ModelLink2/modellink
wucong 0f0e6245f6 !1624 新增llama3-8b微调脚本+readme更新
Merge pull request !1624 from wucong/addllama3
2024-09-12 06:38:36 +00:00
..
core !1658 【mcore】添加baichuan2-13B和deepseek2-coder适配 2024-09-12 01:04:21 +00:00
data !1478 微调与Llamafactory差异修改 2024-08-20 02:43:08 +00:00
model !1642 mod partial_rope to glm_rope 2024-09-10 06:14:35 +00:00
patchs !1642 mod partial_rope to glm_rope 2024-09-10 06:14:35 +00:00
tasks !1658 【mcore】添加baichuan2-13B和deepseek2-coder适配 2024-09-12 01:04:21 +00:00
tokenizer !1455 prompt-type推理适配 2024-08-13 06:53:26 +00:00
__init__.py !1218 迁移megatron patch 2024-04-23 01:57:03 +00:00
arguments.py !1624 新增llama3-8b微调脚本+readme更新 2024-09-12 06:38:36 +00:00
checkpointing.py !1605 [mcore-llm]类deepseekv2模型增加hf2mg,mg2hf权重转换与预训练数据集处理 2024-09-05 01:00:05 +00:00
error_utils.py !1448 删除冗余内容 2024-07-23 09:54:53 +00:00
initialize.py !1448 删除冗余内容 2024-07-23 09:54:53 +00:00
training.py !1584 TTP/UCE支持bf16、fp32+接口更新 2024-09-03 02:15:34 +00:00
utils.py !1591 【varlen场景适配】适配layout=TND格式的FA 2024-09-03 11:11:41 +00:00