Download quantization_config.json from nota-ai/Solar-Open2-250B-Nota-INT4: direct link, hf CLI and curl.
- Browser
- Download file 65.5 kB
-
https://huggingface.co/nota-ai/Solar-Open2-250B-Nota-INT4/resolve/main/quantization_config.json
- Command line
-
hf download hf://nota-ai/Solar-Open2-250B-Nota-INT4/quantization_config.json
-
curl -L -o quantization_config.json https://huggingface.co/nota-ai/Solar-Open2-250B-Nota-INT4/resolve/main/quantization_config.json
65.5 kB
| { | |
| "bits": 4, | |
| "data_type": "int", | |
| "group_size": 128, | |
| "sym": true, | |
| "low_gpu_mem_usage": true, | |
| "block_name_to_quantize": "model.layers", | |
| "quant_method": "auto-round", | |
| "packing_format": "auto_round:auto_gptq", | |
| "extra_config": { | |
| "model.layers.0.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.self_attn.g_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.q_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.k_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.v_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.f_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.f_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.g_a_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.g_b_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.self_attn.o_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.0.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.1.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.2.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.3.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.4.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.5.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.6.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.7.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.8.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.9.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.10.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.11.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.12.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.13.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.14.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.15.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.16.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.17.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.18.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.19.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.20.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.21.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.22.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.23.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.24.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.25.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.26.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.27.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.28.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.29.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.30.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.31.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.32.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.33.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.34.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.35.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.36.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.37.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.38.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.39.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.40.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.41.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.42.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.43.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.44.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.45.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.46.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.mlp.shared_experts.gate_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.mlp.shared_experts.up_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| "model.layers.47.mlp.shared_experts.down_proj": { | |
| "bits": 16, | |
| "data_type": "fp", | |
| "act_data_type": "fp" | |
| }, | |
| ".*model\\.layers\\.[0-47]\\.mlp\\.gate.*": { | |
| "bits": 16, | |
| "data_type": "float" | |
| } | |
| } | |
| } |