/opt/conda/envs/py310/bin/python -m mlc_llm gen_config /models/Qwen1.5-0.5B-Chat --quantization q4f32_1 --conv-template chatml --output /models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC
[2024-06-04 03:31:01] INFO auto_config.py:116: [92mFound[0m model configuration: /models/Qwen1.5-0.5B-Chat/config.json
[2024-06-04 03:31:01] INFO auto_config.py:154: [92mFound[0m model type: [1mqwen2[0m. Use `--model-type` to override.
[2024-06-04 03:31:01] INFO qwen2_model.py:49: [1mcontext_window_size[0m not found in config.json. Falling back to [1mmax_position_embeddings[0m (32768)
[2024-06-04 03:31:01] INFO qwen2_model.py:66: [1mprefill_chunk_size[0m defaults to 2048
[2024-06-04 03:31:01] INFO config.py:107: Overriding [1mmax_batch_size[0m from 1 to 80
[2024-06-04 03:31:01] INFO gen_config.py:143: [generation_config.json] Setting [1mbos_token_id[0m: 151643
[2024-06-04 03:31:01] INFO gen_config.py:143: [generation_config.json] Setting [1mpad_token_id[0m: 151643
[2024-06-04 03:31:01] INFO gen_config.py:143: [generation_config.json] Setting [1meos_token_id[0m: [151645, 151643]
[2024-06-04 03:31:01] INFO gen_config.py:143: [generation_config.json] Setting [1mrepetition_penalty[0m: 1.1
[2024-06-04 03:31:01] INFO gen_config.py:143: [generation_config.json] Setting [1mtop_p[0m: 0.8
[2024-06-04 03:31:01] INFO gen_config.py:157: [91mNot found[0m tokenizer config: /models/Qwen1.5-0.5B-Chat/tokenizer.model
[2024-06-04 03:31:01] INFO gen_config.py:155: [92mFound[0m tokenizer config: /models/Qwen1.5-0.5B-Chat/tokenizer.json. Copying to [1m/models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/tokenizer.json[0m
[2024-06-04 03:31:01] INFO gen_config.py:155: [92mFound[0m tokenizer config: /models/Qwen1.5-0.5B-Chat/vocab.json. Copying to [1m/models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/vocab.json[0m
[2024-06-04 03:31:01] INFO gen_config.py:155: [92mFound[0m tokenizer config: /models/Qwen1.5-0.5B-Chat/merges.txt. Copying to [1m/models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/merges.txt[0m
[2024-06-04 03:31:01] INFO gen_config.py:157: [91mNot found[0m tokenizer config: /models/Qwen1.5-0.5B-Chat/added_tokens.json
[2024-06-04 03:31:01] INFO gen_config.py:155: [92mFound[0m tokenizer config: /models/Qwen1.5-0.5B-Chat/tokenizer_config.json. Copying to [1m/models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/tokenizer_config.json[0m
[2024-06-04 03:31:01] INFO gen_config.py:216: Detected tokenizer info: {'token_postproc_method': 'byte_level', 'prepend_space_in_encode': False, 'strip_space_in_decode': False}
[2024-06-04 03:31:01] INFO gen_config.py:32: [System default] Setting [1mtemperature[0m: 1.0
[2024-06-04 03:31:01] INFO gen_config.py:32: [System default] Setting [1mpresence_penalty[0m: 0.0
[2024-06-04 03:31:01] INFO gen_config.py:32: [System default] Setting [1mfrequency_penalty[0m: 0.0
[2024-06-04 03:31:01] INFO gen_config.py:32: [System default] Setting [1mmean_gen_len[0m: 128
[2024-06-04 03:31:01] INFO gen_config.py:32: [System default] Setting [1mmax_gen_len[0m: 512
[2024-06-04 03:31:01] INFO gen_config.py:32: [System default] Setting [1mshift_fill_factor[0m: 0.3
[2024-06-04 03:31:01] INFO gen_config.py:223: Dumping configuration file to: [1m/models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/mlc-chat-config.json[0m
/opt/conda/envs/py310/bin/python -m mlc_llm convert_weight /models/Qwen1.5-0.5B-Chat --quantization q4f32_1 --output /models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC
[2024-06-04 03:31:03] INFO auto_config.py:116: [92mFound[0m model configuration: /models/Qwen1.5-0.5B-Chat/config.json
[2024-06-04 03:31:04] INFO auto_device.py:79: [92mFound[0m device: cuda:0
[2024-06-04 03:31:06] INFO auto_device.py:88: [91mNot found[0m device: rocm:0
[2024-06-04 03:31:07] INFO auto_device.py:88: [91mNot found[0m device: metal:0
[2024-06-04 03:31:09] INFO auto_device.py:79: [92mFound[0m device: vulkan:0
[2024-06-04 03:31:09] INFO auto_device.py:79: [92mFound[0m device: vulkan:1
[2024-06-04 03:31:09] INFO auto_device.py:79: [92mFound[0m device: vulkan:2
[2024-06-04 03:31:09] INFO auto_device.py:79: [92mFound[0m device: vulkan:3
[2024-06-04 03:31:10] INFO auto_device.py:88: [91mNot found[0m device: opencl:0
[2024-06-04 03:31:10] INFO auto_device.py:35: Using device: [1mcuda:0[0m
[2024-06-04 03:31:10] INFO auto_weight.py:71: Finding weights in: /models/Qwen1.5-0.5B-Chat
[2024-06-04 03:31:10] INFO auto_weight.py:137: [91mNot found[0m Huggingface PyTorch
[2024-06-04 03:31:11] INFO auto_weight.py:161: [92mFound[0m source weight format: huggingface-safetensor. Source configuration: /models/Qwen1.5-0.5B-Chat/model.safetensors.index.json
[2024-06-04 03:31:11] INFO auto_weight.py:107: Using source weight configuration: [1m/models/Qwen1.5-0.5B-Chat/model.safetensors.index.json[0m. Use `--source` to override.
[2024-06-04 03:31:11] INFO auto_weight.py:111: Using source weight format: [1mhuggingface-safetensor[0m. Use `--source-format` to override.
[2024-06-04 03:31:11] INFO auto_config.py:154: [92mFound[0m model type: [1mqwen2[0m. Use `--model-type` to override.
[2024-06-04 03:31:11] INFO qwen2_model.py:49: [1mcontext_window_size[0m not found in config.json. Falling back to [1mmax_position_embeddings[0m (32768)
[2024-06-04 03:31:11] INFO qwen2_model.py:66: [1mprefill_chunk_size[0m defaults to 2048
[1mWeight conversion with arguments:[0m
[1m--config[0m /models/Qwen1.5-0.5B-Chat/config.json
[1m--quantization[0m GroupQuantize(name='q4f32_1', kind='group-quant', group_size=32, quantize_dtype='int4', storage_dtype='uint32', model_dtype='float32', linear_weight_layout='NK', quantize_embedding=True, quantize_final_fc=True, num_elem_per_storage=8, num_storage_per_group=4, max_int_value=7)
[1m--model-type[0m qwen2
[1m--device[0m cuda:0
[1m--source[0m /models/Qwen1.5-0.5B-Chat/model.safetensors.index.json
[1m--source-format[0m huggingface-safetensor
[1m--output[0m /models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC
Start storing to cache /models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC
0%| | 0/171 [00:00<?, ?it/s]
[2024-06-04 03:31:13] INFO huggingface_loader.py:185: Loading HF parameters from: /models/Qwen1.5-0.5B-Chat/model.safetensors
0%| | 0/171 [00:00<?, ?it/s]
[2024-06-04 03:31:15] INFO group_quantization.py:217: Compiling quantize function for key: ((151936, 1024), float32, cuda, axis=1, output_transpose=False)
0%| | 0/171 [00:02<?, ?it/s]
[2024-06-04 03:31:16] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mlm_head.q_weight[0m", shape: (151936, 128), dtype: uint32
0%| | 0/171 [00:02<?, ?it/s]
[2024-06-04 03:31:16] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mlm_head.q_scale[0m", shape: (151936, 32), dtype: float32
0%| | 0/171 [00:03<?, ?it/s]
1%| | 1/171 [00:03<09:04, 3.20s/it]
[2024-06-04 03:31:16] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.embed_tokens.q_weight[0m", shape: (151936, 128), dtype: uint32
1%| | 1/171 [00:03<09:04, 3.20s/it]
[2024-06-04 03:31:16] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.embed_tokens.q_scale[0m", shape: (151936, 32), dtype: float32
1%| | 1/171 [00:03<09:04, 3.20s/it]
1%| | 2/171 [00:03<04:24, 1.56s/it]
[2024-06-04 03:31:16] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.0.input_layernorm.weight[0m", shape: (1024,), dtype: float32
1%| | 2/171 [00:03<04:24, 1.56s/it]
[2024-06-04 03:31:16] INFO group_quantization.py:217: Compiling quantize function for key: ((1024, 2816), float32, cuda, axis=1, output_transpose=False)
1%| | 2/171 [00:03<04:24, 1.56s/it]
[2024-06-04 03:31:17] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
1%| | 2/171 [00:04<04:24, 1.56s/it]
[2024-06-04 03:31:17] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
1%| | 2/171 [00:04<04:24, 1.56s/it]
2%|▏ | 4/171 [00:04<02:01, 1.38it/s]
[2024-06-04 03:31:17] INFO group_quantization.py:217: Compiling quantize function for key: ((5632, 1024), float32, cuda, axis=1, output_transpose=False)
2%|▏ | 4/171 [00:04<02:01, 1.38it/s]
[2024-06-04 03:31:17] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
2%|▏ | 4/171 [00:04<02:01, 1.38it/s]
[2024-06-04 03:31:17] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
2%|▏ | 4/171 [00:04<02:01, 1.38it/s]
3%|▎ | 5/171 [00:04<01:44, 1.58it/s]
[2024-06-04 03:31:17] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.0.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
3%|▎ | 5/171 [00:04<01:44, 1.58it/s]
[2024-06-04 03:31:17] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.0.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
3%|▎ | 5/171 [00:04<01:44, 1.58it/s]
[2024-06-04 03:31:17] INFO group_quantization.py:217: Compiling quantize function for key: ((3072, 1024), float32, cuda, axis=1, output_transpose=False)
3%|▎ | 5/171 [00:04<01:44, 1.58it/s]
[2024-06-04 03:31:17] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
3%|▎ | 5/171 [00:04<01:44, 1.58it/s]
[2024-06-04 03:31:17] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
3%|▎ | 5/171 [00:04<01:44, 1.58it/s]
5%|▍ | 8/171 [00:04<00:57, 2.84it/s]
[2024-06-04 03:31:17] INFO group_quantization.py:217: Compiling quantize function for key: ((1024, 1024), float32, cuda, axis=1, output_transpose=False)
5%|▍ | 8/171 [00:04<00:57, 2.84it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
5%|▍ | 8/171 [00:05<00:57, 2.84it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.0.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
5%|▍ | 8/171 [00:05<00:57, 2.84it/s]
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.1.input_layernorm.weight[0m", shape: (1024,), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.1.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.1.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.1.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.10.input_layernorm.weight[0m", shape: (1024,), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
5%|▌ | 9/171 [00:05<00:59, 2.72it/s]
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.10.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.10.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.10.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.11.input_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.11.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.11.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.11.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.12.input_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.12.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.12.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.12.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.13.input_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.13.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.13.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.13.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
11%|█ | 19/171 [00:05<00:15, 9.64it/s]
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.14.input_layernorm.weight[0m", shape: (1024,), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.14.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.14.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.14.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.15.input_layernorm.weight[0m", shape: (1024,), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
26%|██▌ | 44/171 [00:05<00:03, 32.21it/s]
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.15.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.15.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.15.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.16.input_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.16.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.16.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.16.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.17.input_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.17.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.17.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.17.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.18.input_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.18.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.18.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
32%|███▏ | 54/171 [00:05<00:02, 39.74it/s]
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.18.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.19.input_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.19.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.19.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.19.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.2.input_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:18] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.2.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.2.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.2.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.20.input_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.20.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.20.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.20.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.21.input_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.21.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.21.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.21.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.22.input_layernorm.weight[0m", shape: (1024,), dtype: float32
46%|████▌ | 78/171 [00:05<00:01, 62.38it/s]
63%|██████▎ | 108/171 [00:05<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
63%|██████▎ | 108/171 [00:05<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
63%|██████▎ | 108/171 [00:05<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.22.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.22.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.22.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.23.input_layernorm.weight[0m", shape: (1024,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.23.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.23.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.23.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.3.input_layernorm.weight[0m", shape: (1024,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.3.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
63%|██████▎ | 108/171 [00:06<00:00, 99.69it/s]
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.3.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.3.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.4.input_layernorm.weight[0m", shape: (1024,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.4.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.4.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.4.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.5.input_layernorm.weight[0m", shape: (1024,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.5.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.5.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
73%|███████▎ | 125/171 [00:06<00:00, 106.34it/s]
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.5.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.6.input_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.6.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.6.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.6.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.7.input_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.7.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.7.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.7.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.8.input_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.8.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.8.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.8.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.9.input_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.mlp.down_proj.q_weight[0m", shape: (1024, 352), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.mlp.down_proj.q_scale[0m", shape: (1024, 88), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.mlp.gate_up_proj.q_weight[0m", shape: (5632, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.mlp.gate_up_proj.q_scale[0m", shape: (5632, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.9.post_attention_layernorm.weight[0m", shape: (1024,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.layers.9.self_attn.c_attn.bias[0m", shape: (3072,), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.self_attn.c_attn.q_weight[0m", shape: (3072, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.self_attn.c_attn.q_scale[0m", shape: (3072, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.self_attn.o_proj.q_weight[0m", shape: (1024, 128), dtype: uint32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:167: [Quantized] Parameter: "[1mmodel.layers.9.self_attn.o_proj.q_scale[0m", shape: (1024, 32), dtype: float32
82%|████████▏ | 141/171 [00:06<00:00, 112.83it/s]
99%|█████████▉| 170/171 [00:06<00:00, 150.62it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:175: [Not quantized] Parameter: "[1mmodel.norm.weight[0m", shape: (1024,), dtype: float32
99%|█████████▉| 170/171 [00:06<00:00, 150.62it/s]
100%|██████████| 171/171 [00:06<00:00, 26.94it/s]
[2024-06-04 03:31:19] INFO huggingface_loader.py:197: Unloading HF weight file: /models/Qwen1.5-0.5B-Chat/model.safetensors
[2024-06-04 03:31:19] INFO stats.py:77: [92mTime usage[0m: HF loading: 1.637 sec; Pre-quantization mapping: 0.799 sec; Quantization: 2.662 sec
[2024-06-04 03:31:19] INFO stats.py:91: [92mRAM usage[0m: Peak RAM: 2.308 GB. Total bytes loaded from disk: 2.308 GB
[2024-06-04 03:31:19] INFO convert_weight.py:155: [92mParameter size[0m after quantization: 0.361 GB
[2024-06-04 03:31:19] INFO convert_weight.py:160: [92mTotal parameters[0m: 619,570,176
[2024-06-04 03:31:19] INFO convert_weight.py:161: [92mBits per parameter[0m: 5.005
[2024-06-04 03:31:19] INFO convert_weight.py:166: Saved to directory: [1m/models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC[0m
All finished, 8 total shards committed, record saved to /models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/ndarray-cache.json
Also saved a bf16 record to /models/mlc-delivery/hf/mlc-ai/Qwen1.5-0.5B-Chat-q4f32_1-MLC/ndarray-cache-b16.json