deepseek-ai/DeepSeek-R1

90+ tokens per second for MI300x8 using batch_size = 1

#166 opened about 8 hours ago by

ghostplant

RytryR1

#165 opened about 17 hours ago by

Rocka01

"aha moment" comment deleted by Perplexity (recovered)

3

#164 opened about 18 hours ago by

FalconNet

输出乱码

1

#163 opened about 22 hours ago by

cell22

'num_hidden_layers': 61, but layer 62 has weights.

#162 opened 1 day ago by

xinhe

Upload GTG Breaking every Limit

#161 opened 1 day ago by

GTGenesis

support prefix complete

#158 opened 3 days ago by

HuggineAllen

Create app.py

#157 opened 5 days ago by

SpaceAgeRobotics

Create 1

#156 opened 5 days ago by

madevii

Brokersponsor

#155 opened 6 days ago by

Brokersponsor

Update README.md

#154 opened 6 days ago by

egegvner

Upload IMG_4530.png

#152 opened 6 days ago by

Noemie202586

Upload IMG_1745.JPG

#151 opened 6 days ago by

Ladib

Create Clara

1

#150 opened 7 days ago by

Clblinks

If I understand correctly, evaluating MATH-500 requires 64*500 model calls?

1

#149 opened 8 days ago by

Rorschaaaach

Request: DOI

#148 opened 8 days ago by

Tarush-Appreciate

Update README.md

#147 opened 9 days ago by

tekno-power

Update README.md

#146 opened 9 days ago by

Ekimnedops6969

Adding <think>\n after chat template will cause vllm to not return reasoning_content (null) when reasoning

6

#144 opened 9 days ago by

kebeliu

Update README.md

1

#143 opened 9 days ago by

MuhammadEhsan

Request for Information on Purchasing Reasoning API Key

1

#142 opened 10 days ago by

brahamaandai

ssss

1

#140 opened 11 days ago by

DZGT

Update model_max_length in tokenizer_config.json

#139 opened 12 days ago by

kkokkie2360

Host of the model

3

#138 opened 12 days ago by

henrycwf

Lite version for DeepSeek-R1?

1

#137 opened 12 days ago by

haili-tian

[Bug] assert not self.training

1

#136 opened 12 days ago by

Gaie

Upload IMG_0253.HEIC

#134 opened 12 days ago by

rynty

Upload comment-sample.xlsx

#133 opened 13 days ago by

faham123

non-reasoning data

#132 opened 14 days ago by

cmgzy

能不能放一些 4bit的权重，现在手里面的卡都不支持FP8

1

#131 opened 14 days ago by

zhnagchenchne

For the universe! DeepPhaser.py DeepCoralX.py and DeepSynapse.py

3

#129 opened 14 days ago by

karmikovic

Request: Create distill of Mistral Small 24B

3

#128 opened 14 days ago by

Kenshiro-28

which vision model is R1 using for text extraction from image or pdfs.

2

#127 opened 15 days ago by

ashutoshroy02

PRII

#126 opened 15 days ago by

bajramani

Request: DOI

#125 opened 15 days ago by

Yungchizzy

Little brother(s) of big DeepSeek-R1 ?

2

#124 opened 16 days ago by

MrDevolver

Upload gugagagaggagagagga.pdf

1

#123 opened 16 days ago by

HahahhahH

Tool / Function Calling

3

#122 opened 16 days ago by

smcleod

Change quant_method to bitsandbytes_4bit

#121 opened 16 days ago by

ngoc24794

Unknown quantization type

5

#120 opened 17 days ago by

Reewaz321

UPdate config.json

#119 opened 17 days ago by

keerthanaOfficial2001

所以部署一个671B的模型显存需要多少有什么基准的硬件配置？

24

#118 opened 17 days ago by

cena163

nodejs

#117 opened 17 days ago by

k1de

Distill Compatibility for PC w/ Ryzen 7 Pro 8840HS w/ 780M Graphics 2x32GB RAM 1TB DDR5 SSD

1

#115 opened 18 days ago by

arzx

Upload gitattributes.txt

#114 opened 18 days ago by

SafeerChalil

Introducing Deepseek's TinyZero

1

#113 opened 18 days ago by

DeepSeekModerator

Create Kuch v

1

#112 opened 19 days ago by

gamerdowntown

Request: DOI

#111 opened 19 days ago by

Hassanabbas2975

quantization fp8 error occuring while using pipeline approach or transformer based approach

1

#110 opened 19 days ago by

neethuvm

Deepseek-R1

#109 opened 19 days ago by

KudanTao

90+ tokens per second for MI300x8 using batch_size = 1

RytryR1

"aha moment" comment deleted by Perplexity (recovered)

输出乱码

'num_hidden_​​layers': 61, but layer 62 has weights.

Upload GTG Breaking every Limit

support prefix complete

Create app.py

Create 1

Brokersponsor

Update README.md

Upload IMG_4530.png

Upload IMG_1745.JPG

Create Clara

If I understand correctly, evaluating MATH-500 requires 64*500 model calls?

Request: DOI

Update README.md

Update README.md

Adding <think>\n after chat template will cause vllm to not return reasoning_content (null) when reasoning

Update README.md

Request for Information on Purchasing Reasoning API Key

ssss

Update model_max_length in tokenizer_config.json

Host of the model

Lite version for DeepSeek-R1?

[Bug] assert not self.training

Upload IMG_0253.HEIC

Upload comment-sample.xlsx

non-reasoning data

能不能放一些 4bit的权重，现在手里面的卡都不支持FP8

For the universe! DeepPhaser.py DeepCoralX.py and DeepSynapse.py

Request: Create distill of Mistral Small 24B

which vision model is R1 using for text extraction from image or pdfs.

PRII

Request: DOI

Little brother(s) of big DeepSeek-R1 ?

Upload gugagagaggagagagga.pdf

Tool / Function Calling

Change quant_method to bitsandbytes_4bit

Unknown quantization type

UPdate config.json

所以部署一个671B的模型 显存需要多少 有什么基准的硬件配置？

nodejs

Distill Compatibility for PC w/ Ryzen 7 Pro 8840HS w/ 780M Graphics 2x32GB RAM 1TB DDR5 SSD

Upload gitattributes.txt

Introducing Deepseek's TinyZero

Create Kuch v

Request: DOI

quantization fp8 error occuring while using pipeline approach or transformer based approach

Deepseek-R1

'num_hidden_layers': 61, but layer 62 has weights.

所以部署一个671B的模型显存需要多少有什么基准的硬件配置？