tweetindex
DE

Leonie ✓

@helloiamleonie · Latent space · joined 16 Aug 2022

Post-training @liquidai

20 901Followers
1 297Following
4 324Posts total
185.9KViews on collected posts

Neueste Beiträge

100% local audio-to-text transcription CLI we just updated the demo to use the LFM2.5 family: serve LFM2.5-Audio-1.5B with llama.cpp on your laptop and let it transcribe any audio input code: https://t.co/tKmLsn9O7i https://t.co/39Tg9VZCWY
0:16
17.6K views · 324 likes · 30 reposts · 10 replies Open on X →
great overview of some vision evals 👀 2.9K views · 20 likes · 2 reposts · 2 replies Open on X →
"look at the data" has to be the highest alpha advice for working on LLMs; I'm currently working on optimizing LFM VL models and I realized I could improve my intuitions on vision datasets; a thread where I go through the popular vision benchmarks, 1/x https://t.co/CHOKR7KEW5
9.4K views · 44 likes · 8 reposts · 4 replies Open on X →
we published a blog on hugging face at possibly the worst time yesterday lol congrats to the HF team on the big news! 💚 here's a fine-tuning tutorial showing how to • fine-tune a tiny LFM2.5-350M model • in 100 GRPO steps using TRL • for better structured outputs blog: https:
50.6K views · 992 likes · 114 reposts · 13 replies Open on X →
A tier: having a blog and x presence 1.9K views · 27 likes · 0 reposts · 1 replies Open on X →
Work your ass off that you become "the guy" https://t.co/WB32ImZ7dR
15.5K views · 410 likes · 14 reposts · 23 replies Open on X →
@helloiamleonie sorry to schmidhuber you here but speculative decoding was already explored by noam back in 2018 https://t.co/ZSegJlSBln 382 views · 4 likes · 0 reposts · 0 replies Open on X →
@helloiamleonie One thing from the practice side: the win tracks acceptance rate, and acceptance rate is workload dependent. Someone benchmarked DFlash 2 on Qwen3.8-27B this week and found an added n-gram drafter went from +1% on real coding tasks to -30% on prose. 896 views · 4 likes · 0 reposts · 1 replies Open on X →
@helloiamleonie Now I feel bad for taking about it for a week 😅 1.2K views · 5 likes · 0 reposts · 1 replies Open on X →
study notes on speculative decoding https://t.co/tsIYoxzsqd https://t.co/LUCSi5Cl2y
85.6K views · 1.2K likes · 147 reposts · 15 replies Open on X →

Im Vergleich zu Konten gleicher Größe

10 Beiträge aus den letzten 90 Tagen, verglichen mit der Größenklasse 10K–100K Follower. wird mehr Menschen gezeigt als Konten gleicher Größe.

Medianaufrufe6 172dieses Konto924Median für 10K–100K
Reichweite, %29.53%dieses Konto3.62%Median für 10K–100K
Interaktion, %1.28%dieses Konto1.52%Median für 10K–100K
KennzahlDieses KontoMedian für 10K–100KVerhältnis
Medianaufrufe pro Beitrag6 1729246.68×
Reichweite (Aufrufe ÷ Follower)29.53%3.62%8.16×
Interaktionsrate1.28%1.52%0.84×

Weitere Konten dieser Größe →   Mit einem anderen Konto vergleichen →   Wie diese Vergleichswerte entstehen →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

85.6K23 Aug
1.2K
896
38224 Aug
15.5K4 Sep
1.9K
50.6K
9.4K
2.9K
17.6K7 Sep

Last 10 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

1.59%23 Aug
0.49%
0.56%
1.05%24 Aug
2.88%4 Sep
1.51%
2.21%
0.59%
0.83%
2.07%7 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes50.7%3 025 in total
Reposts5.3%315 in total
Replies1.2%70 in total
Bookmarks42.9%2 560 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Ähnliche Konten