Ceaser1717 21a919d79f

Fix: prevent infinite “A5A5A5...” repetition loop during text generation (Issue #1008) ...

Added a simple repetition detection inside the generate() loop to stop the model 
from endlessly producing patterns like “A5A5A5...”. 

- Imported `re` for regex pattern matching.
- Stops generation if “A5” repeats 10+ times or the same token appears 10 times.
- Prints a warning and exits safely instead of looping infinitely.

Fixes: #1008

2025-10-21 16:02:03 +05:30

..

configs

support scale_fmt=ue8m0 (#964)

2025-08-27 15:30:21 +08:00

convert.py

clarify assertion error

2025-01-28 13:16:54 +01:00

fp8_cast_bf16.py

Enhance documentation and update .gitignore for model conversion scripts

2025-01-05 18:18:18 +00:00

generate.py

Fix: prevent infinite “A5A5A5...” repetition loop during text generation (Issue #1008)

2025-10-21 16:02:03 +05:30

kernel.py

fix rmsnorm and act_quant_kernel

2025-08-27 17:12:13 +08:00

model.py

fix rmsnorm and act_quant_kernel

2025-08-27 17:12:13 +08:00

requirements.txt

Release DeepSeek-V3

2024-12-26 19:01:57 +08:00