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": 2889.388916015625, | |
| "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": 0.09053317736834288, | |
| "model_dtype": "torch.bfloat16", | |
| "move_to_cuda_seconds": 1.2328650671988726, | |
| "num_blocks": 50, | |
| "orbitquant_linear_count": 300, | |
| "orbitquant_linear_cuda_ms": 1062.045917749405, | |
| "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": 225.03200006484985 | |
| }, | |
| "5376x28672": { | |
| "calls": 50, | |
| "total_ms": 347.8003191947937 | |
| }, | |
| "5376x7168": { | |
| "calls": 150, | |
| "total_ms": 365.89542269706726 | |
| }, | |
| "7168x5376": { | |
| "calls": 50, | |
| "total_ms": 123.31817579269409 | |
| } | |
| }, | |
| "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": 18.906143188476562, | |
| "name": "transformer_blocks.0.attn.to_q", | |
| "shape": "5376x7168" | |
| }, | |
| { | |
| "cuda_ms": 9.0982084274292, | |
| "name": "transformer_blocks.0.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 8.691136360168457, | |
| "name": "transformer_blocks.0.ff.net.2", | |
| "shape": "14336x5376" | |
| }, | |
| { | |
| "cuda_ms": 7.211008071899414, | |
| "name": "transformer_blocks.0.attn.to_out.0", | |
| "shape": "7168x5376" | |
| }, | |
| { | |
| "cuda_ms": 7.1143999099731445, | |
| "name": "transformer_blocks.6.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.098368167877197, | |
| "name": "transformer_blocks.1.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0820159912109375, | |
| "name": "transformer_blocks.3.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.071263790130615, | |
| "name": "transformer_blocks.47.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0655999183654785, | |
| "name": "transformer_blocks.2.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0641279220581055, | |
| "name": "transformer_blocks.9.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.058080196380615, | |
| "name": "transformer_blocks.24.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.055903911590576, | |
| "name": "transformer_blocks.34.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.050911903381348, | |
| "name": "transformer_blocks.4.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.049727916717529, | |
| "name": "transformer_blocks.19.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.049215793609619, | |
| "name": "transformer_blocks.5.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.046656131744385, | |
| "name": "transformer_blocks.44.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.04585599899292, | |
| "name": "transformer_blocks.7.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0456318855285645, | |
| "name": "transformer_blocks.39.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.0415358543396, | |
| "name": "transformer_blocks.12.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.039487838745117, | |
| "name": "transformer_blocks.21.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.036448001861572, | |
| "name": "transformer_blocks.29.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.036448001861572, | |
| "name": "transformer_blocks.42.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.028736114501953, | |
| "name": "transformer_blocks.26.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 7.02566385269165, | |
| "name": "transformer_blocks.14.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 6.987775802612305, | |
| "name": "transformer_blocks.37.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 6.983263969421387, | |
| "name": "transformer_blocks.32.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 6.972415924072266, | |
| "name": "transformer_blocks.17.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 6.934175968170166, | |
| "name": "transformer_blocks.27.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 6.928383827209473, | |
| "name": "transformer_blocks.22.ff.net.0.proj", | |
| "shape": "5376x28672" | |
| }, | |
| { | |
| "cuda_ms": 6.850048065185547, | |
| "name": "transformer_blocks.10.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": 311.4284362792969, | |
| "name": "transformer_blocks.0" | |
| }, | |
| { | |
| "cuda_ms": 48.51638412475586, | |
| "name": "transformer_blocks.1" | |
| }, | |
| { | |
| "cuda_ms": 48.574527740478516, | |
| "name": "transformer_blocks.2" | |
| }, | |
| { | |
| "cuda_ms": 48.55510330200195, | |
| "name": "transformer_blocks.3" | |
| }, | |
| { | |
| "cuda_ms": 48.56742477416992, | |
| "name": "transformer_blocks.4" | |
| }, | |
| { | |
| "cuda_ms": 48.5162239074707, | |
| "name": "transformer_blocks.5" | |
| }, | |
| { | |
| "cuda_ms": 48.63600158691406, | |
| "name": "transformer_blocks.6" | |
| }, | |
| { | |
| "cuda_ms": 48.58127975463867, | |
| "name": "transformer_blocks.7" | |
| }, | |
| { | |
| "cuda_ms": 48.1910400390625, | |
| "name": "transformer_blocks.8" | |
| }, | |
| { | |
| "cuda_ms": 48.545345306396484, | |
| "name": "transformer_blocks.9" | |
| }, | |
| { | |
| "cuda_ms": 48.06208038330078, | |
| "name": "transformer_blocks.10" | |
| }, | |
| { | |
| "cuda_ms": 47.823455810546875, | |
| "name": "transformer_blocks.11" | |
| }, | |
| { | |
| "cuda_ms": 48.4901123046875, | |
| "name": "transformer_blocks.12" | |
| }, | |
| { | |
| "cuda_ms": 47.92486572265625, | |
| "name": "transformer_blocks.13" | |
| }, | |
| { | |
| "cuda_ms": 48.53456115722656, | |
| "name": "transformer_blocks.14" | |
| }, | |
| { | |
| "cuda_ms": 48.27961730957031, | |
| "name": "transformer_blocks.15" | |
| }, | |
| { | |
| "cuda_ms": 47.93564987182617, | |
| "name": "transformer_blocks.16" | |
| }, | |
| { | |
| "cuda_ms": 48.035457611083984, | |
| "name": "transformer_blocks.17" | |
| }, | |
| { | |
| "cuda_ms": 47.76406478881836, | |
| "name": "transformer_blocks.18" | |
| }, | |
| { | |
| "cuda_ms": 48.521854400634766, | |
| "name": "transformer_blocks.19" | |
| }, | |
| { | |
| "cuda_ms": 48.037376403808594, | |
| "name": "transformer_blocks.20" | |
| }, | |
| { | |
| "cuda_ms": 48.364158630371094, | |
| "name": "transformer_blocks.21" | |
| }, | |
| { | |
| "cuda_ms": 48.13004684448242, | |
| "name": "transformer_blocks.22" | |
| }, | |
| { | |
| "cuda_ms": 47.60380935668945, | |
| "name": "transformer_blocks.23" | |
| }, | |
| { | |
| "cuda_ms": 48.526206970214844, | |
| "name": "transformer_blocks.24" | |
| }, | |
| { | |
| "cuda_ms": 48.04182434082031, | |
| "name": "transformer_blocks.25" | |
| }, | |
| { | |
| "cuda_ms": 47.844512939453125, | |
| "name": "transformer_blocks.26" | |
| }, | |
| { | |
| "cuda_ms": 48.17929458618164, | |
| "name": "transformer_blocks.27" | |
| }, | |
| { | |
| "cuda_ms": 47.58198547363281, | |
| "name": "transformer_blocks.28" | |
| }, | |
| { | |
| "cuda_ms": 48.438785552978516, | |
| "name": "transformer_blocks.29" | |
| }, | |
| { | |
| "cuda_ms": 48.02822494506836, | |
| "name": "transformer_blocks.30" | |
| }, | |
| { | |
| "cuda_ms": 46.90595245361328, | |
| "name": "transformer_blocks.31" | |
| }, | |
| { | |
| "cuda_ms": 48.04880142211914, | |
| "name": "transformer_blocks.32" | |
| }, | |
| { | |
| "cuda_ms": 47.76803207397461, | |
| "name": "transformer_blocks.33" | |
| }, | |
| { | |
| "cuda_ms": 48.554656982421875, | |
| "name": "transformer_blocks.34" | |
| }, | |
| { | |
| "cuda_ms": 47.988670349121094, | |
| "name": "transformer_blocks.35" | |
| }, | |
| { | |
| "cuda_ms": 47.18297576904297, | |
| "name": "transformer_blocks.36" | |
| }, | |
| { | |
| "cuda_ms": 48.055423736572266, | |
| "name": "transformer_blocks.37" | |
| }, | |
| { | |
| "cuda_ms": 47.75484848022461, | |
| "name": "transformer_blocks.38" | |
| }, | |
| { | |
| "cuda_ms": 48.46246337890625, | |
| "name": "transformer_blocks.39" | |
| }, | |
| { | |
| "cuda_ms": 47.13225555419922, | |
| "name": "transformer_blocks.40" | |
| }, | |
| { | |
| "cuda_ms": 47.67193603515625, | |
| "name": "transformer_blocks.41" | |
| }, | |
| { | |
| "cuda_ms": 47.967838287353516, | |
| "name": "transformer_blocks.42" | |
| }, | |
| { | |
| "cuda_ms": 47.16921615600586, | |
| "name": "transformer_blocks.43" | |
| }, | |
| { | |
| "cuda_ms": 47.7097282409668, | |
| "name": "transformer_blocks.44" | |
| }, | |
| { | |
| "cuda_ms": 47.17932891845703, | |
| "name": "transformer_blocks.45" | |
| }, | |
| { | |
| "cuda_ms": 47.825439453125, | |
| "name": "transformer_blocks.46" | |
| }, | |
| { | |
| "cuda_ms": 48.40099334716797, | |
| "name": "transformer_blocks.47" | |
| }, | |
| { | |
| "cuda_ms": 47.14352035522461, | |
| "name": "transformer_blocks.48" | |
| }, | |
| { | |
| "cuda_ms": 46.874176025390625, | |
| "name": "transformer_blocks.49" | |
| } | |
| ], | |
| "video_tokens": 10545, | |
| "wall_seconds": 2.890391333028674 | |
| } | |