Machine-Learning
/
AIIA
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects Releases Wiki Activity

downsized trainingdata from 20k to 1k

This commit is contained in:
Falko Victor Habel 2025-01-27 08:44:46 +01:00
parent 3749ba9c5f
commit d205346741
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/pretrain.py
View File

@ -9,13 +9,13 @@ import copy
def pretrain_model(data_path1, data_path2, num_epochs=3): def pretrain_model(data_path1, data_path2, num_epochs=3):
# Read and merge datasets # Read and merge datasets
df1 = pd.read_parquet(data_path1).head(10000) df1 = pd.read_parquet(data_path1).head(5000)
df2 = pd.read_parquet(data_path2).head(10000) df2 = pd.read_parquet(data_path2).head(5000)
merged_df = pd.concat([df1, df2], ignore_index=True) merged_df = pd.concat([df1, df2], ignore_index=True)
# Model configuration # Model configuration
config = AIIAConfig( config = AIIAConfig(
model_name="AIIA-Base-512x20k", model_name="AIIA-Base-512x10k",
) )
# Initialize model and data loader # Initialize model and data loader