| /home/cfruan/.conda/envs/mlc-source-311/bin/python -m mlc_chat gen_config /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5 --quantization q4f16_1 --conv-template phi-2 --output /tmp/tmpxe445xtc | |
| [2023-12-28 23:33:19] INFO auto_config.py:115: [92mFound[0m model configuration: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/config.json | |
| [2023-12-28 23:33:19] INFO auto_config.py:151: [92mFound[0m model type: [1mphi-msft[0m. Use `--model-type` to override. | |
| [2023-12-28 23:33:19] INFO phi_model.py:59: [1mcontext_window_size[0m not found in config.json. Falling back to n_positions (2048) | |
| [2023-12-28 23:33:19] INFO gen_config.py:129: [91mNot found[0m tokenizer config: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/tokenizer.model | |
| [2023-12-28 23:33:19] INFO gen_config.py:127: [92mFound[0m tokenizer config: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/tokenizer.json. Copying to [1m/tmp/tmpxe445xtc/tokenizer.json[0m | |
| [2023-12-28 23:33:19] INFO gen_config.py:127: [92mFound[0m tokenizer config: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/vocab.json. Copying to [1m/tmp/tmpxe445xtc/vocab.json[0m | |
| [2023-12-28 23:33:19] INFO gen_config.py:127: [92mFound[0m tokenizer config: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/merges.txt. Copying to [1m/tmp/tmpxe445xtc/merges.txt[0m | |
| [2023-12-28 23:33:19] INFO gen_config.py:127: [92mFound[0m tokenizer config: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/added_tokens.json. Copying to [1m/tmp/tmpxe445xtc/added_tokens.json[0m | |
| [2023-12-28 23:33:19] INFO gen_config.py:127: [92mFound[0m tokenizer config: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/tokenizer_config.json. Copying to [1m/tmp/tmpxe445xtc/tokenizer_config.json[0m | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mpad_token_id[0m: 0 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mbos_token_id[0m: 1 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1meos_token_id[0m: 2 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mtemperature[0m: 0.7 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mrepetition_penalty[0m: 1.0 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mtop_p[0m: 0.95 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mmean_gen_len[0m: 128 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mmax_gen_len[0m: 512 | |
| [2023-12-28 23:33:19] INFO gen_config.py:69: [System default] Setting [1mshift_fill_factor[0m: 0.3 | |
| [2023-12-28 23:33:19] INFO gen_config.py:157: Dumping configuration file to: [1m/tmp/tmpxe445xtc/mlc-chat-config.json[0m | |
| /home/cfruan/.conda/envs/mlc-source-311/bin/python -m mlc_chat convert_weight /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5 --quantization q4f16_1 --source-format auto --output /tmp/tmpxe445xtc | |
| [2023-12-28 23:33:20] INFO auto_config.py:115: [92mFound[0m model configuration: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/config.json | |
| [2023-12-28 23:33:20] INFO auto_device.py:76: [92mFound[0m device: cuda:0 | |
| [2023-12-28 23:33:20] INFO auto_device.py:76: [92mFound[0m device: cuda:1 | |
| [2023-12-28 23:33:20] INFO auto_device.py:85: [91mNot found[0m device: rocm:0 | |
| [2023-12-28 23:33:20] INFO auto_device.py:85: [91mNot found[0m device: metal:0 | |
| [2023-12-28 23:33:21] INFO auto_device.py:76: [92mFound[0m device: vulkan:0 | |
| [2023-12-28 23:33:21] INFO auto_device.py:76: [92mFound[0m device: vulkan:1 | |
| [2023-12-28 23:33:21] INFO auto_device.py:76: [92mFound[0m device: vulkan:2 | |
| [2023-12-28 23:33:21] INFO auto_device.py:85: [91mNot found[0m device: opencl:0 | |
| [2023-12-28 23:33:21] INFO auto_device.py:33: Using device: [1mcuda:0[0m | |
| [2023-12-28 23:33:21] INFO auto_weight.py:70: Finding weights in: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5 | |
| [2023-12-28 23:33:21] INFO auto_weight.py:129: [92mFound[0m source weight format: huggingface-torch. Source configuration: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/pytorch_model.bin | |
| [2023-12-28 23:33:21] INFO auto_weight.py:149: [91mNot found[0m Huggingface Safetensor | |
| [2023-12-28 23:33:21] INFO auto_weight.py:106: Using source weight configuration: [1m/ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/pytorch_model.bin[0m. Use `--source` to override. | |
| [2023-12-28 23:33:21] INFO auto_weight.py:110: Using source weight format: [1mhuggingface-torch[0m. Use `--source-format` to override. | |
| [2023-12-28 23:33:21] INFO auto_config.py:151: [92mFound[0m model type: [1mphi-msft[0m. Use `--model-type` to override. | |
| [2023-12-28 23:33:21] INFO phi_model.py:59: [1mcontext_window_size[0m not found in config.json. Falling back to n_positions (2048) | |
| [2023-12-28 23:33:24] INFO huggingface_loader.py:169: Loading HF parameters from: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/pytorch_model.bin | |
| [1mWeight conversion with arguments:[0m | |
| [1m--config[0m /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/config.json | |
| [1m--quantization[0m GroupQuantize(name='q4f16_1', kind='group-quant', group_size=32, quantize_dtype='int4', storage_dtype='uint32', model_dtype='float16', num_elem_per_storage=8, num_storage_per_group=4, max_int_value=7) | |
| [1m--model-type[0m phi-msft | |
| [1m--device[0m cuda:0 | |
| [1m--source[0m /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/pytorch_model.bin | |
| [1m--source-format[0m huggingface-torch | |
| [1m--output[0m /tmp/tmpxe445xtc | |
| 0%| | 0/245 [00:00<?, ?it/s] [2023-12-28 23:33:25] INFO group_quantization.py:200: Compiling quantize function for key: (51200, 2048, 'float16', 'cuda') | |
| 0%| | 0/245 [00:00<?, ?it/s] [2023-12-28 23:33:26] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.embd.q_weight[0m", shape: (51200, 256), dtype: uint32 | |
| 0%| | 0/245 [00:01<?, ?it/s] [2023-12-28 23:33:26] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.embd.q_scale[0m", shape: (51200, 64), dtype: float16 | |
| 0%| | 0/245 [00:01<?, ?it/s] 0%|▍ | 1/245 [00:01<04:43, 1.16s/it] [2023-12-28 23:33:26] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.0.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 0%|▍ | 1/245 [00:01<04:43, 1.16s/it] [2023-12-28 23:33:26] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.0.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 0%|▍ | 1/245 [00:01<04:43, 1.16s/it] [2023-12-28 23:33:26] INFO group_quantization.py:200: Compiling quantize function for key: (6144, 2048, 'float16', 'cuda') | |
| 0%|▍ | 1/245 [00:01<04:43, 1.16s/it] [2023-12-28 23:33:27] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 0%|▍ | 1/245 [00:01<04:43, 1.16s/it] [2023-12-28 23:33:27] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 0%|▍ | 1/245 [00:01<04:43, 1.16s/it] 2%|█▉ | 4/245 [00:01<01:17, 3.10it/s] [2023-12-28 23:33:27] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.0.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 2%|█▉ | 4/245 [00:01<01:17, 3.10it/s] [2023-12-28 23:33:27] INFO group_quantization.py:200: Compiling quantize function for key: (2048, 2048, 'float16', 'cuda') | |
| 2%|█▉ | 4/245 [00:01<01:17, 3.10it/s] [2023-12-28 23:33:27] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 2%|█▉ | 4/245 [00:01<01:17, 3.10it/s] [2023-12-28 23:33:27] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 2%|█▉ | 4/245 [00:01<01:17, 3.10it/s] 2%|██▉ | 6/245 [00:01<01:02, 3.83it/s] [2023-12-28 23:33:27] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.0.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 2%|██▉ | 6/245 [00:01<01:02, 3.83it/s] [2023-12-28 23:33:27] INFO group_quantization.py:200: Compiling quantize function for key: (8192, 2048, 'float16', 'cuda') | |
| 2%|██▉ | 6/245 [00:01<01:02, 3.83it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 2%|██▉ | 6/245 [00:02<01:02, 3.83it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 2%|██▉ | 6/245 [00:02<01:02, 3.83it/s] 3%|███▉ | 8/245 [00:02<00:55, 4.29it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.0.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 3%|███▉ | 8/245 [00:02<00:55, 4.29it/s] [2023-12-28 23:33:28] INFO group_quantization.py:200: Compiling quantize function for key: (2048, 8192, 'float16', 'cuda') | |
| 3%|███▉ | 8/245 [00:02<00:55, 4.29it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 3%|███▉ | 8/245 [00:02<00:55, 4.29it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.0.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 3%|███▉ | 8/245 [00:02<00:55, 4.29it/s] 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.0.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.1.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.1.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.1.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.1.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.1.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.1.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.1.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.2.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.2.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.2.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.2.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.2.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.2.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.2.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.3.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.3.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.3.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.3.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.3.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.3.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.3.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.4.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.4.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 4%|████▉ | 10/245 [00:02<00:51, 4.59it/s] 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.4.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.4.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.4.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.4.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.4.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.5.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.5.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.5.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.5.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.5.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.5.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.5.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.6.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.6.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.6.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.6.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.6.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.6.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.6.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.7.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.7.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.7.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.7.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 18%|█████████████████████▌ | 44/245 [00:02<00:05, 39.49it/s] 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.7.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.7.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.7.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.8.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.8.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.8.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.8.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.8.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.8.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.8.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.9.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.9.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.9.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.9.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.9.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.9.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.9.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.10.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.10.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.10.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.10.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.10.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.10.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.10.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 32%|██████████████████████████████████████▏ | 78/245 [00:02<00:02, 78.26it/s] 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.11.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.11.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.11.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.11.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.11.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.11.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.11.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.12.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.12.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.12.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:02<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.12.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.12.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.12.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.12.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.13.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.13.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.13.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.13.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.13.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.13.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.13.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.14.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.14.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.14.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.14.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 45%|█████████████████████████████████████████████████████▍ | 111/245 [00:03<00:01, 117.47it/s] 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.14.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.14.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.14.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.15.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.15.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.15.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.15.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.15.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.15.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.15.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.16.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.16.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.16.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.16.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.16.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.16.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.16.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.17.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.17.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.17.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.17.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.17.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.17.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 60%|██████████████████████████████████████████████████████████████████████▊ | 147/245 [00:03<00:00, 161.54it/s] 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.17.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.18.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.18.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.18.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.18.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.18.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.18.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.18.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.19.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.19.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.19.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.19.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.19.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.19.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.19.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.20.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.20.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.20.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.20.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:28] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.20.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.20.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.20.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.21.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.21.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.21.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.21.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 73%|██████████████████████████████████████████████████████████████████████████████████████▋ | 180/245 [00:03<00:00, 192.97it/s] 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.21.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.21.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.21.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.22.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.22.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.22.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.22.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.22.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.22.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.22.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.23.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.23.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mixer.Wqkv.q_weight[0m", shape: (6144, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mixer.Wqkv.q_scale[0m", shape: (6144, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.23.mixer.Wqkv.bias[0m", shape: (6144,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mixer.out_proj.q_weight[0m", shape: (2048, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mixer.out_proj.q_scale[0m", shape: (2048, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.23.mixer.out_proj.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mlp.fc1.q_weight[0m", shape: (8192, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mlp.fc1.q_scale[0m", shape: (8192, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.23.mlp.fc1.bias[0m", shape: (8192,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mlp.fc2.q_weight[0m", shape: (2048, 1024), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mtransformer.h.23.mlp.fc2.q_scale[0m", shape: (2048, 256), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mtransformer.h.23.mlp.fc2.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mlm_head.ln.weight[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mlm_head.ln.bias[0m", shape: (2048,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mlm_head.linear.q_weight[0m", shape: (51200, 256), dtype: uint32 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:121: [Quantized] Parameter: "[1mlm_head.linear.q_scale[0m", shape: (51200, 64), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] [2023-12-28 23:33:29] INFO huggingface_loader.py:129: [Not quantized] Parameter: "[1mlm_head.linear.bias[0m", shape: (51200,), dtype: float16 | |
| 89%|████████████████████████████████████████████████████████████████████████████████████████████████████████▉ | 218/245 [00:03<00:00, 230.43it/s] 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 245/245 [00:03<00:00, 71.86it/s] | |
| [2023-12-28 23:33:29] INFO huggingface_loader.py:179: Unloading HF weight file: /ssd1/cfruan/mlc-llm-repos/mlc-llm-head/dist/models/phi-1_5/pytorch_model.bin | |
| [2023-12-28 23:33:29] INFO stats.py:71: [92mTime usage[0m: HF loading: 1.868 sec; Pre-quantization mapping: 0.359 sec; Quantization: 2.522 sec | |
| [2023-12-28 23:33:29] INFO stats.py:85: [92mRAM usage[0m: Peak RAM: 2.642 GB. Total bytes loaded from disk: 2.642 GB | |
| [2023-12-28 23:33:29] INFO convert_weight.py:110: [92mParameter size[0m after quantization: 0.744 GB | |
| [2023-12-28 23:33:29] INFO convert_weight.py:115: [92mTotal parameters[0m: 1,418,270,720 | |
| [2023-12-28 23:33:29] INFO convert_weight.py:116: [92mBits per parameter[0m: 4.505 | |
| Start storing to cache /tmp/tmpxe445xtc | |
| [0001/0343] saving transformer.embd.q_weight [0002/0343] saving transformer.embd.q_scale [0003/0343] saving transformer.h.0.ln.weight [0004/0343] saving transformer.h.0.ln.bias [0005/0343] saving transformer.h.0.mixer.Wqkv.q_weight [0006/0343] saving transformer.h.0.mixer.Wqkv.q_scale [0007/0343] saving transformer.h.0.mixer.Wqkv.bias [0008/0343] saving transformer.h.0.mixer.out_proj.q_weight [0009/0343] saving transformer.h.0.mixer.out_proj.q_scale [0010/0343] saving transformer.h.0.mixer.out_proj.bias [0011/0343] saving transformer.h.0.mlp.fc1.q_weight [0012/0343] saving transformer.h.0.mlp.fc1.q_scale [0013/0343] saving transformer.h.0.mlp.fc1.bias [0014/0343] saving transformer.h.0.mlp.fc2.q_weight [0015/0343] saving transformer.h.0.mlp.fc2.q_scale [0016/0343] saving transformer.h.0.mlp.fc2.bias [0017/0343] saving transformer.h.1.ln.weight [0018/0343] saving transformer.h.1.ln.bias [0019/0343] saving transformer.h.1.mixer.Wqkv.q_weight [0020/0343] saving transformer.h.1.mixer.Wqkv.q_scale [0021/0343] saving transformer.h.1.mixer.Wqkv.bias [0022/0343] saving transformer.h.1.mixer.out_proj.q_weight [0023/0343] saving transformer.h.1.mixer.out_proj.q_scale [0024/0343] saving transformer.h.1.mixer.out_proj.bias [0025/0343] saving transformer.h.1.mlp.fc1.q_weight [0026/0343] saving transformer.h.1.mlp.fc1.q_scale [0027/0343] saving transformer.h.1.mlp.fc1.bias [0028/0343] saving transformer.h.1.mlp.fc2.q_weight [0029/0343] saving transformer.h.1.mlp.fc2.q_scale [0030/0343] saving transformer.h.1.mlp.fc2.bias [0031/0343] saving transformer.h.2.ln.weight [0032/0343] saving transformer.h.2.ln.bias [0033/0343] saving transformer.h.2.mixer.Wqkv.q_weight [0034/0343] saving transformer.h.2.mixer.Wqkv.q_scale [0035/0343] saving transformer.h.2.mixer.Wqkv.bias [0036/0343] saving transformer.h.2.mixer.out_proj.q_weight [0037/0343] saving transformer.h.2.mixer.out_proj.q_scale [0038/0343] saving transformer.h.2.mixer.out_proj.bias [0039/0343] saving transformer.h.2.mlp.fc1.q_weight [0040/0343] saving transformer.h.2.mlp.fc1.q_scale [0041/0343] saving transformer.h.2.mlp.fc1.bias [0042/0343] saving transformer.h.2.mlp.fc2.q_weight [0043/0343] saving transformer.h.2.mlp.fc2.q_scale [0044/0343] saving transformer.h.2.mlp.fc2.bias [0045/0343] saving transformer.h.3.ln.weight [0046/0343] saving transformer.h.3.ln.bias [0047/0343] saving transformer.h.3.mixer.Wqkv.q_weight [0048/0343] saving transformer.h.3.mixer.Wqkv.q_scale [0049/0343] saving transformer.h.3.mixer.Wqkv.bias [0050/0343] saving transformer.h.3.mixer.out_proj.q_weight [0051/0343] saving transformer.h.3.mixer.out_proj.q_scale [0052/0343] saving transformer.h.3.mixer.out_proj.bias [0053/0343] saving transformer.h.3.mlp.fc1.q_weight [0054/0343] saving transformer.h.3.mlp.fc1.q_scale [0055/0343] saving transformer.h.3.mlp.fc1.bias [0056/0343] saving transformer.h.3.mlp.fc2.q_weight [0057/0343] saving transformer.h.3.mlp.fc2.q_scale [0058/0343] saving transformer.h.3.mlp.fc2.bias [0059/0343] saving transformer.h.4.ln.weight [0060/0343] saving transformer.h.4.ln.bias [0061/0343] saving transformer.h.4.mixer.Wqkv.q_weight [0062/0343] saving transformer.h.4.mixer.Wqkv.q_scale [0063/0343] saving transformer.h.4.mixer.Wqkv.bias [0064/0343] saving transformer.h.4.mixer.out_proj.q_weight [0065/0343] saving transformer.h.4.mixer.out_proj.q_scale [0066/0343] saving transformer.h.4.mixer.out_proj.bias [0067/0343] saving transformer.h.4.mlp.fc1.q_weight [0068/0343] saving transformer.h.4.mlp.fc1.q_scale [0069/0343] saving transformer.h.4.mlp.fc1.bias [0070/0343] saving transformer.h.4.mlp.fc2.q_weight [0071/0343] saving transformer.h.4.mlp.fc2.q_scale [0072/0343] saving transformer.h.4.mlp.fc2.bias [0073/0343] saving transformer.h.5.ln.weight [0074/0343] saving transformer.h.5.ln.bias [0075/0343] saving transformer.h.5.mixer.Wqkv.q_weight [0076/0343] saving transformer.h.5.mixer.Wqkv.q_scale [0077/0343] saving transformer.h.5.mixer.Wqkv.bias [0078/0343] saving transformer.h.5.mixer.out_proj.q_weight [0079/0343] saving transformer.h.5.mixer.out_proj.q_scale [0080/0343] saving transformer.h.5.mixer.out_proj.bias [0081/0343] saving transformer.h.5.mlp.fc1.q_weight [0082/0343] saving transformer.h.5.mlp.fc1.q_scale [0083/0343] saving transformer.h.5.mlp.fc1.bias [0084/0343] saving transformer.h.5.mlp.fc2.q_weight [0085/0343] saving transformer.h.5.mlp.fc2.q_scale [0086/0343] saving transformer.h.5.mlp.fc2.bias [0087/0343] saving transformer.h.6.ln.weight [0088/0343] saving transformer.h.6.ln.bias [0089/0343] saving transformer.h.6.mixer.Wqkv.q_weight [0090/0343] saving transformer.h.6.mixer.Wqkv.q_scale [0091/0343] saving transformer.h.6.mixer.Wqkv.bias [0092/0343] saving transformer.h.6.mixer.out_proj.q_weight [0093/0343] saving transformer.h.6.mixer.out_proj.q_scale [0094/0343] saving transformer.h.6.mixer.out_proj.bias [0095/0343] saving transformer.h.6.mlp.fc1.q_weight [0096/0343] saving transformer.h.6.mlp.fc1.q_scale [0097/0343] saving transformer.h.6.mlp.fc1.bias [0098/0343] saving transformer.h.6.mlp.fc2.q_weight [0099/0343] saving transformer.h.6.mlp.fc2.q_scale [0100/0343] saving transformer.h.6.mlp.fc2.bias [0101/0343] saving transformer.h.7.ln.weight [0102/0343] saving transformer.h.7.ln.bias [0103/0343] saving transformer.h.7.mixer.Wqkv.q_weight [0104/0343] saving transformer.h.7.mixer.Wqkv.q_scale [0105/0343] saving transformer.h.7.mixer.Wqkv.bias [0106/0343] saving transformer.h.7.mixer.out_proj.q_weight [0107/0343] saving transformer.h.7.mixer.out_proj.q_scale [0108/0343] saving transformer.h.7.mixer.out_proj.bias [0109/0343] saving transformer.h.7.mlp.fc1.q_weight [0110/0343] saving transformer.h.7.mlp.fc1.q_scale [0111/0343] saving transformer.h.7.mlp.fc1.bias [0112/0343] saving transformer.h.7.mlp.fc2.q_weight [0113/0343] saving transformer.h.7.mlp.fc2.q_scale [0114/0343] saving transformer.h.7.mlp.fc2.bias [0115/0343] saving transformer.h.8.ln.weight [0116/0343] saving transformer.h.8.ln.bias [0117/0343] saving transformer.h.8.mixer.Wqkv.q_weight [0118/0343] saving transformer.h.8.mixer.Wqkv.q_scale [0119/0343] saving transformer.h.8.mixer.Wqkv.bias [0120/0343] saving transformer.h.8.mixer.out_proj.q_weight [0121/0343] saving transformer.h.8.mixer.out_proj.q_scale [0122/0343] saving transformer.h.8.mixer.out_proj.bias [0123/0343] saving transformer.h.8.mlp.fc1.q_weight [0124/0343] saving transformer.h.8.mlp.fc1.q_scale [0125/0343] saving transformer.h.8.mlp.fc1.bias [0126/0343] saving transformer.h.8.mlp.fc2.q_weight [0127/0343] saving transformer.h.8.mlp.fc2.q_scale [0128/0343] saving transformer.h.8.mlp.fc2.bias [0129/0343] saving transformer.h.9.ln.weight [0130/0343] saving transformer.h.9.ln.bias [0131/0343] saving transformer.h.9.mixer.Wqkv.q_weight [0132/0343] saving transformer.h.9.mixer.Wqkv.q_scale [0133/0343] saving transformer.h.9.mixer.Wqkv.bias [0134/0343] saving transformer.h.9.mixer.out_proj.q_weight [0135/0343] saving transformer.h.9.mixer.out_proj.q_scale [0136/0343] saving transformer.h.9.mixer.out_proj.bias [0137/0343] saving transformer.h.9.mlp.fc1.q_weight [0138/0343] saving transformer.h.9.mlp.fc1.q_scale [0139/0343] saving transformer.h.9.mlp.fc1.bias [0140/0343] saving transformer.h.9.mlp.fc2.q_weight [0141/0343] saving transformer.h.9.mlp.fc2.q_scale [0142/0343] saving transformer.h.9.mlp.fc2.bias [0143/0343] saving transformer.h.10.ln.weight [0144/0343] saving transformer.h.10.ln.bias [0145/0343] saving transformer.h.10.mixer.Wqkv.q_weight [0146/0343] saving transformer.h.10.mixer.Wqkv.q_scale [0147/0343] saving transformer.h.10.mixer.Wqkv.bias [0148/0343] saving transformer.h.10.mixer.out_proj.q_weight [0149/0343] saving transformer.h.10.mixer.out_proj.q_scale [0150/0343] saving transformer.h.10.mixer.out_proj.bias [0151/0343] saving transformer.h.10.mlp.fc1.q_weight [0152/0343] saving transformer.h.10.mlp.fc1.q_scale [0153/0343] saving transformer.h.10.mlp.fc1.bias [0154/0343] saving transformer.h.10.mlp.fc2.q_weight [0155/0343] saving transformer.h.10.mlp.fc2.q_scale [0156/0343] saving transformer.h.10.mlp.fc2.bias [0157/0343] saving transformer.h.11.ln.weight [0158/0343] saving transformer.h.11.ln.bias [0159/0343] saving transformer.h.11.mixer.Wqkv.q_weight [0160/0343] saving transformer.h.11.mixer.Wqkv.q_scale [0161/0343] saving transformer.h.11.mixer.Wqkv.bias [0162/0343] saving transformer.h.11.mixer.out_proj.q_weight [0163/0343] saving transformer.h.11.mixer.out_proj.q_scale [0164/0343] saving transformer.h.11.mixer.out_proj.bias [0165/0343] saving transformer.h.11.mlp.fc1.q_weight [0166/0343] saving transformer.h.11.mlp.fc1.q_scale [0167/0343] saving transformer.h.11.mlp.fc1.bias [0168/0343] saving transformer.h.11.mlp.fc2.q_weight [0169/0343] saving transformer.h.11.mlp.fc2.q_scale [0170/0343] saving transformer.h.11.mlp.fc2.bias [0171/0343] saving transformer.h.12.ln.weight [0172/0343] saving transformer.h.12.ln.bias [0173/0343] saving transformer.h.12.mixer.Wqkv.q_weight [0174/0343] saving transformer.h.12.mixer.Wqkv.q_scale [0175/0343] saving transformer.h.12.mixer.Wqkv.bias [0176/0343] saving transformer.h.12.mixer.out_proj.q_weight [0177/0343] saving transformer.h.12.mixer.out_proj.q_scale [0178/0343] saving transformer.h.12.mixer.out_proj.bias [0179/0343] saving transformer.h.12.mlp.fc1.q_weight [0180/0343] saving transformer.h.12.mlp.fc1.q_scale [0181/0343] saving transformer.h.12.mlp.fc1.bias [0182/0343] saving transformer.h.12.mlp.fc2.q_weight [0183/0343] saving transformer.h.12.mlp.fc2.q_scale [0184/0343] saving transformer.h.12.mlp.fc2.bias [0185/0343] saving transformer.h.13.ln.weight [0186/0343] saving transformer.h.13.ln.bias [0187/0343] saving transformer.h.13.mixer.Wqkv.q_weight [0188/0343] saving transformer.h.13.mixer.Wqkv.q_scale [0189/0343] saving transformer.h.13.mixer.Wqkv.bias [0190/0343] saving transformer.h.13.mixer.out_proj.q_weight [0191/0343] saving transformer.h.13.mixer.out_proj.q_scale [0192/0343] saving transformer.h.13.mixer.out_proj.bias [0193/0343] saving transformer.h.13.mlp.fc1.q_weight [0194/0343] saving transformer.h.13.mlp.fc1.q_scale [0195/0343] saving transformer.h.13.mlp.fc1.bias [0196/0343] saving transformer.h.13.mlp.fc2.q_weight [0197/0343] saving transformer.h.13.mlp.fc2.q_scale [0198/0343] saving transformer.h.13.mlp.fc2.bias [0199/0343] saving transformer.h.14.ln.weight [0200/0343] saving transformer.h.14.ln.bias [0201/0343] saving transformer.h.14.mixer.Wqkv.q_weight [0202/0343] saving transformer.h.14.mixer.Wqkv.q_scale [0203/0343] saving transformer.h.14.mixer.Wqkv.bias [0204/0343] saving transformer.h.14.mixer.out_proj.q_weight [0205/0343] saving transformer.h.14.mixer.out_proj.q_scale [0206/0343] saving transformer.h.14.mixer.out_proj.bias [0207/0343] saving transformer.h.14.mlp.fc1.q_weight [0208/0343] saving transformer.h.14.mlp.fc1.q_scale [0209/0343] saving transformer.h.14.mlp.fc1.bias [0210/0343] saving transformer.h.14.mlp.fc2.q_weight [0211/0343] saving transformer.h.14.mlp.fc2.q_scale [0212/0343] saving transformer.h.14.mlp.fc2.bias [0213/0343] saving transformer.h.15.ln.weight [0214/0343] saving transformer.h.15.ln.bias [0215/0343] saving transformer.h.15.mixer.Wqkv.q_weight [0216/0343] saving transformer.h.15.mixer.Wqkv.q_scale [0217/0343] saving transformer.h.15.mixer.Wqkv.bias [0218/0343] saving transformer.h.15.mixer.out_proj.q_weight [0219/0343] saving transformer.h.15.mixer.out_proj.q_scale [0220/0343] saving transformer.h.15.mixer.out_proj.bias [0221/0343] saving transformer.h.15.mlp.fc1.q_weight [0222/0343] saving transformer.h.15.mlp.fc1.q_scale [0223/0343] saving transformer.h.15.mlp.fc1.bias [0224/0343] saving transformer.h.15.mlp.fc2.q_weight [0225/0343] saving transformer.h.15.mlp.fc2.q_scale [0226/0343] saving transformer.h.15.mlp.fc2.bias [0227/0343] saving transformer.h.16.ln.weight [0228/0343] saving transformer.h.16.ln.bias [0229/0343] saving transformer.h.16.mixer.Wqkv.q_weight [0230/0343] saving transformer.h.16.mixer.Wqkv.q_scale [0231/0343] saving transformer.h.16.mixer.Wqkv.bias [0232/0343] saving transformer.h.16.mixer.out_proj.q_weight [0233/0343] saving transformer.h.16.mixer.out_proj.q_scale [0234/0343] saving transformer.h.16.mixer.out_proj.bias [0235/0343] saving transformer.h.16.mlp.fc1.q_weight [0236/0343] saving transformer.h.16.mlp.fc1.q_scale [0237/0343] saving transformer.h.16.mlp.fc1.bias [0238/0343] saving transformer.h.16.mlp.fc2.q_weight [0239/0343] saving transformer.h.16.mlp.fc2.q_scale [0240/0343] saving transformer.h.16.mlp.fc2.bias [0241/0343] saving transformer.h.17.ln.weight [0242/0343] saving transformer.h.17.ln.bias [0243/0343] saving transformer.h.17.mixer.Wqkv.q_weight [0244/0343] saving transformer.h.17.mixer.Wqkv.q_scale [0245/0343] saving transformer.h.17.mixer.Wqkv.bias [0246/0343] saving transformer.h.17.mixer.out_proj.q_weight [0247/0343] saving transformer.h.17.mixer.out_proj.q_scale [0248/0343] saving transformer.h.17.mixer.out_proj.bias [0249/0343] saving transformer.h.17.mlp.fc1.q_weight [0250/0343] saving transformer.h.17.mlp.fc1.q_scale [0251/0343] saving transformer.h.17.mlp.fc1.bias [0252/0343] saving transformer.h.17.mlp.fc2.q_weight [0253/0343] saving transformer.h.17.mlp.fc2.q_scale [0254/0343] saving transformer.h.17.mlp.fc2.bias [0255/0343] saving transformer.h.18.ln.weight [0256/0343] saving transformer.h.18.ln.bias [0257/0343] saving transformer.h.18.mixer.Wqkv.q_weight [0258/0343] saving transformer.h.18.mixer.Wqkv.q_scale [0259/0343] saving transformer.h.18.mixer.Wqkv.bias [0260/0343] saving transformer.h.18.mixer.out_proj.q_weight [0261/0343] saving transformer.h.18.mixer.out_proj.q_scale [0262/0343] saving transformer.h.18.mixer.out_proj.bias [0263/0343] saving transformer.h.18.mlp.fc1.q_weight [0264/0343] saving transformer.h.18.mlp.fc1.q_scale [0265/0343] saving transformer.h.18.mlp.fc1.bias [0266/0343] saving transformer.h.18.mlp.fc2.q_weight [0267/0343] saving transformer.h.18.mlp.fc2.q_scale [0268/0343] saving transformer.h.18.mlp.fc2.bias [0269/0343] saving transformer.h.19.ln.weight [0270/0343] saving transformer.h.19.ln.bias [0271/0343] saving transformer.h.19.mixer.Wqkv.q_weight [0272/0343] saving transformer.h.19.mixer.Wqkv.q_scale [0273/0343] saving transformer.h.19.mixer.Wqkv.bias [0274/0343] saving transformer.h.19.mixer.out_proj.q_weight [0275/0343] saving transformer.h.19.mixer.out_proj.q_scale [0276/0343] saving transformer.h.19.mixer.out_proj.bias [0277/0343] saving transformer.h.19.mlp.fc1.q_weight [0278/0343] saving transformer.h.19.mlp.fc1.q_scale [0279/0343] saving transformer.h.19.mlp.fc1.bias [0280/0343] saving transformer.h.19.mlp.fc2.q_weight [0281/0343] saving transformer.h.19.mlp.fc2.q_scale [0282/0343] saving transformer.h.19.mlp.fc2.bias [0283/0343] saving transformer.h.20.ln.weight [0284/0343] saving transformer.h.20.ln.bias [0285/0343] saving transformer.h.20.mixer.Wqkv.q_weight [0286/0343] saving transformer.h.20.mixer.Wqkv.q_scale [0287/0343] saving transformer.h.20.mixer.Wqkv.bias [0288/0343] saving transformer.h.20.mixer.out_proj.q_weight [0289/0343] saving transformer.h.20.mixer.out_proj.q_scale [0290/0343] saving transformer.h.20.mixer.out_proj.bias [0291/0343] saving transformer.h.20.mlp.fc1.q_weight [0292/0343] saving transformer.h.20.mlp.fc1.q_scale[2023-12-28 23:33:30] INFO convert_weight.py:132: Saved to directory: [1m/tmp/tmpxe445xtc[0m | |
| [0293/0343] saving transformer.h.20.mlp.fc1.bias [0294/0343] saving transformer.h.20.mlp.fc2.q_weight [0295/0343] saving transformer.h.20.mlp.fc2.q_scale [0296/0343] saving transformer.h.20.mlp.fc2.bias [0297/0343] saving transformer.h.21.ln.weight [0298/0343] saving transformer.h.21.ln.bias [0299/0343] saving transformer.h.21.mixer.Wqkv.q_weight [0300/0343] saving transformer.h.21.mixer.Wqkv.q_scale [0301/0343] saving transformer.h.21.mixer.Wqkv.bias [0302/0343] saving transformer.h.21.mixer.out_proj.q_weight [0303/0343] saving transformer.h.21.mixer.out_proj.q_scale [0304/0343] saving transformer.h.21.mixer.out_proj.bias [0305/0343] saving transformer.h.21.mlp.fc1.q_weight [0306/0343] saving transformer.h.21.mlp.fc1.q_scale [0307/0343] saving transformer.h.21.mlp.fc1.bias [0308/0343] saving transformer.h.21.mlp.fc2.q_weight [0309/0343] saving transformer.h.21.mlp.fc2.q_scale [0310/0343] saving transformer.h.21.mlp.fc2.bias [0311/0343] saving transformer.h.22.ln.weight [0312/0343] saving transformer.h.22.ln.bias [0313/0343] saving transformer.h.22.mixer.Wqkv.q_weight [0314/0343] saving transformer.h.22.mixer.Wqkv.q_scale [0315/0343] saving transformer.h.22.mixer.Wqkv.bias [0316/0343] saving transformer.h.22.mixer.out_proj.q_weight [0317/0343] saving transformer.h.22.mixer.out_proj.q_scale [0318/0343] saving transformer.h.22.mixer.out_proj.bias [0319/0343] saving transformer.h.22.mlp.fc1.q_weight [0320/0343] saving transformer.h.22.mlp.fc1.q_scale [0321/0343] saving transformer.h.22.mlp.fc1.bias [0322/0343] saving transformer.h.22.mlp.fc2.q_weight [0323/0343] saving transformer.h.22.mlp.fc2.q_scale [0324/0343] saving transformer.h.22.mlp.fc2.bias [0325/0343] saving transformer.h.23.ln.weight [0326/0343] saving transformer.h.23.ln.bias [0327/0343] saving transformer.h.23.mixer.Wqkv.q_weight [0328/0343] saving transformer.h.23.mixer.Wqkv.q_scale [0329/0343] saving transformer.h.23.mixer.Wqkv.bias [0330/0343] saving transformer.h.23.mixer.out_proj.q_weight [0331/0343] saving transformer.h.23.mixer.out_proj.q_scale [0332/0343] saving transformer.h.23.mixer.out_proj.bias [0333/0343] saving transformer.h.23.mlp.fc1.q_weight [0334/0343] saving transformer.h.23.mlp.fc1.q_scale [0335/0343] saving transformer.h.23.mlp.fc1.bias [0336/0343] saving transformer.h.23.mlp.fc2.q_weight [0337/0343] saving transformer.h.23.mlp.fc2.q_scale [0338/0343] saving transformer.h.23.mlp.fc2.bias [0339/0343] saving lm_head.ln.weight [0340/0343] saving lm_head.ln.bias [0341/0343] saving lm_head.linear.q_weight [0342/0343] saving lm_head.linear.q_scale [0343/0343] saving lm_head.linear.bias | |
| All finished, 27 total shards committed, record saved to /tmp/tmpxe445xtc/ndarray-cache.json | |