foss

foss22's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in croqaz/Piston-and-Prose-sm about 2 months ago

Reproducibility appreciated

1

#1 opened about 2 months ago by

foss22

New activity in croqaz/vintage-words about 2 months ago

Extend vocab with songs phrases: 32k => 65k tokens

#7 opened about 2 months ago by

foss22

New activity in croqaz/tiny-vintage-completions about 2 months ago

1025 - ~2000 token text vs 2-3 words

👍 1

3

#2 opened 2 months ago by

foss22

New activity in croqaz/Sprocket-and-Say 2 months ago

Agentic 0,07 B LM tasks + complete list of use-cases for Vintage LMs

3

#3 opened 2 months ago by

foss22

77 million parameters is equal to 0.077 billion, not 0.007

❤️ 1

#2 opened 2 months ago by

foss22

New activity in croqaz/vintage-words 2 months ago

Per-profession vocab

👍 1

6

#4 opened 3 months ago by

foss22

RedBook tokenizier. Biodiversity-consetvation inspired

#6 opened 2 months ago by

foss22

New activity in croqaz/experiment-llama-77 3 months ago

Tokens indicate unbalanced subsets, that do not represent general English

#8 opened 3 months ago by

foss22

Å|į|Ã|§|©

1

#7 opened 3 months ago by

foss22

Remove ---- .... === ### dupes from datasets and tokenizer

👍 1

4

#3 opened 3 months ago by

foss22

Capitalization not normalized

#6 opened 3 months ago by

foss22

When capital letter matters?

#5 opened 3 months ago by

foss22

Page numbers dominate 645 num tokens

#4 opened 3 months ago by

foss22

New activity in croqaz/Sprocket-and-Say 3 months ago

Collect test inference or no value of collecting?

6

#1 opened 3 months ago by

foss22

New activity in croqaz/vintage-words 3 months ago

29 of 798 Wikipedia 1900-1990s neologisms

5

#1 opened 3 months ago by

foss22

New activity in croqaz/Sprocket-n-Say 3 months ago

Questions in 454k vs 783k without questions

1

#5 opened 3 months ago by

foss22

Redundancy

1

#4 opened 3 months ago by

foss22

Under-filtering.

1

#3 opened 3 months ago by

foss22

Over-filtering problem with current banned.txt

1

#2 opened 3 months ago by

foss22

New activity in croqaz/experiment-llama-77 3 months ago

F16 NaN in 10 steps workaround

1

#2 opened 3 months ago by

foss22