不同LLM分析SDC得效果

everkevin2天前未分类3



5256 input token

tokens消耗


效果

QWQ-32B

5K

2.22 tok/sec • 547   tokens  •  151.99s to first token • Stop reason: User   Stopped

勉强

grok-3-gemma3-12b-distilled

7K

1.91 tok/sec • 296   tokens • 74.96s to first token •Stop reason: User Stopped

5.95 tok/sec • 371   tokens • 102.78s to first token • Stop reason: User Stopped

 

勉强

claude-3.7-sonnet-reasoning-gemma3-12b.Q8_0

14K

3.42 tok/sec•1096   tokens•326.91s to first token•Stop reason: EOS Token Found

小错误

QWEN3   14B

8K

21.32 tok/sec•3545   tokens•5.63s to first token•Stop reason: EOS Token Found

输出最好的

输出速度感觉不错

 

Lmstudio/deepseek-r1-0528-qwen3-8b

12K

3.56 tok/sec•7456   tokens•51.63s to first token•Stop reason: EOS Token Found

有点答非所问

错误很少

google/gemma-3-27b

9K

0.95 tok/sec •   2292 tokens • 179.37s to first token •Stop reason: EOS Token Found

错误太多,简直没法用

qwen/qwen3-32b 总体感觉不如30b-a3b

9K

3.62 tok/sec •   4390 tokens • 24.33s to first token • Stop reason: EOS Token Found

输出结果好



11.23 token/s•2761 token•首个token用时 11.66 s•停止原因: 检测到 EOS token

双卡

qwen/qwen3-32b 8bit


5.04   token/s•2225 token•首个token用时 32.17 s•停止原因: 检测到 EOS token

AI MAX 395

google/gemma-3-12b

14k

2.63 tok/sec•417   tokens•304.28s to first token•Stop reason: User Stopped

错误太多

tifa-deepsex-14b-cot

7k

1.30 tok/sec •2129   tokens•96.02s to first token•Stop reason: EOS Token Found

小错误

Qwen3-8b

7K

32.84 tok/sec•2293   tokens•3.32s to first token•Stop reason: EOS Token Found

错误较多速度快呀

处理token也快

MaziyarPanahi/DeepSeek-R1-0528-Qwen3-8B-GGUF/DeepSeek-R1-0528-Qwen3-8B.Q4_K_M

8K

4.38 tok/sec•3039   tokens•40.89s to first token•Stop reason: EOS Token Found

计算错误,勉强

microsoft/phi-4-reasoning-plus

16K

17.64   tok/sec•10947 tokens•5.63s to first token•Stop reason: EOS Token Found

输出结果好,速度不错

microsoft/phi-4-mini-reasoning

8K

15.83 tok/sec•3698   tokens•2.01s to first token•Stop reason: EOS Token Found

答非所问

deepseek-r1-distill-qwen-14b

6K

1.40 tok/sec•1267   tokens•72.01s to first token•Stop reason: User Stopped


bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF

6K

1.74 tok/sec•859   tokens•71.33s to first token•Stop reason: User Stopped


mistralai/mistral-7b-instruct-v0.3

8K

38.04   tok/sec •1008 tokens•4.75s to   first token•Stop reason: EOS Token Found

速度快回答简单但是没有错误

Hack337/ChatGPT-5-Q8_0-GGUF/chatgpt-5-q8_0.gguf

6K

74.52 tok/sec•835   tokens•0.77s to first token•Stop reason: EOS Token Found

响应速度飞快,英语好,计算能力几乎为0

deepseek-r1-distill-llama-70b

5K

2.47 tok/sec•779   tokens•272.53s to first token•Stop reason: EOS Token Found

回答很差,半半拉拉

qwen/qwen3-30b-a3b 4bit

9K

24.18 tok/sec•4040   tokens•5.62s to first token•Stop reason: EOS Token Found

输出最好的

输出速度感觉不错



33.35   token/s•4239 token•首个token用时 48.21 s•停止原因: 检测到 EOS token

AI MAX 395   volkun

qwen/qwen3-30b-a3b 8bit


27.14   token/s•3425 token•首个token用时 50.99 s•停止原因: 检测到 EOS token

AI MAX 395   volkun

qwen3-30b-a3b@q8_0

7k

7.18   tok/sec•2370 词元•23.28 到第一个词元停止原因: 发现EOS词元

没有错误,结果还行

gemma-3n-e4b-it-text

13K

31.42 token/s 1303 token•首个token用时 4.37 s•停止原因: 检测到 EOS token

结果不行,双卡

google/gemma-3-27b


15.32   token/s•1593 token•首个token用时 10.46 s•停止原因: 检测到 EOS token

结果还行,没有错误,双卡

llama4-dolphin-8b


39.72   token/s•154 token•首个token用时 4.60 s•停止原因: 检测到 EOS token

答非所问

goedel-prover-v2-8b@f16


8.95   token/s•2586 token•首个token用时 21.65 s•停止原因: 检测到 EOS token

答非所问

goedel-prover-v2-8b@q8_0


15.51   token/s•2715 token•首个token用时 10.00 s•停止原因: 检测到 EOS token

答非所问


返回列表

没有更早的文章了...

没有最新的文章了...

发表评论    

◎欢迎参与讨论,请在这里发表您的看法、交流您的观点。