facebookresearch/fvcore issues and pull requests

#150 - RuntimeError: Detected that you are using FX to torch.jit.trace a dynamo-optimized function. This is not supported at the moment.

Issue - State: open - Opened by arghavan-kpm 2 months ago

#150 - RuntimeError: Detected that you are using FX to torch.jit.trace a dynamo-optimized function. This is not supported at the moment.

Issue - State: open - Opened by arghavan-kpm 2 months ago

#149 - ValueError: You have to specify either decoder_input_ids or decoder_inputs_embeds

Issue - State: open - Opened by Aafiya-H 4 months ago

#148 - Unsupported operator aten::max_pool2d encountered 5 time(s)

Issue - State: open - Opened by Drglitch791 5 months ago

#147 - Counting FLOPS for a custom op with set_op_handle: a toy example that doesn't work.

Issue - State: open - Opened by guynich 6 months ago

#146 - Delete .circleci directory

Pull Request - State: closed - Opened by bigfootjon 7 months ago - 2 comments
Labels: CLA Signed, Merged

#145 - Conflict with torch.distributed?

Issue - State: open - Opened by Wuzimeng 9 months ago - 1 comment

#144 - Update test_flop_count.py

Pull Request - State: open - Opened by anonymousdouble 9 months ago
Labels: CLA Signed

#143 - Unsupported operators

Issue - State: open - Opened by qsimeon 10 months ago

#142 - support div mul add

Issue - State: open - Opened by yokosyun 10 months ago

#141 - iopath dependency

Issue - State: open - Opened by BagelOrb 11 months ago

#140 - creating an example of functionality in the README>=.md

Pull Request - State: open - Opened by Kushal0217 about 1 year ago - 1 comment

#139 - Quantized model size

Issue - State: open - Opened by prinshul about 1 year ago

#138 - Unsupported operator aten::add_

Issue - State: open - Opened by peterzheng98 about 1 year ago

#137 - NotImplementedError during FlopCountAnalysis if submodule decorated by torch.compile

Issue - State: open - Opened by flishwang about 1 year ago

#136 - Sort state_dict.keys() in _strip_prefix_if_present?

Issue - State: open - Opened by isdanni about 1 year ago

#135 - Unsupport the high dimension inputs that larger than 5.

Issue - State: open - Opened by hailong23-jin about 1 year ago

#134 - Updated Github & CircleCI actions

Pull Request - State: open - Opened by johnnynunez over 1 year ago - 3 comments
Labels: CLA Signed

#133 - Clarification in the nomenclature of FLOPs / MACs

Issue - State: closed - Opened by Vinayaka2k over 1 year ago - 1 comment

#132 - Feed two batches and then dividing the ended FLOPs by 2 to avoid the batch_norm error?

Issue - State: open - Opened by bakalaugh over 1 year ago

#131 - Fix jit analysis

Pull Request - State: closed - Opened by wushirong over 1 year ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#130 - Flop counter for matmul does not support matrix-vector product.

Issue - State: open - Opened by AllenYolk over 1 year ago

#129 - Error while calculating flops of transformer model.

Issue - State: closed - Opened by NeelKanwal over 1 year ago - 9 comments

#128 - support loading checkpoint from non-local file directly.

Pull Request - State: closed - Opened by ppwwyyxx over 1 year ago - 5 comments
Labels: CLA Signed, Merged

#127 - Compatible with numpy 1.24

Pull Request - State: closed - Opened by ppwwyyxx over 1 year ago - 2 comments
Labels: CLA Signed, Merged

#126 - Why is the sigmoid function applied to logits before calling F.binary_cross_entropy_with_logits?

Issue - State: closed - Opened by kirill-menke over 1 year ago

#125 - This error occurred when I calculated flops. What is the reason?

Issue - State: open - Opened by fxooooooooo almost 2 years ago - 1 comment

#124 - Formula for batch normalization

Issue - State: open - Opened by unipdjahanifar almost 2 years ago

#123 - Fix Shapely deprecation

Pull Request - State: closed - Opened by normster almost 2 years ago - 3 comments
Labels: CLA Signed, Merged

#122 - Update deprecated Ubuntu image to v20.

Pull Request - State: closed - Opened by ericmintun almost 2 years ago - 2 comments
Labels: CLA Signed

#121 - DeformConv2d not supported in FLOPs counting?

Issue - State: open - Opened by kenmbkr almost 2 years ago - 2 comments

#120 - add flop counter for mul operation

Pull Request - State: closed - Opened by jixiege almost 2 years ago - 2 comments
Labels: CLA Signed

#119 - Does a single multiply op count one flop?

Issue - State: closed - Opened by jixiege almost 2 years ago - 2 comments

#118 - support specifying matching_heuristics from filename

Pull Request - State: closed - Opened by wat3rBro almost 2 years ago - 9 comments
Labels: CLA Signed, Merged, fb-exported

#117 - baddbmm not supported

Issue - State: open - Opened by rwightman almost 2 years ago

#116 - fix shapely deprecation warning

Pull Request - State: open - Opened by jveitchmichaelis almost 2 years ago - 2 comments
Labels: CLA Signed

#115 - Memory leak?

Issue - State: open - Opened by josecannete about 2 years ago - 1 comment

#114 - don't use fp16 in giou loss

Pull Request - State: open - Opened by lbin about 2 years ago
Labels: CLA Signed

#113 - don't use fp16 in giou loss

Pull Request - State: closed - Opened by lbin about 2 years ago
Labels: CLA Signed

#112 - Meaning of Class <ActivationCountAnalysis>

Issue - State: open - Opened by go-ahead-maker about 2 years ago

#111 - Change ConstantParamScheduler to exclude 1

Pull Request - State: open - Opened by xvjiarui about 2 years ago - 3 comments
Labels: CLA Signed

#110 - Clipping only applied to image arrays that arrive in uint8 type

Issue - State: open - Opened by Trolomoll over 2 years ago

#109 - Wrong FLOPs count if some submodules of the model are wrapped by DataParallel

Issue - State: open - Opened by luowyang over 2 years ago

#108 - Matmul Flops

Issue - State: open - Opened by breuera over 2 years ago - 2 comments
Labels: question

#107 - docs: add social GH button in support of Ukraine

Pull Request - State: closed - Opened by dmitryvinn over 2 years ago - 1 comment
Labels: CLA Signed

#106 - Fix circleci gpu resource name

Pull Request - State: closed - Opened by ppwwyyxx over 2 years ago - 2 comments
Labels: CLA Signed

#105 - The flopcounter does not support LeakyReLU?

Issue - State: open - Opened by Windaway over 2 years ago - 2 comments
Labels: potential-bug

#104 - ValueError: Invalid type <class 'numpy.int32'> for the flop count! Please use a wider type to avoid overflow.

Issue - State: open - Opened by WYHZQ over 2 years ago - 5 comments
Labels: potential-bug

#103 - ShapelyDeprecationWarning in CropTransform

Issue - State: open - Opened by bowenc0221 over 2 years ago
Labels: question

#102 - Counting bias term

Issue - State: closed - Opened by noamgot over 2 years ago - 2 comments

#101 - fix pypi package release

Pull Request - State: closed - Opened by ppwwyyxx over 2 years ago - 4 comments
Labels: CLA Signed

#100 - Why not return the variable `save_file` in Line 184 directly?

Issue - State: closed - Opened by DapengFeng almost 3 years ago

#99 - add lstm layer support

Pull Request - State: open - Opened by HendrikKlug-synthara almost 3 years ago - 2 comments

#98 - flop count analysis of LSTM layers

Issue - State: open - Opened by HendrikKlug-synthara almost 3 years ago - 2 comments

#97 - Add device parameter to update_bn_stats() in precise_bn.py

Issue - State: open - Opened by hankyul2 almost 3 years ago

#96 - flop count: capture output tensors from all layers so that unused layers are correctly counted.

Pull Request - State: open - Opened by ppwwyyxx almost 3 years ago - 1 comment
Labels: CLA Signed, fb-exported

#95 - AttributeError: module 'fvcore' has no attribute 'nn'

Issue - State: open - Opened by yongjie-lv almost 3 years ago - 2 comments
Labels: installation

#94 - New import breaks compatibility with python 3.6/7

Issue - State: closed - Opened by radao almost 3 years ago

#93 - The number of FLOPs turned out to be a negative number

Issue - State: closed - Opened by Asagami-Fujino almost 3 years ago - 5 comments

#92 - can't calculate the flops of loop in def forward

Issue - State: open - Opened by wwHwo almost 3 years ago - 3 comments

#91 - GitHub Actions: Test on current versions

Pull Request - State: closed - Opened by cclauss almost 3 years ago
Labels: CLA Signed

#90 - support multiple files as _BASE_ in CfgNode

Pull Request - State: closed - Opened by wat3rBro almost 3 years ago - 3 comments
Labels: CLA Signed, Merged, fb-exported

#89 - Getting:- ImportError: cannot import name 'FakeQuantizeBase' from 'torch.quantization'

Issue - State: closed - Opened by patrickn699 almost 3 years ago - 5 comments
Labels: duplicate

#88 - fixing import bug and gating by torch version

Pull Request - State: closed - Opened by HDCharles almost 3 years ago - 4 comments
Labels: CLA Signed, Merged, fb-exported

#87 - BUG: ImportError: cannot import name 'FakeQuantizeBase' from 'torch.quantization'

Issue - State: closed - Opened by hasansalimkanmaz almost 3 years ago - 9 comments

#86 - torch.quantization --> torch.ao.quantization in vision for facebookresearch/fvcore

Pull Request - State: closed - Opened by HDCharles about 3 years ago - 9 comments
Labels: CLA Signed, Merged, fb-exported

#85 - torch.quantization --> torch.ao.quantization in vision

Pull Request - State: closed - Opened by HDCharles about 3 years ago - 4 comments
Labels: CLA Signed, fb-exported

#84 - "flop_count" for Detectron2 DeformConv

Issue - State: closed - Opened by bowenc0221 about 3 years ago - 3 comments

#83 - How can I support custom operators?

Issue - State: closed - Opened by xunleiwu about 3 years ago - 4 comments

#82 - How do I cite this repo?

Issue - State: closed - Opened by ekosman about 3 years ago - 2 comments

#81 - RuntimeError in ParamScheduler

Issue - State: open - Opened by krishnamoorthybabu about 3 years ago - 9 comments

#80 - Fixed flop counting for transposed convolution.

Pull Request - State: closed - Opened by ericmintun about 3 years ago - 2 comments
Labels: CLA Signed, Merged

#79 - Fix Edge Case when iter is at MAX_ITER

Pull Request - State: open - Opened by johnanthonyjose about 3 years ago - 2 comments
Labels: CLA Signed

#78 - Wrong ConvTranspose flops

Issue - State: closed - Opened by manipopopo about 3 years ago

#77 - add option to count MAC (default to True) instead of flops

Pull Request - State: open - Opened by ppwwyyxx about 3 years ago - 3 comments
Labels: CLA Signed, fb-exported

#76 - Overall Flops are less than the modules

Issue - State: closed - Opened by DuanHuiyu over 3 years ago - 5 comments

#75 - Fix usage example typo

Pull Request - State: closed - Opened by ShoufaChen over 3 years ago - 2 comments
Labels: CLA Signed, Merged

#74 - FlopCountAnalysis: ignore modules or modules with specified type

Issue - State: closed - Opened by senysenyseny16 over 3 years ago - 1 comment

#73 - Update README.md

Pull Request - State: closed - Opened by haooooooqi over 3 years ago - 2 comments
Labels: CLA Signed, Merged

#72 - calculate param problem

Issue - State: closed - Opened by WZMIAOMIAO over 3 years ago - 2 comments

#71 - Module never called

Issue - State: closed - Opened by WZMIAOMIAO over 3 years ago - 1 comment
Labels: enhancement

#70 - preciseBN example

Issue - State: closed - Opened by abhiagwl4262 over 3 years ago - 1 comment

#69 - Flops count is actually MAC count

Issue - State: open - Opened by voldemortX over 3 years ago - 2 comments
Labels: enhancement

#68 - flop count support more complicated einsum equations

Pull Request - State: closed - Opened by ppwwyyxx over 3 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#67 - fvcore flop-counting Error, ValueError: Expected more than 1 value per channel when training, got input size torch.Size([1, 256, 1, 1])

Issue - State: closed - Opened by JamesLee789 over 3 years ago - 1 comment

GitHub / facebookresearch/fvcore issues and pull requests