初始化项目,由ModelHub XC社区提供模型

Model: bcoding/deepseek-R1-0528-Qwen3-8b-base-16-llvm-sft-trained-16bit-epoch-1-merged
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-09-05 07:16:16 +08:00
commit 4b74f7c14b
13 changed files with 929 additions and 0 deletions

8
generation_config.json Normal file
View File

@@ -0,0 +1,8 @@
{
"_from_model_config": true,
"bos_token_id": 151643,
"eos_token_id": 151645,
"max_length": 131072,
"pad_token_id": 151654,
"transformers_version": "4.57.6"
}