/pytorch/examples/issues/518 | closed | [] | 2019-02-26T10:19:24Z | 2022-03-10T05:28:08Z | 1 | jennyzu |
huggingface/transformers | 320 | what is the batch size we can use for SQUAD task? | I am running the squad example.
I have a Tesla M60 GPU which has about 8GB of memory. For bert-large-uncased model, I can only take batch size as 2, even after I used --fp16. Is it normal?
| https://github.com/huggingface/transformers/issues/320 | closed | [] | 2019-02-26T08:56:20Z | 2019-03-03T00:21:25Z | null | leonwyang |
pytorch/examples | 517 | How to save model in mnist.cpp? | How to save the model in cpp api mnist.cpp?
model.save and torch::save(model,"mnisttrain.pkl")
All error | https://github.com/pytorch/examples/issues/517 | open | [
"c++"
] | 2019-02-26T06:58:05Z | 2022-03-09T20:49:34Z | 5 | engineer1109 |
pytorch/tutorials | 437 | A short tutorial showing the input arguments for NLL loss/ cross entropy loss would be incredibly helpful | The arguments NLL loss (and by proxy cross entropy loss) take are in a relatively weird format. The documentation for the function does all it can within reason of the original documentation, but there's an incredible number of questions posted about weird problems giving them the kind of arguments. Much more than for ... | https://github.com/pytorch/tutorials/issues/437 | open | [] | 2019-02-26T05:35:50Z | 2019-02-26T05:35:50Z | 0 | jkterry1 |
pytorch/pytorch | 17,368 | What is a version/git-hash of nightly build? | ## 🚀 Feature
Nightly build does not include git-hash of pytorch.
So we can not know what the build is.
https://download.pytorch.org/libtorch/nightly/cpu/libtorch-shared-with-deps-latest.zip
I know the zip includes build_version which is like "1.0.0dev20190221".
Could you add git-hash of pytorch in build_ver... | https://github.com/pytorch/pytorch/issues/17368 | closed | [
"awaiting response (this tag is deprecated)"
] | 2019-02-21T19:54:41Z | 2019-02-28T22:34:49Z | null | junjihashimoto |
pytorch/ELF | 142 | what is meaning of outputs at verbose mode? | after I input quit at the df_console , it was still calulateing
```
D:\elfv2\play_opengo_v2\elf_gpu_full\elf>df_console --load d:/pretrained-go-19x19-v1.bin --num_block 20 --dim 224 --ver
bose
[2019-02-21 10:35:22.103] [elfgames::go::common::GoGameBase-12] [info] [0] Seed: 62127748, thread_id: 156604934899288204
... | https://github.com/pytorch/ELF/issues/142 | open | [] | 2019-02-21T02:51:58Z | 2019-02-21T02:51:58Z | null | l1t1 |
pytorch/examples | 514 | Why does discriminator's output change between batchsize:64 and batchsize:1 on inference. | I'm trying to get the discriminator output using train finished discriminator.
Procedure is below.
1. training dcgan.
2. preparing my image data and resize it 64 * 64.
3. load my image data using dataloader(same as training's one.)
4. I change only batch size at inference.
5. I got small Discriminator outputs(aft... | https://github.com/pytorch/examples/issues/514 | closed | [] | 2019-02-20T09:07:47Z | 2019-02-20T10:55:48Z | 2 | y-shirai-r |
pytorch/examples | 507 | Is there a plan to make the imagenet example in this repository support `fp16`? | Thanks! :) | https://github.com/pytorch/examples/issues/507 | closed | [] | 2019-02-14T19:38:58Z | 2022-03-10T03:12:22Z | 1 | deepakn94 |
pytorch/pytorch | 17,111 | where is the code for the implement of loss? | i want to find the implement of nn.BCEWithLogitsLoss, but it returns F functions , and i cannot find where F functions is , i want to modify the loss | https://github.com/pytorch/pytorch/issues/17111 | closed | [] | 2019-02-14T12:29:06Z | 2019-02-14T15:29:47Z | null | Jasperty |
pytorch/examples | 503 | I have some basic questions about training | Hi, I have some questions about training the model.
1. "neural_style.py train --dataset /Users/me/Downloads/examples-master/fast_neural_style/me0 --style-image /Users/met/Downloads/examples-master/fast_neural_style/images/content-images/amber.jpg --save-model-dir /Users/umit/Downloads/examples-master/fast_neural_s... | https://github.com/pytorch/examples/issues/503 | open | [
"onnx"
] | 2019-02-03T08:40:43Z | 2022-03-10T03:14:41Z | 0 | Umity |
pytorch/examples | 502 | neural_style.py: error: unrecognized arguments: --export_onnx | I am getting this error when i use:
python neural_style/neural_style.py train --dataset /Users/me/Downloads/examples-master/fast_neural_style/me0 --style-image /Users/me/Downloads/examples-master/fast_neural_style/images/content-images/amber.jpg --save-model-dir /Users/me/Downloads/examples-master/fast_neural_style/... | https://github.com/pytorch/examples/issues/502 | closed | [] | 2019-02-02T21:19:03Z | 2019-02-03T08:30:05Z | 0 | Umity |
pytorch/tutorials | 431 | cpp extension tutorial: not device agnostic? | Would the kerne calll in `lltm_cuda_forward` in the tutorial `tutorials/advanced_source/cpp_extension.rst` fail on multi gpu systems if the inputs are not on the default device, i.e., `device:0`?
To my understanding, some "magic" takes care of setting the right context if we add functionality do pytorch via custom... | https://github.com/pytorch/tutorials/issues/431 | open | [
"C++"
] | 2019-01-31T07:39:07Z | 2023-03-15T02:14:36Z | 1 | c-hofer |
pytorch/examples | 501 | why you set epoch to the sampler in the distributed example? | Hi,
Thanks for providing this helpful tutorial series. I am reading the part of training imagenet with distributed mode:
At [this line](https://github.com/pytorch/examples/blob/fe8abc3c810420df2856c6e668258f396b154cee/imagenet/main.py#L208), I do not understand the reason why shall I set epoch it the sampler. Wh... | https://github.com/pytorch/examples/issues/501 | closed | [] | 2019-01-30T03:44:09Z | 2023-01-05T08:01:47Z | 4 | CoinCheung |
huggingface/transformers | 233 | What is get_lr() meaning in the optimizer.py | I use a Model based on BertModel, and when I use the BertAdam the learning rate isn't changed. And when I use `get_lr()`, the return result is `[0]`. And I see the length of state isn't 0, but why I get that? | https://github.com/huggingface/transformers/issues/233 | closed | [] | 2019-01-28T13:19:06Z | 2019-02-05T16:12:33Z | null | kugwzk |
pytorch/pytorch | 16,439 | What is the difference between F.cross_entropy() and F.nll_loss() ?? | ## ❓ Questions and Help
### Please note that this issue tracker is not a help form and this issue will be closed.
We have a set of [listed resources available on the website](https://pytorch.org/resources). Our primary means of support is our discussion forum:
- [Discussion Forum](https://discuss.pytorch.org/)... | https://github.com/pytorch/pytorch/issues/16439 | closed | [] | 2019-01-28T11:06:23Z | 2019-01-28T13:25:31Z | null | lakshmiumenon |
pytorch/pytorch | 16,438 | What is the difference between F.cross_entropy() and F.nll_loss() ?? | ## ❓ Questions and Help
### Please note that this issue tracker is not a help form and this issue will be closed.
We have a set of [listed resources available on the website](https://pytorch.org/resources). Our primary means of support is our discussion forum:
- [Discussion Forum](https://discuss.pytorch.org/)... | https://github.com/pytorch/pytorch/issues/16438 | closed | [] | 2019-01-28T11:06:21Z | 2019-01-28T13:45:23Z | null | lakshmiumenon |
pytorch/examples | 500 | _pickle.UnpicklingError: invalid load key, '\xff'. | May I know how to fix the following error
```
mahmood@orca:fast_neural_style$ python3.7 neural_style/neural_style.py eval --content-image images/content-images/amber.jpg --model images/style-images/mosaic.jpg --output-image a1.jpg --cuda 1
Traceback (most recent call last):
File "neural_style/neural_style.py", ... | https://github.com/pytorch/examples/issues/500 | open | [
"bug",
"vision",
"pickle"
] | 2019-01-26T15:37:28Z | 2022-03-10T05:20:57Z | 0 | mahmoodn |
pytorch/examples | 499 | Crash in mnist example with num_workers > 0 | I'm getting a crash in the mnist example at the end of the 1st epoch when I run with any num_workers > 0 I'm running the python code in PyCharm debugger on a Ubuntu 16.04 system with PyTorch 1.0 with CUDA enabled.
raceback (most recent call last):
File "/snap/pycharm-community/108/helpers/pydev/pydevd.py", line... | https://github.com/pytorch/examples/issues/499 | open | [
"help wanted"
] | 2019-01-25T22:24:02Z | 2022-03-10T06:03:51Z | 0 | ankur6ue |
huggingface/transformers | 205 | What is the meaning of Attention Mask | Hi, I noticed that there is something called `Attention Mask` in the model.
In the annotation of class `BertForQuestionAnswering`,
```python
`attention_mask`: an optional torch.LongTensor of shape [batch_size, sequence_length] with indices
selected in [0, 1]. It's a mask to be used if the input seq... | https://github.com/huggingface/transformers/issues/205 | closed | [] | 2019-01-18T14:04:11Z | 2022-08-19T19:37:44Z | null | jianyucai |
huggingface/neuralcoref | 127 | Can't find mention type in doc class | I can't find the mention type of a span, so I just copy get_span_type function to get mention types as follows.
Maybe it could be merged into doc object.
```
ACCEPTED_ENTS = ["PERSON", "NORP", "FACILITY", "ORG", "GPE", "LOC", "PRODUCT", "EVENT", "WORK_OF_ART", "LANGUAGE"]
MENTION_TYPE = {"PRONOMINAL": 0, "NOMINAL"... | https://github.com/huggingface/neuralcoref/issues/127 | closed | [
"question",
"wontfix"
] | 2019-01-16T07:46:38Z | 2019-06-21T08:31:42Z | null | joe32140 |
pytorch/tutorials | 411 | ValueError low>=high in RandomCrop | Hello everyone,
First off, thanks for such detailed PyTorch tutorials! Recently, I was going through the data loading and processing tutorial [here](https://pytorch.org/tutorials/beginner/data_loading_tutorial.html). Maybe there's some misunderstanding from my side but in `beginner_source/data_loading_tutorial.py`, ... | https://github.com/pytorch/tutorials/issues/411 | closed | [] | 2019-01-11T23:54:39Z | 2019-09-12T04:24:36Z | 1 | Demfier |
pytorch/tutorials | 410 | TypeError : filename should be a str in beginner/nn_tutorial.py | I got a `TypeError` error in `beginner_source/nn_tutorial.py` while building tutorials with Python 3.5 on Ubuntu 16.04.
It seems like the type of parameter passed in `gzip.open()` in ["beginner_source/nn_tutorial.py(L64)"](https://github.com/pytorch/tutorials/blob/master/beginner_source/nn_tutorial.py#L64) should be... | https://github.com/pytorch/tutorials/issues/410 | closed | [] | 2019-01-11T06:37:53Z | 2019-02-08T20:26:28Z | 1 | 9bow |
pytorch/examples | 483 | different node has different parameters | I have tried it, but if I found that each model in all node has different gradients,so it results to different model among GPUs, At last I do like this:
#something to do###############
loss.backward()
self.average_gradients()
self.optimizer.step():
#other ... | https://github.com/pytorch/examples/issues/483 | closed | [] | 2018-12-24T11:54:04Z | 2018-12-30T03:53:49Z | 1 | YihengJiang |
pytorch/examples | 482 | Encountered IsADirectoryError at neural style eval | Hello, I am new in python and machine learning related field.
I caught the following error when trying to test a style from the examples:
```
/cygdrive/d/Downloaded Programs/git/examples/fast_neural_style
$ python neural_style/neural_style.py eval --content-image <images/content-images/amber.jpg> --model <save... | https://github.com/pytorch/examples/issues/482 | closed | [] | 2018-12-24T09:19:06Z | 2022-03-10T05:47:27Z | 1 | HosinLau |
pytorch/examples | 481 | the imagenet main when is use multi gpu(not set gpu args) then the input will not call input.cuda() why? | 
if i do't set args.gpu, the only target.cuda() call, why do this kind, but the code run success | https://github.com/pytorch/examples/issues/481 | closed | [] | 2018-12-24T08:42:29Z | 2018-12-30T03:45:44Z | 1 | mmxuan18 |
pytorch/tutorials | 400 | C++ Frontend Tutorial with GPU Support | I am following [this tutorial](https://pytorch.org/cppdocs/installing.html) on using PyTorch with C++ frontend. However, I would like to have a CUDA support, not a CPU only. I have also the `torch` package installed using `conda` but I guess it is not enough to compile C++ sources because I am getting the following err... | https://github.com/pytorch/tutorials/issues/400 | closed | [] | 2018-12-24T07:50:37Z | 2021-06-16T20:45:56Z | 0 | i-zaitsev |
pytorch/examples | 479 | Share dataloader in multi node multi gpus training with multiprocessing-distributed | In the example of [imagenet](https://github.com/pytorch/examples/blob/master/imagenet/main.py), `ngpus` process is created, so if I am training on 4 nodes with 4 gpus on each, there would be 16 processes in total.
Is there any way I could share the dataloader for the processes on the same node? Since I implemented a ... | https://github.com/pytorch/examples/issues/479 | closed | [] | 2018-12-19T13:41:36Z | 2020-09-11T12:57:16Z | 0 | xvjiarui |
pytorch/examples | 478 | in imagenet example why the val need to first resize to 256 and then crop 224, if the input is 299 how to set the resize input? | 
when the model is inceptionv3 the input size is 299, while others is 224. so the resize parameter counld set to what? and why in val stage there need to first resize to a bigger size then crop, some example... | https://github.com/pytorch/examples/issues/478 | closed | [] | 2018-12-19T11:48:52Z | 2019-03-27T18:01:58Z | 1 | mmxuan18 |
pytorch/examples | 476 | --resume fails after 1 epoch with Pytorch 1.0 release | Using --resume fails after 1 epoch with Pytorch 1.0 release with error below. I tried this with resnet50 and resnet18
```
Traceback (most recent call last):
File "main.py", line 398, in <module>
main()
File "main.py", line 110, in main
mp.spawn(main_worker, nprocs=ngpus_per_node, args=(ngpus_per_node... | https://github.com/pytorch/examples/issues/476 | open | [
"help wanted",
"vision"
] | 2018-12-17T17:04:15Z | 2022-03-10T06:04:01Z | 1 | mdflickner |
pytorch/examples | 473 | the sum of doc_topics is not equal 1 | Hi, I have a question. when I run lda.py, I find the sum of `doc_topics` is not equal 1. In fact, they are decreasing in the training process. Is there something wrong? | https://github.com/pytorch/examples/issues/473 | closed | [] | 2018-12-14T14:50:43Z | 2018-12-19T22:17:36Z | 2 | dongfeng951 |
pytorch/examples | 472 | How can I find a reference to understand the meaning of pro.sample? | what is the meaning of `pyro.sample( ........, infer={"enumerate": "parallel"})`? How can I find a reference to understand the meaning of `pro.sample`? I cannot find it.
Thanks! | https://github.com/pytorch/examples/issues/472 | closed | [] | 2018-12-14T13:33:42Z | 2018-12-15T04:02:46Z | 1 | dongfeng951 |
pytorch/examples | 471 | How to invoke GPU? | Hi, when I run the example codes such as vae.py, the GPU cannot be invoked automatically and therefore the training is very slow.
But the pytorch can invoke GPU automatically.
So how to invoke GPU?
Thank you! | https://github.com/pytorch/examples/issues/471 | closed | [] | 2018-12-14T08:14:46Z | 2018-12-15T04:18:49Z | 2 | dongfeng951 |
pytorch/examples | 470 | The Volatile GPU-Util is always 0, in examples/imagenet | I run the example of imagenet in https://github.com/pytorch/examples/tree/master/imagenet, althougt I can run it successfully, but it is slow, and the Volatile GPU-Util is always 0 with command 'nvidia-smi'
```
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.87 ... | https://github.com/pytorch/examples/issues/470 | open | [
"question"
] | 2018-12-13T06:49:15Z | 2022-03-10T16:46:35Z | 10 | wangxianrui |
huggingface/transformers | 114 | What is the best dataset structure for BERT? | First I want to say thanks for setting up all this!
I am using BertForSequenceClassification and am wondering what the optimal way is to structure my sequences.
Right now my sequences are blog post which could be upwards to 400 words long.
Would it be better to split my blog posts in sentences and use the se... | https://github.com/huggingface/transformers/issues/114 | closed | [] | 2018-12-11T16:28:00Z | 2018-12-11T20:57:45Z | null | wahlforss |
pytorch/pytorch | 14,889 | what is the algorithm theory of torch.nn.AdaptiveMaxPool2d? | what is the algorithm theory of torch.nn.AdaptiveMaxPool2d?
Is there any papers about torch.nn.AdaptiveMaxPool2d?
And how to find the c++ of implementing torch.nn.AdaptiveMaxPool2d in pytorch? | https://github.com/pytorch/pytorch/issues/14889 | closed | [] | 2018-12-07T10:16:15Z | 2018-12-07T15:46:19Z | null | zsf23 |
pytorch/pytorch | 14,850 | Document what is C10 | C10 seems to have an increasingly important role throughout the PyTorch code base (e.g., see #6325 or count the number of open issues containing "c10") yet I was unable to find a high-level description about it. There are only "rumors" to be found about C10, see for example [this post](https://discuss.pytorch.org/t/pyt... | https://github.com/pytorch/pytorch/issues/14850 | closed | [
"module: docs",
"triaged"
] | 2018-12-06T16:00:51Z | 2024-03-13T05:40:46Z | null | christoph-conrads |
pytorch/examples | 456 | How can I get the name of each image in the whole imagenet training process? | I want to obtain the name and the true label of each image, how can I modify the code to do that ? I find the data_loader just return the input tensor and the label without the image name. | https://github.com/pytorch/examples/issues/456 | closed | [] | 2018-11-30T06:40:48Z | 2018-11-30T06:54:39Z | 1 | lith0613 |
huggingface/neuralcoref | 113 | what is the different between en_coref models? | for three models, en_coref_lg, en_coref_md, en_coref_sm, which one has best performance? only consider the performance, is lg best? | https://github.com/huggingface/neuralcoref/issues/113 | closed | [] | 2018-11-30T06:36:43Z | 2019-04-11T12:14:11Z | null | Jasperty |
pytorch/pytorch | 14,460 | C++ API use model.pt in GPU . When I use lstm in model, there is what(): Expected object of backend CPU but got backend CUDA for argument #2 'mat2' (checked_tensor_unwrap at /pytorch/aten/src/ATen/Utils.h:70) | ## 🐛 Bug
<!-- A clear and concise description of what the bug is. -->
## To Reproduce
C++ code：
module->to(at::kCUDA);
auto gpu_tensor = img_var.to(at::kCUDA);
vector<torch::jit::IValue> inputs;
inputs.push_back(gpu_tensor);
auto out_tensor = module->forward(inputs).toTensor();
model：
# LSTM
... | https://github.com/pytorch/pytorch/issues/14460 | closed | [] | 2018-11-28T07:58:33Z | 2021-06-01T21:26:27Z | null | joy-yjl |
pytorch/pytorch | 14,456 | What is wrong with my model? It slows many times after switching from version 0.4.1 to 1.0 | This is the definition of my model:
```python
import torchvision
import torch
import torch.nn as nn
import torch.nn.functional as F
class Model(nn.Module):
def __init__(self, in_dim, out_dim, *args, **kwargs):
super(Model, self).__init__(*args, **kwargs)
vgg16 = torchvision.models.vg... | https://github.com/pytorch/pytorch/issues/14456 | closed | [
"module: performance"
] | 2018-11-28T06:44:26Z | 2019-06-09T02:44:25Z | null | CoinCheung |
pytorch/examples | 453 | Is the loss of the first word covered during the language model evaluation? | In the language model example, it seems that during the evaluation, the code starts from computing the loss of the second word. Thus, skipping the loss of the first word.
https://github.com/pytorch/examples/blob/537f6971872b839b36983ff40dafe688276fe6c3/word_language_model/main.py#L136
https://github.com/pytorch/exam... | https://github.com/pytorch/examples/issues/453 | open | [
"good first issue",
"nlp"
] | 2018-11-26T10:28:03Z | 2022-03-10T06:08:08Z | 0 | khassanoff |
pytorch/examples | 450 | How to use trained model to classifier pictures? | I have trained a best model by imagenet,but code repo has given does not have test option,so how can I use the model have trained to classifier pictures with labels? | https://github.com/pytorch/examples/issues/450 | open | [
"help wanted",
"vision"
] | 2018-11-25T03:27:13Z | 2022-03-10T06:07:49Z | 2 | mohhao |
pytorch/examples | 448 | which pytorch version can run the fast rcnn demo? | https://github.com/pytorch/examples/issues/448 | closed | [] | 2018-11-21T10:42:41Z | 2022-03-10T00:26:13Z | 2 | Bigwode | |
huggingface/neuralcoref | 110 | Doesn't work when span is merged. | ```python
nlp = spacy.load('en_coref_sm')
text = nlp("Michelle Obama is the wife of former U.S. President Barack Obama. Prior to her role as first lady, she was a lawyer.")
spans = list(text.noun_chunks)
for span in spans:
span.merge()
for word in text:
print(word)
if(word._.in_coref)... | https://github.com/huggingface/neuralcoref/issues/110 | closed | [
"question",
"wontfix"
] | 2018-11-21T10:12:43Z | 2019-06-17T14:22:21Z | null | lahsuk |
pytorch/examples | 443 | DCGAN: Generate more number of images | Is there a way we can generate an arbitrary number of images? Right now the fake sample is outputting to 64 images with default settings. My goal is to get 250 fake images. Is this possible? | https://github.com/pytorch/examples/issues/443 | closed | [] | 2018-11-16T04:49:33Z | 2018-11-16T04:50:14Z | 1 | MonojitBanerjee |
pytorch/pytorch | 13,460 | What is the net *.pb file encoding? | Hi there,
I am running the following code:
```python
with open(EXPORT_PATH + "mnist_init_net.pb", encoding="utf-8") as f:
init_net = f.read()
```
I get the following error:
```python
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xf1 in position 24: invalid continuation byte
```
It seems lik... | https://github.com/pytorch/pytorch/issues/13460 | closed | [] | 2018-11-01T18:26:20Z | 2018-11-07T16:07:03Z | null | Suhail |
pytorch/examples | 431 | How to run distributed training on multiple Node using ImageNet using ResNet model | The script mentioned in https://github.com/pytorch/examples/tree/master/imagenet does provides good guideline on single node training however it doesn't have good documentation on Distributed training on multiple Node.
I tried to use two machines with 8 gpus with below command
Machine-1 script
```
HOST_PORT="tc... | https://github.com/pytorch/examples/issues/431 | open | [
"distributed"
] | 2018-10-31T06:11:37Z | 2022-06-15T10:40:29Z | 12 | goswamig |
pytorch/examples | 430 | error in the backward pass while using the pytorch roi pooling | I am using [link](https://github.com/pytorch/examples/blob/d8d378c31d2766009db400ac03f41dd837a56c2a/fast_rcnn/roi_pooling.py#L38-L53) but i get error while doing the backward pass
```
File "/home/alireza/anaconda3/lib/python3.6/site-packages/spyder_kernels/customize/spydercustomize.py", line 668, in runfile
... | https://github.com/pytorch/examples/issues/430 | closed | [] | 2018-10-30T20:00:14Z | 2018-10-30T23:31:42Z | 2 | isalirezag |
pytorch/examples | 428 | how to deal with backward pass in pytorch version of ROI Pooling | I am trying to make position sensitive roi pooling (PSROIPooling)which is proposed in RFCN work.
PSROIPooling is basically ROIPooling + average pooling.
I am using the `roi_pooling.py` that is written in pytorch and provided [here](https://github.com/pytorch/examples/blob/d8d378c31d2766009db400ac03f41dd837a56c2a/fa... | https://github.com/pytorch/examples/issues/428 | closed | [] | 2018-10-30T02:08:40Z | 2018-10-30T02:21:14Z | 1 | isalirezag |
pytorch/examples | 425 | DCGAN: code and paper don't have the same feature maps? | ## From the code
Input(100\*1\*1) --->((ngf\*8) \*4\*4)--->((ngf\*4) \*8\*8)--->((ngf\*2) \*16\*16)--->(ngf \*32\*32)--->(3\*64\*64)
according to the code, **ngf=64**. Therefore we have
**Input(100\*1\*1) --->(512\*4\*4)--->(256\*8\*8)--->(128\*16\*16)--->(64\*32\*32)--->(3\*64\*64)**
```python
class Generator(n... | https://github.com/pytorch/examples/issues/425 | closed | [] | 2018-10-25T07:36:28Z | 2018-11-03T04:24:48Z | 1 | zhibo-liu |
pytorch/examples | 421 | what is spatial_scale in roi pooling | can you please explain to me what is spatial_scale here:
[Link](https://github.com/pytorch/examples/blob/d8d378c31d2766009db400ac03f41dd837a56c2a/fast_rcnn/roi_pooling.py#L38-L53)
also in ```[..., roi[2]:(roi[4]+1), roi[1]:(roi[3]+1)]```, what does `...` in the begining of the list do?
Thanks | https://github.com/pytorch/examples/issues/421 | closed | [] | 2018-10-11T14:39:24Z | 2018-10-11T19:06:28Z | null | isalirezag |
pytorch/text | 424 | What is "parse_field" for? | In torchtext.datasets.SNLI.splits, there is a parameter named "parse_field". I found if setting this field with a "datasets.snli.ShiftReduceField" object, the vocabulary becomes much smaller and SNLI accuracy always improves (compared with default value). It is amazing! But I can't find any description about it...
... | https://github.com/pytorch/text/issues/424 | closed | [] | 2018-09-27T04:45:10Z | 2018-10-02T03:39:58Z | null | jueliangguke |
pytorch/examples | 412 | RuntimeError: Found 0 images in subfolders of in AWS | Have anyone used torchvision.datasets.ImageFolder in AWS?
I met this error in predicting my pictures in my own folder
RuntimeError: Found 0 images in subfolders of: mine/1/2
Supported extensions are: .jpg,.jpeg,.png,.ppm,.bmp,.pgm,.tif
But, I have uploaded 3 jpg images in mine/1/2.
It also have found the fo... | https://github.com/pytorch/examples/issues/412 | closed | [] | 2018-09-13T11:23:47Z | 2021-09-07T03:11:15Z | 2 | Aaron4Fun |
pytorch/examples | 411 | AlexNet code | Where can I find the AlexNet code? I would like to implement it in a distributed mode using MPI. | https://github.com/pytorch/examples/issues/411 | closed | [] | 2018-09-11T13:08:24Z | 2022-03-10T00:27:48Z | 1 | abidmalikwaterloo |
pytorch/pytorch | 11,130 | where is the caffe2 folder? | Hi,
In the old version of caffe2, I could find the caffe2 folder ( "/usr/local/caffe2"). Where is the the caffe2 folder(within PYTORCH ) right now?
| https://github.com/pytorch/pytorch/issues/11130 | closed | [
"caffe2"
] | 2018-08-31T02:25:04Z | 2018-09-07T02:59:19Z | null | ddeeppnneett |
pytorch/examples | 409 | How to extract a trained model | Hi,
I have trained a model of resnet 152 using the code provided in 'examples/imagenet/main.py' I understand that it saves a checkpoint after every epoch, and at the end of the training it will save the best trained model.
My question is how can i extract this model?
| https://github.com/pytorch/examples/issues/409 | closed | [] | 2018-08-29T04:33:42Z | 2022-03-10T05:45:07Z | 3 | mvk07 |
pytorch/examples | 406 | UserWarning: nn.Upsampling is deprecated. Use nn.functional.interpolate instead. warnings.warn("nn.Upsampling is deprecated. Use nn.functional.interpolate instead.") | UserWarning: nn.Upsampling is deprecated. Use nn.functional.interpolate instead.
warnings.warn("nn.Upsampling is deprecated. Use nn.functional.interpolate instead.")
How can I solve this problem? | https://github.com/pytorch/examples/issues/406 | closed | [] | 2018-08-26T09:23:08Z | 2022-03-10T06:01:35Z | 1 | u0251077 |
pytorch/tutorials | 281 | Question: neural_style_tutorial：how to adjust different input image size to achieve this? | I'm new to dl and pytorch.
neural_style_tutorial
this tutorial is about fixed size image.
but most image cant have the same size defined,how to adjust different input image size to this model?
many thanks, if you can help!!! | https://github.com/pytorch/tutorials/issues/281 | closed | [] | 2018-08-10T06:02:04Z | 2021-06-16T21:11:13Z | 0 | aohan237 |
pytorch/examples | 399 | Why don't we use MSE as a reconstruction loss for VAE ? | Hi,
I am wondering if there is a theoretical reason for using BCE as a reconstruction loss for variation auto-encoders ? Can't we simply use MSE or norm-based reconstruction loss instead ?
Best Regards | https://github.com/pytorch/examples/issues/399 | open | [
"good first issue"
] | 2018-08-07T11:23:11Z | 2022-03-10T06:02:04Z | 7 | ahmed-fau |
pytorch/examples | 393 | How large batch size should I set for imagenet training | I just use the default setting of batch size 256 and 8 TiTAN XP gpus on resnet34， it takes about 1.5 hours for one epoch, I want to speed up the training process, Can I increase the batch size ? | https://github.com/pytorch/examples/issues/393 | closed | [] | 2018-07-26T02:51:18Z | 2018-07-27T04:04:14Z | 1 | lith0613 |
pytorch/examples | 384 | lm example：iteration over a 0-d tensor | I run example code , give me this error. I don't know how to solve this .
All error give from function repackage_hidden.
my pytorch version is .4 | https://github.com/pytorch/examples/issues/384 | closed | [] | 2018-07-13T10:53:35Z | 2019-06-24T11:15:58Z | 2 | EricAugust |
pytorch/pytorch | 9,207 | Where is the include and lib path for caffe2? | i installed pytorch with caffe2 from source by using 'python setup_caffe2.py install' command.
Can anyone tell that where is the default include and lib path for caffe2? | https://github.com/pytorch/pytorch/issues/9207 | open | [
"caffe2"
] | 2018-07-06T14:44:20Z | 2018-07-14T03:58:25Z | null | universewill |
huggingface/pytorch-openai-transformer-lm | 19 | what is the use of dropout in the Transformer? | https://github.com/huggingface/pytorch-openai-transformer-lm/blob/55ba4d78407ae12c7454dc8f3342f476be3dece5/model_pytorch.py#L161 | https://github.com/huggingface/pytorch-openai-transformer-lm/issues/19 | open | [] | 2018-07-05T16:18:48Z | 2018-07-09T13:59:41Z | null | teucer |
pytorch/examples | 376 | trying to understand the meaning of model.train() and model.eval() | Hi
So i see in the main.py we have model.train() and model.val(), i dont understand how to use them. can someone explain it to me please.
For example in here:
`python main.py -a resnet18 [imagenet-folder with train and val folders]` we did not specify train or eval, so how do we know which one to use.
I know my ... | https://github.com/pytorch/examples/issues/376 | closed | [] | 2018-06-23T22:14:02Z | 2018-06-23T22:18:09Z | 1 | isalirezag |
pytorch/examples | 374 | About distributed training of Imagenet, I am confused there is no operation to collect grads from machines and average them before update grads. | I write a distributed training model refer to the code imagenet/main.py , and the models on different machine own their independent optimizer. But I noticed that after backward() there is no operation to collect param grads from other processes and average them to get new grads for update. Does pytorch accomplish the ... | https://github.com/pytorch/examples/issues/374 | closed | [] | 2018-06-14T11:41:15Z | 2019-05-21T21:49:59Z | 4 | TobeyYang |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.