matthewchung74

python -m transformers.onnx -m="microsoft/trocr-large-printed" --framework="pt" --feature="vision2seq-lm" . 
Some weights of VisionEncoderDecoderModel were not initialized from the model checkpoint at microsoft/trocr-large-printed and are newly initialized: ['encoder.pooler.dense.weight', 'encoder.pooler.dense.bias']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
Using framework PyTorch: 1.13.1
/Users/mattc/Desktop/onnx/venv/lib/python3.10/site-packages/transformers/models/vit/modeling_vit.py:176: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if num_channels != self.num_channels:
/Users/mattc/Desktop/onnx/venv/lib/python3.10/site-packages/transformers/models/vit/modeling_vit.py:181: TracerWarning: Converting a tensor to a Python 

matthewchung74

#read all files, concat
from pathlib import Path
import pandas as pd
pathlist = Path('./Reports_Base').glob('*.tsv')
df_array = []
for path in pathlist:
    df = pd.read_csv(path, sep='\t', encoding='iso-8859-1')
    df_array.append(df)

df = pd.concat(df_array)

matthewchung74

Launching lib/main.dart on iPhone 12 Pro Max in debug mode...
Xcode build done.                                           35.1s
Failed to build iOS app
Error output from Xcode build:
↳
    ** BUILD FAILED **
Xcode's output:
↳
    /Users/mattc/Development/tutorials/day18_todo/ios/Pods/SQLite.swift/Sources/SQLite/Core/Connection.swift:431:13: warning: 'sqlite3_trace' was deprecated in iOS 10.0: renamed to 'sqlite3_trace_v2(_:_:_:_:)'
                sqlite3_trace(handle, nil /* xCallback */, nil /* pCtx */)

matthewchung74

class Mlp(nn.Module):
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.drop = nn.Dropout(drop)

matthewchung74

embed_dim = 768
num_heads = 8
block = Block(embed_dim, 8)

batch_size = 1
class_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
class_tokens = class_token.expand(batch_size, -1, -1)
pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))

x = torch.cat((class_tokens, patch_output), dim=1)

matthewchung74

x = self.patch_embed(x)
cls_tokens = self.cls_token.expand(B, -1, -1)
x = torch.cat((cls_tokens, x), dim=1)
x = x + self.pos_embed

matthewchung74

from PIL import Image
loader = DataLoader(dataset = train_dataset, batch_size=1, shuffle=False )
first_batch = next(iter(loader))[0]
model = PatchEmbed()
patch_output = model(first_batch)
batch_size, num_patches, embedding = patch_output.shape
num_patches, embedding
(196, 768)

matthewchung74

from itertools import repeat
import pdb
class PatchEmbed(nn.Module):
  def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768):
    super().__init__()
    img_size = tuple(repeat(img_size, 2))
    patch_size = tuple(repeat(patch_size, 2))
    num_patches = (img_size[1] // patch_size[1]) * (img_size[0] //   patch_size[0])
    self.img_size = img_size
    self.patch_size = patch_size

matthewchung74

self.blocks = nn.ModuleList([
    Block(
        dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
        drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer)
    for i in range(depth)])

matthewchung74

class Block(nn.Module):
  def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0., drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm):
    super().__init__()
    self.norm1 = nn.LayerNorm(dim)
    self.attn = Attention(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
    self.norm2 = nn.LayerNorm(dim)
    mlp_hidden_dim = int(dim * mlp_ratio)
    self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
def forward(self, x):
    x = x + self.attn(self.norm1(x))