Hugging Face's logo

Spaces:
steampunque
/
benchlm
Running

App Files Community
benchlm

Commit History

Add phi-4 Q4_K_H result in general model section
b401a9e

steampunque commited on

Refresh ultravox-v0_5-deepseek-r1-llama-3_1-8b result with new quant
1e44766

steampunque commited on

Add refreshed ultravox llama3.1 8B results, add llama3.1 8B Q4_K_H result.
664f945

steampunque commited on

Add BBHA results to BBA, refresh Voxtral Mini 3B result
6268b17

steampunque commited on

Add Ministral 3 3B Instruct 2512 Q4_K_H and Q6_K_H vision results
c7bef93

steampunque commited on

Add Ministral 3 8B Instruct 2512 vision result
e18cafd

steampunque commited on

Add GLM Z1 9B 0414 Q4_P_H partial result
356f924

steampunque commited on

Add Ministral-3-14B-Instruct-2512.Q4_K_H vision result
d3181cc

steampunque commited on

Add GLM Z1 9B 0414 Q4_K_H math partial results
7754d96

steampunque commited on

Add Qwen3 VL 8B Instruct Q4_K_H vision result
3069f5c

steampunque commited on

fix vison version numbering error
abe00f8

steampunque commited on

Correct engine version for MiniCPM V Q6_K_H rerun.
838ef77

steampunque commited on

Add MiniCPM-V-4_5 Q4_K_H vision result, rerun Q6_K_H vision result. Rerun LFM2-VL-1.6B vision result.
5733735

steampunque commited on

Add Q4_K_H and Q6_K_H quant results for Qwen 2.5 Omni 3B and VL 3B,
1094d35

steampunque commited on

Add Q4_K_H quant results for Qwen 2.5 VL 7B and Qwen 2.5 Omni 7B
e5824d1

steampunque commited on

Update Qwen 2.5 VL 32B result
4c2ec8f

steampunque commited on

Correct engine version for Qwen3 VL 32B Instruct run
70ee647

steampunque commited on

Add Qwen3 VL 32B Instruct Q4_K_H vision result
1cd23e2

steampunque commited on

Add Qwen 2.5 Coder 14B Instruct Q4_K_H results
6e9ac68

steampunque commited on

Refresh CRUXEVALFIM bench, rerun for Qwen 2.5 Coder 7B Instruct Q4_K_H/Q6_K_H
593811c

steampunque commited on

Add Qwen 2.5 Coder 7B Instruct Q4_K_H/Q6_K_H results. Fix model size
c308807

steampunque commited on

Add Qwen2.5 Omni 3B vision and audio results. Add Qwen3 VL 8B Thinking vision result.
5ad9c28

steampunque commited on

Add Qwen3-VL-30B-A3B-Instruct vision result
ac21810

steampunque commited on

Update Qwen 2.5 Omni 7B vision result with mtmd fixes for Qwen.
1ffcb48

steampunque commited on

Update Qwen 2.5 VL 3B result with mtmd fixes for Qwen.
21285fa

steampunque commited on

Add Qwen3 VL 2B/4B Instruct vision results
ee874eb

steampunque commited on

Add Qwen3-VL-8B-Instruct vision result
d7d4780

steampunque commited on

Add REALWORLDQA vision test. Update Qwen 2.5 VL 7B result with mtmd fixes for Qwen.
f24e884

steampunque commited on

Add Ring 2.0 mini thinking model result
d6077a9

steampunque commited on

Add LFM2-VL-1.6B results, update MiniCPM-V-4_5 results
ad825ba

steampunque commited on

Add Qwen3 thinking 2507 results, clarify think model test methodology in description
9fee3e0

steampunque commited on

Add Qwen3 4B Instruct 2507
329867a

steampunque commited on

Correct parameters for minicpm v 4_5
42ba7fe

steampunque commited on

restore vision: header
f8db2b2

steampunque commited on

Add minicpm V 4_5 vision results
488685d

steampunque commited on

Correct sports understanding for QwQ 32B
67c8dc8

steampunque commited on

Add QwQ 32B Q4_K_H full bench results
0f4b3a3

steampunque commited on

Update QwQ 32b Q4_K_H result
cccfefe

steampunque commited on

Add Qwen3 Coder 30B A3B instruct code results. Remove obsolete code models.
a6f6477

steampunque commited on

Add final llama scout vision results
99222be

steampunque commited on

Add Llama scout vision, QwQ hybrid quants. Remove more obsolete models.
00ada6c

steampunque commited on

Remove invalid Llama 3.2 1B ultravox result since llama3.2 1B cannot reliably self grade.
66f8a4a

steampunque commited on

Blank invalid self test results for Llama 3.2 ultravox.
ee18787

steampunque commited on

Update self grade prompt so self grading for Llama 3.1 8b works, regen result.
75f318c

steampunque commited on

Add math categores, GLMZ1 9B results
491f3c6

steampunque commited on

Add Deepseek R1 Distill Llama 8B
658840c

steampunque commited on

Update Qwen2.5 VL 32B results
3e26048

steampunque commited on

New vision results
ea3b3c2

steampunque commited on

Add deepseek hybrid quant ultravox audio results
45b41ed

steampunque commited on

update all audio results using CoT and add some new ultravox models
65c038e

steampunque commited on