novarobot
/
RWKV-LM
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

better chat

Browse Source
main
BlinkDL 3 years ago
parent cf340264dc
commit 59e6deeb58
1 changed files with 92 additions and 42 deletions
Show Stats Download Patch File Download Diff File

134
RWKV-v4neo/chat.py
Unescape Escape View File

@ -17,6 +17,8 @@ torch.backends.cudnn.allow_tf32 = True
torch.backends.cuda.matmul.allow_tf32 = True
np.set_printoptions(precision=4, suppress=True, linewidth=200)
CHAT_LANG = 'English' # English Chinese
WORD_NAME = [
"20B_tokenizer.json",
"20B_tokenizer.json",
@ -48,17 +50,15 @@ args.ctx_len = 1024
# args.n_embd = 2560
# args.ctx_len = 1024
os.environ["RWKV_RUN_DEVICE"] = args.RUN_DEVICE
MODEL_NAME = args.MODEL_NAME
user = "User"
bot = "Bot"
interface = ":"
if CHAT_LANG == 'English':
user = "User"
bot = "Bot"
interface = ":"
# The following is a verbose and detailed conversation between an AI assistant called {bot}, and a human user called {user}. {bot} is intelligent, knowledgeable, wise and polite.
# The following is a conversation between a highly knowledgeable and intelligent AI called {bot}, and a human called {user}. In the following interactions, {user} and {bot} converse in natural language, and {bot} do its best to answer {user}'s questions. {bot} is respectful, polite and inclusive. {bot} knows a lot, and always tells the truth.
# The following is a verbose and detailed conversation between an AI assistant called {bot}, and a human user called {user}. {bot} is intelligent, knowledgeable, wise and polite.
# The following is a conversation between a highly knowledgeable and intelligent AI called {bot}, and a human called {user}. In the following interactions, {user} and {bot} converse in natural language, and {bot} do its best to answer {user}'s questions. {bot} is respectful, polite and inclusive. {bot} knows a lot, and always tells the truth.
init_prompt = f'''
init_prompt = f'''
The following is a verbose and detailed conversation between an AI assistant called {bot}, and a human user called {user}. {bot} is intelligent, knowledgeable, wise and polite.
{user}{interface} french revolution what year
@ -81,10 +81,58 @@ The following is a verbose and detailed conversation between an AI assistant cal
{bot}{interface} LHC is a high-energy particle collider, built by CERN, and completed in 2008. They used it to confirm the existence of the Higgs boson in 2012.
'''
HELP_MSG = '''Commands:
say something --> chat with bot. use \\n for new line.
+alt --> alternate chat reply
+reset --> reset chat
+gen YOUR PROMPT --> free generation with any prompt. use \\n for new line.
+qa YOUR QUESTION --> free generation - ask any question (just ask the question). use \\n for new line.
+more --> continue last free generation (only for +gen / +qa)
+retry --> retry last free generation (only for +gen / +qa)
Now talk with the bot and enjoy. Remember to +reset periodically to clean up the bot's memory. Use RWKV-4 14B for best results.
This is not instruct-tuned for conversation yet, so don't expect good quality. Better use +gen for free generation.
'''
elif CHAT_LANG == 'Chinese':
args.MODEL_NAME = '/fsx/BlinkDL/CODE/_PUBLIC_/RWKV-LM/RWKV-v4neo/7-run3z/rwkv-293'
args.n_layer = 32
args.n_embd = 4096
args.ctx_len = 1024
user = "Q"
bot = "A"
interface = ":"
init_prompt = '''
Q: 企鹅会飞吗？
A: 企鹅是不会飞的。它们的翅膀主要用于游泳和平衡，而不是飞行。
Q: 西瓜是什么
A: 西瓜是一种常见的水果，是一种多年生蔓生藤本植物。西瓜的果实呈圆形或卵形，通常是绿色的，里面有红色或黄色的肉和很多的籽。西瓜味甜，多吃可以增加水分，是夏季非常受欢迎的水果之一。
'''
HELP_MSG = '''指令:
直接输入内容 --> 和机器人聊天，用\\n代表换行
+alt --> 让机器人换个回答
+reset --> 重置对话
+gen 某某内容 --> 续写任何中英文内容，用\\n代表换行
+qa 某某问题 --> 问独立的问题（忽略上下文），用\\n代表换行
+more --> 继续 +gen / +qa 的回答
+retry --> 换个 +gen / +qa 的回答
现在可以输入内容和机器人聊天（注意它不怎么懂中文，它可能更懂英文）。请经常使用 +reset 重置机器人记忆。
'''
# Load Model
os.environ["RWKV_RUN_DEVICE"] = args.RUN_DEVICE
MODEL_NAME = args.MODEL_NAME
print(f'loading... {MODEL_NAME}')
model = RWKV_RNN(args)
@ -107,8 +155,8 @@ def run_rnn(tokens, newline_adj = 0):
out[0] = -999999999 # disable <|endoftext|>
out[187] += newline_adj
if newline_adj > 0:
out[15] += newline_adj / 2 # '.'
# if newline_adj > 0:
# out[15] += newline_adj / 2 # '.'
return out
all_state = {}
@ -144,14 +192,14 @@ for s in srv_list:
print(f'### prompt ###\n[{tokenizer.tokenizer.decode(model_tokens)}]\n')
def reply_msg(msg):
print('Bot:', msg + '\n')
print(f'{bot}{interface} {msg}\n')
def on_message(message):
global model_tokens, current_state
srv = 'dummy_server'
msg = message.strip()
msg = message.replace('\\n','\n').strip()
if len(msg) > 1000:
reply_msg('your message is too long (max 1000 tokens)')
return
@ -179,16 +227,16 @@ def on_message(message):
reply_msg("Chat reset.")
return
elif msg[:5] == '+gen ' or msg[:4] == '+qa ' or msg == '+more' or msg == '+retry':
elif msg[:5].lower() == '+gen ' or msg[:4].lower() == '+qa ' or msg.lower() == '+more' or msg.lower() == '+retry':
if msg[:5] == '+gen ':
if msg[:5].lower() == '+gen ':
new = '\n' + msg[5:].strip()
# print(f'### prompt ###\n[{new}]')
current_state = None
out = run_rnn(tokenizer.tokenizer.encode(new))
save_all_stat(srv, 'gen_0', out)
elif msg[:4] == '+qa ':
elif msg[:4].lower() == '+qa ':
out = load_all_stat('', 'chat_init')
real_msg = msg[4:].strip()
@ -204,21 +252,22 @@ def on_message(message):
# out = run_rnn(tokenizer.tokenizer.encode(new))
# save_all_stat(srv, 'gen_0', out)
elif msg == '+more':
elif msg.lower() == '+more':
try:
out = load_all_stat(srv, 'gen_1')
save_all_stat(srv, 'gen_0', out)
except:
return
elif msg == '+retry':
elif msg.lower() == '+retry':
try:
out = load_all_stat(srv, 'gen_0')
except:
return
begin = len(model_tokens)
for i in range(100):
out_last = begin
for i in range(150):
token = tokenizer.sample_logits(
out,
model_tokens,
@ -227,17 +276,23 @@ def on_message(message):
top_p_usual=x_top_p,
top_p_newline=x_top_p,
)
if msg[:4] == '+qa ':
if msg[:4].lower() == '+qa ':
out = run_rnn([token], newline_adj=-1)
else:
out = run_rnn([token])
send_msg = tokenizer.tokenizer.decode(model_tokens[begin:]).strip()
xxx = tokenizer.tokenizer.decode(model_tokens[out_last:])
if '\ufffd' not in xxx:
print(xxx, end='', flush=True)
out_last = begin + i + 1
print('\n')
# send_msg = tokenizer.tokenizer.decode(model_tokens[begin:]).strip()
# print(f'### send ###\n[{send_msg}]')
reply_msg(send_msg)
# reply_msg(send_msg)
save_all_stat(srv, 'gen_1', out)
else:
if msg == '+alt':
if msg.lower() == '+alt':
try:
out = load_all_stat(srv, 'chat_pre')
except:
@ -250,17 +305,17 @@ def on_message(message):
save_all_stat(srv, 'chat_pre', out)
begin = len(model_tokens)
for i in range(120):
out_last = begin
print(f'{bot}{interface}', end='', flush=True)
for i in range(999):
if i <= 0:
newline_adj = -999999999
elif i <= 30:
newline_adj = (i - 30) / 10
elif i <= 80:
elif i <= 130:
newline_adj = 0
elif i <= 117:
newline_adj = (i - 80) * 0.5
else:
newline_adj = 999999999
newline_adj = (i - 130) * 0.25 # MUST END THE GENERATION
token = tokenizer.sample_logits(
out,
model_tokens,
@ -270,6 +325,12 @@ def on_message(message):
top_p_newline=x_top_p,
)
out = run_rnn([token], newline_adj=newline_adj)
xxx = tokenizer.tokenizer.decode(model_tokens[out_last:])
if '\ufffd' not in xxx:
print(xxx, end='', flush=True)
out_last = begin + i + 1
send_msg = tokenizer.tokenizer.decode(model_tokens[begin:])
if '\n\n' in send_msg:
send_msg = send_msg.strip()
@ -287,24 +348,13 @@ def on_message(message):
# print(f'[{tokenizer.tokenizer.decode(model_tokens)}]')
# print(f'### send ###\n[{send_msg}]')
reply_msg(send_msg)
# reply_msg(send_msg)
save_all_stat(srv, 'chat', out)
print('''Commands:
+alt --> alternate chat reply
+reset --> reset chat
+gen YOUR PROMPT --> free generation with your prompt
+qa YOUR QUESTION --> free generation - ask any question and get answer (just ask the question)
+more --> continue last free generation [does not work for chat]
+retry --> retry last free generation
Now talk with the bot and enjoy. Remember to +reset periodically to clean up the bot's memory. Use RWKV-4 14B for best results.
This is not instruct-tuned for conversation yet, so don't expect good quality. Better use +rwkv_gen for free generation.
''')
print(HELP_MSG)
while True:
msg = input('User: ')
msg = input(f'{user}{interface} ')
if len(msg.strip()) > 0:
on_message(msg)
else:

Write Preview
Loading…
Cancel
Save