Florian Zimmermeister's picture

Florian Zimmermeister

flozi00

·

AI & ML interests

ASR, German LLM

Recent Activity

updated a dataset 1 day ago

flozi00/InstructGer-exp

Reacted to m-ric's post with 👀 5 days ago

𝐒𝐜𝐚𝐥𝐢𝐧𝐠 𝐥𝐚𝐰𝐬 𝐚𝐫𝐞 𝐧𝐨𝐭 𝐝𝐞𝐚𝐝 𝐲𝐞𝐭! New blog post suggests Anthropic might have an extremely strong Opus-3.5 already available, but is not releasing it to keep their edge over the competition. 🧐 ❓Since the release of Opus-3.5 has been delayed indefinitely, there have been lots of rumors and articles about LLMs plateauing. Scaling laws, the main powering factor of the LLM competence increase, could have stopped, according to these rumors, being the cause of this stalling of progress. These rumors were quickly denied by many people at the leading LLM labs, including OpenAI and Anthropic. But these people would be expected to hype the future of LLMs even if scaling laws really plateaued, so the jury is still out. 🗞️ This new article by Semianalysis (generally a good source, specifically on hardware) provides a counter-rumor that I find more convincing: ➡️ Maybe scaling laws still work, Opus-3.5 is ready and as good as planned, but they just don't release it because the synthetic data it helps provide can bring cheaper/smaller models Claude and Haiku up in performance, without risking to leak this precious high-quality synthetic data to competitors. Time will tell! I feel like we'll know more soon. Read the article: https://semianalysis.com/2024/12/11/scaling-laws-o1-pro-architecture-reasoning-infrastructure-orion-and-claude-3-5-opus-failures/

liked a dataset 8 days ago

nvidia/OpenMathInstruct-2

View all activity

Organizations

$A\\Ware's profile picture$

flozi00's activity

New activity in primeline/whisper-large-v3-turbo-german about 1 month ago

Convert to .bin?

#4 opened about 2 months ago by

New activity in primeline/whisper-large-v3-turbo-german about 2 months ago

german or swiss-german

#5 opened about 2 months ago by

New activity in primeline/whisper-large-v3-turbo-german 2 months ago

Comparison with the distilled model

#3 opened 2 months ago by

Evaluations are a bit disingenuous

#1 opened 2 months ago by

New activity in primeline/whisper-large-v3-german 2 months ago

Nur ca. die erste Minute wird transkribiert

#6 opened 3 months ago by

New activity in pL-Community/SauerkrautLM-Mixtral-8x7B-Instruct-FP8-Dynamic 3 months ago

Update config.json

#1 opened 3 months ago by

New activity in pL-Community/SauerkrautLM-Nemo-12b-Instruct-FP8-Dynamic 3 months ago

Update config.json

#1 opened 3 months ago by

New activity in pL-Community/Llama-3.1-SauerkrautLM-70b-Instruct-FP8-Dynamic 3 months ago

Update config.json

#1 opened 3 months ago by

New activity in pL-Community/Qwen2.5-Coder-7B-Instruct-FP8-Dynamic 3 months ago

Update config.json

#1 opened 3 months ago by

New activity in pL-Community/Qwen2.5-72B-Instruct-FP8-Dynamic 3 months ago

Update config.json

#1 opened 3 months ago by

New activity in primeline/whisper-tiny-german 4 months ago

Upload tokenizer.json

#2 opened 4 months ago by

New activity in primeline/whisper-large-v3-german 5 months ago

Compatibility with other Clients

#5 opened 5 months ago by

New activity in flozi00/whisper-large-german-lora-cv13 7 months ago

Adding `safetensors` variant of this model

#2 opened 7 months ago by

New activity in primeline/whisper-tiny-german 8 months ago

Adding ONNX file of this model

#1 opened 8 months ago by

New activity in flozi00/falcon-7b-german-assistant-v2 8 months ago

Adding `safetensors` variant of this model

#1 opened 8 months ago by

New activity in Vezora/Mistral-22B-v0.1 8 months ago

Base Model or Finetuned Version?

#2 opened 8 months ago by

New activity in flozi00/distilwhisper-german-canary 9 months ago

Adding `safetensors` variant of this model

#1 opened 9 months ago by

New activity in aware-ai/whisper-tiny-german 10 months ago

Upload tokenizer.json

#3 opened 10 months ago by

Adding ONNX file of this model

#2 opened 10 months ago by

New activity in primeline/whisper-large-v3-german 11 months ago

Correct long-form generation config parameters 'max_initial_timestamp_index' and 'prev_sot_token_id'.

#4 opened 11 months ago by

patrickvonplaten