zchen/b2txt25
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

tpu

Browse Source
This commit is contained in:
Zchen
2025-10-12 21:20:08 +08:00
parent c6fc211b00
commit 00c94fd48b
2 changed files with 48 additions and 125 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
model_training_nnn/rnn_args.yaml
View File

@@ -20,7 +20,7 @@ use_amp: true # whether to use automatic mixed precision (AMP) for training
# TPU and distributed training settings
use_tpu: true # whether to use TPU for training (set to true for TPU)
num_tpu_cores: 8 # number of TPU cores to use (typically 8 for v3-8 or v4-8)
num_tpu_cores: 8 # number of TPU cores to use (full TPU v3-8)
gradient_accumulation_steps: 1 # number of gradient accumulation steps for distributed training
output_dir: trained_models/baseline_rnn # directory to save the trained model and logs