Comments (2)
HansBambel
commented on July 26, 2024
2
When using the generate_unconditioned_samples.py or the other sampling files you can set a limit to the number of tokens you want to have using --length <number of tokens>. By default this is set to the value of n_ctx in the hparams.json --> 1024 for the 117M/124M model.
The <|endoftext|> shows that the message is done, but it still generates the previously set amount of tokens. Resulting in multiple messages.
To limit messages to stop until an <|endoftext|> occurs would require you to create it yourself. Since it is easier for me I am using PyTorch for that to sample until a . is found (could be adapted to <|endoftext|>. (I fine-tune using TF but then convert to pytorch. Quite annoying :-/)
from gpt-2.
ProtoxiDe22
commented on July 26, 2024
1
i think i got my answers here, thanks
from gpt-2.
Related Issues (20)
- gpt2 translation task HOT 1
- How to train in multiple gpu HOT 1
- Early Stopping
- OOM on 345M with GPU HOT 4
- "ModelNotFoundError": No model named "encoder"
- ModuleNotFoundError: No module named 'encoder' HOT 7
- Where to enter model name as parameter HOT 1
- Cannot download the models HOT 4
- File "encode.py", line 23 HOT 1
- GPT2 Fine Tuning does not support Ampere (RTX 3000s) Cards
- Dockerfile no longer accurate? HOT 2
- ModuleNotFoundError: No module named 'Sampler' HOT 3
- I can't train my dataset HOT 3
- Need some help getting Tensor Rematerialization to work HOT 7
- this parameter cannot be zero when i using my file history.npz
- Encoder doesn't work HOT 1
- avg stays in 2.6-2.9 range
- how to use trained data to generate text? HOT 1
- GPT-2 doesn't have Encode.py or Train.py.
- Training on TPU HOT 1
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 gpt-2.