Iwọn otutu ati Idanwo Iṣapẹẹrẹ

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Kini igbega iwọn otutu ṣe si iṣelọpọ awoṣe kan?
   • Mu ki o ṣe ipinnu diẹ sii ati atunwi
   • Yipada awọn iwuwo awoṣe nigbagbogbo
   • Mu iṣẹ softmax ṣiṣẹ
   • Mu ki o siwaju sii ID ati orisirisi, pẹlu diẹ anfani ti dani àmi
2. Ṣaaju iṣapẹẹrẹ, awọn nọmba ami ami aise (awọn iwe-ipamọ) ti awoṣe ti yipada si awọn iṣeeṣe nipasẹ:
   • Ajọ oke-k
   • Awọn iṣẹ softmax
   • Itankalẹ afẹyinti
   • Awọn tokenizer
3. What determines the candidate set in top-p sampling, compared with top-k?
   • Top-p always keeps exactly ten tokens
   • Both methods always keep every vocabulary token
   • Top-p uses the smallest leading set reaching a cumulative probability threshold; top-k uses a count
   • Top-k uses a cumulative threshold while top-p uses a fixed count
4. Mekaniki, kini iwọn otutu ṣe gangan si awọn logiti?
   • Pin kọọkan logit nipasẹ T ṣaaju ki o to softmax
   • Ṣe afikun nọmba laileto si logt kọọkan
   • Npa awọn logiti ti o kere julọ kuro
   • Ṣe isodipupo awọn iṣeeṣe ikẹhin nipasẹ T
5. For a structured-data task, what does choosing more predictable token selection fail to establish?
   • That a decoding setting was chosen
   • That the generated content is correct or valid
   • That token scores can be compared
   • That a model can produce an output
6. How does greedy decoding choose the next token?
   • It literally divides every logit by zero
   • It uniformly samples every vocabulary token
   • It permanently retrains the model
   • It selects a highest-scoring eligible token rather than sampling
7. Two token logits differ by six before temperature scaling. With positive temperature two, what is their scaled difference?
   • Mẹta
   • Mejila
   • Mẹfa
   • Odo
8. For probabilities 0.55, 0.30, 0.10, and 0.05, how many leading tokens does top-p 0.80 keep?
   • Ọkan
   • Mẹta
   • Meji
   • Mẹrin