[Model] Support DeepSeek-V4.1-Flash (#56214)
Signed-off-by: Yongye Zhu <zyy1102000@gmail.com> Signed-off-by: Jee Jee Li <jeejeelee@inferact.ai> Signed-off-by: andyluo7 <andy.luo@amd.com> Signed-off-by: zjy0516 <riverclouds.zhu@qq.com> Co-authored-by: Jee Jee Li <jeejeelee@inferact.ai> Co-authored-by: andyluo7 <andy.luo@amd.com> Co-authored-by: OpenAI Codex <codex@openai.com> Co-authored-by: zjy0516 <riverclouds.zhu@qq.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Jee Jee Li
andyluo7
OpenAI Codex
zjy0516
Claude Opus 5
parent
e6821ceac9
commit
e77daef89e
@@ -330,7 +330,7 @@ class DeepseekV4ForCausalLMConfig(VerifyAndUpdateConfig):
|
||||
quant_config = getattr(model_config.hf_config, "quantization_config", None)
|
||||
if quant_config is not None and quant_config.get("quant_method") == "fp8":
|
||||
model_type = getattr(model_config.hf_config, "model_type", None)
|
||||
if model_type == "deepseek_v4":
|
||||
if model_type in ("deepseek_v4", "deepseek_v41"):
|
||||
model_config.hf_config.quantization_config["quant_method"] = (
|
||||
"deepseek_v4_fp8"
|
||||
)
|
||||
@@ -343,7 +343,7 @@ class DeepseekV4ForCausalLMConfig(VerifyAndUpdateConfig):
|
||||
and hf_text_quant_config.get("quant_method") == "fp8"
|
||||
):
|
||||
model_type = getattr(model_config.hf_text_config, "model_type", None)
|
||||
if model_type == "deepseek_v4":
|
||||
if model_type in ("deepseek_v4", "deepseek_v41"):
|
||||
model_config.hf_text_config.quantization_config["quant_method"] = (
|
||||
"deepseek_v4_fp8"
|
||||
)
|
||||
@@ -1003,6 +1003,7 @@ MODELS_CONFIG_MAP: dict[str, type[VerifyAndUpdateConfig]] = {
|
||||
"ColQwen3_5": ColQwen3_5Config,
|
||||
"DeepseekV4ForCausalLM": DeepseekV4ForCausalLMConfig,
|
||||
"DeepseekV4ForConditionalGeneration": DeepseekV4ForCausalLMConfig,
|
||||
"DeepseekV41ForCausalLM": DeepseekV4ForCausalLMConfig,
|
||||
"DeepseekV32ForCausalLM": DeepseekV32ForCausalLM,
|
||||
"DiffusionGemmaForBlockDiffusion": DiffusionGemmaModelForBlockDiffusionConfig, # noqa: E501
|
||||
"Ernie4_5_VLMoeForConditionalGeneration": Ernie4_5_VLMoeForConditionalGenerationConfig, # noqa: E501
|
||||
|
||||
@@ -376,6 +376,10 @@ _MULTIMODAL_MODELS = {
|
||||
"vllm.models.deepseek_v4",
|
||||
"DeepseekV4ForConditionalGeneration",
|
||||
),
|
||||
"DeepseekV41ForCausalLM": (
|
||||
"vllm.models.deepseek_v4_1",
|
||||
"DeepseekV41ForCausalLM",
|
||||
),
|
||||
"Dots3NoteForCausalLM": (
|
||||
"vllm.models.dots3_note",
|
||||
"Dots3NoteForCausalLM",
|
||||
@@ -640,6 +644,10 @@ _SPECULATIVE_DECODING_MODELS = {
|
||||
"MuseGlimmerAssistantModel": ("qwen3_dflash", "DFlashQwen3ForCausalLM"),
|
||||
"DFlashMuseGlimmerAssistantModel": ("qwen3_dflash", "DFlashQwen3ForCausalLM"),
|
||||
"DSparkDraftModel": ("vllm.models.deepseek_v4", "DSparkDeepseekV4ForCausalLM"),
|
||||
"DSparkV41DraftModel": (
|
||||
"vllm.models.deepseek_v4_1",
|
||||
"DSparkDeepseekV4ForCausalLM",
|
||||
),
|
||||
"Qwen3DSparkModel": ("qwen3_dspark", "Qwen3DSparkForCausalLM"),
|
||||
"Qwen3OmniDSparkModel": ("qwen3_dspark", "Qwen3DSparkForCausalLM"),
|
||||
"K3DSparkModel": (
|
||||
|
||||
Reference in New Issue
Block a user