Rob

rdtand's activity

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

New activity in rdtand/Qwen3.8-27B-PrismaAQUA-5.5bit-vllm 21 days ago

Amazing humaneval/+ result!

1

#5 opened 22 days ago by

swasslikeme

New activity in rdtand/Qwen3.8-27B-PrismaScout-AQUA-Vision-20GB about 1 month ago

RTX Pro 4500 start up variables

1

#1 opened about 1 month ago by

Gabt123

New activity in rdtand/Qwen3.8-27B-PrismaScout-AQUA-20GB about 1 month ago

Thanks for the 3.8 PrismaScount!

❤️ 1

8

#1 opened about 1 month ago by

Gabt123

New activity in rdtand/DeepSeek-V4-Flash-0731-PrismaQuant-AQUA-gridbook-87GB-spark-vllm about 2 months ago

Single DGX Spark (GB10/sm_121) serving OK at 6.54 tok/s — decode profile + tuning questions

8

#2 opened about 2 months ago by

trosfy

New activity in rdtand/Qwen3.8-27B-PrismaAQUA-5.5bit-vllm about 2 months ago

de-restricted model?

➕ 1

2

#4 opened about 2 months ago by

swasslikeme

New activity in rdtand/DeepSeek-V4-Flash-0731-PrismaQuant-AQUA-gridbook-87GB-spark-vllm about 2 months ago

without gridbook

3

#1 opened about 2 months ago by

spectator2026

New activity in rdtand/Qwen3.8-27B-PrismaAQUA-5.5bit-vllm about 2 months ago

PrismaSCOUT-like model in the future?

❤️ 1

6

#3 opened about 2 months ago by

Gabt123

tried it myself...

1

#1 opened about 2 months ago by

cboergermann

New activity in rdtand/Qwen3.5-122B-A10B-PrismaQuant-4.75bit-vllm 3 months ago

Do I need a special version of VLLM to run this?

3

#1 opened 3 months ago by

dicksondickson

New activity in rdtand/Hy3-295B-A21B-PrismaQuant-5.3bit-2xSpark-vllm 3 months ago

Results on two DGX Spark

5

#1 opened 3 months ago by

cboergermann

New activity in rdtand/Ornith-1.0-35B-PrismaAURA-4.75bit-vllm-MTP 3 months ago

MTP?

5

#1 opened 3 months ago by

budman789

New activity in rdtand/Qwen3.6-35B-A3B-PrismaQuant-4.75bit-vllm 3 months ago

Heretic

2

#5 opened 3 months ago by

budman789

New activity in rdtand/Qwen3.6-27B-PrismaSCOUT-Blackwell-NVFP4-BF16-vllm 4 months ago

llama-eval results

❤️ 1

5

#3 opened 5 months ago by

danym

How much ram required with vllm?

1

#6 opened 4 months ago by

Wint3rmu7e

New activity in rdtand/Qwen3.6-27B-PrismaSCOUT-Blackwell-NVFP4-BF16-vllm 5 months ago

Does not work on 3090 GPUs

3

#2 opened 5 months ago by

Neiko2002

New activity in rdtand/Qwen3.6-27B-PrismaQuant-5.5bit-vllm 5 months ago

Issue running on WSL and RTX 6000 Blackwell

3

#3 opened 5 months ago by

markonimakaroni

Unable to find recipe

4

#2 opened 5 months ago by

SantoBSD

New activity in rdtand/Mistral-Medium-3.5-128B-PrismaQuant-4.75-vllm 5 months ago

Fix Transformers config

1

#1 opened 5 months ago by

juliendenize

New activity in rdtand/Qwen3.6-35B-A3B-PrismaQuant-4.75bit-vllm 5 months ago

Got AssertionError: last dim has to be multiple of 16, but got 2152. when running on dual DGX Spark

7

#3 opened 5 months ago by

hfyiu

New activity in rdtand/Qwen3.6-27B-PrismaQuant-5.5bit-vllm 6 months ago

Unable to run on 3090

1

#1 opened 6 months ago by

mancub