{ "time_text_embed.timestep_embedder.linear_1": {}, "time_text_embed.timestep_embedder.linear_2": {}, "time_text_embed.guidance_embedder.linear_1": {}, "time_text_embed.guidance_embedder.linear_2": {}, "time_text_embed.text_embedder.linear_1": {}, "time_text_embed.text_embedder.linear_2": {}, "context_embedder": {}, "x_embedder": {}, "transformer_blocks.0.norm1.linear": {}, "transformer_blocks.0.norm1_context.linear": {}, "transformer_blocks.0.attn.to_q": {}, "transformer_blocks.0.attn.to_k": {}, "transformer_blocks.0.attn.to_v": {}, "transformer_blocks.0.attn.add_k_proj": {}, "transformer_blocks.0.attn.add_v_proj": {}, "transformer_blocks.0.attn.add_q_proj": {}, "transformer_blocks.0.attn.to_out.0": {}, "transformer_blocks.0.attn.to_add_out": {}, "transformer_blocks.0.attn.to_qkv": {}, "transformer_blocks.0.attn.to_added_qkv": {}, "transformer_blocks.0.attn.output_softmax_quant": { "act_scale": 0.0030924479942768812, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.0.attn.out_q": { "act_scale": 0.1770833283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.0.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.0.attn.out_v": { "act_scale": 0.0520833320915699, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.0.ff.net.0.proj": {}, "transformer_blocks.0.ff.net.2": {}, "transformer_blocks.0.ff_context.net.0.proj": {}, "transformer_blocks.0.ff_context.net.2": {}, "transformer_blocks.1.norm1.linear": {}, "transformer_blocks.1.norm1_context.linear": {}, "transformer_blocks.1.attn.to_q": {}, "transformer_blocks.1.attn.to_k": {}, "transformer_blocks.1.attn.to_v": {}, "transformer_blocks.1.attn.add_k_proj": {}, "transformer_blocks.1.attn.add_v_proj": {}, "transformer_blocks.1.attn.add_q_proj": {}, "transformer_blocks.1.attn.to_out.0": {}, "transformer_blocks.1.attn.to_add_out": {}, "transformer_blocks.1.attn.to_qkv": {}, "transformer_blocks.1.attn.to_added_qkv": {}, "transformer_blocks.1.attn.output_softmax_quant": { "act_scale": 0.0030273436568677425, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.1.attn.out_q": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.1.attn.out_k": { "act_scale": 0.11041666567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.1.attn.out_v": { "act_scale": 0.109375, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.1.ff.net.0.proj": {}, "transformer_blocks.1.ff.net.2": {}, "transformer_blocks.1.ff_context.net.0.proj": {}, "transformer_blocks.1.ff_context.net.2": {}, "transformer_blocks.2.norm1.linear": {}, "transformer_blocks.2.norm1_context.linear": {}, "transformer_blocks.2.attn.to_q": {}, "transformer_blocks.2.attn.to_k": {}, "transformer_blocks.2.attn.to_v": {}, "transformer_blocks.2.attn.add_k_proj": {}, "transformer_blocks.2.attn.add_v_proj": {}, "transformer_blocks.2.attn.add_q_proj": {}, "transformer_blocks.2.attn.to_out.0": {}, "transformer_blocks.2.attn.to_add_out": {}, "transformer_blocks.2.attn.to_qkv": {}, "transformer_blocks.2.attn.to_added_qkv": {}, "transformer_blocks.2.attn.output_softmax_quant": { "act_scale": 0.0034993488807231188, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.2.attn.out_q": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.2.attn.out_k": { "act_scale": 0.08020833134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.2.attn.out_v": { "act_scale": 0.05416666716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.2.ff.net.0.proj": {}, "transformer_blocks.2.ff.net.2": {}, "transformer_blocks.2.ff_context.net.0.proj": {}, "transformer_blocks.2.ff_context.net.2": {}, "transformer_blocks.3.norm1.linear": {}, "transformer_blocks.3.norm1_context.linear": {}, "transformer_blocks.3.attn.to_q": {}, "transformer_blocks.3.attn.to_k": {}, "transformer_blocks.3.attn.to_v": {}, "transformer_blocks.3.attn.add_k_proj": {}, "transformer_blocks.3.attn.add_v_proj": {}, "transformer_blocks.3.attn.add_q_proj": {}, "transformer_blocks.3.attn.to_out.0": {}, "transformer_blocks.3.attn.to_add_out": {}, "transformer_blocks.3.attn.to_qkv": {}, "transformer_blocks.3.attn.to_added_qkv": {}, "transformer_blocks.3.attn.output_softmax_quant": { "act_scale": 0.003971354104578495, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.3.attn.out_q": { "act_scale": 0.12395833432674408, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.3.attn.out_k": { "act_scale": 0.09375, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.3.attn.out_v": { "act_scale": 0.04322916641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.3.ff.net.0.proj": {}, "transformer_blocks.3.ff.net.2": {}, "transformer_blocks.3.ff_context.net.0.proj": {}, "transformer_blocks.3.ff_context.net.2": {}, "transformer_blocks.4.norm1.linear": {}, "transformer_blocks.4.norm1_context.linear": {}, "transformer_blocks.4.attn.to_q": {}, "transformer_blocks.4.attn.to_k": {}, "transformer_blocks.4.attn.to_v": {}, "transformer_blocks.4.attn.add_k_proj": {}, "transformer_blocks.4.attn.add_v_proj": {}, "transformer_blocks.4.attn.add_q_proj": {}, "transformer_blocks.4.attn.to_out.0": {}, "transformer_blocks.4.attn.to_add_out": {}, "transformer_blocks.4.attn.to_qkv": {}, "transformer_blocks.4.attn.to_added_qkv": {}, "transformer_blocks.4.attn.output_softmax_quant": { "act_scale": 0.004101562313735485, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.4.attn.out_q": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.4.attn.out_k": { "act_scale": 0.10572917014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.4.attn.out_v": { "act_scale": 0.05156249925494194, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.4.ff.net.0.proj": {}, "transformer_blocks.4.ff.net.2": {}, "transformer_blocks.4.ff_context.net.0.proj": {}, "transformer_blocks.4.ff_context.net.2": {}, "transformer_blocks.5.norm1.linear": {}, "transformer_blocks.5.norm1_context.linear": {}, "transformer_blocks.5.attn.to_q": {}, "transformer_blocks.5.attn.to_k": {}, "transformer_blocks.5.attn.to_v": {}, "transformer_blocks.5.attn.add_k_proj": {}, "transformer_blocks.5.attn.add_v_proj": {}, "transformer_blocks.5.attn.add_q_proj": {}, "transformer_blocks.5.attn.to_out.0": {}, "transformer_blocks.5.attn.to_add_out": {}, "transformer_blocks.5.attn.to_qkv": {}, "transformer_blocks.5.attn.to_added_qkv": {}, "transformer_blocks.5.attn.output_softmax_quant": { "act_scale": 0.0038736979477107525, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.5.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.5.attn.out_k": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.5.attn.out_v": { "act_scale": 0.04713541641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.5.ff.net.0.proj": {}, "transformer_blocks.5.ff.net.2": {}, "transformer_blocks.5.ff_context.net.0.proj": {}, "transformer_blocks.5.ff_context.net.2": {}, "transformer_blocks.6.norm1.linear": {}, "transformer_blocks.6.norm1_context.linear": {}, "transformer_blocks.6.attn.to_q": {}, "transformer_blocks.6.attn.to_k": {}, "transformer_blocks.6.attn.to_v": {}, "transformer_blocks.6.attn.add_k_proj": {}, "transformer_blocks.6.attn.add_v_proj": {}, "transformer_blocks.6.attn.add_q_proj": {}, "transformer_blocks.6.attn.to_out.0": {}, "transformer_blocks.6.attn.to_add_out": {}, "transformer_blocks.6.attn.to_qkv": {}, "transformer_blocks.6.attn.to_added_qkv": {}, "transformer_blocks.6.attn.output_softmax_quant": { "act_scale": 0.0037434895057231188, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.6.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.6.attn.out_k": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.6.attn.out_v": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.6.ff.net.0.proj": {}, "transformer_blocks.6.ff.net.2": {}, "transformer_blocks.6.ff_context.net.0.proj": {}, "transformer_blocks.6.ff_context.net.2": {}, "transformer_blocks.7.norm1.linear": {}, "transformer_blocks.7.norm1_context.linear": {}, "transformer_blocks.7.attn.to_q": {}, "transformer_blocks.7.attn.to_k": {}, "transformer_blocks.7.attn.to_v": {}, "transformer_blocks.7.attn.add_k_proj": {}, "transformer_blocks.7.attn.add_v_proj": {}, "transformer_blocks.7.attn.add_q_proj": {}, "transformer_blocks.7.attn.to_out.0": {}, "transformer_blocks.7.attn.to_add_out": {}, "transformer_blocks.7.attn.to_qkv": {}, "transformer_blocks.7.attn.to_added_qkv": {}, "transformer_blocks.7.attn.output_softmax_quant": { "act_scale": 0.00390625, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.7.attn.out_q": { "act_scale": 0.10104166716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.7.attn.out_k": { "act_scale": 0.15729166567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.7.attn.out_v": { "act_scale": 0.06640625, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.7.ff.net.0.proj": {}, "transformer_blocks.7.ff.net.2": {}, "transformer_blocks.7.ff_context.net.0.proj": {}, "transformer_blocks.7.ff_context.net.2": {}, "transformer_blocks.8.norm1.linear": {}, "transformer_blocks.8.norm1_context.linear": {}, "transformer_blocks.8.attn.to_q": {}, "transformer_blocks.8.attn.to_k": {}, "transformer_blocks.8.attn.to_v": {}, "transformer_blocks.8.attn.add_k_proj": {}, "transformer_blocks.8.attn.add_v_proj": {}, "transformer_blocks.8.attn.add_q_proj": {}, "transformer_blocks.8.attn.to_out.0": {}, "transformer_blocks.8.attn.to_add_out": {}, "transformer_blocks.8.attn.to_qkv": {}, "transformer_blocks.8.attn.to_added_qkv": {}, "transformer_blocks.8.attn.output_softmax_quant": { "act_scale": 0.003841145895421505, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.8.attn.out_q": { "act_scale": 0.08385416865348816, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.8.attn.out_k": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.8.attn.out_v": { "act_scale": 0.05494791641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.8.ff.net.0.proj": {}, "transformer_blocks.8.ff.net.2": {}, "transformer_blocks.8.ff_context.net.0.proj": {}, "transformer_blocks.8.ff_context.net.2": {}, "transformer_blocks.9.norm1.linear": {}, "transformer_blocks.9.norm1_context.linear": {}, "transformer_blocks.9.attn.to_q": {}, "transformer_blocks.9.attn.to_k": {}, "transformer_blocks.9.attn.to_v": {}, "transformer_blocks.9.attn.add_k_proj": {}, "transformer_blocks.9.attn.add_v_proj": {}, "transformer_blocks.9.attn.add_q_proj": {}, "transformer_blocks.9.attn.to_out.0": {}, "transformer_blocks.9.attn.to_add_out": {}, "transformer_blocks.9.attn.to_qkv": {}, "transformer_blocks.9.attn.to_added_qkv": {}, "transformer_blocks.9.attn.output_softmax_quant": { "act_scale": 0.0035807292442768812, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.9.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.9.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.9.attn.out_v": { "act_scale": 0.05651041492819786, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.9.ff.net.0.proj": {}, "transformer_blocks.9.ff.net.2": {}, "transformer_blocks.9.ff_context.net.0.proj": {}, "transformer_blocks.9.ff_context.net.2": {}, "transformer_blocks.10.norm1.linear": {}, "transformer_blocks.10.norm1_context.linear": {}, "transformer_blocks.10.attn.to_q": {}, "transformer_blocks.10.attn.to_k": {}, "transformer_blocks.10.attn.to_v": {}, "transformer_blocks.10.attn.add_k_proj": {}, "transformer_blocks.10.attn.add_v_proj": {}, "transformer_blocks.10.attn.add_q_proj": {}, "transformer_blocks.10.attn.to_out.0": {}, "transformer_blocks.10.attn.to_add_out": {}, "transformer_blocks.10.attn.to_qkv": {}, "transformer_blocks.10.attn.to_added_qkv": {}, "transformer_blocks.10.attn.output_softmax_quant": { "act_scale": 0.0028483073692768812, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.10.attn.out_q": { "act_scale": 0.11145833134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.10.attn.out_k": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.10.attn.out_v": { "act_scale": 0.06328125298023224, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.10.ff.net.0.proj": {}, "transformer_blocks.10.ff.net.2": {}, "transformer_blocks.10.ff_context.net.0.proj": {}, "transformer_blocks.10.ff_context.net.2": {}, "transformer_blocks.11.norm1.linear": {}, "transformer_blocks.11.norm1_context.linear": {}, "transformer_blocks.11.attn.to_q": {}, "transformer_blocks.11.attn.to_k": {}, "transformer_blocks.11.attn.to_v": {}, "transformer_blocks.11.attn.add_k_proj": {}, "transformer_blocks.11.attn.add_v_proj": {}, "transformer_blocks.11.attn.add_q_proj": {}, "transformer_blocks.11.attn.to_out.0": {}, "transformer_blocks.11.attn.to_add_out": {}, "transformer_blocks.11.attn.to_qkv": {}, "transformer_blocks.11.attn.to_added_qkv": {}, "transformer_blocks.11.attn.output_softmax_quant": { "act_scale": 0.003597005270421505, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.11.attn.out_q": { "act_scale": 0.09947916865348816, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.11.attn.out_k": { "act_scale": 0.16458334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.11.attn.out_v": { "act_scale": 0.05885416641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.11.ff.net.0.proj": {}, "transformer_blocks.11.ff.net.2": {}, "transformer_blocks.11.ff_context.net.0.proj": {}, "transformer_blocks.11.ff_context.net.2": {}, "transformer_blocks.12.norm1.linear": {}, "transformer_blocks.12.norm1_context.linear": {}, "transformer_blocks.12.attn.to_q": {}, "transformer_blocks.12.attn.to_k": {}, "transformer_blocks.12.attn.to_v": {}, "transformer_blocks.12.attn.add_k_proj": {}, "transformer_blocks.12.attn.add_v_proj": {}, "transformer_blocks.12.attn.add_q_proj": {}, "transformer_blocks.12.attn.to_out.0": {}, "transformer_blocks.12.attn.to_add_out": {}, "transformer_blocks.12.attn.to_qkv": {}, "transformer_blocks.12.attn.to_added_qkv": {}, "transformer_blocks.12.attn.output_softmax_quant": { "act_scale": 0.0038085938431322575, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.12.attn.out_q": { "act_scale": 0.08697916567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.12.attn.out_k": { "act_scale": 0.08749999850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.12.attn.out_v": { "act_scale": 0.0598958320915699, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.12.ff.net.0.proj": {}, "transformer_blocks.12.ff.net.2": {}, "transformer_blocks.12.ff_context.net.0.proj": {}, "transformer_blocks.12.ff_context.net.2": {}, "transformer_blocks.13.norm1.linear": {}, "transformer_blocks.13.norm1_context.linear": {}, "transformer_blocks.13.attn.to_q": {}, "transformer_blocks.13.attn.to_k": {}, "transformer_blocks.13.attn.to_v": {}, "transformer_blocks.13.attn.add_k_proj": {}, "transformer_blocks.13.attn.add_v_proj": {}, "transformer_blocks.13.attn.add_q_proj": {}, "transformer_blocks.13.attn.to_out.0": {}, "transformer_blocks.13.attn.to_add_out": {}, "transformer_blocks.13.attn.to_qkv": {}, "transformer_blocks.13.attn.to_added_qkv": {}, "transformer_blocks.13.attn.output_softmax_quant": { "act_scale": 0.0037760415580123663, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.13.attn.out_q": { "act_scale": 0.0833333358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.13.attn.out_k": { "act_scale": 0.11770833283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.13.attn.out_v": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.13.ff.net.0.proj": {}, "transformer_blocks.13.ff.net.2": {}, "transformer_blocks.13.ff_context.net.0.proj": {}, "transformer_blocks.13.ff_context.net.2": {}, "transformer_blocks.14.norm1.linear": {}, "transformer_blocks.14.norm1_context.linear": {}, "transformer_blocks.14.attn.to_q": {}, "transformer_blocks.14.attn.to_k": {}, "transformer_blocks.14.attn.to_v": {}, "transformer_blocks.14.attn.add_k_proj": {}, "transformer_blocks.14.attn.add_v_proj": {}, "transformer_blocks.14.attn.add_q_proj": {}, "transformer_blocks.14.attn.to_out.0": {}, "transformer_blocks.14.attn.to_add_out": {}, "transformer_blocks.14.attn.to_qkv": {}, "transformer_blocks.14.attn.to_added_qkv": {}, "transformer_blocks.14.attn.output_softmax_quant": { "act_scale": 0.003955078311264515, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.14.attn.out_q": { "act_scale": 0.13333334028720856, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.14.attn.out_k": { "act_scale": 0.2666666805744171, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.14.attn.out_v": { "act_scale": 0.1067708358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.14.ff.net.0.proj": {}, "transformer_blocks.14.ff.net.2": {}, "transformer_blocks.14.ff_context.net.0.proj": {}, "transformer_blocks.14.ff_context.net.2": {}, "transformer_blocks.15.norm1.linear": {}, "transformer_blocks.15.norm1_context.linear": {}, "transformer_blocks.15.attn.to_q": {}, "transformer_blocks.15.attn.to_k": {}, "transformer_blocks.15.attn.to_v": {}, "transformer_blocks.15.attn.add_k_proj": {}, "transformer_blocks.15.attn.add_v_proj": {}, "transformer_blocks.15.attn.add_q_proj": {}, "transformer_blocks.15.attn.to_out.0": {}, "transformer_blocks.15.attn.to_add_out": {}, "transformer_blocks.15.attn.to_qkv": {}, "transformer_blocks.15.attn.to_added_qkv": {}, "transformer_blocks.15.attn.output_softmax_quant": { "act_scale": 0.003971354104578495, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.15.attn.out_q": { "act_scale": 0.15833333134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.15.attn.out_k": { "act_scale": 0.20624999701976776, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.15.attn.out_v": { "act_scale": 0.07239583134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.15.ff.net.0.proj": {}, "transformer_blocks.15.ff.net.2": {}, "transformer_blocks.15.ff_context.net.0.proj": {}, "transformer_blocks.15.ff_context.net.2": {}, "transformer_blocks.16.norm1.linear": {}, "transformer_blocks.16.norm1_context.linear": {}, "transformer_blocks.16.attn.to_q": {}, "transformer_blocks.16.attn.to_k": {}, "transformer_blocks.16.attn.to_v": {}, "transformer_blocks.16.attn.add_k_proj": {}, "transformer_blocks.16.attn.add_v_proj": {}, "transformer_blocks.16.attn.add_q_proj": {}, "transformer_blocks.16.attn.to_out.0": {}, "transformer_blocks.16.attn.to_add_out": {}, "transformer_blocks.16.attn.to_qkv": {}, "transformer_blocks.16.attn.to_added_qkv": {}, "transformer_blocks.16.attn.output_softmax_quant": { "act_scale": 0.00390625, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.16.attn.out_q": { "act_scale": 0.16875000298023224, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.16.attn.out_k": { "act_scale": 0.18333333730697632, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.16.attn.out_v": { "act_scale": 0.08906249701976776, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.16.ff.net.0.proj": {}, "transformer_blocks.16.ff.net.2": {}, "transformer_blocks.16.ff_context.net.0.proj": {}, "transformer_blocks.16.ff_context.net.2": {}, "transformer_blocks.17.norm1.linear": {}, "transformer_blocks.17.norm1_context.linear": {}, "transformer_blocks.17.attn.to_q": {}, "transformer_blocks.17.attn.to_k": {}, "transformer_blocks.17.attn.to_v": {}, "transformer_blocks.17.attn.add_k_proj": {}, "transformer_blocks.17.attn.add_v_proj": {}, "transformer_blocks.17.attn.add_q_proj": {}, "transformer_blocks.17.attn.to_out.0": {}, "transformer_blocks.17.attn.to_add_out": {}, "transformer_blocks.17.attn.to_qkv": {}, "transformer_blocks.17.attn.to_added_qkv": {}, "transformer_blocks.17.attn.output_softmax_quant": { "act_scale": 0.0039388020522892475, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.17.attn.out_q": { "act_scale": 0.20624999701976776, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.17.attn.out_k": { "act_scale": 0.20729166269302368, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.17.attn.out_v": { "act_scale": 0.08281250298023224, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.17.ff.net.0.proj": {}, "transformer_blocks.17.ff.net.2": {}, "transformer_blocks.17.ff_context.net.0.proj": {}, "transformer_blocks.17.ff_context.net.2": {}, "transformer_blocks.18.norm1.linear": {}, "transformer_blocks.18.norm1_context.linear": {}, "transformer_blocks.18.attn.to_q": {}, "transformer_blocks.18.attn.to_k": {}, "transformer_blocks.18.attn.to_v": {}, "transformer_blocks.18.attn.add_k_proj": {}, "transformer_blocks.18.attn.add_v_proj": {}, "transformer_blocks.18.attn.add_q_proj": {}, "transformer_blocks.18.attn.to_out.0": {}, "transformer_blocks.18.attn.to_add_out": {}, "transformer_blocks.18.attn.to_qkv": {}, "transformer_blocks.18.attn.to_added_qkv": {}, "transformer_blocks.18.attn.output_softmax_quant": { "act_scale": 0.004166666883975267, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.18.attn.out_q": { "act_scale": 0.15833333134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.18.attn.out_k": { "act_scale": 0.18645833432674408, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.18.attn.out_v": { "act_scale": 0.09583333134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "transformer_blocks.18.ff.net.0.proj": {}, "transformer_blocks.18.ff.net.2": {}, "transformer_blocks.18.ff_context.net.0.proj": {}, "transformer_blocks.18.ff_context.net.2": {}, "single_transformer_blocks.0.norm.linear": {}, "single_transformer_blocks.0.proj_mlp": {}, "single_transformer_blocks.0.proj_out": {}, "single_transformer_blocks.0.attn.to_q": {}, "single_transformer_blocks.0.attn.to_k": {}, "single_transformer_blocks.0.attn.to_v": {}, "single_transformer_blocks.0.attn.to_qkv": {}, "single_transformer_blocks.0.attn.output_softmax_quant": { "act_scale": 0.003483072854578495, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.0.attn.out_q": { "act_scale": 0.12343750149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.0.attn.out_k": { "act_scale": 0.10104166716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.0.attn.out_v": { "act_scale": 0.05885416641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.1.norm.linear": {}, "single_transformer_blocks.1.proj_mlp": {}, "single_transformer_blocks.1.proj_out": {}, "single_transformer_blocks.1.attn.to_q": {}, "single_transformer_blocks.1.attn.to_k": {}, "single_transformer_blocks.1.attn.to_v": {}, "single_transformer_blocks.1.attn.to_qkv": {}, "single_transformer_blocks.1.attn.output_softmax_quant": { "act_scale": 0.0034505208022892475, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.1.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.1.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.1.attn.out_v": { "act_scale": 0.05781250074505806, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.2.norm.linear": {}, "single_transformer_blocks.2.proj_mlp": {}, "single_transformer_blocks.2.proj_out": {}, "single_transformer_blocks.2.attn.to_q": {}, "single_transformer_blocks.2.attn.to_k": {}, "single_transformer_blocks.2.attn.to_v": {}, "single_transformer_blocks.2.attn.to_qkv": {}, "single_transformer_blocks.2.attn.output_softmax_quant": { "act_scale": 0.00390625, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.2.attn.out_q": { "act_scale": 0.06484375149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.2.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.2.attn.out_v": { "act_scale": 0.0481770820915699, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.3.norm.linear": {}, "single_transformer_blocks.3.proj_mlp": {}, "single_transformer_blocks.3.proj_out": {}, "single_transformer_blocks.3.attn.to_q": {}, "single_transformer_blocks.3.attn.to_k": {}, "single_transformer_blocks.3.attn.to_v": {}, "single_transformer_blocks.3.attn.to_qkv": {}, "single_transformer_blocks.3.attn.output_softmax_quant": { "act_scale": 0.0036783854011446238, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.3.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.3.attn.out_k": { "act_scale": 0.06197916716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.3.attn.out_v": { "act_scale": 0.04322916641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.4.norm.linear": {}, "single_transformer_blocks.4.proj_mlp": {}, "single_transformer_blocks.4.proj_out": {}, "single_transformer_blocks.4.attn.to_q": {}, "single_transformer_blocks.4.attn.to_k": {}, "single_transformer_blocks.4.attn.to_v": {}, "single_transformer_blocks.4.attn.to_qkv": {}, "single_transformer_blocks.4.attn.output_softmax_quant": { "act_scale": 0.0037923178169876337, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.4.attn.out_q": { "act_scale": 0.06588541716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.4.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.4.attn.out_v": { "act_scale": 0.04270833358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.5.norm.linear": {}, "single_transformer_blocks.5.proj_mlp": {}, "single_transformer_blocks.5.proj_out": {}, "single_transformer_blocks.5.attn.to_q": {}, "single_transformer_blocks.5.attn.to_k": {}, "single_transformer_blocks.5.attn.to_v": {}, "single_transformer_blocks.5.attn.to_qkv": {}, "single_transformer_blocks.5.attn.output_softmax_quant": { "act_scale": 0.0036295573227107525, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.5.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.5.attn.out_k": { "act_scale": 0.06640625, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.5.attn.out_v": { "act_scale": 0.04192708432674408, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.6.norm.linear": {}, "single_transformer_blocks.6.proj_mlp": {}, "single_transformer_blocks.6.proj_out": {}, "single_transformer_blocks.6.attn.to_q": {}, "single_transformer_blocks.6.attn.to_k": {}, "single_transformer_blocks.6.attn.to_v": {}, "single_transformer_blocks.6.attn.to_qkv": {}, "single_transformer_blocks.6.attn.output_softmax_quant": { "act_scale": 0.003841145895421505, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.6.attn.out_q": { "act_scale": 0.08385416865348816, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.6.attn.out_k": { "act_scale": 0.0638020858168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.6.attn.out_v": { "act_scale": 0.04453124850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.7.norm.linear": {}, "single_transformer_blocks.7.proj_mlp": {}, "single_transformer_blocks.7.proj_out": {}, "single_transformer_blocks.7.attn.to_q": {}, "single_transformer_blocks.7.attn.to_k": {}, "single_transformer_blocks.7.attn.to_v": {}, "single_transformer_blocks.7.attn.to_qkv": {}, "single_transformer_blocks.7.attn.output_softmax_quant": { "act_scale": 0.0035481771919876337, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.7.attn.out_q": { "act_scale": 0.06406249850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.7.attn.out_k": { "act_scale": 0.0611979179084301, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.7.attn.out_v": { "act_scale": 0.04270833358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.8.norm.linear": {}, "single_transformer_blocks.8.proj_mlp": {}, "single_transformer_blocks.8.proj_out": {}, "single_transformer_blocks.8.attn.to_q": {}, "single_transformer_blocks.8.attn.to_k": {}, "single_transformer_blocks.8.attn.to_v": {}, "single_transformer_blocks.8.attn.to_qkv": {}, "single_transformer_blocks.8.attn.output_softmax_quant": { "act_scale": 0.0037760415580123663, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.8.attn.out_q": { "act_scale": 0.08958332985639572, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.8.attn.out_k": { "act_scale": 0.07916666567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.8.attn.out_v": { "act_scale": 0.0416666679084301, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.9.norm.linear": {}, "single_transformer_blocks.9.proj_mlp": {}, "single_transformer_blocks.9.proj_out": {}, "single_transformer_blocks.9.attn.to_q": {}, "single_transformer_blocks.9.attn.to_k": {}, "single_transformer_blocks.9.attn.to_v": {}, "single_transformer_blocks.9.attn.to_qkv": {}, "single_transformer_blocks.9.attn.output_softmax_quant": { "act_scale": 0.0038574219215661287, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.9.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.9.attn.out_k": { "act_scale": 0.06197916716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.9.attn.out_v": { "act_scale": 0.04062499850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.10.norm.linear": {}, "single_transformer_blocks.10.proj_mlp": {}, "single_transformer_blocks.10.proj_out": {}, "single_transformer_blocks.10.attn.to_q": {}, "single_transformer_blocks.10.attn.to_k": {}, "single_transformer_blocks.10.attn.to_v": {}, "single_transformer_blocks.10.attn.to_qkv": {}, "single_transformer_blocks.10.attn.output_softmax_quant": { "act_scale": 0.003922526258975267, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.10.attn.out_q": { "act_scale": 0.08645833283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.10.attn.out_k": { "act_scale": 0.0572916679084301, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.10.attn.out_v": { "act_scale": 0.04062499850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.11.norm.linear": {}, "single_transformer_blocks.11.proj_mlp": {}, "single_transformer_blocks.11.proj_out": {}, "single_transformer_blocks.11.attn.to_q": {}, "single_transformer_blocks.11.attn.to_k": {}, "single_transformer_blocks.11.attn.to_v": {}, "single_transformer_blocks.11.attn.to_qkv": {}, "single_transformer_blocks.11.attn.output_softmax_quant": { "act_scale": 0.00402018241584301, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.11.attn.out_q": { "act_scale": 0.10000000149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.11.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.11.attn.out_v": { "act_scale": 0.04088541492819786, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.12.norm.linear": {}, "single_transformer_blocks.12.proj_mlp": {}, "single_transformer_blocks.12.proj_out": {}, "single_transformer_blocks.12.attn.to_q": {}, "single_transformer_blocks.12.attn.to_k": {}, "single_transformer_blocks.12.attn.to_v": {}, "single_transformer_blocks.12.attn.to_qkv": {}, "single_transformer_blocks.12.attn.output_softmax_quant": { "act_scale": 0.0038085938431322575, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.12.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.12.attn.out_k": { "act_scale": 0.06458333134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.12.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.13.norm.linear": {}, "single_transformer_blocks.13.proj_mlp": {}, "single_transformer_blocks.13.proj_out": {}, "single_transformer_blocks.13.attn.to_q": {}, "single_transformer_blocks.13.attn.to_k": {}, "single_transformer_blocks.13.attn.to_v": {}, "single_transformer_blocks.13.attn.to_qkv": {}, "single_transformer_blocks.13.attn.output_softmax_quant": { "act_scale": 0.0039876303635537624, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.13.attn.out_q": { "act_scale": 0.10520832985639572, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.13.attn.out_k": { "act_scale": 0.05703125149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.13.attn.out_v": { "act_scale": 0.04218750074505806, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.14.norm.linear": {}, "single_transformer_blocks.14.proj_mlp": {}, "single_transformer_blocks.14.proj_out": {}, "single_transformer_blocks.14.attn.to_q": {}, "single_transformer_blocks.14.attn.to_k": {}, "single_transformer_blocks.14.attn.to_v": {}, "single_transformer_blocks.14.attn.to_qkv": {}, "single_transformer_blocks.14.attn.output_softmax_quant": { "act_scale": 0.0040690102614462376, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.14.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.14.attn.out_k": { "act_scale": 0.11302082985639572, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.14.attn.out_v": { "act_scale": 0.0377604179084301, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.15.norm.linear": {}, "single_transformer_blocks.15.proj_mlp": {}, "single_transformer_blocks.15.proj_out": {}, "single_transformer_blocks.15.attn.to_q": {}, "single_transformer_blocks.15.attn.to_k": {}, "single_transformer_blocks.15.attn.to_v": {}, "single_transformer_blocks.15.attn.to_qkv": {}, "single_transformer_blocks.15.attn.output_softmax_quant": { "act_scale": 0.00390625, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.15.attn.out_q": { "act_scale": 0.08802083134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.15.attn.out_k": { "act_scale": 0.08906249701976776, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.15.attn.out_v": { "act_scale": 0.04791666567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.16.norm.linear": {}, "single_transformer_blocks.16.proj_mlp": {}, "single_transformer_blocks.16.proj_out": {}, "single_transformer_blocks.16.attn.to_q": {}, "single_transformer_blocks.16.attn.to_k": {}, "single_transformer_blocks.16.attn.to_v": {}, "single_transformer_blocks.16.attn.to_qkv": {}, "single_transformer_blocks.16.attn.output_softmax_quant": { "act_scale": 0.003971354104578495, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.16.attn.out_q": { "act_scale": 0.09531249850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.16.attn.out_k": { "act_scale": 0.09375, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.16.attn.out_v": { "act_scale": 0.05078125, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.17.norm.linear": {}, "single_transformer_blocks.17.proj_mlp": {}, "single_transformer_blocks.17.proj_out": {}, "single_transformer_blocks.17.attn.to_q": {}, "single_transformer_blocks.17.attn.to_k": {}, "single_transformer_blocks.17.attn.to_v": {}, "single_transformer_blocks.17.attn.to_qkv": {}, "single_transformer_blocks.17.attn.output_softmax_quant": { "act_scale": 0.004166666883975267, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.17.attn.out_q": { "act_scale": 0.0963541641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.17.attn.out_k": { "act_scale": 0.10572917014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.17.attn.out_v": { "act_scale": 0.05416666716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.18.norm.linear": {}, "single_transformer_blocks.18.proj_mlp": {}, "single_transformer_blocks.18.proj_out": {}, "single_transformer_blocks.18.attn.to_q": {}, "single_transformer_blocks.18.attn.to_k": {}, "single_transformer_blocks.18.attn.to_v": {}, "single_transformer_blocks.18.attn.to_qkv": {}, "single_transformer_blocks.18.attn.output_softmax_quant": { "act_scale": 0.003922526258975267, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.18.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.18.attn.out_k": { "act_scale": 0.12343750149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.18.attn.out_v": { "act_scale": 0.04635416716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.19.norm.linear": {}, "single_transformer_blocks.19.proj_mlp": {}, "single_transformer_blocks.19.proj_out": {}, "single_transformer_blocks.19.attn.to_q": {}, "single_transformer_blocks.19.attn.to_k": {}, "single_transformer_blocks.19.attn.to_v": {}, "single_transformer_blocks.19.attn.to_qkv": {}, "single_transformer_blocks.19.attn.output_softmax_quant": { "act_scale": 0.00403645820915699, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.19.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.19.attn.out_k": { "act_scale": 0.07968749850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.19.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.20.norm.linear": {}, "single_transformer_blocks.20.proj_mlp": {}, "single_transformer_blocks.20.proj_out": {}, "single_transformer_blocks.20.attn.to_q": {}, "single_transformer_blocks.20.attn.to_k": {}, "single_transformer_blocks.20.attn.to_v": {}, "single_transformer_blocks.20.attn.to_qkv": {}, "single_transformer_blocks.20.attn.output_softmax_quant": { "act_scale": 0.004134114366024733, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.20.attn.out_q": { "act_scale": 0.08385416865348816, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.20.attn.out_k": { "act_scale": 0.0885416641831398, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.20.attn.out_v": { "act_scale": 0.04661458358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.21.norm.linear": {}, "single_transformer_blocks.21.proj_mlp": {}, "single_transformer_blocks.21.proj_out": {}, "single_transformer_blocks.21.attn.to_q": {}, "single_transformer_blocks.21.attn.to_k": {}, "single_transformer_blocks.21.attn.to_v": {}, "single_transformer_blocks.21.attn.to_qkv": {}, "single_transformer_blocks.21.attn.output_softmax_quant": { "act_scale": 0.0037760415580123663, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.21.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.21.attn.out_k": { "act_scale": 0.06562499701976776, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.21.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.22.norm.linear": {}, "single_transformer_blocks.22.proj_mlp": {}, "single_transformer_blocks.22.proj_out": {}, "single_transformer_blocks.22.attn.to_q": {}, "single_transformer_blocks.22.attn.to_k": {}, "single_transformer_blocks.22.attn.to_v": {}, "single_transformer_blocks.22.attn.to_qkv": {}, "single_transformer_blocks.22.attn.output_softmax_quant": { "act_scale": 0.0040527344681322575, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.22.attn.out_q": { "act_scale": 0.07916666567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.22.attn.out_k": { "act_scale": 0.07864583283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.22.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.23.norm.linear": {}, "single_transformer_blocks.23.proj_mlp": {}, "single_transformer_blocks.23.proj_out": {}, "single_transformer_blocks.23.attn.to_q": {}, "single_transformer_blocks.23.attn.to_k": {}, "single_transformer_blocks.23.attn.to_v": {}, "single_transformer_blocks.23.attn.to_qkv": {}, "single_transformer_blocks.23.attn.output_softmax_quant": { "act_scale": 0.0040527344681322575, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.23.attn.out_q": { "act_scale": 0.09218750149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.23.attn.out_k": { "act_scale": 0.07968749850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.23.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.24.norm.linear": {}, "single_transformer_blocks.24.proj_mlp": {}, "single_transformer_blocks.24.proj_out": {}, "single_transformer_blocks.24.attn.to_q": {}, "single_transformer_blocks.24.attn.to_k": {}, "single_transformer_blocks.24.attn.to_v": {}, "single_transformer_blocks.24.attn.to_qkv": {}, "single_transformer_blocks.24.attn.output_softmax_quant": { "act_scale": 0.003922526258975267, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.24.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.24.attn.out_k": { "act_scale": 0.05781250074505806, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.24.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.25.norm.linear": {}, "single_transformer_blocks.25.proj_mlp": {}, "single_transformer_blocks.25.proj_out": {}, "single_transformer_blocks.25.attn.to_q": {}, "single_transformer_blocks.25.attn.to_k": {}, "single_transformer_blocks.25.attn.to_v": {}, "single_transformer_blocks.25.attn.to_qkv": {}, "single_transformer_blocks.25.attn.output_softmax_quant": { "act_scale": 0.0040690102614462376, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.25.attn.out_q": { "act_scale": 0.08802083134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.25.attn.out_k": { "act_scale": 0.07604166865348816, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.25.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.26.norm.linear": {}, "single_transformer_blocks.26.proj_mlp": {}, "single_transformer_blocks.26.proj_out": {}, "single_transformer_blocks.26.attn.to_q": {}, "single_transformer_blocks.26.attn.to_k": {}, "single_transformer_blocks.26.attn.to_v": {}, "single_transformer_blocks.26.attn.to_qkv": {}, "single_transformer_blocks.26.attn.output_softmax_quant": { "act_scale": 0.0040039061568677425, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.26.attn.out_q": { "act_scale": 0.08124999701976776, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.26.attn.out_k": { "act_scale": 0.07760416716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.26.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.27.norm.linear": {}, "single_transformer_blocks.27.proj_mlp": {}, "single_transformer_blocks.27.proj_out": {}, "single_transformer_blocks.27.attn.to_q": {}, "single_transformer_blocks.27.attn.to_k": {}, "single_transformer_blocks.27.attn.to_v": {}, "single_transformer_blocks.27.attn.to_qkv": {}, "single_transformer_blocks.27.attn.output_softmax_quant": { "act_scale": 0.004101562313735485, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.27.attn.out_q": { "act_scale": 0.0859375, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.27.attn.out_k": { "act_scale": 0.09375, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.27.attn.out_v": { "act_scale": 0.04114583507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.28.norm.linear": {}, "single_transformer_blocks.28.proj_mlp": {}, "single_transformer_blocks.28.proj_out": {}, "single_transformer_blocks.28.attn.to_q": {}, "single_transformer_blocks.28.attn.to_k": {}, "single_transformer_blocks.28.attn.to_v": {}, "single_transformer_blocks.28.attn.to_qkv": {}, "single_transformer_blocks.28.attn.output_softmax_quant": { "act_scale": 0.0040690102614462376, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.28.attn.out_q": { "act_scale": 0.09427083283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.28.attn.out_k": { "act_scale": 0.109375, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.28.attn.out_v": { "act_scale": 0.03333333507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.29.norm.linear": {}, "single_transformer_blocks.29.proj_mlp": {}, "single_transformer_blocks.29.proj_out": {}, "single_transformer_blocks.29.attn.to_q": {}, "single_transformer_blocks.29.attn.to_k": {}, "single_transformer_blocks.29.attn.to_v": {}, "single_transformer_blocks.29.attn.to_qkv": {}, "single_transformer_blocks.29.attn.output_softmax_quant": { "act_scale": 0.004134114366024733, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.29.attn.out_q": { "act_scale": 0.078125, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.29.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.29.attn.out_v": { "act_scale": 0.03984374925494194, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.30.norm.linear": {}, "single_transformer_blocks.30.proj_mlp": {}, "single_transformer_blocks.30.proj_out": {}, "single_transformer_blocks.30.attn.to_q": {}, "single_transformer_blocks.30.attn.to_k": {}, "single_transformer_blocks.30.attn.to_v": {}, "single_transformer_blocks.30.attn.to_qkv": {}, "single_transformer_blocks.30.attn.output_softmax_quant": { "act_scale": 0.004134114366024733, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.30.attn.out_q": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.30.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.30.attn.out_v": { "act_scale": 0.04505208507180214, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.31.norm.linear": {}, "single_transformer_blocks.31.proj_mlp": {}, "single_transformer_blocks.31.proj_out": {}, "single_transformer_blocks.31.attn.to_q": {}, "single_transformer_blocks.31.attn.to_k": {}, "single_transformer_blocks.31.attn.to_v": {}, "single_transformer_blocks.31.attn.to_qkv": {}, "single_transformer_blocks.31.attn.output_softmax_quant": { "act_scale": 0.004166666883975267, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.31.attn.out_q": { "act_scale": 0.0989583358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.31.attn.out_k": { "act_scale": 0.06197916716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.31.attn.out_v": { "act_scale": 0.046875, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.32.norm.linear": {}, "single_transformer_blocks.32.proj_mlp": {}, "single_transformer_blocks.32.proj_out": {}, "single_transformer_blocks.32.attn.to_q": {}, "single_transformer_blocks.32.attn.to_k": {}, "single_transformer_blocks.32.attn.to_v": {}, "single_transformer_blocks.32.attn.to_qkv": {}, "single_transformer_blocks.32.attn.output_softmax_quant": { "act_scale": 0.0040690102614462376, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.32.attn.out_q": { "act_scale": 0.10312499850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.32.attn.out_k": { "act_scale": 0.09270833432674408, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.32.attn.out_v": { "act_scale": 0.05156249925494194, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.33.norm.linear": {}, "single_transformer_blocks.33.proj_mlp": {}, "single_transformer_blocks.33.proj_out": {}, "single_transformer_blocks.33.attn.to_q": {}, "single_transformer_blocks.33.attn.to_k": {}, "single_transformer_blocks.33.attn.to_v": {}, "single_transformer_blocks.33.attn.to_qkv": {}, "single_transformer_blocks.33.attn.output_softmax_quant": { "act_scale": 0.0040690102614462376, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.33.attn.out_q": { "act_scale": 0.08541666716337204, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.33.attn.out_k": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.33.attn.out_v": { "act_scale": 0.04843749850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.34.norm.linear": {}, "single_transformer_blocks.34.proj_mlp": {}, "single_transformer_blocks.34.proj_out": {}, "single_transformer_blocks.34.attn.to_q": {}, "single_transformer_blocks.34.attn.to_k": {}, "single_transformer_blocks.34.attn.to_v": {}, "single_transformer_blocks.34.attn.to_qkv": {}, "single_transformer_blocks.34.attn.output_softmax_quant": { "act_scale": 0.0040690102614462376, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.34.attn.out_q": { "act_scale": 0.09427083283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.34.attn.out_k": { "act_scale": 0.12656250596046448, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.34.attn.out_v": { "act_scale": 0.05520833283662796, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.35.norm.linear": {}, "single_transformer_blocks.35.proj_mlp": {}, "single_transformer_blocks.35.proj_out": {}, "single_transformer_blocks.35.attn.to_q": {}, "single_transformer_blocks.35.attn.to_k": {}, "single_transformer_blocks.35.attn.to_v": {}, "single_transformer_blocks.35.attn.to_qkv": {}, "single_transformer_blocks.35.attn.output_softmax_quant": { "act_scale": 0.00403645820915699, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.35.attn.out_q": { "act_scale": 0.0989583358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.35.attn.out_k": { "act_scale": 0.08020833134651184, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.35.attn.out_v": { "act_scale": 0.05624999850988388, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.36.norm.linear": {}, "single_transformer_blocks.36.proj_mlp": {}, "single_transformer_blocks.36.proj_out": {}, "single_transformer_blocks.36.attn.to_q": {}, "single_transformer_blocks.36.attn.to_k": {}, "single_transformer_blocks.36.attn.to_v": {}, "single_transformer_blocks.36.attn.to_qkv": {}, "single_transformer_blocks.36.attn.output_softmax_quant": { "act_scale": 0.00402018241584301, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.36.attn.out_q": { "act_scale": 0.12604166567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.36.attn.out_k": { "act_scale": 0.10000000149011612, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.36.attn.out_v": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.37.norm.linear": {}, "single_transformer_blocks.37.proj_mlp": {}, "single_transformer_blocks.37.proj_out": {}, "single_transformer_blocks.37.attn.to_q": {}, "single_transformer_blocks.37.attn.to_k": {}, "single_transformer_blocks.37.attn.to_v": {}, "single_transformer_blocks.37.attn.to_qkv": {}, "single_transformer_blocks.37.attn.output_softmax_quant": { "act_scale": 0.0015055338153615594, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.37.attn.out_q": { "act_scale": 0.0989583358168602, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.37.attn.out_k": { "act_scale": 0.08697916567325592, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "single_transformer_blocks.37.attn.out_v": { "act_scale": 0.06666667014360428, "act_scale_shape": [], "act_zp": 0.0, "act_zp_shape": [], "act_zp_dtype": "torch.float8_e4m3fnuz" }, "norm_out.linear": {}, "proj_out": {} }