Update app.py

app.py

@@ -6,40 +6,30 @@ from pynvml import *
 nvmlInit()
 gpu_h = nvmlDeviceGetHandleByIndex(0)
 ctx_limit = 1536
-title = "RWKV-4-Raven-7B-v12-Eng98%-Other2%-20230521-ctx8192"
+title = "RWKV-4-World-7B-v1-OnlyForTest_84%_trained-20230618-ctx4096"
 
 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="BlinkDL/rwkv-4-raven", filename=f"{title}.pth")
+model_path = hf_hub_download(repo_id="BlinkDL/rwkv-4-world", filename=f"{title}.pth")
 model = RWKV(model=model_path, strategy='cuda fp16i8 *8 -> cuda fp16')
-# model = RWKV(model='D:/ChatRWKV/RWKV-4-Raven-7B-v9-Eng99%-Other1%-20230412-ctx8192.pth', strategy='cuda fp16i8 *10 -> cuda fp16')
 from rwkv.utils import PIPELINE, PIPELINE_ARGS
 pipeline = PIPELINE(model, "20B_tokenizer.json")
 
 def generate_prompt(instruction, input=None):
-    instruction = instruction.strip().replace('\r\n','\n').replace('\n\n','\n')
-    input = input.strip().replace('\r\n','\n').replace('\n\n','\n')
+    instruction = instruction.strip().replace('\r\n','\n').replace('\n\n','\n').replace('\n\n','\n')
+    input = input.strip().replace('\r\n','\n').replace('\n\n','\n').replace('\n\n','\n')
     if input:
-        return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
+        return f"""Instruction: {instruction}
 
-# Instruction:
-{instruction}
+Input: {input}
 
-# Input:
-{input}
-
-# Response:
-"""
+Response:"""
     else:
-        return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.
-
-# Instruction:
-{instruction}
+        return f"""Question: {instruction}
 
-# Response:
-"""
+Answer:"""
 
 def evaluate(
     instruction,
@@ -56,8 +46,8 @@ def evaluate(
                      token_ban = [], # ban the generation of some tokens
                      token_stop = [0]) # stop generation whenever you see any token here
 
-    instruction = instruction.strip().replace('\r\n','\n').replace('\n\n','\n')
-    input = input.strip().replace('\r\n','\n').replace('\n\n','\n')
+    instruction = instruction.strip().replace('\r\n','\n').replace('\n\n','\n').replace('\n\n','\n')
+    input = input.strip().replace('\r\n','\n').replace('\n\n','\n').replace('\n\n','\n')
     ctx = generate_prompt(instruction, input)
     
     all_tokens = []
@@ -243,9 +233,9 @@ def chat(
 ##########################################################################
 
 with gr.Blocks(title=title) as demo:
-    gr.HTML(f"<div style=\"text-align: center;\">\n<h1>🐦Raven - {title}</h1>\n</div>")
+    gr.HTML(f"<div style=\"text-align: center;\">\n<h1>🌍World - {title}</h1>\n</div>")
     with gr.Tab("Instruct mode"):
-        gr.Markdown(f"Raven is [RWKV 7B](https://github.com/BlinkDL/ChatRWKV) 100% RNN [RWKV-LM](https://github.com/BlinkDL/RWKV-LM) finetuned to follow instructions. *** Please try examples first (bottom of page) *** (edit them to use your question). Demo limited to ctxlen {ctx_limit}. Finetuned on alpaca, gpt4all, codealpaca and more. For best results, *** keep you prompt short and clear ***. <b>UPDATE: now with Chat (see above, as a tab) ==> turn off as of now due to VRAM leak caused by buggy code.</b>.")
+        gr.Markdown(f"World is [RWKV 7B](https://github.com/BlinkDL/ChatRWKV) 100% RNN [RWKV-LM](https://github.com/BlinkDL/RWKV-LM) trained on 100+ world languages. *** Please try examples first (bottom of page) *** (edit them to use your question). Demo limited to ctxlen {ctx_limit}. Finetuned on alpaca, gpt4all, codealpaca and more. For best results, *** keep you prompt short and clear ***.</b>.") # <b>UPDATE: now with Chat (see above, as a tab) ==> turn off as of now due to VRAM leak caused by buggy code.
         with gr.Row():
             with gr.Column():
                 instruction = gr.Textbox(lines=2, label="Instruction", value="Tell me about ravens.")