Mikhail's picture

Mikhail

RefalMachine

·

RefalMachine

AI & ML interests

NLP

Recent Activity

upvoted a paper about 11 hours ago

FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language

liked a dataset 1 day ago

bulatovv/utmn-study-feedbacks-absa

updated a model 5 days ago

RefalMachine/RuadaptQwen3-32B-Instruct-GGUF

View all activity

Organizations

RefalMachine 's models 178

RefalMachine/ruadapt_qwen2.5_3B_ext_cl100k_unigram_32000_mean_init

3B • Updated Oct 9, 2024 • 8

RefalMachine/ruadapt_qwen2.5_3B_ext_cl100k_bpe_32000_mean_init

3B • Updated Oct 9, 2024 • 8

RefalMachine/ruadapt_qwen2.5_3B_unigram_32000_mean_init

3B • Updated Oct 9, 2024 • 8

RefalMachine/ruadapt_qwen2.5_3B_bpe_32000_mean_init

3B • Updated Oct 9, 2024 • 8

RefalMachine/ruadapt_qwen2.5_3B_darulm_cl100k_extended_u60k_full_lr3e4_bs256

3B • Updated Oct 9, 2024 • 7

RefalMachine/ruadapt_qwen2.5_3B_darulm_cl100k_extended_u60k_full_lr2e4_bs256

3B • Updated Oct 8, 2024 • 7

RefalMachine/ruadapt_qwen2.5_3B_darulm_cl100k_extended_u60k_mean_init

3B • Updated Oct 8, 2024 • 6

RefalMachine/ruadapt_qwen2.5_3B_darulm_bpe_128000_full_lr2e4_bs256

3B • Updated Oct 8, 2024 • 13

RefalMachine/ruadapt_qwen2.5_3B_darulm_unigram_128000_full_lr2e4_bs256

3B • Updated Oct 8, 2024 • 12

RefalMachine/ruadapt_qwen2.5_3B_darulm_unigram_128000_mean_init

3B • Updated Oct 7, 2024 • 13

RefalMachine/ruadapt_qwen2.5_3B_darulm_bpe_128000_mean_init

3B • Updated Oct 7, 2024 • 13

RefalMachine/ruadapt_qwen2.5_3B_darulm_bpe_64000_full_lr2e4_bs256

3B • Updated Oct 7, 2024 • 11

RefalMachine/ruadapt_qwen2.5_3B_darulm_bpe_64000_mean_init

3B • Updated Oct 7, 2024 • 13

RefalMachine/ruadapt_llama3_instruct_lep_saiga_kto_ablitirated

8B • Updated Aug 30, 2024 • 8 • 2

RefalMachine/ruadapt_mistral_7b_openchat_extended_lep_ft

7B • Updated Aug 22, 2024 • 46

RefalMachine/ruadapt_llama3_8b_instruct_extended_lep_ft

8B • Updated Aug 22, 2024 • 11 • 4

RefalMachine/Meta-Llama-3-8B_eval

Updated Aug 15, 2024

RefalMachine/llama3_darulm_20_05_24_part1-2_128000_unigram_full_lr2e4_bs256_v2

Text Generation • 8B • Updated Jul 18, 2024 • 13

RefalMachine/llama3_darulm_20_05_24_part1-2_128000_bpe_full_lr2e4_bs256_v2

Text Generation • 8B • Updated Jul 18, 2024 • 13

RefalMachine/llama3_extended_darulm_20_05_24_part1-2_64000_bpe_full_lr2e4_bs256

Text Generation • 8B • Updated Jul 17, 2024 • 12

RefalMachine/llama3_darulm_20_05_24_part1-2_128000_bpe_full_lr2e4_bs256

Text Generation • 8B • Updated Jul 17, 2024 • 13

RefalMachine/llama3_darulm_20_05_24_part1-2_128000_unigram_full_lr2e4_bs256

Text Generation • 8B • Updated Jul 17, 2024 • 13

RefalMachine/llama3_cut_extended_darulm_20_05_24_part1-2_64000_64000_bpe_full_lr2e4_bs256

Text Generation • 8B • Updated Jul 14, 2024 • 15

RefalMachine/llama3_cut_extended_darulm_20_05_24_part1-2_64000_64000_bpe_full_lr1e4_bs256

Text Generation • 8B • Updated Jul 14, 2024 • 15

RefalMachine/llama3_extended_darulm_20_05_24_part1-2_64000_bpe_full_lr1e4_bs256

Text Generation • 8B • Updated Jul 14, 2024 • 14

RefalMachine/llama3_darulm_20_05_24_part1-2_128000_unigram_full_lr1e4_bs256

Text Generation • 8B • Updated Jul 14, 2024 • 13

RefalMachine/llama3_darulm_20_05_24_part1-2_128000_bpe_full_lr1e4_bs256

Text Generation • 8B • Updated Jul 14, 2024 • 13

RefalMachine/openchat-3.5-0106__eval_bs_experiments

Updated Jul 13, 2024

RefalMachine/Meta-Llama-3-8B-Instruct_eval_bs_experiments

Updated Jul 13, 2024

RefalMachine/saiga_llama3_8b_v7_eval

Updated Jul 12, 2024