| { |
| "format_version": 1, |
| "artifact_type": "compressingvlm_w8a8_smoothquant_bundle", |
| "base_model": "/content/compressingvlm_nuextract2/artifacts/merged_full_checkpoint", |
| "dataset_id": "letxbe/BoundingDocs", |
| "smoothquant_alpha": 0.5, |
| "converted_layers": 327, |
| "skipped_layers": 0, |
| "quant_meta": { |
| "model.visual.blocks.0.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.011774783059367983, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.0.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.004796601186587116, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.0.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.020350467501662847, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.0.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.01206465593473179, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.1.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.013214042806249904, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.1.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.00883366930203175, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.1.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.0195614765948198, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.1.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.006876293599136232, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.2.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.016577895232072964, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.2.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.0066642273129440665, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.2.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.010180345670444759, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.2.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.007410509849157859, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.3.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.020379057080726924, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.3.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006451350028120626, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.3.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.010919001158766858, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.3.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.010403279244430422, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.4.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014471732725308635, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.4.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.007603761248701201, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.4.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.011414487530866007, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.4.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.008864887117400883, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.5.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.009759143581540565, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.5.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.005876585723846916, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.5.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.0104806751716794, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.5.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.0067109634557108245, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.6.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.011332700571675939, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.6.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.005560288279075322, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.6.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.014120476452384408, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.6.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.009994748070484071, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.7.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.013232500534357988, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.7.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.005875116258155642, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.7.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.013313860405148484, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.7.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.008737835358446978, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.8.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.021730391059334823, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.8.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.010192882357619878, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.8.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.016515345085324266, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.8.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.009404166476933037, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.9.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014161588638786256, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.9.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.005805544027193325, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.9.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.0160279217667467, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.9.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.012330186648631659, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.10.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.01268474323543038, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.10.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.005399912830412857, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.10.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.015144372549582653, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.10.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.012361452335447777, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.11.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015269914949972799, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.11.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006234276482439417, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.11.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.014402702098756324, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.11.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.01632130615354523, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.12.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014100967429754302, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.12.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006146921416905921, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.12.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.016758920639518678, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.12.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.01607548533462164, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.13.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.01547414486802469, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.13.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.0061614091002096345, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.13.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.01329935817267951, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.13.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.0183927956528551, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.14.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.016157283557681586, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.14.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.0056524656889006845, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.14.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.014731704719423309, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.14.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.0164302334072083, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.15.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015307413311455194, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.15.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006376280559329536, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.15.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.015562512743191457, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.15.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.025342841786662424, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.16.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015025780895563562, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.16.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.005419237407173697, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.16.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.016547708060797743, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.16.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.018387858323224888, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.17.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.01265057522480882, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.17.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006702738484059732, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.17.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.02144298027819536, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.17.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.02630904340368556, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.18.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014642170095068264, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.18.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.00903098601994552, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.18.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.019541541422445942, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.18.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.04239599348053219, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.19.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014888684580645224, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.19.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006763935089111328, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.19.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.01819218237569013, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.19.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.01527449746770183, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.20.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014723354437219815, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.20.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006135502199488362, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.20.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.019845838621845396, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.20.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.013107329841673844, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.21.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.013580497794263945, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.21.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.012507303493229423, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.21.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.018828945835744303, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.21.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.019425249475193775, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.22.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.014368930200892171, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.22.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.008607041178725836, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.22.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.020806425199733945, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.22.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.0203628277215432, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.23.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015795058152807042, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.23.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006380434111347349, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.23.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.023131884927824725, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.23.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.025056778915285124, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.24.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.016849420202060008, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.24.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.007182589196783351, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.24.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.02979039394949365, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.24.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.02414914003507359, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.25.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015637225053441808, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.25.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.00804293437266913, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.25.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.0279038853532686, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.25.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.019783076338880642, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.26.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015581223908371813, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.26.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.008657125037486158, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.26.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.0347223582230215, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.26.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.019559443466306673, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.27.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.015775372662882165, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.27.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.006993105561714473, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.27.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.0269776197869008, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.27.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.024150317109475926, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.28.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.01766562649584192, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.28.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.008466911128186804, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.28.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.03087514404236801, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.28.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.026798460427231677, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.29.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.017204213330126183, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.29.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.01030280834107887, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.29.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.03364264495729461, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.29.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.039391221023920016, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.30.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.01684388964194951, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.30.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.013944210968618318, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.30.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.026909640454870508, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.30.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.032339970896563194, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.31.attn.qkv": { |
| "in_features": 1280, |
| "out_features": 3840, |
| "act_scale": 0.013929445912518839, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.31.attn.proj": { |
| "in_features": 1280, |
| "out_features": 1280, |
| "act_scale": 0.01802794576629879, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.31.mlp.fc1": { |
| "in_features": 1280, |
| "out_features": 5120, |
| "act_scale": 0.01583776699276421, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.blocks.31.mlp.fc2": { |
| "in_features": 5120, |
| "out_features": 1280, |
| "act_scale": 0.16578437024214138, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.merger.mlp.0": { |
| "in_features": 5120, |
| "out_features": 5120, |
| "act_scale": 0.009428932910829078, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.visual.merger.mlp.2": { |
| "in_features": 5120, |
| "out_features": 1536, |
| "act_scale": 0.015239587918979915, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.038241622954841674, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.019962245085107997, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.008818717453423448, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.0067325891472223235, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.011143252605528344, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.008632640200337088, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.0.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.00833494362868662, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.015791344830370323, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.012920356172276295, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.007185041904449463, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.006423686902354083, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014584682119174266, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.021518359972736028, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.1.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.2898354718065637, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.019355892196414978, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01587051293981357, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.006699156104110357, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.006997298067948949, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014152612273148664, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.015616326820193312, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.2.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.025589364720141793, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.016779912738349494, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.014279944690193717, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.007599234111665741, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.008523108452323853, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.015464061827171506, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.021024826004749208, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.3.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.057624861950010764, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.016623001399002676, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.013275729389641229, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.008648515686275452, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007065747197218767, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.015994242795809046, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.016403669447410764, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.4.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.01744353865075299, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.018114063683457263, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.013239755405215766, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.009236081378666435, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.0064572513572813016, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.018577444271778498, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014611972598578986, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.5.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.011265405519740788, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.017126639058270793, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.014979274254145586, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.007247885381142924, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.0068520542204849365, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.0188901424407959, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.0178241279181533, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.6.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.01482172275152732, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.01707872255580632, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.012455570416187677, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.011186249612823246, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007465832346067654, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.017271993667121947, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.012164750437098226, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.7.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.02268905527009739, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.01781220698919822, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.011503820344219057, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01285932852527288, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007614903093322994, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.017289646028533695, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.013838967000405619, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.8.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.03071435417715959, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.0197468404694805, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016635203924704724, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.009908878897118755, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.009123546870674674, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.017440991138848734, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.015315322425421767, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.9.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.01320920966741607, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.019672572143434538, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.015871153103085015, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.00930871738223579, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.008404247404083492, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014881523575369768, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014982253547728532, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.10.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.0114670991897583, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.01762019555399737, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016357397469948597, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.010812059162169929, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.008168115390567329, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014244061755383109, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.010984372904920202, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.11.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.012976615447697676, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.015788528862900622, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016916130471417285, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.009433514489902286, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.009469956863583543, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014436105104881948, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014943877543051412, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.12.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.010102545182535968, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.017111718185304655, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016115950787161278, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.008848630537198284, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.008366376396239274, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.013314998994662068, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.013658712229390782, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.13.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.012686569859662393, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.015640403342059277, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01659770274725486, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01202462789580578, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.006960612582409475, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014151355413001353, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.01234702707275631, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.14.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.015763763367660402, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.017084838837150514, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016725386221577803, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.008186301847142497, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007004190617658961, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.020452302271925557, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.01974195945919968, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.15.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.01462966156756784, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.02654338633920264, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.017971128929318406, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.007890122143302376, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.009483855540358176, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.015570687496756006, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014243651562788355, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.16.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.014674472057913232, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.017570152057437446, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.02153455366299847, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.012271659580741341, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007414881169326662, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.019823444171214667, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.017401490624495378, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.17.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.013188638086394063, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.018093604741134042, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01484214977955255, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.012393878200861413, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007763971963266688, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.018470108978391634, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.012738921510891652, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.18.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.016149494591660388, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.018284270143884375, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01721131895470807, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.009106295315299447, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.0067378348252904695, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.019664180560374823, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.014785750644413505, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.19.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.017295666566983923, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.0167858112515427, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.021686362469290184, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.009510900091937207, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.006677778687064103, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.018273759075975792, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.01493667993019885, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.20.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.016903460495115267, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.016047697367630606, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.018048720096978617, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.011401965862184059, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.007243935986766665, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.01880555265531765, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.017204089427557516, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.21.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.02033349097244383, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.017547941583348072, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.017770733420304427, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.012843428634283111, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.009375796543331597, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.01807523149205005, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.016792299240592896, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.22.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.02282524484349048, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.017924070358276367, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.018123487787922537, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01064312739635077, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.008809326201911986, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.018356037890817238, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.016144844490712084, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.23.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.020913634713240495, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.020576118484256774, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016015707977174772, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.010398974568825069, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.006741659378442239, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.01695194582300862, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.016808357764416793, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.24.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.02610544332369106, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.01670930517001415, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.016933991214421792, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01482135104382132, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.010464924526965524, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.03200663168599287, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.016040488490908163, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.25.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.03664320097194882, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.019067401961078794, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01596015081630917, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.015354877381812869, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.010470243889515795, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.024390772571713907, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.024410642038180132, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.26.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.2042921621968427, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.self_attn.q_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.016146177381981076, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.self_attn.k_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.01712871348764014, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.self_attn.v_proj": { |
| "in_features": 1536, |
| "out_features": 256, |
| "act_scale": 0.013760393060098483, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.self_attn.o_proj": { |
| "in_features": 1536, |
| "out_features": 1536, |
| "act_scale": 0.014997026113074595, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.mlp.gate_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.026463501096710446, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.mlp.up_proj": { |
| "in_features": 1536, |
| "out_features": 8960, |
| "act_scale": 0.03419333931029312, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "model.language_model.layers.27.mlp.down_proj": { |
| "in_features": 8960, |
| "out_features": 1536, |
| "act_scale": 0.07352166663943313, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| }, |
| "lm_head": { |
| "in_features": 1536, |
| "out_features": 151936, |
| "act_scale": 0.019375630251065954, |
| "smoothquant_alpha": 0.5, |
| "out_dtype": "torch.float16" |
| } |
| }, |
| "self_contained_bundle": true, |
| "requires_base_model_download": false |
| } |
