new methot of generation

ChatAI/chat_ai.py

@@ -1,3 +1,3 @@
-from transformers import pipeline
+from huggingface_hub import InferenceClient
 
-pipe = pipeline("text2text-generation", model="facebook/blenderbot-400M-distill")
+pipe = InferenceClient("mistralai/Mistral-7B-Instruct-v0.3")

app.py

@@ -32,10 +32,11 @@ async def on_message(message):
         async for message in channel.history(limit=4):
             messages.append(message.content)
 
-        response = ai("\n".join(messages), max_length=1900, truncation=True, pad_token_id=50256)
+        previous_messages = ("\n".join(messages))
 
-
-        await channel.send(response[0]['generated_text'])
+        x = generate(previous_messages, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0)
+        print(x)
+        await channel.send(x)
 
 
         message_counts[message.channel.id] = 0 # Reset the counter
@@ -43,6 +44,42 @@ async def on_message(message):
 
     await bot.process_commands(message) # Ensure commands still work
 
+def format_prompt(message, history):
+  prompt = "<s>"
+  for user_prompt, bot_response in history:
+    prompt += f"[INST] {user_prompt} [/INST]"
+    prompt += f" {bot_response}</s> "
+  prompt += f"[INST] {message} [/INST]"
+  return prompt
+
+def generate(
+    prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0,
+):
+    temperature = float(temperature)
+    if temperature < 1e-2:
+        temperature = 1e-2
+    top_p = float(top_p)
+
+    generate_kwargs = dict(
+        temperature=temperature,
+        max_new_tokens=max_new_tokens,
+        top_p=top_p,
+        repetition_penalty=repetition_penalty,
+        do_sample=True,
+        seed=42,
+    )
+
+    formatted_prompt = format_prompt(prompt)
+
+    stream = ai.text_generation(formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
+    output = ""
+
+    for response in stream:
+        output += response.token.text
+        yield output
+    return output
+
+
 @bot.event
 async def on_ready():
     print(f'Logged in as {bot.user}')  # Logs bot login in console