vrr/kvcache-ai-ktransformers
2
0
Fork 0
mirror of https://github.com/kvcache-ai/ktransformers.git synced 2025-09-09 13:55:27 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

support qwen3

Browse source
This commit is contained in:
djw 2025-04-28 18:15:35 +00:00
parent 68c2b2e6e6
commit 33cbd47086
8 changed files with 195 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

4
ktransformers/server/backend/interfaces/balance_serve.py
View file

@ -56,8 +56,8 @@ ktransformer_rules_dir = (
os.path.join(os.path.dirname(os.path.abspath(__file__)), "..", "..", "..", "./optimize/optimize_rules/")
)
default_optimize_rules = {
"DeepseekV3ForCausalLM": ktransformer_rules_dir + "Moonlight-16B-A3B-serve.yaml",
# "DeepseekV3ForCausalLM": ktransformer_rules_dir + "DeepSeek-V3-Chat-serve.yaml",
# "DeepseekV3ForCausalLM": ktransformer_rules_dir + "Moonlight-16B-A3B-serve.yaml",
"DeepseekV3ForCausalLM": ktransformer_rules_dir + "DeepSeek-V3-Chat-serve.yaml",
"Qwen2MoeForCausalLM": ktransformer_rules_dir + "Qwen2-serve.yaml",
"Qwen3MoeForCausalLM": ktransformer_rules_dir + "Qwen3Moe-serve.yaml",
}