/usr/bin/python3 -m mlc_llm gen_config dist/models/stablelm-2-zephyr-1_6b --quantization q4f32_1 --conv-template stablelm-2 --output /tmp/tmpt74_sq0w --context-window-size 4096 [2024-05-24 18:29:33] INFO auto_config.py:115: Found model configuration: dist/models/stablelm-2-zephyr-1_6b/config.json [2024-05-24 18:29:33] INFO auto_config.py:153: Found model type: stablelm. Use `--model-type` to override. [2024-05-24 18:29:33] INFO stablelm_model.py:49: context_window_size not found in config.json. Falling back to max_position_embeddings (4096) [2024-05-24 18:29:33] INFO stablelm_model.py:66: prefill_chunk_size defaults to 2048 [2024-05-24 18:29:33] INFO config.py:106: Overriding context_window_size from 4096 to 4096 [2024-05-24 18:29:33] INFO config.py:106: Overriding max_batch_size from 1 to 80 [2024-05-24 18:29:33] INFO gen_config.py:255: [generation_config.json] Setting bos_token_id: 100257 [2024-05-24 18:29:33] INFO gen_config.py:255: [generation_config.json] Setting eos_token_id: 100257 [2024-05-24 18:29:33] INFO gen_config.py:269: Not found tokenizer config: dist/models/stablelm-2-zephyr-1_6b/tokenizer.model [2024-05-24 18:29:33] INFO gen_config.py:267: Found tokenizer config: dist/models/stablelm-2-zephyr-1_6b/tokenizer.json. Copying to /tmp/tmpt74_sq0w/tokenizer.json [2024-05-24 18:29:33] INFO gen_config.py:267: Found tokenizer config: dist/models/stablelm-2-zephyr-1_6b/vocab.json. Copying to /tmp/tmpt74_sq0w/vocab.json [2024-05-24 18:29:33] INFO gen_config.py:267: Found tokenizer config: dist/models/stablelm-2-zephyr-1_6b/merges.txt. Copying to /tmp/tmpt74_sq0w/merges.txt [2024-05-24 18:29:33] INFO gen_config.py:269: Not found tokenizer config: dist/models/stablelm-2-zephyr-1_6b/added_tokens.json [2024-05-24 18:29:33] INFO gen_config.py:267: Found tokenizer config: dist/models/stablelm-2-zephyr-1_6b/tokenizer_config.json. Copying to /tmp/tmpt74_sq0w/tokenizer_config.json [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting pad_token_id: 0 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting temperature: 0.7 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting presence_penalty: 0.0 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting frequency_penalty: 0.0 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting repetition_penalty: 1.0 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting top_p: 0.95 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting mean_gen_len: 128 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting max_gen_len: 512 [2024-05-24 18:29:33] INFO gen_config.py:80: [System default] Setting shift_fill_factor: 0.3 [2024-05-24 18:29:33] INFO gen_config.py:335: Dumping configuration file to: /tmp/tmpt74_sq0w/mlc-chat-config.json /usr/bin/python3 -m mlc_llm convert_weight dist/models/stablelm-2-zephyr-1_6b --quantization q4f32_1 --source-format auto --output /tmp/tmpt74_sq0w [2024-05-24 18:29:34] INFO auto_config.py:115: Found model configuration: dist/models/stablelm-2-zephyr-1_6b/config.json [2024-05-24 18:29:36] INFO auto_device.py:79: Found device: cuda:0 [2024-05-24 18:29:37] INFO auto_device.py:88: Not found device: rocm:0 [2024-05-24 18:29:38] INFO auto_device.py:88: Not found device: metal:0 [2024-05-24 18:29:39] INFO auto_device.py:88: Not found device: vulkan:0 [2024-05-24 18:29:40] INFO auto_device.py:88: Not found device: opencl:0 [2024-05-24 18:29:40] INFO auto_device.py:35: Using device: cuda:0 [2024-05-24 18:29:40] INFO auto_weight.py:70: Finding weights in: dist/models/stablelm-2-zephyr-1_6b [2024-05-24 18:29:40] INFO auto_weight.py:136: Not found Huggingface PyTorch [2024-05-24 18:29:40] INFO auto_weight.py:143: Found source weight format: huggingface-safetensor. Source configuration: dist/models/stablelm-2-zephyr-1_6b/model.safetensors.index.json [2024-05-24 18:29:40] INFO auto_weight.py:106: Using source weight configuration: dist/models/stablelm-2-zephyr-1_6b/model.safetensors.index.json. Use `--source` to override. [2024-05-24 18:29:40] INFO auto_weight.py:110: Using source weight format: huggingface-safetensor. Use `--source-format` to override. [2024-05-24 18:29:40] INFO auto_config.py:153: Found model type: stablelm. Use `--model-type` to override. [2024-05-24 18:29:40] INFO stablelm_model.py:49: context_window_size not found in config.json. Falling back to max_position_embeddings (4096) [2024-05-24 18:29:40] INFO stablelm_model.py:66: prefill_chunk_size defaults to 2048 Weight conversion with arguments: --config dist/models/stablelm-2-zephyr-1_6b/config.json --quantization 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) --model-type stablelm --device cuda:0 --source dist/models/stablelm-2-zephyr-1_6b/model.safetensors.index.json --source-format huggingface-safetensor --output /tmp/tmpt74_sq0w Start storing to cache /tmp/tmpt74_sq0w 0%| | 0/220 [00:00 type is zero. setattr(self, word, getattr(machar, word).flat[0]) /home/tlopex/.local/lib/python3.8/site-packages/numpy/core/getlimits.py:89: UserWarning: The value of the smallest subnormal for type is zero. return self._float_to_str(self.smallest_subnormal) 0%| | 1/220 [00:03<12:53, 3.53s/it] [2024-05-24 18:29:47] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.embed_tokens.q_weight", shape: (100352, 256), dtype: uint32 0%| | 1/220 [00:04<12:53, 3.53s/it] [2024-05-24 18:29:47] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.embed_tokens.q_scale", shape: (100352, 64), dtype: float32 0%| | 1/220 [00:04<12:53, 3.53s/it] 1%| | 2/220 [00:04<07:50, 2.16s/it] [2024-05-24 18:29:47] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.0.input_layernorm.bias", shape: (2048,), dtype: float32 1%| | 2/220 [00:04<07:50, 2.16s/it] [2024-05-24 18:29:47] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.0.input_layernorm.weight", shape: (2048,), dtype: float32 1%| | 2/220 [00:04<07:50, 2.16s/it] [2024-05-24 18:29:47] INFO group_quantization.py:217: Compiling quantize function for key: ((2048, 5632), float32, cuda, axis=1, output_transpose=False) 1%| | 2/220 [00:04<07:50, 2.16s/it] [2024-05-24 18:29:48] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 1%| | 2/220 [00:05<07:50, 2.16s/it] [2024-05-24 18:29:48] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 1%| | 2/220 [00:05<07:50, 2.16s/it] 2%|▏ | 5/220 [00:05<02:28, 1.45it/s] [2024-05-24 18:29:48] INFO group_quantization.py:217: Compiling quantize function for key: ((11264, 2048), float32, cuda, axis=1, output_transpose=False) 2%|▏ | 5/220 [00:05<02:28, 1.45it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 2%|▏ | 5/220 [00:05<02:28, 1.45it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 2%|▏ | 5/220 [00:05<02:28, 1.45it/s] 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.0.post_attention_layernorm.bias", shape: (2048,), dtype: float32 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.0.post_attention_layernorm.weight", shape: (2048,), dtype: float32 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.0.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] [2024-05-24 18:29:48] INFO group_quantization.py:217: Compiling quantize function for key: ((6144, 2048), float32, cuda, axis=1, output_transpose=False) 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] [2024-05-24 18:29:48] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 3%|▎ | 6/220 [00:05<02:09, 1.65it/s] 5%|▍ | 10/220 [00:05<01:00, 3.46it/s] [2024-05-24 18:29:48] INFO group_quantization.py:217: Compiling quantize function for key: ((2048, 2048), float32, cuda, axis=1, output_transpose=False) 5%|▍ | 10/220 [00:05<01:00, 3.46it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 5%|▍ | 10/220 [00:05<01:00, 3.46it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.0.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 5%|▍ | 10/220 [00:05<01:00, 3.46it/s] 5%|▌ | 11/220 [00:05<00:59, 3.54it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.1.input_layernorm.bias", shape: (2048,), dtype: float32 5%|▌ | 11/220 [00:05<00:59, 3.54it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.1.input_layernorm.weight", shape: (2048,), dtype: float32 5%|▌ | 11/220 [00:05<00:59, 3.54it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 5%|▌ | 11/220 [00:05<00:59, 3.54it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 5%|▌ | 11/220 [00:05<00:59, 3.54it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 5%|▌ | 11/220 [00:06<00:59, 3.54it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 5%|▌ | 11/220 [00:06<00:59, 3.54it/s] 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.1.post_attention_layernorm.bias", shape: (2048,), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.1.post_attention_layernorm.weight", shape: (2048,), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.1.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.1.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.10.input_layernorm.bias", shape: (2048,), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.10.input_layernorm.weight", shape: (2048,), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 7%|▋ | 15/220 [00:06<00:32, 6.24it/s] 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.10.post_attention_layernorm.bias", shape: (2048,), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.10.post_attention_layernorm.weight", shape: (2048,), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.10.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.10.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.11.input_layernorm.bias", shape: (2048,), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.11.input_layernorm.weight", shape: (2048,), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 11%|█ | 24/220 [00:06<00:14, 13.50it/s] 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.11.post_attention_layernorm.bias", shape: (2048,), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.11.post_attention_layernorm.weight", shape: (2048,), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.11.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.11.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.12.input_layernorm.bias", shape: (2048,), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.12.input_layernorm.weight", shape: (2048,), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 15%|█▌ | 33/220 [00:06<00:09, 20.58it/s] 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.12.post_attention_layernorm.bias", shape: (2048,), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.12.post_attention_layernorm.weight", shape: (2048,), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.12.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.12.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.13.input_layernorm.bias", shape: (2048,), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.13.input_layernorm.weight", shape: (2048,), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 19%|█▉ | 42/220 [00:06<00:06, 27.17it/s] 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.13.post_attention_layernorm.bias", shape: (2048,), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.13.post_attention_layernorm.weight", shape: (2048,), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.13.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.13.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.14.input_layernorm.bias", shape: (2048,), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.14.input_layernorm.weight", shape: (2048,), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:49] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 23%|██▎ | 51/220 [00:06<00:05, 32.93it/s] 27%|██▋ | 60/220 [00:06<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.14.post_attention_layernorm.bias", shape: (2048,), dtype: float32 27%|██▋ | 60/220 [00:06<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.14.post_attention_layernorm.weight", shape: (2048,), dtype: float32 27%|██▋ | 60/220 [00:06<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.14.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 27%|██▋ | 60/220 [00:06<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.14.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.15.input_layernorm.bias", shape: (2048,), dtype: float32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.15.input_layernorm.weight", shape: (2048,), dtype: float32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 27%|██▋ | 60/220 [00:07<00:04, 37.69it/s] 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.15.post_attention_layernorm.bias", shape: (2048,), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.15.post_attention_layernorm.weight", shape: (2048,), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.15.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.15.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.16.input_layernorm.bias", shape: (2048,), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.16.input_layernorm.weight", shape: (2048,), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 31%|███▏ | 69/220 [00:07<00:03, 41.52it/s] 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.16.post_attention_layernorm.bias", shape: (2048,), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.16.post_attention_layernorm.weight", shape: (2048,), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.16.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.16.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.17.input_layernorm.bias", shape: (2048,), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.17.input_layernorm.weight", shape: (2048,), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 35%|███▌ | 78/220 [00:07<00:03, 44.44it/s] 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.17.post_attention_layernorm.bias", shape: (2048,), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.17.post_attention_layernorm.weight", shape: (2048,), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.17.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.17.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.18.input_layernorm.bias", shape: (2048,), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.18.input_layernorm.weight", shape: (2048,), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 40%|███▉ | 87/220 [00:07<00:02, 46.64it/s] 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.18.post_attention_layernorm.bias", shape: (2048,), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.18.post_attention_layernorm.weight", shape: (2048,), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.18.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.18.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.19.input_layernorm.bias", shape: (2048,), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.19.input_layernorm.weight", shape: (2048,), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 44%|████▎ | 96/220 [00:07<00:02, 48.27it/s] 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.19.post_attention_layernorm.bias", shape: (2048,), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.19.post_attention_layernorm.weight", shape: (2048,), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.19.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:50] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.19.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.2.input_layernorm.bias", shape: (2048,), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.2.input_layernorm.weight", shape: (2048,), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 48%|████▊ | 105/220 [00:07<00:02, 49.46it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 48%|████▊ | 105/220 [00:08<00:02, 49.46it/s] 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.2.post_attention_layernorm.bias", shape: (2048,), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.2.post_attention_layernorm.weight", shape: (2048,), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.2.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.2.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.20.input_layernorm.bias", shape: (2048,), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.20.input_layernorm.weight", shape: (2048,), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 52%|█████▏ | 114/220 [00:08<00:02, 50.30it/s] 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.20.post_attention_layernorm.bias", shape: (2048,), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.20.post_attention_layernorm.weight", shape: (2048,), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.20.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.20.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.21.input_layernorm.bias", shape: (2048,), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.21.input_layernorm.weight", shape: (2048,), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 56%|█████▌ | 123/220 [00:08<00:01, 50.90it/s] 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.21.post_attention_layernorm.bias", shape: (2048,), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.21.post_attention_layernorm.weight", shape: (2048,), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.21.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.21.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.22.input_layernorm.bias", shape: (2048,), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.22.input_layernorm.weight", shape: (2048,), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 60%|██████ | 132/220 [00:08<00:01, 51.32it/s] 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.22.post_attention_layernorm.bias", shape: (2048,), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.22.post_attention_layernorm.weight", shape: (2048,), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.22.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.22.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.23.input_layernorm.bias", shape: (2048,), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.23.input_layernorm.weight", shape: (2048,), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 64%|██████▍ | 141/220 [00:08<00:01, 51.60it/s] 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.23.post_attention_layernorm.bias", shape: (2048,), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.23.post_attention_layernorm.weight", shape: (2048,), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.23.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.23.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.3.input_layernorm.bias", shape: (2048,), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.3.input_layernorm.weight", shape: (2048,), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] [2024-05-24 18:29:51] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 68%|██████▊ | 150/220 [00:08<00:01, 51.82it/s] 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.3.post_attention_layernorm.bias", shape: (2048,), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.3.post_attention_layernorm.weight", shape: (2048,), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.3.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.3.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.4.input_layernorm.bias", shape: (2048,), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.4.input_layernorm.weight", shape: (2048,), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 72%|███████▏ | 159/220 [00:08<00:01, 51.97it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 72%|███████▏ | 159/220 [00:09<00:01, 51.97it/s] 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.4.post_attention_layernorm.bias", shape: (2048,), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.4.post_attention_layernorm.weight", shape: (2048,), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.4.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.4.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.5.input_layernorm.bias", shape: (2048,), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.5.input_layernorm.weight", shape: (2048,), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 76%|███████▋ | 168/220 [00:09<00:00, 52.08it/s] 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.5.post_attention_layernorm.bias", shape: (2048,), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.5.post_attention_layernorm.weight", shape: (2048,), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.5.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.5.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.6.input_layernorm.bias", shape: (2048,), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.6.input_layernorm.weight", shape: (2048,), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 80%|████████ | 177/220 [00:09<00:00, 52.14it/s] 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.6.post_attention_layernorm.bias", shape: (2048,), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.6.post_attention_layernorm.weight", shape: (2048,), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.6.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.6.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.7.input_layernorm.bias", shape: (2048,), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.7.input_layernorm.weight", shape: (2048,), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 85%|████████▍ | 186/220 [00:09<00:00, 52.23it/s] 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.7.post_attention_layernorm.bias", shape: (2048,), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.7.post_attention_layernorm.weight", shape: (2048,), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.7.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.7.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.8.input_layernorm.bias", shape: (2048,), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.8.input_layernorm.weight", shape: (2048,), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 89%|████████▊ | 195/220 [00:09<00:00, 52.25it/s] 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.8.post_attention_layernorm.bias", shape: (2048,), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.8.post_attention_layernorm.weight", shape: (2048,), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.8.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.8.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.9.input_layernorm.bias", shape: (2048,), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.9.input_layernorm.weight", shape: (2048,), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.mlp.down_proj.q_weight", shape: (2048, 704), dtype: uint32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.mlp.down_proj.q_scale", shape: (2048, 176), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:52] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.mlp.gate_up_proj.q_weight", shape: (11264, 256), dtype: uint32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.mlp.gate_up_proj.q_scale", shape: (11264, 64), dtype: float32 93%|█████████▎| 204/220 [00:09<00:00, 52.28it/s] 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.9.post_attention_layernorm.bias", shape: (2048,), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.9.post_attention_layernorm.weight", shape: (2048,), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.layers.9.self_attn.qkv_proj.bias", shape: (6144,), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.self_attn.qkv_proj.q_weight", shape: (6144, 256), dtype: uint32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.self_attn.qkv_proj.q_scale", shape: (6144, 64), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.self_attn.o_proj.q_weight", shape: (2048, 256), dtype: uint32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:166: [Quantized] Parameter: "model.layers.9.self_attn.o_proj.q_scale", shape: (2048, 64), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.norm.bias", shape: (2048,), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:174: [Not quantized] Parameter: "model.norm.weight", shape: (2048,), dtype: float32 97%|█████████▋| 213/220 [00:09<00:00, 52.30it/s] 100%|██████████| 220/220 [00:09<00:00, 22.13it/s] [2024-05-24 18:29:53] INFO huggingface_loader.py:196: Unloading HF weight file: dist/models/stablelm-2-zephyr-1_6b/model.safetensors [2024-05-24 18:29:53] INFO stats.py:76: Time usage: HF loading: 1.007 sec; Pre-quantization mapping: 3.312 sec; Quantization: 2.344 sec [2024-05-24 18:29:53] INFO stats.py:90: RAM usage: Peak RAM: 3.063 GB. Total bytes loaded from disk: 3.063 GB [2024-05-24 18:29:53] INFO convert_weight.py:155: Parameter size after quantization: 0.958 GB [2024-05-24 18:29:53] INFO convert_weight.py:160: Total parameters: 1,644,515,328 [2024-05-24 18:29:53] INFO convert_weight.py:161: Bits per parameter: 5.006 [2024-05-24 18:29:53] INFO convert_weight.py:166: Saved to directory: /tmp/tmpt74_sq0w All finished, 27 total shards committed, record saved to /tmp/tmpt74_sq0w/ndarray-cache.json Also saved a bf16 record to /tmp/tmpt74_sq0w/ndarray-cache-b16.json