-
Notifications
You must be signed in to change notification settings - Fork 3.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fixed minor typos in documentation #1959
Changes from all commits
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -390,3 +390,6 @@ R-package/src/src/ | |
lightgbm_r/* | ||
lightgbm*.tar.gz | ||
lightgbm.Rcheck/ | ||
|
||
# Files generated by aspell | ||
**/*.bak |
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -5,7 +5,7 @@ GPU Targets Table | |
================= | ||
|
||
OpenCL is a universal massively parallel programming framework that targets to multiple backends (GPU, CPU, FPGA, etc). | ||
Bascially, to use a device from a vendor, you have to install drivers from that specific vendor. | ||
Basically, to use a device from a vendor, you have to install drivers from that specific vendor. | ||
Intel's and AMD's OpenCL runtime also include x86 CPU target support. | ||
NVIDIA's OpenCL runtime only supports NVIDIA GPU (no CPU support). | ||
In general, OpenCL CPU backends are quite slow, and should be used for testing and debugging only. | ||
|
@@ -32,7 +32,7 @@ Legend: | |
Query OpenCL Devices in Your System | ||
=================================== | ||
|
||
Your system might have multiple GPUs from different vendors ("platforms") installed. Setting up LightGBM GPU device requries two parameters: `OpenCL Platform ID <./Parameters.rst#gpu_platform_id>`__ (``gpu_platform_id``) and `OpenCL Device ID <./Parameters.rst#gpu_device_id>`__ (``gpu_device_id``). Generally speaking, each vendor provides an OpenCL platform, and devices from the same vendor have different device IDs under that platform. For example, if your system has an Intel integrated GPU and two discrete GPUs from AMD, you will have two OpenCL platforms (with ``gpu_platform_id=0`` and ``gpu_platform_id=1``). If the platform 0 is Intel, it has one device (``gpu_device_id=0``) representing the Intel GPU; if the platform 1 is AMD, it has two devices (``gpu_device_id=0``, ``gpu_device_id=1``) representing the two AMD GPUs. If you have a discrete GPU by AMD/NVIDIA and an integrated GPU by Intel, make sure to select the correct ``gpu_platform_id`` to use the discrete GPU as it usually provides better performance. | ||
Your system might have multiple GPUs from different vendors ("platforms") installed. Setting up LightGBM GPU device requires two parameters: `OpenCL Platform ID <./Parameters.rst#gpu_platform_id>`__ (``gpu_platform_id``) and `OpenCL Device ID <./Parameters.rst#gpu_device_id>`__ (``gpu_device_id``). Generally speaking, each vendor provides an OpenCL platform, and devices from the same vendor have different device IDs under that platform. For example, if your system has an Intel integrated GPU and two discrete GPUs from AMD, you will have two OpenCL platforms (with ``gpu_platform_id=0`` and ``gpu_platform_id=1``). If the platform 0 is Intel, it has one device (``gpu_device_id=0``) representing the Intel GPU; if the platform 1 is AMD, it has two devices (``gpu_device_id=0``, ``gpu_device_id=1``) representing the two AMD GPUs. If you have a discrete GPU by AMD/NVIDIA and an integrated GPU by Intel, make sure to select the correct ``gpu_platform_id`` to use the discrete GPU as it usually provides better performance. | ||
|
||
On Windows, OpenCL devices can be queried using `GPUCapsViewer`_, under the OpenCL tab. Note that the platform and device IDs reported by this utility start from 1. So you should minus the reported IDs by 1. | ||
|
||
|
@@ -102,7 +102,7 @@ Example of using GPU (``gpu_platform_id = 0`` and ``gpu_device_id = 0`` in our s | |
[LightGBM] [Info] Compiling OpenCL Kernel with 16 bins... | ||
[LightGBM] [Info] GPU programs have been built | ||
[LightGBM] [Info] Size of histogram bin entry: 12 | ||
[LightGBM] [Info] 40 dense feature groups (0.12 MB) transfered to GPU in 0.004211 secs. 76 sparse feature groups. | ||
[LightGBM] [Info] 40 dense feature groups (0.12 MB) transferred to GPU in 0.004211 secs. 76 sparse feature groups. | ||
[LightGBM] [Info] No further splits with positive gain, best gain: -inf | ||
[LightGBM] [Info] Trained a tree with leaves=16 and max_depth=8 | ||
[1]: test's rmse:1.10643e-17 | ||
|
@@ -140,7 +140,7 @@ Example of using CPU (``gpu_platform_id = 0``, ``gpu_device_id = 1``). The GPU d | |
[LightGBM] [Info] Compiling OpenCL Kernel with 16 bins... | ||
[LightGBM] [Info] GPU programs have been built | ||
[LightGBM] [Info] Size of histogram bin entry: 12 | ||
[LightGBM] [Info] 40 dense feature groups (0.12 MB) transfered to GPU in 0.004540 secs. 76 sparse feature groups. | ||
[LightGBM] [Info] 40 dense feature groups (0.12 MB) transferred to GPU in 0.004540 secs. 76 sparse feature groups. | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. same as previous comment There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I'll check! |
||
[LightGBM] [Info] No further splits with positive gain, best gain: -inf | ||
[LightGBM] [Info] Trained a tree with leaves=16 and max_depth=8 | ||
[1]: test's rmse:1.10643e-17 | ||
|
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -538,7 +538,7 @@ void GPUTreeLearner::AllocateGPUMemory() { | |
} | ||
// data transfer time | ||
std::chrono::duration<double, std::milli> end_time = std::chrono::steady_clock::now() - start_time; | ||
Log::Info("%d dense feature groups (%.2f MB) transfered to GPU in %f secs. %d sparse feature groups", | ||
Log::Info("%d dense feature groups (%.2f MB) transferred to GPU in %f secs. %d sparse feature groups", | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. @Laurae2 good call. This is the only one I found (with |
||
dense_feature_group_map_.size(), ((dense_feature_group_map_.size() + (dword_features_ - 1)) / dword_features_) * num_data_ * sizeof(Feature4) / (1024.0 * 1024.0), | ||
end_time * 1e-3, sparse_feature_group_map_.size()); | ||
#if GPU_DEBUG >= 1 | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
transfered -> transferred => is it also in the C++ side?