@wapiflapi

python expert, wannabe hacker, dungeon master, he/him

 France & Europe
wapiflapi.com/
Joined August 2010
  • Tweets1.8K
  • Following483
  • Followers790
118Media
I made an easy decision tree for anyone considering using git submodules or a monorepo.
137
398
59
4,260
554,653
We solved *Semantic Search-as-you-Type* ⌨️ It is faster than embedding inference ⚡ It is Open-Source 👐 It is in Rust 🦀 We use it for our docs 📜 Read more how 👇
7
52
9
426
85,392
wapiflapi retweeted
Paul Graham
@paulg6 Aug 2023
An unexpected result: Cutting emissions by ships has increased global warming. science.org/content/article/…

‘We’re changing the clouds.’ An unintended test of geoengineering is fueling record ocean warmth

Pollution cuts have diminished “ship track” clouds, adding to global warming

science.org
116
273
71
1,927
791,701
Dites-moi si vous les avez déjà lus 😎 Idem si vous en recommandez d’autres ou si vous en voulez plus !
1
138
New settings on this platform, by default it switched to only allow messages from verified users. This sucks if you want people to reach out to you. Also sucks if you want to reach out to people who haven't seen this change.
2
1
5
3,966
Ah yes, the well known "except you, FAANG" clause that's so common in *open source* licenses like GPL, MIT, BSD, Apache2, ... Here I go again, this can't be for real lol
Yann LeCun
@ylecun18 Jul 2023
This is huge: Llama-v2 is open source, with a license that authorizes commercial use! This is going to change the landscape of the LLM market. Llama-v2 is available on Microsoft Azure and will be available on AWS, Hugging Face and other providers Pretrained and fine-tuned models are available with 7B, 13B and 70B parameters. Llama-2 website: ai.meta.com/llama/ Llama-2 paper: ai.meta.com/research/publica… A number of personalities from industry and academia have endorsed our open source approach: about.fb.com/news/2023/07/ll…
57
138
50
1,510
733,783
wapiflapi retweeted
Paul Graham
@paulg12 Jul 2023
It's so stupid to try to teach teenagers "entrepreneurship" by having them pitch made-up startup ideas. You make them focus on the one thing they shouldn't be focusing on, and then you rate them by their appeal to investors, instead of users.
20
100
12
1,294
122,830
wapiflapi retweeted
Paul Graham
@paulg12 Jul 2023
Replying to @callmeed
It is if you want to start a startup. It might be good if you wanted to start a small business.
5
4
1
164
22,680
this is wild — kNN using a gzip-based distance metric outperforms BERT and other neural methods for OOD sentence classification intuition: 2 texts similar if cat-ing one to the other barely increases gzip size no training, no tuning, no params — this is the entire algorithm:
for (x1, _) in test_set: Cx1 = len(gzip.compress(x1.encode())) distance_from_x1 = [] for (x2, _) in training_set: Cx2 = len(gzip.compress(x2. encode()) x1x2 = " ".join([x1, x2]) Cx1x2 = len(gzip.compress(x1x2. encode()) ncd = (Cx1x2 - min(Cx1,Cx2)) / max(Cx1, Cx2) distance_from_x1.append(ncd) sorted_idx = np.argsort(np.array(distance_from_x1)) top_k_class = training_set[sorted_idx[:k], 1] predict_class = max(set(top_k_class), key=top_k_class.count)

ALT for (x1, _) in test_set: Cx1 = len(gzip.compress(x1.encode())) distance_from_x1 = [] for (x2, _) in training_set: Cx2 = len(gzip.compress(x2. encode()) x1x2 = " ".join([x1, x2]) Cx1x2 = len(gzip.compress(x1x2. encode()) ncd = (Cx1x2 - min(Cx1,Cx2)) / max(Cx1, Cx2) distance_from_x1.append(ncd) sorted_idx = np.argsort(np.array(distance_from_x1)) top_k_class = training_set[sorted_idx[:k], 1] predict_class = max(set(top_k_class), key=top_k_class.count)

Luke Gessler@LukeGessler12 Jul 2023
this paper's nuts. for sentence classification on out-of-domain datasets, all neural (Transformer or not) approaches lose to good old kNN on representations generated by.... gzip aclanthology.org/2023.findin…
134
1,091
334
6,866
2,062,245
A 14-line Python script using gzip outperforming a 345m parameter transformer model is probably the most hilarious result I've seen all year. aclanthology.org/2023.findin…
65
937
148
6,523
1,398,601
RIP create-react-app. The current meta in web apps changes so quickly. When I need to throw together a simple app I always need to figure it out again. A good combo for me right now seems to be vitejs.dev/ with daisyui.com/.
151
Relancer les meet-ups #python a #Toulouse ca vous tente ? Laissez moi un message si ca vous intéresse. On prévoit des choses super cool !
2
1
406
"Hey {name}".format(name="twitter") Templates written by users are not safe when using something like jinja or the standard `format`. If you know of a #python library that can help with this, please share ! With AI prompts everywhere, I'm sure this is an issue.
1
265
Turns out `string.Template` is pretty much what I was looking for :-)
125
don’t be misled by low-order statistics
11
322
27
1,642
156,054
I had some problems deploying monorepos to something that "just runs it". Turns out railway.app/ has been really great at doing exactly that. You don't even need to create an account to try it out. Just spin up a database and let it run some code it gets from @github
165
Python has a different meaning for "precision" than me. A more accurate one actually :-) It's interesting how I completely missunderstand documentation sometimes...
Screenshot of a python interpreter showing: >>> import decimal >>> context = decimal.Context(prec=2) >>> context.create_decimal_from_float(123.456) Decimal('1.2E+2') >>> # I expected 123.45, what about you?

ALT Screenshot of a python interpreter showing: >>> import decimal >>> context = decimal.Context(prec=2) >>> context.create_decimal_from_float(123.456) Decimal('1.2E+2') >>> # I expected 123.45, what about you?

1
1
1
220
As usual, it's in the FAQ. docs.python.org/3/library/de… I was actually looking for docs.python.org/3/library/de… , it doesn't look like there's a way to set that on the context,

decimal — Decimal fixed-point and floating-point arithmetic

Source code: Lib/decimal.py The decimal module provides support for fast correctly rounded decimal floating-point arithmetic. It offers several advantages over the float datatype: Decimal “is based...

docs.python.org
1
122
Here we go.
Screenshot of a python interpreter showing: >>> context.create_decimal_from_float( ... 123.456 ... ).quantize(decimal.Decimal(10) ** -2) Decimal('123.46') >>> # It's not 123.45 but I'll take it.

ALT Screenshot of a python interpreter showing: >>> context.create_decimal_from_float( ... 123.456 ... ).quantize(decimal.Decimal(10) ** -2) Decimal('123.46') >>> # It's not 123.45 but I'll take it.

92
My favorite HTTP client for python is github.com/encode/httpx A trick that did wonders for me is combining it with pydantic to load HTTPX PROXY settings from the environment through JSON. It's extremely flexible.
A python dictionary containing configuration for complex routing to different HTTP proxies. There are examples with different protocols, domains, ports and combination of those.

ALT A python dictionary containing configuration for complex routing to different HTTP proxies. There are examples with different protocols, domains, ports and combination of those.

186
Load more