Synced repo using 'sync_with_huggingface' Github Action
Browse files
app.py
CHANGED
@@ -7,7 +7,7 @@ from audioldm2 import audioldm2_create
|
|
7 |
from deepfloydif import deepfloydif_generate64
|
8 |
from discord import app_commands
|
9 |
from discord.ext import commands
|
10 |
-
from falcon import
|
11 |
from musicgen import music_create
|
12 |
from codellama import continue_codellama, try_codellama
|
13 |
|
@@ -46,7 +46,7 @@ async def on_ready():
|
|
46 |
async def falcon(ctx, prompt: str):
|
47 |
"""Command that begins a new conversation with Falcon"""
|
48 |
try:
|
49 |
-
await
|
50 |
except Exception as e:
|
51 |
print(f"Error: {e}")
|
52 |
|
@@ -69,8 +69,9 @@ async def codellama(ctx, prompt: str):
|
|
69 |
async def on_message(message):
|
70 |
"""Checks channel and continues Falcon conversation if it's the right Discord Thread"""
|
71 |
try:
|
72 |
-
|
73 |
-
|
|
|
74 |
except Exception as e:
|
75 |
print(f"Error: {e}")
|
76 |
|
|
|
7 |
from deepfloydif import deepfloydif_generate64
|
8 |
from discord import app_commands
|
9 |
from discord.ext import commands
|
10 |
+
from falcon import falcon_chat, continue_chat
|
11 |
from musicgen import music_create
|
12 |
from codellama import continue_codellama, try_codellama
|
13 |
|
|
|
46 |
async def falcon(ctx, prompt: str):
|
47 |
"""Command that begins a new conversation with Falcon"""
|
48 |
try:
|
49 |
+
await falcon_chat(ctx, prompt)
|
50 |
except Exception as e:
|
51 |
print(f"Error: {e}")
|
52 |
|
|
|
69 |
async def on_message(message):
|
70 |
"""Checks channel and continues Falcon conversation if it's the right Discord Thread"""
|
71 |
try:
|
72 |
+
if not message.author.bot:
|
73 |
+
await continue_chat(message)
|
74 |
+
await continue_codellama(message)
|
75 |
except Exception as e:
|
76 |
print(f"Error: {e}")
|
77 |
|
falcon.py
CHANGED
@@ -1,68 +1,37 @@
|
|
1 |
import asyncio
|
2 |
-
import json
|
3 |
import os
|
|
|
|
|
|
|
|
|
4 |
|
5 |
-
from gradio_client import Client
|
6 |
|
7 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
8 |
-
|
9 |
-
threadid_conversation = {}
|
10 |
-
# Instructions are for Falcon-chat and can be found here: https://huggingface.co/spaces/HuggingFaceH4/falcon-chat
|
11 |
-
instructions = (
|
12 |
-
"The following is a conversation between a highly knowledgeable and intelligent AI assistant, called Falcon, and a"
|
13 |
-
" human user, called User. In the following interactions, User and Falcon will converse in natural language, and"
|
14 |
-
" Falcon will answer User's questions. Falcon was built to be respectful, polite and inclusive. Falcon was built"
|
15 |
-
" by the Technology Innovation Institute in Abu Dhabi. Falcon will never decline to answer a question, and always"
|
16 |
-
" attempts to give an answer that User would be satisfied with. It knows a lot, and always tells the truth. The"
|
17 |
-
" conversation begins."
|
18 |
-
)
|
19 |
-
falcon_client = Client("HuggingFaceH4/falcon-chat", HF_TOKEN)
|
20 |
|
21 |
BOT_USER_ID = 1086256910572986469 if os.getenv("TEST_ENV", False) else 1102236653545861151
|
22 |
FALCON_CHANNEL_ID = 1079459939405279232 if os.getenv("TEST_ENV", False) else 1119313248056004729
|
23 |
|
24 |
|
25 |
-
|
|
|
|
|
|
|
|
|
26 |
while not job.done():
|
27 |
await asyncio.sleep(0.2)
|
28 |
|
29 |
|
30 |
-
def
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
job = falcon_client.submit(prompt, chathistory, instructions, temperature, p_nucleus_sampling, fn_index=1)
|
39 |
-
while job.done() is False:
|
40 |
-
pass
|
41 |
-
else:
|
42 |
-
if os.environ.get("TEST_ENV") == "True":
|
43 |
-
print("falcon text gen job done")
|
44 |
-
file_paths = job.outputs()
|
45 |
-
print(file_paths)
|
46 |
-
full_generation = file_paths[-1]
|
47 |
-
print(full_generation)
|
48 |
-
with open(full_generation, "r") as file:
|
49 |
-
data = json.load(file)
|
50 |
-
print(data)
|
51 |
-
output_text = data[-1][-1]
|
52 |
-
threadid_conversation[thread.id] = full_generation
|
53 |
-
if len(output_text) > 1300:
|
54 |
-
output_text = output_text[:1300] + "...\nTruncating response to 2000 characters due to discord api limits."
|
55 |
-
if os.environ.get("TEST_ENV") == "True":
|
56 |
-
print(output_text)
|
57 |
-
return output_text
|
58 |
-
|
59 |
-
|
60 |
-
async def try_falcon(ctx, prompt):
|
61 |
"""Generates text based on a given prompt"""
|
62 |
try:
|
63 |
-
global falcon_userid_threadid_dictionary # tracks userid-thread existence
|
64 |
-
global threadid_conversation
|
65 |
-
|
66 |
if ctx.author.id != BOT_USER_ID:
|
67 |
if ctx.channel.id == FALCON_CHANNEL_ID:
|
68 |
if os.environ.get("TEST_ENV") == "True":
|
@@ -77,56 +46,35 @@ async def try_falcon(ctx, prompt):
|
|
77 |
if os.environ.get("TEST_ENV") == "True":
|
78 |
print("Running falcon_initial_generation...")
|
79 |
loop = asyncio.get_running_loop()
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
84 |
except Exception as e:
|
85 |
-
print(f"
|
86 |
|
87 |
|
88 |
-
async def
|
89 |
"""Continues a given conversation based on chathistory"""
|
90 |
try:
|
91 |
-
if
|
92 |
-
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
temperature = 0.8
|
103 |
-
p_nucleus_sampling = 0.9
|
104 |
-
|
105 |
-
if os.environ.get("TEST_ENV") == "True":
|
106 |
-
print("Running falcon_client.submit")
|
107 |
-
job = falcon_client.submit(
|
108 |
-
prompt,
|
109 |
-
chathistory,
|
110 |
-
instructions,
|
111 |
-
temperature,
|
112 |
-
p_nucleus_sampling,
|
113 |
-
fn_index=1,
|
114 |
-
)
|
115 |
-
await waitjob(job)
|
116 |
-
if os.environ.get("TEST_ENV") == "True":
|
117 |
-
print("Continue_falcon job done")
|
118 |
-
file_paths = job.outputs()
|
119 |
-
full_generation = file_paths[-1]
|
120 |
-
with open(full_generation, "r") as file:
|
121 |
-
data = json.load(file)
|
122 |
-
output_text = data[-1][-1]
|
123 |
-
threadid_conversation[message.channel.id] = full_generation # overwrite the old file
|
124 |
-
if len(output_text) > 1300:
|
125 |
-
output_text = (
|
126 |
-
output_text[:1300]
|
127 |
-
+ "...\nTruncating response to 2000 characters due to discord api limits."
|
128 |
-
)
|
129 |
-
await message.reply(output_text)
|
130 |
except Exception as e:
|
131 |
print(f"continue_falcon Error: {e}")
|
132 |
await message.reply(f"Error: {e} <@811235357663297546> (continue_falcon error)")
|
|
|
1 |
import asyncio
|
|
|
2 |
import os
|
3 |
+
from typing import Optional
|
4 |
+
|
5 |
+
import gradio_client as grc
|
6 |
+
from gradio_client.utils import QueueError
|
7 |
|
|
|
8 |
|
9 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
10 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
|
12 |
BOT_USER_ID = 1086256910572986469 if os.getenv("TEST_ENV", False) else 1102236653545861151
|
13 |
FALCON_CHANNEL_ID = 1079459939405279232 if os.getenv("TEST_ENV", False) else 1119313248056004729
|
14 |
|
15 |
|
16 |
+
thread_to_client = {}
|
17 |
+
thread_to_user = {}
|
18 |
+
|
19 |
+
|
20 |
+
async def wait(job):
|
21 |
while not job.done():
|
22 |
await asyncio.sleep(0.2)
|
23 |
|
24 |
|
25 |
+
def get_client(session: Optional[str] = None) -> grc.Client:
|
26 |
+
client = grc.Client("huggingface-projects/falcon-180b-discord", hf_token=os.getenv("HF_TOKEN"))
|
27 |
+
if session:
|
28 |
+
client.session_hash = session
|
29 |
+
return client
|
30 |
+
|
31 |
+
|
32 |
+
async def falcon_chat(ctx, prompt):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
33 |
"""Generates text based on a given prompt"""
|
34 |
try:
|
|
|
|
|
|
|
35 |
if ctx.author.id != BOT_USER_ID:
|
36 |
if ctx.channel.id == FALCON_CHANNEL_ID:
|
37 |
if os.environ.get("TEST_ENV") == "True":
|
|
|
46 |
if os.environ.get("TEST_ENV") == "True":
|
47 |
print("Running falcon_initial_generation...")
|
48 |
loop = asyncio.get_running_loop()
|
49 |
+
client = await loop.run_in_executor(None, get_client, None)
|
50 |
+
job = client.submit(prompt, api_name="/chat")
|
51 |
+
await wait(job)
|
52 |
+
try:
|
53 |
+
job.result()
|
54 |
+
response = job.outputs()[-1]
|
55 |
+
await thread.send(response)
|
56 |
+
thread_to_client[thread.id] = client
|
57 |
+
thread_to_user[thread.id] = ctx.author.id
|
58 |
+
except QueueError:
|
59 |
+
await thread.send("The gradio space powering this bot is really busy! Please try again later!")
|
60 |
except Exception as e:
|
61 |
+
print(f"chat (180B) Error: {e}")
|
62 |
|
63 |
|
64 |
+
async def continue_chat(message):
|
65 |
"""Continues a given conversation based on chathistory"""
|
66 |
try:
|
67 |
+
if message.channel.id in thread_to_user:
|
68 |
+
if thread_to_user[message.channel.id] == message.author.id:
|
69 |
+
client = thread_to_client[message.channel.id]
|
70 |
+
job = client.submit(message.content, api_name="/chat")
|
71 |
+
await wait(job)
|
72 |
+
try:
|
73 |
+
job.result()
|
74 |
+
response = job.outputs()[-1]
|
75 |
+
await message.reply(response)
|
76 |
+
except QueueError:
|
77 |
+
await message.reply("The gradio space powering this bot is really busy! Please try again later!")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
78 |
except Exception as e:
|
79 |
print(f"continue_falcon Error: {e}")
|
80 |
await message.reply(f"Error: {e} <@811235357663297546> (continue_falcon error)")
|