removing quant and kv-cache fp8 from deepseek run instructions #509
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
arakowsk-amd
requested review from
shajrawi,
gshtras,
maleksan85,
sunway513 and
hongxiayang
as code owners
shajrawi
approved these changes
Apr 9, 2025
shajrawi
requested changes
Apr 9, 2025
docs/dev-docker/README.md
Outdated
| @@ -377,7 +377,7 @@ python3 /app/vllm/benchmarks/benchmark_serving.py \ | |||
| # Offline throughput | |||
| python3 /app/vllm/benchmarks/benchmark_throughput.py --model deepseek-ai/DeepSeek-V3 \ | |||
| --input-len <> --output-len <> --tensor-parallel-size 8 \ | |||
| --quantization fp8 --kv-cache-dtype fp8 --dtype float16 \ | |||
| --dtype float16 \ | |||
There was a problem hiding this comment.
Raises error an error:
export VLLM_MLA_DISABLE=0
export VLLM_USE_AITER=1
export VLLM_USE_TRITON_FLASH_ATTN=1
python3 /app/vllm/benchmarks/benchmark_throughput.py --model /data/DeepSeek-R1/ --input-len 128 --output-len 128 --tensor-parallel-size 8 --quantization fp8 --kv-cache-dtype fp8 --dtype bfloat16 --max-model-len 32768 --block-size=1 --trust-remote-code
[rank0]: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]: File "/usr/local/lib/python3.12/dist-packages/vllm/attention/backends/triton_mla.py", line 63, in __init__
[rank0]: raise NotImplementedError(
[rank0]: NotImplementedError: TritonMLA with FP8 KV cache not yet supportedThere was a problem hiding this comment.
Why is Triton MLA being used with AITER? cc @qli88
There was a problem hiding this comment.
@arakowsk-amd are you using the latest version? If you'd like we can discuss through Teams.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.