Comments (3)
Hi, this is not a bug. We need to backdrop gradient signal from these two losses to the embedding function in order to jointly train.
from diffusion-lm.
@XiangLi1999 Hi, I'm curious about the loss funtion too. I can not understand why we need to compute decoder_nll
using x_start
in decoder_nll = self.token_discrete_loss(x_start, get_logits, input_ids)
. I think x_start
is the word embedding added with extra noise, and this decode loss is trying to recover the noise, and this has no relation with diffusion model. Besides, this is not consistent with the formulation in loss function, in the \log p_theta (w|x_0) part. Why can't we replace x_start
to the predicted model_out_x_start
? Is this more reasonable? (But the experiment results is not good)
from diffusion-lm.
@XiangLi1999 Hi, I'm curious about the loss funtion too. I can not understand why we need to compute
decoder_nll
usingx_start
indecoder_nll = self.token_discrete_loss(x_start, get_logits, input_ids)
. I thinkx_start
is the word embedding added with extra noise, and this decode loss is trying to recover the noise, and this has no relation with diffusion model. Besides, this is not consistent with the formulation in loss function, in the \log p_theta (w|x_0) part. Why can't we replacex_start
to the predictedmodel_out_x_start
? Is this more reasonable? (But the experiment results is not good)
Hi @summmeer ,
I also share the same feelings about this problem. It seems that decoder_nll
has less correlation with the training of diffusion models. And I wonder how is it performing during your experiments? As I was training the model, I found that NLL loss equals to zero for a long training period (about 8k iterations). At about 10k
training steps, the NLL loss occurs with increasing values. Have you ever encountered the similar situation? How is the NLL loss during your experiments?
Thanks for your reply in advance. It would help me a lot.
Best,
from diffusion-lm.
Related Issues (20)
- I wander where to find the model in the predictability HOT 1
- Training on A100
- Separate weights for word embedding and lm-head?
- Questions about the result of success rate of PPLM? HOT 2
- Why not directly use Emb(W) as X_0? HOT 2
- Error when running training script on Google Colab HOT 2
- Fail to load GPT2 pretrained model for attribute controled generation
- Reproducing Table 5: Sentence Infilling - CIDEr / BLEU-4 metrics HOT 1
- Baseline reproduction
- error when runing:Exception in thread Thread-4:·······ValueError: signal number 32 out of range
- Which classifier to use in custom_trainer.py for controllable generation?
- About the tT_loss HOT 1
- The difference between this code and the paper "IDDPM" in the run_loop function in train_util.py.
- The relevant code that caused the error is in the Controllable Text Generation section, after the model trained for 6 epochs and started evaluating, it raised a KeyError: 'eval_loss' HOT 2
- Questions about the NLL loss
- E2E training procedure
- Issue while generating controllable text generation
- How to Execute the Semantic Content Subtask with infill.py
- Seq2Seq tasks with Diffusion LM
- Difficulty in running code
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from diffusion-lm.