宋小猫

SongXiaoMao's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in z-lab/Qwen3.5-27B-DFlash 3 days ago

FP8 work for base model or is 16-bit of 27B required?

14

#2 opened 17 days ago by

unoid

New activity in Jackrong/Qwopus3.5-27B-v3 12 days ago

Is there anyone who can tell me how to run this model with vllm correctly?

😔 3

7

#8 opened 14 days ago by

beginor

New activity in olka-fi/Qwen3.5-27B-MXFP4 15 days ago

Can the big guy quantify this model into MXFP4? Thank you!!

#3 opened 15 days ago by

SongXiaoMao

New activity in Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled 15 days ago

How does the VLLM start this model?

2

#4 opened about 1 month ago by

SongXiaoMao

New activity in olka-fi/Qwen3.5-122B-A10B-MXFP4 17 days ago

This quantization model is amzing

❤️👍 2

5

#1 opened about 1 month ago by

hyunw55

New activity in olka-fi/Qwen3.5-27B-MXFP4 18 days ago

Why is the file size of 4bit similar to FP8?

3

#2 opened 18 days ago by

SongXiaoMao

New activity in edp1096/Huihui-Qwen3.5-27B-abliterated-FP8 18 days ago

Sensitive information is not a question

2

#3 opened 20 days ago by

SongXiaoMao

New activity in win10/Huihui-Qwen3.5-27B-abliterated-FP8 20 days ago

VLLM 0.18.0 runs with an error

#2 opened 20 days ago by

SongXiaoMao

New activity in groxaxo/Huihui-Qwen3.5-27B-W8A8-INT8 23 days ago

I get an error using vllm0.18.0

1

#1 opened 23 days ago by

SongXiaoMao

New activity in huihui-ai/Huihui-Qwen3.5-27B-Claude-4.6-Opus-abliterated 28 days ago

使用VLLM启动会报错

#3 opened 28 days ago by

SongXiaoMao

New activity in Jackrong/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled about 1 month ago

Tokenizer class TokenizersBackend does not exist in vllm v0.17.1

12

#26 opened about 1 month ago by

putcn

New activity in huihui-ai/Huihui-Qwen3.5-122B-A10B-abliterated-GGUF about 1 month ago

Can you make a quantitative model? Qwen3.5-122B-A10B-GPTQ-Int4

#2 opened about 1 month ago by

SongXiaoMao

New activity in Qwen/QwQ-32B about 1 year ago

When will you fix the model replies missing</think>\n start tags

18

#19 opened about 1 year ago by

xldistance

When answering questions in Chinese, the model frequently terminates prematurely (outputs the end token). Is this a common problem?

1

#40 opened about 1 year ago by

zhangw355

missing opening <think>

20

#4 opened about 1 year ago by

chriswritescode

New activity in Valdemardi/DeepSeek-R1-Distill-Llama-70B-AWQ about 1 year ago

AWQ q6

1

#1 opened about 1 year ago by

D-r-e

New activity in unsloth/DeepSeek-R1-GGUF about 1 year ago

I tested dynamic 1.58bit and 2.22bit, All thoughts are empty?

9

#24 opened about 1 year ago by

SongXiaoMao

No think tokens visible

6

#15 opened about 1 year ago by

sudkamath

New activity in Tiiny/SmallThinker-3B-Preview over 1 year ago

How to Pair with Larger Models

4

#7 opened over 1 year ago by

windkkk

New activity in Qwen/QwQ-32B-Preview over 1 year ago

multi GPU inferencing

2

#18 opened over 1 year ago by

cjj2003