Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
llm: Enable flash attention by default for qwen3 and qwen3moe
Jesse Gross0bda72892cc8
diff --git a/fs/ggml/ggml.go b/fs/ggml/ggml.goindex 58803f58f..d34a69138 100644--- a/fs/ggml/ggml.go+++ b/fs/ggml/ggml.go@@ -899,6 +899,8 @@ func (f GGML) SupportsFlashAttention() bool {func (f GGML) FlashAttention() bool {return slices.Contains([]string{"gptoss", "gpt-oss",+ "qwen3",+ "qwen3moe",}, f.KV().String("general.architecture"))}