Quiz collegato alla guida · Medio Livello
Quiz sulla quantizzazione
Esamina i valori di precisione inferiore, il carico utile del peso ideale, l'arrotondamento, la calibrazione GPTQ e AWQ, i kernel, i pesi sensibili all'attivazione, i codici affini e QAT.
Domanda 1 di 8
In a weight-focused quantization scheme, what representation property is changed?
Anteprima di ogni domanda8 domande
Le risposte e le spiegazioni vengono visualizzate mentre rispondi al quiz.
- In a weight-focused quantization scheme, what representation property is changed?
- The mandatory number of network layers
- The identity of every original training document
- The numerical levels or precision used for selected weights
- Only the displayed model title
- Ignoring metadata and other working memory, what fraction of raw packed-weight bits remains when moving the same weights from 16-bit to 4-bit?
- The same amount
- Twice as much
- One-hundredth in every case
- One-quarter
- Qual è il principale svantaggio della quantizzazione aggressiva?
- L'errore di arrotondamento può ridurre la precisione del modello
- Il modello non può più essere scaricato
- Cancella permanentemente i dati di allenamento
- Rende sempre il modello più lento
- Why do the cited GPTQ and AWQ methods inspect calibration inputs?
- To guarantee identical performance on every future input
- To guide their respective weight-quantization choices and error management
- To translate the vocabulary into another language
- To add a required number of new transformer layers
- Why must runtime speed be measured rather than inferred from a lower weight bitwidth alone?
- A bitwidth label guarantees an identical speedup on every device
- All quantized values remove the need for computation
- Every compressed model must be slower by definition
- Kernels, hardware, and conversion overhead also affect runtime
- What does activation-aware mean in the cited AWQ weight-quantization method?
- Every activation must be stored at the same low bitwidth as every weight
- The weights are never quantized
- Activation observations help choose salient weight scaling
- The method uses only parameter names without input observations
- Using the stated reconstruction rule scale times code minus zero point with subtraction inside, what value comes from code six, scale 0.5, and zero point two?
- Quattro
- Due
- Tre
- Sei
- How does quantization-aware training differ in timing from post-training quantization?
- Training incorporates quantization effects so learned values can adapt
- It is defined as changing only a fitted model after every training stage ends
- It forbids all training examples
- It guarantees no rounding or approximation error in any representation