Glossary entry
Pre-training
पूर्व-तालिम
Definition (English)
The initial large-scale training of a model on broad data before any task-specific tuning.
अर्थ (नेपालीमा)
कुनै निश्चित कामअघि व्यापक डाटामा मोडेललाई दिइने प्रारम्भिक ठूलो तालिम।
Related terms
Other terms related to Pre-training.
-
Fine-Tuning
फाइन-ट्युनिङ
Further training a pre-trained model on specific data to specialize it for a task.
-
LoRA
लोरा (LoRA)
A lightweight fine-tuning method that adapts a large model by training small added layers.
-
Distributed Training
वितरित तालिम
Training a model across many machines or chips at once to handle large workloads.
-
Existential Risk
अस्तित्वगत जोखिम
The concern that highly advanced AI could pose severe, large-scale risks to humanity.
-
Foundation Model
आधार मोडेल
A large model trained on broad data that can be adapted to many downstream tasks.
-
Hyperparameter
हाइपरप्यारामिटर
A configuration value set before training that controls how a model learns, like learning rate.
-
Narrow AI
सीमित एआई
AI designed and trained for one specific task, like spam filtering or face recognition.
-
Next Token Prediction
अर्को टोकन अनुमान
The core training task of LLMs: predicting the next piece of text given prior context.