FineTune_withPlots1082275.out

Loading pytorch-gpu/py3/2.1.1
  Loading requirement: cuda/11.8.0 nccl/2.18.5-1-cuda cudnn/8.7.0.84-cuda
    gcc/8.5.0 openmpi/4.1.5-cuda intel-mkl/2020.4 magma/2.7.1-cuda sox/14.4.2
    sparsehash/2.0.3 libjpeg-turbo/2.1.3 ffmpeg/4.4.4
+ HF_DATASETS_OFFLINE=1
+ TRANSFORMERS_OFFLINE=1
+ python3 FIneTune_withPlots.py

Checking label assignment:

Domain: Mathematics
Categories: math.KT math.RT
Abstract: we compute the hochschild cohomology and homology of a class of quantum exterior algebras with coeff...

Domain: Computer Science
Categories: cs.AI cs.LO
Abstract: this paper presents experiments on common knowledge logic conducted with the help of the proof assis...

Domain: Physics
Categories: physics.ins-det physics.gen-ph
Abstract: soil bulk density affects water storage water and nutrient movement and plant root activity in the s...

Domain: Chemistry
Categories: nlin.CD
Abstract: two chaotic systems which interact by mutually exchanging a signal built from their delayed internal...

Domain: Statistics
Categories: stat.ME stat.AP
Abstract: it is difficult to accurately estimate the rates of rape and domestic violence due to the sensitive ...

Domain: Biology
Categories: q-bio.PE
Abstract: the distribution of genetic polymorphisms in a population contains information about the mutation ra...
/linkhome/rech/genrug01/uft12cr/.local/lib/python3.11/site-packages/transformers/tokenization_utils_base.py:2057: FutureWarning: Calling BertTokenizer.from_pretrained() with the path to a single file or url is deprecated and won't be possible anymore in v5. Use a model identifier or the path to a directory instead.
  warnings.warn(

Training with All Cluster tokenizer:
Vocabulary size: 16005
Could not load pretrained weights from /gpfswork/rech/fmr/uft12cr/finetuneAli/Bert_Model. Starting with random weights. Error: Error while deserializing header: HeaderTooLarge
Initialized model with vocabulary size: 16005
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:173: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  scaler = amp.GradScaler()
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Epoch 1/5:
Train Loss: 0.8860, Train Accuracy: 0.7123
Val Loss: 0.6624, Val Accuracy: 0.7811, Val F1: 0.7137
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Epoch 2/5:
Train Loss: 0.6292, Train Accuracy: 0.7928
Val Loss: 0.6377, Val Accuracy: 0.7942, Val F1: 0.7572
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Epoch 3/5:
Train Loss: 0.5420, Train Accuracy: 0.8283
Val Loss: 0.6224, Val Accuracy: 0.7983, Val F1: 0.7744
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Epoch 4/5:
Train Loss: 0.4496, Train Accuracy: 0.8583
Val Loss: 0.6285, Val Accuracy: 0.8109, Val F1: 0.7863
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 16003
Vocab size: 16005
Epoch 5/5:
Train Loss: 0.3687, Train Accuracy: 0.8816
Val Loss: 0.6460, Val Accuracy: 0.8111, Val F1: 0.7860

Test Results for All Cluster tokenizer:
Accuracy: 0.8111
F1 Score: 0.7860
AUC-ROC: 0.8681

Training with Final tokenizer:
Vocabulary size: 18524
Could not load pretrained weights from /gpfswork/rech/fmr/uft12cr/finetuneAli/Bert_Model. Starting with random weights. Error: Error while deserializing header: HeaderTooLarge
Initialized model with vocabulary size: 18524
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:173: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  scaler = amp.GradScaler()
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Epoch 1/5:
Train Loss: 0.9291, Train Accuracy: 0.6943
Val Loss: 0.7526, Val Accuracy: 0.7593, Val F1: 0.6923
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Epoch 2/5:
Train Loss: 0.6952, Train Accuracy: 0.7752
Val Loss: 0.6884, Val Accuracy: 0.7705, Val F1: 0.7291
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Epoch 3/5:
Train Loss: 0.6147, Train Accuracy: 0.7993
Val Loss: 0.6780, Val Accuracy: 0.7874, Val F1: 0.7596
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Epoch 4/5:
Train Loss: 0.5494, Train Accuracy: 0.8242
Val Loss: 0.6878, Val Accuracy: 0.7920, Val F1: 0.7655
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 18523
Vocab size: 18524
Epoch 5/5:
Train Loss: 0.4703, Train Accuracy: 0.8558
Val Loss: 0.7217, Val Accuracy: 0.8046, Val F1: 0.7712

Test Results for Final tokenizer:
Accuracy: 0.8043
F1 Score: 0.7709
AUC-ROC: 0.8254

Training with General tokenizer:
Vocabulary size: 30522
Could not load pretrained weights from /gpfswork/rech/fmr/uft12cr/finetuneAli/Bert_Model. Starting with random weights. Error: Error while deserializing header: HeaderTooLarge
Initialized model with vocabulary size: 30522
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:173: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
  scaler = amp.GradScaler()
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29464
Vocab size: 30522
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29521
Vocab size: 30522
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29446
Vocab size: 30522
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29320
Vocab size: 30522
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29336
Vocab size: 30522
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29280
Vocab size: 30522
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29130
Vocab size: 30522
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29536
Vocab size: 30522
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29445
Vocab size: 30522
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29469
Vocab size: 30522
Epoch 1/5:
Train Loss: 0.9230, Train Accuracy: 0.6966
Val Loss: 0.7881, Val Accuracy: 0.7465, Val F1: 0.6718
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29462
Vocab size: 30522
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29464
Vocab size: 30522
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29477
Vocab size: 30522
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29464
Vocab size: 30522
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29402
Vocab size: 30522
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 28993
Vocab size: 30522
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29238
Vocab size: 30522
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29558
Vocab size: 30522
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29433
Vocab size: 30522
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29339
Vocab size: 30522
Epoch 2/5:
Train Loss: 0.6269, Train Accuracy: 0.7939
Val Loss: 0.6425, Val Accuracy: 0.7959, Val F1: 0.7705
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29160
Vocab size: 30522
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29464
Vocab size: 30522
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29535
Vocab size: 30522
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29160
Vocab size: 30522
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29536
Vocab size: 30522
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29458
Vocab size: 30522
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29560
Vocab size: 30522
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29605
Vocab size: 30522
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29513
Vocab size: 30522
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29532
Vocab size: 30522
Epoch 3/5:
Train Loss: 0.5377, Train Accuracy: 0.8242
Val Loss: 0.6742, Val Accuracy: 0.7797, Val F1: 0.7674
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29494
Vocab size: 30522
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29461
Vocab size: 30522
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29454
Vocab size: 30522
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29536
Vocab size: 30522
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29602
Vocab size: 30522
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29238
Vocab size: 30522
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29536
Vocab size: 30522
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29292
Vocab size: 30522
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29390
Vocab size: 30522
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29464
Vocab size: 30522
Epoch 4/5:
Train Loss: 0.4776, Train Accuracy: 0.8478
Val Loss: 0.5951, Val Accuracy: 0.8095, Val F1: 0.7732
Batch 0:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 28987
Vocab size: 30522
/gpfsdswork/projects/rech/fmr/uft12cr/finetuneAli/FIneTune_withPlots.py:202: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
  with amp.autocast():
Batch 100:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29605
Vocab size: 30522
Batch 200:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29083
Vocab size: 30522
Batch 300:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29532
Vocab size: 30522
Batch 400:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29605
Vocab size: 30522
Batch 500:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29417
Vocab size: 30522
Batch 600:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29280
Vocab size: 30522
Batch 700:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29464
Vocab size: 30522
Batch 800:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29390
Vocab size: 30522
Batch 900:
input_ids shape: torch.Size([16, 256])
attention_mask shape: torch.Size([16, 256])
labels shape: torch.Size([16])
input_ids max value: 29441
Vocab size: 30522
Epoch 5/5:
Train Loss: 0.3833, Train Accuracy: 0.8814
Val Loss: 0.6523, Val Accuracy: 0.7882, Val F1: 0.7792

Test Results for General tokenizer:
Accuracy: 0.7885
F1 Score: 0.7796
AUC-ROC: 0.8664

Summary of Results:

All Cluster Tokenizer:
Accuracy: 0.8111
F1 Score: 0.7860
AUC-ROC: 0.8681

Final Tokenizer:
Accuracy: 0.8043
F1 Score: 0.7709
AUC-ROC: 0.8254

General Tokenizer:
Accuracy: 0.7885
F1 Score: 0.7796
AUC-ROC: 0.8664

Class distribution in training set:
Class Biology: 439 samples
Class Chemistry: 454 samples
Class Computer Science: 1358 samples
Class Mathematics: 9480 samples
Class Physics: 2733 samples
Class Statistics: 200 samples