15 1 51

Soshyant PRO

Respair

AI & ML interests

Speech and Audio Processing, Natural Language Processing

Recent Activity

new activity about 12 hours ago

Respair/Tsukasa_Speech:ValueError when attempting to train

updated a model 1 day ago

Respair/Voc

published a model 1 day ago

Respair/Voc

View all activity

Organizations

Respair's activity

New activity in Respair/Tsukasa_Speech about 12 hours ago

ValueError when attempting to train

#2 opened about 12 hours ago by

sumomomomomo

updated a model 1 day ago

Respair/Voc

Updated 1 day ago

published a model 1 day ago

Respair/Voc

Updated 1 day ago

updated 2 models 2 days ago

Respair/stts_test_ckpt

Updated 2 days ago

Respair/Text_Aligners

Updated 2 days ago • 1

updated a model 3 days ago

Respair/ar_cache

Updated 3 days ago

liked a Space 4 days ago

Running

🥰🎤🤔

PartAI/Tooka-SBERT

VITA-MLLM/VITA-1.5

Video-Text-to-Text • Updated 15 days ago • 845 • 33

yl4579/StyleTTS

Updated 17 days ago • 2

New activity in Respair/Tsukasa_Speech 7 days ago

Can u please opensource the right weights?

#1 opened 15 days ago by

lucasjin

liked a model 7 days ago

novateur/aaa

Updated 16 days ago • 1

updated a model 8 days ago

Respair/Tsukasa_Speech

Text-to-Speech • Updated 8 days ago • 30

reacted to leonardlin's post with 👍 8 days ago

Post

1971

My weekened project ended up being doing some testing between torchtune, axolotl, and unsloth. I *think* it's a 1:1 comparison of what LoRA fine-tuning performance looks like between the different hardware I have in my dev boxes (4090, 3090, 7900 XTX, W7900) with a few other interesting tidbits.

Tonight I wrote up a WandB report (the panel editor is super broken in Firefox 😔) that sums up some of the more interesting bits from the results: https://wandb.ai/augmxnt/train-bench/reports/torchtune-vs-axolotl-vs-unsloth-Trainer-Comparison--Vmlldzo4MzU3NTAx