Fix dataset name in baseline commands

- Change tiny-imagenet to tiny_imagenet to match argparse choices
- Fixes argument validation error when running experiments

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

Author: dariocazzani

PROPER_BASELINE_COMMANDS.sh

@@ -31,9 +31,9 @@ CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dat
 CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset cifar100 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 # ResNet-18 Tiny-ImageNet (3 seeds)
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny-imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny-imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny-imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny_imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny_imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny_imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 # ResNet-50 CIFAR-10 (3 seeds)
 CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset cifar10 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
@@ -46,9 +46,9 @@ CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dat
 CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset cifar100 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 # ResNet-50 Tiny-ImageNet (3 seeds)
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny-imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny-imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny-imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny_imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny_imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny_imagenet --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 
 ================================================================================
@@ -71,9 +71,9 @@ CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --
 CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset cifar100 --teacher-path results/raw/cifar100/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
 
 # ResNet-18 Tiny-ImageNet FP32+KD (3 seeds)
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet18 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet18 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 123
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet18 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet18 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet18 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 123
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet18 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --student-is-fp32 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
 
 
 ================================================================================
@@ -94,9 +94,9 @@ CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dat
 CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset cifar100 --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 # ResNet-18 Tiny-ImageNet BitNet (3 seeds)
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny-imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny-imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny-imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny_imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny_imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train --model resnet18 --dataset tiny_imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 # ResNet-50 CIFAR-10 BitNet (3 seeds)
 CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset cifar10 --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
@@ -109,9 +109,9 @@ CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dat
 CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset cifar100 --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 # ResNet-50 Tiny-ImageNet BitNet (3 seeds)
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny-imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny-imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny-imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny_imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 42
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny_imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 123
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train --model resnet50 --dataset tiny_imagenet --bit-version --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --mixup-alpha 0.2 --label-smoothing 0.1 --seed 456
 
 
 ================================================================================
@@ -134,9 +134,9 @@ CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --
 CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset cifar100 --teacher-path results/raw/cifar100/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
 
 # ResNet-18 Tiny-ImageNet BitNet+Recipe (3 seeds)
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 123
-CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 123
+CUDA_VISIBLE_DEVICES=0 uv run python -m experiments.train_kd --model resnet18 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet18/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
 
 # ResNet-50 CIFAR-10 BitNet+Recipe (3 seeds)
 CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset cifar10 --teacher-path results/raw/cifar10/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
@@ -149,9 +149,9 @@ CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --
 CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset cifar100 --teacher-path results/raw/cifar100/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
 
 # ResNet-50 Tiny-ImageNet BitNet+Recipe (3 seeds)
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 123
-CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset tiny-imagenet --teacher-path results/raw/tiny-imagenet/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 42
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 123
+CUDA_VISIBLE_DEVICES=1 uv run python -m experiments.train_kd --model resnet50 --dataset tiny_imagenet --teacher-path results/raw/tiny-imagenet/resnet50/std_s42/best_model.pth --ablation keep_conv1 --epochs 300 --warmup-epochs 5 --min-lr 1e-5 --seed 456
 
 
 ================================================================================