Update app.py
Browse files
app.py
CHANGED
@@ -96,9 +96,9 @@ from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
|
|
96 |
#####
|
97 |
from huggingface_hub import InferenceClient
|
98 |
|
99 |
-
repo_id = "meta-llama/Llama-3.2-1B-Instruct"
|
100 |
|
101 |
-
llm = InferenceClient(model=repo_id, timeout=120)
|
102 |
|
103 |
# Test your LLM client
|
104 |
#llm_client.text_generation(prompt="How are you today?", max_new_tokens=20)
|
@@ -108,7 +108,7 @@ def download_nltk_resources():
|
|
108 |
resources = ['punkt', 'stopwords', 'snowball_data']
|
109 |
for resource in resources:
|
110 |
try:
|
111 |
-
nltk.download(resource, quiet=
|
112 |
except Exception as e:
|
113 |
print(f"Failed to download {resource}: {str(e)}")
|
114 |
|
@@ -331,21 +331,25 @@ import nltk
|
|
331 |
|
332 |
def optimize_query(
|
333 |
query: str,
|
|
|
334 |
chunks: List[str],
|
335 |
embedding_model: str,
|
|
|
|
|
336 |
top_k: int = 3,
|
337 |
-
|
338 |
-
use_gpu: bool = False # Default to CPU
|
339 |
) -> str:
|
340 |
"""
|
341 |
CPU-optimized version of query expansion using a small language model.
|
342 |
|
343 |
Args:
|
344 |
query: Original search query
|
|
|
345 |
chunks: List of text chunks to search through
|
346 |
embedding_model: Name of the embedding model being used
|
|
|
|
|
347 |
top_k: Number of expansion terms to add
|
348 |
-
model_name: Name of the small language model to use
|
349 |
use_gpu: Whether to use GPU if available (defaults to False for CPU)
|
350 |
|
351 |
Returns:
|
@@ -367,42 +371,42 @@ def optimize_query(
|
|
367 |
# Limit number of lemmas
|
368 |
expanded_terms.update([lemma.name() for lemma in syn.lemmas()[:2]])
|
369 |
|
370 |
-
# 3. Use
|
371 |
try:
|
372 |
# Load model with reduced memory footprint
|
373 |
tokenizer = AutoTokenizer.from_pretrained(
|
374 |
-
|
375 |
-
model_max_length=128,
|
376 |
-
cache_dir="./model_cache"
|
377 |
)
|
378 |
-
model =
|
379 |
-
|
380 |
-
low_cpu_mem_usage=True,
|
381 |
-
device_map="cpu"
|
382 |
)
|
383 |
|
384 |
# Move model to CPU and eval mode
|
385 |
model = model.to(device)
|
386 |
-
model.eval()
|
387 |
|
388 |
# Prepare input with reduced length
|
389 |
prompt = f"Enhance this search query with relevant terms: {query}"
|
390 |
inputs = tokenizer(
|
391 |
prompt,
|
392 |
return_tensors="pt",
|
393 |
-
max_length=64,
|
394 |
truncation=True,
|
395 |
padding=True
|
396 |
)
|
397 |
|
398 |
# Generate with minimal parameters
|
399 |
-
with torch.no_grad():
|
400 |
outputs = model.generate(
|
401 |
inputs.input_ids.to(device),
|
402 |
-
max_length=32,
|
403 |
num_return_sequences=1,
|
404 |
temperature=0.7,
|
405 |
-
do_sample=False,
|
406 |
early_stopping=True
|
407 |
)
|
408 |
|
@@ -414,12 +418,12 @@ def optimize_query(
|
|
414 |
|
415 |
except Exception as model_error:
|
416 |
print(f"Model-based expansion failed: {str(model_error)}")
|
417 |
-
enhanced_query = query
|
418 |
|
419 |
# 4. Combine original and expanded terms
|
420 |
final_terms = set(tokens)
|
421 |
final_terms.update(expanded_terms)
|
422 |
-
if enhanced_query != query:
|
423 |
final_terms.update(word_tokenize(enhanced_query.lower()))
|
424 |
|
425 |
# 5. Remove stopwords and select top_k most relevant terms
|
@@ -434,13 +438,15 @@ def optimize_query(
|
|
434 |
del tokenizer
|
435 |
if device == "cuda":
|
436 |
torch.cuda.empty_cache()
|
437 |
-
|
438 |
return expanded_query.strip()
|
439 |
|
440 |
except Exception as e:
|
441 |
print(f"Query optimization failed: {str(e)}")
|
442 |
return query # Return original query if optimization fails
|
443 |
|
|
|
|
|
444 |
# Example usage
|
445 |
"""
|
446 |
chunks = ["sample text chunk 1", "sample text chunk 2"]
|
@@ -843,6 +849,7 @@ def compare_embeddings(file, query, embedding_models, custom_embedding_model, sp
|
|
843 |
"apply_phonetic": apply_phonetic,
|
844 |
"phonetic_weight": phonetic_weight,
|
845 |
"use_query_optimization": use_query_optimization,
|
|
|
846 |
"use_reranking": use_reranking
|
847 |
}
|
848 |
|
@@ -1337,7 +1344,7 @@ def launch_interface(share=True):
|
|
1337 |
'apply_phonetic': [False], # Default phonetic settings
|
1338 |
'phonetic_weight': [0.5],
|
1339 |
'custom_separators': [None],
|
1340 |
-
'query_optimization_model': ['
|
1341 |
}
|
1342 |
|
1343 |
# Run automated tests
|
|
|
96 |
#####
|
97 |
from huggingface_hub import InferenceClient
|
98 |
|
99 |
+
#repo_id = "meta-llama/Llama-3.2-1B-Instruct"
|
100 |
|
101 |
+
#llm = InferenceClient(model=repo_id, timeout=120)
|
102 |
|
103 |
# Test your LLM client
|
104 |
#llm_client.text_generation(prompt="How are you today?", max_new_tokens=20)
|
|
|
108 |
resources = ['punkt', 'stopwords', 'snowball_data']
|
109 |
for resource in resources:
|
110 |
try:
|
111 |
+
nltk.download(resource, quiet=False)
|
112 |
except Exception as e:
|
113 |
print(f"Failed to download {resource}: {str(e)}")
|
114 |
|
|
|
331 |
|
332 |
def optimize_query(
|
333 |
query: str,
|
334 |
+
query_optimization_model: str, # Added to match your signature = "google/flan-t5-small"
|
335 |
chunks: List[str],
|
336 |
embedding_model: str,
|
337 |
+
vector_store_type: str, # Added to match your signature
|
338 |
+
search_type: str, # Added to match your signature
|
339 |
top_k: int = 3,
|
340 |
+
use_gpu: bool = False
|
|
|
341 |
) -> str:
|
342 |
"""
|
343 |
CPU-optimized version of query expansion using a small language model.
|
344 |
|
345 |
Args:
|
346 |
query: Original search query
|
347 |
+
query_optimization_model: Name or path of the model to use for optimization
|
348 |
chunks: List of text chunks to search through
|
349 |
embedding_model: Name of the embedding model being used
|
350 |
+
vector_store_type: Type of vector store being used
|
351 |
+
search_type: Type of search being performed
|
352 |
top_k: Number of expansion terms to add
|
|
|
353 |
use_gpu: Whether to use GPU if available (defaults to False for CPU)
|
354 |
|
355 |
Returns:
|
|
|
371 |
# Limit number of lemmas
|
372 |
expanded_terms.update([lemma.name() for lemma in syn.lemmas()[:2]])
|
373 |
|
374 |
+
# 3. Use provided model with reduced complexity
|
375 |
try:
|
376 |
# Load model with reduced memory footprint
|
377 |
tokenizer = AutoTokenizer.from_pretrained(
|
378 |
+
query_optimization_model, # Use the provided model name
|
379 |
+
model_max_length=128,
|
380 |
+
cache_dir="./model_cache"
|
381 |
)
|
382 |
+
model = AutoModelForSeq2Gen.from_pretrained(
|
383 |
+
query_optimization_model, # Use the provided model name
|
384 |
+
low_cpu_mem_usage=True,
|
385 |
+
device_map="cpu"
|
386 |
)
|
387 |
|
388 |
# Move model to CPU and eval mode
|
389 |
model = model.to(device)
|
390 |
+
model.eval()
|
391 |
|
392 |
# Prepare input with reduced length
|
393 |
prompt = f"Enhance this search query with relevant terms: {query}"
|
394 |
inputs = tokenizer(
|
395 |
prompt,
|
396 |
return_tensors="pt",
|
397 |
+
max_length=64,
|
398 |
truncation=True,
|
399 |
padding=True
|
400 |
)
|
401 |
|
402 |
# Generate with minimal parameters
|
403 |
+
with torch.no_grad():
|
404 |
outputs = model.generate(
|
405 |
inputs.input_ids.to(device),
|
406 |
+
max_length=32,
|
407 |
num_return_sequences=1,
|
408 |
temperature=0.7,
|
409 |
+
do_sample=False,
|
410 |
early_stopping=True
|
411 |
)
|
412 |
|
|
|
418 |
|
419 |
except Exception as model_error:
|
420 |
print(f"Model-based expansion failed: {str(model_error)}")
|
421 |
+
enhanced_query = query
|
422 |
|
423 |
# 4. Combine original and expanded terms
|
424 |
final_terms = set(tokens)
|
425 |
final_terms.update(expanded_terms)
|
426 |
+
if enhanced_query != query:
|
427 |
final_terms.update(word_tokenize(enhanced_query.lower()))
|
428 |
|
429 |
# 5. Remove stopwords and select top_k most relevant terms
|
|
|
438 |
del tokenizer
|
439 |
if device == "cuda":
|
440 |
torch.cuda.empty_cache()
|
441 |
+
|
442 |
return expanded_query.strip()
|
443 |
|
444 |
except Exception as e:
|
445 |
print(f"Query optimization failed: {str(e)}")
|
446 |
return query # Return original query if optimization fails
|
447 |
|
448 |
+
|
449 |
+
|
450 |
# Example usage
|
451 |
"""
|
452 |
chunks = ["sample text chunk 1", "sample text chunk 2"]
|
|
|
849 |
"apply_phonetic": apply_phonetic,
|
850 |
"phonetic_weight": phonetic_weight,
|
851 |
"use_query_optimization": use_query_optimization,
|
852 |
+
"query_optimization_model": query_optimization_model
|
853 |
"use_reranking": use_reranking
|
854 |
}
|
855 |
|
|
|
1344 |
'apply_phonetic': [False], # Default phonetic settings
|
1345 |
'phonetic_weight': [0.5],
|
1346 |
'custom_separators': [None],
|
1347 |
+
'query_optimization_model': ['google/flan-t5-base'] # Default query optimization model
|
1348 |
}
|
1349 |
|
1350 |
# Run automated tests
|