Steve Li

CHNtentes's activity

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

New activity in cyankiwi/Qwen3.8-27B-AWQ-INT4 about 2 months ago

ship the MTP head as a standalone mtp.safetensors (like official FP8)

🔥👍 2

2

#2 opened about 2 months ago by

NKLAR5

New activity in deepseek-ai/DeepSeek-V4-Flash-0731 about 2 months ago

# 0731 KV cache per-token is ~8× larger than preview (56 vs 7.2 bytes/token) - is this expected?

👀 1

2

#40 opened 2 months ago by

dfsafdsafxcvzxv

New activity in unsloth/DeepSeek-V4-Flash-0731-GGUF 2 months ago

Pulling UD_Q8 right now, MTP questions lol

5

#3 opened 2 months ago by

Chuck572

New activity in unsloth/Kimi-K3-GGUF 2 months ago

1-bit Kimi K3 vs Claude Opus 5 vs GPT 5.6

👀🚀 20

25

#12 opened 2 months ago by

danielhanchen

Can i run this on my macbook?

👀🤯 4

5

#5 opened 2 months ago by

hugginface536

New activity in prism-ml/Bonsai-27B-gguf 3 months ago

Disappointed by inference speed compared to a usual MoE on my laptop.

🧠 1

5

#10 opened 3 months ago by

extrabigmehdi

New activity in unsloth/Qwen-AgentWorld-35B-A3B-GGUF 3 months ago

mmproj file so small!

7

#1 opened 3 months ago by

CHHORVORN

New activity in Comfy-Org/Ideogram-4 4 months ago

Flash attention optimization for significant speedup. - old title: Optimization tips to maximize generation speed?

16

#6 opened 4 months ago by

eepos

New activity in tencent/Hy-MT2-1.8B 4 months ago

You open-sourced my ass - 你“开源”我的屁吧！

🚀👍 3

3

#1 opened 4 months ago by

JLouisBiz

New activity in unsloth/Qwen3.6-27B-MTP-GGUF 5 months ago

Noticeable Performance Decrease

👍 3

4

#23 opened 5 months ago by

WebWeaverWraith

New activity in canada-quant/DeepSeek-V4-Flash-W4A16-FP8 5 months ago

Can I run this model on 2x H20 141GB?

3

#1 opened 5 months ago by

CHNtentes

New activity in havenoammo/Qwen3.6-35B-A3B-MTP-GGUF 5 months ago

Is it possible to only download the mtp gguf (<1GB one) to use with existing ggufs?

4

#3 opened 5 months ago by

CHNtentes

New activity in deepseek-ai/DeepSeek-V4-Pro 5 months ago

Will there be small models like 12b?

👍👀 5

15

#164 opened 5 months ago by

Crownelius

New activity in deepseek-ai/DeepSeek-V4-Flash 5 months ago

Too big to run locally.

🤯👍 12

21

#12 opened 5 months ago by

Dampfinchen

New activity in deepseek-ai/DeepSeek-V4-Pro 5 months ago

所以我猜是混合精度加模型太大导致暂时还没有量化的模型出来

6

#96 opened 5 months ago by

lzm1066258

New activity in deepseek-ai/DeepSeek-V4-Flash 5 months ago

May I ask if there is a deployment document?

2

#10 opened 5 months ago by

jerryliujiawei

New activity in Qwen/Qwen3.6-27B-FP8 5 months ago

Parameter model.layers.15.mlp.gate_gate_up_proj.weight_scale_inv not found in params_dict

5

#3 opened 5 months ago by

CHNtentes

New activity in Qwen/Qwen3.6-35B-A3B 5 months ago

太吃显存啦

6

#21 opened 6 months ago by

yukojiangjiang

New activity in cyankiwi/MiniMax-M2.7-AWQ-4bit 6 months ago

These are NOT actual AWQ-quantized models.

4

#2 opened 6 months ago by

cai-cai

New activity in unsloth/MiniMax-M2.7-GGUF 6 months ago

larger file size for same quant

5

#4 opened 6 months ago by

CHNtentes