Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -3,11 +3,9 @@ import spaces
|
|
3 |
import torch
|
4 |
import subprocess
|
5 |
import sys
|
6 |
-
from threading import Thread
|
7 |
-
from transformers import TextIteratorStreamer
|
8 |
|
9 |
# Install required packages
|
10 |
-
subprocess.check_call([sys.executable, "-m", "pip", "install", "-U", "--force-reinstall", "
|
11 |
subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
|
12 |
|
13 |
from transformers import OlmoeForCausalLM, AutoTokenizer
|
@@ -20,7 +18,7 @@ try:
|
|
20 |
model = OlmoeForCausalLM.from_pretrained(
|
21 |
model_name,
|
22 |
trust_remote_code=True,
|
23 |
-
torch_dtype=torch.
|
24 |
low_cpu_mem_usage=True,
|
25 |
device_map="auto",
|
26 |
_attn_implementation="flash_attention_2" # Enable Flash Attention 2
|
@@ -53,26 +51,32 @@ def generate_response(message, history, temperature, max_new_tokens):
|
|
53 |
inputs = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt").to(DEVICE)
|
54 |
|
55 |
try:
|
56 |
-
streamer = TextIteratorStreamer(tokenizer,
|
57 |
generation_kwargs = dict(
|
58 |
inputs=inputs,
|
59 |
-
streamer=streamer,
|
60 |
max_new_tokens=max_new_tokens,
|
61 |
do_sample=True,
|
62 |
temperature=temperature,
|
63 |
eos_token_id=tokenizer.eos_token_id,
|
|
|
64 |
)
|
65 |
|
66 |
-
thread =
|
67 |
thread.start()
|
68 |
|
69 |
-
|
70 |
for new_text in streamer:
|
71 |
-
|
72 |
-
yield
|
73 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
74 |
except Exception as e:
|
75 |
-
yield f"An error occurred: {str(e)}"
|
76 |
|
77 |
css = """
|
78 |
#output {
|
@@ -83,12 +87,12 @@ css = """
|
|
83 |
"""
|
84 |
|
85 |
with gr.Blocks(css=css) as demo:
|
86 |
-
gr.Markdown("# Nisten's Karpathy Chatbot with OSS OLMoE (Now with Flash Attention 2
|
87 |
chatbot = gr.Chatbot(elem_id="output")
|
88 |
msg = gr.Textbox(label="Meow")
|
89 |
with gr.Row():
|
90 |
temperature = gr.Slider(minimum=0.1, maximum=1.0, value=0.7, step=0.1, label="Temperature")
|
91 |
-
max_new_tokens = gr.Slider(minimum=50, maximum=
|
92 |
clear = gr.Button("Clear")
|
93 |
|
94 |
def user(user_message, history):
|
|
|
3 |
import torch
|
4 |
import subprocess
|
5 |
import sys
|
|
|
|
|
6 |
|
7 |
# Install required packages
|
8 |
+
subprocess.check_call([sys.executable, "-m", "pip", "install", "-U", "--force-reinstall", "einops", "accelerate", "git+https://github.com/Muennighoff/transformers.git@olmoe"])
|
9 |
subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
|
10 |
|
11 |
from transformers import OlmoeForCausalLM, AutoTokenizer
|
|
|
18 |
model = OlmoeForCausalLM.from_pretrained(
|
19 |
model_name,
|
20 |
trust_remote_code=True,
|
21 |
+
torch_dtype=torch.float16, # Using float16 for lower precision
|
22 |
low_cpu_mem_usage=True,
|
23 |
device_map="auto",
|
24 |
_attn_implementation="flash_attention_2" # Enable Flash Attention 2
|
|
|
51 |
inputs = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt").to(DEVICE)
|
52 |
|
53 |
try:
|
54 |
+
streamer = gr.TextIteratorStreamer(tokenizer, skip_special_tokens=True)
|
55 |
generation_kwargs = dict(
|
56 |
inputs=inputs,
|
|
|
57 |
max_new_tokens=max_new_tokens,
|
58 |
do_sample=True,
|
59 |
temperature=temperature,
|
60 |
eos_token_id=tokenizer.eos_token_id,
|
61 |
+
streamer=streamer
|
62 |
)
|
63 |
|
64 |
+
thread = torch.multiprocessing.Process(target=model.generate, kwargs=generation_kwargs)
|
65 |
thread.start()
|
66 |
|
67 |
+
generated_text = ""
|
68 |
for new_text in streamer:
|
69 |
+
generated_text += new_text
|
70 |
+
yield generated_text.strip()
|
71 |
+
|
72 |
+
thread.join()
|
73 |
+
except RuntimeError as e:
|
74 |
+
if "CUDA out of memory" in str(e):
|
75 |
+
yield "GPU memory exceeded. Try reducing the max tokens or using a smaller model."
|
76 |
+
else:
|
77 |
+
yield f"An error occurred: {str(e)}"
|
78 |
except Exception as e:
|
79 |
+
yield f"An unexpected error occurred: {str(e)}"
|
80 |
|
81 |
css = """
|
82 |
#output {
|
|
|
87 |
"""
|
88 |
|
89 |
with gr.Blocks(css=css) as demo:
|
90 |
+
gr.Markdown("# Nisten's Karpathy Chatbot with OSS OLMoE (Now with Flash Attention 2!)")
|
91 |
chatbot = gr.Chatbot(elem_id="output")
|
92 |
msg = gr.Textbox(label="Meow")
|
93 |
with gr.Row():
|
94 |
temperature = gr.Slider(minimum=0.1, maximum=1.0, value=0.7, step=0.1, label="Temperature")
|
95 |
+
max_new_tokens = gr.Slider(minimum=50, maximum=4000, value=2000, step=50, label="Max New Tokens")
|
96 |
clear = gr.Button("Clear")
|
97 |
|
98 |
def user(user_message, history):
|