tokenspace / calculate-distancesXL.py
ppbrown's picture
Upload 2 files
2edceda verified
#!/bin/env python
"""
Plan:
Read in "dictionary" for list of words
Read in pre-calculated "proper" embedding for each word from safetensor file
Prompt user for a word from the list
Generate a tensor array of distance to all the other known words
Print out the 20 closest ones
"""
import sys
import torch
from safetensors import safe_open
from transformers import CLIPProcessor,CLIPModel, CLIPTextModelWithProjection
processor=None
tmodel2=None
model_path2=None
model_config2=None
if len(sys.argv) == 4:
model_path2=sys.argv[1]
model_config2=sys.argv[2]
embed_file=sys.argv[3]
else:
print("You have to give name of textencoder modelfile,config file, and embeddings file")
sys.exit(1)
device=torch.device("cuda")
def init():
global tmodel2,processor
# yes, oddly they all use the same tokenizer, basically
processor = CLIPProcessor.from_pretrained("openai/clip-vit-large-patch14")
print("loading",model_path)
tmodel2 = CLIPTextModelWithProjection.from_pretrained(model_path,config=model_config,local_files_only=True,use_safetensors=True)
tmodel2.to(device)
print("read in words from dictionary now",file=sys.stderr)
with open("dictionary","r") as f:
tokendict = f.readlines()
wordlist = [token.strip() for token in tokendict] # Remove trailing newlines
print(len(wordlist),"lines read")
print("read in embeddings now",file=sys.stderr)
model = safe_open(embed_file,framework="pt",device="cuda")
embs=model.get_tensor("embeddings")
embs.to(device)
print("Shape of loaded embeds =",embs.shape)
def standard_embed_calc(text):
global processor,tmodel2
inputs = processor(text=text, return_tensors="pt")
inputs.to(device)
with torch.no_grad():
outputs = tmodel2(**inputs)
embeddings = outputs.text_embeds
return embeddings[0]
def print_distances(targetemb):
targetdistances = torch.cdist( targetemb.unsqueeze(0), embs, p=2)
print("shape of distances...",targetdistances.shape)
smallest_distances, smallest_indices = torch.topk(targetdistances[0], 20, largest=False)
smallest_distances=smallest_distances.tolist()
smallest_indices=smallest_indices.tolist()
for d,i in zip(smallest_distances,smallest_indices):
print(wordlist[i],"(",d,")")
# Find 10 closest tokens to targetword.
# Will include the word itself
def find_closest(targetword):
try:
targetindex=wordlist.index(targetword)
targetemb=embs[targetindex]
print_distances(targetemb)
return
except ValueError:
print(targetword,"not found in cache")
print("Now doing with full calc embed")
targetemb=standard_embed_calc(targetword)
print_distances(targetemb)
while True:
input_text=input("Input a word now:")
find_closest(input_text)