tiny llama tp2 confirm performance

Created on August 13|Last edited on August 13
Comment
﻿
Section 1﻿
validation_bleu_covost_v2.fr_en.dev.fr
validation_bleu_covost_v2.fr_en.dev.fr
20k40k60k80k100k120ktrainer/global_step28303234363840
prompt_lhmerge5_p2c_tp2_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs512_mbs16_ep200
prompt_lhmerge5_p2a_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs4096_mbs16_ep200
validation_bleu_covost_v2.es_en.dev.es
validation_bleu_covost_v2.es_en.dev.es
20k40k60k80k100k120ktrainer/global_step2830323436
prompt_lhmerge5_p2c_tp2_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs512_mbs16_ep200
prompt_lhmerge5_p2a_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs4096_mbs16_ep200
validation_bleu_europarl_st.en_es.dev
validation_bleu_europarl_st.en_es.dev
20k40k60k80k100k120ktrainer/global_step26283032343638
prompt_lhmerge5_p2c_tp2_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs512_mbs16_ep200
prompt_lhmerge5_p2a_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs4096_mbs16_ep200
validation_bleu_europarl_st.en_de.dev
validation_bleu_europarl_st.en_de.dev
20k40k60k80k100k120ktrainer/global_step18202224262830
prompt_lhmerge5_p2c_tp2_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs512_mbs16_ep200
prompt_lhmerge5_p2a_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs4096_mbs16_ep200
validation_bleu_covost_v2.de_en.dev.de
validation_bleu_covost_v2.de_en.dev.de
20k40k60k80k100k120ktrainer/global_step253035
prompt_lhmerge5_p2c_tp2_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs512_mbs16_ep200
prompt_lhmerge5_p2a_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs4096_mbs16_ep200
validation_bleu_europarl_st.en_fr.dev
validation_bleu_europarl_st.en_fr.dev
20k40k60k80k100k120ktrainer/global_step2022242628
prompt_lhmerge5_p2c_tp2_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs512_mbs16_ep200
prompt_lhmerge5_p2a_oci_FC-GPT_llama_tiny_canaryset_b6s4kf-sunolong_noCC_langtemp0.5_dsettemp0.5_lr1e-4wd1e-3_CosineAnnealing_warmup2500_minlr1e-6_gbs4096_mbs16_ep200
Run set465
﻿
﻿
Add a comment