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": 5536.7392578125, | |
| "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": "default", | |
| "load_seconds": 0.17170354118570685, | |
| "model_dtype": "torch.bfloat16", | |
| "move_to_cuda_seconds": 1.235624415334314, | |
| "num_blocks": 50, | |
| "orbitquant_linear_count": 300, | |
| "orbitquant_linear_cuda_ms": 3368.3247916698456, | |
| "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": 658.479362487793 | |
| }, | |
| "5376x28672": { | |
| "calls": 50, | |
| "total_ms": 493.38377046585083 | |
| }, | |
| "5376x7168": { | |
| "calls": 150, | |
| "total_ms": 1740.172526359558 | |
| }, | |
| "7168x5376": { | |
| "calls": 50, | |
| "total_ms": 476.2891323566437 | |
| } | |
| }, | |
| "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": 1387.7381591796875, | |
| "name": "transformer_blocks.0.attn.to_q", | |
| "shape": "5376x7168" | |
| }, | |
| { | |
| "cuda_ms": 439.46435546875, | |
| "name": "transformer_blocks.0.ff.net.2", | |
| "shape": "14336x5376" | |
| }, | |
| { | |
| "cuda_ms": 359.4147644042969, | |
| "name": "transformer_blocks.0.attn.to_out.0", | |
| "shape": "7168x5376" | |
| }, | |
| { | |
| "cuda_ms": 152.1229705810547, | |
| "name": "transformer_blocks.0.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.239264011383057, | |
| "name": "transformer_blocks.12.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.144959926605225, | |
| "name": "transformer_blocks.9.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.108128070831299, | |
| "name": "transformer_blocks.19.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.105184078216553, | |
| "name": "transformer_blocks.7.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0979838371276855, | |
| "name": "transformer_blocks.4.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0952959060668945, | |
| "name": "transformer_blocks.6.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.088128089904785, | |
| "name": "transformer_blocks.14.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.078400135040283, | |
| "name": "transformer_blocks.20.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.073760032653809, | |
| "name": "transformer_blocks.15.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.069248199462891, | |
| "name": "transformer_blocks.13.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.067359924316406, | |
| "name": "transformer_blocks.27.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.066656112670898, | |
| "name": "transformer_blocks.1.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.066112041473389, | |
| "name": "transformer_blocks.46.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.065248012542725, | |
| "name": "transformer_blocks.34.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.061535835266113, | |
| "name": "transformer_blocks.2.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0609917640686035, | |
| "name": "transformer_blocks.10.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.057951927185059, | |
| "name": "transformer_blocks.32.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.054912090301514, | |
| "name": "transformer_blocks.44.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.054336071014404, | |
| "name": "transformer_blocks.36.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.053855895996094, | |
| "name": "transformer_blocks.39.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.053311824798584, | |
| "name": "transformer_blocks.3.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.05299186706543, | |
| "name": "transformer_blocks.24.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.052864074707031, | |
| "name": "transformer_blocks.29.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0502400398254395, | |
| "name": "transformer_blocks.22.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.048736095428467, | |
| "name": "transformer_blocks.41.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.04307222366333, | |
| "name": "transformer_blocks.17.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| } | |
| ], | |
| "status": "pass", | |
| "text_tokens": 512, | |
| "torch": "2.8.0+cu128", | |
| "transformer": "/workspace/orbitquant-h3-ref/model/transformer_ref", | |
| "transformer_block_cuda_ms": [ | |
| { | |
| "cuda_ms": 2941.07666015625, | |
| "name": "transformer_blocks.0" | |
| }, | |
| { | |
| "cuda_ms": 48.6420783996582, | |
| "name": "transformer_blocks.1" | |
| }, | |
| { | |
| "cuda_ms": 48.49164962768555, | |
| "name": "transformer_blocks.2" | |
| }, | |
| { | |
| "cuda_ms": 48.473182678222656, | |
| "name": "transformer_blocks.3" | |
| }, | |
| { | |
| "cuda_ms": 48.64112091064453, | |
| "name": "transformer_blocks.4" | |
| }, | |
| { | |
| "cuda_ms": 48.53055953979492, | |
| "name": "transformer_blocks.5" | |
| }, | |
| { | |
| "cuda_ms": 48.63804626464844, | |
| "name": "transformer_blocks.6" | |
| }, | |
| { | |
| "cuda_ms": 48.6923828125, | |
| "name": "transformer_blocks.7" | |
| }, | |
| { | |
| "cuda_ms": 48.50038528442383, | |
| "name": "transformer_blocks.8" | |
| }, | |
| { | |
| "cuda_ms": 48.72953414916992, | |
| "name": "transformer_blocks.9" | |
| }, | |
| { | |
| "cuda_ms": 48.53916931152344, | |
| "name": "transformer_blocks.10" | |
| }, | |
| { | |
| "cuda_ms": 48.08662414550781, | |
| "name": "transformer_blocks.11" | |
| }, | |
| { | |
| "cuda_ms": 48.662113189697266, | |
| "name": "transformer_blocks.12" | |
| }, | |
| { | |
| "cuda_ms": 48.73068618774414, | |
| "name": "transformer_blocks.13" | |
| }, | |
| { | |
| "cuda_ms": 48.63391876220703, | |
| "name": "transformer_blocks.14" | |
| }, | |
| { | |
| "cuda_ms": 48.57455825805664, | |
| "name": "transformer_blocks.15" | |
| }, | |
| { | |
| "cuda_ms": 48.3389778137207, | |
| "name": "transformer_blocks.16" | |
| }, | |
| { | |
| "cuda_ms": 48.74409484863281, | |
| "name": "transformer_blocks.17" | |
| }, | |
| { | |
| "cuda_ms": 48.08147048950195, | |
| "name": "transformer_blocks.18" | |
| }, | |
| { | |
| "cuda_ms": 48.471649169921875, | |
| "name": "transformer_blocks.19" | |
| }, | |
| { | |
| "cuda_ms": 48.72867202758789, | |
| "name": "transformer_blocks.20" | |
| }, | |
| { | |
| "cuda_ms": 48.13267135620117, | |
| "name": "transformer_blocks.21" | |
| }, | |
| { | |
| "cuda_ms": 48.483360290527344, | |
| "name": "transformer_blocks.22" | |
| }, | |
| { | |
| "cuda_ms": 48.04719924926758, | |
| "name": "transformer_blocks.23" | |
| }, | |
| { | |
| "cuda_ms": 48.55311965942383, | |
| "name": "transformer_blocks.24" | |
| }, | |
| { | |
| "cuda_ms": 48.2775993347168, | |
| "name": "transformer_blocks.25" | |
| }, | |
| { | |
| "cuda_ms": 47.991905212402344, | |
| "name": "transformer_blocks.26" | |
| }, | |
| { | |
| "cuda_ms": 48.53667068481445, | |
| "name": "transformer_blocks.27" | |
| }, | |
| { | |
| "cuda_ms": 48.02115249633789, | |
| "name": "transformer_blocks.28" | |
| }, | |
| { | |
| "cuda_ms": 48.52803039550781, | |
| "name": "transformer_blocks.29" | |
| }, | |
| { | |
| "cuda_ms": 48.025054931640625, | |
| "name": "transformer_blocks.30" | |
| }, | |
| { | |
| "cuda_ms": 47.83679962158203, | |
| "name": "transformer_blocks.31" | |
| }, | |
| { | |
| "cuda_ms": 48.23612976074219, | |
| "name": "transformer_blocks.32" | |
| }, | |
| { | |
| "cuda_ms": 49.4167366027832, | |
| "name": "transformer_blocks.33" | |
| }, | |
| { | |
| "cuda_ms": 48.58224105834961, | |
| "name": "transformer_blocks.34" | |
| }, | |
| { | |
| "cuda_ms": 48.018463134765625, | |
| "name": "transformer_blocks.35" | |
| }, | |
| { | |
| "cuda_ms": 47.99913787841797, | |
| "name": "transformer_blocks.36" | |
| }, | |
| { | |
| "cuda_ms": 48.30006408691406, | |
| "name": "transformer_blocks.37" | |
| }, | |
| { | |
| "cuda_ms": 47.80400085449219, | |
| "name": "transformer_blocks.38" | |
| }, | |
| { | |
| "cuda_ms": 48.554622650146484, | |
| "name": "transformer_blocks.39" | |
| }, | |
| { | |
| "cuda_ms": 48.00057601928711, | |
| "name": "transformer_blocks.40" | |
| }, | |
| { | |
| "cuda_ms": 48.38880157470703, | |
| "name": "transformer_blocks.41" | |
| }, | |
| { | |
| "cuda_ms": 47.99033737182617, | |
| "name": "transformer_blocks.42" | |
| }, | |
| { | |
| "cuda_ms": 47.718048095703125, | |
| "name": "transformer_blocks.43" | |
| }, | |
| { | |
| "cuda_ms": 48.561824798583984, | |
| "name": "transformer_blocks.44" | |
| }, | |
| { | |
| "cuda_ms": 48.02803039550781, | |
| "name": "transformer_blocks.45" | |
| }, | |
| { | |
| "cuda_ms": 48.63555145263672, | |
| "name": "transformer_blocks.46" | |
| }, | |
| { | |
| "cuda_ms": 47.13932800292969, | |
| "name": "transformer_blocks.47" | |
| }, | |
| { | |
| "cuda_ms": 47.67804718017578, | |
| "name": "transformer_blocks.48" | |
| }, | |
| { | |
| "cuda_ms": 48.166561126708984, | |
| "name": "transformer_blocks.49" | |
| } | |
| ], | |
| "video_tokens": 10545, | |
| "wall_seconds": 5.5382678848691285 | |
| } | |