File size: 1,011 Bytes
85b32b8
 
 
 
 
 
 
 
 
 
 
 
7b03907
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ffb726a
7b03907
ffb726a
7b03907
ffb726a
7b03907
ffb726a
7b03907
ffb726a
 
 
 
 
 
7b03907
 
 
 
 
 
 
85b32b8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
---
license: mit
language:
- en
metrics:
- accuracy
- f1
- precision
- recall
library_name: transformers
pipeline_tag: text-classification
---
# Total Samples
Samples: 716017(Train+Test)

Training Samples: 579973

Validation Samples :64442

Test Samples :71602

# Overall Metrics
Accuracy :92%

F1 Score:92%

Recall:92%

Precisison: 92%

# Fine Tune Parameters
No of epochs: 3

Batch Size: 16

evaluation startegy: epoch

optimiser:Adamw

learning_rate:2e-5

max_steps:1000

warmup_step: 100

Monitoring Train & Evaluation:WANDB API

# Train 
train_runtime': 1594.4072, 'train_samples_per_second': 80.281, 'train_steps_per_second': 0.627, 'total_flos': 5589761482241280.0, 'train_loss': 0.26639655661582945, 'epoch': 0.22

# Validation 
'eval_loss': 0.22991116344928741,'eval_accuracy': 0.9211073523478477,'eval_precision': 0.9213582014463746,'eval_recall': 0.921107352347847'eval_f1': 0.9210970707304227,

'eval_runtime': 238.5409,'eval_samples_per_second': 270.151,'eval_steps_per_second': 8.443,'epoch': 0.22