Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

DeepSeek V4.1 Flash 1017 on TensorFold ⚡️ - 16 concurrent session - Each with full native 1M context - Non-greedy, thinking on. OFFICIAL WEIGHTS. Recipe dropping very soon!

MiaAI_lab's profile picture

Mia

43,880 subscribers

21,711 Aufrufe • vor 4 Tagen •via X (Twitter)

Wissenschaft & Technologie
LIVE

Anya Rossi• Live Now

Private livecam show

43 Kommentare

Profilbild von Mia
Miavor 4 Tagen

@volatilemarkts 4x DGX Sparks SUPPORTS SWITCHLESS

Profilbild von Mia
Miavor 4 Tagen

@volatilemarkts Out now

Profilbild von Stary Libra
Stary Libravor 4 Tagen

@ashxhart @volatilemarkts Can I change the setting to TP3 and use it on 3 Spark?

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts Initial release supports only 4

Profilbild von Stary Libra
Stary Libravor 4 Tagen

@ashxhart @volatilemarkts Okay. I'll wait. Thanks as always for sharing your awesome work!

Profilbild von Kevin Theo
Kevin Theovor 4 Tagen

@ashxhart @volatilemarkts running your SGLang recipe on 4 Sparks in a switchless ring. stock DeepSeek V4.1 Flash here: 70.0 tok/s on code single stream, 72.9 total at 4 streams (26.2 each). 16 sessions at full 1M is a different league. will rerun my prompts on TensorFold when the recipe drops

Profilbild von HealthRanger
HealthRangervor 4 Tagen

@ashxhart @volatilemarkts Oh man, this is most definitely worth trying out. Can't wait for the recipe!

Profilbild von vr8vr8
vr8vr8vor 4 Tagen

@ashxhart @volatilemarkts Nice achievement 👍

Profilbild von TechMD
TechMDvor 4 Tagen

@ashxhart @volatilemarkts Did you get the 4 spark @MiaAI_lab

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts Yes a while ago :)

Profilbild von TechMD
TechMDvor 4 Tagen

@ashxhart @volatilemarkts nice Im happy, Hope it made a difference to your workflow

Profilbild von GP3611 Phoenix
GP3611 Phoenixvor 4 Tagen

@ashxhart @volatilemarkts how about single stream peak?

Profilbild von Simo
Simovor 4 Tagen

@ashxhart @volatilemarkts Publish sustained throughput latency and memory under full load

Profilbild von Twon | 00.sol | 1TB VRAM
Twon | 00.sol | 1TB VRAMvor 4 Tagen

@ashxhart @volatilemarkts Ablit recipe too please 🙏

Profilbild von Mr. Galt
Mr. Galtvor 4 Tagen

@ashxhart @volatilemarkts Nice, now I want two extra Sparks. :(

Profilbild von BullBear.News
BullBear.Newsvor 4 Tagen

@ashxhart @volatilemarkts running 1M context natively on a 16gb mac mini via ollama sounds wild, how does tensorfold handle KV cache paging for those concurrent sessions?

Profilbild von lostinazeroy
lostinazeroyvor 4 Tagen

@ashxhart @volatilemarkts 16 sessions each at full 1m context on one box is wild

Profilbild von Mike Hutu
Mike Hutuvor 4 Tagen

@ashxhart @volatilemarkts Sayy whaa 2x DGX?

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts Nah... this is 4

Profilbild von Chris Frisby
Chris Frisbyvor 4 Tagen

@ashxhart @volatilemarkts Dang, what's the best quant on TP=3? 4bit?

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts OFFICIAL WEIGHTS !!!!

Profilbild von Chris Frisby
Chris Frisbyvor 4 Tagen

@ashxhart @volatilemarkts Dang, looking forward to it! 🤞🏼🙌🏼

Profilbild von Surrender ⚡️
Surrender ⚡️vor 4 Tagen

@ashxhart @volatilemarkts have you tried minimax yet?

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts Yes. I skipped it.

Profilbild von Ponzi
Ponzivor 4 Tagen

@ashxhart @volatilemarkts

Profilbild von Pragnyan Ramtha
Pragnyan Ramthavor 4 Tagen

@ashxhart @volatilemarkts ok where is the kv cache even living

Profilbild von Sage Byte
Sage Bytevor 4 Tagen

@ashxhart @volatilemarkts Offical weights and 16 concurrent sessions is crazy 🔥

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts The only way to do it is crazy.

Profilbild von Whispers of Wisdom
Whispers of Wisdomvor 4 Tagen

@ashxhart @volatilemarkts If you have time, can you please update the 2x DGX Spark version too please ♥️

Profilbild von FrontierMind
FrontierMindvor 4 Tagen

@ashxhart @volatilemarkts running 16 concurrent sessions at 1m context on flash weights sounds wild for local infra

Profilbild von steddy
steddyvor 4 Tagen

@ashxhart @volatilemarkts I don’t see a graphical front end for TensorFold on Spark. Only Mac

Profilbild von Peter Morris
Peter Morrisvor 4 Tagen

@ashxhart @volatilemarkts I assume you need 4×GB10 for 16 concurrent?

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts With official weights, yes, but it will support 3 sparks later on too

Profilbild von Steven Cheng
Steven Chengvor 4 Tagen

@ashxhart @volatilemarkts 16 sessions with full 1M context running simultaneously? That memory bandwidth must be absolutely insane.

Profilbild von Novaris
Novarisvor 4 Tagen

@ashxhart @volatilemarkts 16 concurrent sessions, each with full 1M context — that's the kind of spec that makes 'just run it locally' sound almost reasonable.

Profilbild von The SI Maximalist
The SI Maximalistvor 4 Tagen

@ashxhart @volatilemarkts the number I’m watching for in the recipe: tokens/sec per session with all 16 contexts filled. A 1M window tells us capacity; sustained decoding tells us how responsive those sessions are

Profilbild von Random Libertarian Tech Lead
Random Libertarian Tech Leadvor 4 Tagen

@ashxhart @volatilemarkts Should be interesting to see how this compares single-stream compared to my ground-up Rust stack.

Profilbild von 汉卿 | hanqing
汉卿 | hanqingvor 4 Tagen

@ashxhart @volatilemarkts 1M上下文+16并发+非贪心，这配方一出直接封神。

Profilbild von Ryan Feng
Ryan Fengvor 4 Tagen

@ashxhart @volatilemarkts prefil? sglang can do 5k+/s on TP4

Profilbild von Mia
Miavor 4 Tagen

@ashxhart @volatilemarkts I can only test switchless for now, don't have a switch. Switchless gets 4k prefill on 256k

Profilbild von Dave
Davevor 4 Tagen

@ashxhart @volatilemarkts I'm on TF 1.0.12 with DS 4 spark and I have images. But sad face TF does not cache images so reads are expeeeeensive.

Profilbild von AI Doc
AI Docvor 4 Tagen

@ashxhart @volatilemarkts But is it better than glm 5.3 flash? @MiaAI_lab

Profilbild von Dr.-Ing. Manuel Ramsaier
Dr.-Ing. Manuel Ramsaiervor 4 Tagen

@ashxhart @volatilemarkts ❤️❤️❤️ will this be significantly better than the exl3 on 2 sparks ?!

Ähnliche Videos

DeepSeek V4.1 Flash 1017 on 4x DGX Sparks ⚡️ - Official weights on TensorFold - 8 concurrent streams with thinking on
0:35

DeepSeek V4.1 Flash 1017 on 4x DGX Sparks ⚡️ - Official weights on TensorFold - 8 concurrent streams with thinking on

Mia

22,649 Aufrufe • vor 6 Tagen

How DeepSeek v4.1 Flash looks on 3x DGX Sparks with 4 concurrent streams
0:50

How DeepSeek v4.1 Flash looks on 3x DGX Sparks with 4 concurrent streams

Mia

19,730 Aufrufe • vor 1 Monat

GLM 5.3 Flash vs DeepSeek V4.1 Flash 1017 Can't believe we're abusing gummy bears to test models, but here we are! Both models were run locally. - GLM 5.3 Flash: my own EXL3 quant (link below) - DeepSeek V4.1 Flash 1017: official quant - Same harness.Both on Max effort. Who won?
0:33

GLM 5.3 Flash vs DeepSeek V4.1 Flash 1017 Can't believe we're abusing gummy bears to test models, but here we are! Both models were run locally. - GLM 5.3 Flash: my own EXL3 quant (link below) - DeepSeek V4.1 Flash 1017: official quant - Same harness.Both on Max effort. Who won?

Mia

32,512 Aufrufe • vor 11 Tagen

DeepSeek-V4.1-Flash-1017 is great at tutorial-style motion videos. We made this with the model using OpenDesign Motion: a quick guide to using the OpenDesign API in other coding agents to access DeepSeek V4.1 Flash 1017 and more models.
1:07

DeepSeek-V4.1-Flash-1017 is great at tutorial-style motion videos. We made this with the model using OpenDesign Motion: a quick guide to using the OpenDesign API in other coding agents to access DeepSeek V4.1 Flash 1017 and more models.

OpenDesign

41,432 Aufrufe • vor 12 Tagen

DeepSeek V4.1 Flash 1017 on 4x DGX Sparks runs so well, feels too good to be true. SOTA AI at home.
0:28

DeepSeek V4.1 Flash 1017 on 4x DGX Sparks runs so well, feels too good to be true. SOTA AI at home.

Mia

52,979 Aufrufe • vor 2 Tagen

Introducing Cline Desktop - a native interface for working with open weights models. Use with ClinePass and all our free models like DeepSeek-V4.1-Flash, Musespark-1.3, or BYOK with any provider!
0:32

Introducing Cline Desktop - a native interface for working with open weights models. Use with ClinePass and all our free models like DeepSeek-V4.1-Flash, Musespark-1.3, or BYOK with any provider!

Cline

1,023,648 Aufrufe • vor 1 Monat

DeepSeek V4.1 Flash 1017 everyone's talking about it for a reason. It does great with 3d built this honeybee on a sunflower like it was nothing. Used Deepseek Harness, 1 turn, looks so good!
1:30

DeepSeek V4.1 Flash 1017 everyone's talking about it for a reason. It does great with 3d built this honeybee on a sunflower like it was nothing. Used Deepseek Harness, 1 turn, looks so good!

Loktar 🇺🇸

43,671 Aufrufe • vor 12 Tagen

Minimax M3 is excellent at SVG generation, reaching close to Gemini 3.5 Flash levels and beating Opus 4.7 on SVG-Bench. With 1M context, native multimodality, strong agentic/coding ability and open weights coming soon, the closed-source moat is thinning fast. Full Video:
0:22

Minimax M3 is excellent at SVG generation, reaching close to Gemini 3.5 Flash levels and beating Opus 4.7 on SVG-Bench. With 1M context, native multimodality, strong agentic/coding ability and open weights coming soon, the closed-source moat is thinning fast. Full Video:

WorldofAI

16,499 Aufrufe • vor 5 Monaten

450 tok/sec ⚡ Abliterated DeepSeek V4.1 Flash 1017 $0.22 / $0.75 / $0.045 cache available on Refuseless API Watch it do cyber work at light speed
0:44

450 tok/sec ⚡ Abliterated DeepSeek V4.1 Flash 1017 $0.22 / $0.75 / $0.045 cache available on Refuseless API Watch it do cyber work at light speed

Refuseless

32,428 Aufrufe • vor 3 Tagen

DeepSeek v4.1 Flash running on 2x DGX Sparks
0:35

DeepSeek v4.1 Flash running on 2x DGX Sparks

Mia

59,572 Aufrufe • vor 1 Monat

DeepSeek-V4.1-Flash-1017 decodes code at 80.5 tok/s on one Mac Studio M5 Ultra 256 GB with BobClawblaw's recipe on TensorFold. Setup: – Recipe and TensorFold branch deepseek-v41-main (72a552ed), his own tool and settings – Context 524,288 tokens, 4 streams, DSpark drafts on, temperature 0 – Speed and quality with thinking off, except GPQA with thinking Speed: – Decode, code: 80.5 tok/s – Decode, prose: 70.5 tok/s – 500 words: about 9.3 s – First token after a 32K prompt: 18.6 s – 8 requests at once: 88.3 tok/s in total Quality: – MMLU-Pro: 67.5% ± 4.6 points – GPQA Diamond: 45.0% ± 6.9 points – IFEval: 90.5% ± 4.1 points – LiveCodeBench: 60.0% ± 14.5 points – GPQA with thinking: 76.7% ± 14.6 points Voxel world: – Built as an agent in 45.7 min, DeepSeek Harness 0.2.1-rc.2, one prompt, one try, no help, thinking on – Tokens written: 112,065 – Browser check: 8.5/10 Good to know: – The author runs temperature 0, the DeepSeek card suggests 1.0. I kept the author's setting. – The browser check ran on a Linux box from the same files. – Answers cut at the cap count as wrong: LiveCodeBench 14 of 40, GPQA Diamond 9 of 198, MMLU-Pro 1 of 400. Play the world in your browser. All numbers, the exact command and the raw output: Recipe by BobClawblaw on Ash Hart's TensorFold, model by DeepSeek.
0:57

DeepSeek-V4.1-Flash-1017 decodes code at 80.5 tok/s on one Mac Studio M5 Ultra 256 GB with BobClawblaw's recipe on TensorFold. Setup: – Recipe and TensorFold branch deepseek-v41-main (72a552ed), his own tool and settings – Context 524,288 tokens, 4 streams, DSpark drafts on, temperature 0 – Speed and quality with thinking off, except GPQA with thinking Speed: – Decode, code: 80.5 tok/s – Decode, prose: 70.5 tok/s – 500 words: about 9.3 s – First token after a 32K prompt: 18.6 s – 8 requests at once: 88.3 tok/s in total Quality: – MMLU-Pro: 67.5% ± 4.6 points – GPQA Diamond: 45.0% ± 6.9 points – IFEval: 90.5% ± 4.1 points – LiveCodeBench: 60.0% ± 14.5 points – GPQA with thinking: 76.7% ± 14.6 points Voxel world: – Built as an agent in 45.7 min, DeepSeek Harness 0.2.1-rc.2, one prompt, one try, no help, thinking on – Tokens written: 112,065 – Browser check: 8.5/10 Good to know: – The author runs temperature 0, the DeepSeek card suggests 1.0. I kept the author's setting. – The browser check ran on a Linux box from the same files. – Answers cut at the cap count as wrong: LiveCodeBench 14 of 40, GPQA Diamond 9 of 198, MMLU-Pro 1 of 400. Play the world in your browser. All numbers, the exact command and the raw output: Recipe by BobClawblaw on Ash Hart's TensorFold, model by DeepSeek.

Niklas

11,028 Aufrufe • vor 4 Tagen

You have Deepseek-v4.1-flash-1017 for free at amazing speeds and you're dooming? Opencode is going god's work.
1:58

You have Deepseek-v4.1-flash-1017 for free at amazing speeds and you're dooming? Opencode is going god's work.

0xSero

49,642 Aufrufe • vor 8 Tagen

- GLM 5.3 Flash EXL3 - 2x DGX Sparks - TensorFold - 4 concurrent prose streams Pay attention to tok/s and TTFT on each stream. Buttery smooth experience!
0:37

- GLM 5.3 Flash EXL3 - 2x DGX Sparks - TensorFold - 4 concurrent prose streams Pay attention to tok/s and TTFT on each stream. Buttery smooth experience!

Mia

32,203 Aufrufe • vor 28 Tagen

FREE UNCENSORED DEEPSEEK V4.1 FLASH 💀 check here: • 552B parameters • 8B/16B active per token • 1M context • vision support • MIT license check details before installing
0:25

FREE UNCENSORED DEEPSEEK V4.1 FLASH 💀 check here: • 552B parameters • 8B/16B active per token • 1M context • vision support • MIT license check details before installing

Peng

118,451 Aufrufe • vor 20 Tagen

Jumpgate - hyperspace jump DS4.1F-1017 vs GLM 5.3 Flash vs Fable 5.1 Who won imo: 🥇 Fable 5.1 🥈 DeepSeek V4.1 Flash 1017 🥉 GLM 5.3 Flash You can run these by yourself here: Prompt below.
0:24

Jumpgate - hyperspace jump DS4.1F-1017 vs GLM 5.3 Flash vs Fable 5.1 Who won imo: 🥇 Fable 5.1 🥈 DeepSeek V4.1 Flash 1017 🥉 GLM 5.3 Flash You can run these by yourself here: Prompt below.

Mia

12,939 Aufrufe • vor 9 Tagen

GLM-5.3-Flash streaming at 32 tok/s on a single Spark via Tensorfold ⚡️ i will share the recipe later today
0:30

GLM-5.3-Flash streaming at 32 tok/s on a single Spark via Tensorfold ⚡️ i will share the recipe later today

filipe

15,230 Aufrufe • vor 27 Tagen

DeepSeek V4.1 Flash 1017 Abliterated just got a major update. • Served with FP8 quant • 70-120 tok/sec • Nearly 0% refusals We're proud to have now the only DeepSeek Uncensored model that ACTUALLY doesn't refuse! Come try it on Refuseless.
0:33

DeepSeek V4.1 Flash 1017 Abliterated just got a major update. • Served with FP8 quant • 70-120 tok/sec • Nearly 0% refusals We're proud to have now the only DeepSeek Uncensored model that ACTUALLY doesn't refuse! Come try it on Refuseless.

Refuseless

31,269 Aufrufe • vor 6 Tagen

DeepSeek V4.1 Flash 1017 vs GPT-6 Astra The Space Train 🚂🪐 Am I crazy or did DeepSeek just beat Astra on this Space Train benchmark??? 😱 Astra took 40m 18s and cost $14.37. DeepSeek took 2h 07m and cost $0.40. That's 36x cheaper. Forty cents!! 🤯 DeepSeek V4.1 Flash 1017. Remember that name. Watch them side by side and tell me: is Astra worth 36x the price here? 👀 Google Gemini Google OpenAI ChatGPT
0:25

DeepSeek V4.1 Flash 1017 vs GPT-6 Astra The Space Train 🚂🪐 Am I crazy or did DeepSeek just beat Astra on this Space Train benchmark??? 😱 Astra took 40m 18s and cost $14.37. DeepSeek took 2h 07m and cost $0.40. That's 36x cheaper. Forty cents!! 🤯 DeepSeek V4.1 Flash 1017. Remember that name. Watch them side by side and tell me: is Astra worth 36x the price here? 👀 Google Gemini Google OpenAI ChatGPT

Tony

49,306 Aufrufe • vor 11 Tagen

deepseek v4.1 flash 1017 just mogged opus 5.5 and it was 32x cheaper: deepseek: $0.31 opus 5.5: ~$10 with the right harness and a few iterations, ds4.1 flash is a model you can actually use every day
0:40

deepseek v4.1 flash 1017 just mogged opus 5.5 and it was 32x cheaper: deepseek: $0.31 opus 5.5: ~$10 with the right harness and a few iterations, ds4.1 flash is a model you can actually use every day

filipe

33,003 Aufrufe • vor 11 Tagen

$25 free credits on River AI right now 😳 DeepSeek V4.1 Flash + GLM-5.3 Flash at 20% cheaper than official rates. → Go grab it.
0:27

$25 free credits on River AI right now 😳 DeepSeek V4.1 Flash + GLM-5.3 Flash at 20% cheaper than official rates. → Go grab it.

Punch AI

13,434 Aufrufe • vor 18 Tagen