Update app.py

app.py

@@ -6,14 +6,14 @@ from pynvml import *
 nvmlInit()
 gpu_h = nvmlDeviceGetHandleByIndex(0)
 ctx_limit = 3000
-title = "RWKV-v5-Eagle-World-7B-v2-20240128-ctx4096"
+title = "EagleX 1.7T Demo"
 
 os.environ["RWKV_JIT_ON"] = '1'
 os.environ["RWKV_CUDA_ON"] = '1' # if '1' then use CUDA kernel for seq mode (much faster)
 
 from rwkv.model import RWKV
-model_path = hf_hub_download(repo_id="RWKV/v5-Eagle-7B", filename=f"{title}.pth")
-model = RWKV(model=model_path, strategy='cuda fp16i8 *8 -> cuda fp16')
+model_path = hf_hub_download(repo_id="recursal/EagleX_1-7T", filename="EagleX-1_7T.pth")
+model = RWKV(model=model_path, strategy='cuda bf16')
 from rwkv.utils import PIPELINE, PIPELINE_ARGS
 pipeline = PIPELINE(model, "rwkv_vocab_v20230424")
 
@@ -27,13 +27,13 @@ def generate_prompt(instruction, input=""):
 
 ### Response:"""
     else:
-        return f"""### User: hi
+        return f"""User: hi
 
-### Assistant: Hi. I am your assistant and I will provide expert full response in full details. Please feel free to ask any question and I will always answer it.
+Assistant: Hi. I am your assistant and I will provide expert full response in full details. Please feel free to ask any question and I will always answer it.
 
-### User: {instruction}
+User: {instruction}
 
-### Assistant:"""
+Assistant:"""
 
 def evaluate(
     ctx,