Skip to content

Adapt Eagle3 for Deepseek architecture#186

Open
xuhaojie-2025 wants to merge 1 commit intosgl-project:mainfrom
bytedance-iaas:feat/dpsk_v3
Open

Adapt Eagle3 for Deepseek architecture#186
xuhaojie-2025 wants to merge 1 commit intosgl-project:mainfrom
bytedance-iaas:feat/dpsk_v3

Conversation

@xuhaojie-2025
Copy link

Motivation

Implement adaptation of this framework for DeepseekV3ForCausalLM

Modifications

Created a new file specforge/modeling/draft/deepseekv3_eagle.py under specforge/modeling/draft, designed DeepseekV3ForCausalLMEagle3 for Deepseek and registered it in relevant files.
Added deepseek_r1 and deepseek_v3 conversation templates in specforge/data/template.py.

Related Issues

Accuracy Test

Benchmark & Profiling

Checklist

@gemini-code-assist
Copy link
Contributor

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant