Image-Text-to-Video
Diffusers
Safetensors
orbitquant
comfyui
w4
w4a4
native-w4a4-transformer-runtime
text-to-video
audio-video-generation
8-bit precision
Instructions to use WaveCut/MiniMax-H3-OrbitQuant-W4A4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use WaveCut/MiniMax-H3-OrbitQuant-W4A4 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("WaveCut/MiniMax-H3-OrbitQuant-W4A4", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
| { | |
| "activation_backends": { | |
| "triton_cuda_packed_w4": 300 | |
| }, | |
| "audio_tokens": 207, | |
| "capability": [ | |
| 12, | |
| 0 | |
| ], | |
| "cuda": "12.8", | |
| "cuda_after_load_bytes": 18328776192, | |
| "cuda_forward_ms": 2867.617919921875, | |
| "cuda_peak_bytes": 46381307904, | |
| "effective_runtime_modes": { | |
| "native_packed_matmul": 300 | |
| }, | |
| "gpu": "NVIDIA RTX PRO 6000 Blackwell Server Edition", | |
| "h3_dtype_views": 50, | |
| "load_dtype": "bfloat16", | |
| "load_seconds": 0.28624553605914116, | |
| "model_dtype": "torch.bfloat16", | |
| "move_to_cuda_seconds": 1.1773170102387667, | |
| "num_blocks": 50, | |
| "orbitquant_linear_count": 300, | |
| "orbitquant_linear_cuda_ms": 1077.578048467636, | |
| "orbitquant_linear_inputs": { | |
| "transformer_blocks.0.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.0.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.0.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.0.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.0.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.0.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.1.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.1.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.1.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.1.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.1.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.1.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.10.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.10.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.10.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.10.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.10.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.10.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.11.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.11.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.11.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.11.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.11.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.11.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.12.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.12.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.12.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.12.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.12.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.12.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.13.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.13.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.13.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.13.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.13.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.13.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.14.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.14.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.14.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.14.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.14.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.14.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.15.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.15.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.15.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.15.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.15.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.15.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.16.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.16.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.16.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.16.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.16.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.16.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.17.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.17.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.17.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.17.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.17.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.17.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.18.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.18.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.18.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.18.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.18.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.18.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.19.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.19.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.19.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.19.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.19.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.19.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.2.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.2.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.2.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.2.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.2.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.2.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.20.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.20.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.20.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.20.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.20.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.20.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.21.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.21.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.21.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.21.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.21.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.21.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.22.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.22.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.22.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.22.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.22.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.22.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.23.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.23.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.23.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.23.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.23.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.23.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.24.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.24.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.24.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.24.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.24.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.24.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.25.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.25.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.25.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.25.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.25.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.25.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.26.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.26.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.26.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.26.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.26.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.26.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.27.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.27.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.27.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.27.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.27.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.27.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.28.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.28.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.28.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.28.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.28.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.28.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.29.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.29.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.29.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.29.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.29.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.29.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.3.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.3.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.3.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.3.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.3.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.3.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.30.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.30.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.30.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.30.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.30.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.30.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.31.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.31.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.31.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.31.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.31.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.31.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.32.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.32.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.32.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.32.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.32.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.32.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.33.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.33.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.33.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.33.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.33.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.33.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.34.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.34.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.34.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.34.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.34.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.34.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.35.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.35.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.35.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.35.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.35.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.35.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.36.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.36.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.36.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.36.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.36.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.36.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.37.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.37.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.37.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.37.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.37.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.37.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.38.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.38.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.38.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.38.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.38.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.38.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.39.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.39.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.39.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.39.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.39.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.39.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.4.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.4.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.4.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.4.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.4.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.4.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.40.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.40.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.40.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.40.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.40.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.40.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.41.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.41.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.41.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.41.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.41.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.41.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.42.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.42.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.42.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.42.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.42.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.42.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.43.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.43.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.43.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.43.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.43.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.43.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.44.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.44.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.44.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.44.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.44.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.44.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.45.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.45.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.45.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.45.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.45.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.45.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.46.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.46.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.46.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.46.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.46.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.46.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.47.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.47.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.47.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.47.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.47.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.47.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.48.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.48.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.48.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.48.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.48.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.48.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.49.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.49.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.49.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.49.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.49.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.49.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.5.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.5.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.5.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.5.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.5.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.5.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.6.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.6.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.6.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.6.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.6.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.6.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.7.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.7.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.7.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.7.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.7.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.7.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.8.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.8.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.8.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.8.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.8.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.8.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.9.attn.to_k": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.9.attn.to_out.0": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 7168 | |
| ], | |
| "stride": [ | |
| 80740352, | |
| 7168, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.9.attn.to_q": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.9.attn.to_v": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.9.ff.net.0.proj": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 5376 | |
| ], | |
| "stride": [ | |
| 60555264, | |
| 5376, | |
| 1 | |
| ] | |
| }, | |
| "transformer_blocks.9.ff.net.2": { | |
| "contiguous": true, | |
| "dtype": "torch.bfloat16", | |
| "shape": [ | |
| 1, | |
| 11264, | |
| 14336 | |
| ], | |
| "stride": [ | |
| 161480704, | |
| 14336, | |
| 1 | |
| ] | |
| } | |
| }, | |
| "orbitquant_shape_cuda_ms": { | |
| "14336x5376": { | |
| "calls": 50, | |
| "total_ms": 229.30192136764526 | |
| }, | |
| "5376x28672": { | |
| "calls": 50, | |
| "total_ms": 352.2663993835449 | |
| }, | |
| "5376x7168": { | |
| "calls": 150, | |
| "total_ms": 371.99267172813416 | |
| }, | |
| "7168x5376": { | |
| "calls": 50, | |
| "total_ms": 124.01705598831177 | |
| } | |
| }, | |
| "orbitquant_shapes": { | |
| "14336x5376": 50, | |
| "5376x28672": 50, | |
| "5376x7168": 150, | |
| "7168x5376": 50 | |
| }, | |
| "output_finite": [ | |
| true, | |
| true | |
| ], | |
| "output_shapes": [ | |
| [ | |
| 1, | |
| 10545, | |
| 96 | |
| ], | |
| [ | |
| 1, | |
| 207, | |
| 32 | |
| ] | |
| ], | |
| "runtime_mode": "native_packed_matmul", | |
| "slowest_orbitquant_linears": [ | |
| { | |
| "cuda_ms": 18.641632080078125, | |
| "name": "transformer_blocks.0.attn.to_q", | |
| "shape": "5376x7168" | |
| }, | |
| { | |
| "cuda_ms": 9.849472045898438, | |
| "name": "transformer_blocks.0.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 8.686016082763672, | |
| "name": "transformer_blocks.0.ff.net.2", | |
| "shape": "14336x5376" | |
| }, | |
| { | |
| "cuda_ms": 7.292928218841553, | |
| "name": "transformer_blocks.0.attn.to_out.0", | |
| "shape": "7168x5376" | |
| }, | |
| { | |
| "cuda_ms": 7.155712127685547, | |
| "name": "transformer_blocks.14.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.124512195587158, | |
| "name": "transformer_blocks.24.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.1244797706604, | |
| "name": "transformer_blocks.2.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.116352081298828, | |
| "name": "transformer_blocks.7.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.115295886993408, | |
| "name": "transformer_blocks.9.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.111199855804443, | |
| "name": "transformer_blocks.5.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.111199855804443, | |
| "name": "transformer_blocks.11.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.103007793426514, | |
| "name": "transformer_blocks.19.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.101952075958252, | |
| "name": "transformer_blocks.12.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.098879814147949, | |
| "name": "transformer_blocks.25.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.094880104064941, | |
| "name": "transformer_blocks.6.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0866241455078125, | |
| "name": "transformer_blocks.4.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.085055828094482, | |
| "name": "transformer_blocks.17.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.08355188369751, | |
| "name": "transformer_blocks.18.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.080480098724365, | |
| "name": "transformer_blocks.30.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.074431896209717, | |
| "name": "transformer_blocks.29.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.068192005157471, | |
| "name": "transformer_blocks.10.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.064576148986816, | |
| "name": "transformer_blocks.15.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.062047958374023, | |
| "name": "transformer_blocks.27.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0609917640686035, | |
| "name": "transformer_blocks.1.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.059999942779541, | |
| "name": "transformer_blocks.3.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.059999942779541, | |
| "name": "transformer_blocks.34.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.059967994689941, | |
| "name": "transformer_blocks.20.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.058976173400879, | |
| "name": "transformer_blocks.32.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.056896209716797, | |
| "name": "transformer_blocks.13.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.055903911590576, | |
| "name": "transformer_blocks.46.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| } | |
| ], | |
| "status": "pass", | |
| "text_tokens": 512, | |
| "torch": "2.8.0+cu128", | |
| "transformer": "/workspace/orbitquant-h3-perf/model/transformer", | |
| "transformer_block_cuda_ms": [ | |
| { | |
| "cuda_ms": 309.7416687011719, | |
| "name": "transformer_blocks.0" | |
| }, | |
| { | |
| "cuda_ms": 48.530784606933594, | |
| "name": "transformer_blocks.1" | |
| }, | |
| { | |
| "cuda_ms": 48.752193450927734, | |
| "name": "transformer_blocks.2" | |
| }, | |
| { | |
| "cuda_ms": 48.56070327758789, | |
| "name": "transformer_blocks.3" | |
| }, | |
| { | |
| "cuda_ms": 48.65910339355469, | |
| "name": "transformer_blocks.4" | |
| }, | |
| { | |
| "cuda_ms": 48.712257385253906, | |
| "name": "transformer_blocks.5" | |
| }, | |
| { | |
| "cuda_ms": 48.625247955322266, | |
| "name": "transformer_blocks.6" | |
| }, | |
| { | |
| "cuda_ms": 48.70915222167969, | |
| "name": "transformer_blocks.7" | |
| }, | |
| { | |
| "cuda_ms": 48.56073760986328, | |
| "name": "transformer_blocks.8" | |
| }, | |
| { | |
| "cuda_ms": 48.7092170715332, | |
| "name": "transformer_blocks.9" | |
| }, | |
| { | |
| "cuda_ms": 48.58531188964844, | |
| "name": "transformer_blocks.10" | |
| }, | |
| { | |
| "cuda_ms": 48.66614532470703, | |
| "name": "transformer_blocks.11" | |
| }, | |
| { | |
| "cuda_ms": 48.691776275634766, | |
| "name": "transformer_blocks.12" | |
| }, | |
| { | |
| "cuda_ms": 48.55561447143555, | |
| "name": "transformer_blocks.13" | |
| }, | |
| { | |
| "cuda_ms": 48.70761489868164, | |
| "name": "transformer_blocks.14" | |
| }, | |
| { | |
| "cuda_ms": 48.58272171020508, | |
| "name": "transformer_blocks.15" | |
| }, | |
| { | |
| "cuda_ms": 48.22224044799805, | |
| "name": "transformer_blocks.16" | |
| }, | |
| { | |
| "cuda_ms": 48.6544303894043, | |
| "name": "transformer_blocks.17" | |
| }, | |
| { | |
| "cuda_ms": 48.637569427490234, | |
| "name": "transformer_blocks.18" | |
| }, | |
| { | |
| "cuda_ms": 48.6978874206543, | |
| "name": "transformer_blocks.19" | |
| }, | |
| { | |
| "cuda_ms": 48.58736038208008, | |
| "name": "transformer_blocks.20" | |
| }, | |
| { | |
| "cuda_ms": 48.09062576293945, | |
| "name": "transformer_blocks.21" | |
| }, | |
| { | |
| "cuda_ms": 48.501312255859375, | |
| "name": "transformer_blocks.22" | |
| }, | |
| { | |
| "cuda_ms": 48.043521881103516, | |
| "name": "transformer_blocks.23" | |
| }, | |
| { | |
| "cuda_ms": 48.79324722290039, | |
| "name": "transformer_blocks.24" | |
| }, | |
| { | |
| "cuda_ms": 48.69996643066406, | |
| "name": "transformer_blocks.25" | |
| }, | |
| { | |
| "cuda_ms": 48.07846450805664, | |
| "name": "transformer_blocks.26" | |
| }, | |
| { | |
| "cuda_ms": 48.556575775146484, | |
| "name": "transformer_blocks.27" | |
| }, | |
| { | |
| "cuda_ms": 48.14499282836914, | |
| "name": "transformer_blocks.28" | |
| }, | |
| { | |
| "cuda_ms": 48.63030242919922, | |
| "name": "transformer_blocks.29" | |
| }, | |
| { | |
| "cuda_ms": 48.61299133300781, | |
| "name": "transformer_blocks.30" | |
| }, | |
| { | |
| "cuda_ms": 48.09686279296875, | |
| "name": "transformer_blocks.31" | |
| }, | |
| { | |
| "cuda_ms": 48.560672760009766, | |
| "name": "transformer_blocks.32" | |
| }, | |
| { | |
| "cuda_ms": 48.092735290527344, | |
| "name": "transformer_blocks.33" | |
| }, | |
| { | |
| "cuda_ms": 48.55769729614258, | |
| "name": "transformer_blocks.34" | |
| }, | |
| { | |
| "cuda_ms": 48.0081901550293, | |
| "name": "transformer_blocks.35" | |
| }, | |
| { | |
| "cuda_ms": 48.5715217590332, | |
| "name": "transformer_blocks.36" | |
| }, | |
| { | |
| "cuda_ms": 48.09116744995117, | |
| "name": "transformer_blocks.37" | |
| }, | |
| { | |
| "cuda_ms": 47.97200012207031, | |
| "name": "transformer_blocks.38" | |
| }, | |
| { | |
| "cuda_ms": 48.52592086791992, | |
| "name": "transformer_blocks.39" | |
| }, | |
| { | |
| "cuda_ms": 47.92879867553711, | |
| "name": "transformer_blocks.40" | |
| }, | |
| { | |
| "cuda_ms": 48.56079864501953, | |
| "name": "transformer_blocks.41" | |
| }, | |
| { | |
| "cuda_ms": 48.176639556884766, | |
| "name": "transformer_blocks.42" | |
| }, | |
| { | |
| "cuda_ms": 48.074241638183594, | |
| "name": "transformer_blocks.43" | |
| }, | |
| { | |
| "cuda_ms": 48.551456451416016, | |
| "name": "transformer_blocks.44" | |
| }, | |
| { | |
| "cuda_ms": 47.84288024902344, | |
| "name": "transformer_blocks.45" | |
| }, | |
| { | |
| "cuda_ms": 48.65087890625, | |
| "name": "transformer_blocks.46" | |
| }, | |
| { | |
| "cuda_ms": 47.98934555053711, | |
| "name": "transformer_blocks.47" | |
| }, | |
| { | |
| "cuda_ms": 48.15216064453125, | |
| "name": "transformer_blocks.48" | |
| }, | |
| { | |
| "cuda_ms": 48.495201110839844, | |
| "name": "transformer_blocks.49" | |
| } | |
| ], | |
| "video_tokens": 10545, | |
| "wall_seconds": 2.8687175642699003 | |
| } | |