Snapshot
Browse files- completions.py +32 -42
- main.py +5 -11
completions.py
CHANGED
@@ -6,14 +6,9 @@ from transformers.generation.utils import GenerateOutput
|
|
6 |
from typing import cast
|
7 |
from dataclasses import dataclass
|
8 |
|
9 |
-
|
10 |
|
11 |
-
|
12 |
-
class Word:
|
13 |
-
tokens: list[int]
|
14 |
-
text: str
|
15 |
-
logprob: float
|
16 |
-
context: list[int]
|
17 |
|
18 |
def starts_with_space(token: str) -> bool:
|
19 |
return token.startswith(chr(9601)) or token.startswith(chr(288))
|
@@ -107,47 +102,42 @@ def extract_replacements(outputs: GenerateOutput | torch.LongTensor, tokenizer:
|
|
107 |
|
108 |
#%%
|
109 |
|
110 |
-
|
111 |
-
|
112 |
-
# model_name = "mistralai/Mistral-7B-v0.1"
|
113 |
-
model_name = "unsloth/Llama-3.2-1B"
|
114 |
-
model, tokenizer = load_model_and_tokenizer(model_name, device)
|
115 |
|
116 |
-
|
117 |
-
|
118 |
-
|
|
|
119 |
|
|
|
120 |
#%%
|
121 |
-
|
122 |
|
123 |
-
#%%
|
124 |
-
|
125 |
-
log_prob_threshold = -5.0
|
126 |
-
low_prob_words = [word for word in words if word.logprob < log_prob_threshold]
|
127 |
|
128 |
-
#%%
|
129 |
-
|
130 |
-
|
131 |
-
|
132 |
|
133 |
-
#%%
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
end_time = time.time()
|
138 |
-
print(f"Total time taken for replacements: {end_time - start_time:.4f} seconds")
|
139 |
-
|
140 |
-
#%%
|
141 |
-
replacements_batch = extract_replacements(outputs, tokenizer, input_ids.shape[0], input_ids.shape[1], num_samples)
|
142 |
|
143 |
-
#%%
|
144 |
-
|
145 |
-
|
146 |
-
|
|
|
|
|
147 |
|
148 |
-
|
149 |
-
|
150 |
-
for g in generated_ids:
|
151 |
-
print(tokenizer.convert_ids_to_tokens(g.tolist()))
|
152 |
|
153 |
-
|
|
|
|
|
|
|
|
6 |
from typing import cast
|
7 |
from dataclasses import dataclass
|
8 |
|
9 |
+
from models import ApiWord, Word
|
10 |
|
11 |
+
type Tokenizer = PreTrainedTokenizer | PreTrainedTokenizerFast
|
|
|
|
|
|
|
|
|
|
|
12 |
|
13 |
def starts_with_space(token: str) -> bool:
|
14 |
return token.startswith(chr(9601)) or token.startswith(chr(288))
|
|
|
102 |
|
103 |
#%%
|
104 |
|
105 |
+
def load_model() -> tuple[PreTrainedModel, Tokenizer, torch.device]:
|
106 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
|
|
|
|
|
|
107 |
|
108 |
+
# model_name = "mistralai/Mistral-7B-v0.1"
|
109 |
+
model_name = "unsloth/Llama-3.2-1B"
|
110 |
+
model, tokenizer = load_model_and_tokenizer(model_name, device)
|
111 |
+
return model, tokenizer, device
|
112 |
|
113 |
+
def check_text(input_text: str, model: PreTrainedModel, tokenizer: Tokenizer, device: torch.device) -> list[ApiWord]:
|
114 |
#%%
|
115 |
+
inputs: BatchEncoding = tokenize(input_text, tokenizer, device)
|
116 |
|
117 |
+
#%%
|
118 |
+
token_probs: list[tuple[int, float]] = calculate_log_probabilities(model, tokenizer, inputs)
|
|
|
|
|
119 |
|
120 |
+
#%%
|
121 |
+
words = split_into_words(token_probs, tokenizer)
|
122 |
+
log_prob_threshold = -5.0
|
123 |
+
low_prob_words = [word for word in words if word.logprob < log_prob_threshold]
|
124 |
|
125 |
+
#%%
|
126 |
+
contexts = [word.context for word in low_prob_words]
|
127 |
+
inputs = prepare_inputs(contexts, tokenizer, device)
|
128 |
+
input_ids = inputs["input_ids"]
|
|
|
|
|
|
|
|
|
|
|
129 |
|
130 |
+
#%%
|
131 |
+
num_samples = 5
|
132 |
+
start_time = time.time()
|
133 |
+
outputs = generate_outputs(model, inputs, num_samples)
|
134 |
+
end_time = time.time()
|
135 |
+
print(f"Total time taken for replacements: {end_time - start_time:.4f} seconds")
|
136 |
|
137 |
+
#%%
|
138 |
+
replacements = extract_replacements(outputs, tokenizer, input_ids.shape[0], input_ids.shape[1], num_samples)
|
|
|
|
|
139 |
|
140 |
+
#%%
|
141 |
+
for word, replacements in zip(low_prob_words, replacements):
|
142 |
+
print(f"Original word: {word.text}, Log Probability: {word.logprob:.4f}")
|
143 |
+
print(f"Proposed replacements: {replacements}")
|
main.py
CHANGED
@@ -2,22 +2,16 @@ from fastapi import FastAPI
|
|
2 |
from fastapi.staticfiles import StaticFiles
|
3 |
from pydantic import BaseModel
|
4 |
|
5 |
-
|
6 |
-
|
7 |
-
start: int
|
8 |
-
end: int
|
9 |
-
logprob: float
|
10 |
-
suggestions: list[str]
|
11 |
-
|
12 |
-
class CheckResponse(BaseModel):
|
13 |
-
text: str
|
14 |
-
words: list[Word]
|
15 |
|
16 |
app = FastAPI()
|
17 |
|
|
|
|
|
18 |
@app.get("/check", response_model=CheckResponse)
|
19 |
def check(text: str):
|
20 |
-
return CheckResponse(text=text, words=
|
21 |
|
22 |
# serve files from frontend/public
|
23 |
app.mount("/", StaticFiles(directory="frontend/public", html=True))
|
|
|
2 |
from fastapi.staticfiles import StaticFiles
|
3 |
from pydantic import BaseModel
|
4 |
|
5 |
+
from models import CheckResponse, ApiWord
|
6 |
+
from completions import check_text, load_model
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
|
8 |
app = FastAPI()
|
9 |
|
10 |
+
model, tokenizer, device = load_model()
|
11 |
+
|
12 |
@app.get("/check", response_model=CheckResponse)
|
13 |
def check(text: str):
|
14 |
+
return CheckResponse(text=text, words=check_text(text, model, tokenizer, device))
|
15 |
|
16 |
# serve files from frontend/public
|
17 |
app.mount("/", StaticFiles(directory="frontend/public", html=True))
|