User profile

rocm-rider

18 runsFirst seen 2026-04-10Avg 66.3

Best PipelineScore

81.5MAINLINE

on Qwen 3 72B Instruct

Total tokens98.0KAcross every task this user has run

Avg latency918msPer task, across all submissions

Tasks run45018 submissions x ~25 tasks

Rigs used12Distinct hardware tags

Category signature

Average score per category across all 18 runs.

Code

66.0

Reason

66.3

Write

67.5

Tool Use

66.3

RAG

67.5

Speed

64.6

Hardware mix

Rigs this user benchmarked on.

a100-80gb3 (17%)

ryzen-7950x-rtx-30902 (11%)

rtx-4070-12gb2 (11%)

ryzen-7950x-cpu-only2 (11%)

rtx-3060-12gb2 (11%)

a6000-48gb1 (6%)

dgx-h1001 (6%)

m3-max-64gb1 (6%)

rtx-4080-16gb1 (6%)

rtx-3080-10gb1 (6%)

m2-pro-16gb1 (6%)

m2-air-16gb1 (6%)

Provider mix

Where they spend their tokens.

deepseek3 (17%)

alibaba2 (11%)

meta2 (11%)

openchat2 (11%)

moonshot1 (6%)

mistral1 (6%)

yi1 (6%)

ibm1 (6%)

cognitivecomputations1 (6%)

bigcode1 (6%)

cohere1 (6%)

community1 (6%)

huggingface1 (6%)

Models tried

Best score per model. Click a model to see its full page.

#	Model	Provider	Best Score	Tier	Achieved
1	Qwen 3 72B Instruct	alibaba	81.5	MAINLINE	2026-05-04
2	DeepSeek V2.5	deepseek	77.5	MAINLINE	2026-05-16
3	DeepSeek Coder V2 16B	deepseek	76.9	MAINLINE	2026-04-18
4	Kimi K2 Instruct	moonshot	76.4	MAINLINE	2026-04-26
5	Llama 3.1 70B Instruct	meta	75.2	MAINLINE	2026-05-07
6	Codestral 22B	mistral	73.4	FEEDER	2026-05-06
7	Yi 1.5 34B Chat	yi	71.3	FEEDER	2026-04-14
8	Qwen 2.5 VL 7B	alibaba	70.3	FEEDER	2026-04-17
9	Granite 3.1 8B Instruct	ibm	69.6	FEEDER	2026-05-02
10	OpenChat 3.6 8B	openchat	67.0	FEEDER	2026-04-28
11	Dolphin 3.0 Llama 3.1 8B	cognitivecomputations	63.7	FEEDER	2026-04-11
12	StarCoder2 7B	bigcode	61.3	FEEDER	2026-04-20
13	Aya 23 8B	cohere	59.6	TAP	2026-05-23
14	LLaVA 1.6 Mistral 7B	community	52.4	TAP	2026-05-14
15	SmolLM2 1.7B	huggingface	43.0	TAP	2026-04-14
16	Llama 3.2 1B Instruct	meta	38.3	DRIP	2026-05-08

All submissions

Every run, ordered by score.

#	Model	Hardware	Score	Tier	Code	Reason	Write	Tool Use	RAG	Speed	Tokens	Avg ms	Date
1	Qwen 3 72B Instruct	a100-80gb	81.5	MAINLINE	81.6	79.0	81.1	79.5	86.9	82.8	5.2K	869	2026-05-04
2	DeepSeek V2.5	a100-80gb	77.5	MAINLINE	77.8	71.7	76.9	77.7	82.0	81.8	5.2K	899	2026-05-16
3	DeepSeek Coder V2 16B	a6000-48gb	76.9	MAINLINE	81.7	75.7	80.8	73.5	78.5	67.3	6.1K	1510	2026-04-18
4	Kimi K2 Instruct	dgx-h100	76.4	MAINLINE	69.6	79.7	80.7	80.8	81.0	70.1	5.7K	905	2026-04-26
5	Llama 3.1 70B Instruct	a100-80gb	75.2	MAINLINE	73.7	78.2	67.1	78.8	77.7	76.3	6.4K	851	2026-05-07
6	DeepSeek Coder V2 16B	m3-max-64gb	74.0	FEEDER	72.0	75.0	80.7	74.0	71.1	71.2	5.3K	1479	2026-05-10
7	Codestral 22B	ryzen-7950x-rtx-3090	73.4	FEEDER	69.4	80.0	73.0	72.4	78.6	67.6	5.5K	1507	2026-05-06
8	Yi 1.5 34B Chat	ryzen-7950x-rtx-3090	71.3	FEEDER	71.0	69.7	77.9	66.5	75.7	67.8	4.9K	1402	2026-04-14
9	Qwen 2.5 VL 7B	rtx-4080-16gb	70.3	FEEDER	67.6	77.0	68.0	71.8	67.6	68.7	4.7K	824	2026-04-17
10	Granite 3.1 8B Instruct	rtx-4070-12gb	69.6	FEEDER	67.0	70.5	67.2	69.3	76.3	70.1	5.2K	797	2026-05-02
11	OpenChat 3.6 8B	ryzen-7950x-cpu-only	67.0	FEEDER	69.8	66.2	71.6	63.5	63.4	64.5	5.8K	730	2026-04-28
12	Dolphin 3.0 Llama 3.1 8B	rtx-3060-12gb	63.7	FEEDER	68.4	58.8	66.0	62.1	63.5	61.6	5.8K	818	2026-04-11
13	OpenChat 3.6 8B	rtx-3060-12gb	63.0	FEEDER	60.1	65.7	56.6	67.9	67.2	62.3	5.0K	779	2026-04-10
14	StarCoder2 7B	rtx-4070-12gb	61.3	FEEDER	63.2	60.5	63.8	59.7	61.2	58.2	5.0K	844	2026-04-20
15	Aya 23 8B	rtx-3080-10gb	59.6	TAP	58.9	57.9	64.5	65.5	54.5	56.1	5.6K	823	2026-05-23
16	LLaVA 1.6 Mistral 7B	m2-pro-16gb	52.4	TAP	51.8	52.4	52.7	51.4	53.3	53.4	5.6K	813	2026-05-14
17	SmolLM2 1.7B	m2-air-16gb	43.0	TAP	45.9	43.3	43.5	38.6	39.0	45.0	5.3K	346	2026-04-14
18	Llama 3.2 1B Instruct	ryzen-7950x-cpu-only	38.3	DRIP	38.7	32.7	43.6	39.5	38.1	38.8	5.6K	334	2026-05-08