previous commit broke flax model, fixed now

This commit is contained in:
Brett Kuprel 2022-06-28 12:54:58 -04:00
parent 5aa6fe49bf
commit 9d6b6dcc92
4 changed files with 16 additions and 17 deletions

View File

@ -3,6 +3,7 @@ import json
import numpy
from PIL import Image
from typing import Tuple, List
import torch
from min_dalle.load_params import load_dalle_bart_flax_params
from min_dalle.text_tokenizer import TextTokenizer
@ -53,25 +54,23 @@ def generate_image_from_text(
text_tokens = tokenize_text(text, config, vocab, merges)
params_dalle_bart = load_dalle_bart_flax_params(model_path)
image_tokens = numpy.zeros(config['image_length'])
if is_torch:
image_tokens[:image_token_count] = generate_image_tokens_torch(
image_tokens = generate_image_tokens_torch(
text_tokens = text_tokens,
seed = seed,
config = config,
params = params_dalle_bart,
image_token_count = image_token_count
)
if image_token_count == config['image_length']:
image = detokenize_torch(image_tokens)
return Image.fromarray(image)
else:
image_tokens[...] = generate_image_tokens_flax(
image_tokens = generate_image_tokens_flax(
text_tokens = text_tokens,
seed = seed,
config = config,
params = params_dalle_bart,
)
if image_token_count == config['image_length']:
image = detokenize_torch(image_tokens)
return Image.fromarray(image)
else:
return None
image = detokenize_torch(torch.tensor(image_tokens))
return Image.fromarray(image)

View File

@ -30,7 +30,7 @@ def load_vqgan_torch_params(path: str) -> Dict[str, torch.Tensor]:
for i in P:
P[i] = torch.tensor(P[i])
if torch.cuda.is_available(): P[i] = P[i].cuda()
# if torch.cuda.is_available(): P[i] = P[i].cuda()
P['embedding.weight'] = P.pop('quantize.embedding.embedding')
@ -87,7 +87,7 @@ def convert_dalle_bart_torch_from_flax_params(
for i in P:
P[i] = torch.tensor(P[i])
if torch.cuda.is_available(): P[i] = P[i].cuda()
# if torch.cuda.is_available(): P[i] = P[i].cuda()
for i in list(P):
if 'kernel' in i:

View File

@ -85,7 +85,7 @@ def generate_image_tokens_torch(
image_token_count: int
) -> LongTensor:
text_tokens = torch.tensor(text_tokens).to(torch.long)
if torch.cuda.is_available(): text_tokens = text_tokens.cuda()
# if torch.cuda.is_available(): text_tokens = text_tokens.cuda()
encoder_state = encode_torch(
text_tokens,
config,

View File

@ -127,10 +127,10 @@ class DalleBartDecoderTorch(nn.Module):
self.start_token = torch.tensor([start_token]).to(torch.long)
self.pad_token = torch.tensor([1]).to(torch.long)
self.condition_factor = torch.tensor([10]).to(torch.float)
if torch.cuda.is_available():
self.start_token = self.start_token.cuda()
self.pad_token = self.pad_token.cuda()
self.condition_factor = self.condition_factor.cuda()
# if torch.cuda.is_available():
# self.start_token = self.start_token.cuda()
# self.pad_token = self.pad_token.cuda()
# self.condition_factor = self.condition_factor.cuda()
self.image_token_count = image_token_count
self.embed_tokens = nn.Embedding(image_vocab_size + 1, embed_count)
self.embed_positions = nn.Embedding(image_token_count, embed_count)
@ -204,7 +204,7 @@ class DalleBartDecoderTorch(nn.Module):
for i in range(self.sample_token_count):
token_index = torch.tensor([i]).to(torch.long)
if torch.cuda.is_available(): token_index = token_index.cuda()
# if torch.cuda.is_available(): token_index = token_index.cuda()
probs, keys_values_state = self.decode_step(
text_tokens = text_tokens,
encoder_state = encoder_state,