Paraphrases of the max 500 tokens subset of the MMLU dataset. We train models on both paraphrases and not paraphrases to increase robustness.
Róbert Belanec
rbelanec
AI & ML interests
None yet
Recent Activity
published
a model
25 minutes ago
rbelanec/train_2025-04-09-17-42-00
updated
a model
about 1 hour ago
rbelanec/train_2025-04-09-14-52-53
published
a model
about 1 hour ago
rbelanec/train_2025-04-09-14-52-53
Organizations
Collections
2
models
126
rbelanec/train_2025-04-09-17-42-00
Updated
rbelanec/train_2025-04-09-14-52-53
Updated
rbelanec/lora_04032025080604_mmlu_adv_gemma-2-9b-it
Updated
rbelanec/lora_04012025151205_mmlu_adv_meta-llama-3.1-8b-instruct
Updated
rbelanec/pt_04012025151019_mmlu_adv_gemma-2-9b-it
Updated
rbelanec/pt_04012025142553_mmlu_adv_meta-llama-3.1-8b-instruct
Updated
rbelanec/lora_03182025094215_mmlu_paraphrases_meta-llama-3.1-8b-instruct
Updated
rbelanec/lora_03182025100228_mmlu_paraphrases_gemma-2-9b-it
Updated
rbelanec/lora_03182025100228_mmlu_gemma-2-9b-it
Updated
rbelanec/lora_03182025094215_mmlu_meta-llama-3.1-8b-instruct
Updated