File size: 3,372 Bytes
e8e23b0 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 | # BERT: fill-mask with Transformers
This page is the Transformers contract of the BERT MIC and also a usage example. [doc-builder](https://github.com/huggingface/doc-builder/blob/main/docs/runnable-code-blocks.md) renders the `runnable` blocks below as documentation, and its pytest plugin runs them as written against `google-bert/bert-base-uncased`. The MIC runner runs the same blocks against each slim fixture and each checkpoint in [`integration.json`](../../integration.json). Lines marked `# doc-builder: hide` only matter there: they swap in the pinned checkpoint or variant and print the result record.
## Predict the masked word
```py runnable:fill_mask
import torch
from transformers import AutoModelForMaskedLM, AutoTokenizer, set_seed
import json # doc-builder: hide
import os # doc-builder: hide
model_id = "google-bert/bert-base-uncased"
model_id = os.environ.get("MIC_MODEL", model_id) # doc-builder: hide
variant = json.loads(os.environ.get("MIC_MODEL_KWARGS", "{}")) # doc-builder: hide
device = "cuda" if torch.cuda.is_available() else "cpu"
set_seed(0) # seeds torch, NumPy, and Python: loading, dropout, and data augmentation draw random numbers
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForMaskedLM.from_pretrained(model_id).to(device)
if variant: # doc-builder: hide
model = AutoModelForMaskedLM.from_pretrained(model_id, **variant).to(device)
countries = ["France", "Germany", "Italy", "the United Kingdom"]
prompts = [f"The capital of {country} is {tokenizer.mask_token}." for country in countries]
inputs = tokenizer(prompts, return_tensors="pt", padding=True).to(device)
with torch.no_grad():
logits = model(**inputs).logits
masked = inputs["input_ids"] == tokenizer.mask_token_id
predicted = logits[masked].argmax(dim=-1)
print(tokenizer.convert_ids_to_tokens(predicted))
```
## One training step
```py runnable:fill_mask:2
answers = ["paris", "berlin", "rome", "london"]
answer_ids = tokenizer.convert_tokens_to_ids(answers)
labels = torch.full_like(inputs["input_ids"], -100)
labels[masked] = torch.tensor(answer_ids, device=device)
model.train()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)
loss = model(**inputs, labels=labels).loss
loss.backward()
before = {name: p.detach().clone() for name, p in model.named_parameters()} # doc-builder: hide
optimizer.step()
print(f"loss: {loss.item():.4f}")
if os.environ.get("MIC_RUNNER"): # doc-builder: hide
named = dict(model.named_parameters())
checked = ("bert.encoder.layer.0.attention.self.query.weight", "bert.encoder.layer.0.intermediate.dense.weight")
result = {
"runtime": {
"device": model.device.type,
"dtype": str(model.dtype).removeprefix("torch."),
"attn_implementation": model.config._attn_implementation,
},
"predicted_tokens": predicted.tolist(),
"predicted_text": tokenizer.convert_ids_to_tokens(predicted),
"answer_logits": logits[masked][:, answer_ids].float().tolist(),
"training": {
"loss": loss.item(),
"gradients": {n: bool(named[n].grad is not None and torch.count_nonzero(named[n].grad) > 0 and torch.isfinite(named[n].grad).all()) for n in checked},
"updated": any(not torch.equal(p, before[n]) for n, p in named.items()),
},
}
print(json.dumps(result))
```
|