Ctrl+K

1 contributor

History: 4 commits

demon-zombie

Rewrite README: minimal style, drop perf data, match nvidia M2.5 recipe details

1075a4d verified 29 days ago

.gitattributes

1.58 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
README.md

3.11 kB
Rewrite README: minimal style, drop perf data, match nvidia M2.5 recipe details 29 days ago
added_tokens.json

1.46 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
chat_template.jinja

6.52 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
config.json

8.8 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
configuration_minimax_m2.py

10.2 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
generation_config.json

166 Bytes
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
hf_quant_config.json

6.14 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
merges.txt

2.41 MB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00001-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00002-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00003-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00004-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00005-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00006-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00007-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00008-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00009-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00010-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00011-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00012-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00013-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00014-of-00029.safetensors

4.96 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00015-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00016-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00017-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00018-of-00029.safetensors

4.95 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00019-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00020-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00021-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00022-of-00029.safetensors

4.95 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00023-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00024-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00025-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00026-of-00029.safetensors

4.95 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00027-of-00029.safetensors

5 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00028-of-00029.safetensors

3.87 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model-00029-of-00029.safetensors

1.23 GB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
model.safetensors.index.json

19.5 MB
xet

NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
modeling_minimax_m2.py

30.9 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
special_tokens_map.json

1.53 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
tokenizer.json

9.73 MB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
tokenizer_config.json

10.9 kB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago
vocab.json

4.71 MB
NVFP4 quant via modelopt 0.43.0rc3 + TensorRT-Model-Optimizer examples, recipe general/ptq/nvfp4_experts_only-fp8_kv (calib_size=1024, moe_calib_experts_ratio=1.0, cnn_dailymail + Nemotron-Post-Training-Dataset-v2, batch_size=64->32 auto-fallback, source=FP8 M2.7, FP8->bf16 on-forward dequant, 4x H100 SXM) 29 days ago