diff --git a/server/text_generation_server/models/causal_lm.py b/server/text_generation_server/models/causal_lm.py index 831b0b5e..f814a8e2 100644 --- a/server/text_generation_server/models/causal_lm.py +++ b/server/text_generation_server/models/causal_lm.py @@ -691,9 +691,9 @@ class CausalLM(Model): "return_dict": True, } - if model.config.model_type in ["llama", "mistral", "starcoder2"]: + if model.config.model_type in ["llama", "mistral", "starcoder2", "qwen2"]: - if model.config.model_type in ["llama", "mistral"]: + if model.config.model_type in ["llama", "mistral", "qwen2"]: kwargs["attn_softmax_bf16"] = True kwargs["trim_logits"] = True