diff --git a/.gitattributes b/.gitattributes
index a6344aac8c09253b3b630fb776ae94478aa0275b..52373fe24473b1aa44333d318f578ae6bf04b49b 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/logs.txt b/logs.txt
new file mode 100644
index 0000000000000000000000000000000000000000..433341121083e09f5d31a853860e6e35988f9cab
--- /dev/null
+++ b/logs.txt
@@ -0,0 +1,241 @@
+/home/cfruan/.conda/envs/mlc-source-311/bin/python -m mlc_chat gen_config /models/gemma-7b-it --quantization q0f16 --conv-template gemma_instruction --output /tmp/tmp2aw6bqmy --context-window-size 8192 --prefill-chunk-size 1024
+[2024-02-21 23:28:16] INFO auto_config.py:115: Found model configuration: /models/gemma-7b-it/config.json
+[2024-02-21 23:28:16] INFO auto_config.py:153: Found model type: gemma. Use `--model-type` to override.
+[2024-02-21 23:28:16] INFO gemma_model.py:55: context_window_size not found in config.json. Falling back to max_position_embeddings (8192)
+[2024-02-21 23:28:16] INFO gemma_model.py:70: prefill_chunk_size defaults to context_window_size (8192)
+[2024-02-21 23:28:16] INFO config.py:106: Overriding context_window_size from 8192 to 8192
+[2024-02-21 23:28:16] INFO config.py:106: Overriding prefill_chunk_size from 8192 to 1024
+[2024-02-21 23:28:16] INFO config.py:106: Overriding max_batch_size from 1 to 80
+[2024-02-21 23:28:16] INFO gen_config.py:121: [generation_config.json] Setting bos_token_id: 2
+[2024-02-21 23:28:16] INFO gen_config.py:121: [generation_config.json] Setting eos_token_id: 1
+[2024-02-21 23:28:16] INFO gen_config.py:121: [generation_config.json] Setting pad_token_id: 0
+[2024-02-21 23:28:16] INFO gen_config.py:133: Found tokenizer config: /models/gemma-7b-it/tokenizer.model. Copying to /tmp/tmp2aw6bqmy/tokenizer.model
+[2024-02-21 23:28:16] INFO gen_config.py:133: Found tokenizer config: /models/gemma-7b-it/tokenizer.json. Copying to /tmp/tmp2aw6bqmy/tokenizer.json
+[2024-02-21 23:28:16] INFO gen_config.py:135: Not found tokenizer config: /models/gemma-7b-it/vocab.json
+[2024-02-21 23:28:16] INFO gen_config.py:135: Not found tokenizer config: /models/gemma-7b-it/merges.txt
+[2024-02-21 23:28:16] INFO gen_config.py:135: Not found tokenizer config: /models/gemma-7b-it/added_tokens.json
+[2024-02-21 23:28:16] INFO gen_config.py:133: Found tokenizer config: /models/gemma-7b-it/tokenizer_config.json. Copying to /tmp/tmp2aw6bqmy/tokenizer_config.json
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting temperature: 0.7
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting presence_penalty: 0.0
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting frequency_penalty: 0.0
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting repetition_penalty: 1.0
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting top_p: 0.95
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting mean_gen_len: 128
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting max_gen_len: 512
+[2024-02-21 23:28:16] INFO gen_config.py:74: [System default] Setting shift_fill_factor: 0.3
+[2024-02-21 23:28:16] INFO gen_config.py:186: Dumping configuration file to: /tmp/tmp2aw6bqmy/mlc-chat-config.json
+/home/cfruan/.conda/envs/mlc-source-311/bin/python -m mlc_chat convert_weight /models/gemma-7b-it --quantization q0f16 --source-format auto --output /tmp/tmp2aw6bqmy
+[2024-02-21 23:28:18] INFO auto_config.py:115: Found model configuration: /models/gemma-7b-it/config.json
+[2024-02-21 23:28:19] INFO auto_device.py:76: Found device: cuda:0
+[2024-02-21 23:28:19] INFO auto_device.py:76: Found device: cuda:1
+[2024-02-21 23:28:20] INFO auto_device.py:85: Not found device: rocm:0
+[2024-02-21 23:28:21] INFO auto_device.py:85: Not found device: metal:0
+[2024-02-21 23:28:31] INFO auto_device.py:76: Found device: vulkan:0
+[2024-02-21 23:28:31] INFO auto_device.py:76: Found device: vulkan:1
+[2024-02-21 23:28:31] INFO auto_device.py:76: Found device: vulkan:2
+[2024-02-21 23:28:33] INFO auto_device.py:85: Not found device: opencl:0
+[2024-02-21 23:28:33] INFO auto_device.py:33: Using device: cuda:0
+[2024-02-21 23:28:33] INFO auto_weight.py:70: Finding weights in: /models/gemma-7b-it
+[2024-02-21 23:28:33] INFO auto_weight.py:136: Not found Huggingface PyTorch
+[2024-02-21 23:28:33] INFO auto_weight.py:143: Found source weight format: huggingface-safetensor. Source configuration: /models/gemma-7b-it/model.safetensors.index.json
+[2024-02-21 23:28:33] INFO auto_weight.py:106: Using source weight configuration: /models/gemma-7b-it/model.safetensors.index.json. Use `--source` to override.
+[2024-02-21 23:28:33] INFO auto_weight.py:110: Using source weight format: huggingface-safetensor. Use `--source-format` to override.
+[2024-02-21 23:28:33] INFO auto_config.py:153: Found model type: gemma. Use `--model-type` to override.
+[2024-02-21 23:28:33] INFO gemma_model.py:55: context_window_size not found in config.json. Falling back to max_position_embeddings (8192)
+[2024-02-21 23:28:33] INFO gemma_model.py:70: prefill_chunk_size defaults to context_window_size (8192)
+Weight conversion with arguments:
+  --config          /models/gemma-7b-it/config.json
+  --quantization    NoQuantize(name='q0f16', kind='no-quant', model_dtype='float16')
+  --model-type      gemma
+  --device          cuda:0
+  --source          /models/gemma-7b-it/model.safetensors.index.json
+  --source-format   huggingface-safetensor
+  --output          /tmp/tmp2aw6bqmy
+
  0%|                                                                                                                                                                                                                                                                  | 0/170 [00:00<?, ?it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:34] INFO huggingface_loader.py:182: Loading HF parameters from: /models/gemma-7b-it/model-00001-of-00004.safetensors
+
  0%|                                                                                                                                                                                                                                                                  | 0/170 [00:00<?, ?it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:49] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.embed_tokens.weight", shape: (256000, 3072), dtype: float16
+
  0%|                                                                                                                                                                                                                                                                  | 0/170 [00:14<?, ?it/s]
  1%|█▍                                                                                                                                                                                                                                                        | 1/170 [00:15<44:13, 15.70s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:50] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.0.input_layernorm.weight", shape: (3072,), dtype: float16
+
  1%|█▍                                                                                                                                                                                                                                                        | 1/170 [00:15<44:13, 15.70s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:50] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.0.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
  1%|█▍                                                                                                                                                                                                                                                        | 1/170 [00:16<44:13, 15.70s/it]
  2%|████▍                                                                                                                                                                                                                                                     | 3/170 [00:16<11:45,  4.23s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:51] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.0.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
  2%|████▍                                                                                                                                                                                                                                                     | 3/170 [00:16<11:45,  4.23s/it]
  2%|█████▉                                                                                                                                                                                                                                                    | 4/170 [00:17<08:37,  3.12s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:51] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.0.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
  2%|█████▉                                                                                                                                                                                                                                                    | 4/170 [00:17<08:37,  3.12s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:51] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.0.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
  2%|█████▉                                                                                                                                                                                                                                                    | 4/170 [00:17<08:37,  3.12s/it]
  4%|████████▊                                                                                                                                                                                                                                                 | 6/170 [00:17<04:28,  1.63s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:52] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.0.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
  4%|████████▊                                                                                                                                                                                                                                                 | 6/170 [00:17<04:28,  1.63s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:52] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.1.input_layernorm.weight", shape: (3072,), dtype: float16
+
  4%|████████▊                                                                                                                                                                                                                                                 | 6/170 [00:17<04:28,  1.63s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:52] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.1.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
  4%|████████▊                                                                                                                                                                                                                                                 | 6/170 [00:17<04:28,  1.63s/it]
  5%|█████████████▏                                                                                                                                                                                                                                            | 9/170 [00:17<02:21,  1.14it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.1.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
  5%|█████████████▏                                                                                                                                                                                                                                            | 9/170 [00:18<02:21,  1.14it/s]
  6%|██████████████▋                                                                                                                                                                                                                                          | 10/170 [00:18<02:25,  1.10it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.1.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
  6%|██████████████▋                                                                                                                                                                                                                                          | 10/170 [00:18<02:25,  1.10it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.1.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
  6%|██████████████▋                                                                                                                                                                                                                                          | 10/170 [00:19<02:25,  1.10it/s]
  7%|█████████████████▌                                                                                                                                                                                                                                       | 12/170 [00:19<01:39,  1.59it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.1.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
  7%|█████████████████▌                                                                                                                                                                                                                                       | 12/170 [00:19<01:39,  1.59it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.2.input_layernorm.weight", shape: (3072,), dtype: float16
+
  7%|█████████████████▌                                                                                                                                                                                                                                       | 12/170 [00:19<01:39,  1.59it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:54] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.2.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
  7%|█████████████████▌                                                                                                                                                                                                                                       | 12/170 [00:19<01:39,  1.59it/s]
  9%|█████████████████████▉                                                                                                                                                                                                                                   | 15/170 [00:19<01:06,  2.34it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:55] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.2.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
  9%|█████████████████████▉                                                                                                                                                                                                                                   | 15/170 [00:20<01:06,  2.34it/s]
  9%|███████████████████████▍                                                                                                                                                                                                                                 | 16/170 [00:20<01:21,  1.89it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:55] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.2.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
  9%|███████████████████████▍                                                                                                                                                                                                                                 | 16/170 [00:20<01:21,  1.89it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:55] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.2.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
  9%|███████████████████████▍                                                                                                                                                                                                                                 | 16/170 [00:20<01:21,  1.89it/s]
 11%|██████████████████████████▎                                                                                                                                                                                                                              | 18/170 [00:20<01:00,  2.51it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:55] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.2.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 11%|██████████████████████████▎                                                                                                                                                                                                                              | 18/170 [00:21<01:00,  2.51it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:55] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.3.input_layernorm.weight", shape: (3072,), dtype: float16
+
 11%|██████████████████████████▎                                                                                                                                                                                                                              | 18/170 [00:21<01:00,  2.51it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:55] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.3.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 11%|██████████████████████████▎                                                                                                                                                                                                                              | 18/170 [00:21<01:00,  2.51it/s]
 12%|██████████████████████████████▊                                                                                                                                                                                                                          | 21/170 [00:21<00:44,  3.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:56] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.3.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 12%|██████████████████████████████▊                                                                                                                                                                                                                          | 21/170 [00:22<00:44,  3.33it/s]
 13%|████████████████████████████████▏                                                                                                                                                                                                                        | 22/170 [00:22<01:02,  2.37it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.3.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 13%|████████████████████████████████▏                                                                                                                                                                                                                        | 22/170 [00:22<01:02,  2.37it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.3.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 13%|████████████████████████████████▏                                                                                                                                                                                                                        | 22/170 [00:22<01:02,  2.37it/s]
 14%|███████████████████████████████████▏                                                                                                                                                                                                                     | 24/170 [00:22<00:47,  3.04it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.3.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 14%|███████████████████████████████████▏                                                                                                                                                                                                                     | 24/170 [00:22<00:47,  3.04it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.4.input_layernorm.weight", shape: (3072,), dtype: float16
+
 14%|███████████████████████████████████▏                                                                                                                                                                                                                     | 24/170 [00:22<00:47,  3.04it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.4.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 14%|███████████████████████████████████▏                                                                                                                                                                                                                     | 24/170 [00:23<00:47,  3.04it/s]
 16%|███████████████████████████████████████▌                                                                                                                                                                                                                 | 27/170 [00:23<00:37,  3.82it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:58] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.4.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 16%|███████████████████████████████████████▌                                                                                                                                                                                                                 | 27/170 [00:24<00:37,  3.82it/s]
 16%|█████████████████████████████████████████                                                                                                                                                                                                                | 28/170 [00:24<00:55,  2.58it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:58] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.4.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 16%|█████████████████████████████████████████                                                                                                                                                                                                                | 28/170 [00:24<00:55,  2.58it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:59] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.4.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 16%|█████████████████████████████████████████                                                                                                                                                                                                                | 28/170 [00:24<00:55,  2.58it/s]
 18%|███████████████████████████████████████████▉                                                                                                                                                                                                             | 30/170 [00:24<00:43,  3.26it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:59] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.4.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 18%|███████████████████████████████████████████▉                                                                                                                                                                                                             | 30/170 [00:24<00:43,  3.26it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:59] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.5.input_layernorm.weight", shape: (3072,), dtype: float16
+
 18%|███████████████████████████████████████████▉                                                                                                                                                                                                             | 30/170 [00:24<00:43,  3.26it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:28:59] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.5.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 18%|███████████████████████████████████████████▉                                                                                                                                                                                                             | 30/170 [00:24<00:43,  3.26it/s]
 19%|████████████████████████████████████████████████▎                                                                                                                                                                                                        | 33/170 [00:25<00:34,  4.01it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.5.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 19%|████████████████████████████████████████████████▎                                                                                                                                                                                                        | 33/170 [00:25<00:34,  4.01it/s]
 20%|█████████████████████████████████████████████████▊                                                                                                                                                                                                       | 34/170 [00:26<00:51,  2.65it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.5.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 20%|█████████████████████████████████████████████████▊                                                                                                                                                                                                       | 34/170 [00:26<00:51,  2.65it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.5.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 20%|█████████████████████████████████████████████████▊                                                                                                                                                                                                       | 34/170 [00:26<00:51,  2.65it/s]
 21%|████████████████████████████████████████████████████▋                                                                                                                                                                                                    | 36/170 [00:26<00:40,  3.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.5.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 21%|████████████████████████████████████████████████████▋                                                                                                                                                                                                    | 36/170 [00:26<00:40,  3.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:01] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.6.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 21%|████████████████████████████████████████████████████▋                                                                                                                                                                                                    | 36/170 [00:26<00:40,  3.33it/s]
 22%|███████████████████████████████████████████████████████▋                                                                                                                                                                                                 | 38/170 [00:26<00:34,  3.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:01] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.6.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 22%|███████████████████████████████████████████████████████▋                                                                                                                                                                                                 | 38/170 [00:26<00:34,  3.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:01] INFO huggingface_loader.py:194: Unloading HF weight file: /models/gemma-7b-it/model-00001-of-00004.safetensors
+
 22%|███████████████████████████████████████████████████████▋                                                                                                                                                                                                 | 38/170 [00:26<00:34,  3.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:02] INFO huggingface_loader.py:182: Loading HF parameters from: /models/gemma-7b-it/model-00002-of-00004.safetensors
+
 22%|███████████████████████████████████████████████████████▋                                                                                                                                                                                                 | 38/170 [00:27<00:34,  3.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:11] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.10.input_layernorm.weight", shape: (3072,), dtype: float16
+
 22%|███████████████████████████████████████████████████████▋                                                                                                                                                                                                 | 38/170 [00:37<00:34,  3.88it/s]
 24%|██████████████████████████████████████████████████████████▌                                                                                                                                                                                              | 40/170 [00:37<03:56,  1.82s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:12] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.10.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 24%|██████████████████████████████████████████████████████████▌                                                                                                                                                                                              | 40/170 [00:37<03:56,  1.82s/it]
 24%|████████████████████████████████████████████████████████████                                                                                                                                                                                             | 41/170 [00:37<03:22,  1.57s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:13] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.10.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 24%|████████████████████████████████████████████████████████████                                                                                                                                                                                             | 41/170 [00:38<03:22,  1.57s/it]
 25%|█████████████████████████████████████████████████████████████▌                                                                                                                                                                                           | 42/170 [00:38<03:07,  1.46s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:13] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.10.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 25%|█████████████████████████████████████████████████████████████▌                                                                                                                                                                                           | 42/170 [00:38<03:07,  1.46s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:13] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.10.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 25%|█████████████████████████████████████████████████████████████▌                                                                                                                                                                                           | 42/170 [00:39<03:07,  1.46s/it]
 26%|████████████████████████████████████████████████████████████████▍                                                                                                                                                                                        | 44/170 [00:39<02:02,  1.03it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:13] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.10.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 26%|████████████████████████████████████████████████████████████████▍                                                                                                                                                                                        | 44/170 [00:39<02:02,  1.03it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:13] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.11.input_layernorm.weight", shape: (3072,), dtype: float16
+
 26%|████████████████████████████████████████████████████████████████▍                                                                                                                                                                                        | 44/170 [00:39<02:02,  1.03it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:14] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.11.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 26%|████████████████████████████████████████████████████████████████▍                                                                                                                                                                                        | 44/170 [00:39<02:02,  1.03it/s]
 28%|████████████████████████████████████████████████████████████████████▊                                                                                                                                                                                    | 47/170 [00:39<01:15,  1.63it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:15] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.11.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 28%|████████████████████████████████████████████████████████████████████▊                                                                                                                                                                                    | 47/170 [00:40<01:15,  1.63it/s]
 28%|██████████████████████████████████████████████████████████████████████▎                                                                                                                                                                                  | 48/170 [00:40<01:23,  1.46it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:15] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.11.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 28%|██████████████████████████████████████████████████████████████████████▎                                                                                                                                                                                  | 48/170 [00:40<01:23,  1.46it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:15] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.11.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 28%|██████████████████████████████████████████████████████████████████████▎                                                                                                                                                                                  | 48/170 [00:40<01:23,  1.46it/s]
 29%|█████████████████████████████████████████████████████████████████████████▏                                                                                                                                                                               | 50/170 [00:40<01:00,  1.99it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:15] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.11.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 29%|█████████████████████████████████████████████████████████████████████████▏                                                                                                                                                                               | 50/170 [00:40<01:00,  1.99it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:15] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.12.input_layernorm.weight", shape: (3072,), dtype: float16
+
 29%|█████████████████████████████████████████████████████████████████████████▏                                                                                                                                                                               | 50/170 [00:40<01:00,  1.99it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:15] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.12.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 29%|█████████████████████████████████████████████████████████████████████████▏                                                                                                                                                                               | 50/170 [00:41<01:00,  1.99it/s]
 31%|█████████████████████████████████████████████████████████████████████████████▋                                                                                                                                                                           | 53/170 [00:41<00:42,  2.77it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:16] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.12.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 31%|█████████████████████████████████████████████████████████████████████████████▋                                                                                                                                                                           | 53/170 [00:42<00:42,  2.77it/s]
 32%|███████████████████████████████████████████████████████████████████████████████                                                                                                                                                                          | 54/170 [00:42<00:54,  2.12it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:16] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.12.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 32%|███████████████████████████████████████████████████████████████████████████████                                                                                                                                                                          | 54/170 [00:42<00:54,  2.12it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:17] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.12.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 32%|███████████████████████████████████████████████████████████████████████████████                                                                                                                                                                          | 54/170 [00:42<00:54,  2.12it/s]
 33%|██████████████████████████████████████████████████████████████████████████████████                                                                                                                                                                       | 56/170 [00:42<00:41,  2.77it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:17] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.12.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 33%|██████████████████████████████████████████████████████████████████████████████████                                                                                                                                                                       | 56/170 [00:42<00:41,  2.77it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:17] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.13.input_layernorm.weight", shape: (3072,), dtype: float16
+
 33%|██████████████████████████████████████████████████████████████████████████████████                                                                                                                                                                       | 56/170 [00:42<00:41,  2.77it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:17] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.13.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 33%|██████████████████████████████████████████████████████████████████████████████████                                                                                                                                                                       | 56/170 [00:43<00:41,  2.77it/s]
 35%|██████████████████████████████████████████████████████████████████████████████████████▍                                                                                                                                                                  | 59/170 [00:43<00:31,  3.57it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:18] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.13.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 35%|██████████████████████████████████████████████████████████████████████████████████████▍                                                                                                                                                                  | 59/170 [00:44<00:31,  3.57it/s]
 35%|███████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                                                                 | 60/170 [00:44<00:44,  2.49it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:18] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.13.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 35%|███████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                                                                 | 60/170 [00:44<00:44,  2.49it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:18] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.13.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 35%|███████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                                                                 | 60/170 [00:44<00:44,  2.49it/s]
 36%|██████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                                                              | 62/170 [00:44<00:34,  3.16it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.13.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 36%|██████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                                                              | 62/170 [00:44<00:34,  3.16it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.14.input_layernorm.weight", shape: (3072,), dtype: float16
+
 36%|██████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                                                              | 62/170 [00:44<00:34,  3.16it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.14.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 36%|██████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                                                              | 62/170 [00:44<00:34,  3.16it/s]
 38%|███████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                                         | 65/170 [00:44<00:26,  3.93it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:20] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.14.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 38%|███████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                                         | 65/170 [00:45<00:26,  3.93it/s]
 39%|████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                                                                                        | 66/170 [00:45<00:39,  2.63it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:20] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.14.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 39%|████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                                                                                        | 66/170 [00:45<00:39,  2.63it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:20] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.14.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 39%|████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                                                                                        | 66/170 [00:46<00:39,  2.63it/s]
 40%|███████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                     | 68/170 [00:46<00:30,  3.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:20] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.14.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 40%|███████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                     | 68/170 [00:46<00:30,  3.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:21] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.15.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 40%|███████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                     | 68/170 [00:46<00:30,  3.33it/s]
 41%|██████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                  | 70/170 [00:46<00:25,  3.89it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:21] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.15.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 41%|██████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                  | 70/170 [00:46<00:25,  3.89it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:21] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.6.input_layernorm.weight", shape: (3072,), dtype: float16
+
 41%|██████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                  | 70/170 [00:46<00:25,  3.89it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:21] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.6.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 41%|██████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                                                  | 70/170 [00:46<00:25,  3.89it/s]
 43%|██████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                                              | 73/170 [00:47<00:21,  4.57it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:22] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.6.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 43%|██████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                                              | 73/170 [00:47<00:21,  4.57it/s]
 44%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                                                                            | 74/170 [00:48<00:33,  2.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:22] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.6.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 44%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                                                                            | 74/170 [00:48<00:33,  2.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:22] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.7.input_layernorm.weight", shape: (3072,), dtype: float16
+
 44%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                                                                            | 74/170 [00:48<00:33,  2.88it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:22] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.7.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 44%|████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                                                                            | 74/170 [00:48<00:33,  2.88it/s]
 45%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                                        | 77/170 [00:48<00:24,  3.80it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:23] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.7.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 45%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                                        | 77/170 [00:49<00:24,  3.80it/s]
 46%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                      | 78/170 [00:49<00:35,  2.59it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:24] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.7.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 46%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                      | 78/170 [00:49<00:35,  2.59it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:24] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.7.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 46%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                      | 78/170 [00:49<00:35,  2.59it/s]
 47%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                   | 80/170 [00:49<00:27,  3.28it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:24] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.7.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 47%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                   | 80/170 [00:49<00:27,  3.28it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:24] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.8.input_layernorm.weight", shape: (3072,), dtype: float16
+
 47%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                   | 80/170 [00:49<00:27,  3.28it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:24] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.8.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 47%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                                                   | 80/170 [00:50<00:27,  3.28it/s]
 49%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                               | 83/170 [00:50<00:21,  4.05it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:25] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.8.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 49%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                                               | 83/170 [00:51<00:21,  4.05it/s]
 49%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                                                                              | 84/170 [00:51<00:32,  2.68it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:25] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.8.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 49%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                                                                              | 84/170 [00:51<00:32,  2.68it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.8.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 49%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                                                                              | 84/170 [00:51<00:32,  2.68it/s]
 51%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                           | 86/170 [00:51<00:24,  3.37it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.8.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 51%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                           | 86/170 [00:51<00:24,  3.37it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.9.input_layernorm.weight", shape: (3072,), dtype: float16
+
 51%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                           | 86/170 [00:51<00:24,  3.37it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.9.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 51%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                                           | 86/170 [00:51<00:24,  3.37it/s]
 52%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                                                                                                      | 89/170 [00:52<00:19,  4.13it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:27] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.9.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 52%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                                                                                                      | 89/170 [00:52<00:19,  4.13it/s]
 53%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                     | 90/170 [00:53<00:29,  2.71it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:27] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.9.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 53%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                     | 90/170 [00:53<00:29,  2.71it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:27] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.9.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 53%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                     | 90/170 [00:53<00:29,  2.71it/s]
 54%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                  | 92/170 [00:53<00:22,  3.41it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:27] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.9.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 54%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                  | 92/170 [00:53<00:22,  3.41it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:27] INFO huggingface_loader.py:194: Unloading HF weight file: /models/gemma-7b-it/model-00002-of-00004.safetensors
+
 54%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                  | 92/170 [00:53<00:22,  3.41it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:28] INFO huggingface_loader.py:182: Loading HF parameters from: /models/gemma-7b-it/model-00003-of-00004.safetensors
+
 54%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                  | 92/170 [00:54<00:22,  3.41it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:38] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.15.input_layernorm.weight", shape: (3072,), dtype: float16
+
 54%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                                  | 92/170 [01:03<00:22,  3.41it/s]
 55%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                                               | 94/170 [01:03<02:20,  1.84s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:39] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.15.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 55%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                                               | 94/170 [01:04<02:20,  1.84s/it]
 56%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                             | 95/170 [01:04<02:04,  1.66s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:41] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.15.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 56%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                                             | 95/170 [01:06<02:04,  1.66s/it]
 56%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                            | 96/170 [01:06<02:10,  1.76s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:41] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.15.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 56%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                            | 96/170 [01:06<02:10,  1.76s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:41] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.16.input_layernorm.weight", shape: (3072,), dtype: float16
+
 56%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                            | 96/170 [01:06<02:10,  1.76s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:41] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.16.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 56%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                                            | 96/170 [01:07<02:10,  1.76s/it]
 58%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                                                        | 99/170 [01:07<01:12,  1.02s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:43] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.16.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 58%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                                                        | 99/170 [01:09<01:12,  1.02s/it]
 59%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                      | 100/170 [01:09<01:22,  1.18s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:43] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.16.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 59%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                      | 100/170 [01:09<01:22,  1.18s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:44] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.16.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 59%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                                      | 100/170 [01:09<01:22,  1.18s/it]
 60%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                   | 102/170 [01:09<00:58,  1.17it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:44] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.16.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 60%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                   | 102/170 [01:09<00:58,  1.17it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:44] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.17.input_layernorm.weight", shape: (3072,), dtype: float16
+
 60%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                   | 102/170 [01:09<00:58,  1.17it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:44] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.17.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 60%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                                                   | 102/170 [01:10<00:58,  1.17it/s]
 62%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                              | 105/170 [01:10<00:37,  1.73it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:46] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.17.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 62%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                                              | 105/170 [01:12<00:37,  1.73it/s]
 62%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                             | 106/170 [01:12<00:52,  1.23it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:46] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.17.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 62%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                             | 106/170 [01:12<00:52,  1.23it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:47] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.17.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 62%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                                             | 106/170 [01:12<00:52,  1.23it/s]
 64%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                          | 108/170 [01:12<00:38,  1.61it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:47] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.17.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 64%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                          | 108/170 [01:12<00:38,  1.61it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:47] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.18.input_layernorm.weight", shape: (3072,), dtype: float16
+
 64%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                          | 108/170 [01:12<00:38,  1.61it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:47] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.18.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 64%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                                                          | 108/170 [01:13<00:38,  1.61it/s]
 65%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                      | 111/170 [01:13<00:26,  2.21it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:49] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.18.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 65%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                                      | 111/170 [01:15<00:26,  2.21it/s]
 66%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                    | 112/170 [01:15<00:40,  1.42it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:49] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.18.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 66%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                    | 112/170 [01:15<00:40,  1.42it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:50] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.18.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 66%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                                    | 112/170 [01:15<00:40,  1.42it/s]
 67%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                                                                 | 114/170 [01:15<00:31,  1.79it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:50] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.18.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 67%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                                                                 | 114/170 [01:15<00:31,  1.79it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:50] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.19.input_layernorm.weight", shape: (3072,), dtype: float16
+
 67%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                                                                 | 114/170 [01:15<00:31,  1.79it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:51] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.19.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 67%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                                                                 | 114/170 [01:16<00:31,  1.79it/s]
 69%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                             | 117/170 [01:16<00:22,  2.40it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.19.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 69%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                                             | 117/170 [01:18<00:22,  2.40it/s]
 69%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                           | 118/170 [01:18<00:37,  1.38it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.19.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 69%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                           | 118/170 [01:18<00:37,  1.38it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.19.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 69%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                                           | 118/170 [01:19<00:37,  1.38it/s]
 71%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                         | 120/170 [01:19<00:28,  1.74it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.19.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 71%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                         | 120/170 [01:19<00:28,  1.74it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:53] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.20.input_layernorm.weight", shape: (3072,), dtype: float16
+
 71%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                         | 120/170 [01:19<00:28,  1.74it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:54] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.20.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 71%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                         | 120/170 [01:19<00:28,  1.74it/s]
 72%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                    | 123/170 [01:19<00:19,  2.38it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:56] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.20.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 72%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                                    | 123/170 [01:22<00:19,  2.38it/s]
 73%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                   | 124/170 [01:22<00:33,  1.36it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:56] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.20.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 73%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                   | 124/170 [01:22<00:33,  1.36it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.20.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 73%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                                   | 124/170 [01:22<00:33,  1.36it/s]
 74%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                | 126/170 [01:22<00:25,  1.74it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.20.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 74%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                | 126/170 [01:22<00:25,  1.74it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.21.input_layernorm.weight", shape: (3072,), dtype: float16
+
 74%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                | 126/170 [01:22<00:25,  1.74it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:57] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.21.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 74%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                                                                | 126/170 [01:23<00:25,  1.74it/s]
 76%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                           | 129/170 [01:23<00:17,  2.38it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:29:59] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.21.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 76%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                                           | 129/170 [01:25<00:17,  2.38it/s]
 76%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                          | 130/170 [01:25<00:28,  1.40it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.21.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 76%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                          | 130/170 [01:25<00:28,  1.40it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.21.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 76%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                                          | 130/170 [01:26<00:28,  1.40it/s]
 78%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                       | 132/170 [01:26<00:21,  1.79it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.21.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 78%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                       | 132/170 [01:26<00:21,  1.79it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:00] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.22.input_layernorm.weight", shape: (3072,), dtype: float16
+
 78%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                       | 132/170 [01:26<00:21,  1.79it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:01] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.22.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 78%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                                                       | 132/170 [01:26<00:21,  1.79it/s]
 79%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                   | 135/170 [01:26<00:14,  2.45it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:03] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.22.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 79%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                                   | 135/170 [01:28<00:14,  2.45it/s]
 80%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                 | 136/170 [01:28<00:23,  1.47it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:03] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.22.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 80%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                 | 136/170 [01:28<00:23,  1.47it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:03] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.22.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 80%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍                                                 | 136/170 [01:29<00:23,  1.47it/s]
 81%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                              | 138/170 [01:29<00:17,  1.87it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:03] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.22.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 81%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                              | 138/170 [01:29<00:17,  1.87it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:03] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.23.input_layernorm.weight", shape: (3072,), dtype: float16
+
 81%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                              | 138/170 [01:29<00:17,  1.87it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:04] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.23.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 81%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                                              | 138/170 [01:29<00:17,  1.87it/s]
 83%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                          | 141/170 [01:29<00:11,  2.54it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:06] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.23.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 83%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                                          | 141/170 [01:31<00:11,  2.54it/s]
 84%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                        | 142/170 [01:31<00:18,  1.50it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:06] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.23.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 84%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                        | 142/170 [01:31<00:18,  1.50it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:06] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.23.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 84%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                                        | 142/170 [01:32<00:18,  1.50it/s]
 85%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                      | 144/170 [01:32<00:13,  1.92it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:06] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.23.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 85%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                      | 144/170 [01:32<00:13,  1.92it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:07] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.24.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 85%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                      | 144/170 [01:32<00:13,  1.92it/s]
 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                   | 146/170 [01:32<00:10,  2.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:07] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.24.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                   | 146/170 [01:32<00:10,  2.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:07] INFO huggingface_loader.py:194: Unloading HF weight file: /models/gemma-7b-it/model-00003-of-00004.safetensors
+
 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                   | 146/170 [01:32<00:10,  2.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:07] INFO huggingface_loader.py:182: Loading HF parameters from: /models/gemma-7b-it/model-00004-of-00004.safetensors
+
 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                   | 146/170 [01:33<00:10,  2.33it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:12] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.24.input_layernorm.weight", shape: (3072,), dtype: float16
+
 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                   | 146/170 [01:37<00:10,  2.33it/s]
 87%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                | 148/170 [01:37<00:23,  1.05s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:12] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.24.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 87%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                                | 148/170 [01:38<00:23,  1.05s/it]
 88%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                              | 149/170 [01:38<00:20,  1.03it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:16] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.24.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 88%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎                              | 149/170 [01:41<00:20,  1.03it/s]
 88%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                             | 150/170 [01:41<00:30,  1.53s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:16] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.24.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 88%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                             | 150/170 [01:41<00:30,  1.53s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:16] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.25.input_layernorm.weight", shape: (3072,), dtype: float16
+
 88%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                             | 150/170 [01:41<00:30,  1.53s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:16] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.25.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 88%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▊                             | 150/170 [01:42<00:30,  1.53s/it]
 90%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                        | 153/170 [01:42<00:14,  1.14it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.25.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 90%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏                        | 153/170 [01:44<00:14,  1.14it/s]
 91%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                       | 154/170 [01:44<00:19,  1.20s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.25.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 91%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                       | 154/170 [01:44<00:19,  1.20s/it]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.25.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 91%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋                       | 154/170 [01:45<00:19,  1.20s/it]
 92%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                    | 156/170 [01:45<00:11,  1.20it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.25.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 92%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                    | 156/170 [01:45<00:11,  1.20it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:19] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.26.input_layernorm.weight", shape: (3072,), dtype: float16
+
 92%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                    | 156/170 [01:45<00:11,  1.20it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:20] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.26.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 92%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▌                    | 156/170 [01:45<00:11,  1.20it/s]
 94%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                | 159/170 [01:45<00:06,  1.81it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:22] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.26.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 94%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▉                | 159/170 [01:48<00:06,  1.81it/s]
 94%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍              | 160/170 [01:48<00:09,  1.10it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:22] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.26.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 94%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍              | 160/170 [01:48<00:09,  1.10it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:23] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.26.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 94%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▍              | 160/170 [01:48<00:09,  1.10it/s]
 95%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎           | 162/170 [01:48<00:05,  1.52it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:23] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.26.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 95%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎           | 162/170 [01:48<00:05,  1.52it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:23] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.27.input_layernorm.weight", shape: (3072,), dtype: float16
+
 95%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎           | 162/170 [01:48<00:05,  1.52it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:23] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.27.mlp.down_proj.weight", shape: (3072, 24576), dtype: float16
+
 95%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎           | 162/170 [01:49<00:05,  1.52it/s]
 97%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋       | 165/170 [01:49<00:02,  2.20it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:25] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.27.mlp.gate_up_proj.weight", shape: (49152, 3072), dtype: float16
+
 97%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▋       | 165/170 [01:51<00:02,  2.20it/s]
 98%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏     | 166/170 [01:51<00:02,  1.35it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:25] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.27.post_attention_layernorm.weight", shape: (3072,), dtype: float16
+
 98%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏     | 166/170 [01:51<00:02,  1.35it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.27.self_attn.qkv_proj.weight", shape: (12288, 3072), dtype: float16
+
 98%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏     | 166/170 [01:51<00:02,  1.35it/s]
 99%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████   | 168/170 [01:51<00:01,  1.82it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.layers.27.self_attn.o_proj.weight", shape: (3072, 4096), dtype: float16
+
 99%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████   | 168/170 [01:51<00:01,  1.82it/s]
                                                                                                                                                                                                                                                                                               
[2024-02-21 23:30:26] INFO huggingface_loader.py:172: [Not quantized] Parameter: "model.norm.weight", shape: (3072,), dtype: float16
+
 99%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████   | 168/170 [01:51<00:01,  1.82it/s]
100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 170/170 [01:51<00:00,  1.52it/s]
+[2024-02-21 23:30:26] INFO huggingface_loader.py:194: Unloading HF weight file: /models/gemma-7b-it/model-00004-of-00004.safetensors
+[2024-02-21 23:30:26] INFO stats.py:76: Time usage: HF loading: 37.572 sec; Pre-quantization mapping: 61.186 sec; Quantization: 0.000 sec
+[2024-02-21 23:30:26] INFO stats.py:90: RAM usage: Peak RAM: 9.305 GB. Total bytes loaded from disk: 31.805 GB
+[2024-02-21 23:30:26] INFO convert_weight.py:132: Parameter size after quantization: 15.903 GB
+[2024-02-21 23:30:26] INFO convert_weight.py:137: Total parameters: 8,537,680,896
+[2024-02-21 23:30:26] INFO convert_weight.py:138: Bits per parameter: 16.000
+Start storing to cache /tmp/tmp2aw6bqmy
+

[0001/0170] saving model.embed_tokens.weight
                                            
[0002/0170] saving model.layers.0.input_layernorm.weight
                                                        
[0003/0170] saving model.layers.0.mlp.down_proj.weight
                                                        
[0004/0170] saving model.layers.0.mlp.gate_up_proj.weight
                                                         
[0005/0170] saving model.layers.0.post_attention_layernorm.weight
                                                                 
[0006/0170] saving model.layers.0.self_attn.qkv_proj.weight
                                                                 
[0007/0170] saving model.layers.0.self_attn.o_proj.weight
                                                                 
[0008/0170] saving model.layers.1.input_layernorm.weight
                                                                 
[0009/0170] saving model.layers.1.mlp.down_proj.weight
                                                                 
[0010/0170] saving model.layers.1.mlp.gate_up_proj.weight
                                                                 
[0011/0170] saving model.layers.1.post_attention_layernorm.weight
                                                                 
[0012/0170] saving model.layers.1.self_attn.qkv_proj.weight
                                                                 
[0013/0170] saving model.layers.1.self_attn.o_proj.weight
                                                                 
[0014/0170] saving model.layers.2.input_layernorm.weight
                                                                 
[0015/0170] saving model.layers.2.mlp.down_proj.weight
                                                                 
[0016/0170] saving model.layers.2.mlp.gate_up_proj.weight
                                                                 
[0017/0170] saving model.layers.2.post_attention_layernorm.weight
                                                                 
[0018/0170] saving model.layers.2.self_attn.qkv_proj.weight
                                                                 
[0019/0170] saving model.layers.2.self_attn.o_proj.weight
                                                                 
[0020/0170] saving model.layers.3.input_layernorm.weight
                                                                 
[0021/0170] saving model.layers.3.mlp.down_proj.weight
                                                                 
[0022/0170] saving model.layers.3.mlp.gate_up_proj.weight
                                                                 
[0023/0170] saving model.layers.3.post_attention_layernorm.weight
                                                                 
[0024/0170] saving model.layers.3.self_attn.qkv_proj.weight
                                                                 
[0025/0170] saving model.layers.3.self_attn.o_proj.weight
                                                                 
[0026/0170] saving model.layers.4.input_layernorm.weight
                                                                 
[0027/0170] saving model.layers.4.mlp.down_proj.weight
                                                                 
[0028/0170] saving model.layers.4.mlp.gate_up_proj.weight
                                                                 
[0029/0170] saving model.layers.4.post_attention_layernorm.weight
                                                                 
[0030/0170] saving model.layers.4.self_attn.qkv_proj.weight
                                                                 
[0031/0170] saving model.layers.4.self_attn.o_proj.weight
                                                                 
[0032/0170] saving model.layers.5.input_layernorm.weight
                                                                 
[0033/0170] saving model.layers.5.mlp.down_proj.weight
                                                                 
[0034/0170] saving model.layers.5.mlp.gate_up_proj.weight
                                                                 
[0035/0170] saving model.layers.5.post_attention_layernorm.weight
                                                                 
[0036/0170] saving model.layers.5.self_attn.qkv_proj.weight
                                                                 
[0037/0170] saving model.layers.5.self_attn.o_proj.weight
                                                                 
[0038/0170] saving model.layers.6.self_attn.qkv_proj.weight
                                                                 
[0039/0170] saving model.layers.6.self_attn.o_proj.weight
                                                                 
[0040/0170] saving model.layers.10.input_layernorm.weight
                                                                 
[0041/0170] saving model.layers.10.mlp.down_proj.weight
                                                                 
[0042/0170] saving model.layers.10.mlp.gate_up_proj.weight
                                                                 
[0043/0170] saving model.layers.10.post_attention_layernorm.weight
                                                                  
[0044/0170] saving model.layers.10.self_attn.qkv_proj.weight
                                                                  
[0045/0170] saving model.layers.10.self_attn.o_proj.weight
                                                                  
[0046/0170] saving model.layers.11.input_layernorm.weight
                                                                  
[0047/0170] saving model.layers.11.mlp.down_proj.weight
                                                                  
[0048/0170] saving model.layers.11.mlp.gate_up_proj.weight
                                                                  
[0049/0170] saving model.layers.11.post_attention_layernorm.weight
                                                                  
[0050/0170] saving model.layers.11.self_attn.qkv_proj.weight
                                                                  
[0051/0170] saving model.layers.11.self_attn.o_proj.weight
                                                                  
[0052/0170] saving model.layers.12.input_layernorm.weight
                                                                  
[0053/0170] saving model.layers.12.mlp.down_proj.weight
                                                                  
[0054/0170] saving model.layers.12.mlp.gate_up_proj.weight
                                                                  
[0055/0170] saving model.layers.12.post_attention_layernorm.weight
                                                                  
[0056/0170] saving model.layers.12.self_attn.qkv_proj.weight
                                                                  
[0057/0170] saving model.layers.12.self_attn.o_proj.weight
                                                                  
[0058/0170] saving model.layers.13.input_layernorm.weight
                                                                  
[0059/0170] saving model.layers.13.mlp.down_proj.weight
                                                                  
[0060/0170] saving model.layers.13.mlp.gate_up_proj.weight
                                                                  
[0061/0170] saving model.layers.13.post_attention_layernorm.weight
                                                                  
[0062/0170] saving model.layers.13.self_attn.qkv_proj.weight
                                                                  
[0063/0170] saving model.layers.13.self_attn.o_proj.weight
                                                                  
[0064/0170] saving model.layers.14.input_layernorm.weight
                                                                  
[0065/0170] saving model.layers.14.mlp.down_proj.weight
                                                                  
[0066/0170] saving model.layers.14.mlp.gate_up_proj.weight
                                                                  
[0067/0170] saving model.layers.14.post_attention_layernorm.weight
                                                                  
[0068/0170] saving model.layers.14.self_attn.qkv_proj.weight
                                                                  
[0069/0170] saving model.layers.14.self_attn.o_proj.weight
                                                                  
[0070/0170] saving model.layers.15.self_attn.qkv_proj.weight
                                                                  
[0071/0170] saving model.layers.15.self_attn.o_proj.weight
                                                                  
[0072/0170] saving model.layers.6.input_layernorm.weight
                                                                  
[0073/0170] saving model.layers.6.mlp.down_proj.weight
                                                                  
[0074/0170] saving model.layers.6.mlp.gate_up_proj.weight
                                                                  
[0075/0170] saving model.layers.6.post_attention_layernorm.weight
                                                                  
[0076/0170] saving model.layers.7.input_layernorm.weight
                                                                  
[0077/0170] saving model.layers.7.mlp.down_proj.weight
                                                                  
[0078/0170] saving model.layers.7.mlp.gate_up_proj.weight
                                                                  
[0079/0170] saving model.layers.7.post_attention_layernorm.weight
                                                                  
[0080/0170] saving model.layers.7.self_attn.qkv_proj.weight
                                                                  
[0081/0170] saving model.layers.7.self_attn.o_proj.weight
                                                                  
[0082/0170] saving model.layers.8.input_layernorm.weight
                                                                  
[0083/0170] saving model.layers.8.mlp.down_proj.weight
                                                                  
[0084/0170] saving model.layers.8.mlp.gate_up_proj.weight
                                                                  
[0085/0170] saving model.layers.8.post_attention_layernorm.weight
                                                                  
[0086/0170] saving model.layers.8.self_attn.qkv_proj.weight
                                                                  
[0087/0170] saving model.layers.8.self_attn.o_proj.weight
                                                                  
[0088/0170] saving model.layers.9.input_layernorm.weight
                                                                  
[0089/0170] saving model.layers.9.mlp.down_proj.weight
                                                                  
[0090/0170] saving model.layers.9.mlp.gate_up_proj.weight
                                                                  
[0091/0170] saving model.layers.9.post_attention_layernorm.weight
                                                                  
[0092/0170] saving model.layers.9.self_attn.qkv_proj.weight
                                                                  
[0093/0170] saving model.layers.9.self_attn.o_proj.weight
                                                                  
[0094/0170] saving model.layers.15.input_layernorm.weight
                                                                  
[0095/0170] saving model.layers.15.mlp.down_proj.weight
                                                                  
[0096/0170] saving model.layers.15.mlp.gate_up_proj.weight
                                                                  
[0097/0170] saving model.layers.15.post_attention_layernorm.weight
                                                                  
[0098/0170] saving model.layers.16.input_layernorm.weight
                                                                  
[0099/0170] saving model.layers.16.mlp.down_proj.weight
                                                                  
[0100/0170] saving model.layers.16.mlp.gate_up_proj.weight
                                                                  
[0101/0170] saving model.layers.16.post_attention_layernorm.weight
                                                                  
[0102/0170] saving model.layers.16.self_attn.qkv_proj.weight
                                                                  
[0103/0170] saving model.layers.16.self_attn.o_proj.weight
                                                                  
[0104/0170] saving model.layers.17.input_layernorm.weight
                                                                  
[0105/0170] saving model.layers.17.mlp.down_proj.weight
                                                                  
[0106/0170] saving model.layers.17.mlp.gate_up_proj.weight
                                                                  
[0107/0170] saving model.layers.17.post_attention_layernorm.weight
                                                                  
[0108/0170] saving model.layers.17.self_attn.qkv_proj.weight
                                                                  
[0109/0170] saving model.layers.17.self_attn.o_proj.weight
                                                                  
[0110/0170] saving model.layers.18.input_layernorm.weight
                                                                  
[0111/0170] saving model.layers.18.mlp.down_proj.weight
                                                                  
[0112/0170] saving model.layers.18.mlp.gate_up_proj.weight
                                                                  
[0113/0170] saving model.layers.18.post_attention_layernorm.weight
                                                                  
[0114/0170] saving model.layers.18.self_attn.qkv_proj.weight
                                                                  
[0115/0170] saving model.layers.18.self_attn.o_proj.weight
                                                                  
[0116/0170] saving model.layers.19.input_layernorm.weight
                                                                  
[0117/0170] saving model.layers.19.mlp.down_proj.weight
                                                                  
[0118/0170] saving model.layers.19.mlp.gate_up_proj.weight
                                                                  
[0119/0170] saving model.layers.19.post_attention_layernorm.weight
                                                                  
[0120/0170] saving model.layers.19.self_attn.qkv_proj.weight
                                                                  
[0121/0170] saving model.layers.19.self_attn.o_proj.weight
                                                                  
[0122/0170] saving model.layers.20.input_layernorm.weight
                                                                  
[0123/0170] saving model.layers.20.mlp.down_proj.weight
                                                                  
[0124/0170] saving model.layers.20.mlp.gate_up_proj.weight
                                                                  
[0125/0170] saving model.layers.20.post_attention_layernorm.weight
                                                                  
[0126/0170] saving model.layers.20.self_attn.qkv_proj.weight
                                                                  
[0127/0170] saving model.layers.20.self_attn.o_proj.weight
                                                                  
[0128/0170] saving model.layers.21.input_layernorm.weight
                                                                  
[0129/0170] saving model.layers.21.mlp.down_proj.weight[2024-02-21 23:31:13] INFO convert_weight.py:154: Saved to directory: /tmp/tmp2aw6bqmy
+
                                                                  
[0130/0170] saving model.layers.21.mlp.gate_up_proj.weight
                                                                  
[0131/0170] saving model.layers.21.post_attention_layernorm.weight
                                                                  
[0132/0170] saving model.layers.21.self_attn.qkv_proj.weight
                                                                  
[0133/0170] saving model.layers.21.self_attn.o_proj.weight
                                                                  
[0134/0170] saving model.layers.22.input_layernorm.weight
                                                                  
[0135/0170] saving model.layers.22.mlp.down_proj.weight
                                                                  
[0136/0170] saving model.layers.22.mlp.gate_up_proj.weight
                                                                  
[0137/0170] saving model.layers.22.post_attention_layernorm.weight
                                                                  
[0138/0170] saving model.layers.22.self_attn.qkv_proj.weight
                                                                  
[0139/0170] saving model.layers.22.self_attn.o_proj.weight
                                                                  
[0140/0170] saving model.layers.23.input_layernorm.weight
                                                                  
[0141/0170] saving model.layers.23.mlp.down_proj.weight
                                                                  
[0142/0170] saving model.layers.23.mlp.gate_up_proj.weight
                                                                  
[0143/0170] saving model.layers.23.post_attention_layernorm.weight
                                                                  
[0144/0170] saving model.layers.23.self_attn.qkv_proj.weight
                                                                  
[0145/0170] saving model.layers.23.self_attn.o_proj.weight
                                                                  
[0146/0170] saving model.layers.24.self_attn.qkv_proj.weight
                                                                  
[0147/0170] saving model.layers.24.self_attn.o_proj.weight
                                                                  
[0148/0170] saving model.layers.24.input_layernorm.weight
                                                                  
[0149/0170] saving model.layers.24.mlp.down_proj.weight
                                                                  
[0150/0170] saving model.layers.24.mlp.gate_up_proj.weight
                                                                  
[0151/0170] saving model.layers.24.post_attention_layernorm.weight
                                                                  
[0152/0170] saving model.layers.25.input_layernorm.weight
                                                                  
[0153/0170] saving model.layers.25.mlp.down_proj.weight
                                                                  
[0154/0170] saving model.layers.25.mlp.gate_up_proj.weight
                                                                  
[0155/0170] saving model.layers.25.post_attention_layernorm.weight
                                                                  
[0156/0170] saving model.layers.25.self_attn.qkv_proj.weight
                                                                  
[0157/0170] saving model.layers.25.self_attn.o_proj.weight
                                                                  
[0158/0170] saving model.layers.26.input_layernorm.weight
                                                                  
[0159/0170] saving model.layers.26.mlp.down_proj.weight
                                                                  
[0160/0170] saving model.layers.26.mlp.gate_up_proj.weight
                                                                  
[0161/0170] saving model.layers.26.post_attention_layernorm.weight
                                                                  
[0162/0170] saving model.layers.26.self_attn.qkv_proj.weight
                                                                  
[0163/0170] saving model.layers.26.self_attn.o_proj.weight
                                                                  
[0164/0170] saving model.layers.27.input_layernorm.weight
                                                                  
[0165/0170] saving model.layers.27.mlp.down_proj.weight
                                                                  
[0166/0170] saving model.layers.27.mlp.gate_up_proj.weight
                                                                  
[0167/0170] saving model.layers.27.post_attention_layernorm.weight
                                                                  
[0168/0170] saving model.layers.27.self_attn.qkv_proj.weight
                                                                  
[0169/0170] saving model.layers.27.self_attn.o_proj.weight
                                                                  
[0170/0170] saving model.norm.weight
+All finished, 113 total shards committed, record saved to /tmp/tmp2aw6bqmy/ndarray-cache.json
diff --git a/mlc-chat-config.json b/mlc-chat-config.json
new file mode 100644
index 0000000000000000000000000000000000000000..912522cbe32203f1e2fc4ef6f06a6b6dbc602512
--- /dev/null
+++ b/mlc-chat-config.json
@@ -0,0 +1,45 @@
+{
+  "model_type": "gemma",
+  "quantization": "q0f16",
+  "model_config": {
+    "hidden_size": 3072,
+    "hidden_act": "gelu",
+    "intermediate_size": 24576,
+    "attention_bias": false,
+    "num_attention_heads": 16,
+    "num_key_value_heads": 16,
+    "head_dim": 256,
+    "num_hidden_layers": 28,
+    "rms_norm_eps": 1e-06,
+    "vocab_size": 256000,
+    "position_embedding_base": 10000.0,
+    "context_window_size": 8192,
+    "prefill_chunk_size": 1024,
+    "tensor_parallel_shards": 1,
+    "max_batch_size": 80
+  },
+  "vocab_size": 256000,
+  "context_window_size": 8192,
+  "sliding_window_size": -1,
+  "prefill_chunk_size": 1024,
+  "attention_sink_size": -1,
+  "tensor_parallel_shards": 1,
+  "mean_gen_len": 128,
+  "max_gen_len": 512,
+  "shift_fill_factor": 0.3,
+  "temperature": 0.7,
+  "presence_penalty": 0.0,
+  "frequency_penalty": 0.0,
+  "repetition_penalty": 1.0,
+  "top_p": 0.95,
+  "conv_template": "gemma_instruction",
+  "pad_token_id": 0,
+  "bos_token_id": 2,
+  "eos_token_id": 1,
+  "tokenizer_files": [
+    "tokenizer.model",
+    "tokenizer.json",
+    "tokenizer_config.json"
+  ],
+  "version": "0.1.0"
+}
\ No newline at end of file
diff --git a/ndarray-cache.json b/ndarray-cache.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b304f9a82a07f365dd900b5ed21ecd34b20e980
--- /dev/null
+++ b/ndarray-cache.json
@@ -0,0 +1,2726 @@
+{
+    "metadata": {
+        "ParamSize": 170,
+        "ParamBytes": 17075361792.0,
+        "BitsPerParam": 16.0
+    },
+    "records": [
+        {
+            "dataPath": "params_shard_0.bin",
+            "format": "raw-shard",
+            "nbytes": 1572864000,
+            "records": [
+                {
+                    "name": "model.embed_tokens.weight",
+                    "shape": [
+                        256000,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 1572864000,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c60082db0e4d8f35e7289fd7de2d6953"
+        },
+        {
+            "dataPath": "params_shard_1.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.0.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "0f4e2e963889085e8afe6bc6c49f5572"
+        },
+        {
+            "dataPath": "params_shard_2.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.0.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "14fb73c20a944f08d1ca3fbc249a4dcb"
+        },
+        {
+            "dataPath": "params_shard_3.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.0.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "86a00af176de654eddef55171a294f02"
+        },
+        {
+            "dataPath": "params_shard_4.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.1.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "f95cbe4ce39a600c2e2ea3b4a79d3506"
+        },
+        {
+            "dataPath": "params_shard_5.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.1.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "58c5b02dbdfb6f4ad7dc07a987487d81"
+        },
+        {
+            "dataPath": "params_shard_6.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.1.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "f824e83e1c09800fed5d9e501cbef9ed"
+        },
+        {
+            "dataPath": "params_shard_7.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.1.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "6b4b1df277cfbc3b70d620be9b0fd82f"
+        },
+        {
+            "dataPath": "params_shard_8.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.2.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "aa05d7ae7d4eb220f37747d841c39f41"
+        },
+        {
+            "dataPath": "params_shard_9.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.2.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "a04a7a8dfba3225e5c9e4d08f5d991c8"
+        },
+        {
+            "dataPath": "params_shard_10.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.2.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "69642bf15d15e8e3d0a50df2c87b3c79"
+        },
+        {
+            "dataPath": "params_shard_11.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.2.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "1bd0562941d5344f2f2981de3eda76b5"
+        },
+        {
+            "dataPath": "params_shard_12.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.3.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "5241c2d095d69394ec4ea68ba6f23f9e"
+        },
+        {
+            "dataPath": "params_shard_13.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.3.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "1b1b08abf4a0d04a65bf68736d8673f7"
+        },
+        {
+            "dataPath": "params_shard_14.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.3.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "6510c5f58389931a2d5e1943a4f6aa2e"
+        },
+        {
+            "dataPath": "params_shard_15.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.3.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "4ba676881606031a5ee0b74e33fb88de"
+        },
+        {
+            "dataPath": "params_shard_16.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.4.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "40d0b41abee65ea73e3a4c82a6fa65c3"
+        },
+        {
+            "dataPath": "params_shard_17.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.4.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "f3187d1dc182a8f95ceab7df293cb152"
+        },
+        {
+            "dataPath": "params_shard_18.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.4.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "0a1d0e2b8358cd2b4856c73c250dd3a3"
+        },
+        {
+            "dataPath": "params_shard_19.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.4.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "aa2041a5903e8b1079fa61a8cd603429"
+        },
+        {
+            "dataPath": "params_shard_20.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.5.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "09e715068b771738ffa53ace091b7f0b"
+        },
+        {
+            "dataPath": "params_shard_21.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.5.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "4054e3e3ba3296ba506ab84288192da9"
+        },
+        {
+            "dataPath": "params_shard_22.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.5.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "194504b7b8ab52be14b2581a772ba2d1"
+        },
+        {
+            "dataPath": "params_shard_23.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.5.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "19d6c58020599c295a7c72322240d15f"
+        },
+        {
+            "dataPath": "params_shard_24.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.6.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "02f69e4867691d3b47d95927a56854ab"
+        },
+        {
+            "dataPath": "params_shard_25.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.6.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "6ae680ce108a1e6d457056c82ae18bcb"
+        },
+        {
+            "dataPath": "params_shard_26.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.10.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "547cbecf7fd4ee5e09053084470211c8"
+        },
+        {
+            "dataPath": "params_shard_27.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.10.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "78a402b92e7a0b524531fc7229cfe3eb"
+        },
+        {
+            "dataPath": "params_shard_28.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.10.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "bbe991c38a3544a1a32286d300efdf3d"
+        },
+        {
+            "dataPath": "params_shard_29.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.10.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "eff50d4dc40d27ff7f0b45d9361ee8fa"
+        },
+        {
+            "dataPath": "params_shard_30.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.11.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "1e65975a0acf68aa3dead36688fae7c6"
+        },
+        {
+            "dataPath": "params_shard_31.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.11.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "66d5d789c9a68fc43f0432dee2025af1"
+        },
+        {
+            "dataPath": "params_shard_32.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.11.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "21b8e6c6330437b35c12bac84fc5aa55"
+        },
+        {
+            "dataPath": "params_shard_33.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.11.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "297b687fc59ee54860e1e086a189d049"
+        },
+        {
+            "dataPath": "params_shard_34.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.12.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "64c4395080e11b49919f611fac8f5bd0"
+        },
+        {
+            "dataPath": "params_shard_35.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.12.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "2168bb7b517618736468495b64de6492"
+        },
+        {
+            "dataPath": "params_shard_36.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.12.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "7601fdd9f5c4172ad568a88b92780d4f"
+        },
+        {
+            "dataPath": "params_shard_37.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.12.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "6dced480e397d0e4b0f4e229e59168b3"
+        },
+        {
+            "dataPath": "params_shard_38.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.13.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "b7a2901c6cb2f4e1a84b2559bee2c205"
+        },
+        {
+            "dataPath": "params_shard_39.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.13.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "668d44bcf32609f674422885adbb2568"
+        },
+        {
+            "dataPath": "params_shard_40.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.13.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "9f7f06099c20e9561b85bbbd71083a9d"
+        },
+        {
+            "dataPath": "params_shard_41.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.13.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "36e8659919fcc7c0e9cd15bf7ee20638"
+        },
+        {
+            "dataPath": "params_shard_42.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.14.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "69272f810a63454b7e493ca8db2e4e01"
+        },
+        {
+            "dataPath": "params_shard_43.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.14.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "97ec6c6098706829215b9775c002b32b"
+        },
+        {
+            "dataPath": "params_shard_44.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.14.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "1b79bfda3ec65213084ac12ceee29f33"
+        },
+        {
+            "dataPath": "params_shard_45.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.14.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "b807f3ab603cd37a846676d193f9d277"
+        },
+        {
+            "dataPath": "params_shard_46.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.15.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "e5dd0bf094a56e3d175d2fdf163b8b7d"
+        },
+        {
+            "dataPath": "params_shard_47.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.15.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "b5bcc5621ca7a57eb12c483feeb43307"
+        },
+        {
+            "dataPath": "params_shard_48.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.6.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "8217f638eeb1e46fe3868fd8e8bf100e"
+        },
+        {
+            "dataPath": "params_shard_49.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.6.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "dba591e829aac359033b6e918382810e"
+        },
+        {
+            "dataPath": "params_shard_50.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.7.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "dbc527f7fa51b6884d6653380f293d2a"
+        },
+        {
+            "dataPath": "params_shard_51.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.7.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "f1d1a0bb4f98bd51126c57fe77684dce"
+        },
+        {
+            "dataPath": "params_shard_52.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.7.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "f58da53e30d94e45a0a369e0a1240ee2"
+        },
+        {
+            "dataPath": "params_shard_53.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.7.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "1e27f4032aabc28bbe3cd547f8e186d7"
+        },
+        {
+            "dataPath": "params_shard_54.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.8.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "408fc336ff7f18a7c83fb7fa603a4998"
+        },
+        {
+            "dataPath": "params_shard_55.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.8.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "538df5b59596e2902efe2bc0af1568b6"
+        },
+        {
+            "dataPath": "params_shard_56.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.8.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "e814405b553d7ec957cad8f3d616ebc7"
+        },
+        {
+            "dataPath": "params_shard_57.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.8.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "3f81f8bc3e1646d20574a7bf8d96b0d5"
+        },
+        {
+            "dataPath": "params_shard_58.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.9.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "e1a1953eb2ccac5af2d7590e6d2b5ec5"
+        },
+        {
+            "dataPath": "params_shard_59.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.9.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "ce270223bc558db0a5d27af2116fba53"
+        },
+        {
+            "dataPath": "params_shard_60.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.9.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "59e777a792b4913307bfda4cf4075a71"
+        },
+        {
+            "dataPath": "params_shard_61.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.9.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c7a7d717f3b9f832dcaef9749a97c74f"
+        },
+        {
+            "dataPath": "params_shard_62.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.15.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "385607fea47c6f19a9fd39de55901040"
+        },
+        {
+            "dataPath": "params_shard_63.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.15.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c7c9818761d650fea693ab80b98a9d1d"
+        },
+        {
+            "dataPath": "params_shard_64.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.16.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "48ea87d4d016f945999069724103aac9"
+        },
+        {
+            "dataPath": "params_shard_65.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.16.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "0ff42527fb75bd5513d8c18ab7c59d98"
+        },
+        {
+            "dataPath": "params_shard_66.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.16.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "2d87584e9446fb1ebf02c1f1709f3ab1"
+        },
+        {
+            "dataPath": "params_shard_67.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.16.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "a952eddcef19e87acba07a874b8f37e7"
+        },
+        {
+            "dataPath": "params_shard_68.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.17.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "12fda1b30881783dbe2e19b782ad94fa"
+        },
+        {
+            "dataPath": "params_shard_69.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.17.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "2b7afa8bf48f008e0ebed5abcd2f15ea"
+        },
+        {
+            "dataPath": "params_shard_70.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.17.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "260f73d27ed48ba8a642b899eec416bc"
+        },
+        {
+            "dataPath": "params_shard_71.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.17.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "9d65ef1f2796ef68e935d6014ca5ba1a"
+        },
+        {
+            "dataPath": "params_shard_72.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.18.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "3171884795345165c875daa55a17fe94"
+        },
+        {
+            "dataPath": "params_shard_73.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.18.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "2cfb567bc6879de46040a7f8680d8066"
+        },
+        {
+            "dataPath": "params_shard_74.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.18.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "7c1f33a25629a88883fd6c1526fda4a8"
+        },
+        {
+            "dataPath": "params_shard_75.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.18.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "dcd7f2a5e7045ff5cf5f581832219172"
+        },
+        {
+            "dataPath": "params_shard_76.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.19.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "fbaa14547db7f71b1885972de54cf52a"
+        },
+        {
+            "dataPath": "params_shard_77.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.19.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "b73203d60d4898d6f198982b94ef17f3"
+        },
+        {
+            "dataPath": "params_shard_78.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.19.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "206abc6ecd1f274098f872f4881e3cca"
+        },
+        {
+            "dataPath": "params_shard_79.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.19.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "4428c7ab551722a23e8d07042b9f88d1"
+        },
+        {
+            "dataPath": "params_shard_80.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.20.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "00c3310f6c68029e3681e6e29acea1f3"
+        },
+        {
+            "dataPath": "params_shard_81.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.20.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "6e8f193ec13343dd3282e3c72cf16498"
+        },
+        {
+            "dataPath": "params_shard_82.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.20.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "84e566f57f2c1c687fe22e5a709c75d4"
+        },
+        {
+            "dataPath": "params_shard_83.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.20.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "0b4e188c7aab9c318bc26bc00fa6eaa7"
+        },
+        {
+            "dataPath": "params_shard_84.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.21.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "6b956e65718e312b8f64d65480a7eb73"
+        },
+        {
+            "dataPath": "params_shard_85.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.21.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "aa1eba42e21c4a3a209ed123807a5cc5"
+        },
+        {
+            "dataPath": "params_shard_86.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.21.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "9c593249fc8574d8d11888f36a4ebc8f"
+        },
+        {
+            "dataPath": "params_shard_87.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.21.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "9a9767c0741c14fb34c278847729ca41"
+        },
+        {
+            "dataPath": "params_shard_88.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.22.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "fefa508ad96e075de33755d3e77c0b02"
+        },
+        {
+            "dataPath": "params_shard_89.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.22.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "272780eda760a1b632ee0898bd822bed"
+        },
+        {
+            "dataPath": "params_shard_90.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.22.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "0c1787fdf9ada9a3cbc2615a3557dc6e"
+        },
+        {
+            "dataPath": "params_shard_91.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.22.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "0549b7e823aae895993f740fb5f92d1f"
+        },
+        {
+            "dataPath": "params_shard_92.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.23.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "947ff6f419800cdad216c95bb6eae64b"
+        },
+        {
+            "dataPath": "params_shard_93.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.23.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c03fc0850c2564e0c0b7c2420601f1e9"
+        },
+        {
+            "dataPath": "params_shard_94.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.23.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "a6ff8555eede8310413c7f0edc374b2f"
+        },
+        {
+            "dataPath": "params_shard_95.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.23.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "3f9577771286399bee6784def87c1a18"
+        },
+        {
+            "dataPath": "params_shard_96.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.24.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "cf1fbf37bec4183e93eef194cc87cc26"
+        },
+        {
+            "dataPath": "params_shard_97.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.24.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "31edc9bcc3145aece6c8ee2d7b0583d7"
+        },
+        {
+            "dataPath": "params_shard_98.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.24.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "a90e87939fe6ca58ddf0de8ebd52c62e"
+        },
+        {
+            "dataPath": "params_shard_99.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.24.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c0b0e83fa9c6d85c2b92efc97bb9dcd6"
+        },
+        {
+            "dataPath": "params_shard_100.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.25.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "74f19307deb63ec97ff3aa8f5e1ac00a"
+        },
+        {
+            "dataPath": "params_shard_101.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.25.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "e9d88cc3c3265e7b073f039e020e3fba"
+        },
+        {
+            "dataPath": "params_shard_102.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.25.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c3832849eba8cbf73f87dead864a76d0"
+        },
+        {
+            "dataPath": "params_shard_103.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.25.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "a51bb7fd741c29ef91dd4fadc787e03d"
+        },
+        {
+            "dataPath": "params_shard_104.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.26.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "2ea4aeb190ee39979e1099d1856a93c9"
+        },
+        {
+            "dataPath": "params_shard_105.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.26.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "23e6ae0d6b3145458e7285adee8c25e3"
+        },
+        {
+            "dataPath": "params_shard_106.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.26.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "cd7a693ca49d3883fb57d5064f5aaa5d"
+        },
+        {
+            "dataPath": "params_shard_107.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.26.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "4e8b4c437c8a8c1893756bfe27dd50f8"
+        },
+        {
+            "dataPath": "params_shard_108.bin",
+            "format": "raw-shard",
+            "nbytes": 150994944,
+            "records": [
+                {
+                    "name": "model.layers.27.mlp.down_proj.weight",
+                    "shape": [
+                        3072,
+                        24576
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 150994944,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "b592e53fa6d6c73c73924b4ba2944b56"
+        },
+        {
+            "dataPath": "params_shard_109.bin",
+            "format": "raw-shard",
+            "nbytes": 301989888,
+            "records": [
+                {
+                    "name": "model.layers.27.mlp.gate_up_proj.weight",
+                    "shape": [
+                        49152,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 301989888,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "c6d781f9c489a4bab30b084193c37b1c"
+        },
+        {
+            "dataPath": "params_shard_110.bin",
+            "format": "raw-shard",
+            "nbytes": 75497472,
+            "records": [
+                {
+                    "name": "model.layers.27.self_attn.qkv_proj.weight",
+                    "shape": [
+                        12288,
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 75497472,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "984e89674ec1c07fc87a9a8b3a3e72bb"
+        },
+        {
+            "dataPath": "params_shard_111.bin",
+            "format": "raw-shard",
+            "nbytes": 25165824,
+            "records": [
+                {
+                    "name": "model.layers.27.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 0
+                }
+            ],
+            "md5sum": "ad1ef7e668a5d8ded355ef2621df9998"
+        },
+        {
+            "dataPath": "params_shard_112.bin",
+            "format": "raw-shard",
+            "nbytes": 25516032,
+            "records": [
+                {
+                    "name": "model.layers.0.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 0
+                },
+                {
+                    "name": "model.layers.0.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 6144
+                },
+                {
+                    "name": "model.layers.0.self_attn.o_proj.weight",
+                    "shape": [
+                        3072,
+                        4096
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 25165824,
+                    "byteOffset": 12288
+                },
+                {
+                    "name": "model.layers.1.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25178112
+                },
+                {
+                    "name": "model.layers.1.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25184256
+                },
+                {
+                    "name": "model.layers.2.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25190400
+                },
+                {
+                    "name": "model.layers.2.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25196544
+                },
+                {
+                    "name": "model.layers.3.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25202688
+                },
+                {
+                    "name": "model.layers.3.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25208832
+                },
+                {
+                    "name": "model.layers.4.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25214976
+                },
+                {
+                    "name": "model.layers.4.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25221120
+                },
+                {
+                    "name": "model.layers.5.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25227264
+                },
+                {
+                    "name": "model.layers.5.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25233408
+                },
+                {
+                    "name": "model.layers.10.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25239552
+                },
+                {
+                    "name": "model.layers.10.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25245696
+                },
+                {
+                    "name": "model.layers.11.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25251840
+                },
+                {
+                    "name": "model.layers.11.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25257984
+                },
+                {
+                    "name": "model.layers.12.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25264128
+                },
+                {
+                    "name": "model.layers.12.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25270272
+                },
+                {
+                    "name": "model.layers.13.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25276416
+                },
+                {
+                    "name": "model.layers.13.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25282560
+                },
+                {
+                    "name": "model.layers.14.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25288704
+                },
+                {
+                    "name": "model.layers.14.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25294848
+                },
+                {
+                    "name": "model.layers.6.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25300992
+                },
+                {
+                    "name": "model.layers.6.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25307136
+                },
+                {
+                    "name": "model.layers.7.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25313280
+                },
+                {
+                    "name": "model.layers.7.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25319424
+                },
+                {
+                    "name": "model.layers.8.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25325568
+                },
+                {
+                    "name": "model.layers.8.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25331712
+                },
+                {
+                    "name": "model.layers.9.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25337856
+                },
+                {
+                    "name": "model.layers.9.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25344000
+                },
+                {
+                    "name": "model.layers.15.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25350144
+                },
+                {
+                    "name": "model.layers.15.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25356288
+                },
+                {
+                    "name": "model.layers.16.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25362432
+                },
+                {
+                    "name": "model.layers.16.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25368576
+                },
+                {
+                    "name": "model.layers.17.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25374720
+                },
+                {
+                    "name": "model.layers.17.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25380864
+                },
+                {
+                    "name": "model.layers.18.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25387008
+                },
+                {
+                    "name": "model.layers.18.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25393152
+                },
+                {
+                    "name": "model.layers.19.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25399296
+                },
+                {
+                    "name": "model.layers.19.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25405440
+                },
+                {
+                    "name": "model.layers.20.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25411584
+                },
+                {
+                    "name": "model.layers.20.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25417728
+                },
+                {
+                    "name": "model.layers.21.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25423872
+                },
+                {
+                    "name": "model.layers.21.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25430016
+                },
+                {
+                    "name": "model.layers.22.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25436160
+                },
+                {
+                    "name": "model.layers.22.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25442304
+                },
+                {
+                    "name": "model.layers.23.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25448448
+                },
+                {
+                    "name": "model.layers.23.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25454592
+                },
+                {
+                    "name": "model.layers.24.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25460736
+                },
+                {
+                    "name": "model.layers.24.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25466880
+                },
+                {
+                    "name": "model.layers.25.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25473024
+                },
+                {
+                    "name": "model.layers.25.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25479168
+                },
+                {
+                    "name": "model.layers.26.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25485312
+                },
+                {
+                    "name": "model.layers.26.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25491456
+                },
+                {
+                    "name": "model.layers.27.input_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25497600
+                },
+                {
+                    "name": "model.layers.27.post_attention_layernorm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25503744
+                },
+                {
+                    "name": "model.norm.weight",
+                    "shape": [
+                        3072
+                    ],
+                    "dtype": "float16",
+                    "format": "f32-to-bf16",
+                    "nbytes": 6144,
+                    "byteOffset": 25509888
+                }
+            ],
+            "md5sum": "e8b4f06390ce04a29734c710dfe87aa4"
+        }
+    ]
+}
\ No newline at end of file
diff --git a/params_shard_0.bin b/params_shard_0.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ad9198ecf0f9278607a69dc5dacd5c0914a0a216
--- /dev/null
+++ b/params_shard_0.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2be3ee8010ace2bde14cf19c7a0c3d397674fcea96fcc119fbe99953b4cae48
+size 1572864000
diff --git a/params_shard_1.bin b/params_shard_1.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f5aba509094f8b20c962c571bc83185603bd0dc5
--- /dev/null
+++ b/params_shard_1.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b63aba686c39d40c1900f6bd49a7af165f31987c5a721d2ba4fb0a4d116f996a
+size 150994944
diff --git a/params_shard_10.bin b/params_shard_10.bin
new file mode 100644
index 0000000000000000000000000000000000000000..79cc4994aba49ac6c16b3b6190b868b3095709e1
--- /dev/null
+++ b/params_shard_10.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f495eb463333483ddae4e8e90d58300aa1026db1ed14173bf90f40d00e7c456
+size 75497472
diff --git a/params_shard_100.bin b/params_shard_100.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c5fe49255c8fa4fe1a204790d73504e4bc46a619
--- /dev/null
+++ b/params_shard_100.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e77edb14de946d2713aa63ac299c5be10cc4acff7350349b053a6131057c09b7
+size 150994944
diff --git a/params_shard_101.bin b/params_shard_101.bin
new file mode 100644
index 0000000000000000000000000000000000000000..6614106c684c8be6b5edee4188ba39cd72f7ebb4
--- /dev/null
+++ b/params_shard_101.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e94f28d2eead7d2e978499a0ba47a5989e8943c823dcb2a7c4bc990d751defb7
+size 301989888
diff --git a/params_shard_102.bin b/params_shard_102.bin
new file mode 100644
index 0000000000000000000000000000000000000000..a92aabaf4c363443c515f51ff03ab5e0156bce32
--- /dev/null
+++ b/params_shard_102.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7957f893e6de2dec4a1a7e53907fbd93c858b61c38334ad761a3fbde1c99b8d5
+size 75497472
diff --git a/params_shard_103.bin b/params_shard_103.bin
new file mode 100644
index 0000000000000000000000000000000000000000..aaf44a8f388be9f61d8fc49b0f7287a0eac95574
--- /dev/null
+++ b/params_shard_103.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2865b6a083f81a3c7b0084c4ed4de5d2995259a81da1585fa7addd574e8e8614
+size 25165824
diff --git a/params_shard_104.bin b/params_shard_104.bin
new file mode 100644
index 0000000000000000000000000000000000000000..44e60b5438352b78c66ae306ec97bb7566e39439
--- /dev/null
+++ b/params_shard_104.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23a8be7c40ac9f6a13e7f394c415b5db741152111bb7f3c28243b6063fda5226
+size 150994944
diff --git a/params_shard_105.bin b/params_shard_105.bin
new file mode 100644
index 0000000000000000000000000000000000000000..5964882697095da091820584077bff8e9b235b7f
--- /dev/null
+++ b/params_shard_105.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b8a92c488dedfdd039c7db9f8e2b0f0a775179d9207db6fa125b0f6faebc1ad
+size 301989888
diff --git a/params_shard_106.bin b/params_shard_106.bin
new file mode 100644
index 0000000000000000000000000000000000000000..52706d292bbdb6d4cc4112a3f6460bf5025968e7
--- /dev/null
+++ b/params_shard_106.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f9846ce82f525fca9302fa4cd1ab2ff77e343ac72c27b84110e4c61b20d7dc4
+size 75497472
diff --git a/params_shard_107.bin b/params_shard_107.bin
new file mode 100644
index 0000000000000000000000000000000000000000..806d49743131f27199361fc7846371d71b40afc5
--- /dev/null
+++ b/params_shard_107.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:331a03482989dd78e8db33de63adf5813e9966b79839e9831399af7c63da4d00
+size 25165824
diff --git a/params_shard_108.bin b/params_shard_108.bin
new file mode 100644
index 0000000000000000000000000000000000000000..14f1d20b5c5513748e65c5d2c2cff572310c3c64
--- /dev/null
+++ b/params_shard_108.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0d80d380afa2d664ec8e14e979183d8370e629ae277477736e63463af9e1f2c
+size 150994944
diff --git a/params_shard_109.bin b/params_shard_109.bin
new file mode 100644
index 0000000000000000000000000000000000000000..d576c0f4080881184850fa08e07e819dfdbd7aef
--- /dev/null
+++ b/params_shard_109.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26f95791da0f338a8b8ff371c4291fcb572c58b92a7bdf3dca69eeff794e078c
+size 301989888
diff --git a/params_shard_11.bin b/params_shard_11.bin
new file mode 100644
index 0000000000000000000000000000000000000000..decb1a20348332330e32da7dc0fcecee25e1d975
--- /dev/null
+++ b/params_shard_11.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e67d3e80eb0abbf51ec71c271de42fcd9df95e5bf2b0955c452b3b85a34bcaf2
+size 25165824
diff --git a/params_shard_110.bin b/params_shard_110.bin
new file mode 100644
index 0000000000000000000000000000000000000000..fde5d97591dd93dd69219dff67f45840d21ec6ff
--- /dev/null
+++ b/params_shard_110.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02960ae58f4b1ec29ef42da1d765e6dc6d2923ac5e264b4b1c50d5282279eb64
+size 75497472
diff --git a/params_shard_111.bin b/params_shard_111.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7fab7548484de120dd1aeba013a61bcd7e4e3ca6
--- /dev/null
+++ b/params_shard_111.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f407f0fbe370718390b20ba882c6ac5d7824ef7ff5290121daa87ceac8fd422
+size 25165824
diff --git a/params_shard_112.bin b/params_shard_112.bin
new file mode 100644
index 0000000000000000000000000000000000000000..bbd9ff7c6bd606955afd158d696e05eb4b0d01b8
--- /dev/null
+++ b/params_shard_112.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:315087c1b5601a7f39845ef375467a2f1201ddde1f5f000c5eee0c3f059b589e
+size 25516032
diff --git a/params_shard_12.bin b/params_shard_12.bin
new file mode 100644
index 0000000000000000000000000000000000000000..2fce4f4e410672671af0c01b0ee5864657b8e2a6
--- /dev/null
+++ b/params_shard_12.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c45207b1c3d735eea4d40f023d5028c9bde3a52534fec34a0899f2fd5f72f7ee
+size 150994944
diff --git a/params_shard_13.bin b/params_shard_13.bin
new file mode 100644
index 0000000000000000000000000000000000000000..80534655b683fc060bfb8645bc9a9f2aad1feda7
--- /dev/null
+++ b/params_shard_13.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:370ab0e337c9ba4cd4ad8c57a8b49c119e315f054666f115710b1e2b4339852e
+size 301989888
diff --git a/params_shard_14.bin b/params_shard_14.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ab414716e8295edbca0f7eb28f6948b797c6a2b7
--- /dev/null
+++ b/params_shard_14.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bdb922fc2bd70dd767789ae7514f292fce11af8ba8a37a623d4d74514fb7ead
+size 75497472
diff --git a/params_shard_15.bin b/params_shard_15.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8f2e9e1e2e29b77c5bacfe44410e228eec97b8fb
--- /dev/null
+++ b/params_shard_15.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3874a8054f4bf5254c2f75c30f71dcdf1d67786b70c409ff6ca25d264075bf3
+size 25165824
diff --git a/params_shard_16.bin b/params_shard_16.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7cdd89d18f9756f79a986319a88baf2337c875d2
--- /dev/null
+++ b/params_shard_16.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3eef07184feaae3b5c6a3f7ebfcac69937c0881f6ee97263f014dddfea1e3e41
+size 150994944
diff --git a/params_shard_17.bin b/params_shard_17.bin
new file mode 100644
index 0000000000000000000000000000000000000000..cc04295fa59c0a3a16b82f08a83880492fc01d11
--- /dev/null
+++ b/params_shard_17.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a05e3105f0e74b75239583869d2704f969be23ad9b142623481ec68b597805b1
+size 301989888
diff --git a/params_shard_18.bin b/params_shard_18.bin
new file mode 100644
index 0000000000000000000000000000000000000000..0e74fd5c05cc770c6cf517c819e8933dc1c8564d
--- /dev/null
+++ b/params_shard_18.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b1d79cb6e2dee6661f100c503306df2bcd5f05020ffb1a6cba3ea51b5fc0990
+size 75497472
diff --git a/params_shard_19.bin b/params_shard_19.bin
new file mode 100644
index 0000000000000000000000000000000000000000..03850807916b065984b72994af4799e8bcdfef62
--- /dev/null
+++ b/params_shard_19.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:815ff7b9e9b9ca24bdc9ef369c7a7ec58dd9d8129f0ecc5a13457ac92352bca4
+size 25165824
diff --git a/params_shard_2.bin b/params_shard_2.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f7b160ad78db72f2d72ccfc3620e69ec855f852c
--- /dev/null
+++ b/params_shard_2.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33420175d82ed8ac5a9fb33c402f84a5f2f5d8dbeaa25430bbfb4dcc679b914e
+size 301989888
diff --git a/params_shard_20.bin b/params_shard_20.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c0b450ad84f543418f45d9775261a523b65b51a0
--- /dev/null
+++ b/params_shard_20.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a81464bdc4c5f52054b7de4631efd631abe6129b5d2405d6fdb3143df6d61f0
+size 150994944
diff --git a/params_shard_21.bin b/params_shard_21.bin
new file mode 100644
index 0000000000000000000000000000000000000000..feb4796c45226a15a06dd7857777b72a6d00c0ff
--- /dev/null
+++ b/params_shard_21.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92320ffeaefaf928b948e5532932cd03e2385de28d2f2039e75be8ae1b806ef0
+size 301989888
diff --git a/params_shard_22.bin b/params_shard_22.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ae731cc1ca4c6c5ddca9d3efd57d7fcb9328ee7e
--- /dev/null
+++ b/params_shard_22.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9edd71672f7d83d8b0b5d532a69e2853bbdf0c8f891b14b62221c92aa4f4e1e
+size 75497472
diff --git a/params_shard_23.bin b/params_shard_23.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9ce71c17fa2488c5681512339f5ba1e78d127545
--- /dev/null
+++ b/params_shard_23.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b30b1ea827a35f1a11b9499d3d06d404693cb8f868926ebf91bb4826d7f031f
+size 25165824
diff --git a/params_shard_24.bin b/params_shard_24.bin
new file mode 100644
index 0000000000000000000000000000000000000000..44dd6be96e9971d704ced5ab0c3e04853af695c2
--- /dev/null
+++ b/params_shard_24.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f45d8549d5724b3b2429df849fe0b43b9a3ba16a78b0280dd26c047a42092293
+size 75497472
diff --git a/params_shard_25.bin b/params_shard_25.bin
new file mode 100644
index 0000000000000000000000000000000000000000..20de1babf043b34c8a1fa8c526cc87e67bd05516
--- /dev/null
+++ b/params_shard_25.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2258782464b090e5aa8ac8bd80f24b0d795c965c95d9e44e8b11a0dec6864e45
+size 25165824
diff --git a/params_shard_26.bin b/params_shard_26.bin
new file mode 100644
index 0000000000000000000000000000000000000000..de84727014c978bc246d6d60ae96e30a1081fe10
--- /dev/null
+++ b/params_shard_26.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21fd4fc3fe70e7fdc7a445893ffa1f8ea0e0a734963b92abf8c6971b2bf16f32
+size 150994944
diff --git a/params_shard_27.bin b/params_shard_27.bin
new file mode 100644
index 0000000000000000000000000000000000000000..74ad6111e00258801c9fb65a2192d08492efe526
--- /dev/null
+++ b/params_shard_27.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc337a2b5d750d414cf181bd10a9db1a471e87368198556274df0514991fedd7
+size 301989888
diff --git a/params_shard_28.bin b/params_shard_28.bin
new file mode 100644
index 0000000000000000000000000000000000000000..28bb881720f66502c314f62ae66991d361b1f408
--- /dev/null
+++ b/params_shard_28.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d7435717d68ff5a285eeb40df5131ecb6f129fd0eea330dfe9bbf3030818d3e
+size 75497472
diff --git a/params_shard_29.bin b/params_shard_29.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8040bcc48049da4831337221736677ef4f8c35e
--- /dev/null
+++ b/params_shard_29.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3ff1c006811268f00687ec24afa52dbcecd46f0a26fa309b95a5f882b3506f3
+size 25165824
diff --git a/params_shard_3.bin b/params_shard_3.bin
new file mode 100644
index 0000000000000000000000000000000000000000..5e1bad2eb29dd1e2baf62c1dc3f2209402e6a4ad
--- /dev/null
+++ b/params_shard_3.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:895b39f41fa1d329b78cdef54ef513b4bdb4b66012b8870e4b432e49f97bb8df
+size 75497472
diff --git a/params_shard_30.bin b/params_shard_30.bin
new file mode 100644
index 0000000000000000000000000000000000000000..2f4e8e417a49ed3e3f2ba75dce751237d6c1aca7
--- /dev/null
+++ b/params_shard_30.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:335b12431ddf3e62ac831542cb38c665511f712c01b23107ea6a51d07d3dcd87
+size 150994944
diff --git a/params_shard_31.bin b/params_shard_31.bin
new file mode 100644
index 0000000000000000000000000000000000000000..69b5c6ecf1983ce2cef0f2a0f074be83439b6ca4
--- /dev/null
+++ b/params_shard_31.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47eb0196f808b61ce4645f186172ea32404047ac41e18105fa61232632072281
+size 301989888
diff --git a/params_shard_32.bin b/params_shard_32.bin
new file mode 100644
index 0000000000000000000000000000000000000000..349f6eaa39654673ea16e7206c25a25c7f9be7f6
--- /dev/null
+++ b/params_shard_32.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1fe00114c9a5c81ba92167a83fdccfc4b8d4d7616b3c57b4e8137114b9cbe1d
+size 75497472
diff --git a/params_shard_33.bin b/params_shard_33.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ca6b43ff64dd75008f3eb4922f3ef8b3af388f36
--- /dev/null
+++ b/params_shard_33.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28bca728997a0de46a6d227049c9fabd0470bf0c6e0a16871d0b67ec5aefb941
+size 25165824
diff --git a/params_shard_34.bin b/params_shard_34.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1d2715a4d3b8baea0654ed99c2881157a57131bb
--- /dev/null
+++ b/params_shard_34.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc2537b5584183d5dd78ef7ad52a654eb5013fbec6c72b1397ffebfcfc2277da
+size 150994944
diff --git a/params_shard_35.bin b/params_shard_35.bin
new file mode 100644
index 0000000000000000000000000000000000000000..d0031c5c9bc038bcf2cc3c0d2291b5e3c5fce24c
--- /dev/null
+++ b/params_shard_35.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57b62f2e4bc80405a5c005c9c45e28b851e8624ca3d3d5b6810956af5338b2f0
+size 301989888
diff --git a/params_shard_36.bin b/params_shard_36.bin
new file mode 100644
index 0000000000000000000000000000000000000000..6842ca32d932137e659c7e50ed8af29336c6f7b2
--- /dev/null
+++ b/params_shard_36.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce7ba63a7eff5da2902de541fda5e89c7cfcbf5778b5b9eab127ef6ad30e2cb0
+size 75497472
diff --git a/params_shard_37.bin b/params_shard_37.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7210a5b16e9152a5fe2a8edffe374eaa84760f29
--- /dev/null
+++ b/params_shard_37.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0be92750790ee9eafbe01cb17bf66039ba951b5ec7be3b1d4567b6d0bf8d2cd
+size 25165824
diff --git a/params_shard_38.bin b/params_shard_38.bin
new file mode 100644
index 0000000000000000000000000000000000000000..cf889ba660645c36552cfd45575f51e4fa538533
--- /dev/null
+++ b/params_shard_38.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84db54f8b3b3e0e75550bdbb36dd7409eebace73c2375412204ffb0442f3e6c8
+size 150994944
diff --git a/params_shard_39.bin b/params_shard_39.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ab35019107583222e6343298a9194f624f779e1c
--- /dev/null
+++ b/params_shard_39.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b533513a1e9def6f32690967d9f23cf7a648c3f149fa315159ed79172ec6bfcd
+size 301989888
diff --git a/params_shard_4.bin b/params_shard_4.bin
new file mode 100644
index 0000000000000000000000000000000000000000..b45a46471d4aab66f9728993767706e25e1181e3
--- /dev/null
+++ b/params_shard_4.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf15483ce6dd08c10f6cd4cf3d657117d9405a11ea4812e378be0f249baca074
+size 150994944
diff --git a/params_shard_40.bin b/params_shard_40.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9b61cf3a157d201e467e3e224c98b66925c8f5af
--- /dev/null
+++ b/params_shard_40.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c18bef757b621dbea17bc989e6fe303571a6cc83586b3a7d4695df2a2bb1f730
+size 75497472
diff --git a/params_shard_41.bin b/params_shard_41.bin
new file mode 100644
index 0000000000000000000000000000000000000000..42f3183f8c35c5938a48b7e2092f226b1c473447
--- /dev/null
+++ b/params_shard_41.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8367f30859a789d51a90cb80f83c4539152144d841672e0e99fdf7b2838257b0
+size 25165824
diff --git a/params_shard_42.bin b/params_shard_42.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ae514d307b1448b83b268b6ae4908eb29919e01a
--- /dev/null
+++ b/params_shard_42.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c41853740c2e6e69c2bf60b80a2bd6d2fc535ebd8c453516be851765cadeefd
+size 150994944
diff --git a/params_shard_43.bin b/params_shard_43.bin
new file mode 100644
index 0000000000000000000000000000000000000000..472406b8fd5e88f8d97a0ae86af91bfe1e9b29e8
--- /dev/null
+++ b/params_shard_43.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76b81828d4f5250ecbe487daf4c4a5afdf103d7357f2b5ca76948efcffa1e518
+size 301989888
diff --git a/params_shard_44.bin b/params_shard_44.bin
new file mode 100644
index 0000000000000000000000000000000000000000..6ceb4038cf432267e87edc1f85181a36455c08ea
--- /dev/null
+++ b/params_shard_44.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d4cc5b02d48eaab10cf9e4bfde99afbf10e6bcc7c846379e79d123dea16efb28
+size 75497472
diff --git a/params_shard_45.bin b/params_shard_45.bin
new file mode 100644
index 0000000000000000000000000000000000000000..045f7931382d2d6f50d9e8adcc852cdd9ac24595
--- /dev/null
+++ b/params_shard_45.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6907f94bd2de42e805777f59a17285bb4459d9fa0891393f8675c4b33b01c11
+size 25165824
diff --git a/params_shard_46.bin b/params_shard_46.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8dc533de62f54af07dc499de14a0aa39f92e856c
--- /dev/null
+++ b/params_shard_46.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:489e434c32f3f5f94c1ac58f23d4b1535af7c9166acd456573b36a4bde051e27
+size 75497472
diff --git a/params_shard_47.bin b/params_shard_47.bin
new file mode 100644
index 0000000000000000000000000000000000000000..85f5bb8aa654f0c387e52fcc82d404c8ce8bc8c9
--- /dev/null
+++ b/params_shard_47.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69124045ed87aab093d3f2fe8de97dc7d201bde5596d575038ee8cd57cceeed3
+size 25165824
diff --git a/params_shard_48.bin b/params_shard_48.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1e5afff76c4249d8bb881661267339970cf99df9
--- /dev/null
+++ b/params_shard_48.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f16f7210df8214425a0d540945747fe4ba6bb71df32bb766d06982146163a4e8
+size 150994944
diff --git a/params_shard_49.bin b/params_shard_49.bin
new file mode 100644
index 0000000000000000000000000000000000000000..a61eb703a14c78f752d76cbc3c4897a7ea85293c
--- /dev/null
+++ b/params_shard_49.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:482c3632242a12a5e6fba73e209f4bdd73b3d4ebbcb8dc5717c4bbca6f4d7a5e
+size 301989888
diff --git a/params_shard_5.bin b/params_shard_5.bin
new file mode 100644
index 0000000000000000000000000000000000000000..0de7973d8f637616d555df86cd6377c297ea79e3
--- /dev/null
+++ b/params_shard_5.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e454a1eca400f951efdaa1a6ade90c3e092ccc8713b66df9cb8c8e10760e61d
+size 301989888
diff --git a/params_shard_50.bin b/params_shard_50.bin
new file mode 100644
index 0000000000000000000000000000000000000000..01bd71cecf76a00ed90688939b1b09baffba9f88
--- /dev/null
+++ b/params_shard_50.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:940a406c7be5c866af6e3455006e959b5abdeab9b5e0dd06ea124a74b321b0c4
+size 150994944
diff --git a/params_shard_51.bin b/params_shard_51.bin
new file mode 100644
index 0000000000000000000000000000000000000000..844474c5266ad9ba8b9e491817531fb443ee678a
--- /dev/null
+++ b/params_shard_51.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:196fce2d3992420a73631bc35257d9b1f691824bf49395a44fd1b172bcb494a9
+size 301989888
diff --git a/params_shard_52.bin b/params_shard_52.bin
new file mode 100644
index 0000000000000000000000000000000000000000..46d87f1d70315a8e6d3c5efc250ee505fbc4b978
--- /dev/null
+++ b/params_shard_52.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44863a778e6d7af7b6263f08ce3dcff07849cdf98b35172d811cac32cea5de4a
+size 75497472
diff --git a/params_shard_53.bin b/params_shard_53.bin
new file mode 100644
index 0000000000000000000000000000000000000000..87b7a65f9e2bb21a444e93d3d38adf3a8b8e81f3
--- /dev/null
+++ b/params_shard_53.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a9eef2e7d38888863bc3bc5efd54309543b97c2e1c7562633a7c87802cc56ed
+size 25165824
diff --git a/params_shard_54.bin b/params_shard_54.bin
new file mode 100644
index 0000000000000000000000000000000000000000..445f6b35a664aced41505d2a01a8e95e99fd95f9
--- /dev/null
+++ b/params_shard_54.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e37c2775a572635bbbf29097604494c7435ab99eb97d2f8cebcff9cb6ecd230
+size 150994944
diff --git a/params_shard_55.bin b/params_shard_55.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9c09846ec075c9ae432077959bbb7f3f13570748
--- /dev/null
+++ b/params_shard_55.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ca216606ea1df4cd175916f04d105419e3aa937822b1ac8312423be83238eb4
+size 301989888
diff --git a/params_shard_56.bin b/params_shard_56.bin
new file mode 100644
index 0000000000000000000000000000000000000000..3878357a76668d1042a658ddf650b1ecae396861
--- /dev/null
+++ b/params_shard_56.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fc0f27879be833d1dcfffcbf9f5dda8f89acbe64ae6db3f912b60e6201193ea
+size 75497472
diff --git a/params_shard_57.bin b/params_shard_57.bin
new file mode 100644
index 0000000000000000000000000000000000000000..b5657df8b1b90c8833aa30b8113d9e533051b4b7
--- /dev/null
+++ b/params_shard_57.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb563a62c6e8a505c1ae5997429a5059cf62c9301f5ea942a93f59422b5213aa
+size 25165824
diff --git a/params_shard_58.bin b/params_shard_58.bin
new file mode 100644
index 0000000000000000000000000000000000000000..dfb0d9a59d6213113ec023892dace844278db4e2
--- /dev/null
+++ b/params_shard_58.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17664f068f4a0a3e9f0106ccb37662711146be2144ff0a713169f5fdf690e4ed
+size 150994944
diff --git a/params_shard_59.bin b/params_shard_59.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8894dbc8a19b5acc81bcfcf1b38fa06f4ca283b8
--- /dev/null
+++ b/params_shard_59.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c3082b61415cb1dc5e6a79edeb78d2c00a1d146e2df6bbf90b99011638b5cd0
+size 301989888
diff --git a/params_shard_6.bin b/params_shard_6.bin
new file mode 100644
index 0000000000000000000000000000000000000000..89923cf611deedab5047ecd68072c02f0a8fcdaa
--- /dev/null
+++ b/params_shard_6.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32f45a826cb38c20abdcc4f4a479545edcd048ee35a1677bb27b61f7e0d3b177
+size 75497472
diff --git a/params_shard_60.bin b/params_shard_60.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8651bcfc1039a666d48ff305901805f2f4391db
--- /dev/null
+++ b/params_shard_60.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0926c082b6acf0af78fed7016ce80dd25f81d10fc6a3fa1613a40af4ff7b4ea7
+size 75497472
diff --git a/params_shard_61.bin b/params_shard_61.bin
new file mode 100644
index 0000000000000000000000000000000000000000..409720436fbbb96950c50457327fb9b406b19582
--- /dev/null
+++ b/params_shard_61.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d684c1452f9686be2b5000032c77fb2ce3a6e323a97cf509459d3b6f95adeb13
+size 25165824
diff --git a/params_shard_62.bin b/params_shard_62.bin
new file mode 100644
index 0000000000000000000000000000000000000000..3a417981a0ab9a908820ab9e7a359e3e3c250b23
--- /dev/null
+++ b/params_shard_62.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b952a4090ad89749f8b4b9c1ce475701e85bfe99cbe4b312382d753850f4fff5
+size 150994944
diff --git a/params_shard_63.bin b/params_shard_63.bin
new file mode 100644
index 0000000000000000000000000000000000000000..e30b596161556068fe8372072c8cc08af5127357
--- /dev/null
+++ b/params_shard_63.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fea0a9ff021a357246e1c6e08e1c8aeffd06ac7d398739b231854e4682ada4f1
+size 301989888
diff --git a/params_shard_64.bin b/params_shard_64.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ed2fd4a36ad2dcf506fe48303027317550d692c6
--- /dev/null
+++ b/params_shard_64.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b03985f174ef59742eca58f3ffe312d6f7bbaa693b0f9c3ca5554f2e5ad9a78e
+size 150994944
diff --git a/params_shard_65.bin b/params_shard_65.bin
new file mode 100644
index 0000000000000000000000000000000000000000..3488c94241966732e16822560229bde20f114c99
--- /dev/null
+++ b/params_shard_65.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48e3b44602afdeb71bba569b83ca53305e20e47da66b8b87671ae38b8b4339a5
+size 301989888
diff --git a/params_shard_66.bin b/params_shard_66.bin
new file mode 100644
index 0000000000000000000000000000000000000000..e34939e39272bc2aaf1f0a277beb700cee8fe93e
--- /dev/null
+++ b/params_shard_66.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18906d3e0c16bda4bc18f35d2dfeaced4b00ea2d128a151468b5fc74ff3e4320
+size 75497472
diff --git a/params_shard_67.bin b/params_shard_67.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ded9ccd1d60224edc3839694df3c2009f9ee2ca0
--- /dev/null
+++ b/params_shard_67.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ae4e9668014f6f16254c869a44dccb55c7f5e6482e5a8e0b7a4101562a63815
+size 25165824
diff --git a/params_shard_68.bin b/params_shard_68.bin
new file mode 100644
index 0000000000000000000000000000000000000000..b436e7acec736ef21116a860790f974cf91e8a1a
--- /dev/null
+++ b/params_shard_68.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc2f8600b0686c1a139c158a2e72c09210c83774cb1960814e5b6fd0711ae375
+size 150994944
diff --git a/params_shard_69.bin b/params_shard_69.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8fbe2ab536821009c3b0042a04d4d5c94f25f9f3
--- /dev/null
+++ b/params_shard_69.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e82b5aa40550a94462caf1ce4076dd8a4e34698bda25559712ddd90c29d3564
+size 301989888
diff --git a/params_shard_7.bin b/params_shard_7.bin
new file mode 100644
index 0000000000000000000000000000000000000000..4d8243f84d251ada654e7d8b76d4f6f46f9925f3
--- /dev/null
+++ b/params_shard_7.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff674047a5df0437d8d86e3561947f62be6d4906c00c08542e99ffe668ad3aae
+size 25165824
diff --git a/params_shard_70.bin b/params_shard_70.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ac20bc61ee4b04325c0bd22621828c11b3309e70
--- /dev/null
+++ b/params_shard_70.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2f37e29f9c8c66e20000f4e7bed1b17efababeb26ee60e5eecbf5f1d6bd0bab
+size 75497472
diff --git a/params_shard_71.bin b/params_shard_71.bin
new file mode 100644
index 0000000000000000000000000000000000000000..51f02752eba98b44f85bba496dd620437e3f02bb
--- /dev/null
+++ b/params_shard_71.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fc1f0a69e156b06415ac01a3338a4496fe86d75fa3a4d33ac1c71aff966f665
+size 25165824
diff --git a/params_shard_72.bin b/params_shard_72.bin
new file mode 100644
index 0000000000000000000000000000000000000000..4a8fe29dcb32d18e39a94fb1c9eb3ce83f843b82
--- /dev/null
+++ b/params_shard_72.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1819f1ddb3f588475d2928786c16cc893663e8994dd9ffd1cba35f7a14f9be2
+size 150994944
diff --git a/params_shard_73.bin b/params_shard_73.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9760e72f751a6639f9954179b0d14974b96f71e5
--- /dev/null
+++ b/params_shard_73.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02f0a3f66b16a9dd52d5493caea4eda6297a8f02f47f38b9c81e1884ea8f6c39
+size 301989888
diff --git a/params_shard_74.bin b/params_shard_74.bin
new file mode 100644
index 0000000000000000000000000000000000000000..81948cd1214aa2007ef78523165909cdca479195
--- /dev/null
+++ b/params_shard_74.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99d447e3d03be2b4203a9f6d17cea432b23e1a1faadc4aec123eea7469495449
+size 75497472
diff --git a/params_shard_75.bin b/params_shard_75.bin
new file mode 100644
index 0000000000000000000000000000000000000000..75c3d7031a18a289b2df23d6a99f765f63d2fdb6
--- /dev/null
+++ b/params_shard_75.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de0ff87d048ca250b2d69340a16407144ea73227ec39d0c26002f2b654f1aa00
+size 25165824
diff --git a/params_shard_76.bin b/params_shard_76.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8ea37d7982499df1532c49d29aee4c6ac06f1aba
--- /dev/null
+++ b/params_shard_76.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:498b7faed32c867688fc45ecb4288d76fee0f923cc35feb3f4e6491031f35fac
+size 150994944
diff --git a/params_shard_77.bin b/params_shard_77.bin
new file mode 100644
index 0000000000000000000000000000000000000000..5290e468b4e32a20f0049e69e643c1a0f8cbb12b
--- /dev/null
+++ b/params_shard_77.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2466e6c1cec6dd794e781ffd96c262c06ccad070ed5d626b7b8ecdcc116cf00
+size 301989888
diff --git a/params_shard_78.bin b/params_shard_78.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8e7242cd1c42a64480ff16155cd95aafbff94a8f
--- /dev/null
+++ b/params_shard_78.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b938b63723a137b89206f8ddcc1653da91a3532c0c1e4f324661e290b698362c
+size 75497472
diff --git a/params_shard_79.bin b/params_shard_79.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8dba0d4a73e8ba90e9105151c67d29eb83d79160
--- /dev/null
+++ b/params_shard_79.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c5a1ea25288c41f93550da53ede8898f330fc4d11d6d9a5fd86c2a234d9fcc3
+size 25165824
diff --git a/params_shard_8.bin b/params_shard_8.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8df42c4be9ba32dd11d1b6fba0712db59e3a8b54
--- /dev/null
+++ b/params_shard_8.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1df233272619ab16053f98a7c03cfbb6df39fb323b99166522942f3c0d95fb2f
+size 150994944
diff --git a/params_shard_80.bin b/params_shard_80.bin
new file mode 100644
index 0000000000000000000000000000000000000000..afc2d216680ffd1b76fddaf7670168f70e114df7
--- /dev/null
+++ b/params_shard_80.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8bbdd54aa8dadff531fe3e02f020266d16440650a5e702f47ab8824366164f01
+size 150994944
diff --git a/params_shard_81.bin b/params_shard_81.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9f58a956fc8e81cce3b0181d7cd52d2b405758cc
--- /dev/null
+++ b/params_shard_81.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86f3637c947863e040fa96fa6ba6aee31367ef168f9e0c5785d820798d1b1aa2
+size 301989888
diff --git a/params_shard_82.bin b/params_shard_82.bin
new file mode 100644
index 0000000000000000000000000000000000000000..e23c562ecd68b5d0f66443fa657466817cf454bd
--- /dev/null
+++ b/params_shard_82.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83b58bf9f9af7442b7693b400aa93ab4dff4f64253489951be4e265a3f3c457d
+size 75497472
diff --git a/params_shard_83.bin b/params_shard_83.bin
new file mode 100644
index 0000000000000000000000000000000000000000..4b1280f09c28fbb0669291ec9c027f7d1d63ac8d
--- /dev/null
+++ b/params_shard_83.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57eddabd17a9175d0a5c0eda4e01fb091b72a3660889f3b0c966c3ba75044e2a
+size 25165824
diff --git a/params_shard_84.bin b/params_shard_84.bin
new file mode 100644
index 0000000000000000000000000000000000000000..463872cea17929d459d2f75d77c23dc46b97ae1c
--- /dev/null
+++ b/params_shard_84.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffadabf61174bfe1664c66473598f1be8a96d1739de0f6ad486e3412a67ba3a6
+size 150994944
diff --git a/params_shard_85.bin b/params_shard_85.bin
new file mode 100644
index 0000000000000000000000000000000000000000..0f7b9dea8b819dca047b34dd8ddd493eacf102e7
--- /dev/null
+++ b/params_shard_85.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03d7a6696dd4fc9081a5aa1894d4891d6816679fed8c0fb294fbaf439243a010
+size 301989888
diff --git a/params_shard_86.bin b/params_shard_86.bin
new file mode 100644
index 0000000000000000000000000000000000000000..49b8f6cb7683bcb7674a99d28762422e4bc0f953
--- /dev/null
+++ b/params_shard_86.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fe0c7931b93530df484d63717226f71bc3f8482b639a5fbf3741d83bbb0af67
+size 75497472
diff --git a/params_shard_87.bin b/params_shard_87.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7f00a7f174803d032ae032d45281309b78d95b7d
--- /dev/null
+++ b/params_shard_87.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb7233e2a1d6fb06c15f4512346f3718db0a1ecb59f3848a73034c36000e31e7
+size 25165824
diff --git a/params_shard_88.bin b/params_shard_88.bin
new file mode 100644
index 0000000000000000000000000000000000000000..b6b5d7fabf7412284db635e73a6bab0120616595
--- /dev/null
+++ b/params_shard_88.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17a49f515b6842904e09294dafe9b788e8ef3b4629fbb6642eb39bbf2acd7643
+size 150994944
diff --git a/params_shard_89.bin b/params_shard_89.bin
new file mode 100644
index 0000000000000000000000000000000000000000..0ac508459abd9f8523af269ffe51ff4ad5422ef4
--- /dev/null
+++ b/params_shard_89.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcb50556cf3aa8a1ebd15c419fa6c3524b1d8db96b461b98fb964a17377f5784
+size 301989888
diff --git a/params_shard_9.bin b/params_shard_9.bin
new file mode 100644
index 0000000000000000000000000000000000000000..25001f00ab1e310487a95829b19a9d024898160f
--- /dev/null
+++ b/params_shard_9.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:990b81b620b0161c9fb63394b92af77b22d6198d62e973498a0352c43a598be7
+size 301989888
diff --git a/params_shard_90.bin b/params_shard_90.bin
new file mode 100644
index 0000000000000000000000000000000000000000..e48fca37654f10850b7485410e198c49d2bcf019
--- /dev/null
+++ b/params_shard_90.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76c7314508aa5802359c457be087aaaeebd44a5e058a4c9d9e2db504a4d82b06
+size 75497472
diff --git a/params_shard_91.bin b/params_shard_91.bin
new file mode 100644
index 0000000000000000000000000000000000000000..7378ed96cd232dc17c26719380703b627287ed27
--- /dev/null
+++ b/params_shard_91.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:756b32024d8606c5acc68312511b9db4d939afef0578a1e3ffd1c81293f77b56
+size 25165824
diff --git a/params_shard_92.bin b/params_shard_92.bin
new file mode 100644
index 0000000000000000000000000000000000000000..dbd792230f8038b56dc9461966b8a0ce0e8a03a9
--- /dev/null
+++ b/params_shard_92.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c91c2ab50b444ca72e94c75c4ea8bd44dfd177dc99f47479630e45fc6308e1e
+size 150994944
diff --git a/params_shard_93.bin b/params_shard_93.bin
new file mode 100644
index 0000000000000000000000000000000000000000..ccfe1f71dcac2eae4453250e5aa98bf65635f97e
--- /dev/null
+++ b/params_shard_93.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75a8d7085d86e964755299c1ac769ae7f4a95ead72bdf2d230af5804aae6f87c
+size 301989888
diff --git a/params_shard_94.bin b/params_shard_94.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f7428c2750db02c3947af3b2b39d4a8d2213d08c
--- /dev/null
+++ b/params_shard_94.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37a1815d7003fef13cb9611ad3071254bc6d9e388dad14a4295fcc9573d83433
+size 75497472
diff --git a/params_shard_95.bin b/params_shard_95.bin
new file mode 100644
index 0000000000000000000000000000000000000000..641a0f78d460b0951d6b3d44d87182a32b2fe098
--- /dev/null
+++ b/params_shard_95.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c613fee063dcc6983ad540981bdda6cfda4622be79f625599fb533ee31a0d8ab
+size 25165824
diff --git a/params_shard_96.bin b/params_shard_96.bin
new file mode 100644
index 0000000000000000000000000000000000000000..9408308aed61f02c4f96344a3a076c9746e2dd9c
--- /dev/null
+++ b/params_shard_96.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f2c6dc78a02b67797f728f2e98075cd6a81f1be360a051f28a3c5b429431122
+size 75497472
diff --git a/params_shard_97.bin b/params_shard_97.bin
new file mode 100644
index 0000000000000000000000000000000000000000..49c51772a8356d03b4d570c4520850a59e34b483
--- /dev/null
+++ b/params_shard_97.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b37a42c05fd635c90bcf1c964441e73454d83b07069e38c690af409ebdc69b8
+size 25165824
diff --git a/params_shard_98.bin b/params_shard_98.bin
new file mode 100644
index 0000000000000000000000000000000000000000..fb25bb61d32402b60fa88275cf97246c6a80d8ba
--- /dev/null
+++ b/params_shard_98.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70fb64a68e65d1e7748fd42515ffbdafa3d39fd8b552c9d8dba0007ceacd2409
+size 150994944
diff --git a/params_shard_99.bin b/params_shard_99.bin
new file mode 100644
index 0000000000000000000000000000000000000000..05455246ffa3e1399ae5eb27109d995d6ac818cb
--- /dev/null
+++ b/params_shard_99.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff7fc097507e7b0b741606b5018aa5c3dcb127fb01919d5b359c98ceb00c7a7a
+size 301989888
diff --git a/tokenizer.json b/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..10d1ca087d5e78cbfc41ae083201df86aa031044
--- /dev/null
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05e97791a5e007260de1db7e1692e53150e08cea481e2bf25435553380c147ee
+size 17477929
diff --git a/tokenizer.model b/tokenizer.model
new file mode 100644
index 0000000000000000000000000000000000000000..0fb99fadba40bda6e559565d903f62cfde63b88a
--- /dev/null
+++ b/tokenizer.model
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6969e64047744a44bb3abfb5c50f8de0f7ed8b571d5444426ef931f651d1a0ef
+size 4241111
diff --git a/tokenizer_config.json b/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f66eacb3b9c5afce19c15caa4f822b39bc0aac45
--- /dev/null
+++ b/tokenizer_config.json
@@ -0,0 +1,70 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<eos>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<bos>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "106": {
+      "content": "<start_of_turn>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "107": {
+      "content": "<end_of_turn>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<start_of_turn>",
+    "<end_of_turn>"
+  ],
+  "bos_token": "<bos>",
+  "chat_template": "{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '<start_of_turn>' + role + '\n' + message['content'] | trim + '<end_of_turn>\n' }}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model\n'}}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<eos>",
+  "legacy": null,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}