Comments (6)
Hi @philschmid. We have reproduced the issue and have a fix which will be available in an upcoming release.
from transformers-neuronx.
Can you confirm this is fixed with latest release ?
This is working for me with batch_size=2
and tp_degree=4
. tp_degree should be a divisor of number of attention heads, just fyi.
from transformers-neuronx.
FYI. Running the sample with a batch_size=2
works.
from transformers-neuronx.
Hi @aws-ennst any updates on this?
from transformers-neuronx.
Can you confirm this is fixed with latest release ?
from transformers-neuronx.
Hi @dacorvo , I have confirmed that the GPT-2 code snippet from README.md is now working with release 2.12:
(aws_neuron_venv_pytorch) ubuntu@ip-10-0-10-149:~$ python test.py
.
Compiler status PASS
.
Compiler status PASS
Downloading (…)olve/main/vocab.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1.04M/1.04M [00:00<00:00, 64.1MB/s]
Downloading (…)olve/main/merges.txt: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 456k/456k [00:00<00:00, 38.3MB/s]
Downloading (…)/main/tokenizer.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1.36M/1.36M [00:00<00:00, 34.0MB/s]
Setting `pad_token_id` to `eos_token_id`:50256 for open-end generation.
['Hello, I\'m a language model, you know. So I\'ll give you an example.\n\nNow this is really important, and you know: if you\'re a programmer, you\'re going to have to understand these things. I don\'t mean programming, or programming about languages. I mean writing. And so the first thing I need to do is write a program for the program, and then I need to write a function. And that\'s a lot of stuff, and it\'s not something that you could just throw away.\n\nThen you don\'t have to write a compiler. You can write a language, and you can write a function, and they\'ll look at your code and say, "Okay, this is going to be a function."\n\nNow, because languages are so powerful, and because they\'re so complicated, if it\'s a function, then you can write a function, and you\'re going to be able to write a function, and you\'re going to be able to write a function. And so that\'s a big problem, and it\'s one that\'s really hard to fix.\n\nSo the first thing I need to do is write a program. And then I need to write a function, and I need to write']
Packages:
(aws_neuron_venv_pytorch) ubuntu@ip-10-0-10-149:~$ pip list | grep neuron
aws-neuronx-runtime-discovery 2.9
libneuronxla 0.5.391
neuronx-cc 2.8.0.25+a3ad0f342
neuronx-distributed 0.1.0
neuronx-hwm 2.8.0.3+2b7c6da39
torch-neuronx 1.13.1.1.9.0
torch-xla 1.13.1+torchneuron8
transformers-neuronx 0.5.58
from transformers-neuronx.
Related Issues (20)
- Llama2 inference overhead time way too long HOT 6
- llama-2/codellama benchmark for inf2.xlarge HOT 4
- Mixtral Model support HOT 2
- Vicuna13B model support HOT 1
- Inf2 Modified Llama 2 Loading Issue HOT 11
- Skipping generation for useless tokens, and modiying cacheids HOT 3
- How to use generate() with inputs_embeds HOT 2
- Mixtral config issue -- not handling null well HOT 8
- Generate Llama 2 from Embeddings HOT 5
- Infering logits from `model.forward` for the entire batch instead of the last forward's output. HOT 6
- Support for MPT model HOT 1
- `stopping_criteria_list(input_ids, probs)` does not check for the correct sequence. HOT 4
- User feedback when compiling and reloading a large model HOT 1
- Issue while compiling Mistral 7B 0.2 Instruct HOT 5
- Backward compatibility with saved llama 2 compiled artifacts HOT 1
- NaN outputs when masking llama model inputs HOT 8
- Improve Neuron model loading time HOT 4
- Add support for `gemma` models HOT 1
- Add support for Baichuan-13B model
- Latest changes introduced for continuous batching break Mixtral model HOT 5
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 transformers-neuronx.