Feat/add nemotron nano v3 tests (#33345)
This commit is contained in:
@@ -0,0 +1,15 @@
|
|||||||
|
model_name: "nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16"
|
||||||
|
tasks:
|
||||||
|
- name: "gsm8k"
|
||||||
|
metrics:
|
||||||
|
- name: "exact_match,strict-match"
|
||||||
|
value: 0.695
|
||||||
|
- name: "exact_match,flexible-extract"
|
||||||
|
value: 0.447
|
||||||
|
limit: 1319
|
||||||
|
num_fewshot: 5
|
||||||
|
max_model_len: 262144
|
||||||
|
enforce_eager: false
|
||||||
|
apply_chat_template: true
|
||||||
|
fewshot_as_multiturn: true
|
||||||
|
trust_remote_code: true
|
||||||
@@ -0,0 +1,19 @@
|
|||||||
|
model_name: "nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-FP8"
|
||||||
|
tasks:
|
||||||
|
- name: "gsm8k"
|
||||||
|
metrics:
|
||||||
|
- name: "exact_match,strict-match"
|
||||||
|
value: 0.7142
|
||||||
|
- name: "exact_match,flexible-extract"
|
||||||
|
value: 0.4579
|
||||||
|
env_vars:
|
||||||
|
VLLM_USE_FLASHINFER_MOE_FP8: "1"
|
||||||
|
VLLM_FLASHINFER_MOE_BACKEND: "throughput"
|
||||||
|
limit: 1319
|
||||||
|
num_fewshot: 5
|
||||||
|
max_model_len: 262144
|
||||||
|
kv_cache_dtype: fp8
|
||||||
|
enforce_eager: false
|
||||||
|
apply_chat_template: true
|
||||||
|
fewshot_as_multiturn: true
|
||||||
|
trust_remote_code: true
|
||||||
@@ -1 +1,2 @@
|
|||||||
Qwen3-235B-A22B-Instruct-2507-FP8.yaml
|
Qwen3-235B-A22B-Instruct-2507-FP8.yaml
|
||||||
|
NVIDIA-Nemotron-3-Nano-30B-A3B-FP8.yaml
|
||||||
|
|||||||
@@ -3,3 +3,4 @@ Meta-Llama-3-70B-Instruct.yaml
|
|||||||
Mixtral-8x7B-Instruct-v0.1.yaml
|
Mixtral-8x7B-Instruct-v0.1.yaml
|
||||||
Qwen2-57B-A14-Instruct.yaml
|
Qwen2-57B-A14-Instruct.yaml
|
||||||
DeepSeek-V2-Lite-Chat.yaml
|
DeepSeek-V2-Lite-Chat.yaml
|
||||||
|
NVIDIA-Nemotron-3-Nano-30B-A3B-BF16.yaml
|
||||||
|
|||||||
@@ -355,5 +355,22 @@
|
|||||||
"is_deepseek_mla": true,
|
"is_deepseek_mla": true,
|
||||||
"is_multimodal_model": false,
|
"is_multimodal_model": false,
|
||||||
"dtype": "torch.float32"
|
"dtype": "torch.float32"
|
||||||
|
},
|
||||||
|
"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16": {
|
||||||
|
"architectures": [
|
||||||
|
"NemotronHForCausalLM"
|
||||||
|
],
|
||||||
|
"model_type": "nemotron_h",
|
||||||
|
"text_model_type": "nemotron_h",
|
||||||
|
"hidden_size": 2688,
|
||||||
|
"total_num_hidden_layers": 52,
|
||||||
|
"total_num_attention_heads": 32,
|
||||||
|
"head_size": 128,
|
||||||
|
"vocab_size": 131072,
|
||||||
|
"total_num_kv_heads": 2,
|
||||||
|
"num_experts": 128,
|
||||||
|
"is_deepseek_mla": false,
|
||||||
|
"is_multimodal_model": false,
|
||||||
|
"dtype": "torch.bfloat16"
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|||||||
@@ -14,6 +14,7 @@ from vllm.transformers_utils.model_arch_config_convertor import (
|
|||||||
|
|
||||||
BASE_TRUST_REMOTE_CODE_MODELS = {
|
BASE_TRUST_REMOTE_CODE_MODELS = {
|
||||||
"nvidia/Llama-3_3-Nemotron-Super-49B-v1",
|
"nvidia/Llama-3_3-Nemotron-Super-49B-v1",
|
||||||
|
"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16",
|
||||||
"XiaomiMiMo/MiMo-7B-RL",
|
"XiaomiMiMo/MiMo-7B-RL",
|
||||||
# Excluded: Not available online right now
|
# Excluded: Not available online right now
|
||||||
# "FreedomIntelligence/openPangu-Ultra-MoE-718B-V1.1",
|
# "FreedomIntelligence/openPangu-Ultra-MoE-718B-V1.1",
|
||||||
|
|||||||
Reference in New Issue
Block a user