-
-
Save avidale/c6b19687d333655da483421880441950 to your computer and use it in GitHub Desktop.
do you have a two file ? .txt and .csv ? then may be the length of list of sentences is not the same as your csv file.
i thought that your sentences are from the same CSV file so the length of date and para are the same
but now I'm sorry I can't help you
you need information how to link that sentences and dates ( for each sentence the appropriate date ) then you can solve the pb
It is the same CSV file. I first convert, CSV to list and text as in the original example.
df = pd.read_csv("/content/df3.csv",parse_dates= True)
df = df.set_index("content")
df.head(1)
text_dict = df.to_dict()
len_text = len(text_dict["date"])
df = df["date"].to_dict()
df_sentences_list = list(df.keys())
len(df_sentences_list)
df_sentences_list = [str(d) for d in tqdm(df_sentences_list)]
file_content = "\n".join(df_sentences_list)
with open("input_text.txt","w") as f:
f.write(file_content)
with open("/content/input_text.txt","r") as f:
lines1 = f.readlines()
lines1[0]
all_sentences = [l.split('\t')[0] for l in lines1]
Again for the date, We use the same csv file, I only distable this part df = df.set_index("content")
If I can have your email Id I will send refine codes with small sample data
Thanks a lot!!!
@avidale Hi thanks for the amazing work.. I want to implement that, but need few points to understand..
- I want it to work for large dataset, can it handle it or need to implement something else on it?
- After getting embeddings, can I save it so next time I just load the data and get result by query?
@sridhardev07 yes and yes
@sridhardev07 yes and yes
Can you tell me how? that will be really helpful for me!!
Hi @avidale I tried this with some bigger dataset to test the accuracy. Dataset having sentences about 37126, it is showing me memory error: numpy.core._exceptions.MemoryError: Unable to allocate 2.35 GiB for an array with shape (819827, 768) and data type float32
I am having 16GB of RAM, can you tell any alternate way to do, which uses less RAM or retrieve the data from the disk??
Hi @sridhardev07!
The simplest trick I could suggest is to convert all vectors from float32 to float16, this will reduce memory requirements by half without significantly affecting the quality.
If this does not suffice, you could look at https://github.com/facebookresearch/faiss - a library for fast vector similarity search that allegedly can work with very large sets. Specifically, they implement product quantization for lossy compression of the vectors. If you choose to use Faiss, you should rewrite my solution: unite process_sentences and build_search_index that processes the sentences incrementally and adds their vectors to a faiss.IndexIVFPQ instead of a KDTree.
Hi @avidale ! Thanks for the answer!
I tried converting the vectors to float16 it does help to reduce the size but not that much as I am working with large dataset.
I tried the second approach of Faiss, it worked good when I tried Flat index, so I can add the index incrementally. But on saving that to disk taking lots of storage. Approx 1 GB of 15K sentences. here is what I did:
def __init__(self, sentences, model):
self.sentences = sentences
self.model = model
self.index = faiss.IndexFlatL2(768)
def process_sentences(self):
result = self.model(self.sentences)
self.sentence_ids = []
self.token_ids = []
self.all_tokens = []
for i, (toks, embs) in enumerate(tqdm(result)):
# initialize all_embeddings for every new sentence
all_embeddings = []
for j, (tok, emb) in enumerate(zip(toks, embs)):
self.sentence_ids.append(i)
self.token_ids.append(j)
self.all_tokens.append(tok)
all_embeddings.append(emb)
all_embeddings = np.stack(all_embeddings) # Add embeddings after every sentence
self.index.add(all_embeddings)
faiss.write_index(self.index, "faiss_Model")
Then I tried with faiss.IndexIVFPQ, it works good, but did not works for incremental index as it needs the training data too. So need to calculate all the embeddings and then train and add. Again the size is small but its taking too much RAM that is causing issue while working with large data. here is what I did:
def __init__(self, sentences, model):
self.sentences = sentences
self.model = model
self.quantizer = faiss.IndexFlatL2(768)
self.index = faiss.IndexIVFPQ(self.quantizer, 768, 1000, 16, 8)
def process_sentences(self):
result = self.model(self.sentences)
self.sentence_ids = []
self.token_ids = []
self.all_tokens = []
all_embeddings = []
for i, (toks, embs) in enumerate(tqdm(result)):
for j, (tok, emb) in enumerate(zip(toks, embs)):
self.sentence_ids.append(i)
self.token_ids.append(j)
self.all_tokens.append(tok)
all_embeddings.append(emb)
all_embeddings = np.stack(all_embeddings)
self.index.train(all_embeddings) # Train
self.index.add(all_embeddings) # Add to index
faiss.write_index(self.index, "faiss_Model_mini")
the length of the liste date is 8804 and you have an index of 10759 that's why you have in out of index .
i need to think