GitLab

Co-authored-by: Michael Goin <michael@neuralmagic.com>

Signed-off-by: Ye Cao <caoye.cao@alibaba-inc.com>

Co-authored-by: Cody Yu <hao.yu.cody@gmail.com>

Co-authored-by: Varun Sundar Rabindranath <varunsundar08@gmail.com>
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com>

Allow dummy load format for fp8,
torch.uniform_ doesn't support FP8 at the moment
Co-authored-by: Mor Zusman <morz@ai21.com>

Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>

[Frontend] [Core] perf: Automatically detect vLLM-tensorized model, update `tensorizer` to version 2.9.0 (#4208)

Storing exception frame is extremely prone to circular refernece because it contains the reference to objects.

When tensorizer is not installed, it leaks llm instance because error frame has references to various modules which cause circular reference problem.

I also found spec decoding has a circular reference issue, and I solved it using weakref.proxy.

Signed-off-by: Prashant Gupta <prashantgupta@us.ibm.com>
Co-authored-by: Travis Johnson <tjohnson31415@gmail.com>

Co-authored-by: Danny Guinther <dguinther@neuralmagic.com>

This PR is the first step towards fixing https://github.com/vllm-project/vllm/pull/3208

It implements dynamic per-tensor scaling (see https://github.com/vllm-project/vllm/pull/4118), so users do not need to compute activation scales on a calibration dataset and they also don't need to convert their model checkpoints. It is enough to specify the `quantization="fp8"` argument. You can try out the PR like this:

```python
from vllm import LLM, SamplingParams

prompts = [
    "Hello, my name is",
    "The president of the United States is",
    "The capital of France is",
    "The future of AI is",
]
sampling_params = SamplingParams(temperature=0.8, top_p=0.95)

llm = LLM(model="mistralai/Mixtral-8x7B-Instruct-v0.1", tensor_parallel_size=2, quantization="fp8")

outputs = llm.generate(prompts, sampling_params)

# Print the outputs.
for output in outputs:
    prompt = output.prompt
    generated_text = output.outputs[0].text
    print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
```

**Performance**: For this PR, the focus is on making the code clean (while still trying to get reasonable performance), there is a bunch of optimizations that we will submit as a follow up PR that significantly improve the performance (similar to the numbers in https://github.com/vllm-project/vllm/pull/3954). With this PR, the results are as follows:

<img width="725" alt="Screenshot 2024-04-21 at 1 31 50 PM" src="https://github.com/vllm-project/vllm/assets/113316/d8fe1118-07a0-4d4e-8530-37a77d465a03">


**Accuracy**: The accuracy with this PR on MMLU on `mistralai/Mixtral-8x7B-v0.1` is as follows:

```
|      Groups      |Version|Filter|n-shot|Metric|Value |   |Stderr|
|------------------|-------|------|-----:|------|-----:|---|-----:|
|mmlu              |N/A    |none  |     0|acc   |0.7018|±  |0.0036|
| - humanities     |N/A    |none  |     5|acc   |0.6472|±  |0.0065|
| - other          |N/A    |none  |     5|acc   |0.7673|±  |0.0072|
| - social_sciences|N/A    |none  |     5|acc   |0.8099|±  |0.0070|
| - stem           |N/A    |none  |     5|acc   |0.6131|±  |0.0083|
```
this compares favorably with the fp16 results which are
```
|      Groups      |Version|Filter|n-shot|Metric|Value |   |Stderr|
|------------------|-------|------|-----:|------|-----:|---|-----:|
|mmlu              |N/A    |none  |     0|acc   |0.7020|±  |0.1313|
| - humanities     |N/A    |none  |     5|acc   |0.6425|±  |0.1349|
| - other          |N/A    |none  |     5|acc   |0.7744|±  |0.1038|
| - social_sciences|N/A    |none  |     5|acc   |0.8131|±  |0.0695|
| - stem           |N/A    |none  |     5|acc   |0.6108|±  |0.1383|
```

Happy hacking!

Provide an initial support to FP8 computation. This PR is inspired by HuggingFace TGI: huggingface/text-generation-inference#1726

This feature can be enabled with --quantization fp8 or -q fp8 when launching an engine.

Algorithm:
We still load a model checkpoint in FP16/BF16. After the weights are loaded, Fp8LinearMethod calculates the per-tensor scaling factor of weights and quantizes the weights accordingly. The scaling factor will then be stored for future use. Meanwhile, the per-tensor scaling factor for activations is calculated in every forward pass.

Initial Results:
Currently tested Mistral-7B on 1xH100. With prompt length ~5 and decoding length 128:

BF16: 1.47s
FP8: 1.66s
I'll try to use larger models and try to find more performance bottleneck. Meanwhile, you're welcome to try this code.

Co-authored-by: SangBin Cho <sangcho@sangcho-LT93GQWG9C.local>