Skip to content

fix dynamic Cfp8 for RL load #4144

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Nov 3, 2025
Merged

fix dynamic Cfp8 for RL load #4144

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
449587b
fix Cfp8 for RL load
rsmallblue Sep 16, 2025
8c74f53
Merge branch 'develop' into RL-c8
rsmallblue Sep 25, 2025
19c7930
Merge branch 'develop' into RL-c8
rsmallblue Oct 31, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 9 additions & 0 deletions fastdeploy/model_executor/layers/attention/attention.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,9 @@
from paddleformers.utils.log import logger

from fastdeploy.config import FDConfig
from fastdeploy.model_executor.layers.quantization.kv_cache import (
KvCacheQuantzationTypes,
)
from fastdeploy.model_executor.layers.quantization.quant_base import QuantMethodBase

if TYPE_CHECKING:
Expand Down Expand Up @@ -107,6 +110,12 @@ def __init__(

if fd_config.quant_config and hasattr(fd_config.quant_config, "kv_cache_quant_type"):
self.quant_method: QuantMethodBase = fd_config.quant_config.get_quant_method(self)

# set for RL model, as RL do not need load state dict
if fd_config.quant_config.kv_cache_quant_type == KvCacheQuantzationTypes.BLOCK_WISE_FP8:
self.cache_quant_type_str = "block_wise_fp8"
self.quant_max_bound = 448.0
self.quant_min_bound = -448.0
else:
self.quant_method = None

Expand Down