Spaces:
Sleeping
Sleeping
Commit History
-
disable cache for better compability
ea7c44f
-
make qwen2.5-taiwan 1.5b the default
493107e -
upgrade transforemrs to 4.53.3+
46df16c -
upgrade transformers to 4.46.1+
2a1d352 -
add timm to requirements
0aa7a80 -
add access_token to model loader
882aef2 -
add gemma 3n, smallthinker 4b0.6b and falcon h1-1.5b
de303d7
-
rename duckduckgo_search to ddgs
b858214 -
remove qwen3 30ba3b & 32b, but add 14b
e883ef3 -
use fp8 version of qwen3 30ba3b & 32b
9ece009 -
add autoawq
df9e599 -
add qwe3 30ba3b awq & 30b awq
3a690c1
-
update prompt
8d42201 -
update prompt
5398786 -
fix err in prompt
b1fb4c7 -
improve prompt for better web search result merging
6a87eb4 -
Add D-R1-0528 8B, Nemotron 1.5B, Qwen2.5 Taiwan 1.5B
12d3fe3
-
reduce rep penality to 1.2
079e166 -
increase rep penality as on many slms, reptition is frequent.
8a7c97d -
revert device_map to audo
cc12a51 -
Add smolLM2-135M GRPO and SmolLM2-360M
cb16ac6
-
Update README.md
Verified signature371669a -
update readme
076c1f2
-
add Qwen2.5-Omni-3B & MiMo-7B-RL
6a4537b -
disable models unrunnable on hf spaces
7c5f318 -
switch Phi-4-mini-Instruct" from unsloth to microsoft
23b3848 -
extend max tokens
d730ffe -
adjust layout
4911925 -
add "Granite-4.0-Tiny-Preview" model
09d9700 -
add phi-4 reasonning and phi-4-mini reasonning
68e6569 -
bugfix: set device to xpu by mistake
1242438 -
add all qwen3 variants
2882063 -
user can define search timeout
e2ee907 -
give 5 second for web earch to gather reults
c00d442 -
support thinking models and streamingly display thought
8c3c2b9 -
do not preview prompt at error return from chat response
c09049b -
inject assistant placeholder at right time
12dd3f3 -
disable L142 which is not needed
3c176a1 -
fix bug in prompt preview display
41ee8bf -
add prompt preview for debug
5fc0117 -
fix: prevent self-talking issue by using tokenizer chat_template formatting
- Updated prompt construction to use `tokenizer.apply_chat_template()` for chat-tuned models - Ensured backward compatibility with non-chat models by falling back to manual prompt formatting - Resolves issue where models would echo both user and assistant messages due to incorrect flat-text prompts960db60
-
bugfix to Error: "str" object has no attribute "pad_token_id"
889f080 -
add taiwan elm 1.1b & 270m instruct
c8399e3
-
add type in qwen3 0.6b repo id
76d4d60 -
add qwen3
fe395ab
-
Add Smollm2 360m instruct fine-tuned on TaiwanChat
7308211
-
keep debug message
37f7787 -
add debug to show web resarch result
a2f07a4 -
give 1 second for web search to grab data
9ad3ffd -
inject web search result if web search enabled
bc257ff