Thiago de Arruda Padilha

tarruda's activity

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

New activity in stepfun-ai/Step-3.7-Flash-GGUF about 2 hours ago

Some practical tips for using Step 3.7 GGUFs

🧠👍 2

6

#6 opened 26 days ago by

tarruda

New activity in stepfun-ai/Step-3.7-Flash 2 days ago

Can we add something like Qwen's preserve_thinking to the chat template?

#15 opened 2 days ago by

tarruda

New activity in AesSedai/Step-3.7-Flash-GGUF 3 days ago

PPQ(Q)/PPL(base) and KLD seems significantly worse than Step 3.5 Flash

👀 1

23

#2 opened 27 days ago by

tarruda

New activity in bartowski/prefeitura-rio_Rio-3.5-Open-397B-GGUF 8 days ago

Final model?

👍 1

7

#1 opened 12 days ago by

blankreg

New activity in WithinUsAI/GPT5.5_thinking_max_distill_god_seed_25K 9 days ago

Aren't the thinking traces hidden in the OpenAI API?

1

#2 opened 9 days ago by

tarruda

New activity in nex-agi/Nex-N2-Pro 10 days ago

Model Looping in Reasoning

2

#12 opened 11 days ago by

remichu

New activity in prefeitura-rio/Rio-3.5-Open-397B 11 days ago

Despite the way the project was released, this is a good model.

👍 5

9

#25 opened 11 days ago by

tarruda

New activity in prefeitura-rio/Rio-3.5-Open-397B 13 days ago

Gguf

👍 5

5

#1 opened 14 days ago by

cdani

New activity in stepfun-ai/Step-3.7-Flash 23 days ago

Supers Found in model

❤️ 2

9

#11 opened 25 days ago by

tcclaviger

New activity in stepfun-ai/Step-3.7-Flash 25 days ago

Tool calling bug in the official API

👍 1

#12 opened 25 days ago by

tarruda

New activity in stepfun-ai/Step-3.7-Flash-GGUF 29 days ago

imatrix corpus

3

#3 opened 29 days ago by

ghit72

New activity in AesSedai/Qwen3.5-397B-A17B-GGUF about 1 month ago

MTP support merged in llama.cpp

13

#10 opened about 1 month ago by

tarruda

New activity in ubergarm/Kimi-K2.6-GGUF about 1 month ago

MTP Support

👍 2

9

#13 opened about 1 month ago by

anikifoss

New activity in antirez/deepseek-v4-gguf about 1 month ago

Suggestion: Add IQ3 variants.

👍 2

4

#7 opened about 2 months ago by

tarruda

New activity in AesSedai/MiMo-V2.5-GGUF about 2 months ago

Any llama.cpp parameters to work around the looping?

👍 1

2

#12 opened about 2 months ago by

tarruda

Can you share the command you used to create the BF16 ggus?

2

#10 opened about 2 months ago by

tarruda

IQ3_S doesn't load with "wrong number of tensors"

2

#9 opened about 2 months ago by

tarruda

New activity in llmfan46/Qwen3.6-27B-uncensored-heretic-v2-Native-MTP-Preserved-GGUF about 2 months ago

Can the MTP version currently be loaded in LM Studio?

🚀👍 3

5

#2 opened about 2 months ago by

cclongdevil

New activity in llmfan46/Qwen3.6-35B-A3B-uncensored-heretic-GGUF about 2 months ago

Excellent work on this one

4

#8 opened about 2 months ago by

tarruda

New activity in llmfan46/Qwen3.6-27B-uncensored-heretic-v2 about 2 months ago

Keep the mtp layer weight

👍 2

10

#5 opened about 2 months ago by

Light4Bear