apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 16:48:03 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[CP]Glm45 air 2.2 (#4073)
Some checks failed
CE Compile Job / ce_job_pre_check (push) Has been cancelled
Details
CE Compile Job / print_ce_job_pre_check_outputs (push) Has been cancelled
Details
CE Compile Job / FD-Clone-Linux (push) Has been cancelled
Details
CE Compile Job / Show Code Archive Output (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8090 (push) Has been cancelled
Details
CE Compile Job / BUILD_SM8689 (push) Has been cancelled
Details
CE Compile Job / CE_UPLOAD (push) Has been cancelled
Details

Browse Source 
* [Feature] Support zai-org/GLM-4.5-Air BF16 model (#3928)

* support glm45_air

* [Feature] GLM-45-AIR Support Mix Quantization(Dense wfp8afp8 and wint8 triton_moe_backend) (#4051)

* check

* fix v1 load for mix and wint8

* check --quantizations 'None'

* check

* support RL rollout

* check v1 loader

* check glm rollout_model, change wfp8afp8 per_token_cast_to_fp8 to native impl

* check rollout moe gate begin layer_id

* check rollout e_score_correction_bias

* delete infer_to_train_mapping={}

* code check
This commit is contained in:
chen
2025-09-15 18:52:58 +08:00
committed by GitHub
parent 4e8ba62241
commit fbb4e0f8d1
25 changed files with 1505 additions and 170 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
fastdeploy/model_executor/layers/quantization/weight_only.py
View File

@@ -44,6 +44,7 @@ class WeightOnlyConfig(QuantConfigBase):
def __init__(
self,
algo: str,
is_checkpoint_bf16: bool = False,
) -> None:
super().__init__()
self.algo = algo
@@ -55,6 +56,7 @@ class WeightOnlyConfig(QuantConfigBase):
self.quant_max_bound = 0
self.quant_min_bound = 0
self.quant_round_type = 0
self.is_checkpoint_bf16 = is_checkpoint_bf16
def name(self) -> str:
return "weight_only"
@@ -62,7 +64,8 @@ class WeightOnlyConfig(QuantConfigBase):
@classmethod
def from_config(cls, config: dict) -> "WeightOnlyConfig":
algo = config["algo"]
return cls(algo)
is_checkpoint_bf16 = config.get("is_checkpoint_bf16", False)
return cls(algo, is_checkpoint_bf16=is_checkpoint_bf16)
def get_quant_method(self, layer) -> Optional[QuantMethodBase]:
if current_platform.is_xpu():
@@ -153,12 +156,13 @@ class WINT8Config(WeightOnlyConfig):
weight only int8 config
"""
def __init__(self) -> None:
super().__init__("weight_only_int8")
def __init__(self, is_checkpoint_bf16: bool = False) -> None:
super().__init__("weight_only_int8", is_checkpoint_bf16)
@classmethod
def from_config(cls, config: dict) -> "WINT8Config":
return cls()
is_checkpoint_bf16 = config.get("is_checkpoint_bf16", False)
return cls(is_checkpoint_bf16)
def name(self) -> str:
return "wint8"