brezn + kafka DARE

data_stats_langs.csv

@@ -1,25 +1,27 @@
 model,model_name,de,en,de_frac
 mattshumer_mistral-8x7b-chat,mattshumer/mistral-8x7b-chat (Mistral),0.5600046414510724,0.6436643654583162,
 DiscoResearch_mixtral-7b-8expert,DiscoResearch/mixtral-7b-8expert (Mistral MoE),0.530225066820544,0.6235667963921786,
-VAGOsolutions_SauerkrautLM-SOLAR-Instruct,VAGOsolutions/SauerkrautLM-SOLAR-Instruct (Mistral),0.5264626841230896,0.6448543385417635,0.6872427983539094
-upstage_SOLAR-10.7B-Instruct-v1.0,upstage/SOLAR-10.7B-Instruct-v1.0 (Mistral),0.5185997487442598,0.6412418664969965,0.654320987654321
+VAGOsolutions_SauerkrautLM-SOLAR-Instruct,VAGOsolutions/SauerkrautLM-SOLAR-Instruct (Mistral),0.5264626841230896,0.6448543385417635,0.6831275720164609
+seedboxai_KafkaLM-7B-DARE_TIES-LaserRMT-QLoRA-DPO-v0.5,seedboxai/KafkaLM-7B-DARE_TIES-LaserRMT-QLoRA-DPO-v0.5 (Mistral),0.52427980929471,0.6068824551944333,0.8559670781893004
+mayflowergmbh_Brezn-7b,mayflowergmbh/Brezn-7b (Mistral),0.5188284563593292,0.6217040067980718,0.9711934156378601
+upstage_SOLAR-10.7B-Instruct-v1.0,upstage/SOLAR-10.7B-Instruct-v1.0 (Mistral),0.5185997487442598,0.6412418664969965,0.6502057613168725
 mayflowergmbh_Wiedervereinigung-7b-dpo,mayflowergmbh/Wiedervereinigung-7b-dpo (Mistral),0.5084692636625303,0.5649458654779574,0.9753086419753086
 mayflowergmbh_Wiedervereinigung-7b-dpo-laser,mayflowergmbh/Wiedervereinigung-7b-dpo-laser (Mistral),0.5050259512945204,0.565063617359027,0.9753086419753086
-mlabonne_AlphaMonarch-7B,mlabonne/AlphaMonarch-7B (Mistral),0.501206548961392,0.6591874438622916,0.8518518518518519
-NousResearch_Nous-Hermes-2-Mistral-7B-DPO,NousResearch/Nous-Hermes-2-Mistral-7B-DPO (Mistral),0.5005169712928532,0.6084873546922523,0.897119341563786
-DRXD1000_Phoenix,DRXD1000/Phoenix (Mistral),0.4875727579736855,0.5238922380729938,0.8477366255144033
+mlabonne_AlphaMonarch-7B,mlabonne/AlphaMonarch-7B (Mistral),0.501206548961392,0.6591874438622916,0.8559670781893004
+DRXD1000_Phoenix,DRXD1000/Phoenix (Mistral),0.4875727579736855,0.5238922380729938,0.8395061728395061
 openaccess-ai-collective_DPOpenHermes-7B,openaccess-ai-collective/DPOpenHermes-7B (Mistral),0.4808996941837862,0.6204995440653475,
 VAGOsolutions_SauerkrautLM-7b-HerO,VAGOsolutions/SauerkrautLM-7b-HerO (Mistral),0.4793394346045998,0.5973278911555183,0.9506172839506173
 malteos_hermeo-7b,malteos/hermeo-7b (Mistral),0.4789016907464211,0.5777652767503162,0.9382716049382716
 fblgit_una-cybertron-7b-v2-bf16,fblgit/una-cybertron-7b-v2-bf16 (Mistral),0.4714702998712998,0.6280572487472983,
 aari1995_germeo-7b-awq,aari1995/germeo-7b-awq (Mistral),0.4701767724890486,0.5677674726911679,
+NousResearch_Nous-Hermes-2-Mistral-7B-DPO,NousResearch/Nous-Hermes-2-Mistral-7B-DPO (Mistral),0.469343509872141,0.6084873546922523,0.9012345679012346
 berkeley-nest_Starling-LM-7B-alpha,berkeley-nest/Starling-LM-7B-alpha (Mistral),0.4682092030538821,0.6082623478431886,
-openchat_openchat_3.5,openchat/openchat_3.5 (Mistral),0.4677697134012669,0.6093495408502738,0.9053497942386831
+openchat_openchat_3.5,openchat/openchat_3.5 (Mistral),0.4677697134012669,0.6093495408502738,0.9012345679012346
 mistralai_Mistral-7B-Instruct-v0.2,mistralai/Mistral-7B-Instruct-v0.2 (Mistral),0.4671287660155921,0.5974778867312014,0.48148148148148145
 teknium_OpenHermes-2.5-Mistral-7B,teknium/OpenHermes-2.5-Mistral-7B (Mistral),0.4648073216754412,0.6009262837808728,
-HuggingFaceH4_zephyr-7b-beta,HuggingFaceH4/zephyr-7b-beta (Mistral),0.4641427965971664,0.5990940441734774,0.7530864197530864
+HuggingFaceH4_zephyr-7b-beta,HuggingFaceH4/zephyr-7b-beta (Mistral),0.4641427965971664,0.5990940441734774,0.7489711934156379
 seedboxai_KafkaLM-7B-German-V0.1-DPO,seedboxai/KafkaLM-7B-German-V0.1-DPO (Mistral),0.4594284223032104,0.5042915057319466,0.9547325102880658
-seedboxai_KafkaLM-7B-German-V0.1,seedboxai/KafkaLM-7B-German-V0.1 (Mistral),0.4586080784053851,0.5042391535612452,0.9506172839506173
+seedboxai_KafkaLM-7B-German-V0.1,seedboxai/KafkaLM-7B-German-V0.1 (Mistral),0.4586080784053851,0.5042391535612452,0.9465020576131687
 VAGOsolutions_SauerkrautLM-7b-v1-mistral,VAGOsolutions/SauerkrautLM-7b-v1-mistral (Mistral),0.4582885390631627,0.5730988906371487,
 ehartford_dolphin-2.2.1-mistral-7b,ehartford/dolphin-2.2.1-mistral-7b (Mistral),0.4575957336188839,0.5889248773933949,
 argilla_notus-7b-v1,argilla/notus-7b-v1 (Mistral),0.4538267074114633,0.6071679590684926,
@@ -34,9 +36,9 @@ LeoLM_leo-hessianai-7b-chat,LeoLM/leo-hessianai-7b-chat (Llama2),0.4198959530774
 kaist-ai_prometheus-13b-v1.0,kaist-ai/prometheus-13b-v1.0 (Llama2),0.4137699872914333,0.5276183571587882,
 meta-llama_Llama-2-13b-chat-hf,meta-llama/Llama-2-13b-chat-hf (Llama2),0.4133328486084298,0.5331779969903168,
 Deci_DeciLM-7B-instruct,Deci/DeciLM-7B-instruct (Deci),0.4087346114484392,0.5746784325286699,0.5843621399176955
-LeoLM_leo-hessianai-7b-chat-bilingual,LeoLM/leo-hessianai-7b-chat-bilingual (Llama2),0.4058414394510105,0.4701378898025081,0.43209876543209874
+LeoLM_leo-hessianai-7b-chat-bilingual,LeoLM/leo-hessianai-7b-chat-bilingual (Llama2),0.4058414394510105,0.4701378898025081,0.43621399176954734
 mistralai_Mistral-7B-Instruct-v0.1,mistralai/Mistral-7B-Instruct-v0.1 (Mistral),0.4012720986996964,0.5329324676998503,
-allenai_tulu-2-dpo-7b,allenai/tulu-2-dpo-7b (Llama2),0.3963844596307675,0.5442215865769193,0.7613168724279835
+allenai_tulu-2-dpo-7b,allenai/tulu-2-dpo-7b (Llama2),0.3963844596307675,0.5442215865769193,0.757201646090535
 Deci_DeciLM-7B,Deci/DeciLM-7B (Deci),0.3891524333700293,0.5632148707253728,
 microsoft_Orca-2-7b,microsoft/Orca-2-7b (Llama2),0.387631543457434,0.5275512116539839,
 stabilityai_StableBeluga-7B,stabilityai/StableBeluga-7B (Llama2),0.3866115159394073,0.5308540939333947,
@@ -53,9 +55,9 @@ meta-llama_Llama-2-7b-hf,meta-llama/Llama-2-7b-hf (Llama2),0.3426308784990838,0.
 mosaicml_mpt-7b-chat,mosaicml/mpt-7b-chat (Moasaic),0.3027332136150936,0.456474162289026,
 mosaicml_mpt-7b-instruct,mosaicml/mpt-7b-instruct (Moasaic),0.3010459486867476,0.4486541776873733,
 mosaicml_mpt-7b,mosaicml/mpt-7b (Moasaic),0.2907097637642126,0.4243986375791853,
-microsoft_phi-2,microsoft/phi-2 (Phi),0.2670193027850859,0.5443139382853103,0.7448559670781894
+microsoft_phi-2,microsoft/phi-2 (Phi),0.2670193027850859,0.5443139382853103,0.7407407407407407
 tiiuae_falcon-7b-instruct,tiiuae/falcon-7b-instruct (Falcon),0.2605640889237062,0.3861411107665298,
-google_gemma-2b-it,google/gemma-2b-it (gemma),0.251767466436072,0.2809217087736937,0.8724279835390947
+google_gemma-2b-it,google/gemma-2b-it (gemma),0.251767466436072,0.2809217087736937,0.8765432098765432
 microsoft_phi-1_5,microsoft/phi-1_5 (Phi),0.2440912782008547,0.4476316337588321,
 google_gemma-7b,google/gemma-7b (gemma),0.2424066312098118,0.2337357849227176,
 google_gemma-7b-it,google/gemma-7b-it (gemma),0.2403170773221191,0.2593482051453202,0.8395061728395061

model_stats.csv

@@ -2,17 +2,19 @@ model_name,de,en,model
 mattshumer/mistral-8x7b-chat (Mistral),0.5600046414510724,0.6436643654583162,mattshumer_mistral-8x7b-chat
 DiscoResearch/mixtral-7b-8expert (Mistral MoE),0.530225066820544,0.6235667963921786,DiscoResearch_mixtral-7b-8expert
 VAGOsolutions/SauerkrautLM-SOLAR-Instruct (Mistral),0.5264626841230896,0.6448543385417635,VAGOsolutions_SauerkrautLM-SOLAR-Instruct
+seedboxai/KafkaLM-7B-DARE_TIES-LaserRMT-QLoRA-DPO-v0.5 (Mistral),0.52427980929471,0.6068824551944333,seedboxai_KafkaLM-7B-DARE_TIES-LaserRMT-QLoRA-DPO-v0.5
+mayflowergmbh/Brezn-7b (Mistral),0.5188284563593292,0.6217040067980718,mayflowergmbh_Brezn-7b
 upstage/SOLAR-10.7B-Instruct-v1.0 (Mistral),0.5185997487442598,0.6412418664969965,upstage_SOLAR-10.7B-Instruct-v1.0
 mayflowergmbh/Wiedervereinigung-7b-dpo (Mistral),0.5084692636625303,0.5649458654779574,mayflowergmbh_Wiedervereinigung-7b-dpo
 mayflowergmbh/Wiedervereinigung-7b-dpo-laser (Mistral),0.5050259512945204,0.565063617359027,mayflowergmbh_Wiedervereinigung-7b-dpo-laser
 mlabonne/AlphaMonarch-7B (Mistral),0.501206548961392,0.6591874438622916,mlabonne_AlphaMonarch-7B
-NousResearch/Nous-Hermes-2-Mistral-7B-DPO (Mistral),0.5005169712928532,0.6084873546922523,NousResearch_Nous-Hermes-2-Mistral-7B-DPO
 DRXD1000/Phoenix (Mistral),0.48757275797368554,0.5238922380729938,DRXD1000_Phoenix
 openaccess-ai-collective/DPOpenHermes-7B (Mistral),0.48089969418378625,0.6204995440653475,openaccess-ai-collective_DPOpenHermes-7B
 VAGOsolutions/SauerkrautLM-7b-HerO (Mistral),0.4793394346045998,0.5973278911555183,VAGOsolutions_SauerkrautLM-7b-HerO
 malteos/hermeo-7b (Mistral),0.47890169074642114,0.5777652767503162,malteos_hermeo-7b
 fblgit/una-cybertron-7b-v2-bf16 (Mistral),0.4714702998712998,0.6280572487472983,fblgit_una-cybertron-7b-v2-bf16
 aari1995/germeo-7b-awq (Mistral),0.47017677248904866,0.5677674726911679,aari1995_germeo-7b-awq
+NousResearch/Nous-Hermes-2-Mistral-7B-DPO (Mistral),0.469343509872141,0.6084873546922523,NousResearch_Nous-Hermes-2-Mistral-7B-DPO
 berkeley-nest/Starling-LM-7B-alpha (Mistral),0.4682092030538821,0.6082623478431886,berkeley-nest_Starling-LM-7B-alpha
 openchat/openchat_3.5 (Mistral),0.4677697134012669,0.6093495408502738,openchat_openchat_3.5
 mistralai/Mistral-7B-Instruct-v0.2 (Mistral),0.4671287660155921,0.5974778867312014,mistralai_Mistral-7B-Instruct-v0.2