Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,38 +1,157 @@
|
|
| 1 |
import spaces
|
|
|
|
| 2 |
import gradio as gr
|
| 3 |
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
|
| 4 |
import torch
|
| 5 |
|
| 6 |
title = """# 🙋🏻♂️Welcome to🌟Tonic's🔮DeepSeekMath📉
|
| 7 |
-
You can build with this endpoint using🔮DeepSeekMath
|
| 8 |
You can also use 🔮DeepSeekMath📉 by cloning this space. Simply click here: <a style="display:inline-block" href="https://huggingface.co/spaces/Tonic/Math?duplicate=true"><img src="https://img.shields.io/badge/-Duplicate%20Space-blue?labelColor=white&style=flat&logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAAAXNSR0IArs4c6QAAAP5JREFUOE+lk7FqAkEURY+ltunEgFXS2sZGIbXfEPdLlnxJyDdYB62sbbUKpLbVNhyYFzbrrA74YJlh9r079973psed0cvUD4A+4HoCjsA85X0Dfn/RBLBgBDxnQPfAEJgBY+A9gALA4tcbamSzS4xq4FOQAJgCDwV2CPKV8tZAJcAjMMkUe1vX+U+SMhfAJEHasQIWmXNN3abzDwHUrgcRGmYcgKe0bxrblHEB4E/pndMazNpSZGcsZdBlYJcEL9Afo75molJyM2FxmPgmgPqlWNLGfwZGG6UiyEvLzHYDmoPkDDiNm9JR9uboiONcBXrpY1qmgs21x1QwyZcpvxt9NS09PlsPAAAAAElFTkSuQmCC&logoWidth=14" alt="Duplicate Space"></a></h3>
|
| 9 |
Join us : 🌟TeamTonic🌟 is always making cool demos! Join our active builder's 🛠️community 👻 [](https://discord.gg/GWpVpekp) On 🤗Huggingface: [TeamTonic](https://huggingface.co/TeamTonic) & [MultiTransformer](https://huggingface.co/MultiTransformer) Math with [introspector](https://huggingface.co/introspector) On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to🌟 [SciTonic](https://github.com/Tonic-AI/scitonic)🤗Big thanks to Yuvi Sharma and all the folks at huggingface for the community grant 🤗
|
| 10 |
"""
|
| 11 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
model_name = "deepseek-ai/deepseek-math-7b-instruct"
|
| 13 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
| 14 |
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
|
| 15 |
model.generation_config = GenerationConfig.from_pretrained(model_name)
|
| 16 |
model.generation_config.pad_token_id = model.generation_config.eos_token_id
|
| 17 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
@spaces.GPU
|
| 19 |
-
def solve_math_problem(question, max_tokens):
|
| 20 |
-
prompt = f"User: {question}\
|
| 21 |
input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to(model.device)
|
| 22 |
outputs = model.generate(input_ids, max_length=max_tokens + input_ids.shape[1], pad_token_id=model.generation_config.pad_token_id)
|
| 23 |
result = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
| 24 |
-
|
|
|
|
|
|
|
| 25 |
|
| 26 |
def main():
|
| 27 |
with gr.Blocks() as demo:
|
| 28 |
gr.Markdown(title)
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
submit_button = gr.Button("Solve")
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
|
| 37 |
demo.launch()
|
| 38 |
|
|
|
|
| 1 |
import spaces
|
| 2 |
+
import re
|
| 3 |
import gradio as gr
|
| 4 |
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
|
| 5 |
import torch
|
| 6 |
|
| 7 |
title = """# 🙋🏻♂️Welcome to🌟Tonic's🔮DeepSeekMath📉
|
| 8 |
+
You can build with this endpoint using🔮DeepSeekMath📉 available here : [deepseek-ai/deepseek-math-7b-instruct](https://huggingface.co/deepseek-ai/deepseek-math-7b-instruct). We're using 🤖[introspector/unimath](https://huggingface.co/datasets/introspector/unimath) for cool examples, check it out below ! The demo is still a work in progress and we're looking forward to build downstream tasks that showcase outstanding mathematical reasoning. Have any ideas ? join us below !
|
| 9 |
You can also use 🔮DeepSeekMath📉 by cloning this space. Simply click here: <a style="display:inline-block" href="https://huggingface.co/spaces/Tonic/Math?duplicate=true"><img src="https://img.shields.io/badge/-Duplicate%20Space-blue?labelColor=white&style=flat&logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAAAXNSR0IArs4c6QAAAP5JREFUOE+lk7FqAkEURY+ltunEgFXS2sZGIbXfEPdLlnxJyDdYB62sbbUKpLbVNhyYFzbrrA74YJlh9r079973psed0cvUD4A+4HoCjsA85X0Dfn/RBLBgBDxnQPfAEJgBY+A9gALA4tcbamSzS4xq4FOQAJgCDwV2CPKV8tZAJcAjMMkUe1vX+U+SMhfAJEHasQIWmXNN3abzDwHUrgcRGmYcgKe0bxrblHEB4E/pndMazNpSZGcsZdBlYJcEL9Afo75molJyM2FxmPgmgPqlWNLGfwZGG6UiyEvLzHYDmoPkDDiNm9JR9uboiONcBXrpY1qmgs21x1QwyZcpvxt9NS09PlsPAAAAAElFTkSuQmCC&logoWidth=14" alt="Duplicate Space"></a></h3>
|
| 10 |
Join us : 🌟TeamTonic🌟 is always making cool demos! Join our active builder's 🛠️community 👻 [](https://discord.gg/GWpVpekp) On 🤗Huggingface: [TeamTonic](https://huggingface.co/TeamTonic) & [MultiTransformer](https://huggingface.co/MultiTransformer) Math with [introspector](https://huggingface.co/introspector) On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to🌟 [SciTonic](https://github.com/Tonic-AI/scitonic)🤗Big thanks to Yuvi Sharma and all the folks at huggingface for the community grant 🤗
|
| 11 |
"""
|
| 12 |
|
| 13 |
+
unimath1 = """Goal:
|
| 14 |
+
|
| 15 |
+
X : UU
|
| 16 |
+
Y : UU
|
| 17 |
+
P : UU
|
| 18 |
+
xp : (X → P) → P
|
| 19 |
+
yp : (Y → P) → P
|
| 20 |
+
X0 : X × Y → P
|
| 21 |
+
x : X
|
| 22 |
+
============================
|
| 23 |
+
(Y → P)
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
DEBUG:Going to execute:
|
| 27 |
+
PTRDEBUGTAC<coq-core.plugins.ltac::intro@1> $1
|
| 28 |
+
DEBUG LTAC Evaluated term: yp
|
| 29 |
+
|
| 30 |
+
TcDebug (0) > NONE??? LtacNotationCall (no location),/mnt/data1/2024/01/05/UniMath/UniMath/Foundations/PartA.v:234
|
| 31 |
+
"""
|
| 32 |
+
# source : unimath/unimath/batch2/data08
|
| 33 |
+
|
| 34 |
+
unimath2 = """Goal:
|
| 35 |
+
R : ring M : module R
|
| 36 |
+
============================
|
| 37 |
+
(islinear (idfun M))
|
| 38 |
+
|
| 39 |
+
|
| 40 |
+
DEBUG:Going to execute:
|
| 41 |
+
PTRDEBUGTACapply pathsinv0; trivial
|
| 42 |
+
Level 0: Backtrace:
|
| 43 |
+
|
| 44 |
+
Proof is not complete.
|
| 45 |
+
Level 0: Backtrace:
|
| 46 |
+
|
| 47 |
+
Proof is not complete.
|
| 48 |
+
|
| 49 |
+
TcDebug (0) > NONE??? LtacNotationCall (no location),/mnt/data1/2024/01/05/UniMath/UniMath/MoreFoundations/Tactics.veasy:19, Tactics (UniMath.MoreFoundations),/mnt/data1/2024/01/05/UniMath/UniMath/Algebra/Modules/Examples.v:27
|
| 50 |
+
|
| 51 |
+
"""
|
| 52 |
+
# source : unimath/unimath/batch2/data_22/BATCH122007
|
| 53 |
+
|
| 54 |
+
unimath3 = """Goal:
|
| 55 |
+
X : UU i : nat b : hProptoType (i < S i) x : Vector X (S i) r : i = i
|
| 56 |
+
============================
|
| 57 |
+
(pr1 lastelement = pr1 (i,, b))
|
| 58 |
+
|
| 59 |
+
|
| 60 |
+
DEBUG:Going to execute:
|
| 61 |
+
PTRDEBUGTACsimpl
|
| 62 |
+
DEBUG LTAC Evaluated term: isinjstntonat
|
| 63 |
+
|
| 64 |
+
TcDebug (0) > /mnt/data1/2024/01/05/UniMath/UniMath/Combinatorics/FiniteSequences.v:114
|
| 65 |
+
"""
|
| 66 |
+
# source : unimath/unimath/batch2/data_12/BATCH112026
|
| 67 |
+
|
| 68 |
+
unimath4 = """Goal:
|
| 69 |
+
X : dcpo CX : continuous_dcpo_struct X x : pr1hSet X y : pr1hSet X
|
| 70 |
+
============================
|
| 71 |
+
(x ⊑ y
|
| 72 |
+
≃ (∀ i : approximating_family CX x, approximating_family CX x i ⊑ y))
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
DEBUG:Going to execute:
|
| 76 |
+
PTRDEBUGTACsimple refine (p _ _ _) ||
|
| 77 |
+
simple refine (p _ _ _ _) ||
|
| 78 |
+
simple refine (p _ _ _ _ _) ||
|
| 79 |
+
simple refine (p _ _ _ _ _ _) ||
|
| 80 |
+
simple refine (p _ _ _ _ _ _ _) ||
|
| 81 |
+
simple refine (p _ _ _ _ _ _ _ _) ||
|
| 82 |
+
simple refine (p _ _ _ _ _ _ _ _ _) ||
|
| 83 |
+
simple refine (p _ _ _ _ _ _ _ _ _ _) ||
|
| 84 |
+
simple refine (p _ _ _ _ _ _ _ _ _ _ _) ||
|
| 85 |
+
simple refine (p _ _ _ _ _ _ _ _ _ _ _ _) ||
|
| 86 |
+
simple refine (p _ _ _ _ _ _ _ _ _ _ _ _ _) ||
|
| 87 |
+
simple refine
|
| 88 |
+
(p _ _ _ _ _ _ _ _ _ _ _ _ _ _) || simple
|
| 89 |
+
refine (p _ _ _ _ _ _ _ _ _ _ _ _ _ _ _)
|
| 90 |
+
Level 0: Backtrace:
|
| 91 |
+
|
| 92 |
+
In environment
|
| 93 |
+
X : dcpo
|
| 94 |
+
CX : continuous_dcpo_struct X
|
| 95 |
+
x, y : X
|
| 96 |
+
The term "weqimplimpl ?f ?g" has type "isaprop ?X → isaprop ?Y → ?X ≃ ?Y"
|
| 97 |
+
while it is expected to have type
|
| 98 |
+
"x ⊑ y ≃ (∀ i : approximating_family CX x, approximating_family CX x i ⊑ y)".
|
| 99 |
+
Level 0: Backtrace:
|
| 100 |
+
|
| 101 |
+
In environment
|
| 102 |
+
X : dcpo
|
| 103 |
+
CX : continuous_dcpo_struct X
|
| 104 |
+
x, y : X
|
| 105 |
+
The term "weqimplimpl ?f ?g" has type "isaprop ?X → isaprop ?Y → ?X ≃ ?Y"
|
| 106 |
+
while it is expected to have type
|
| 107 |
+
"x ⊑ y ≃ (∀ i : approximating_family CX x, approximating_family CX x i ⊑ y)".
|
| 108 |
+
|
| 109 |
+
TcDebug (0) > NONE??? LtacNotationCall (no location),/mnt/data1/2024/01/05/UniMath/UniMath/Foundations/Init.vsimple_rapply:174, Init (UniMath.Foundations),/mnt/data1/2024/01/05/UniMath/UniMath/Foundations/Init.v??? LtacNotationCall:189, Init (UniMath.Foundations),/mnt/data1/2024/01/05/UniMath/UniMath/OrderTheory/DCPOs/Basis/Continuous.v:166
|
| 110 |
+
"""
|
| 111 |
+
# source : unimath/unimath/batch2/data_42/BATCH142042
|
| 112 |
+
|
| 113 |
+
unimath_examples = [unimath1, unimath2, unimath3, unimath4]
|
| 114 |
+
|
| 115 |
model_name = "deepseek-ai/deepseek-math-7b-instruct"
|
| 116 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
| 117 |
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
|
| 118 |
model.generation_config = GenerationConfig.from_pretrained(model_name)
|
| 119 |
model.generation_config.pad_token_id = model.generation_config.eos_token_id
|
| 120 |
|
| 121 |
+
def parse_full_answer(answer):
|
| 122 |
+
"""Parses the assistant's answer, excluding any text before 'Assistant :'."""
|
| 123 |
+
match = re.search(r"Assistant\s*:\s*(.*)", answer, re.DOTALL)
|
| 124 |
+
return match.group(1).strip() if match else "No assistant answer found."
|
| 125 |
+
|
| 126 |
+
def parse_final_answer(answer):
|
| 127 |
+
"""Extracts the final answer enclosed within \boxed{}."""
|
| 128 |
+
match = re.search(r"\\boxed\{([^}]+)\}", answer)
|
| 129 |
+
return match.group(1) if match else "No final answer found."
|
| 130 |
+
|
| 131 |
@spaces.GPU
|
| 132 |
+
def solve_math_problem(question, additional_info, max_tokens):
|
| 133 |
+
prompt = f"User: {question}\n{additional_info}.\nAssistant:"
|
| 134 |
input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to(model.device)
|
| 135 |
outputs = model.generate(input_ids, max_length=max_tokens + input_ids.shape[1], pad_token_id=model.generation_config.pad_token_id)
|
| 136 |
result = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
| 137 |
+
full_answer = parse_full_answer(result)
|
| 138 |
+
final_answer = parse_final_answer(result)
|
| 139 |
+
return full_answer, final_answer
|
| 140 |
|
| 141 |
def main():
|
| 142 |
with gr.Blocks() as demo:
|
| 143 |
gr.Markdown(title)
|
| 144 |
+
final_answer_output = gr.Textbox(label="🔮DeepSeekMath📉")
|
| 145 |
+
full_answer_output = gr.Code(label="🔮TonicsMathAssistant📉", interactive=False)
|
| 146 |
+
max_tokens = gr.Slider(minimum=150, maximum=1200, value=250, label="Max Tokens")
|
| 147 |
+
submit_button = gr.Button("📉Solve")
|
| 148 |
+
question = gr.Code(language='python', value='what is the integral of x^2 from 0 to 2?', label="🤔Enter your math problem")
|
| 149 |
+
additional_info = gr.Text(value='Please reason step by step, and put your final answer within \\boxed{{}}', label="🪜Optional train-of-thought")
|
| 150 |
+
with gr.Accordion("🤖metaintrospector/UniMath Examples"):
|
| 151 |
+
for example in unimath_examples:
|
| 152 |
+
gr.Textbox(value=example, label="Example", readonly=True, lines=10)
|
| 153 |
+
|
| 154 |
+
submit_button.click(fn=solve_math_problem, inputs=[question, additional_info, max_tokens], outputs=[full_answer_output, final_answer_output])
|
| 155 |
|
| 156 |
demo.launch()
|
| 157 |
|