Mirostat Perplexity Kudzora Mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

1. What does Mirostat try to control during generation?
   • A target level of output perplexity or surprise
   • Nhamba yemuenzaniso parameters
   • The training dataset size
   • Iyo tembiricha yeGPU
2. Mirostat yakanyatsosiyana sei neyakajairwa top-k kana top-p sampling?
   • Hazvimboshandisi zvakangoerekana zvaitika
   • Iyo inogadzirisa zvine simba truncation uchishandisa mhinduro pane yakagadziriswa cutoff
   • Inoshanda chete pamifananidzo
   • Zvinoda kudzidzisazve modhi
3. MuMirostat, ibasa rei iro parameter eta inoita?
   • Inoisa saizi yemazwi
   • Inogadzirisa kureba kwekubuda
   • Ndiwo mwero wekudzidza unodzora kuti chikumbaridzo chinochinja sei
   • Inodzima sampling
4. Ndeipi kuvandudzwa kukuru kweMirostat vhezheni 2 pamusoro pevhezheni 1?
   • Inogadzira mifananidzo zvakare
   • Zvinoda imwe ndangariro
   • Inobvisa tau parameter
   • Iyo inodonhedza fungidziro yekugovera kweZipfian kuti ive nyore, yakasimba kutonga
5. Kana Mirostat ikaona kuti kuburitsa kuri kuita kunge kusinga fungidzike, ichaita:
   • Sunungura truncation kubvuma mamwe ma tokens akasiyana
   • Simbisa truncation kuti ubvumire zviratidzo zvishoma
   • Rega kugadzira nekukurumidza
   • Wedzera saizi yemuenzaniso
6. Which Mirostat settings does current llama.cpp expose?
   • Tokenizer vocabulary and number of layers
   • Mode, target tau, and learning-rate eta
   • Only the prompt length
   • Training epochs and validation loss
7. What does eta control in Mirostat’s feedback update?
   • The size of the model vocabulary
   • The target number of output tokens
   • How strongly the threshold changes in response to the observed error
   • The probability of a correct factual answer
8. What does a Mirostat target not guarantee by itself?
   • That the sampler reads token probabilities
   • That the setting can influence surprise
   • That the output contains tokens
   • That the generated text is factually correct