Comments (4)
zch-cc
commented on August 18, 2024
find similar issue here #1957 and it looks like versions after 2.0.2 is not working. I tried all versions after 2.0.2, all fails
from text-generation-inference.
ErikKaum
commented on August 18, 2024
Hi!
Thanks for reporting the issue 👍 could you share a bit on how to reproduce this?
E.g. which model are you using, what's the command to launch the docker container?
from text-generation-inference.
zch-cc
commented on August 18, 2024
Hi @ErikKaum,
Thanks for responding. The model is
deepseek-ai/deepseek-coder-6.7b-base
and you can use the official docker container command to reproduce this
docker run --gpus all --shm-size 1g -p 8080:80 -v $volume:/data \
ghcr.io/huggingface/text-generation-inference:2.1.1 --model-id $model
from text-generation-inference.
ErikKaum
commented on August 18, 2024
Thank you.
Yeah, I'm able to reproduce this on my machine. Also quickly checking seems that deepseek-ai/deepseek-coder-6.7b-base is working with the transformers library. So most likely a bug in our end.
At the moment, I unfortunately don't have bandwidth to start debugging.
I'm seeing a lot of warnings like this:
2024-07-16T15:00:17.309591Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'õ' was expected to have ID '32000' but was given ID 'None'
2024-07-16T15:00:17.309615Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '÷' was expected to have ID '32001' but was given ID 'None'
2024-07-16T15:00:17.309618Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'Á' was expected to have ID '32002' but was given ID 'None'
2024-07-16T15:00:17.309621Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ý' was expected to have ID '32003' but was given ID 'None'
2024-07-16T15:00:17.309624Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'À' was expected to have ID '32004' but was given ID 'None'
2024-07-16T15:00:17.309626Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ÿ' was expected to have ID '32005' but was given ID 'None'
2024-07-16T15:00:17.309629Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ø' was expected to have ID '32006' but was given ID 'None'
2024-07-16T15:00:17.309631Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ú' was expected to have ID '32007' but was given ID 'None'
2024-07-16T15:00:17.309641Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'þ' was expected to have ID '32008' but was given ID 'None'
2024-07-16T15:00:17.309643Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ü' was expected to have ID '32009' but was given ID 'None'
2024-07-16T15:00:17.309646Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ù' was expected to have ID '32010' but was given ID 'None'
2024-07-16T15:00:17.309648Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'ö' was expected to have ID '32011' but was given ID 'None'
2024-07-16T15:00:17.309651Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token 'û' was expected to have ID '32012' but was given ID 'None'
2024-07-16T15:00:17.309653Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<｜begin▁of▁sentence｜>' was expected to have ID '32013' but was given ID 'None'
2024-07-16T15:00:17.309656Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<｜end▁of▁sentence｜>' was expected to have ID '32014' but was given ID 'None'
2024-07-16T15:00:17.309658Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<｜fim▁hole｜>' was expected to have ID '32015' but was given ID 'None'
2024-07-16T15:00:17.309661Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<｜fim▁begin｜>' was expected to have ID '32016' but was given ID 'None'
2024-07-16T15:00:17.309664Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<｜fim▁end｜>' was expected to have ID '32017' but was given ID 'None'
2024-07-16T15:00:17.309666Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<pad>' was expected to have ID '32018' but was given ID 'None'
2024-07-16T15:00:17.309669Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<|User|>' was expected to have ID '32019' but was given ID 'None'
2024-07-16T15:00:17.309672Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<|Assistant|>' was expected to have ID '32020' but was given ID 'None'
2024-07-16T15:00:17.309674Z WARN tokenizers::tokenizer::serialization: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/tokenizers-0.19.1/src/tokenizer/serialization.rs:159: Warning: Token '<|EOT|>' was expected to have ID '32021' but was given ID 'None'
2024-07-16T15:00:17.310063Z INFO text_generation_router: router/src/main.rs:330: Overriding LlamaTokenizer with TemplateProcessing to follow python override defined in https://github.com/huggingface/transformers/blob/4aa17d00690b7f82c95bb2949ea57e22c35b4336/src/transformers/models/llama/tokenization_llama_fast.py#L203-L205
which makes me think it might be a tokenization issue.
from text-generation-inference.
Related Issues (20)
- load eetq model HOT 1
- error warming up cohere/aya-23-35b model: error in `reshape_and_cache` function with TGI 2.1.1 HOT 1
- How to clean the TGI guidance cache? HOT 5
- TGI crashes on Multi GPUs HOT 1
- Tgi crash on multi GPUs HOT 10
- Description of Tools format within Documentation HOT 2
- OPENAI Format Completion endpoint does not include finish reason and logprobs HOT 1
- `bitsandbytes==0.43.0` not supported on MacOS HOT 1
- Support for no_repeat_ngram_size generation parameter HOT 3
- OpenAI supports `top_p = 0.0` and `top_p = 1.0` but TGI fails with a validation error with either of these values. HOT 4
- CPU and Memory Utilization for TGI HOT 1
- Model offloading (preparing for Llama3 405b) HOT 3
- Nightly load test results HOT 3
- Regression: get_weights_col_packed_qkv() quantize parameter not included in calls so fails with error missing positional parfameter - merge error?
- Shard process was signaled to shutdown with signal 4 rank=0 Error: ShardCannotStart HOT 13
- Can I somehow change attention type from 'FlashAttention' in the text-server-launcher? HOT 1
- `tool_calls` gives sporadic EoF parsing errors HOT 4
- max_batch_size limit doesn't work well at queue.next_batch() HOT 2
- can't start server with small --max-total-tokens. But works fine with big stting HOT 4
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 text-generation-inference.