Exo Labs @exolabs Twitter profile

Pinned Tweet

Exo Labs

@exolabs

12 days

The future of AI is open source and decentralized

AI at Meta

@AIatMeta

12 days

"Exo's use of Llama 405B and consumer-grade devices to run inference at scale on the edge shows that the future of AI is open source and decentralized." - @mo_baioumy

43

218

1K

6

30

218

Last Seen Profiles

@survivorturkiye

@ArinFishb

@CoachMartin901

@stw_pdg

@gertsofficial

@bokeplokalmalam

@puts_seila

@kemsveto

@bhy_ch48127

@bokeplokalmalam

@apstrout18

@bartenderjack1

@stw_pdg

@SSN_JMU

@kacha7d

@EllaRobinson80

@kobikwelu

@pringlepaam

@stw_pdg

@myDrip_io

@darksx_

@HomiKSC

@GligaZoran

@AmayahMoen49711

@Sukasuk3299711

@milesishere_

@Leah__vtuber

@ahlam794749

@AllynHylin

@AliAli980460

@EthanHershaft

@elite_gz

@Matt_hubbard40

@MaryWatts399526

@JungjohannB

@xp65535

Exo Labs

@exolabs

3 hours

Meta Connect is happening today. exo features in the keynote at 11am PT.

3

2

32

Exo Labs

@exolabs

1 day

AI on any device.

Alex Cheema - e/acc

@ac_crypto

1 day

Home AI cluster on Lenovo Tab M10's using @exolabs

3

4

59

1

4

45

Exo Labs

@exolabs

4 days

Welcome @pmarca

Big Tech Alert

@BigTechAlert

5 days

🆕 @pmarca has started following @exolabs

0

2

12

1

2

73

Exo Labs Retweeted

Mohamed Baioumy

@mo_baioumy

6 days

@exolabs Github passed 6.9k stars

5

28

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

7 days

Running Qwen 2.5 Math 72B distributed across 2 MacBooks. Uses @exolabs_ with the MLX backend.

5

12

102

Exo Labs Retweeted

Beff – e/acc

@BasedBeffJezos

12 days

The future of AI is open source and decentralized. Amen.

22

58

553

Exo Labs Retweeted

kache

@yacineMTB

12 days

yeah, this is genuinely good news for hardware to consumer providers

0

3

21

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

12 days

This figure from the o1 research report makes the case for a setup like this much stronger. We can scale LLM capability with more inference time compute. Search is sync-lite: you don't need ultra low latency / high bandwidth interconnect between your @__tinygrad__ tinyboxes.

4

44

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

12 days

Think twice about trading in that old MacBook

8

7

165

Exo Labs Retweeted

Marvin Gabler

@MarvinGabler

13 days

Idle MacBooks got a job now @exolabs_

4

57

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

16 days

Yes, exo will support Apple TV's.

21

18

368

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

25 days

On the power efficiency of Apple M chips. @exolabs_ with the MLX backend can get ~100% GPU utilisation out of each MacBook Pro M3 Max. The GPU has 128GB of memory (bandwidth ~400GB/s) and consumes ~40 Watts of power with ~30 TFLOPs of fp16 performance. MacBook Pro M3 Max GPU: -

25

114

821

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

25 days

23 iMacs running distributed llama 3 405b with @exolabs_

30

75

878

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

25 days

free on-premise llama 3 405b on 23 iMacs using @exolabs_

3

48

Exo Labs Retweeted

Alex Cheema - e/acc

@ac_crypto

26 days

is there a name for this? when someone is shaking from using your product?

4

1

28

Exo Labs Retweeted

luffy

@0xluffyb

28 days

peak website design @exolabs_ @ac_crypto

1

4

54

Exo Labs Retweeted

vincent

@t0kenl1mit

28 days

More progress in the @exolabs_ pytorch interface - Tested with Qwen2 Instruct, TinyLlama and Llama3 8b and was able to produce tokens - Currently not producing good output and gibberish. Looking at my top_p, top_k and temp along with my logit_sample method - Testing on