Datasets:

hlzhang109
/

proteus-2k

model_id stringlengths 10 96	leaderboard_acc_none float64 0.1 0.65	leaderboard_acc_norm_none float64 0.26 0.68	leaderboard_bbh_acc_norm_none float64 0.25 0.74	leaderboard_bbh_boolean_expressions_acc_norm_none float64 0.42 0.95	leaderboard_bbh_causal_judgement_acc_norm_none float64 0.43 0.7	leaderboard_bbh_date_understanding_acc_norm_none float64 0.12 0.84	leaderboard_bbh_disambiguation_qa_acc_norm_none float64 0.19 0.86	leaderboard_bbh_formal_fallacies_acc_norm_none float64 0.39 0.82	leaderboard_bbh_geometric_shapes_acc_norm_none float64 0 0.7	leaderboard_bbh_hyperbaton_acc_norm_none float64 0.34 0.97	leaderboard_bbh_logical_deduction_five_objects_acc_norm_none float64 0.15 0.78	leaderboard_bbh_logical_deduction_seven_objects_acc_norm_none float64 0.1 0.7	leaderboard_bbh_logical_deduction_three_objects_acc_norm_none float64 0.28 0.98	leaderboard_bbh_movie_recommendation_acc_norm_none float64 0 0.97	leaderboard_bbh_navigate_acc_norm_none float64 0.38 0.84	leaderboard_bbh_object_counting_acc_norm_none float64 0 0.62	leaderboard_bbh_penguins_in_a_table_acc_norm_none float64 0.11 0.84	leaderboard_bbh_reasoning_about_colored_objects_acc_norm_none float64 0.02 0.9	leaderboard_bbh_ruin_names_acc_norm_none float64 0 0.88	leaderboard_bbh_salient_translation_error_detection_acc_norm_none float64 0.1 0.75	leaderboard_bbh_snarks_acc_norm_none float64 0.34 0.88	leaderboard_bbh_sports_understanding_acc_norm_none float64 0.43 0.92	leaderboard_bbh_temporal_sequences_acc_norm_none float64 0.02 1	leaderboard_bbh_tracking_shuffled_objects_five_objects_acc_norm_none float64 0.11 0.35	leaderboard_bbh_tracking_shuffled_objects_seven_objects_acc_norm_none float64 0.06 0.36	leaderboard_bbh_tracking_shuffled_objects_three_objects_acc_norm_none float64 0.2 0.44	leaderboard_bbh_web_of_lies_acc_norm_none float64 0.44 0.79	leaderboard_exact_match_none float64 0 0.93	leaderboard_gpqa_acc_norm_none float64 0.21 0.45	leaderboard_gpqa_diamond_acc_norm_none float64 0.17 0.48	leaderboard_gpqa_extended_acc_norm_none float64 0.21 0.45	leaderboard_gpqa_main_acc_norm_none float64 0.18 0.43	leaderboard_ifeval_inst_level_strict_acc_none float64 0 0.84	leaderboard_ifeval_prompt_level_strict_acc_none float64 0 0.78	leaderboard_inst_level_strict_acc_none float64 0 0.84	leaderboard_math_algebra_hard_exact_match_none float64 0 0.95	leaderboard_math_counting_and_prob_hard_exact_match_none float64 0 0.95	leaderboard_math_geometry_hard_exact_match_none float64 0 0.89	leaderboard_math_hard_exact_match_none float64 0 0.93	leaderboard_math_intermediate_algebra_hard_exact_match_none float64 0 0.9	leaderboard_math_num_theory_hard_exact_match_none float64 0 0.97	leaderboard_math_prealgebra_hard_exact_match_none float64 0 0.93	leaderboard_math_precalculus_hard_exact_match_none float64 0 0.87	leaderboard_mmlu_pro_acc_none float64 0.1 0.65	leaderboard_musr_acc_norm_none float64 0.28 0.58	leaderboard_musr_murder_mysteries_acc_norm_none float64 0.45 0.72	leaderboard_musr_object_placements_acc_norm_none float64 0.2 0.5	leaderboard_musr_team_allocation_acc_norm_none float64 0.13 0.66	leaderboard_prompt_level_strict_acc_none float64 0 0.78	Average ⬆️ float64 17.3 58.7	Architecture stringclasses 32 values	Model sha stringlengths 40 40 ⌀	Hub License stringclasses 26 values	Hub ❤️ float64 0 6.49k ⌀	Hub downloads float64 0 19.6M ⌀	#Params (B) float64 0 57.4 ⌀	Available on the hub bool 1 class	Chat Template bool 2 classes	Base Model stringclasses 204 values	Hub lastModified stringdate 2022-09-26 19:05:19 2026-03-23 02:44:22 ⌀	library_name stringclasses 6 values	pipeline_tag stringclasses 3 values	gated stringclasses 3 values
meta-llama/Meta-Llama-3-8B	0.325632	0.427682	0.457559	0.788	0.55615	0.548	0.456	0.588	0.336	0.712	0.336	0.324	0.46	0.796	0.508	0.452	0.438356	0.376	0.508	0.352	0.567416	0.728	0.1	0.152	0.136	0.324	0.488	0.02719	0.321309	0.282828	0.320513	0.339286	0.227818	0.114603	0.227818	0.039088	0.02439	0.007576	0.02719	0.003571	0.012987	0.072539	0.022222	0.325632	0.367725	0.524	0.320313	0.26	0.114603	28.787213	LlamaForCausalLM	8cde5ca8380496c9a6cc7ef3a8b46a0372a1d920	llama3	6,489	3,509,989	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-09-27T15:52:33.000Z	transformers	text-generation	manual
meta-llama/Meta-Llama-3-8B-Instruct	0.359292	0.450642	0.489325	0.752	0.625668	0.46	0.64	0.556	0.34	0.544	0.408	0.448	0.584	0.584	0.528	0.472	0.472603	0.604	0.592	0.492	0.505618	0.728	0.212	0.216	0.168	0.332	0.512	0.070997	0.308725	0.328283	0.305861	0.303571	0.510791	0.358595	0.510791	0.127036	0.03252	0.045455	0.070997	0.028571	0.038961	0.124352	0.051852	0.359292	0.37963	0.532	0.265625	0.344	0.358595	35.312658	LlamaForCausalLM	8afb486c1db24fe5011ec46dfbe5b5dccdb575c2	llama3	4,428	1,454,679	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2025-06-18T23:49:51.000Z	transformers	text-generation	manual
NousResearch/Meta-Llama-3-8B-Instruct	0.359292	0.450642	0.489325	0.752	0.625668	0.46	0.64	0.556	0.34	0.544	0.408	0.448	0.584	0.584	0.528	0.472	0.472603	0.604	0.592	0.492	0.505618	0.728	0.212	0.216	0.168	0.332	0.512	0.070997	0.308725	0.328283	0.305861	0.303571	0.51199	0.360444	0.51199	0.127036	0.03252	0.045455	0.070997	0.028571	0.038961	0.124352	0.051852	0.359292	0.37963	0.532	0.265625	0.344	0.360444	35.332642	LlamaForCausalLM	53346005fb0ef11d3b6a83b12c895cca40156b6c	other	105	34,292	null	true	true	meta-llama/Meta-Llama-3-8B	2024-07-23T04:40:46.000Z	transformers	text-generation	False
unsloth/llama-3-8b-bnb-4bit	0.314245	0.418731	0.450963	0.772	0.518717	0.452	0.616	0.54	0.424	0.616	0.296	0.228	0.436	0.784	0.52	0.484	0.356164	0.392	0.516	0.396	0.522472	0.684	0.116	0.184	0.148	0.332	0.488	0.023414	0.300336	0.267677	0.300366	0.314732	0.2494	0.147874	0.2494	0.04886	0	0	0.023414	0.014286	0.019481	0.041451	0.007407	0.314245	0.359788	0.516	0.324219	0.24	0.147874	28.30245	LlamaForCausalLM	f20166737ba79ae1129e83194c84e2ec0013bf1b	llama3	204	85,889	8.248929	true	false	meta-llama/Meta-Llama-3-8B	2025-01-07T10:35:03.000Z	transformers	text-generation	False
NousResearch/Meta-Llama-3-8B	0.325632	0.427682	0.457559	0.788	0.55615	0.548	0.456	0.588	0.336	0.712	0.336	0.324	0.46	0.796	0.508	0.452	0.438356	0.376	0.508	0.352	0.567416	0.728	0.1	0.152	0.136	0.324	0.488	0.02719	0.321309	0.282828	0.320513	0.339286	0.227818	0.114603	0.227818	0.039088	0.02439	0.007576	0.02719	0.003571	0.012987	0.072539	0.022222	0.325632	0.367725	0.524	0.320313	0.26	0.114603	28.787213	LlamaForCausalLM	315b20096dc791d381d514deb5f8bd9c8d6d3061	other	102	15,017	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-04-30T04:45:04.000Z	transformers	text-generation	False
elyza/Llama-3-ELYZA-JP-8B	0.287317	0.422234	0.455303	0.74	0.59893	0.476	0.572	0.572	0.388	0.512	0.372	0.372	0.48	0.64	0.52	0.46	0.445205	0.488	0.552	0.4	0.511236	0.656	0.124	0.14	0.14	0.324	0.492	0.028701	0.291107	0.30303	0.283883	0.294643	0.556355	0.439926	0.556355	0.029316	0.00813	0.030303	0.028701	0.021429	0.012987	0.067358	0.022222	0.287317	0.376984	0.552	0.28125	0.3	0.439926	33.26279	LlamaForCausalLM	e6c316496ee7d9a11710c50229e8cb39b6b0a4a3	llama3	146	14,232	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-06-26T02:56:23.000Z	transformers	text-generation	False
NousResearch/Hermes-2-Pro-Llama-3-8B	0.322058	0.475678	0.513279	0.752	0.609626	0.368	0.688	0.6	0.42	0.684	0.432	0.42	0.584	0.776	0.584	0.492	0.527397	0.56	0.62	0.452	0.758427	0.752	0.228	0.204	0.124	0.32	0.464	0.050604	0.315436	0.308081	0.309524	0.325893	0.408873	0.269871	0.408873	0.081433	0.04878	0.022727	0.050604	0.010714	0.032468	0.119171	0.014815	0.322058	0.441799	0.548	0.417969	0.36	0.269871	34.200818	LlamaForCausalLM	f798274b30e7d2d4797c369edcc0cc7473b6e6f2	llama3	442	18,491	8.030523	true	true	NousResearch/Meta-Llama-3-8B	2024-09-14T16:29:41.000Z	transformers	text-generation	False
gradientai/Llama-3-8B-Instruct-Gradient-1048k	0.281582	0.409781	0.430828	0.74	0.572193	0.508	0.524	0.536	0.34	0.652	0.344	0.324	0.444	0.572	0.544	0.436	0.349315	0.368	0.312	0.44	0.47191	0.656	0.12	0.164	0.12	0.316	0.5	0.037009	0.301174	0.318182	0.311355	0.28125	0.268585	0.168207	0.268585	0.058632	0.02439	0.030303	0.037009	0.021429	0.032468	0.056995	0.014815	0.281582	0.420635	0.508	0.34375	0.412	0.168207	28.996901	LlamaForCausalLM	cd3069b65a8eb13da639d332a5f61b0fbb29fa73	llama3	680	11,539	8.030261	true	true	gradientai/Llama-3-8B-Instruct-Gradient-1048k	2024-10-29T16:13:18.000Z	transformers	text-generation	False
migtissera/Tess-2.0-Llama-3-8B	0.329953	0.451939	0.48811	0.82	0.59893	0.456	0.604	0.568	0.348	0.756	0.384	0.352	0.524	0.724	0.476	0.496	0.472603	0.504	0.56	0.452	0.55618	0.704	0.272	0.184	0.132	0.324	0.488	0.039275	0.302852	0.287879	0.296703	0.316964	0.376499	0.236599	0.376499	0.084691	0.00813	0.015152	0.039275	0.014286	0.012987	0.082902	0.007407	0.329953	0.411376	0.508	0.339844	0.388	0.236599	32.467748	LlamaForCausalLM	9a577d7666eb90c13752ada950a0b5f91d3749f1	llama3	17	16,345	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-06-02T01:04:27.000Z	transformers	text-generation	False
DeepMount00/Llama-3-8b-Ita	0.381732	0.457128	0.495053	0.776	0.604278	0.508	0.612	0.576	0.348	0.552	0.38	0.324	0.524	0.512	0.588	0.424	0.527397	0.584	0.532	0.496	0.567416	0.804	0.448	0.22	0.204	0.284	0.548	0.046828	0.302013	0.328283	0.298535	0.294643	0.485612	0.32902	0.485612	0.084691	0.01626	0.015152	0.046828	0.010714	0.025974	0.108808	0.02963	0.381732	0.412698	0.54	0.269531	0.432	0.32902	35.398936	LlamaForCausalLM	f95f9e25647bab95bf7966d8dbef2a6c48b5557a	llama3	31	13,841	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2025-06-11T10:04:06.000Z	transformers	text-generation	False
Groq/Llama-3-Groq-8B-Tool-Use	0.333777	0.444675	0.478389	0.728	0.631016	0.484	0.648	0.532	0.312	0.596	0.42	0.42	0.54	0.672	0.568	0.424	0.486301	0.556	0.392	0.464	0.516854	0.728	0.26	0.188	0.176	0.336	0.456	0.057402	0.317953	0.308081	0.322344	0.316964	0.432854	0.292052	0.432854	0.104235	0.03252	0.037879	0.057402	0.021429	0.019481	0.11399	0.02963	0.333777	0.387566	0.564	0.289063	0.312	0.292052	33.465674	LlamaForCausalLM	3bf6b914d7043d1bbfcfc7a9aa7581a8104eabac	llama3	296	626	8.03	true	true	meta-llama/Meta-Llama-3-8B	2024-08-27T15:50:24.000Z	transformers	text-generation	False
tokyotech-llm/Llama-3-Swallow-8B-Instruct-v0.1	0.312001	0.458814	0.492449	0.8	0.582888	0.428	0.624	0.552	0.272	0.528	0.38	0.416	0.576	0.748	0.556	0.544	0.486301	0.496	0.632	0.412	0.589888	0.732	0.292	0.188	0.16	0.328	0.544	0.043807	0.305369	0.29798	0.300366	0.314732	0.393285	0.229205	0.393285	0.058632	0.03252	0.015152	0.043807	0.028571	0.045455	0.093264	0.007407	0.312001	0.444444	0.576	0.417969	0.34	0.229205	33.189269	LlamaForCausalLM	2ed77ce6f5453bb2f933df18e236ebe0518bee1c	llama3	21	10,650	8.030261	true	true	tokyotech-llm/Llama-3-Swallow-8B-Instruct-v0.1	2024-10-08T14:05:37.000Z	transformers	text-generation	False
nvidia/Minitron-8B-Base	0.319149	0.406278	0.433605	0.812	0.625668	0.42	0.648	0.472	0.288	0.628	0.352	0.28	0.48	0.376	0.58	0.332	0.458904	0.332	0.368	0.376	0.561798	0.74	0.284	0.168	0.104	0.288	0.528	0.035498	0.277685	0.227273	0.289377	0.285714	0.300959	0.173752	0.300959	0.045603	0.02439	0.022727	0.035498	0.021429	0.064935	0.036269	0.02963	0.319149	0.400794	0.544	0.34375	0.316	0.173752	29.461502	NemotronForCausalLM	e2c30f2d4c8751f21b33be2b403d10d5f373ffc1	other	69	12,763	7.22	true	false	nvidia/Minitron-8B-Base	2025-02-14T19:04:05.000Z	transformers	text-generation	False
VAGOsolutions/Llama-3-SauerkrautLM-8b-Instruct	0.381732	0.458944	0.496442	0.776	0.59893	0.504	0.616	0.572	0.348	0.552	0.388	0.332	0.528	0.504	0.588	0.424	0.534247	0.58	0.54	0.5	0.578652	0.804	0.448	0.212	0.212	0.288	0.552	0.052115	0.306208	0.323232	0.305861	0.299107	0.471223	0.321627	0.471223	0.078176	0.04065	0.037879	0.052115	0.014286	0.058442	0.098446	0.022222	0.381732	0.414021	0.544	0.269531	0.432	0.321627	35.362345	LlamaForCausalLM	37127c44d7c0fb56cef817270c4b1a6802d8793a	other	56	13,544	8.030261	true	true	VAGOsolutions/Llama-3-SauerkrautLM-8b-Instruct	2024-04-29T18:28:16.000Z	transformers	text-generation	False
lightblue/suzume-llama-3-8B-multilingual	0.342171	0.452458	0.487763	0.788	0.593583	0.468	0.64	0.556	0.38	0.532	0.4	0.464	0.564	0.6	0.54	0.496	0.5	0.584	0.576	0.452	0.494382	0.752	0.176	0.2	0.12	0.34	0.524	0.073263	0.321309	0.333333	0.324176	0.3125	0.538369	0.404806	0.538369	0.120521	0.04065	0.030303	0.073263	0.042857	0.045455	0.150259	0.022222	0.342171	0.390212	0.532	0.324219	0.316	0.404806	35.884765	LlamaForCausalLM	0cb15aa9ec685eef494f9a15f65aefcfe3c04c66	other	114	14,480	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2024-06-02T02:14:24.000Z	transformers	text-generation	False
winninghealth/WiNGPT2-Llama-3-8B-Chat	0.315076	0.430017	0.468495	0.696	0.529412	0.424	0.708	0.576	0.348	0.656	0.372	0.416	0.576	0.804	0.656	0.492	0.465753	0.388	0.468	0.392	0.5	0.54	0.132	0.156	0.132	0.352	0.488	0.02719	0.288591	0.267677	0.272894	0.316964	0.091127	0.040665	0.091127	0.042345	0.00813	0.007576	0.02719	0.021429	0.019481	0.051813	0.014815	0.315076	0.359788	0.512	0.300781	0.268	0.040665	25.837799	LlamaForCausalLM	54f077fe389090a8e581baa1afe0f919fc300a24	apache-2.0	4	8,017	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2026-01-26T08:13:54.000Z	transformers	text-generation	False
abacusai/Llama-3-Smaug-8B	0.317902	0.455312	0.494359	0.736	0.593583	0.428	0.636	0.552	0.312	0.6	0.436	0.528	0.612	0.676	0.532	0.508	0.493151	0.552	0.62	0.42	0.539326	0.72	0.22	0.192	0.148	0.336	0.512	0.061934	0.302852	0.292929	0.302198	0.308036	0.470024	0.31793	0.470024	0.107492	0.056911	0.022727	0.061934	0.014286	0.045455	0.134715	0.014815	0.317902	0.398148	0.56	0.292969	0.344	0.31793	34.086982	LlamaForCausalLM	fe54a7d42160d3d8fcc3289c8c411fd9dd5e8357	llama2	91	12,367	8.030261	true	true	abacusai/Llama-3-Smaug-8B	2024-05-19T12:26:31.000Z	transformers	text-generation	False
meta-llama/Meta-Llama-Guard-2-8B	0.312084	0.415618	0.450269	0.728	0.604278	0.536	0.72	0.572	0.368	0.516	0.276	0.272	0.392	0.78	0.496	0.436	0.424658	0.408	0.512	0.272	0.657303	0.564	0.232	0.18	0.128	0.348	0.472	0.022659	0.29698	0.343434	0.283883	0.292411	0.03717	0.022181	0.03717	0.029316	0.01626	0.007576	0.022659	0.014286	0.032468	0.025907	0.02963	0.312084	0.338624	0.564	0.296875	0.156	0.022181	24.296443	LlamaForCausalLM	7d257f3c1a0ec6ed99b2cb715027149dfb9784ef	llama3	307	6,266	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-05-13T09:36:04.000Z	transformers	text-generation	manual
shanchen/llama3-8B-slerp-med-chinese	0.324219	0.438708	0.472661	0.704	0.513369	0.512	0.68	0.568	0.348	0.676	0.376	0.408	0.568	0.804	0.612	0.548	0.417808	0.364	0.476	0.412	0.505618	0.648	0.116	0.128	0.116	0.352	0.488	0.024169	0.311242	0.323232	0.298535	0.321429	0.292566	0.171904	0.292566	0.032573	0.03252	0.007576	0.024169	0.010714	0.019481	0.056995	0	0.324219	0.380952	0.504	0.3125	0.328	0.171904	30.096814	LlamaForCausalLM	f234f794019d98ace332115134997ba4224d8883	llama3	6	9,208	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-04-30T20:01:48.000Z	transformers	text-generation	False
MLP-KTLim/llama-3-Korean-Bllossom-8B	0.349235	0.430925	0.464329	0.684	0.57754	0.432	0.596	0.54	0.296	0.62	0.4	0.444	0.564	0.572	0.532	0.444	0.493151	0.576	0.428	0.468	0.550562	0.62	0.216	0.192	0.176	0.336	0.452	0.070997	0.298658	0.292929	0.29304	0.308036	0.476019	0.338262	0.476019	0.14658	0.04878	0.037879	0.070997	0.021429	0.051948	0.098446	0.037037	0.349235	0.384921	0.52	0.273438	0.364	0.338262	34.069317	LlamaForCausalLM	ed9647c18477ee09a03690c613c859eddca24362	llama3	388	4,202	8.03	true	true	MLP-KTLim/llama-3-Korean-Bllossom-8B (Merge)	2024-12-18T12:18:13.000Z	transformers	text-generation	False
nvidia/Llama3-ChatQA-1.5-8B	0.303939	0.436373	0.467801	0.768	0.566845	0.468	0.608	0.532	0.392	0.636	0.328	0.396	0.448	0.716	0.444	0.504	0.438356	0.476	0.512	0.38	0.606742	0.72	0.232	0.18	0.132	0.308	0.488	0.024924	0.297819	0.29798	0.291209	0.305804	0.209832	0.138632	0.209832	0.026059	0.02439	0.007576	0.024924	0.017857	0.019481	0.062176	0.007407	0.303939	0.415344	0.52	0.316406	0.412	0.138632	28.660992	LlamaForCausalLM	3b98162e3f97550d62aeeb19ea50208f968c678a	llama3	555	11,196	null	true	true	meta-llama/Meta-Llama-3-8B	2024-05-24T17:28:49.000Z	transformers	text-generation	False
dphn/dolphin-2.9.1-llama-3-8b	0.279754	0.44221	0.483944	0.74	0.57754	0.404	0.612	0.564	0.436	0.648	0.464	0.428	0.564	0.776	0.46	0.48	0.452055	0.508	0.548	0.352	0.573034	0.712	0.156	0.216	0.156	0.328	0.496	0.049094	0.258389	0.257576	0.252747	0.265625	0.390887	0.251386	0.390887	0.065147	0.01626	0.060606	0.049094	0.021429	0.025974	0.129534	0	0.279754	0.414021	0.56	0.316406	0.368	0.251386	31.268152	LlamaForCausalLM	924427715104c3667868d9297e4069b4c62bfd88	other	38	8,889	8.030278	true	true	meta-llama/Meta-Llama-3-8B	2024-05-20T14:36:52.000Z	transformers	text-generation	False
winninghealth/WiNGPT2-Llama-3-8B-Base	0.324884	0.438967	0.47214	0.708	0.518717	0.504	0.68	0.548	0.348	0.704	0.364	0.396	0.552	0.808	0.624	0.532	0.410959	0.384	0.472	0.424	0.505618	0.644	0.112	0.136	0.104	0.36	0.488	0.021903	0.317114	0.318182	0.304029	0.332589	0.305755	0.190388	0.305755	0.032573	0.02439	0.015152	0.021903	0.007143	0.032468	0.036269	0	0.324884	0.378307	0.5	0.3125	0.324	0.190388	30.33506	LlamaForCausalLM	8261856faa9e9cc7fc224c01289393d29ce98f64	apache-2.0	8	7,929	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2026-01-26T08:14:37.000Z	transformers	text-generation	False
scb10x/llama-3-typhoon-v1.5-8b-instruct	0.327377	0.444026	0.479604	0.772	0.540107	0.436	0.56	0.572	0.38	0.604	0.412	0.38	0.528	0.676	0.524	0.484	0.445205	0.512	0.644	0.444	0.494382	0.744	0.204	0.204	0.14	0.332	0.484	0.056647	0.281879	0.30303	0.272894	0.283482	0.514388	0.371534	0.514388	0.084691	0.02439	0.030303	0.056647	0.017857	0.064935	0.129534	0.014815	0.327377	0.428571	0.532	0.304688	0.452	0.371534	34.807781	LlamaForCausalLM	4c94d29c3c445f5ac324ead7d0a3e34b92fc42c9	llama3	29	9,041	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2025-03-31T10:47:50.000Z	transformers	text-generation	False
vicgalle/Configurable-Hermes-2-Pro-Llama-3-8B	0.324385	0.471657	0.505815	0.764	0.620321	0.368	0.652	0.552	0.384	0.664	0.424	0.42	0.588	0.72	0.632	0.46	0.520548	0.56	0.616	0.412	0.758427	0.764	0.252	0.176	0.14	0.308	0.492	0.064955	0.325503	0.328283	0.315018	0.337054	0.440048	0.288355	0.440048	0.110749	0.056911	0.015152	0.064955	0.032143	0.045455	0.119171	0.02963	0.324385	0.441799	0.548	0.417969	0.36	0.288355	35.041748	LlamaForCausalLM	82e7ca13eaa4e5110aee050f5e7076a1571fe604	apache-2.0	6	10,445	8.030523	true	true	NousResearch/Meta-Llama-3-8B	2024-07-31T16:38:50.000Z	transformers	text-generation	False
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half	0.366356	0.437022	0.476827	0.788	0.55615	0.52	0.5	0.548	0.324	0.74	0.348	0.432	0.532	0.556	0.42	0.516	0.520548	0.58	0.608	0.42	0.52809	0.684	0.196	0.172	0.124	0.38	0.504	0.058912	0.306208	0.292929	0.302198	0.316964	0.585132	0.451017	0.585132	0.104235	0.02439	0.022727	0.058912	0.010714	0.058442	0.124352	0.02963	0.366356	0.339947	0.5	0.242188	0.28	0.451017	35.556379	LlamaForCausalLM	b82150a9840ba5ba93918c745adc70afc6ad2ce1	cc-by-nc-4.0	16	9,149	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2024-05-30T09:58:00.000Z	transformers	text-generation	False
mlabonne/ChimeraLlama-3-8B-v3	0.368684	0.457128	0.497136	0.768	0.614973	0.468	0.64	0.552	0.32	0.568	0.4	0.452	0.596	0.556	0.52	0.468	0.486301	0.608	0.624	0.488	0.511236	0.768	0.312	0.22	0.188	0.32	0.512	0.068731	0.301174	0.328283	0.294872	0.296875	0.444844	0.295749	0.444844	0.100977	0.03252	0.068182	0.068731	0.010714	0.051948	0.15544	0.044444	0.368684	0.398148	0.54	0.289063	0.368	0.295749	34.645289	LlamaForCausalLM	ff1e5ff9844908d3f151e5f01eec2e804d7fa2ac	other	15	7,936	8.030261	true	false	mlabonne/ChimeraLlama-3-8B-v3 (Merge)	2024-09-17T10:16:45.000Z	transformers	text-generation	False
NousResearch/Hermes-2-Theta-Llama-3-8B	0.34857	0.485666	0.529248	0.784	0.614973	0.428	0.668	0.596	0.356	0.716	0.432	0.46	0.612	0.692	0.572	0.508	0.513699	0.62	0.636	0.508	0.662921	0.772	0.4	0.24	0.192	0.324	0.448	0.081571	0.314597	0.333333	0.304029	0.319196	0.444844	0.297597	0.444844	0.123779	0.04065	0.037879	0.081571	0.014286	0.058442	0.222798	0.02963	0.34857	0.42328	0.56	0.382813	0.328	0.297597	35.701862	LlamaForCausalLM	57a73110702e7b05ba3f39fef36297454c680725	apache-2.0	204	10,758	8.030261	true	true	NousResearch/Meta-Llama-3-8B	2024-09-08T08:11:20.000Z	transformers	text-generation	False
QuixiAI/Llama-3-8B-Instruct-abliterated-v2	0.345745	0.45181	0.487068	0.74	0.566845	0.468	0.62	0.544	0.364	0.544	0.42	0.436	0.608	0.592	0.604	0.44	0.493151	0.576	0.528	0.464	0.5	0.712	0.22	0.22	0.22	0.348	0.488	0.070242	0.297819	0.29798	0.305861	0.287946	0.453237	0.306839	0.453237	0.140065	0.056911	0.022727	0.070242	0.010714	0.058442	0.134715	0.014815	0.345745	0.425926	0.568	0.304688	0.408	0.306839	34.667279	LlamaForCausalLM	c99bd5b5e73f1d24504d1145b8a8dfd66c0a2378	llama3	29	8,177	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-05-12T17:50:18.000Z	transformers	text-generation	False
Kukedlc/NeuralLLaMa-3-8b-DT-v0.1	0.379737	0.459204	0.497309	0.788	0.59893	0.496	0.628	0.556	0.328	0.532	0.392	0.432	0.572	0.536	0.536	0.448	0.506849	0.58	0.584	0.488	0.52809	0.768	0.432	0.236	0.196	0.284	0.528	0.079305	0.308725	0.338384	0.29304	0.314732	0.472422	0.334566	0.472422	0.140065	0.02439	0.022727	0.079305	0.028571	0.071429	0.181347	0.014815	0.379737	0.406085	0.556	0.277344	0.388	0.334566	35.726392	LlamaForCausalLM	e69b4f7c30346e5666814a632f6c7af95dd740ae	other	1	7,951	8.030261	true	false	Kukedlc/NeuralLLaMa-3-8b-DT-v0.1 (Merge)	2024-09-22T19:07:24.000Z	transformers	text-generation	False
openchat/openchat-3.6-8b-20240522	0.341257	0.492411	0.53515	0.78	0.609626	0.44	0.768	0.54	0.424	0.64	0.416	0.448	0.58	0.768	0.644	0.484	0.452055	0.52	0.612	0.548	0.792135	0.752	0.616	0.176	0.156	0.264	0.472	0.064955	0.321309	0.338384	0.298535	0.341518	0.446043	0.308688	0.446043	0.107492	0.01626	0.022727	0.064955	0.021429	0.058442	0.145078	0.037037	0.341257	0.436508	0.568	0.378906	0.364	0.308688	35.753688	LlamaForCausalLM	2264eb98558978f708e88ae52afb78e43b832801	llama3	157	11,164	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-05-28T05:23:57.000Z	transformers	text-generation	False
hfl/llama-3-chinese-8b-instruct	0.268534	0.435724	0.469884	0.768	0.625668	0.416	0.644	0.596	0.236	0.488	0.456	0.42	0.56	0.788	0.58	0.48	0.445205	0.448	0.408	0.388	0.550562	0.696	0.24	0.148	0.092	0.32	0.536	0.029456	0.286074	0.262626	0.289377	0.292411	0.392086	0.255083	0.392086	0.045603	0.02439	0.015152	0.029456	0.021429	0.019481	0.051813	0.007407	0.268534	0.411376	0.5	0.328125	0.408	0.255083	30.956827	LlamaForCausalLM	06bd938075968adc98bc4080bfcd65a8c2a25250	apache-2.0	13	7,923	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-05-29T05:14:31.000Z	transformers	text-generation	False
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25	0.369016	0.450772	0.488457	0.796	0.609626	0.512	0.596	0.54	0.352	0.592	0.376	0.46	0.584	0.572	0.448	0.516	0.534247	0.576	0.636	0.488	0.505618	0.696	0.18	0.172	0.132	0.376	0.528	0.063444	0.316275	0.368687	0.315018	0.294643	0.591127	0.463956	0.591127	0.100977	0.00813	0.030303	0.063444	0.021429	0.032468	0.170984	0.02963	0.369016	0.375661	0.54	0.316406	0.272	0.463956	36.73301	LlamaForCausalLM	5a2f17238cc83932e00613d285f8bf6b8f4a0c3a	cc-by-nc-4.0	3	9,172	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2024-05-30T09:57:34.000Z	transformers	text-generation	False
shenzhi-wang/Llama3-8B-Chinese-Chat	0.34267	0.474121	0.518486	0.764	0.620321	0.488	0.676	0.56	0.392	0.684	0.448	0.476	0.612	0.632	0.616	0.488	0.527397	0.576	0.744	0.472	0.601124	0.744	0.116	0.208	0.176	0.352	0.524	0.070242	0.309564	0.323232	0.298535	0.316964	0.450839	0.301294	0.450839	0.104235	0.073171	0.037879	0.070242	0.014286	0.071429	0.150259	0.022222	0.34267	0.395503	0.552	0.339844	0.296	0.301294	34.788389	LlamaForCausalLM	f25f13cb2571e70e285121faceac92926b51e6f5	llama3	687	8,493	null	true	true	meta-llama/Meta-Llama-3-8B	2024-07-04T10:08:56.000Z	transformers	text-generation	False
hfl/llama-3-chinese-8b-instruct-v3	0.339511	0.46076	0.499392	0.768	0.631016	0.46	0.684	0.544	0.34	0.612	0.476	0.476	0.62	0.672	0.452	0.508	0.5	0.612	0.52	0.492	0.52809	0.76	0.196	0.192	0.148	0.316	0.52	0.044562	0.311242	0.323232	0.311355	0.305804	0.440048	0.304991	0.440048	0.058632	0.03252	0.015152	0.044562	0.014286	0.038961	0.124352	0.007407	0.339511	0.402116	0.524	0.328125	0.356	0.304991	33.947861	LlamaForCausalLM	91ca32fc684ce0d48ce7b29a342b16b7f9f1685c	apache-2.0	63	8,078	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2025-07-11T06:12:18.000Z	transformers	text-generation	False
scb10x/llama-3-typhoon-v1.5-8b	0.317736	0.424569	0.456692	0.772	0.518717	0.52	0.348	0.544	0.32	0.664	0.348	0.38	0.52	0.724	0.412	0.48	0.431507	0.468	0.516	0.44	0.595506	0.656	0.188	0.192	0.132	0.348	0.488	0.035498	0.297819	0.267677	0.29304	0.316964	0.231415	0.125693	0.231415	0.058632	0.01626	0.007576	0.035498	0.025	0.038961	0.062176	0.007407	0.317736	0.37963	0.508	0.292969	0.34	0.125693	28.646489	LlamaForCausalLM	c341c4a2b748336a53fa398ac566ec94374ad30a	llama3	9	8,083	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2025-03-31T10:47:10.000Z	transformers	text-generation	False
MaziyarPanahi/Llama-3-8B-Instruct-v0.8	0.382646	0.456739	0.494706	0.78	0.59893	0.5	0.616	0.572	0.344	0.552	0.384	0.348	0.528	0.508	0.58	0.42	0.513699	0.584	0.54	0.508	0.544944	0.796	0.448	0.212	0.204	0.288	0.552	0.062689	0.303691	0.343434	0.291209	0.301339	0.489209	0.343808	0.489209	0.09772	0.02439	0.030303	0.062689	0.025	0.084416	0.11399	0.02963	0.382646	0.40873	0.548	0.265625	0.416	0.343808	35.694516	LlamaForCausalLM	bd80951b7ae97f633ed48b80334af8df96b49f36	other	8	8,518	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2024-07-04T13:17:03.000Z	transformers	text-generation	False
Kukedlc/NeuralLLaMa-3-8b-ORPO-v0.4	0.357796	0.456739	0.490887	0.776	0.620321	0.468	0.68	0.552	0.288	0.696	0.352	0.404	0.56	0.624	0.512	0.464	0.486301	0.532	0.576	0.468	0.561798	0.736	0.284	0.184	0.172	0.344	0.492	0.053625	0.313758	0.318182	0.318681	0.305804	0.432854	0.28281	0.432854	0.100977	0.04065	0	0.053625	0.003571	0.032468	0.145078	0.007407	0.357796	0.421958	0.548	0.34375	0.376	0.28281	34.514634	LlamaForCausalLM	7aa29fb84b6138c3204cf295a2339242d26997af	apache-2.0	1	8,416	8.030278	true	true	meta-llama/Meta-Llama-3-8B	2024-09-22T19:11:58.000Z	transformers	text-generation	False
vicgalle/Configurable-Llama-3-8B-v0.3	0.365858	0.445064	0.486027	0.756	0.593583	0.464	0.612	0.572	0.348	0.544	0.388	0.428	0.528	0.54	0.512	0.444	0.465753	0.612	0.58	0.48	0.516854	0.728	0.312	0.22	0.192	0.348	0.508	0.062689	0.291107	0.282828	0.283883	0.303571	0.489209	0.347505	0.489209	0.114007	0.01626	0.022727	0.062689	0.025	0.058442	0.124352	0.022222	0.365858	0.375661	0.528	0.257813	0.344	0.347505	34.509178	LlamaForCausalLM	73b9c1bfc75301d5bcd84d85be0d995aaeec8fe2	apache-2.0	15	8,473	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-04-23T07:26:41.000Z	transformers	text-generation	False
MaziyarPanahi/Llama-3-8B-Instruct-v0.9	0.383893	0.461149	0.498872	0.78	0.604278	0.496	0.636	0.568	0.344	0.544	0.396	0.332	0.54	0.508	0.588	0.428	0.520548	0.6	0.556	0.516	0.550562	0.796	0.464	0.224	0.196	0.284	0.552	0.067221	0.307886	0.323232	0.305861	0.303571	0.47482	0.332717	0.47482	0.136808	0.03252	0.045455	0.067221	0.021429	0.058442	0.108808	0.007407	0.383893	0.415344	0.56	0.257813	0.432	0.332717	35.800586	LlamaForCausalLM	ddf91fdc0a3ab5e5d76864f1c4cf44e5adacd565	other	6	8,483	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2024-06-03T21:12:44.000Z	transformers	text-generation	False
UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3	0.361287	0.458944	0.498525	0.768	0.620321	0.464	0.68	0.576	0.336	0.56	0.416	0.436	0.604	0.468	0.592	0.472	0.5	0.568	0.68	0.496	0.516854	0.772	0.28	0.204	0.196	0.32	0.476	0.067221	0.309564	0.29798	0.300366	0.325893	0.484412	0.349353	0.484412	0.100977	0.03252	0.037879	0.067221	0.028571	0.051948	0.160622	0.014815	0.361287	0.392857	0.544	0.277344	0.36	0.349353	35.231084	LlamaForCausalLM	48c29bf2d9d68113255df9a47a9dabff6c67a13f	apache-2.0	83	8,435	8.030261	true	true	UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3	2024-06-28T19:36:28.000Z	transformers	text-generation	False
migtissera/Llama-3-8B-Synthia-v3.5	0.31142	0.462057	0.504253	0.824	0.572193	0.48	0.66	0.568	0.312	0.768	0.472	0.496	0.612	0.728	0.444	0.472	0.493151	0.548	0.596	0.404	0.58427	0.76	0.212	0.152	0.136	0.356	0.488	0.042296	0.30453	0.287879	0.307692	0.308036	0.436451	0.299445	0.436451	0.078176	0.02439	0.022727	0.042296	0.003571	0.019481	0.093264	0.02963	0.31142	0.388889	0.52	0.351563	0.296	0.299445	33.130638	LlamaForCausalLM	af4990801a24fee7acf16370cb5aa5643b5e9d6c	llama3	15	8,470	8.030261	true	true	migtissera/Llama-3-8B-Synthia-v3.5	2024-06-02T01:06:30.000Z	transformers	text-generation	False
Magpie-Align/Llama-3-8B-Magpie-Align-SFT-v0.3	0.306017	0.451291	0.492623	0.808	0.614973	0.5	0.576	0.6	0.412	0.68	0.38	0.444	0.532	0.74	0.524	0.456	0.486301	0.464	0.512	0.464	0.573034	0.692	0.284	0.188	0.156	0.308	0.48	0.049094	0.288591	0.242424	0.282051	0.316964	0.352518	0.227357	0.352518	0.071661	0.00813	0.022727	0.049094	0.014286	0.058442	0.124352	0.014815	0.306017	0.392857	0.496	0.339844	0.344	0.227357	31.361658	LlamaForCausalLM	d2578eb754d1c20efe604749296580f680950917	llama3	6	8,476	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-07-19T20:33:26.000Z	transformers	text-generation	False
johnsnowlabs/JSL-MedLlama-3-8B-v2.0	0.341922	0.449345	0.490193	0.764	0.572193	0.496	0.696	0.54	0.328	0.736	0.388	0.38	0.636	0.756	0.568	0.456	0.479452	0.52	0.516	0.428	0.573034	0.744	0.128	0.184	0.14	0.26	0.516	0.056647	0.298658	0.338384	0.282051	0.301339	0.280576	0.177449	0.280576	0.09772	0.03252	0.030303	0.056647	0.025	0.032468	0.129534	0	0.341922	0.375661	0.5	0.339844	0.288	0.177449	30.72759	LlamaForCausalLM	f3d41874ba4998c31727b8a18fae3e680f440f2f	cc-by-nc-nd-4.0	43	8,088	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-04-30T18:06:24.000Z	transformers	text-generation	False
GraySwanAI/Llama-3-8B-Instruct-RR	0.358045	0.449215	0.488283	0.76	0.647059	0.452	0.636	0.56	0.34	0.548	0.428	0.456	0.596	0.58	0.52	0.464	0.472603	0.592	0.58	0.484	0.550562	0.704	0.196	0.216	0.168	0.324	0.496	0.067976	0.302852	0.318182	0.298535	0.301339	0.51199	0.36414	0.51199	0.100977	0.04878	0.037879	0.067976	0.025	0.051948	0.15544	0.022222	0.358045	0.382275	0.536	0.261719	0.352	0.36414	35.19037	LlamaForCausalLM	d92f951d380d3489fb56b08c296376ea61cebef0	null	15	17,541	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-07-09T03:05:36.000Z	transformers	text-generation	False
swap-uniba/LLaMAntino-3-ANITA-8B-Inst-DPO-ITA	0.373753	0.455831	0.491061	0.764	0.561497	0.524	0.624	0.572	0.356	0.584	0.388	0.32	0.588	0.476	0.564	0.376	0.5	0.596	0.62	0.46	0.55618	0.804	0.42	0.196	0.228	0.288	0.46	0.054381	0.295302	0.30303	0.291209	0.296875	0.520384	0.367837	0.520384	0.091205	0.02439	0.030303	0.054381	0.014286	0.090909	0.093264	0.007407	0.373753	0.440476	0.516	0.328125	0.48	0.367837	36.255941	LlamaForCausalLM	5f5c90b197446b7ca4dfbfa099fcf3566edf2e5c	llama3	30	12,023	8.030261	true	true	meta-llama/Meta-Llama-3-8B-Instruct	2025-09-01T11:05:55.000Z	transformers	text-generation	False
Henrychur/MMed-Llama-3-8B	0.311253	0.444675	0.479083	0.776	0.582888	0.48	0.56	0.532	0.396	0.696	0.412	0.388	0.508	0.764	0.58	0.484	0.445205	0.42	0.652	0.372	0.573034	0.74	0.068	0.152	0.124	0.344	0.488	0.042296	0.298658	0.30303	0.294872	0.301339	0.247002	0.142329	0.247002	0.084691	0.01626	0.015152	0.042296	0.017857	0.019481	0.067358	0.037037	0.311253	0.412698	0.516	0.382813	0.34	0.142329	29.849857	LlamaForCausalLM	6c3057bb49ac499970eb2891daaef9b5c14f6943	llama3	32	3,077	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-05-24T14:43:50.000Z	transformers	text-generation	False
Orenguteng/Llama-3-8B-Lexi-Uncensored	0.350898	0.443897	0.483423	0.764	0.609626	0.448	0.632	0.528	0.368	0.556	0.424	0.464	0.576	0.6	0.5	0.44	0.506849	0.552	0.62	0.444	0.522472	0.676	0.2	0.224	0.192	0.324	0.484	0.071752	0.30453	0.323232	0.305861	0.294643	0.482014	0.341959	0.482014	0.107492	0.02439	0.022727	0.071752	0.021429	0.103896	0.124352	0.074074	0.350898	0.362434	0.512	0.285156	0.292	0.341959	34.250856	LlamaForCausalLM	ff95e3bfcd6142759ce82099b58bc7a789ac241b	llama3	273	1,603	8.030261	true	true	meta-llama/Meta-Llama-3-8B	2024-05-27T06:16:40.000Z	transformers	text-generation	False
Iker/Llama-3-Instruct-Neurona-8b	0.309674	0.420158	0.451311	0.776	0.588235	0.324	0.716	0.468	0.328	0.584	0.388	0.332	0.548	0.632	0.516	0.48	0.520548	0.532	0.424	0.384	0.466292	0.46	0.26	0.192	0.16	0.332	0.488	0.048338	0.285235	0.328283	0.274725	0.279018	0.492806	0.362292	0.492806	0.061889	0.02439	0.007576	0.048338	0.014286	0.058442	0.129534	0.022222	0.309674	0.395503	0.528	0.308594	0.352	0.362292	33.047773	LlamaForCausalLM	0087ae77f1907e47ef60be6a9f50b182acf09781	llama3	3	7	8.030327	true	true	meta-llama/Meta-Llama-3-8B	2024-06-02T08:04:03.000Z	transformers	text-generation	False
ruslanmv/Medical-Llama3-8B	0.315824	0.420029	0.452352	0.784	0.524064	0.488	0.6	0.54	0.428	0.632	0.304	0.248	0.416	0.78	0.516	0.476	0.390411	0.368	0.512	0.384	0.52809	0.692	0.092	0.192	0.152	0.336	0.488	0.028701	0.302013	0.282828	0.291209	0.323661	0.217026	0.107209	0.217026	0.029316	0.03252	0.007576	0.028701	0.014286	0.025974	0.072539	0.014815	0.315824	0.359788	0.512	0.324219	0.244	0.107209	27.928426	LlamaForCausalLM	82cb0a9d1d30a9532253d21c2793f7163535357b	apache-2.0	105	508	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-05-15T08:53:33.000Z	transformers	text-generation	False
ytu-ce-cosmos/Turkish-Llama-8b-v0.1	0.299867	0.441302	0.476653	0.76	0.625668	0.528	0.636	0.532	0.4	0.572	0.4	0.372	0.592	0.8	0.58	0.468	0.438356	0.388	0.468	0.368	0.651685	0.776	0.088	0.144	0.108	0.304	0.512	0.033988	0.28943	0.262626	0.29304	0.296875	0.275779	0.138632	0.275779	0.04886	0.01626	0.007576	0.033988	0.021429	0.045455	0.062176	0.014815	0.299867	0.411376	0.504	0.324219	0.408	0.138632	29.784881	LlamaForCausalLM	8493b0c54fb98b35726dc9d500825d696c8cde4d	llama3	59	100	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-12-03T17:52:40.000Z	transformers	text-generation	False
unsloth/DeepSeek-R1-Distill-Llama-8B-bnb-4bit	0.264628	0.393825	0.420413	0.808	0.545455	0.424	0.228	0.54	0.236	0.488	0.368	0.392	0.456	0.692	0.604	0.4	0.273973	0.444	0.428	0.36	0.533708	0.532	0.264	0.152	0.092	0.348	0.484	0.364048	0.275168	0.247475	0.283883	0.276786	0.431655	0.280961	0.431655	0.628664	0.300813	0.219697	0.364048	0.117857	0.480519	0.528497	0.103704	0.264628	0.378307	0.52	0.304688	0.312	0.280961	35.570308	LlamaForCausalLM	80630da672d7df44498d6d58b2da22f3fb00d98b	llama3.1	3	5,463	8.248929	true	true	meta-llama/Meta-Llama-3-8B	2025-07-18T13:05:58.000Z	transformers	text-generation	False
rinna/llama-3-youko-8b	0.277344	0.401998	0.428919	0.788	0.561497	0.536	0.308	0.536	0.36	0.492	0.348	0.28	0.42	0.816	0.58	0.504	0.376712	0.36	0.464	0.356	0.5	0.492	0.056	0.168	0.128	0.312	0.584	0.019637	0.291946	0.277778	0.291209	0.299107	0.229017	0.12939	0.229017	0.032573	0.00813	0.007576	0.019637	0.014286	0.012987	0.031088	0.014815	0.277344	0.37037	0.516	0.246094	0.352	0.12939	26.953888	LlamaForCausalLM	6af890bc6294b8f311e25a1462fcd38554d95b8e	llama3	62	242	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2025-03-23T11:40:25.000Z	transformers	text-generation	False
walid-iguider/Llama-3-8B-4bit-UltraChat-Ita	0.316822	0.436633	0.469884	0.776	0.550802	0.472	0.544	0.576	0.336	0.676	0.344	0.328	0.512	0.792	0.588	0.54	0.39726	0.372	0.448	0.36	0.522472	0.78	0.204	0.224	0.14	0.312	0.488	0.018127	0.303691	0.277778	0.307692	0.310268	0.308153	0.190388	0.308153	0.032573	0.01626	0.007576	0.018127	0.014286	0.006494	0.031088	0	0.316822	0.392857	0.556	0.351563	0.272	0.190388	30.158905	LlamaForCausalLM	473a2ee0296e1471ae635fd8977dd3b44404f0e8	apache-2.0	0	2,232	8.030261	true	false	meta-llama/Meta-Llama-3-8B	2024-05-06T07:26:05.000Z	transformers	text-generation	False

End of preview. Expand in Data Studio

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Proteus-2k

Proteus-2k is a large-scale benchmark table of recent open-weight language models evaluated with the Open LLM Leaderboard v2 pipeline. It was built to extend public leaderboards after freeze dates and to support research on how compute–capability relationships hold up as model families and post-training evolve.

Dataset (full): hlzhang109/proteus-2k — CSV: proteus_2k.csv (~2.4k model checkpoints).
Selected subset: hlzhang109/proteus-selected — CSV: proteus_2k_selected.csv (curated rows for specific analyses).

Overview

Proteus-2k is the language-model evaluation dataset constructed and open-sourced in the companion paper. It is meant to fill gaps when public leaderboards (for example the Open LLM Leaderboard) stop updating or fall behind, and to test whether capability boundaries grounded in compute stay valid and stable over time.

At a glance:

Scale and timeframe: Roughly 2.4k recently released open-weight models, with the earliest from (2022-09-26) and mostly from after the Open LLM Leaderboard v2 cutoff (2025-03-13) through late 2025.
Model families: Many frontier architectures that rarely appeared on older leaderboard tables—for example Qwen3, Gemma 3, and GPT-OSS, plus (per the paper appendix) Llama 3.2, Mistral-7B-v0.3, Allen AI’s OLMo-3, and NVIDIA’s Nemotron, among others.
Evaluation: The authors evaluated all ~2.4k models and followed the Open LLM Leaderboard evaluation pipeline exactly so results stay aligned with historical runs.
Research use: The paper studies upper bounds on performance at a given compute budget. Proteus-2k is a temporal out-of-distribution (OOD) benchmark: fit an envelope on older models, then check whether the latest models still respect compute-based limits as techniques and post-training evolve.
Release: Full table and subset on Hugging Face: hlzhang109/proteus-2k, hlzhang109/proteus-selected.

In short, Proteus-2k is a large-scale, up-to-date open-model table with standardized leaderboard evaluations, aimed at tracing how compute maps to real downstream performance as the field moves forward.

Files in this directory

File	Role
`proteus_2k.csv`	Full table (~2.4k rows); uploaded to `hlzhang109/proteus-2k`.
`proteus_2k_selected.csv`	Selected subset; uploaded to `hlzhang109/proteus-selected`.
`upload_data.py`	Helper to push these CSVs to Hugging Face (`huggingface_hub`).

Schema

Each row is one Hugging Face model repo (or checkpoint) identified by model_id.

Scores — columns prefixed with leaderboard_ are per-task or aggregate metrics from the Open LLM Leaderboard v2 suite (e.g. BBH subtasks, GPQA, IFEval, MATH-hard, MMLU-Pro, MUSR, exact-match aggregates). The column Average ⬆️ is the leaderboard-style overall average used for ranking.

Hub metadata (examples): Architecture, Model sha, Hub License, Hub downloads, #Params (B), Available on the hub, Chat Template, Base Model, Hub lastModified, library_name, pipeline_tag, gated.

Task names follow the leaderboard’s naming; see Open LLM Leaderboard documentation for benchmark definitions.

Loading with Hugging Face Datasets

from datasets import load_dataset

ds = load_dataset("hlzhang109/proteus-2k", data_files="proteus_2k.csv")
# or the selected split/repo:
# ds = load_dataset("hlzhang109/proteus-selected", data_files="proteus_2k_selected.csv")
df = ds["train"].to_pandas()

You can also download the CSV from the dataset repo’s Files tab and work with pandas / any CSV tooling.

Citations and licenses

If you use Proteus-2k, please cite the paper:

@misc{zhang2026prescriptive,
      title={Prescriptive Scaling Reveals the Evolution of Language Model Capabilities},
      author={Hanlin Zhang and Jikai Jin and Vasilis Syrgkanis and Sham Kakade},
      year={2026},
      eprint={2602.15327},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2602.15327},
}

Per-model licensing varies (Hub License column and each model card on the Hub). The dataset aggregates publicly reported evaluation numbers and metadata; it does not redistribute model weights.

Downloads last month: 37

Paper for hlzhang109/proteus-2k

Prescriptive Scaling Reveals the Evolution of Language Model Capabilities

Paper • 2602.15327 • Published Feb 17 • 3