Are there still windows 7 64 exe, being produced?

[Dboingue]

The links within the closed issue are no valid anymore. I think that the XML error page I end up with means that.
#3 (comment)

If however, there are no updated such executable, then I figure that this repository can be used in visual studio, and will compile fine on that plaform? right. I am new to this, (github repository structures, and anything new from the past 10 years in terms of visual studio, but i am making some progress), and using an executable would spare me the delay on compiling, etc...

[Checkmate6659]

i also want to build this tool because i didn't find an executable, and so i tried opening the solution in visual studio. but it says it can't find the .vxcproj file, and so i cant build it.

[KyleGass]

---

Will be updated soon

This Link by the LC0 discord server seems to work:
https://cdn.discordapp.com/attachments/427066771627966466/787663547546861568/trainingdata-tool.exe

For anyone following this LC0 guide: Training and Testing a Net on Google Colab: Beginner Friendly Guide
(https://lczero.org/dev/wiki/training-and-testing-a-net-on-google-colab-beginner-friendly-guide/)

I later also needed to create a proto folder in lczero-training-master\tf and put in this file:
https://github.com/LeelaChessZero/lczero-common/tree/master/proto

Discord fildes maybe too old (https://cdn.discordapp.com/attachments/427066771627966466/822101986597797918/chunk_pb2.py
https://cdn.discordapp.com/attachments/427066771627966466/822101992587395142/net_pb2.py)
(and also created a tmp folder with an empty mymodel4.txt file in that folder)

I also had to edit the example.yaml file in the lczero-training-master\tf\configs folder

%YAML 1.2

name: t1-128x10b
gpu: 0
dataset:
num_chunks: 16865
train_ratio: 0.95
input_train: /content/gdrive/My Drive/lc0/Training Data//
input_test: /content/gdrive/My Drive/lc0/Training Data//
training:
batch_size: 512
test_steps: 2000
train_avg_report_steps: 1
total_steps: 100000
checkpoint_steps: 1000
warmup_steps: 250
shuffle_size: 524288
lr_values:
- 0.02
- 0.002
- 0.0005
lr_boundaries:
- 60000
- 80000
policy_loss_weight: 1
value_loss_weight: 1
moves_left_loss_weight: .01
path: /content/gdrive/My Drive/lc0/
model:
filters: 128
residual_blocks: 10
se_ratio: 8
...

I Created a google drive folder: lc0 put in the lczero-training-master folder, Training Data folder, trainingdata-tool.exe & pgn-extract.exe and the input.pgn as described in https://lczero.org/dev/wiki/training-and-testing-a-net-on-google-colab-beginner-friendly-guide/

with google colab created a notebook & with this commands it finally worked:
from google.colab import drive
drive.mount('/content/gdrive', force_remount=True)
!chmod -R 777 "/content/gdrive/My Drive/lc0/"
!"/content/gdrive/My Drive/lc0/lczero-training-master/tf/train.py" --cfg "/content/gdrive/My Drive/lc0/lczero-training-master/tf/configs/example.yaml" --output "/content/gdrive/My Drive/lc0/lczero-training-master/tf/tmp/mymodel4.txt"

[CallOn84]

The links for the .exe have all been depreciated. Luckily, I downloaded a copy before the link died. It's available in my repository for all my Lc0 nets.