guglielmo

Who am I? Who am I? -per ora un giovane fisico/mountain addicted/fotografo a tempo perso/giocoliere di tradizione/cuoco nel cuore

[guglielmo @diabbolo]

la fibra non mi funziona. riavviato il modem piu' volte e ancora lampeggia rosso. avete notizie di disservizi su milano? @FASTWEBHelp

[guglielmo @diabbolo]

@LambdaAPI I cannot login, I insert my password and username but the page does not update (does not says either my password is wrong) I wrote to tech support but still no answers

[guglielmo @diabbolo]

@LambdaAPI appreciated! thanks so much

[guglielmo @diabbolo]

@LambdaAPI I would like to use your inference API service https://t.co/R3L1sWWoAX but I cannot find any detail about pricing, could you please point me to where I can find that information?

[guglielmo @diabbolo]

@ollama @huggingface is it already in in v 3.13 ?

[guglielmo @diabbolo]

magari un correttore di bozze? @Corriere https://t.co/t1VfVJlnvg "le aprole che" invece di "le parole che" pure word avrebbe autocorretto giusto ... :/

[guglielmo @diabbolo]

@ollama amazing! any ETA on the support for the vision 11B and 90B ? looking forward to that

[guglielmo @diabbolo]

e' una settimana che provo a contattare @PosteItaliane teelfonicamente e' impossibile, via chat anche e provando a scrivergli trovo il seguente messaggio... c'e' altro che posso fare?

[Lorenzo Fares @lorenzofares]

Jannik #Sinner 🇮🇹 ha vinto 8 partite di fila da quando è salito al n°1

Ha raggiunto due di passaggio, Pete Sampras 🇺🇸 e Jimmy Connors, 🇺🇸, in questa statistica

Meglio di lui solo 3:
- Bjorn Borg 🇸🇪 (12 vittorie)
- Novak Djokovic 🇷🇸 (9)
- Andy Murray 🇬🇧 (9)
#tennis #Wimbledon

[guglielmo @diabbolo]

@osanseviero is there a particular reason why the codellama models are from codellama and not from meta-llama organization? in the description https://t.co/OJCYtSNoel says
"This is an unofficial organization for the Code Llama models in the Hugging Face Transformers format. "

[guglielmo @diabbolo]

@_philschmid @huggingface @osanseviero Any additional info on this???

[guglielmo @diabbolo]

HF folks, is there an ETA on the availability of llama3 on inference for PRO ? @huggingface @_philschmid @osanseviero

[guglielmo @diabbolo]

@_philschmid @huggingface @osanseviero However I cannot use "codellama/CodeLlama-70b-hf" which is advertised in the inference pro webpage from hf (and instruct neither) I get "The model ... is too large to be loaded automatically (137GB > 10GB)"

[guglielmo @diabbolo]

@_philschmid @huggingface @osanseviero I just now tried the llama3, have not done it earlier cause they were not mentioned in the inference PRO webpage, and it worked for the llama3-8B-instruct and the 70B version as well. thanks

[guglielmo @diabbolo]

Dear HF! I am using inference pro on HF but I cannot run the codellama70b even though it is listed in the supported model in inference for pro.When trying to use it I get model too big... I use the model `codellama/CodeLlama-70b-Instruct-hf` @huggingface @_philschmid @osanseviero

[guglielmo @diabbolo]

@osanseviero even though it is listed in the supported model in inference for pro the codellama70b is not working for me... I get model too big ... using ```codellama/CodeLlama-70b-Instruct-hf```

[guglielmo @diabbolo]

I have a mac updated to the latest MacOS 14.3.1 sonoma @citrix is there a way to have the app working with my OS? or an ETA on when you will release a version compatible? #sonoma #citrix

[guglielmo @diabbolo]

@ezeferrero well it is not unknown anymore! :)

[guglielmo @diabbolo]

@_philschmid @AIatMeta @huggingface is it available also in inference PRO ? https://t.co/teHNLoxSS4

[Philipp Schmid @_philschmid]

Have you heard of Mixture of Experts (MoE) models? 🤔 With the release of @MistralAI Mixtral 8x7B, MoEs are gaining attention, it is also rumored @OpenAI GPT-4 is an MoE👀 

But what exactly are MoEs, and how do they work?

We created an in-depth blog.

https://t.co/a4DL1A8zCe