Upload 24 files
Browse files- mlc-chat-config.json +77 -0
- ndarray-cache-b16.json +639 -0
- ndarray-cache.json +639 -0
- params_shard_0.bin +3 -0
- params_shard_1.bin +3 -0
- params_shard_10.bin +3 -0
- params_shard_11.bin +3 -0
- params_shard_12.bin +3 -0
- params_shard_13.bin +3 -0
- params_shard_14.bin +3 -0
- params_shard_15.bin +3 -0
- params_shard_16.bin +3 -0
- params_shard_17.bin +3 -0
- params_shard_2.bin +3 -0
- params_shard_3.bin +3 -0
- params_shard_4.bin +3 -0
- params_shard_5.bin +3 -0
- params_shard_6.bin +3 -0
- params_shard_7.bin +3 -0
- params_shard_8.bin +3 -0
- params_shard_9.bin +3 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +35 -0
mlc-chat-config.json
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "llama",
|
| 3 |
+
"quantization": "q4f32_1",
|
| 4 |
+
"model_config": {
|
| 5 |
+
"hidden_size": 4096,
|
| 6 |
+
"intermediate_size": 11008,
|
| 7 |
+
"num_attention_heads": 32,
|
| 8 |
+
"num_hidden_layers": 4,
|
| 9 |
+
"rms_norm_eps": 1e-05,
|
| 10 |
+
"vocab_size": 55296,
|
| 11 |
+
"position_embedding_base": 10000.0,
|
| 12 |
+
"context_window_size": 4096,
|
| 13 |
+
"prefill_chunk_size": 2048,
|
| 14 |
+
"num_key_value_heads": 32,
|
| 15 |
+
"head_dim": 128,
|
| 16 |
+
"tensor_parallel_shards": 1,
|
| 17 |
+
"max_batch_size": 80
|
| 18 |
+
},
|
| 19 |
+
"vocab_size": 55296,
|
| 20 |
+
"context_window_size": 4096,
|
| 21 |
+
"sliding_window_size": -1,
|
| 22 |
+
"prefill_chunk_size": 2048,
|
| 23 |
+
"attention_sink_size": -1,
|
| 24 |
+
"tensor_parallel_shards": 1,
|
| 25 |
+
"mean_gen_len": 128,
|
| 26 |
+
"max_gen_len": 512,
|
| 27 |
+
"shift_fill_factor": 0.3,
|
| 28 |
+
"temperature": 0.2,
|
| 29 |
+
"presence_penalty": 0.0,
|
| 30 |
+
"frequency_penalty": 0.0,
|
| 31 |
+
"repetition_penalty": 1.0,
|
| 32 |
+
"top_p": 0.9,
|
| 33 |
+
"conv_template": {
|
| 34 |
+
"name": "llama-2",
|
| 35 |
+
"system_template": "[INST] <<SYS>>\n{system_message}\n<</SYS>>\n\n",
|
| 36 |
+
"system_message": "You are a helpful, respectful and honest assistant.",
|
| 37 |
+
"system_prefix_token_ids": [
|
| 38 |
+
1
|
| 39 |
+
],
|
| 40 |
+
"add_role_after_system_message": false,
|
| 41 |
+
"roles": {
|
| 42 |
+
"user": "<s>[INST]",
|
| 43 |
+
"assistant": "[/INST]",
|
| 44 |
+
"tool": "[INST]"
|
| 45 |
+
},
|
| 46 |
+
"role_templates": {
|
| 47 |
+
"user": "{user_message}",
|
| 48 |
+
"assistant": "{assistant_message}",
|
| 49 |
+
"tool": "{tool_message}"
|
| 50 |
+
},
|
| 51 |
+
"messages": [],
|
| 52 |
+
"seps": [
|
| 53 |
+
" ",
|
| 54 |
+
" </s>"
|
| 55 |
+
],
|
| 56 |
+
"role_content_sep": " ",
|
| 57 |
+
"role_empty_sep": " ",
|
| 58 |
+
"stop_str": [
|
| 59 |
+
"[INST]"
|
| 60 |
+
],
|
| 61 |
+
"stop_token_ids": [
|
| 62 |
+
2
|
| 63 |
+
],
|
| 64 |
+
"function_string": "",
|
| 65 |
+
"use_function_calling": false
|
| 66 |
+
},
|
| 67 |
+
"pad_token_id": 0,
|
| 68 |
+
"bos_token_id": 1,
|
| 69 |
+
"eos_token_id": 2,
|
| 70 |
+
"tokenizer_files": [
|
| 71 |
+
"tokenizer.model",
|
| 72 |
+
"tokenizer.json",
|
| 73 |
+
"tokenizer_config.json"
|
| 74 |
+
],
|
| 75 |
+
"token_table_postproc_method": "byte_fallback",
|
| 76 |
+
"version": "0.1.0"
|
| 77 |
+
}
|
ndarray-cache-b16.json
ADDED
|
@@ -0,0 +1,639 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"ParamSize": 45,
|
| 4 |
+
"ParamBytes": 789200896.0,
|
| 5 |
+
"BitsPerParam": 5.000788364646225
|
| 6 |
+
},
|
| 7 |
+
"records": [
|
| 8 |
+
{
|
| 9 |
+
"dataPath": "params_shard_0.bin",
|
| 10 |
+
"format": "raw-shard",
|
| 11 |
+
"nbytes": 113246208,
|
| 12 |
+
"records": [
|
| 13 |
+
{
|
| 14 |
+
"name": "model.embed_tokens.q_weight",
|
| 15 |
+
"shape": [
|
| 16 |
+
55296,
|
| 17 |
+
512
|
| 18 |
+
],
|
| 19 |
+
"dtype": "uint32",
|
| 20 |
+
"format": "f32-to-bf16",
|
| 21 |
+
"nbytes": 113246208,
|
| 22 |
+
"byteOffset": 0
|
| 23 |
+
}
|
| 24 |
+
],
|
| 25 |
+
"md5sum": "fe870c24c37d50833f35e5356e23fa21"
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"dataPath": "params_shard_1.bin",
|
| 29 |
+
"format": "raw-shard",
|
| 30 |
+
"nbytes": 25165824,
|
| 31 |
+
"records": [
|
| 32 |
+
{
|
| 33 |
+
"name": "model.layers.0.self_attn.qkv_proj.q_weight",
|
| 34 |
+
"shape": [
|
| 35 |
+
12288,
|
| 36 |
+
512
|
| 37 |
+
],
|
| 38 |
+
"dtype": "uint32",
|
| 39 |
+
"format": "f32-to-bf16",
|
| 40 |
+
"nbytes": 25165824,
|
| 41 |
+
"byteOffset": 0
|
| 42 |
+
}
|
| 43 |
+
],
|
| 44 |
+
"md5sum": "2732fee74ddceeec2a570c77bcf3a75f"
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"dataPath": "params_shard_2.bin",
|
| 48 |
+
"format": "raw-shard",
|
| 49 |
+
"nbytes": 45088768,
|
| 50 |
+
"records": [
|
| 51 |
+
{
|
| 52 |
+
"name": "model.layers.0.mlp.gate_up_proj.q_weight",
|
| 53 |
+
"shape": [
|
| 54 |
+
22016,
|
| 55 |
+
512
|
| 56 |
+
],
|
| 57 |
+
"dtype": "uint32",
|
| 58 |
+
"format": "f32-to-bf16",
|
| 59 |
+
"nbytes": 45088768,
|
| 60 |
+
"byteOffset": 0
|
| 61 |
+
}
|
| 62 |
+
],
|
| 63 |
+
"md5sum": "5583cabee6e1df6bc137f2495e41aacb"
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"dataPath": "params_shard_3.bin",
|
| 67 |
+
"format": "raw-shard",
|
| 68 |
+
"nbytes": 22544384,
|
| 69 |
+
"records": [
|
| 70 |
+
{
|
| 71 |
+
"name": "model.layers.0.mlp.down_proj.q_weight",
|
| 72 |
+
"shape": [
|
| 73 |
+
4096,
|
| 74 |
+
1376
|
| 75 |
+
],
|
| 76 |
+
"dtype": "uint32",
|
| 77 |
+
"format": "f32-to-bf16",
|
| 78 |
+
"nbytes": 22544384,
|
| 79 |
+
"byteOffset": 0
|
| 80 |
+
}
|
| 81 |
+
],
|
| 82 |
+
"md5sum": "cea6f0488cca4a33509e06aeb87e9658"
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"dataPath": "params_shard_4.bin",
|
| 86 |
+
"format": "raw-shard",
|
| 87 |
+
"nbytes": 32374784,
|
| 88 |
+
"records": [
|
| 89 |
+
{
|
| 90 |
+
"name": "model.embed_tokens.q_scale",
|
| 91 |
+
"shape": [
|
| 92 |
+
55296,
|
| 93 |
+
128
|
| 94 |
+
],
|
| 95 |
+
"dtype": "bfloat16",
|
| 96 |
+
"format": "raw",
|
| 97 |
+
"nbytes": 14155776,
|
| 98 |
+
"byteOffset": 0
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"name": "model.layers.0.self_attn.qkv_proj.q_scale",
|
| 102 |
+
"shape": [
|
| 103 |
+
12288,
|
| 104 |
+
128
|
| 105 |
+
],
|
| 106 |
+
"dtype": "bfloat16",
|
| 107 |
+
"format": "raw",
|
| 108 |
+
"nbytes": 3145728,
|
| 109 |
+
"byteOffset": 14155776
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"name": "model.layers.0.self_attn.o_proj.q_weight",
|
| 113 |
+
"shape": [
|
| 114 |
+
4096,
|
| 115 |
+
512
|
| 116 |
+
],
|
| 117 |
+
"dtype": "uint32",
|
| 118 |
+
"format": "f32-to-bf16",
|
| 119 |
+
"nbytes": 8388608,
|
| 120 |
+
"byteOffset": 17301504
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"name": "model.layers.0.self_attn.o_proj.q_scale",
|
| 124 |
+
"shape": [
|
| 125 |
+
4096,
|
| 126 |
+
128
|
| 127 |
+
],
|
| 128 |
+
"dtype": "bfloat16",
|
| 129 |
+
"format": "raw",
|
| 130 |
+
"nbytes": 1048576,
|
| 131 |
+
"byteOffset": 25690112
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"name": "model.layers.0.mlp.gate_up_proj.q_scale",
|
| 135 |
+
"shape": [
|
| 136 |
+
22016,
|
| 137 |
+
128
|
| 138 |
+
],
|
| 139 |
+
"dtype": "bfloat16",
|
| 140 |
+
"format": "raw",
|
| 141 |
+
"nbytes": 5636096,
|
| 142 |
+
"byteOffset": 26738688
|
| 143 |
+
}
|
| 144 |
+
],
|
| 145 |
+
"md5sum": "b6cdce6d80073579dc41fd1cea4b3f2b"
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"dataPath": "params_shard_5.bin",
|
| 149 |
+
"format": "raw-shard",
|
| 150 |
+
"nbytes": 31145984,
|
| 151 |
+
"records": [
|
| 152 |
+
{
|
| 153 |
+
"name": "model.layers.0.mlp.down_proj.q_scale",
|
| 154 |
+
"shape": [
|
| 155 |
+
4096,
|
| 156 |
+
344
|
| 157 |
+
],
|
| 158 |
+
"dtype": "bfloat16",
|
| 159 |
+
"format": "raw",
|
| 160 |
+
"nbytes": 2818048,
|
| 161 |
+
"byteOffset": 0
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"name": "model.layers.0.input_layernorm.weight",
|
| 165 |
+
"shape": [
|
| 166 |
+
4096
|
| 167 |
+
],
|
| 168 |
+
"dtype": "bfloat16",
|
| 169 |
+
"format": "raw",
|
| 170 |
+
"nbytes": 8192,
|
| 171 |
+
"byteOffset": 2818048
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"name": "model.layers.0.post_attention_layernorm.weight",
|
| 175 |
+
"shape": [
|
| 176 |
+
4096
|
| 177 |
+
],
|
| 178 |
+
"dtype": "bfloat16",
|
| 179 |
+
"format": "raw",
|
| 180 |
+
"nbytes": 8192,
|
| 181 |
+
"byteOffset": 2826240
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"name": "model.layers.1.self_attn.qkv_proj.q_weight",
|
| 185 |
+
"shape": [
|
| 186 |
+
12288,
|
| 187 |
+
512
|
| 188 |
+
],
|
| 189 |
+
"dtype": "uint32",
|
| 190 |
+
"format": "f32-to-bf16",
|
| 191 |
+
"nbytes": 25165824,
|
| 192 |
+
"byteOffset": 2834432
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"name": "model.layers.1.self_attn.qkv_proj.q_scale",
|
| 196 |
+
"shape": [
|
| 197 |
+
12288,
|
| 198 |
+
128
|
| 199 |
+
],
|
| 200 |
+
"dtype": "bfloat16",
|
| 201 |
+
"format": "raw",
|
| 202 |
+
"nbytes": 3145728,
|
| 203 |
+
"byteOffset": 28000256
|
| 204 |
+
}
|
| 205 |
+
],
|
| 206 |
+
"md5sum": "24947fc54d4d479312e736080fe47c06"
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"dataPath": "params_shard_6.bin",
|
| 210 |
+
"format": "raw-shard",
|
| 211 |
+
"nbytes": 45088768,
|
| 212 |
+
"records": [
|
| 213 |
+
{
|
| 214 |
+
"name": "model.layers.1.mlp.gate_up_proj.q_weight",
|
| 215 |
+
"shape": [
|
| 216 |
+
22016,
|
| 217 |
+
512
|
| 218 |
+
],
|
| 219 |
+
"dtype": "uint32",
|
| 220 |
+
"format": "f32-to-bf16",
|
| 221 |
+
"nbytes": 45088768,
|
| 222 |
+
"byteOffset": 0
|
| 223 |
+
}
|
| 224 |
+
],
|
| 225 |
+
"md5sum": "f81606af90319436b0c52bc3087a8a75"
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"dataPath": "params_shard_7.bin",
|
| 229 |
+
"format": "raw-shard",
|
| 230 |
+
"nbytes": 22544384,
|
| 231 |
+
"records": [
|
| 232 |
+
{
|
| 233 |
+
"name": "model.layers.1.mlp.down_proj.q_weight",
|
| 234 |
+
"shape": [
|
| 235 |
+
4096,
|
| 236 |
+
1376
|
| 237 |
+
],
|
| 238 |
+
"dtype": "uint32",
|
| 239 |
+
"format": "f32-to-bf16",
|
| 240 |
+
"nbytes": 22544384,
|
| 241 |
+
"byteOffset": 0
|
| 242 |
+
}
|
| 243 |
+
],
|
| 244 |
+
"md5sum": "2f3b104d6012ee96282f1cb68e9a540a"
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"dataPath": "params_shard_8.bin",
|
| 248 |
+
"format": "raw-shard",
|
| 249 |
+
"nbytes": 25165824,
|
| 250 |
+
"records": [
|
| 251 |
+
{
|
| 252 |
+
"name": "model.layers.2.self_attn.qkv_proj.q_weight",
|
| 253 |
+
"shape": [
|
| 254 |
+
12288,
|
| 255 |
+
512
|
| 256 |
+
],
|
| 257 |
+
"dtype": "uint32",
|
| 258 |
+
"format": "f32-to-bf16",
|
| 259 |
+
"nbytes": 25165824,
|
| 260 |
+
"byteOffset": 0
|
| 261 |
+
}
|
| 262 |
+
],
|
| 263 |
+
"md5sum": "92644b7b40d42f6a8f79a2780b4287e0"
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"dataPath": "params_shard_9.bin",
|
| 267 |
+
"format": "raw-shard",
|
| 268 |
+
"nbytes": 45088768,
|
| 269 |
+
"records": [
|
| 270 |
+
{
|
| 271 |
+
"name": "model.layers.2.mlp.gate_up_proj.q_weight",
|
| 272 |
+
"shape": [
|
| 273 |
+
22016,
|
| 274 |
+
512
|
| 275 |
+
],
|
| 276 |
+
"dtype": "uint32",
|
| 277 |
+
"format": "f32-to-bf16",
|
| 278 |
+
"nbytes": 45088768,
|
| 279 |
+
"byteOffset": 0
|
| 280 |
+
}
|
| 281 |
+
],
|
| 282 |
+
"md5sum": "cbc67bc2b9741282cdba7dc02cf94d88"
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"dataPath": "params_shard_10.bin",
|
| 286 |
+
"format": "raw-shard",
|
| 287 |
+
"nbytes": 30490624,
|
| 288 |
+
"records": [
|
| 289 |
+
{
|
| 290 |
+
"name": "model.layers.1.self_attn.o_proj.q_weight",
|
| 291 |
+
"shape": [
|
| 292 |
+
4096,
|
| 293 |
+
512
|
| 294 |
+
],
|
| 295 |
+
"dtype": "uint32",
|
| 296 |
+
"format": "f32-to-bf16",
|
| 297 |
+
"nbytes": 8388608,
|
| 298 |
+
"byteOffset": 0
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"name": "model.layers.1.self_attn.o_proj.q_scale",
|
| 302 |
+
"shape": [
|
| 303 |
+
4096,
|
| 304 |
+
128
|
| 305 |
+
],
|
| 306 |
+
"dtype": "bfloat16",
|
| 307 |
+
"format": "raw",
|
| 308 |
+
"nbytes": 1048576,
|
| 309 |
+
"byteOffset": 8388608
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"name": "model.layers.1.mlp.gate_up_proj.q_scale",
|
| 313 |
+
"shape": [
|
| 314 |
+
22016,
|
| 315 |
+
128
|
| 316 |
+
],
|
| 317 |
+
"dtype": "bfloat16",
|
| 318 |
+
"format": "raw",
|
| 319 |
+
"nbytes": 5636096,
|
| 320 |
+
"byteOffset": 9437184
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"name": "model.layers.1.mlp.down_proj.q_scale",
|
| 324 |
+
"shape": [
|
| 325 |
+
4096,
|
| 326 |
+
344
|
| 327 |
+
],
|
| 328 |
+
"dtype": "bfloat16",
|
| 329 |
+
"format": "raw",
|
| 330 |
+
"nbytes": 2818048,
|
| 331 |
+
"byteOffset": 15073280
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"name": "model.layers.1.input_layernorm.weight",
|
| 335 |
+
"shape": [
|
| 336 |
+
4096
|
| 337 |
+
],
|
| 338 |
+
"dtype": "bfloat16",
|
| 339 |
+
"format": "raw",
|
| 340 |
+
"nbytes": 8192,
|
| 341 |
+
"byteOffset": 17891328
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"name": "model.layers.1.post_attention_layernorm.weight",
|
| 345 |
+
"shape": [
|
| 346 |
+
4096
|
| 347 |
+
],
|
| 348 |
+
"dtype": "bfloat16",
|
| 349 |
+
"format": "raw",
|
| 350 |
+
"nbytes": 8192,
|
| 351 |
+
"byteOffset": 17899520
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"name": "model.layers.2.self_attn.qkv_proj.q_scale",
|
| 355 |
+
"shape": [
|
| 356 |
+
12288,
|
| 357 |
+
128
|
| 358 |
+
],
|
| 359 |
+
"dtype": "bfloat16",
|
| 360 |
+
"format": "raw",
|
| 361 |
+
"nbytes": 3145728,
|
| 362 |
+
"byteOffset": 17907712
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"name": "model.layers.2.self_attn.o_proj.q_weight",
|
| 366 |
+
"shape": [
|
| 367 |
+
4096,
|
| 368 |
+
512
|
| 369 |
+
],
|
| 370 |
+
"dtype": "uint32",
|
| 371 |
+
"format": "f32-to-bf16",
|
| 372 |
+
"nbytes": 8388608,
|
| 373 |
+
"byteOffset": 21053440
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"name": "model.layers.2.self_attn.o_proj.q_scale",
|
| 377 |
+
"shape": [
|
| 378 |
+
4096,
|
| 379 |
+
128
|
| 380 |
+
],
|
| 381 |
+
"dtype": "bfloat16",
|
| 382 |
+
"format": "raw",
|
| 383 |
+
"nbytes": 1048576,
|
| 384 |
+
"byteOffset": 29442048
|
| 385 |
+
}
|
| 386 |
+
],
|
| 387 |
+
"md5sum": "99fa11d45465d2a56e742b32f1f9a1d8"
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"dataPath": "params_shard_11.bin",
|
| 391 |
+
"format": "raw-shard",
|
| 392 |
+
"nbytes": 25165824,
|
| 393 |
+
"records": [
|
| 394 |
+
{
|
| 395 |
+
"name": "model.layers.3.self_attn.qkv_proj.q_weight",
|
| 396 |
+
"shape": [
|
| 397 |
+
12288,
|
| 398 |
+
512
|
| 399 |
+
],
|
| 400 |
+
"dtype": "uint32",
|
| 401 |
+
"format": "f32-to-bf16",
|
| 402 |
+
"nbytes": 25165824,
|
| 403 |
+
"byteOffset": 0
|
| 404 |
+
}
|
| 405 |
+
],
|
| 406 |
+
"md5sum": "e012050a37a0fcec4e06efd599d37d51"
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"dataPath": "params_shard_12.bin",
|
| 410 |
+
"format": "raw-shard",
|
| 411 |
+
"nbytes": 31014912,
|
| 412 |
+
"records": [
|
| 413 |
+
{
|
| 414 |
+
"name": "model.layers.2.mlp.gate_up_proj.q_scale",
|
| 415 |
+
"shape": [
|
| 416 |
+
22016,
|
| 417 |
+
128
|
| 418 |
+
],
|
| 419 |
+
"dtype": "bfloat16",
|
| 420 |
+
"format": "raw",
|
| 421 |
+
"nbytes": 5636096,
|
| 422 |
+
"byteOffset": 0
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"name": "model.layers.2.mlp.down_proj.q_weight",
|
| 426 |
+
"shape": [
|
| 427 |
+
4096,
|
| 428 |
+
1376
|
| 429 |
+
],
|
| 430 |
+
"dtype": "uint32",
|
| 431 |
+
"format": "f32-to-bf16",
|
| 432 |
+
"nbytes": 22544384,
|
| 433 |
+
"byteOffset": 5636096
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"name": "model.layers.2.mlp.down_proj.q_scale",
|
| 437 |
+
"shape": [
|
| 438 |
+
4096,
|
| 439 |
+
344
|
| 440 |
+
],
|
| 441 |
+
"dtype": "bfloat16",
|
| 442 |
+
"format": "raw",
|
| 443 |
+
"nbytes": 2818048,
|
| 444 |
+
"byteOffset": 28180480
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"name": "model.layers.2.input_layernorm.weight",
|
| 448 |
+
"shape": [
|
| 449 |
+
4096
|
| 450 |
+
],
|
| 451 |
+
"dtype": "bfloat16",
|
| 452 |
+
"format": "raw",
|
| 453 |
+
"nbytes": 8192,
|
| 454 |
+
"byteOffset": 30998528
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"name": "model.layers.2.post_attention_layernorm.weight",
|
| 458 |
+
"shape": [
|
| 459 |
+
4096
|
| 460 |
+
],
|
| 461 |
+
"dtype": "bfloat16",
|
| 462 |
+
"format": "raw",
|
| 463 |
+
"nbytes": 8192,
|
| 464 |
+
"byteOffset": 31006720
|
| 465 |
+
}
|
| 466 |
+
],
|
| 467 |
+
"md5sum": "4e0c7558b16811665b346390a73a8911"
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"dataPath": "params_shard_13.bin",
|
| 471 |
+
"format": "raw-shard",
|
| 472 |
+
"nbytes": 45088768,
|
| 473 |
+
"records": [
|
| 474 |
+
{
|
| 475 |
+
"name": "model.layers.3.mlp.gate_up_proj.q_weight",
|
| 476 |
+
"shape": [
|
| 477 |
+
22016,
|
| 478 |
+
512
|
| 479 |
+
],
|
| 480 |
+
"dtype": "uint32",
|
| 481 |
+
"format": "f32-to-bf16",
|
| 482 |
+
"nbytes": 45088768,
|
| 483 |
+
"byteOffset": 0
|
| 484 |
+
}
|
| 485 |
+
],
|
| 486 |
+
"md5sum": "b969fb28468abdae8a625930d384bdec"
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"dataPath": "params_shard_14.bin",
|
| 490 |
+
"format": "raw-shard",
|
| 491 |
+
"nbytes": 22544384,
|
| 492 |
+
"records": [
|
| 493 |
+
{
|
| 494 |
+
"name": "model.layers.3.mlp.down_proj.q_weight",
|
| 495 |
+
"shape": [
|
| 496 |
+
4096,
|
| 497 |
+
1376
|
| 498 |
+
],
|
| 499 |
+
"dtype": "uint32",
|
| 500 |
+
"format": "f32-to-bf16",
|
| 501 |
+
"nbytes": 22544384,
|
| 502 |
+
"byteOffset": 0
|
| 503 |
+
}
|
| 504 |
+
],
|
| 505 |
+
"md5sum": "93ffb5c0be2625c19d8f1411fbb2e187"
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"dataPath": "params_shard_15.bin",
|
| 509 |
+
"format": "raw-shard",
|
| 510 |
+
"nbytes": 113246208,
|
| 511 |
+
"records": [
|
| 512 |
+
{
|
| 513 |
+
"name": "lm_head.q_weight",
|
| 514 |
+
"shape": [
|
| 515 |
+
55296,
|
| 516 |
+
512
|
| 517 |
+
],
|
| 518 |
+
"dtype": "uint32",
|
| 519 |
+
"format": "f32-to-bf16",
|
| 520 |
+
"nbytes": 113246208,
|
| 521 |
+
"byteOffset": 0
|
| 522 |
+
}
|
| 523 |
+
],
|
| 524 |
+
"md5sum": "ad297ed29c5c64ff691e30d12bfaba09"
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"dataPath": "params_shard_16.bin",
|
| 528 |
+
"format": "raw-shard",
|
| 529 |
+
"nbytes": 21061632,
|
| 530 |
+
"records": [
|
| 531 |
+
{
|
| 532 |
+
"name": "model.layers.3.self_attn.qkv_proj.q_scale",
|
| 533 |
+
"shape": [
|
| 534 |
+
12288,
|
| 535 |
+
128
|
| 536 |
+
],
|
| 537 |
+
"dtype": "bfloat16",
|
| 538 |
+
"format": "raw",
|
| 539 |
+
"nbytes": 3145728,
|
| 540 |
+
"byteOffset": 0
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"name": "model.layers.3.self_attn.o_proj.q_weight",
|
| 544 |
+
"shape": [
|
| 545 |
+
4096,
|
| 546 |
+
512
|
| 547 |
+
],
|
| 548 |
+
"dtype": "uint32",
|
| 549 |
+
"format": "f32-to-bf16",
|
| 550 |
+
"nbytes": 8388608,
|
| 551 |
+
"byteOffset": 3145728
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"name": "model.layers.3.self_attn.o_proj.q_scale",
|
| 555 |
+
"shape": [
|
| 556 |
+
4096,
|
| 557 |
+
128
|
| 558 |
+
],
|
| 559 |
+
"dtype": "bfloat16",
|
| 560 |
+
"format": "raw",
|
| 561 |
+
"nbytes": 1048576,
|
| 562 |
+
"byteOffset": 11534336
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"name": "model.layers.3.mlp.gate_up_proj.q_scale",
|
| 566 |
+
"shape": [
|
| 567 |
+
22016,
|
| 568 |
+
128
|
| 569 |
+
],
|
| 570 |
+
"dtype": "bfloat16",
|
| 571 |
+
"format": "raw",
|
| 572 |
+
"nbytes": 5636096,
|
| 573 |
+
"byteOffset": 12582912
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"name": "model.layers.3.mlp.down_proj.q_scale",
|
| 577 |
+
"shape": [
|
| 578 |
+
4096,
|
| 579 |
+
344
|
| 580 |
+
],
|
| 581 |
+
"dtype": "bfloat16",
|
| 582 |
+
"format": "raw",
|
| 583 |
+
"nbytes": 2818048,
|
| 584 |
+
"byteOffset": 18219008
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"name": "model.layers.3.input_layernorm.weight",
|
| 588 |
+
"shape": [
|
| 589 |
+
4096
|
| 590 |
+
],
|
| 591 |
+
"dtype": "bfloat16",
|
| 592 |
+
"format": "raw",
|
| 593 |
+
"nbytes": 8192,
|
| 594 |
+
"byteOffset": 21037056
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"name": "model.layers.3.post_attention_layernorm.weight",
|
| 598 |
+
"shape": [
|
| 599 |
+
4096
|
| 600 |
+
],
|
| 601 |
+
"dtype": "bfloat16",
|
| 602 |
+
"format": "raw",
|
| 603 |
+
"nbytes": 8192,
|
| 604 |
+
"byteOffset": 21045248
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"name": "model.norm.weight",
|
| 608 |
+
"shape": [
|
| 609 |
+
4096
|
| 610 |
+
],
|
| 611 |
+
"dtype": "bfloat16",
|
| 612 |
+
"format": "raw",
|
| 613 |
+
"nbytes": 8192,
|
| 614 |
+
"byteOffset": 21053440
|
| 615 |
+
}
|
| 616 |
+
],
|
| 617 |
+
"md5sum": "3865398255ece47a36de555fa54b952e"
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"dataPath": "params_shard_17.bin",
|
| 621 |
+
"format": "raw-shard",
|
| 622 |
+
"nbytes": 14155776,
|
| 623 |
+
"records": [
|
| 624 |
+
{
|
| 625 |
+
"name": "lm_head.q_scale",
|
| 626 |
+
"shape": [
|
| 627 |
+
55296,
|
| 628 |
+
128
|
| 629 |
+
],
|
| 630 |
+
"dtype": "bfloat16",
|
| 631 |
+
"format": "raw",
|
| 632 |
+
"nbytes": 14155776,
|
| 633 |
+
"byteOffset": 0
|
| 634 |
+
}
|
| 635 |
+
],
|
| 636 |
+
"md5sum": "2b16bda1d0ada369b8f31ed936e081f6"
|
| 637 |
+
}
|
| 638 |
+
]
|
| 639 |
+
}
|
ndarray-cache.json
ADDED
|
@@ -0,0 +1,639 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"ParamSize": 45,
|
| 4 |
+
"ParamBytes": 789200896.0,
|
| 5 |
+
"BitsPerParam": 5.000788364646225
|
| 6 |
+
},
|
| 7 |
+
"records": [
|
| 8 |
+
{
|
| 9 |
+
"dataPath": "params_shard_0.bin",
|
| 10 |
+
"format": "raw-shard",
|
| 11 |
+
"nbytes": 113246208,
|
| 12 |
+
"records": [
|
| 13 |
+
{
|
| 14 |
+
"name": "model.embed_tokens.q_weight",
|
| 15 |
+
"shape": [
|
| 16 |
+
55296,
|
| 17 |
+
512
|
| 18 |
+
],
|
| 19 |
+
"dtype": "uint32",
|
| 20 |
+
"format": "f32-to-bf16",
|
| 21 |
+
"nbytes": 113246208,
|
| 22 |
+
"byteOffset": 0
|
| 23 |
+
}
|
| 24 |
+
],
|
| 25 |
+
"md5sum": "fe870c24c37d50833f35e5356e23fa21"
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"dataPath": "params_shard_1.bin",
|
| 29 |
+
"format": "raw-shard",
|
| 30 |
+
"nbytes": 25165824,
|
| 31 |
+
"records": [
|
| 32 |
+
{
|
| 33 |
+
"name": "model.layers.0.self_attn.qkv_proj.q_weight",
|
| 34 |
+
"shape": [
|
| 35 |
+
12288,
|
| 36 |
+
512
|
| 37 |
+
],
|
| 38 |
+
"dtype": "uint32",
|
| 39 |
+
"format": "f32-to-bf16",
|
| 40 |
+
"nbytes": 25165824,
|
| 41 |
+
"byteOffset": 0
|
| 42 |
+
}
|
| 43 |
+
],
|
| 44 |
+
"md5sum": "2732fee74ddceeec2a570c77bcf3a75f"
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"dataPath": "params_shard_2.bin",
|
| 48 |
+
"format": "raw-shard",
|
| 49 |
+
"nbytes": 45088768,
|
| 50 |
+
"records": [
|
| 51 |
+
{
|
| 52 |
+
"name": "model.layers.0.mlp.gate_up_proj.q_weight",
|
| 53 |
+
"shape": [
|
| 54 |
+
22016,
|
| 55 |
+
512
|
| 56 |
+
],
|
| 57 |
+
"dtype": "uint32",
|
| 58 |
+
"format": "f32-to-bf16",
|
| 59 |
+
"nbytes": 45088768,
|
| 60 |
+
"byteOffset": 0
|
| 61 |
+
}
|
| 62 |
+
],
|
| 63 |
+
"md5sum": "5583cabee6e1df6bc137f2495e41aacb"
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"dataPath": "params_shard_3.bin",
|
| 67 |
+
"format": "raw-shard",
|
| 68 |
+
"nbytes": 22544384,
|
| 69 |
+
"records": [
|
| 70 |
+
{
|
| 71 |
+
"name": "model.layers.0.mlp.down_proj.q_weight",
|
| 72 |
+
"shape": [
|
| 73 |
+
4096,
|
| 74 |
+
1376
|
| 75 |
+
],
|
| 76 |
+
"dtype": "uint32",
|
| 77 |
+
"format": "f32-to-bf16",
|
| 78 |
+
"nbytes": 22544384,
|
| 79 |
+
"byteOffset": 0
|
| 80 |
+
}
|
| 81 |
+
],
|
| 82 |
+
"md5sum": "cea6f0488cca4a33509e06aeb87e9658"
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"dataPath": "params_shard_4.bin",
|
| 86 |
+
"format": "raw-shard",
|
| 87 |
+
"nbytes": 32374784,
|
| 88 |
+
"records": [
|
| 89 |
+
{
|
| 90 |
+
"name": "model.embed_tokens.q_scale",
|
| 91 |
+
"shape": [
|
| 92 |
+
55296,
|
| 93 |
+
128
|
| 94 |
+
],
|
| 95 |
+
"dtype": "float32",
|
| 96 |
+
"format": "f32-to-bf16",
|
| 97 |
+
"nbytes": 14155776,
|
| 98 |
+
"byteOffset": 0
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"name": "model.layers.0.self_attn.qkv_proj.q_scale",
|
| 102 |
+
"shape": [
|
| 103 |
+
12288,
|
| 104 |
+
128
|
| 105 |
+
],
|
| 106 |
+
"dtype": "float32",
|
| 107 |
+
"format": "f32-to-bf16",
|
| 108 |
+
"nbytes": 3145728,
|
| 109 |
+
"byteOffset": 14155776
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"name": "model.layers.0.self_attn.o_proj.q_weight",
|
| 113 |
+
"shape": [
|
| 114 |
+
4096,
|
| 115 |
+
512
|
| 116 |
+
],
|
| 117 |
+
"dtype": "uint32",
|
| 118 |
+
"format": "f32-to-bf16",
|
| 119 |
+
"nbytes": 8388608,
|
| 120 |
+
"byteOffset": 17301504
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"name": "model.layers.0.self_attn.o_proj.q_scale",
|
| 124 |
+
"shape": [
|
| 125 |
+
4096,
|
| 126 |
+
128
|
| 127 |
+
],
|
| 128 |
+
"dtype": "float32",
|
| 129 |
+
"format": "f32-to-bf16",
|
| 130 |
+
"nbytes": 1048576,
|
| 131 |
+
"byteOffset": 25690112
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"name": "model.layers.0.mlp.gate_up_proj.q_scale",
|
| 135 |
+
"shape": [
|
| 136 |
+
22016,
|
| 137 |
+
128
|
| 138 |
+
],
|
| 139 |
+
"dtype": "float32",
|
| 140 |
+
"format": "f32-to-bf16",
|
| 141 |
+
"nbytes": 5636096,
|
| 142 |
+
"byteOffset": 26738688
|
| 143 |
+
}
|
| 144 |
+
],
|
| 145 |
+
"md5sum": "b6cdce6d80073579dc41fd1cea4b3f2b"
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"dataPath": "params_shard_5.bin",
|
| 149 |
+
"format": "raw-shard",
|
| 150 |
+
"nbytes": 31145984,
|
| 151 |
+
"records": [
|
| 152 |
+
{
|
| 153 |
+
"name": "model.layers.0.mlp.down_proj.q_scale",
|
| 154 |
+
"shape": [
|
| 155 |
+
4096,
|
| 156 |
+
344
|
| 157 |
+
],
|
| 158 |
+
"dtype": "float32",
|
| 159 |
+
"format": "f32-to-bf16",
|
| 160 |
+
"nbytes": 2818048,
|
| 161 |
+
"byteOffset": 0
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"name": "model.layers.0.input_layernorm.weight",
|
| 165 |
+
"shape": [
|
| 166 |
+
4096
|
| 167 |
+
],
|
| 168 |
+
"dtype": "float32",
|
| 169 |
+
"format": "f32-to-bf16",
|
| 170 |
+
"nbytes": 8192,
|
| 171 |
+
"byteOffset": 2818048
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"name": "model.layers.0.post_attention_layernorm.weight",
|
| 175 |
+
"shape": [
|
| 176 |
+
4096
|
| 177 |
+
],
|
| 178 |
+
"dtype": "float32",
|
| 179 |
+
"format": "f32-to-bf16",
|
| 180 |
+
"nbytes": 8192,
|
| 181 |
+
"byteOffset": 2826240
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"name": "model.layers.1.self_attn.qkv_proj.q_weight",
|
| 185 |
+
"shape": [
|
| 186 |
+
12288,
|
| 187 |
+
512
|
| 188 |
+
],
|
| 189 |
+
"dtype": "uint32",
|
| 190 |
+
"format": "f32-to-bf16",
|
| 191 |
+
"nbytes": 25165824,
|
| 192 |
+
"byteOffset": 2834432
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"name": "model.layers.1.self_attn.qkv_proj.q_scale",
|
| 196 |
+
"shape": [
|
| 197 |
+
12288,
|
| 198 |
+
128
|
| 199 |
+
],
|
| 200 |
+
"dtype": "float32",
|
| 201 |
+
"format": "f32-to-bf16",
|
| 202 |
+
"nbytes": 3145728,
|
| 203 |
+
"byteOffset": 28000256
|
| 204 |
+
}
|
| 205 |
+
],
|
| 206 |
+
"md5sum": "24947fc54d4d479312e736080fe47c06"
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"dataPath": "params_shard_6.bin",
|
| 210 |
+
"format": "raw-shard",
|
| 211 |
+
"nbytes": 45088768,
|
| 212 |
+
"records": [
|
| 213 |
+
{
|
| 214 |
+
"name": "model.layers.1.mlp.gate_up_proj.q_weight",
|
| 215 |
+
"shape": [
|
| 216 |
+
22016,
|
| 217 |
+
512
|
| 218 |
+
],
|
| 219 |
+
"dtype": "uint32",
|
| 220 |
+
"format": "f32-to-bf16",
|
| 221 |
+
"nbytes": 45088768,
|
| 222 |
+
"byteOffset": 0
|
| 223 |
+
}
|
| 224 |
+
],
|
| 225 |
+
"md5sum": "f81606af90319436b0c52bc3087a8a75"
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"dataPath": "params_shard_7.bin",
|
| 229 |
+
"format": "raw-shard",
|
| 230 |
+
"nbytes": 22544384,
|
| 231 |
+
"records": [
|
| 232 |
+
{
|
| 233 |
+
"name": "model.layers.1.mlp.down_proj.q_weight",
|
| 234 |
+
"shape": [
|
| 235 |
+
4096,
|
| 236 |
+
1376
|
| 237 |
+
],
|
| 238 |
+
"dtype": "uint32",
|
| 239 |
+
"format": "f32-to-bf16",
|
| 240 |
+
"nbytes": 22544384,
|
| 241 |
+
"byteOffset": 0
|
| 242 |
+
}
|
| 243 |
+
],
|
| 244 |
+
"md5sum": "2f3b104d6012ee96282f1cb68e9a540a"
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"dataPath": "params_shard_8.bin",
|
| 248 |
+
"format": "raw-shard",
|
| 249 |
+
"nbytes": 25165824,
|
| 250 |
+
"records": [
|
| 251 |
+
{
|
| 252 |
+
"name": "model.layers.2.self_attn.qkv_proj.q_weight",
|
| 253 |
+
"shape": [
|
| 254 |
+
12288,
|
| 255 |
+
512
|
| 256 |
+
],
|
| 257 |
+
"dtype": "uint32",
|
| 258 |
+
"format": "f32-to-bf16",
|
| 259 |
+
"nbytes": 25165824,
|
| 260 |
+
"byteOffset": 0
|
| 261 |
+
}
|
| 262 |
+
],
|
| 263 |
+
"md5sum": "92644b7b40d42f6a8f79a2780b4287e0"
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"dataPath": "params_shard_9.bin",
|
| 267 |
+
"format": "raw-shard",
|
| 268 |
+
"nbytes": 45088768,
|
| 269 |
+
"records": [
|
| 270 |
+
{
|
| 271 |
+
"name": "model.layers.2.mlp.gate_up_proj.q_weight",
|
| 272 |
+
"shape": [
|
| 273 |
+
22016,
|
| 274 |
+
512
|
| 275 |
+
],
|
| 276 |
+
"dtype": "uint32",
|
| 277 |
+
"format": "f32-to-bf16",
|
| 278 |
+
"nbytes": 45088768,
|
| 279 |
+
"byteOffset": 0
|
| 280 |
+
}
|
| 281 |
+
],
|
| 282 |
+
"md5sum": "cbc67bc2b9741282cdba7dc02cf94d88"
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"dataPath": "params_shard_10.bin",
|
| 286 |
+
"format": "raw-shard",
|
| 287 |
+
"nbytes": 30490624,
|
| 288 |
+
"records": [
|
| 289 |
+
{
|
| 290 |
+
"name": "model.layers.1.self_attn.o_proj.q_weight",
|
| 291 |
+
"shape": [
|
| 292 |
+
4096,
|
| 293 |
+
512
|
| 294 |
+
],
|
| 295 |
+
"dtype": "uint32",
|
| 296 |
+
"format": "f32-to-bf16",
|
| 297 |
+
"nbytes": 8388608,
|
| 298 |
+
"byteOffset": 0
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"name": "model.layers.1.self_attn.o_proj.q_scale",
|
| 302 |
+
"shape": [
|
| 303 |
+
4096,
|
| 304 |
+
128
|
| 305 |
+
],
|
| 306 |
+
"dtype": "float32",
|
| 307 |
+
"format": "f32-to-bf16",
|
| 308 |
+
"nbytes": 1048576,
|
| 309 |
+
"byteOffset": 8388608
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"name": "model.layers.1.mlp.gate_up_proj.q_scale",
|
| 313 |
+
"shape": [
|
| 314 |
+
22016,
|
| 315 |
+
128
|
| 316 |
+
],
|
| 317 |
+
"dtype": "float32",
|
| 318 |
+
"format": "f32-to-bf16",
|
| 319 |
+
"nbytes": 5636096,
|
| 320 |
+
"byteOffset": 9437184
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"name": "model.layers.1.mlp.down_proj.q_scale",
|
| 324 |
+
"shape": [
|
| 325 |
+
4096,
|
| 326 |
+
344
|
| 327 |
+
],
|
| 328 |
+
"dtype": "float32",
|
| 329 |
+
"format": "f32-to-bf16",
|
| 330 |
+
"nbytes": 2818048,
|
| 331 |
+
"byteOffset": 15073280
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"name": "model.layers.1.input_layernorm.weight",
|
| 335 |
+
"shape": [
|
| 336 |
+
4096
|
| 337 |
+
],
|
| 338 |
+
"dtype": "float32",
|
| 339 |
+
"format": "f32-to-bf16",
|
| 340 |
+
"nbytes": 8192,
|
| 341 |
+
"byteOffset": 17891328
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"name": "model.layers.1.post_attention_layernorm.weight",
|
| 345 |
+
"shape": [
|
| 346 |
+
4096
|
| 347 |
+
],
|
| 348 |
+
"dtype": "float32",
|
| 349 |
+
"format": "f32-to-bf16",
|
| 350 |
+
"nbytes": 8192,
|
| 351 |
+
"byteOffset": 17899520
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"name": "model.layers.2.self_attn.qkv_proj.q_scale",
|
| 355 |
+
"shape": [
|
| 356 |
+
12288,
|
| 357 |
+
128
|
| 358 |
+
],
|
| 359 |
+
"dtype": "float32",
|
| 360 |
+
"format": "f32-to-bf16",
|
| 361 |
+
"nbytes": 3145728,
|
| 362 |
+
"byteOffset": 17907712
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"name": "model.layers.2.self_attn.o_proj.q_weight",
|
| 366 |
+
"shape": [
|
| 367 |
+
4096,
|
| 368 |
+
512
|
| 369 |
+
],
|
| 370 |
+
"dtype": "uint32",
|
| 371 |
+
"format": "f32-to-bf16",
|
| 372 |
+
"nbytes": 8388608,
|
| 373 |
+
"byteOffset": 21053440
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"name": "model.layers.2.self_attn.o_proj.q_scale",
|
| 377 |
+
"shape": [
|
| 378 |
+
4096,
|
| 379 |
+
128
|
| 380 |
+
],
|
| 381 |
+
"dtype": "float32",
|
| 382 |
+
"format": "f32-to-bf16",
|
| 383 |
+
"nbytes": 1048576,
|
| 384 |
+
"byteOffset": 29442048
|
| 385 |
+
}
|
| 386 |
+
],
|
| 387 |
+
"md5sum": "99fa11d45465d2a56e742b32f1f9a1d8"
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"dataPath": "params_shard_11.bin",
|
| 391 |
+
"format": "raw-shard",
|
| 392 |
+
"nbytes": 25165824,
|
| 393 |
+
"records": [
|
| 394 |
+
{
|
| 395 |
+
"name": "model.layers.3.self_attn.qkv_proj.q_weight",
|
| 396 |
+
"shape": [
|
| 397 |
+
12288,
|
| 398 |
+
512
|
| 399 |
+
],
|
| 400 |
+
"dtype": "uint32",
|
| 401 |
+
"format": "f32-to-bf16",
|
| 402 |
+
"nbytes": 25165824,
|
| 403 |
+
"byteOffset": 0
|
| 404 |
+
}
|
| 405 |
+
],
|
| 406 |
+
"md5sum": "e012050a37a0fcec4e06efd599d37d51"
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"dataPath": "params_shard_12.bin",
|
| 410 |
+
"format": "raw-shard",
|
| 411 |
+
"nbytes": 31014912,
|
| 412 |
+
"records": [
|
| 413 |
+
{
|
| 414 |
+
"name": "model.layers.2.mlp.gate_up_proj.q_scale",
|
| 415 |
+
"shape": [
|
| 416 |
+
22016,
|
| 417 |
+
128
|
| 418 |
+
],
|
| 419 |
+
"dtype": "float32",
|
| 420 |
+
"format": "f32-to-bf16",
|
| 421 |
+
"nbytes": 5636096,
|
| 422 |
+
"byteOffset": 0
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"name": "model.layers.2.mlp.down_proj.q_weight",
|
| 426 |
+
"shape": [
|
| 427 |
+
4096,
|
| 428 |
+
1376
|
| 429 |
+
],
|
| 430 |
+
"dtype": "uint32",
|
| 431 |
+
"format": "f32-to-bf16",
|
| 432 |
+
"nbytes": 22544384,
|
| 433 |
+
"byteOffset": 5636096
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"name": "model.layers.2.mlp.down_proj.q_scale",
|
| 437 |
+
"shape": [
|
| 438 |
+
4096,
|
| 439 |
+
344
|
| 440 |
+
],
|
| 441 |
+
"dtype": "float32",
|
| 442 |
+
"format": "f32-to-bf16",
|
| 443 |
+
"nbytes": 2818048,
|
| 444 |
+
"byteOffset": 28180480
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"name": "model.layers.2.input_layernorm.weight",
|
| 448 |
+
"shape": [
|
| 449 |
+
4096
|
| 450 |
+
],
|
| 451 |
+
"dtype": "float32",
|
| 452 |
+
"format": "f32-to-bf16",
|
| 453 |
+
"nbytes": 8192,
|
| 454 |
+
"byteOffset": 30998528
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"name": "model.layers.2.post_attention_layernorm.weight",
|
| 458 |
+
"shape": [
|
| 459 |
+
4096
|
| 460 |
+
],
|
| 461 |
+
"dtype": "float32",
|
| 462 |
+
"format": "f32-to-bf16",
|
| 463 |
+
"nbytes": 8192,
|
| 464 |
+
"byteOffset": 31006720
|
| 465 |
+
}
|
| 466 |
+
],
|
| 467 |
+
"md5sum": "4e0c7558b16811665b346390a73a8911"
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"dataPath": "params_shard_13.bin",
|
| 471 |
+
"format": "raw-shard",
|
| 472 |
+
"nbytes": 45088768,
|
| 473 |
+
"records": [
|
| 474 |
+
{
|
| 475 |
+
"name": "model.layers.3.mlp.gate_up_proj.q_weight",
|
| 476 |
+
"shape": [
|
| 477 |
+
22016,
|
| 478 |
+
512
|
| 479 |
+
],
|
| 480 |
+
"dtype": "uint32",
|
| 481 |
+
"format": "f32-to-bf16",
|
| 482 |
+
"nbytes": 45088768,
|
| 483 |
+
"byteOffset": 0
|
| 484 |
+
}
|
| 485 |
+
],
|
| 486 |
+
"md5sum": "b969fb28468abdae8a625930d384bdec"
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"dataPath": "params_shard_14.bin",
|
| 490 |
+
"format": "raw-shard",
|
| 491 |
+
"nbytes": 22544384,
|
| 492 |
+
"records": [
|
| 493 |
+
{
|
| 494 |
+
"name": "model.layers.3.mlp.down_proj.q_weight",
|
| 495 |
+
"shape": [
|
| 496 |
+
4096,
|
| 497 |
+
1376
|
| 498 |
+
],
|
| 499 |
+
"dtype": "uint32",
|
| 500 |
+
"format": "f32-to-bf16",
|
| 501 |
+
"nbytes": 22544384,
|
| 502 |
+
"byteOffset": 0
|
| 503 |
+
}
|
| 504 |
+
],
|
| 505 |
+
"md5sum": "93ffb5c0be2625c19d8f1411fbb2e187"
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"dataPath": "params_shard_15.bin",
|
| 509 |
+
"format": "raw-shard",
|
| 510 |
+
"nbytes": 113246208,
|
| 511 |
+
"records": [
|
| 512 |
+
{
|
| 513 |
+
"name": "lm_head.q_weight",
|
| 514 |
+
"shape": [
|
| 515 |
+
55296,
|
| 516 |
+
512
|
| 517 |
+
],
|
| 518 |
+
"dtype": "uint32",
|
| 519 |
+
"format": "f32-to-bf16",
|
| 520 |
+
"nbytes": 113246208,
|
| 521 |
+
"byteOffset": 0
|
| 522 |
+
}
|
| 523 |
+
],
|
| 524 |
+
"md5sum": "ad297ed29c5c64ff691e30d12bfaba09"
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"dataPath": "params_shard_16.bin",
|
| 528 |
+
"format": "raw-shard",
|
| 529 |
+
"nbytes": 21061632,
|
| 530 |
+
"records": [
|
| 531 |
+
{
|
| 532 |
+
"name": "model.layers.3.self_attn.qkv_proj.q_scale",
|
| 533 |
+
"shape": [
|
| 534 |
+
12288,
|
| 535 |
+
128
|
| 536 |
+
],
|
| 537 |
+
"dtype": "float32",
|
| 538 |
+
"format": "f32-to-bf16",
|
| 539 |
+
"nbytes": 3145728,
|
| 540 |
+
"byteOffset": 0
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"name": "model.layers.3.self_attn.o_proj.q_weight",
|
| 544 |
+
"shape": [
|
| 545 |
+
4096,
|
| 546 |
+
512
|
| 547 |
+
],
|
| 548 |
+
"dtype": "uint32",
|
| 549 |
+
"format": "f32-to-bf16",
|
| 550 |
+
"nbytes": 8388608,
|
| 551 |
+
"byteOffset": 3145728
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"name": "model.layers.3.self_attn.o_proj.q_scale",
|
| 555 |
+
"shape": [
|
| 556 |
+
4096,
|
| 557 |
+
128
|
| 558 |
+
],
|
| 559 |
+
"dtype": "float32",
|
| 560 |
+
"format": "f32-to-bf16",
|
| 561 |
+
"nbytes": 1048576,
|
| 562 |
+
"byteOffset": 11534336
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"name": "model.layers.3.mlp.gate_up_proj.q_scale",
|
| 566 |
+
"shape": [
|
| 567 |
+
22016,
|
| 568 |
+
128
|
| 569 |
+
],
|
| 570 |
+
"dtype": "float32",
|
| 571 |
+
"format": "f32-to-bf16",
|
| 572 |
+
"nbytes": 5636096,
|
| 573 |
+
"byteOffset": 12582912
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"name": "model.layers.3.mlp.down_proj.q_scale",
|
| 577 |
+
"shape": [
|
| 578 |
+
4096,
|
| 579 |
+
344
|
| 580 |
+
],
|
| 581 |
+
"dtype": "float32",
|
| 582 |
+
"format": "f32-to-bf16",
|
| 583 |
+
"nbytes": 2818048,
|
| 584 |
+
"byteOffset": 18219008
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"name": "model.layers.3.input_layernorm.weight",
|
| 588 |
+
"shape": [
|
| 589 |
+
4096
|
| 590 |
+
],
|
| 591 |
+
"dtype": "float32",
|
| 592 |
+
"format": "f32-to-bf16",
|
| 593 |
+
"nbytes": 8192,
|
| 594 |
+
"byteOffset": 21037056
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"name": "model.layers.3.post_attention_layernorm.weight",
|
| 598 |
+
"shape": [
|
| 599 |
+
4096
|
| 600 |
+
],
|
| 601 |
+
"dtype": "float32",
|
| 602 |
+
"format": "f32-to-bf16",
|
| 603 |
+
"nbytes": 8192,
|
| 604 |
+
"byteOffset": 21045248
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"name": "model.norm.weight",
|
| 608 |
+
"shape": [
|
| 609 |
+
4096
|
| 610 |
+
],
|
| 611 |
+
"dtype": "float32",
|
| 612 |
+
"format": "f32-to-bf16",
|
| 613 |
+
"nbytes": 8192,
|
| 614 |
+
"byteOffset": 21053440
|
| 615 |
+
}
|
| 616 |
+
],
|
| 617 |
+
"md5sum": "3865398255ece47a36de555fa54b952e"
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"dataPath": "params_shard_17.bin",
|
| 621 |
+
"format": "raw-shard",
|
| 622 |
+
"nbytes": 14155776,
|
| 623 |
+
"records": [
|
| 624 |
+
{
|
| 625 |
+
"name": "lm_head.q_scale",
|
| 626 |
+
"shape": [
|
| 627 |
+
55296,
|
| 628 |
+
128
|
| 629 |
+
],
|
| 630 |
+
"dtype": "float32",
|
| 631 |
+
"format": "f32-to-bf16",
|
| 632 |
+
"nbytes": 14155776,
|
| 633 |
+
"byteOffset": 0
|
| 634 |
+
}
|
| 635 |
+
],
|
| 636 |
+
"md5sum": "2b16bda1d0ada369b8f31ed936e081f6"
|
| 637 |
+
}
|
| 638 |
+
]
|
| 639 |
+
}
|
params_shard_0.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b924a0640ca78b3b37ea8d83ee06c16f7d4b418f4722f8f87924a6f8ef239d8b
|
| 3 |
+
size 113246208
|
params_shard_1.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba24116ebde38a3d8de7838dcecde6de3f99bb60f73bb244875d5c7e920dcc15
|
| 3 |
+
size 25165824
|
params_shard_10.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b29b19ef21b2dd4f67d879d3ce83f2c090f97aa075447f8294fa523d4b07f5e0
|
| 3 |
+
size 30490624
|
params_shard_11.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d59a750c6837d64bd8f1e40d5af9b15eac719a598d20d6f5ae0120f3da342a18
|
| 3 |
+
size 25165824
|
params_shard_12.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:493f9b539056c6d40957440a9590c96dec1311a3b4bb5b660f6c4e1409abb453
|
| 3 |
+
size 31014912
|
params_shard_13.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:59da04bd8b6c0afa0f9b620a3ee43ab825f0ede87523414ba27bd8b307164add
|
| 3 |
+
size 45088768
|
params_shard_14.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a76faf9b956d98fa99fdbc79e8eb400b4e8b87c7589f6ccef73d8a2310ba71e
|
| 3 |
+
size 22544384
|
params_shard_15.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d59f54e18c377bcd2a1d20e2bbece291ec46f568bc7b7bda212cec60b598eef1
|
| 3 |
+
size 113246208
|
params_shard_16.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb9e5dfa7f88b98e312ac89df5ac83efc7a838ffb092a9e3b92952c354043edf
|
| 3 |
+
size 21061632
|
params_shard_17.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b352ebb41a0e5cbd2fc287d54e71ef67116fd0755e5019d1d5d5591254c03fdb
|
| 3 |
+
size 14155776
|
params_shard_2.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5dc0a54404a0a6cbdabae1fc185fb77d79d2bf78e2fb702289d3e3579be53d97
|
| 3 |
+
size 45088768
|
params_shard_3.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a698caf8a68f48e7c8bbc189dffdcd7628d4f3cb0938e8f08db3bf68304ea921
|
| 3 |
+
size 22544384
|
params_shard_4.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5eca3986a38430361f42baee35d04fd9037324719ba3cd08820fd34f3d02c670
|
| 3 |
+
size 32374784
|
params_shard_5.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41cd8f78595c2519d3ae0c24eae75452c25f85aa7ded823a80515df36f8e4543
|
| 3 |
+
size 31145984
|
params_shard_6.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3de5593138676f78244baabb317692ba9a51478ceea60843d102884ac28f4d77
|
| 3 |
+
size 45088768
|
params_shard_7.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:299537347aea2d23a2f37d283f4ec169739da06d50deff8ea5898d45f926defa
|
| 3 |
+
size 22544384
|
params_shard_8.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aebe29c963aa795121a556295af577766cd0b667c0df2a4e0f2ce705a940a842
|
| 3 |
+
size 25165824
|
params_shard_9.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:289ce81400fe8ef756affe43e44450297e10e5ba3d279511d809b6d7d7eea9e3
|
| 3 |
+
size 45088768
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3b8844863b200dfcca971db228e96ce388290dfcf72c15d7a9d2f604bac787c
|
| 3 |
+
size 844403
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"bos_token": {
|
| 5 |
+
"__type": "AddedToken",
|
| 6 |
+
"content": "<s>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"clean_up_tokenization_spaces": false,
|
| 13 |
+
"eos_token": {
|
| 14 |
+
"__type": "AddedToken",
|
| 15 |
+
"content": "</s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": true,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false
|
| 20 |
+
},
|
| 21 |
+
"legacy": true,
|
| 22 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 23 |
+
"pad_token": null,
|
| 24 |
+
"sp_model_kwargs": {},
|
| 25 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 26 |
+
"unk_token": {
|
| 27 |
+
"__type": "AddedToken",
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": true,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false
|
| 33 |
+
},
|
| 34 |
+
"use_fast": false
|
| 35 |
+
}
|