Add weights and config
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- mlc-chat-config.json +77 -0
- ndarray-cache.json +2395 -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_18.bin +3 -0
- params_shard_19.bin +3 -0
- params_shard_2.bin +3 -0
- params_shard_20.bin +3 -0
- params_shard_21.bin +3 -0
- params_shard_22.bin +3 -0
- params_shard_23.bin +3 -0
- params_shard_24.bin +3 -0
- params_shard_25.bin +3 -0
- params_shard_26.bin +3 -0
- params_shard_27.bin +3 -0
- params_shard_28.bin +3 -0
- params_shard_29.bin +3 -0
- params_shard_3.bin +3 -0
- params_shard_30.bin +3 -0
- params_shard_31.bin +3 -0
- params_shard_32.bin +3 -0
- params_shard_33.bin +3 -0
- params_shard_34.bin +3 -0
- params_shard_35.bin +3 -0
- params_shard_36.bin +3 -0
- params_shard_37.bin +3 -0
- params_shard_38.bin +3 -0
- params_shard_39.bin +3 -0
- params_shard_4.bin +3 -0
- params_shard_40.bin +3 -0
- params_shard_41.bin +3 -0
- params_shard_42.bin +3 -0
- params_shard_43.bin +3 -0
- params_shard_44.bin +3 -0
- params_shard_45.bin +3 -0
- params_shard_46.bin +3 -0
- params_shard_47.bin +3 -0
- params_shard_48.bin +3 -0
- params_shard_49.bin +3 -0
- params_shard_5.bin +3 -0
- params_shard_50.bin +3 -0
- params_shard_51.bin +3 -0
mlc-chat-config.json
ADDED
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"version": "0.1.0",
|
3 |
+
"model_type": "llama",
|
4 |
+
"quantization": "q4f16_1",
|
5 |
+
"model_config": {
|
6 |
+
"hidden_size": 4096,
|
7 |
+
"intermediate_size": 11008,
|
8 |
+
"num_attention_heads": 32,
|
9 |
+
"num_hidden_layers": 17,
|
10 |
+
"rms_norm_eps": 1e-06,
|
11 |
+
"vocab_size": 32000,
|
12 |
+
"position_embedding_base": 10000,
|
13 |
+
"context_window_size": 2048,
|
14 |
+
"prefill_chunk_size": 2048,
|
15 |
+
"num_key_value_heads": 32,
|
16 |
+
"head_dim": 128,
|
17 |
+
"tensor_parallel_shards": 1,
|
18 |
+
"max_batch_size": 80
|
19 |
+
},
|
20 |
+
"vocab_size": 32000,
|
21 |
+
"context_window_size": 2048,
|
22 |
+
"sliding_window_size": -1,
|
23 |
+
"prefill_chunk_size": 2048,
|
24 |
+
"attention_sink_size": -1,
|
25 |
+
"tensor_parallel_shards": 1,
|
26 |
+
"mean_gen_len": 128,
|
27 |
+
"max_gen_len": 512,
|
28 |
+
"shift_fill_factor": 0.3,
|
29 |
+
"temperature": 0.6,
|
30 |
+
"presence_penalty": 0.0,
|
31 |
+
"frequency_penalty": 0.0,
|
32 |
+
"repetition_penalty": 1.0,
|
33 |
+
"top_p": 0.9,
|
34 |
+
"tokenizer_files": [
|
35 |
+
"tokenizer.model",
|
36 |
+
"tokenizer_config.json",
|
37 |
+
"tokenizer.json"
|
38 |
+
],
|
39 |
+
"tokenizer_info": {
|
40 |
+
"token_postproc_method": "byte_fallback",
|
41 |
+
"prepend_space_in_encode": true,
|
42 |
+
"strip_space_in_decode": true
|
43 |
+
},
|
44 |
+
"conv_template": {
|
45 |
+
"name": "st-llm",
|
46 |
+
"system_template": "Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n\n### Instruction:\n{MessagePlaceholders.SYSTEM.value}</s>",
|
47 |
+
"system_message": "You are a helpful, respectful and honest assistant.",
|
48 |
+
"system_prefix_token_ids": null,
|
49 |
+
"add_role_after_system_message": true,
|
50 |
+
"roles": {
|
51 |
+
"user": "### Input:",
|
52 |
+
"assistant": "### Response:"
|
53 |
+
},
|
54 |
+
"role_templates": {
|
55 |
+
"user": "{user_message}",
|
56 |
+
"assistant": "{assistant_message}",
|
57 |
+
"tool": "{tool_message}"
|
58 |
+
},
|
59 |
+
"messages": [],
|
60 |
+
"seps": [
|
61 |
+
"</s>"
|
62 |
+
],
|
63 |
+
"role_content_sep": "\n",
|
64 |
+
"role_empty_sep": "\n",
|
65 |
+
"stop_str": [
|
66 |
+
"</s>"
|
67 |
+
],
|
68 |
+
"stop_token_ids": [
|
69 |
+
2
|
70 |
+
],
|
71 |
+
"function_string": "",
|
72 |
+
"use_function_calling": false
|
73 |
+
},
|
74 |
+
"pad_token_id": 0,
|
75 |
+
"bos_token_id": 1,
|
76 |
+
"eos_token_id": 2
|
77 |
+
}
|
ndarray-cache.json
ADDED
@@ -0,0 +1,2395 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"metadata": {
|
3 |
+
"ParamSize": 175,
|
4 |
+
"ParamBytes": 2082955264.0,
|
5 |
+
"BitsPerParam": 4.50044525764654
|
6 |
+
},
|
7 |
+
"records": [
|
8 |
+
{
|
9 |
+
"dataPath": "params_shard_0.bin",
|
10 |
+
"format": "raw-shard",
|
11 |
+
"nbytes": 65536000,
|
12 |
+
"records": [
|
13 |
+
{
|
14 |
+
"name": "model.embed_tokens.q_weight",
|
15 |
+
"shape": [
|
16 |
+
32000,
|
17 |
+
512
|
18 |
+
],
|
19 |
+
"dtype": "uint32",
|
20 |
+
"format": "f32-to-bf16",
|
21 |
+
"nbytes": 65536000,
|
22 |
+
"byteOffset": 0
|
23 |
+
}
|
24 |
+
],
|
25 |
+
"md5sum": "0c80f07c0e705a2e2c1cd1d241a4b609"
|
26 |
+
},
|
27 |
+
{
|
28 |
+
"dataPath": "params_shard_1.bin",
|
29 |
+
"format": "raw-shard",
|
30 |
+
"nbytes": 33357824,
|
31 |
+
"records": [
|
32 |
+
{
|
33 |
+
"name": "model.embed_tokens.q_scale",
|
34 |
+
"shape": [
|
35 |
+
32000,
|
36 |
+
128
|
37 |
+
],
|
38 |
+
"dtype": "float16",
|
39 |
+
"format": "f32-to-bf16",
|
40 |
+
"nbytes": 8192000,
|
41 |
+
"byteOffset": 0
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"name": "model.layers.0.self_attn.qkv_proj.q_weight",
|
45 |
+
"shape": [
|
46 |
+
12288,
|
47 |
+
512
|
48 |
+
],
|
49 |
+
"dtype": "uint32",
|
50 |
+
"format": "f32-to-bf16",
|
51 |
+
"nbytes": 25165824,
|
52 |
+
"byteOffset": 8192000
|
53 |
+
}
|
54 |
+
],
|
55 |
+
"md5sum": "7cb2cf4d089efc4e8f0398043ee2d587"
|
56 |
+
},
|
57 |
+
{
|
58 |
+
"dataPath": "params_shard_2.bin",
|
59 |
+
"format": "raw-shard",
|
60 |
+
"nbytes": 45088768,
|
61 |
+
"records": [
|
62 |
+
{
|
63 |
+
"name": "model.layers.0.mlp.gate_up_proj.q_weight",
|
64 |
+
"shape": [
|
65 |
+
22016,
|
66 |
+
512
|
67 |
+
],
|
68 |
+
"dtype": "uint32",
|
69 |
+
"format": "f32-to-bf16",
|
70 |
+
"nbytes": 45088768,
|
71 |
+
"byteOffset": 0
|
72 |
+
}
|
73 |
+
],
|
74 |
+
"md5sum": "caa1243424708896db6a17dfe758e06b"
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"dataPath": "params_shard_3.bin",
|
78 |
+
"format": "raw-shard",
|
79 |
+
"nbytes": 22544384,
|
80 |
+
"records": [
|
81 |
+
{
|
82 |
+
"name": "model.layers.0.mlp.down_proj.q_weight",
|
83 |
+
"shape": [
|
84 |
+
4096,
|
85 |
+
1376
|
86 |
+
],
|
87 |
+
"dtype": "uint32",
|
88 |
+
"format": "f32-to-bf16",
|
89 |
+
"nbytes": 22544384,
|
90 |
+
"byteOffset": 0
|
91 |
+
}
|
92 |
+
],
|
93 |
+
"md5sum": "11240ff24505ab642aabdf77448f3d25"
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"dataPath": "params_shard_4.bin",
|
97 |
+
"format": "raw-shard",
|
98 |
+
"nbytes": 25165824,
|
99 |
+
"records": [
|
100 |
+
{
|
101 |
+
"name": "model.layers.1.self_attn.qkv_proj.q_weight",
|
102 |
+
"shape": [
|
103 |
+
12288,
|
104 |
+
512
|
105 |
+
],
|
106 |
+
"dtype": "uint32",
|
107 |
+
"format": "f32-to-bf16",
|
108 |
+
"nbytes": 25165824,
|
109 |
+
"byteOffset": 0
|
110 |
+
}
|
111 |
+
],
|
112 |
+
"md5sum": "c6db109227d5e3f1659a6641f04510cf"
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"dataPath": "params_shard_5.bin",
|
116 |
+
"format": "raw-shard",
|
117 |
+
"nbytes": 32587776,
|
118 |
+
"records": [
|
119 |
+
{
|
120 |
+
"name": "model.layers.0.self_attn.qkv_proj.q_scale",
|
121 |
+
"shape": [
|
122 |
+
12288,
|
123 |
+
128
|
124 |
+
],
|
125 |
+
"dtype": "float16",
|
126 |
+
"format": "f32-to-bf16",
|
127 |
+
"nbytes": 3145728,
|
128 |
+
"byteOffset": 0
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"name": "model.layers.0.self_attn.o_proj.q_weight",
|
132 |
+
"shape": [
|
133 |
+
4096,
|
134 |
+
512
|
135 |
+
],
|
136 |
+
"dtype": "uint32",
|
137 |
+
"format": "f32-to-bf16",
|
138 |
+
"nbytes": 8388608,
|
139 |
+
"byteOffset": 3145728
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"name": "model.layers.0.self_attn.o_proj.q_scale",
|
143 |
+
"shape": [
|
144 |
+
4096,
|
145 |
+
128
|
146 |
+
],
|
147 |
+
"dtype": "float16",
|
148 |
+
"format": "f32-to-bf16",
|
149 |
+
"nbytes": 1048576,
|
150 |
+
"byteOffset": 11534336
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"name": "model.layers.0.mlp.gate_up_proj.q_scale",
|
154 |
+
"shape": [
|
155 |
+
22016,
|
156 |
+
128
|
157 |
+
],
|
158 |
+
"dtype": "float16",
|
159 |
+
"format": "f32-to-bf16",
|
160 |
+
"nbytes": 5636096,
|
161 |
+
"byteOffset": 12582912
|
162 |
+
},
|
163 |
+
{
|
164 |
+
"name": "model.layers.0.mlp.down_proj.q_scale",
|
165 |
+
"shape": [
|
166 |
+
4096,
|
167 |
+
344
|
168 |
+
],
|
169 |
+
"dtype": "float16",
|
170 |
+
"format": "f32-to-bf16",
|
171 |
+
"nbytes": 2818048,
|
172 |
+
"byteOffset": 18219008
|
173 |
+
},
|
174 |
+
{
|
175 |
+
"name": "model.layers.0.input_layernorm.weight",
|
176 |
+
"shape": [
|
177 |
+
4096
|
178 |
+
],
|
179 |
+
"dtype": "float16",
|
180 |
+
"format": "f32-to-bf16",
|
181 |
+
"nbytes": 8192,
|
182 |
+
"byteOffset": 21037056
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"name": "model.layers.0.post_attention_layernorm.weight",
|
186 |
+
"shape": [
|
187 |
+
4096
|
188 |
+
],
|
189 |
+
"dtype": "float16",
|
190 |
+
"format": "f32-to-bf16",
|
191 |
+
"nbytes": 8192,
|
192 |
+
"byteOffset": 21045248
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"name": "model.layers.1.self_attn.qkv_proj.q_scale",
|
196 |
+
"shape": [
|
197 |
+
12288,
|
198 |
+
128
|
199 |
+
],
|
200 |
+
"dtype": "float16",
|
201 |
+
"format": "f32-to-bf16",
|
202 |
+
"nbytes": 3145728,
|
203 |
+
"byteOffset": 21053440
|
204 |
+
},
|
205 |
+
{
|
206 |
+
"name": "model.layers.1.self_attn.o_proj.q_weight",
|
207 |
+
"shape": [
|
208 |
+
4096,
|
209 |
+
512
|
210 |
+
],
|
211 |
+
"dtype": "uint32",
|
212 |
+
"format": "f32-to-bf16",
|
213 |
+
"nbytes": 8388608,
|
214 |
+
"byteOffset": 24199168
|
215 |
+
}
|
216 |
+
],
|
217 |
+
"md5sum": "180a418bc044d5be3e3e3cbd969b7ac8"
|
218 |
+
},
|
219 |
+
{
|
220 |
+
"dataPath": "params_shard_6.bin",
|
221 |
+
"format": "raw-shard",
|
222 |
+
"nbytes": 45088768,
|
223 |
+
"records": [
|
224 |
+
{
|
225 |
+
"name": "model.layers.1.mlp.gate_up_proj.q_weight",
|
226 |
+
"shape": [
|
227 |
+
22016,
|
228 |
+
512
|
229 |
+
],
|
230 |
+
"dtype": "uint32",
|
231 |
+
"format": "f32-to-bf16",
|
232 |
+
"nbytes": 45088768,
|
233 |
+
"byteOffset": 0
|
234 |
+
}
|
235 |
+
],
|
236 |
+
"md5sum": "21d1038acf0a42100d056bd7934e7114"
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"dataPath": "params_shard_7.bin",
|
240 |
+
"format": "raw-shard",
|
241 |
+
"nbytes": 25165824,
|
242 |
+
"records": [
|
243 |
+
{
|
244 |
+
"name": "model.layers.2.self_attn.qkv_proj.q_weight",
|
245 |
+
"shape": [
|
246 |
+
12288,
|
247 |
+
512
|
248 |
+
],
|
249 |
+
"dtype": "uint32",
|
250 |
+
"format": "f32-to-bf16",
|
251 |
+
"nbytes": 25165824,
|
252 |
+
"byteOffset": 0
|
253 |
+
}
|
254 |
+
],
|
255 |
+
"md5sum": "74b85fe2af8fe0bd74a60cdef00dab56"
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"dataPath": "params_shard_8.bin",
|
259 |
+
"format": "raw-shard",
|
260 |
+
"nbytes": 32063488,
|
261 |
+
"records": [
|
262 |
+
{
|
263 |
+
"name": "model.layers.1.self_attn.o_proj.q_scale",
|
264 |
+
"shape": [
|
265 |
+
4096,
|
266 |
+
128
|
267 |
+
],
|
268 |
+
"dtype": "float16",
|
269 |
+
"format": "f32-to-bf16",
|
270 |
+
"nbytes": 1048576,
|
271 |
+
"byteOffset": 0
|
272 |
+
},
|
273 |
+
{
|
274 |
+
"name": "model.layers.1.mlp.gate_up_proj.q_scale",
|
275 |
+
"shape": [
|
276 |
+
22016,
|
277 |
+
128
|
278 |
+
],
|
279 |
+
"dtype": "float16",
|
280 |
+
"format": "f32-to-bf16",
|
281 |
+
"nbytes": 5636096,
|
282 |
+
"byteOffset": 1048576
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"name": "model.layers.1.mlp.down_proj.q_weight",
|
286 |
+
"shape": [
|
287 |
+
4096,
|
288 |
+
1376
|
289 |
+
],
|
290 |
+
"dtype": "uint32",
|
291 |
+
"format": "f32-to-bf16",
|
292 |
+
"nbytes": 22544384,
|
293 |
+
"byteOffset": 6684672
|
294 |
+
},
|
295 |
+
{
|
296 |
+
"name": "model.layers.1.mlp.down_proj.q_scale",
|
297 |
+
"shape": [
|
298 |
+
4096,
|
299 |
+
344
|
300 |
+
],
|
301 |
+
"dtype": "float16",
|
302 |
+
"format": "f32-to-bf16",
|
303 |
+
"nbytes": 2818048,
|
304 |
+
"byteOffset": 29229056
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"name": "model.layers.1.input_layernorm.weight",
|
308 |
+
"shape": [
|
309 |
+
4096
|
310 |
+
],
|
311 |
+
"dtype": "float16",
|
312 |
+
"format": "f32-to-bf16",
|
313 |
+
"nbytes": 8192,
|
314 |
+
"byteOffset": 32047104
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"name": "model.layers.1.post_attention_layernorm.weight",
|
318 |
+
"shape": [
|
319 |
+
4096
|
320 |
+
],
|
321 |
+
"dtype": "float16",
|
322 |
+
"format": "f32-to-bf16",
|
323 |
+
"nbytes": 8192,
|
324 |
+
"byteOffset": 32055296
|
325 |
+
}
|
326 |
+
],
|
327 |
+
"md5sum": "56643a500f2732bdee21c975ff084ae6"
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"dataPath": "params_shard_9.bin",
|
331 |
+
"format": "raw-shard",
|
332 |
+
"nbytes": 45088768,
|
333 |
+
"records": [
|
334 |
+
{
|
335 |
+
"name": "model.layers.2.mlp.gate_up_proj.q_weight",
|
336 |
+
"shape": [
|
337 |
+
22016,
|
338 |
+
512
|
339 |
+
],
|
340 |
+
"dtype": "uint32",
|
341 |
+
"format": "f32-to-bf16",
|
342 |
+
"nbytes": 45088768,
|
343 |
+
"byteOffset": 0
|
344 |
+
}
|
345 |
+
],
|
346 |
+
"md5sum": "f2f2ef6beb194aa2d8a974a4f83c1a1c"
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"dataPath": "params_shard_10.bin",
|
350 |
+
"format": "raw-shard",
|
351 |
+
"nbytes": 22544384,
|
352 |
+
"records": [
|
353 |
+
{
|
354 |
+
"name": "model.layers.2.mlp.down_proj.q_weight",
|
355 |
+
"shape": [
|
356 |
+
4096,
|
357 |
+
1376
|
358 |
+
],
|
359 |
+
"dtype": "uint32",
|
360 |
+
"format": "f32-to-bf16",
|
361 |
+
"nbytes": 22544384,
|
362 |
+
"byteOffset": 0
|
363 |
+
}
|
364 |
+
],
|
365 |
+
"md5sum": "fc02a588f3f4fc8a9d61294c88150097"
|
366 |
+
},
|
367 |
+
{
|
368 |
+
"dataPath": "params_shard_11.bin",
|
369 |
+
"format": "raw-shard",
|
370 |
+
"nbytes": 25165824,
|
371 |
+
"records": [
|
372 |
+
{
|
373 |
+
"name": "model.layers.3.self_attn.qkv_proj.q_weight",
|
374 |
+
"shape": [
|
375 |
+
12288,
|
376 |
+
512
|
377 |
+
],
|
378 |
+
"dtype": "uint32",
|
379 |
+
"format": "f32-to-bf16",
|
380 |
+
"nbytes": 25165824,
|
381 |
+
"byteOffset": 0
|
382 |
+
}
|
383 |
+
],
|
384 |
+
"md5sum": "2f8c920884c554d1c068bd586bbaf716"
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"dataPath": "params_shard_12.bin",
|
388 |
+
"format": "raw-shard",
|
389 |
+
"nbytes": 32587776,
|
390 |
+
"records": [
|
391 |
+
{
|
392 |
+
"name": "model.layers.2.self_attn.qkv_proj.q_scale",
|
393 |
+
"shape": [
|
394 |
+
12288,
|
395 |
+
128
|
396 |
+
],
|
397 |
+
"dtype": "float16",
|
398 |
+
"format": "f32-to-bf16",
|
399 |
+
"nbytes": 3145728,
|
400 |
+
"byteOffset": 0
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"name": "model.layers.2.self_attn.o_proj.q_weight",
|
404 |
+
"shape": [
|
405 |
+
4096,
|
406 |
+
512
|
407 |
+
],
|
408 |
+
"dtype": "uint32",
|
409 |
+
"format": "f32-to-bf16",
|
410 |
+
"nbytes": 8388608,
|
411 |
+
"byteOffset": 3145728
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"name": "model.layers.2.self_attn.o_proj.q_scale",
|
415 |
+
"shape": [
|
416 |
+
4096,
|
417 |
+
128
|
418 |
+
],
|
419 |
+
"dtype": "float16",
|
420 |
+
"format": "f32-to-bf16",
|
421 |
+
"nbytes": 1048576,
|
422 |
+
"byteOffset": 11534336
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"name": "model.layers.2.mlp.gate_up_proj.q_scale",
|
426 |
+
"shape": [
|
427 |
+
22016,
|
428 |
+
128
|
429 |
+
],
|
430 |
+
"dtype": "float16",
|
431 |
+
"format": "f32-to-bf16",
|
432 |
+
"nbytes": 5636096,
|
433 |
+
"byteOffset": 12582912
|
434 |
+
},
|
435 |
+
{
|
436 |
+
"name": "model.layers.2.mlp.down_proj.q_scale",
|
437 |
+
"shape": [
|
438 |
+
4096,
|
439 |
+
344
|
440 |
+
],
|
441 |
+
"dtype": "float16",
|
442 |
+
"format": "f32-to-bf16",
|
443 |
+
"nbytes": 2818048,
|
444 |
+
"byteOffset": 18219008
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"name": "model.layers.2.input_layernorm.weight",
|
448 |
+
"shape": [
|
449 |
+
4096
|
450 |
+
],
|
451 |
+
"dtype": "float16",
|
452 |
+
"format": "f32-to-bf16",
|
453 |
+
"nbytes": 8192,
|
454 |
+
"byteOffset": 21037056
|
455 |
+
},
|
456 |
+
{
|
457 |
+
"name": "model.layers.2.post_attention_layernorm.weight",
|
458 |
+
"shape": [
|
459 |
+
4096
|
460 |
+
],
|
461 |
+
"dtype": "float16",
|
462 |
+
"format": "f32-to-bf16",
|
463 |
+
"nbytes": 8192,
|
464 |
+
"byteOffset": 21045248
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"name": "model.layers.3.self_attn.qkv_proj.q_scale",
|
468 |
+
"shape": [
|
469 |
+
12288,
|
470 |
+
128
|
471 |
+
],
|
472 |
+
"dtype": "float16",
|
473 |
+
"format": "f32-to-bf16",
|
474 |
+
"nbytes": 3145728,
|
475 |
+
"byteOffset": 21053440
|
476 |
+
},
|
477 |
+
{
|
478 |
+
"name": "model.layers.3.self_attn.o_proj.q_weight",
|
479 |
+
"shape": [
|
480 |
+
4096,
|
481 |
+
512
|
482 |
+
],
|
483 |
+
"dtype": "uint32",
|
484 |
+
"format": "f32-to-bf16",
|
485 |
+
"nbytes": 8388608,
|
486 |
+
"byteOffset": 24199168
|
487 |
+
}
|
488 |
+
],
|
489 |
+
"md5sum": "966951e3ee637b3d25f2ccf5c33962ee"
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"dataPath": "params_shard_13.bin",
|
493 |
+
"format": "raw-shard",
|
494 |
+
"nbytes": 45088768,
|
495 |
+
"records": [
|
496 |
+
{
|
497 |
+
"name": "model.layers.3.mlp.gate_up_proj.q_weight",
|
498 |
+
"shape": [
|
499 |
+
22016,
|
500 |
+
512
|
501 |
+
],
|
502 |
+
"dtype": "uint32",
|
503 |
+
"format": "f32-to-bf16",
|
504 |
+
"nbytes": 45088768,
|
505 |
+
"byteOffset": 0
|
506 |
+
}
|
507 |
+
],
|
508 |
+
"md5sum": "feb8d01187dde104bbc36b0015648647"
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"dataPath": "params_shard_14.bin",
|
512 |
+
"format": "raw-shard",
|
513 |
+
"nbytes": 25165824,
|
514 |
+
"records": [
|
515 |
+
{
|
516 |
+
"name": "model.layers.4.self_attn.qkv_proj.q_weight",
|
517 |
+
"shape": [
|
518 |
+
12288,
|
519 |
+
512
|
520 |
+
],
|
521 |
+
"dtype": "uint32",
|
522 |
+
"format": "f32-to-bf16",
|
523 |
+
"nbytes": 25165824,
|
524 |
+
"byteOffset": 0
|
525 |
+
}
|
526 |
+
],
|
527 |
+
"md5sum": "cc381114da3cc1b73f351e768d9c78d0"
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"dataPath": "params_shard_15.bin",
|
531 |
+
"format": "raw-shard",
|
532 |
+
"nbytes": 32063488,
|
533 |
+
"records": [
|
534 |
+
{
|
535 |
+
"name": "model.layers.3.self_attn.o_proj.q_scale",
|
536 |
+
"shape": [
|
537 |
+
4096,
|
538 |
+
128
|
539 |
+
],
|
540 |
+
"dtype": "float16",
|
541 |
+
"format": "f32-to-bf16",
|
542 |
+
"nbytes": 1048576,
|
543 |
+
"byteOffset": 0
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"name": "model.layers.3.mlp.gate_up_proj.q_scale",
|
547 |
+
"shape": [
|
548 |
+
22016,
|
549 |
+
128
|
550 |
+
],
|
551 |
+
"dtype": "float16",
|
552 |
+
"format": "f32-to-bf16",
|
553 |
+
"nbytes": 5636096,
|
554 |
+
"byteOffset": 1048576
|
555 |
+
},
|
556 |
+
{
|
557 |
+
"name": "model.layers.3.mlp.down_proj.q_weight",
|
558 |
+
"shape": [
|
559 |
+
4096,
|
560 |
+
1376
|
561 |
+
],
|
562 |
+
"dtype": "uint32",
|
563 |
+
"format": "f32-to-bf16",
|
564 |
+
"nbytes": 22544384,
|
565 |
+
"byteOffset": 6684672
|
566 |
+
},
|
567 |
+
{
|
568 |
+
"name": "model.layers.3.mlp.down_proj.q_scale",
|
569 |
+
"shape": [
|
570 |
+
4096,
|
571 |
+
344
|
572 |
+
],
|
573 |
+
"dtype": "float16",
|
574 |
+
"format": "f32-to-bf16",
|
575 |
+
"nbytes": 2818048,
|
576 |
+
"byteOffset": 29229056
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"name": "model.layers.3.input_layernorm.weight",
|
580 |
+
"shape": [
|
581 |
+
4096
|
582 |
+
],
|
583 |
+
"dtype": "float16",
|
584 |
+
"format": "f32-to-bf16",
|
585 |
+
"nbytes": 8192,
|
586 |
+
"byteOffset": 32047104
|
587 |
+
},
|
588 |
+
{
|
589 |
+
"name": "model.layers.3.post_attention_layernorm.weight",
|
590 |
+
"shape": [
|
591 |
+
4096
|
592 |
+
],
|
593 |
+
"dtype": "float16",
|
594 |
+
"format": "f32-to-bf16",
|
595 |
+
"nbytes": 8192,
|
596 |
+
"byteOffset": 32055296
|
597 |
+
}
|
598 |
+
],
|
599 |
+
"md5sum": "9c4f4401e504ff51a8670de6b8cd10c9"
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"dataPath": "params_shard_16.bin",
|
603 |
+
"format": "raw-shard",
|
604 |
+
"nbytes": 45088768,
|
605 |
+
"records": [
|
606 |
+
{
|
607 |
+
"name": "model.layers.4.mlp.gate_up_proj.q_weight",
|
608 |
+
"shape": [
|
609 |
+
22016,
|
610 |
+
512
|
611 |
+
],
|
612 |
+
"dtype": "uint32",
|
613 |
+
"format": "f32-to-bf16",
|
614 |
+
"nbytes": 45088768,
|
615 |
+
"byteOffset": 0
|
616 |
+
}
|
617 |
+
],
|
618 |
+
"md5sum": "969a13b8bd14af892f1a47de6f3bdf6f"
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"dataPath": "params_shard_17.bin",
|
622 |
+
"format": "raw-shard",
|
623 |
+
"nbytes": 22544384,
|
624 |
+
"records": [
|
625 |
+
{
|
626 |
+
"name": "model.layers.4.mlp.down_proj.q_weight",
|
627 |
+
"shape": [
|
628 |
+
4096,
|
629 |
+
1376
|
630 |
+
],
|
631 |
+
"dtype": "uint32",
|
632 |
+
"format": "f32-to-bf16",
|
633 |
+
"nbytes": 22544384,
|
634 |
+
"byteOffset": 0
|
635 |
+
}
|
636 |
+
],
|
637 |
+
"md5sum": "bbd77d563bda4d127ad7a44a5234c72e"
|
638 |
+
},
|
639 |
+
{
|
640 |
+
"dataPath": "params_shard_18.bin",
|
641 |
+
"format": "raw-shard",
|
642 |
+
"nbytes": 25165824,
|
643 |
+
"records": [
|
644 |
+
{
|
645 |
+
"name": "model.layers.5.self_attn.qkv_proj.q_weight",
|
646 |
+
"shape": [
|
647 |
+
12288,
|
648 |
+
512
|
649 |
+
],
|
650 |
+
"dtype": "uint32",
|
651 |
+
"format": "f32-to-bf16",
|
652 |
+
"nbytes": 25165824,
|
653 |
+
"byteOffset": 0
|
654 |
+
}
|
655 |
+
],
|
656 |
+
"md5sum": "2f42b0fb6f06cdd44054e42926064765"
|
657 |
+
},
|
658 |
+
{
|
659 |
+
"dataPath": "params_shard_19.bin",
|
660 |
+
"format": "raw-shard",
|
661 |
+
"nbytes": 32587776,
|
662 |
+
"records": [
|
663 |
+
{
|
664 |
+
"name": "model.layers.4.self_attn.qkv_proj.q_scale",
|
665 |
+
"shape": [
|
666 |
+
12288,
|
667 |
+
128
|
668 |
+
],
|
669 |
+
"dtype": "float16",
|
670 |
+
"format": "f32-to-bf16",
|
671 |
+
"nbytes": 3145728,
|
672 |
+
"byteOffset": 0
|
673 |
+
},
|
674 |
+
{
|
675 |
+
"name": "model.layers.4.self_attn.o_proj.q_weight",
|
676 |
+
"shape": [
|
677 |
+
4096,
|
678 |
+
512
|
679 |
+
],
|
680 |
+
"dtype": "uint32",
|
681 |
+
"format": "f32-to-bf16",
|
682 |
+
"nbytes": 8388608,
|
683 |
+
"byteOffset": 3145728
|
684 |
+
},
|
685 |
+
{
|
686 |
+
"name": "model.layers.4.self_attn.o_proj.q_scale",
|
687 |
+
"shape": [
|
688 |
+
4096,
|
689 |
+
128
|
690 |
+
],
|
691 |
+
"dtype": "float16",
|
692 |
+
"format": "f32-to-bf16",
|
693 |
+
"nbytes": 1048576,
|
694 |
+
"byteOffset": 11534336
|
695 |
+
},
|
696 |
+
{
|
697 |
+
"name": "model.layers.4.mlp.gate_up_proj.q_scale",
|
698 |
+
"shape": [
|
699 |
+
22016,
|
700 |
+
128
|
701 |
+
],
|
702 |
+
"dtype": "float16",
|
703 |
+
"format": "f32-to-bf16",
|
704 |
+
"nbytes": 5636096,
|
705 |
+
"byteOffset": 12582912
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"name": "model.layers.4.mlp.down_proj.q_scale",
|
709 |
+
"shape": [
|
710 |
+
4096,
|
711 |
+
344
|
712 |
+
],
|
713 |
+
"dtype": "float16",
|
714 |
+
"format": "f32-to-bf16",
|
715 |
+
"nbytes": 2818048,
|
716 |
+
"byteOffset": 18219008
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"name": "model.layers.4.input_layernorm.weight",
|
720 |
+
"shape": [
|
721 |
+
4096
|
722 |
+
],
|
723 |
+
"dtype": "float16",
|
724 |
+
"format": "f32-to-bf16",
|
725 |
+
"nbytes": 8192,
|
726 |
+
"byteOffset": 21037056
|
727 |
+
},
|
728 |
+
{
|
729 |
+
"name": "model.layers.4.post_attention_layernorm.weight",
|
730 |
+
"shape": [
|
731 |
+
4096
|
732 |
+
],
|
733 |
+
"dtype": "float16",
|
734 |
+
"format": "f32-to-bf16",
|
735 |
+
"nbytes": 8192,
|
736 |
+
"byteOffset": 21045248
|
737 |
+
},
|
738 |
+
{
|
739 |
+
"name": "model.layers.5.self_attn.qkv_proj.q_scale",
|
740 |
+
"shape": [
|
741 |
+
12288,
|
742 |
+
128
|
743 |
+
],
|
744 |
+
"dtype": "float16",
|
745 |
+
"format": "f32-to-bf16",
|
746 |
+
"nbytes": 3145728,
|
747 |
+
"byteOffset": 21053440
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"name": "model.layers.5.self_attn.o_proj.q_weight",
|
751 |
+
"shape": [
|
752 |
+
4096,
|
753 |
+
512
|
754 |
+
],
|
755 |
+
"dtype": "uint32",
|
756 |
+
"format": "f32-to-bf16",
|
757 |
+
"nbytes": 8388608,
|
758 |
+
"byteOffset": 24199168
|
759 |
+
}
|
760 |
+
],
|
761 |
+
"md5sum": "95fee31a0e963a5282a83fbf4903b056"
|
762 |
+
},
|
763 |
+
{
|
764 |
+
"dataPath": "params_shard_20.bin",
|
765 |
+
"format": "raw-shard",
|
766 |
+
"nbytes": 45088768,
|
767 |
+
"records": [
|
768 |
+
{
|
769 |
+
"name": "model.layers.5.mlp.gate_up_proj.q_weight",
|
770 |
+
"shape": [
|
771 |
+
22016,
|
772 |
+
512
|
773 |
+
],
|
774 |
+
"dtype": "uint32",
|
775 |
+
"format": "f32-to-bf16",
|
776 |
+
"nbytes": 45088768,
|
777 |
+
"byteOffset": 0
|
778 |
+
}
|
779 |
+
],
|
780 |
+
"md5sum": "20f7954fd752aa89db92760d3191eddf"
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"dataPath": "params_shard_21.bin",
|
784 |
+
"format": "raw-shard",
|
785 |
+
"nbytes": 25165824,
|
786 |
+
"records": [
|
787 |
+
{
|
788 |
+
"name": "model.layers.6.self_attn.qkv_proj.q_weight",
|
789 |
+
"shape": [
|
790 |
+
12288,
|
791 |
+
512
|
792 |
+
],
|
793 |
+
"dtype": "uint32",
|
794 |
+
"format": "f32-to-bf16",
|
795 |
+
"nbytes": 25165824,
|
796 |
+
"byteOffset": 0
|
797 |
+
}
|
798 |
+
],
|
799 |
+
"md5sum": "05d71f8b2fd3719876737bfba90ee4c0"
|
800 |
+
},
|
801 |
+
{
|
802 |
+
"dataPath": "params_shard_22.bin",
|
803 |
+
"format": "raw-shard",
|
804 |
+
"nbytes": 32063488,
|
805 |
+
"records": [
|
806 |
+
{
|
807 |
+
"name": "model.layers.5.self_attn.o_proj.q_scale",
|
808 |
+
"shape": [
|
809 |
+
4096,
|
810 |
+
128
|
811 |
+
],
|
812 |
+
"dtype": "float16",
|
813 |
+
"format": "f32-to-bf16",
|
814 |
+
"nbytes": 1048576,
|
815 |
+
"byteOffset": 0
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"name": "model.layers.5.mlp.gate_up_proj.q_scale",
|
819 |
+
"shape": [
|
820 |
+
22016,
|
821 |
+
128
|
822 |
+
],
|
823 |
+
"dtype": "float16",
|
824 |
+
"format": "f32-to-bf16",
|
825 |
+
"nbytes": 5636096,
|
826 |
+
"byteOffset": 1048576
|
827 |
+
},
|
828 |
+
{
|
829 |
+
"name": "model.layers.5.mlp.down_proj.q_weight",
|
830 |
+
"shape": [
|
831 |
+
4096,
|
832 |
+
1376
|
833 |
+
],
|
834 |
+
"dtype": "uint32",
|
835 |
+
"format": "f32-to-bf16",
|
836 |
+
"nbytes": 22544384,
|
837 |
+
"byteOffset": 6684672
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"name": "model.layers.5.mlp.down_proj.q_scale",
|
841 |
+
"shape": [
|
842 |
+
4096,
|
843 |
+
344
|
844 |
+
],
|
845 |
+
"dtype": "float16",
|
846 |
+
"format": "f32-to-bf16",
|
847 |
+
"nbytes": 2818048,
|
848 |
+
"byteOffset": 29229056
|
849 |
+
},
|
850 |
+
{
|
851 |
+
"name": "model.layers.5.input_layernorm.weight",
|
852 |
+
"shape": [
|
853 |
+
4096
|
854 |
+
],
|
855 |
+
"dtype": "float16",
|
856 |
+
"format": "f32-to-bf16",
|
857 |
+
"nbytes": 8192,
|
858 |
+
"byteOffset": 32047104
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"name": "model.layers.5.post_attention_layernorm.weight",
|
862 |
+
"shape": [
|
863 |
+
4096
|
864 |
+
],
|
865 |
+
"dtype": "float16",
|
866 |
+
"format": "f32-to-bf16",
|
867 |
+
"nbytes": 8192,
|
868 |
+
"byteOffset": 32055296
|
869 |
+
}
|
870 |
+
],
|
871 |
+
"md5sum": "1f64e37be45d052fe3d264fd700fb53c"
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"dataPath": "params_shard_23.bin",
|
875 |
+
"format": "raw-shard",
|
876 |
+
"nbytes": 45088768,
|
877 |
+
"records": [
|
878 |
+
{
|
879 |
+
"name": "model.layers.6.mlp.gate_up_proj.q_weight",
|
880 |
+
"shape": [
|
881 |
+
22016,
|
882 |
+
512
|
883 |
+
],
|
884 |
+
"dtype": "uint32",
|
885 |
+
"format": "f32-to-bf16",
|
886 |
+
"nbytes": 45088768,
|
887 |
+
"byteOffset": 0
|
888 |
+
}
|
889 |
+
],
|
890 |
+
"md5sum": "2ef9699b390d6e8706188197ec40c8f1"
|
891 |
+
},
|
892 |
+
{
|
893 |
+
"dataPath": "params_shard_24.bin",
|
894 |
+
"format": "raw-shard",
|
895 |
+
"nbytes": 22544384,
|
896 |
+
"records": [
|
897 |
+
{
|
898 |
+
"name": "model.layers.6.mlp.down_proj.q_weight",
|
899 |
+
"shape": [
|
900 |
+
4096,
|
901 |
+
1376
|
902 |
+
],
|
903 |
+
"dtype": "uint32",
|
904 |
+
"format": "f32-to-bf16",
|
905 |
+
"nbytes": 22544384,
|
906 |
+
"byteOffset": 0
|
907 |
+
}
|
908 |
+
],
|
909 |
+
"md5sum": "6cd3ca07a86d12f62ee11ccf2379d86d"
|
910 |
+
},
|
911 |
+
{
|
912 |
+
"dataPath": "params_shard_25.bin",
|
913 |
+
"format": "raw-shard",
|
914 |
+
"nbytes": 25165824,
|
915 |
+
"records": [
|
916 |
+
{
|
917 |
+
"name": "model.layers.7.self_attn.qkv_proj.q_weight",
|
918 |
+
"shape": [
|
919 |
+
12288,
|
920 |
+
512
|
921 |
+
],
|
922 |
+
"dtype": "uint32",
|
923 |
+
"format": "f32-to-bf16",
|
924 |
+
"nbytes": 25165824,
|
925 |
+
"byteOffset": 0
|
926 |
+
}
|
927 |
+
],
|
928 |
+
"md5sum": "ae283ce0ce29814081b952312f0e181c"
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"dataPath": "params_shard_26.bin",
|
932 |
+
"format": "raw-shard",
|
933 |
+
"nbytes": 32587776,
|
934 |
+
"records": [
|
935 |
+
{
|
936 |
+
"name": "model.layers.6.self_attn.qkv_proj.q_scale",
|
937 |
+
"shape": [
|
938 |
+
12288,
|
939 |
+
128
|
940 |
+
],
|
941 |
+
"dtype": "float16",
|
942 |
+
"format": "f32-to-bf16",
|
943 |
+
"nbytes": 3145728,
|
944 |
+
"byteOffset": 0
|
945 |
+
},
|
946 |
+
{
|
947 |
+
"name": "model.layers.6.self_attn.o_proj.q_weight",
|
948 |
+
"shape": [
|
949 |
+
4096,
|
950 |
+
512
|
951 |
+
],
|
952 |
+
"dtype": "uint32",
|
953 |
+
"format": "f32-to-bf16",
|
954 |
+
"nbytes": 8388608,
|
955 |
+
"byteOffset": 3145728
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"name": "model.layers.6.self_attn.o_proj.q_scale",
|
959 |
+
"shape": [
|
960 |
+
4096,
|
961 |
+
128
|
962 |
+
],
|
963 |
+
"dtype": "float16",
|
964 |
+
"format": "f32-to-bf16",
|
965 |
+
"nbytes": 1048576,
|
966 |
+
"byteOffset": 11534336
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"name": "model.layers.6.mlp.gate_up_proj.q_scale",
|
970 |
+
"shape": [
|
971 |
+
22016,
|
972 |
+
128
|
973 |
+
],
|
974 |
+
"dtype": "float16",
|
975 |
+
"format": "f32-to-bf16",
|
976 |
+
"nbytes": 5636096,
|
977 |
+
"byteOffset": 12582912
|
978 |
+
},
|
979 |
+
{
|
980 |
+
"name": "model.layers.6.mlp.down_proj.q_scale",
|
981 |
+
"shape": [
|
982 |
+
4096,
|
983 |
+
344
|
984 |
+
],
|
985 |
+
"dtype": "float16",
|
986 |
+
"format": "f32-to-bf16",
|
987 |
+
"nbytes": 2818048,
|
988 |
+
"byteOffset": 18219008
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"name": "model.layers.6.input_layernorm.weight",
|
992 |
+
"shape": [
|
993 |
+
4096
|
994 |
+
],
|
995 |
+
"dtype": "float16",
|
996 |
+
"format": "f32-to-bf16",
|
997 |
+
"nbytes": 8192,
|
998 |
+
"byteOffset": 21037056
|
999 |
+
},
|
1000 |
+
{
|
1001 |
+
"name": "model.layers.6.post_attention_layernorm.weight",
|
1002 |
+
"shape": [
|
1003 |
+
4096
|
1004 |
+
],
|
1005 |
+
"dtype": "float16",
|
1006 |
+
"format": "f32-to-bf16",
|
1007 |
+
"nbytes": 8192,
|
1008 |
+
"byteOffset": 21045248
|
1009 |
+
},
|
1010 |
+
{
|
1011 |
+
"name": "model.layers.7.self_attn.qkv_proj.q_scale",
|
1012 |
+
"shape": [
|
1013 |
+
12288,
|
1014 |
+
128
|
1015 |
+
],
|
1016 |
+
"dtype": "float16",
|
1017 |
+
"format": "f32-to-bf16",
|
1018 |
+
"nbytes": 3145728,
|
1019 |
+
"byteOffset": 21053440
|
1020 |
+
},
|
1021 |
+
{
|
1022 |
+
"name": "model.layers.7.self_attn.o_proj.q_weight",
|
1023 |
+
"shape": [
|
1024 |
+
4096,
|
1025 |
+
512
|
1026 |
+
],
|
1027 |
+
"dtype": "uint32",
|
1028 |
+
"format": "f32-to-bf16",
|
1029 |
+
"nbytes": 8388608,
|
1030 |
+
"byteOffset": 24199168
|
1031 |
+
}
|
1032 |
+
],
|
1033 |
+
"md5sum": "fcba23e82341dce9cb3ad860da7d2f3f"
|
1034 |
+
},
|
1035 |
+
{
|
1036 |
+
"dataPath": "params_shard_27.bin",
|
1037 |
+
"format": "raw-shard",
|
1038 |
+
"nbytes": 45088768,
|
1039 |
+
"records": [
|
1040 |
+
{
|
1041 |
+
"name": "model.layers.7.mlp.gate_up_proj.q_weight",
|
1042 |
+
"shape": [
|
1043 |
+
22016,
|
1044 |
+
512
|
1045 |
+
],
|
1046 |
+
"dtype": "uint32",
|
1047 |
+
"format": "f32-to-bf16",
|
1048 |
+
"nbytes": 45088768,
|
1049 |
+
"byteOffset": 0
|
1050 |
+
}
|
1051 |
+
],
|
1052 |
+
"md5sum": "699b46d2d1f5da456315aa38db192157"
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"dataPath": "params_shard_28.bin",
|
1056 |
+
"format": "raw-shard",
|
1057 |
+
"nbytes": 25165824,
|
1058 |
+
"records": [
|
1059 |
+
{
|
1060 |
+
"name": "model.layers.8.self_attn.qkv_proj.q_weight",
|
1061 |
+
"shape": [
|
1062 |
+
12288,
|
1063 |
+
512
|
1064 |
+
],
|
1065 |
+
"dtype": "uint32",
|
1066 |
+
"format": "f32-to-bf16",
|
1067 |
+
"nbytes": 25165824,
|
1068 |
+
"byteOffset": 0
|
1069 |
+
}
|
1070 |
+
],
|
1071 |
+
"md5sum": "bd62335afdfb632a45cc2e6c91f1b44b"
|
1072 |
+
},
|
1073 |
+
{
|
1074 |
+
"dataPath": "params_shard_29.bin",
|
1075 |
+
"format": "raw-shard",
|
1076 |
+
"nbytes": 32063488,
|
1077 |
+
"records": [
|
1078 |
+
{
|
1079 |
+
"name": "model.layers.7.self_attn.o_proj.q_scale",
|
1080 |
+
"shape": [
|
1081 |
+
4096,
|
1082 |
+
128
|
1083 |
+
],
|
1084 |
+
"dtype": "float16",
|
1085 |
+
"format": "f32-to-bf16",
|
1086 |
+
"nbytes": 1048576,
|
1087 |
+
"byteOffset": 0
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"name": "model.layers.7.mlp.gate_up_proj.q_scale",
|
1091 |
+
"shape": [
|
1092 |
+
22016,
|
1093 |
+
128
|
1094 |
+
],
|
1095 |
+
"dtype": "float16",
|
1096 |
+
"format": "f32-to-bf16",
|
1097 |
+
"nbytes": 5636096,
|
1098 |
+
"byteOffset": 1048576
|
1099 |
+
},
|
1100 |
+
{
|
1101 |
+
"name": "model.layers.7.mlp.down_proj.q_weight",
|
1102 |
+
"shape": [
|
1103 |
+
4096,
|
1104 |
+
1376
|
1105 |
+
],
|
1106 |
+
"dtype": "uint32",
|
1107 |
+
"format": "f32-to-bf16",
|
1108 |
+
"nbytes": 22544384,
|
1109 |
+
"byteOffset": 6684672
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"name": "model.layers.7.mlp.down_proj.q_scale",
|
1113 |
+
"shape": [
|
1114 |
+
4096,
|
1115 |
+
344
|
1116 |
+
],
|
1117 |
+
"dtype": "float16",
|
1118 |
+
"format": "f32-to-bf16",
|
1119 |
+
"nbytes": 2818048,
|
1120 |
+
"byteOffset": 29229056
|
1121 |
+
},
|
1122 |
+
{
|
1123 |
+
"name": "model.layers.7.input_layernorm.weight",
|
1124 |
+
"shape": [
|
1125 |
+
4096
|
1126 |
+
],
|
1127 |
+
"dtype": "float16",
|
1128 |
+
"format": "f32-to-bf16",
|
1129 |
+
"nbytes": 8192,
|
1130 |
+
"byteOffset": 32047104
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"name": "model.layers.7.post_attention_layernorm.weight",
|
1134 |
+
"shape": [
|
1135 |
+
4096
|
1136 |
+
],
|
1137 |
+
"dtype": "float16",
|
1138 |
+
"format": "f32-to-bf16",
|
1139 |
+
"nbytes": 8192,
|
1140 |
+
"byteOffset": 32055296
|
1141 |
+
}
|
1142 |
+
],
|
1143 |
+
"md5sum": "74c0d61cfa83c3e7082da4f40da88746"
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"dataPath": "params_shard_30.bin",
|
1147 |
+
"format": "raw-shard",
|
1148 |
+
"nbytes": 45088768,
|
1149 |
+
"records": [
|
1150 |
+
{
|
1151 |
+
"name": "model.layers.8.mlp.gate_up_proj.q_weight",
|
1152 |
+
"shape": [
|
1153 |
+
22016,
|
1154 |
+
512
|
1155 |
+
],
|
1156 |
+
"dtype": "uint32",
|
1157 |
+
"format": "f32-to-bf16",
|
1158 |
+
"nbytes": 45088768,
|
1159 |
+
"byteOffset": 0
|
1160 |
+
}
|
1161 |
+
],
|
1162 |
+
"md5sum": "3ef2125e3d6ad0ba41b0d52d4a935cd0"
|
1163 |
+
},
|
1164 |
+
{
|
1165 |
+
"dataPath": "params_shard_31.bin",
|
1166 |
+
"format": "raw-shard",
|
1167 |
+
"nbytes": 22544384,
|
1168 |
+
"records": [
|
1169 |
+
{
|
1170 |
+
"name": "model.layers.8.mlp.down_proj.q_weight",
|
1171 |
+
"shape": [
|
1172 |
+
4096,
|
1173 |
+
1376
|
1174 |
+
],
|
1175 |
+
"dtype": "uint32",
|
1176 |
+
"format": "f32-to-bf16",
|
1177 |
+
"nbytes": 22544384,
|
1178 |
+
"byteOffset": 0
|
1179 |
+
}
|
1180 |
+
],
|
1181 |
+
"md5sum": "2e32a84420247b18d1543a2e2cbb5ca6"
|
1182 |
+
},
|
1183 |
+
{
|
1184 |
+
"dataPath": "params_shard_32.bin",
|
1185 |
+
"format": "raw-shard",
|
1186 |
+
"nbytes": 25165824,
|
1187 |
+
"records": [
|
1188 |
+
{
|
1189 |
+
"name": "model.layers.9.self_attn.qkv_proj.q_weight",
|
1190 |
+
"shape": [
|
1191 |
+
12288,
|
1192 |
+
512
|
1193 |
+
],
|
1194 |
+
"dtype": "uint32",
|
1195 |
+
"format": "f32-to-bf16",
|
1196 |
+
"nbytes": 25165824,
|
1197 |
+
"byteOffset": 0
|
1198 |
+
}
|
1199 |
+
],
|
1200 |
+
"md5sum": "7219fa6ba4178ab2468e66b64eb925a6"
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"dataPath": "params_shard_33.bin",
|
1204 |
+
"format": "raw-shard",
|
1205 |
+
"nbytes": 32587776,
|
1206 |
+
"records": [
|
1207 |
+
{
|
1208 |
+
"name": "model.layers.8.self_attn.qkv_proj.q_scale",
|
1209 |
+
"shape": [
|
1210 |
+
12288,
|
1211 |
+
128
|
1212 |
+
],
|
1213 |
+
"dtype": "float16",
|
1214 |
+
"format": "f32-to-bf16",
|
1215 |
+
"nbytes": 3145728,
|
1216 |
+
"byteOffset": 0
|
1217 |
+
},
|
1218 |
+
{
|
1219 |
+
"name": "model.layers.8.self_attn.o_proj.q_weight",
|
1220 |
+
"shape": [
|
1221 |
+
4096,
|
1222 |
+
512
|
1223 |
+
],
|
1224 |
+
"dtype": "uint32",
|
1225 |
+
"format": "f32-to-bf16",
|
1226 |
+
"nbytes": 8388608,
|
1227 |
+
"byteOffset": 3145728
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"name": "model.layers.8.self_attn.o_proj.q_scale",
|
1231 |
+
"shape": [
|
1232 |
+
4096,
|
1233 |
+
128
|
1234 |
+
],
|
1235 |
+
"dtype": "float16",
|
1236 |
+
"format": "f32-to-bf16",
|
1237 |
+
"nbytes": 1048576,
|
1238 |
+
"byteOffset": 11534336
|
1239 |
+
},
|
1240 |
+
{
|
1241 |
+
"name": "model.layers.8.mlp.gate_up_proj.q_scale",
|
1242 |
+
"shape": [
|
1243 |
+
22016,
|
1244 |
+
128
|
1245 |
+
],
|
1246 |
+
"dtype": "float16",
|
1247 |
+
"format": "f32-to-bf16",
|
1248 |
+
"nbytes": 5636096,
|
1249 |
+
"byteOffset": 12582912
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"name": "model.layers.8.mlp.down_proj.q_scale",
|
1253 |
+
"shape": [
|
1254 |
+
4096,
|
1255 |
+
344
|
1256 |
+
],
|
1257 |
+
"dtype": "float16",
|
1258 |
+
"format": "f32-to-bf16",
|
1259 |
+
"nbytes": 2818048,
|
1260 |
+
"byteOffset": 18219008
|
1261 |
+
},
|
1262 |
+
{
|
1263 |
+
"name": "model.layers.8.input_layernorm.weight",
|
1264 |
+
"shape": [
|
1265 |
+
4096
|
1266 |
+
],
|
1267 |
+
"dtype": "float16",
|
1268 |
+
"format": "f32-to-bf16",
|
1269 |
+
"nbytes": 8192,
|
1270 |
+
"byteOffset": 21037056
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"name": "model.layers.8.post_attention_layernorm.weight",
|
1274 |
+
"shape": [
|
1275 |
+
4096
|
1276 |
+
],
|
1277 |
+
"dtype": "float16",
|
1278 |
+
"format": "f32-to-bf16",
|
1279 |
+
"nbytes": 8192,
|
1280 |
+
"byteOffset": 21045248
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"name": "model.layers.9.self_attn.qkv_proj.q_scale",
|
1284 |
+
"shape": [
|
1285 |
+
12288,
|
1286 |
+
128
|
1287 |
+
],
|
1288 |
+
"dtype": "float16",
|
1289 |
+
"format": "f32-to-bf16",
|
1290 |
+
"nbytes": 3145728,
|
1291 |
+
"byteOffset": 21053440
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"name": "model.layers.9.self_attn.o_proj.q_weight",
|
1295 |
+
"shape": [
|
1296 |
+
4096,
|
1297 |
+
512
|
1298 |
+
],
|
1299 |
+
"dtype": "uint32",
|
1300 |
+
"format": "f32-to-bf16",
|
1301 |
+
"nbytes": 8388608,
|
1302 |
+
"byteOffset": 24199168
|
1303 |
+
}
|
1304 |
+
],
|
1305 |
+
"md5sum": "4ba50eb46a4852cc6eea12850222bef7"
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"dataPath": "params_shard_34.bin",
|
1309 |
+
"format": "raw-shard",
|
1310 |
+
"nbytes": 45088768,
|
1311 |
+
"records": [
|
1312 |
+
{
|
1313 |
+
"name": "model.layers.9.mlp.gate_up_proj.q_weight",
|
1314 |
+
"shape": [
|
1315 |
+
22016,
|
1316 |
+
512
|
1317 |
+
],
|
1318 |
+
"dtype": "uint32",
|
1319 |
+
"format": "f32-to-bf16",
|
1320 |
+
"nbytes": 45088768,
|
1321 |
+
"byteOffset": 0
|
1322 |
+
}
|
1323 |
+
],
|
1324 |
+
"md5sum": "e8a62f76bd551abcd10c3e3e125b7c5b"
|
1325 |
+
},
|
1326 |
+
{
|
1327 |
+
"dataPath": "params_shard_35.bin",
|
1328 |
+
"format": "raw-shard",
|
1329 |
+
"nbytes": 25165824,
|
1330 |
+
"records": [
|
1331 |
+
{
|
1332 |
+
"name": "model.layers.10.self_attn.qkv_proj.q_weight",
|
1333 |
+
"shape": [
|
1334 |
+
12288,
|
1335 |
+
512
|
1336 |
+
],
|
1337 |
+
"dtype": "uint32",
|
1338 |
+
"format": "f32-to-bf16",
|
1339 |
+
"nbytes": 25165824,
|
1340 |
+
"byteOffset": 0
|
1341 |
+
}
|
1342 |
+
],
|
1343 |
+
"md5sum": "2302523927b4273eeb18c56c1976f8cc"
|
1344 |
+
},
|
1345 |
+
{
|
1346 |
+
"dataPath": "params_shard_36.bin",
|
1347 |
+
"format": "raw-shard",
|
1348 |
+
"nbytes": 32063488,
|
1349 |
+
"records": [
|
1350 |
+
{
|
1351 |
+
"name": "model.layers.9.self_attn.o_proj.q_scale",
|
1352 |
+
"shape": [
|
1353 |
+
4096,
|
1354 |
+
128
|
1355 |
+
],
|
1356 |
+
"dtype": "float16",
|
1357 |
+
"format": "f32-to-bf16",
|
1358 |
+
"nbytes": 1048576,
|
1359 |
+
"byteOffset": 0
|
1360 |
+
},
|
1361 |
+
{
|
1362 |
+
"name": "model.layers.9.mlp.gate_up_proj.q_scale",
|
1363 |
+
"shape": [
|
1364 |
+
22016,
|
1365 |
+
128
|
1366 |
+
],
|
1367 |
+
"dtype": "float16",
|
1368 |
+
"format": "f32-to-bf16",
|
1369 |
+
"nbytes": 5636096,
|
1370 |
+
"byteOffset": 1048576
|
1371 |
+
},
|
1372 |
+
{
|
1373 |
+
"name": "model.layers.9.mlp.down_proj.q_weight",
|
1374 |
+
"shape": [
|
1375 |
+
4096,
|
1376 |
+
1376
|
1377 |
+
],
|
1378 |
+
"dtype": "uint32",
|
1379 |
+
"format": "f32-to-bf16",
|
1380 |
+
"nbytes": 22544384,
|
1381 |
+
"byteOffset": 6684672
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"name": "model.layers.9.mlp.down_proj.q_scale",
|
1385 |
+
"shape": [
|
1386 |
+
4096,
|
1387 |
+
344
|
1388 |
+
],
|
1389 |
+
"dtype": "float16",
|
1390 |
+
"format": "f32-to-bf16",
|
1391 |
+
"nbytes": 2818048,
|
1392 |
+
"byteOffset": 29229056
|
1393 |
+
},
|
1394 |
+
{
|
1395 |
+
"name": "model.layers.9.input_layernorm.weight",
|
1396 |
+
"shape": [
|
1397 |
+
4096
|
1398 |
+
],
|
1399 |
+
"dtype": "float16",
|
1400 |
+
"format": "f32-to-bf16",
|
1401 |
+
"nbytes": 8192,
|
1402 |
+
"byteOffset": 32047104
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"name": "model.layers.9.post_attention_layernorm.weight",
|
1406 |
+
"shape": [
|
1407 |
+
4096
|
1408 |
+
],
|
1409 |
+
"dtype": "float16",
|
1410 |
+
"format": "f32-to-bf16",
|
1411 |
+
"nbytes": 8192,
|
1412 |
+
"byteOffset": 32055296
|
1413 |
+
}
|
1414 |
+
],
|
1415 |
+
"md5sum": "873054a18b28147ccbc0bb8ec58c2201"
|
1416 |
+
},
|
1417 |
+
{
|
1418 |
+
"dataPath": "params_shard_37.bin",
|
1419 |
+
"format": "raw-shard",
|
1420 |
+
"nbytes": 45088768,
|
1421 |
+
"records": [
|
1422 |
+
{
|
1423 |
+
"name": "model.layers.10.mlp.gate_up_proj.q_weight",
|
1424 |
+
"shape": [
|
1425 |
+
22016,
|
1426 |
+
512
|
1427 |
+
],
|
1428 |
+
"dtype": "uint32",
|
1429 |
+
"format": "f32-to-bf16",
|
1430 |
+
"nbytes": 45088768,
|
1431 |
+
"byteOffset": 0
|
1432 |
+
}
|
1433 |
+
],
|
1434 |
+
"md5sum": "4e7af0f5b05a0f9c399088231cd684ee"
|
1435 |
+
},
|
1436 |
+
{
|
1437 |
+
"dataPath": "params_shard_38.bin",
|
1438 |
+
"format": "raw-shard",
|
1439 |
+
"nbytes": 22544384,
|
1440 |
+
"records": [
|
1441 |
+
{
|
1442 |
+
"name": "model.layers.10.mlp.down_proj.q_weight",
|
1443 |
+
"shape": [
|
1444 |
+
4096,
|
1445 |
+
1376
|
1446 |
+
],
|
1447 |
+
"dtype": "uint32",
|
1448 |
+
"format": "f32-to-bf16",
|
1449 |
+
"nbytes": 22544384,
|
1450 |
+
"byteOffset": 0
|
1451 |
+
}
|
1452 |
+
],
|
1453 |
+
"md5sum": "5993975a2b6771cbcda9d79ee7f31245"
|
1454 |
+
},
|
1455 |
+
{
|
1456 |
+
"dataPath": "params_shard_39.bin",
|
1457 |
+
"format": "raw-shard",
|
1458 |
+
"nbytes": 25165824,
|
1459 |
+
"records": [
|
1460 |
+
{
|
1461 |
+
"name": "model.layers.11.self_attn.qkv_proj.q_weight",
|
1462 |
+
"shape": [
|
1463 |
+
12288,
|
1464 |
+
512
|
1465 |
+
],
|
1466 |
+
"dtype": "uint32",
|
1467 |
+
"format": "f32-to-bf16",
|
1468 |
+
"nbytes": 25165824,
|
1469 |
+
"byteOffset": 0
|
1470 |
+
}
|
1471 |
+
],
|
1472 |
+
"md5sum": "1951fdb99a782b3dbc318d4f15bd63fa"
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"dataPath": "params_shard_40.bin",
|
1476 |
+
"format": "raw-shard",
|
1477 |
+
"nbytes": 32587776,
|
1478 |
+
"records": [
|
1479 |
+
{
|
1480 |
+
"name": "model.layers.10.self_attn.qkv_proj.q_scale",
|
1481 |
+
"shape": [
|
1482 |
+
12288,
|
1483 |
+
128
|
1484 |
+
],
|
1485 |
+
"dtype": "float16",
|
1486 |
+
"format": "f32-to-bf16",
|
1487 |
+
"nbytes": 3145728,
|
1488 |
+
"byteOffset": 0
|
1489 |
+
},
|
1490 |
+
{
|
1491 |
+
"name": "model.layers.10.self_attn.o_proj.q_weight",
|
1492 |
+
"shape": [
|
1493 |
+
4096,
|
1494 |
+
512
|
1495 |
+
],
|
1496 |
+
"dtype": "uint32",
|
1497 |
+
"format": "f32-to-bf16",
|
1498 |
+
"nbytes": 8388608,
|
1499 |
+
"byteOffset": 3145728
|
1500 |
+
},
|
1501 |
+
{
|
1502 |
+
"name": "model.layers.10.self_attn.o_proj.q_scale",
|
1503 |
+
"shape": [
|
1504 |
+
4096,
|
1505 |
+
128
|
1506 |
+
],
|
1507 |
+
"dtype": "float16",
|
1508 |
+
"format": "f32-to-bf16",
|
1509 |
+
"nbytes": 1048576,
|
1510 |
+
"byteOffset": 11534336
|
1511 |
+
},
|
1512 |
+
{
|
1513 |
+
"name": "model.layers.10.mlp.gate_up_proj.q_scale",
|
1514 |
+
"shape": [
|
1515 |
+
22016,
|
1516 |
+
128
|
1517 |
+
],
|
1518 |
+
"dtype": "float16",
|
1519 |
+
"format": "f32-to-bf16",
|
1520 |
+
"nbytes": 5636096,
|
1521 |
+
"byteOffset": 12582912
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"name": "model.layers.10.mlp.down_proj.q_scale",
|
1525 |
+
"shape": [
|
1526 |
+
4096,
|
1527 |
+
344
|
1528 |
+
],
|
1529 |
+
"dtype": "float16",
|
1530 |
+
"format": "f32-to-bf16",
|
1531 |
+
"nbytes": 2818048,
|
1532 |
+
"byteOffset": 18219008
|
1533 |
+
},
|
1534 |
+
{
|
1535 |
+
"name": "model.layers.10.input_layernorm.weight",
|
1536 |
+
"shape": [
|
1537 |
+
4096
|
1538 |
+
],
|
1539 |
+
"dtype": "float16",
|
1540 |
+
"format": "f32-to-bf16",
|
1541 |
+
"nbytes": 8192,
|
1542 |
+
"byteOffset": 21037056
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"name": "model.layers.10.post_attention_layernorm.weight",
|
1546 |
+
"shape": [
|
1547 |
+
4096
|
1548 |
+
],
|
1549 |
+
"dtype": "float16",
|
1550 |
+
"format": "f32-to-bf16",
|
1551 |
+
"nbytes": 8192,
|
1552 |
+
"byteOffset": 21045248
|
1553 |
+
},
|
1554 |
+
{
|
1555 |
+
"name": "model.layers.11.self_attn.qkv_proj.q_scale",
|
1556 |
+
"shape": [
|
1557 |
+
12288,
|
1558 |
+
128
|
1559 |
+
],
|
1560 |
+
"dtype": "float16",
|
1561 |
+
"format": "f32-to-bf16",
|
1562 |
+
"nbytes": 3145728,
|
1563 |
+
"byteOffset": 21053440
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"name": "model.layers.11.self_attn.o_proj.q_weight",
|
1567 |
+
"shape": [
|
1568 |
+
4096,
|
1569 |
+
512
|
1570 |
+
],
|
1571 |
+
"dtype": "uint32",
|
1572 |
+
"format": "f32-to-bf16",
|
1573 |
+
"nbytes": 8388608,
|
1574 |
+
"byteOffset": 24199168
|
1575 |
+
}
|
1576 |
+
],
|
1577 |
+
"md5sum": "fd6c4c9bbb84ef263171a5614daa76cc"
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"dataPath": "params_shard_41.bin",
|
1581 |
+
"format": "raw-shard",
|
1582 |
+
"nbytes": 45088768,
|
1583 |
+
"records": [
|
1584 |
+
{
|
1585 |
+
"name": "model.layers.11.mlp.gate_up_proj.q_weight",
|
1586 |
+
"shape": [
|
1587 |
+
22016,
|
1588 |
+
512
|
1589 |
+
],
|
1590 |
+
"dtype": "uint32",
|
1591 |
+
"format": "f32-to-bf16",
|
1592 |
+
"nbytes": 45088768,
|
1593 |
+
"byteOffset": 0
|
1594 |
+
}
|
1595 |
+
],
|
1596 |
+
"md5sum": "2cd319ee56313a3ea29c7bfc4dfa2ad3"
|
1597 |
+
},
|
1598 |
+
{
|
1599 |
+
"dataPath": "params_shard_42.bin",
|
1600 |
+
"format": "raw-shard",
|
1601 |
+
"nbytes": 25165824,
|
1602 |
+
"records": [
|
1603 |
+
{
|
1604 |
+
"name": "model.layers.12.self_attn.qkv_proj.q_weight",
|
1605 |
+
"shape": [
|
1606 |
+
12288,
|
1607 |
+
512
|
1608 |
+
],
|
1609 |
+
"dtype": "uint32",
|
1610 |
+
"format": "f32-to-bf16",
|
1611 |
+
"nbytes": 25165824,
|
1612 |
+
"byteOffset": 0
|
1613 |
+
}
|
1614 |
+
],
|
1615 |
+
"md5sum": "9fdb9ea9168460346d1deafa5115a282"
|
1616 |
+
},
|
1617 |
+
{
|
1618 |
+
"dataPath": "params_shard_43.bin",
|
1619 |
+
"format": "raw-shard",
|
1620 |
+
"nbytes": 32063488,
|
1621 |
+
"records": [
|
1622 |
+
{
|
1623 |
+
"name": "model.layers.11.self_attn.o_proj.q_scale",
|
1624 |
+
"shape": [
|
1625 |
+
4096,
|
1626 |
+
128
|
1627 |
+
],
|
1628 |
+
"dtype": "float16",
|
1629 |
+
"format": "f32-to-bf16",
|
1630 |
+
"nbytes": 1048576,
|
1631 |
+
"byteOffset": 0
|
1632 |
+
},
|
1633 |
+
{
|
1634 |
+
"name": "model.layers.11.mlp.gate_up_proj.q_scale",
|
1635 |
+
"shape": [
|
1636 |
+
22016,
|
1637 |
+
128
|
1638 |
+
],
|
1639 |
+
"dtype": "float16",
|
1640 |
+
"format": "f32-to-bf16",
|
1641 |
+
"nbytes": 5636096,
|
1642 |
+
"byteOffset": 1048576
|
1643 |
+
},
|
1644 |
+
{
|
1645 |
+
"name": "model.layers.11.mlp.down_proj.q_weight",
|
1646 |
+
"shape": [
|
1647 |
+
4096,
|
1648 |
+
1376
|
1649 |
+
],
|
1650 |
+
"dtype": "uint32",
|
1651 |
+
"format": "f32-to-bf16",
|
1652 |
+
"nbytes": 22544384,
|
1653 |
+
"byteOffset": 6684672
|
1654 |
+
},
|
1655 |
+
{
|
1656 |
+
"name": "model.layers.11.mlp.down_proj.q_scale",
|
1657 |
+
"shape": [
|
1658 |
+
4096,
|
1659 |
+
344
|
1660 |
+
],
|
1661 |
+
"dtype": "float16",
|
1662 |
+
"format": "f32-to-bf16",
|
1663 |
+
"nbytes": 2818048,
|
1664 |
+
"byteOffset": 29229056
|
1665 |
+
},
|
1666 |
+
{
|
1667 |
+
"name": "model.layers.11.input_layernorm.weight",
|
1668 |
+
"shape": [
|
1669 |
+
4096
|
1670 |
+
],
|
1671 |
+
"dtype": "float16",
|
1672 |
+
"format": "f32-to-bf16",
|
1673 |
+
"nbytes": 8192,
|
1674 |
+
"byteOffset": 32047104
|
1675 |
+
},
|
1676 |
+
{
|
1677 |
+
"name": "model.layers.11.post_attention_layernorm.weight",
|
1678 |
+
"shape": [
|
1679 |
+
4096
|
1680 |
+
],
|
1681 |
+
"dtype": "float16",
|
1682 |
+
"format": "f32-to-bf16",
|
1683 |
+
"nbytes": 8192,
|
1684 |
+
"byteOffset": 32055296
|
1685 |
+
}
|
1686 |
+
],
|
1687 |
+
"md5sum": "4af071255a901797d8e8803f2771596d"
|
1688 |
+
},
|
1689 |
+
{
|
1690 |
+
"dataPath": "params_shard_44.bin",
|
1691 |
+
"format": "raw-shard",
|
1692 |
+
"nbytes": 45088768,
|
1693 |
+
"records": [
|
1694 |
+
{
|
1695 |
+
"name": "model.layers.12.mlp.gate_up_proj.q_weight",
|
1696 |
+
"shape": [
|
1697 |
+
22016,
|
1698 |
+
512
|
1699 |
+
],
|
1700 |
+
"dtype": "uint32",
|
1701 |
+
"format": "f32-to-bf16",
|
1702 |
+
"nbytes": 45088768,
|
1703 |
+
"byteOffset": 0
|
1704 |
+
}
|
1705 |
+
],
|
1706 |
+
"md5sum": "1e23819b13809c063c314faee8f945ac"
|
1707 |
+
},
|
1708 |
+
{
|
1709 |
+
"dataPath": "params_shard_45.bin",
|
1710 |
+
"format": "raw-shard",
|
1711 |
+
"nbytes": 22544384,
|
1712 |
+
"records": [
|
1713 |
+
{
|
1714 |
+
"name": "model.layers.12.mlp.down_proj.q_weight",
|
1715 |
+
"shape": [
|
1716 |
+
4096,
|
1717 |
+
1376
|
1718 |
+
],
|
1719 |
+
"dtype": "uint32",
|
1720 |
+
"format": "f32-to-bf16",
|
1721 |
+
"nbytes": 22544384,
|
1722 |
+
"byteOffset": 0
|
1723 |
+
}
|
1724 |
+
],
|
1725 |
+
"md5sum": "6807ab8932573414e3fd161fe9ab91c7"
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"dataPath": "params_shard_46.bin",
|
1729 |
+
"format": "raw-shard",
|
1730 |
+
"nbytes": 25165824,
|
1731 |
+
"records": [
|
1732 |
+
{
|
1733 |
+
"name": "model.layers.13.self_attn.qkv_proj.q_weight",
|
1734 |
+
"shape": [
|
1735 |
+
12288,
|
1736 |
+
512
|
1737 |
+
],
|
1738 |
+
"dtype": "uint32",
|
1739 |
+
"format": "f32-to-bf16",
|
1740 |
+
"nbytes": 25165824,
|
1741 |
+
"byteOffset": 0
|
1742 |
+
}
|
1743 |
+
],
|
1744 |
+
"md5sum": "31ef0402d240b00fc38c442fb07d3c9c"
|
1745 |
+
},
|
1746 |
+
{
|
1747 |
+
"dataPath": "params_shard_47.bin",
|
1748 |
+
"format": "raw-shard",
|
1749 |
+
"nbytes": 32587776,
|
1750 |
+
"records": [
|
1751 |
+
{
|
1752 |
+
"name": "model.layers.12.self_attn.qkv_proj.q_scale",
|
1753 |
+
"shape": [
|
1754 |
+
12288,
|
1755 |
+
128
|
1756 |
+
],
|
1757 |
+
"dtype": "float16",
|
1758 |
+
"format": "f32-to-bf16",
|
1759 |
+
"nbytes": 3145728,
|
1760 |
+
"byteOffset": 0
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"name": "model.layers.12.self_attn.o_proj.q_weight",
|
1764 |
+
"shape": [
|
1765 |
+
4096,
|
1766 |
+
512
|
1767 |
+
],
|
1768 |
+
"dtype": "uint32",
|
1769 |
+
"format": "f32-to-bf16",
|
1770 |
+
"nbytes": 8388608,
|
1771 |
+
"byteOffset": 3145728
|
1772 |
+
},
|
1773 |
+
{
|
1774 |
+
"name": "model.layers.12.self_attn.o_proj.q_scale",
|
1775 |
+
"shape": [
|
1776 |
+
4096,
|
1777 |
+
128
|
1778 |
+
],
|
1779 |
+
"dtype": "float16",
|
1780 |
+
"format": "f32-to-bf16",
|
1781 |
+
"nbytes": 1048576,
|
1782 |
+
"byteOffset": 11534336
|
1783 |
+
},
|
1784 |
+
{
|
1785 |
+
"name": "model.layers.12.mlp.gate_up_proj.q_scale",
|
1786 |
+
"shape": [
|
1787 |
+
22016,
|
1788 |
+
128
|
1789 |
+
],
|
1790 |
+
"dtype": "float16",
|
1791 |
+
"format": "f32-to-bf16",
|
1792 |
+
"nbytes": 5636096,
|
1793 |
+
"byteOffset": 12582912
|
1794 |
+
},
|
1795 |
+
{
|
1796 |
+
"name": "model.layers.12.mlp.down_proj.q_scale",
|
1797 |
+
"shape": [
|
1798 |
+
4096,
|
1799 |
+
344
|
1800 |
+
],
|
1801 |
+
"dtype": "float16",
|
1802 |
+
"format": "f32-to-bf16",
|
1803 |
+
"nbytes": 2818048,
|
1804 |
+
"byteOffset": 18219008
|
1805 |
+
},
|
1806 |
+
{
|
1807 |
+
"name": "model.layers.12.input_layernorm.weight",
|
1808 |
+
"shape": [
|
1809 |
+
4096
|
1810 |
+
],
|
1811 |
+
"dtype": "float16",
|
1812 |
+
"format": "f32-to-bf16",
|
1813 |
+
"nbytes": 8192,
|
1814 |
+
"byteOffset": 21037056
|
1815 |
+
},
|
1816 |
+
{
|
1817 |
+
"name": "model.layers.12.post_attention_layernorm.weight",
|
1818 |
+
"shape": [
|
1819 |
+
4096
|
1820 |
+
],
|
1821 |
+
"dtype": "float16",
|
1822 |
+
"format": "f32-to-bf16",
|
1823 |
+
"nbytes": 8192,
|
1824 |
+
"byteOffset": 21045248
|
1825 |
+
},
|
1826 |
+
{
|
1827 |
+
"name": "model.layers.13.self_attn.qkv_proj.q_scale",
|
1828 |
+
"shape": [
|
1829 |
+
12288,
|
1830 |
+
128
|
1831 |
+
],
|
1832 |
+
"dtype": "float16",
|
1833 |
+
"format": "f32-to-bf16",
|
1834 |
+
"nbytes": 3145728,
|
1835 |
+
"byteOffset": 21053440
|
1836 |
+
},
|
1837 |
+
{
|
1838 |
+
"name": "model.layers.13.self_attn.o_proj.q_weight",
|
1839 |
+
"shape": [
|
1840 |
+
4096,
|
1841 |
+
512
|
1842 |
+
],
|
1843 |
+
"dtype": "uint32",
|
1844 |
+
"format": "f32-to-bf16",
|
1845 |
+
"nbytes": 8388608,
|
1846 |
+
"byteOffset": 24199168
|
1847 |
+
}
|
1848 |
+
],
|
1849 |
+
"md5sum": "c0eef17b98c90131d5b0628775bfbffa"
|
1850 |
+
},
|
1851 |
+
{
|
1852 |
+
"dataPath": "params_shard_48.bin",
|
1853 |
+
"format": "raw-shard",
|
1854 |
+
"nbytes": 45088768,
|
1855 |
+
"records": [
|
1856 |
+
{
|
1857 |
+
"name": "model.layers.13.mlp.gate_up_proj.q_weight",
|
1858 |
+
"shape": [
|
1859 |
+
22016,
|
1860 |
+
512
|
1861 |
+
],
|
1862 |
+
"dtype": "uint32",
|
1863 |
+
"format": "f32-to-bf16",
|
1864 |
+
"nbytes": 45088768,
|
1865 |
+
"byteOffset": 0
|
1866 |
+
}
|
1867 |
+
],
|
1868 |
+
"md5sum": "7233db81dba12add6ddd1ea9dfe9c3dd"
|
1869 |
+
},
|
1870 |
+
{
|
1871 |
+
"dataPath": "params_shard_49.bin",
|
1872 |
+
"format": "raw-shard",
|
1873 |
+
"nbytes": 25165824,
|
1874 |
+
"records": [
|
1875 |
+
{
|
1876 |
+
"name": "model.layers.14.self_attn.qkv_proj.q_weight",
|
1877 |
+
"shape": [
|
1878 |
+
12288,
|
1879 |
+
512
|
1880 |
+
],
|
1881 |
+
"dtype": "uint32",
|
1882 |
+
"format": "f32-to-bf16",
|
1883 |
+
"nbytes": 25165824,
|
1884 |
+
"byteOffset": 0
|
1885 |
+
}
|
1886 |
+
],
|
1887 |
+
"md5sum": "02a22884484fa383e5cafd9c7dfea8f5"
|
1888 |
+
},
|
1889 |
+
{
|
1890 |
+
"dataPath": "params_shard_50.bin",
|
1891 |
+
"format": "raw-shard",
|
1892 |
+
"nbytes": 32063488,
|
1893 |
+
"records": [
|
1894 |
+
{
|
1895 |
+
"name": "model.layers.13.self_attn.o_proj.q_scale",
|
1896 |
+
"shape": [
|
1897 |
+
4096,
|
1898 |
+
128
|
1899 |
+
],
|
1900 |
+
"dtype": "float16",
|
1901 |
+
"format": "f32-to-bf16",
|
1902 |
+
"nbytes": 1048576,
|
1903 |
+
"byteOffset": 0
|
1904 |
+
},
|
1905 |
+
{
|
1906 |
+
"name": "model.layers.13.mlp.gate_up_proj.q_scale",
|
1907 |
+
"shape": [
|
1908 |
+
22016,
|
1909 |
+
128
|
1910 |
+
],
|
1911 |
+
"dtype": "float16",
|
1912 |
+
"format": "f32-to-bf16",
|
1913 |
+
"nbytes": 5636096,
|
1914 |
+
"byteOffset": 1048576
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"name": "model.layers.13.mlp.down_proj.q_weight",
|
1918 |
+
"shape": [
|
1919 |
+
4096,
|
1920 |
+
1376
|
1921 |
+
],
|
1922 |
+
"dtype": "uint32",
|
1923 |
+
"format": "f32-to-bf16",
|
1924 |
+
"nbytes": 22544384,
|
1925 |
+
"byteOffset": 6684672
|
1926 |
+
},
|
1927 |
+
{
|
1928 |
+
"name": "model.layers.13.mlp.down_proj.q_scale",
|
1929 |
+
"shape": [
|
1930 |
+
4096,
|
1931 |
+
344
|
1932 |
+
],
|
1933 |
+
"dtype": "float16",
|
1934 |
+
"format": "f32-to-bf16",
|
1935 |
+
"nbytes": 2818048,
|
1936 |
+
"byteOffset": 29229056
|
1937 |
+
},
|
1938 |
+
{
|
1939 |
+
"name": "model.layers.13.input_layernorm.weight",
|
1940 |
+
"shape": [
|
1941 |
+
4096
|
1942 |
+
],
|
1943 |
+
"dtype": "float16",
|
1944 |
+
"format": "f32-to-bf16",
|
1945 |
+
"nbytes": 8192,
|
1946 |
+
"byteOffset": 32047104
|
1947 |
+
},
|
1948 |
+
{
|
1949 |
+
"name": "model.layers.13.post_attention_layernorm.weight",
|
1950 |
+
"shape": [
|
1951 |
+
4096
|
1952 |
+
],
|
1953 |
+
"dtype": "float16",
|
1954 |
+
"format": "f32-to-bf16",
|
1955 |
+
"nbytes": 8192,
|
1956 |
+
"byteOffset": 32055296
|
1957 |
+
}
|
1958 |
+
],
|
1959 |
+
"md5sum": "7e0e19e0c2512dff82f2a8c442a16ec8"
|
1960 |
+
},
|
1961 |
+
{
|
1962 |
+
"dataPath": "params_shard_51.bin",
|
1963 |
+
"format": "raw-shard",
|
1964 |
+
"nbytes": 45088768,
|
1965 |
+
"records": [
|
1966 |
+
{
|
1967 |
+
"name": "model.layers.14.mlp.gate_up_proj.q_weight",
|
1968 |
+
"shape": [
|
1969 |
+
22016,
|
1970 |
+
512
|
1971 |
+
],
|
1972 |
+
"dtype": "uint32",
|
1973 |
+
"format": "f32-to-bf16",
|
1974 |
+
"nbytes": 45088768,
|
1975 |
+
"byteOffset": 0
|
1976 |
+
}
|
1977 |
+
],
|
1978 |
+
"md5sum": "1a5169035ab977aaf0e031f1e80dbf06"
|
1979 |
+
},
|
1980 |
+
{
|
1981 |
+
"dataPath": "params_shard_52.bin",
|
1982 |
+
"format": "raw-shard",
|
1983 |
+
"nbytes": 22544384,
|
1984 |
+
"records": [
|
1985 |
+
{
|
1986 |
+
"name": "model.layers.14.mlp.down_proj.q_weight",
|
1987 |
+
"shape": [
|
1988 |
+
4096,
|
1989 |
+
1376
|
1990 |
+
],
|
1991 |
+
"dtype": "uint32",
|
1992 |
+
"format": "f32-to-bf16",
|
1993 |
+
"nbytes": 22544384,
|
1994 |
+
"byteOffset": 0
|
1995 |
+
}
|
1996 |
+
],
|
1997 |
+
"md5sum": "a523a61ca293169dad379555f513f77b"
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"dataPath": "params_shard_53.bin",
|
2001 |
+
"format": "raw-shard",
|
2002 |
+
"nbytes": 25165824,
|
2003 |
+
"records": [
|
2004 |
+
{
|
2005 |
+
"name": "model.layers.15.self_attn.qkv_proj.q_weight",
|
2006 |
+
"shape": [
|
2007 |
+
12288,
|
2008 |
+
512
|
2009 |
+
],
|
2010 |
+
"dtype": "uint32",
|
2011 |
+
"format": "f32-to-bf16",
|
2012 |
+
"nbytes": 25165824,
|
2013 |
+
"byteOffset": 0
|
2014 |
+
}
|
2015 |
+
],
|
2016 |
+
"md5sum": "7f47a4a5ba0c5b1ffc169d4a0210d6e6"
|
2017 |
+
},
|
2018 |
+
{
|
2019 |
+
"dataPath": "params_shard_54.bin",
|
2020 |
+
"format": "raw-shard",
|
2021 |
+
"nbytes": 32587776,
|
2022 |
+
"records": [
|
2023 |
+
{
|
2024 |
+
"name": "model.layers.14.self_attn.qkv_proj.q_scale",
|
2025 |
+
"shape": [
|
2026 |
+
12288,
|
2027 |
+
128
|
2028 |
+
],
|
2029 |
+
"dtype": "float16",
|
2030 |
+
"format": "f32-to-bf16",
|
2031 |
+
"nbytes": 3145728,
|
2032 |
+
"byteOffset": 0
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"name": "model.layers.14.self_attn.o_proj.q_weight",
|
2036 |
+
"shape": [
|
2037 |
+
4096,
|
2038 |
+
512
|
2039 |
+
],
|
2040 |
+
"dtype": "uint32",
|
2041 |
+
"format": "f32-to-bf16",
|
2042 |
+
"nbytes": 8388608,
|
2043 |
+
"byteOffset": 3145728
|
2044 |
+
},
|
2045 |
+
{
|
2046 |
+
"name": "model.layers.14.self_attn.o_proj.q_scale",
|
2047 |
+
"shape": [
|
2048 |
+
4096,
|
2049 |
+
128
|
2050 |
+
],
|
2051 |
+
"dtype": "float16",
|
2052 |
+
"format": "f32-to-bf16",
|
2053 |
+
"nbytes": 1048576,
|
2054 |
+
"byteOffset": 11534336
|
2055 |
+
},
|
2056 |
+
{
|
2057 |
+
"name": "model.layers.14.mlp.gate_up_proj.q_scale",
|
2058 |
+
"shape": [
|
2059 |
+
22016,
|
2060 |
+
128
|
2061 |
+
],
|
2062 |
+
"dtype": "float16",
|
2063 |
+
"format": "f32-to-bf16",
|
2064 |
+
"nbytes": 5636096,
|
2065 |
+
"byteOffset": 12582912
|
2066 |
+
},
|
2067 |
+
{
|
2068 |
+
"name": "model.layers.14.mlp.down_proj.q_scale",
|
2069 |
+
"shape": [
|
2070 |
+
4096,
|
2071 |
+
344
|
2072 |
+
],
|
2073 |
+
"dtype": "float16",
|
2074 |
+
"format": "f32-to-bf16",
|
2075 |
+
"nbytes": 2818048,
|
2076 |
+
"byteOffset": 18219008
|
2077 |
+
},
|
2078 |
+
{
|
2079 |
+
"name": "model.layers.14.input_layernorm.weight",
|
2080 |
+
"shape": [
|
2081 |
+
4096
|
2082 |
+
],
|
2083 |
+
"dtype": "float16",
|
2084 |
+
"format": "f32-to-bf16",
|
2085 |
+
"nbytes": 8192,
|
2086 |
+
"byteOffset": 21037056
|
2087 |
+
},
|
2088 |
+
{
|
2089 |
+
"name": "model.layers.14.post_attention_layernorm.weight",
|
2090 |
+
"shape": [
|
2091 |
+
4096
|
2092 |
+
],
|
2093 |
+
"dtype": "float16",
|
2094 |
+
"format": "f32-to-bf16",
|
2095 |
+
"nbytes": 8192,
|
2096 |
+
"byteOffset": 21045248
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"name": "model.layers.15.self_attn.qkv_proj.q_scale",
|
2100 |
+
"shape": [
|
2101 |
+
12288,
|
2102 |
+
128
|
2103 |
+
],
|
2104 |
+
"dtype": "float16",
|
2105 |
+
"format": "f32-to-bf16",
|
2106 |
+
"nbytes": 3145728,
|
2107 |
+
"byteOffset": 21053440
|
2108 |
+
},
|
2109 |
+
{
|
2110 |
+
"name": "model.layers.15.self_attn.o_proj.q_weight",
|
2111 |
+
"shape": [
|
2112 |
+
4096,
|
2113 |
+
512
|
2114 |
+
],
|
2115 |
+
"dtype": "uint32",
|
2116 |
+
"format": "f32-to-bf16",
|
2117 |
+
"nbytes": 8388608,
|
2118 |
+
"byteOffset": 24199168
|
2119 |
+
}
|
2120 |
+
],
|
2121 |
+
"md5sum": "e5046b82ae71b7ee0b3c13cd8ad87d59"
|
2122 |
+
},
|
2123 |
+
{
|
2124 |
+
"dataPath": "params_shard_55.bin",
|
2125 |
+
"format": "raw-shard",
|
2126 |
+
"nbytes": 45088768,
|
2127 |
+
"records": [
|
2128 |
+
{
|
2129 |
+
"name": "model.layers.15.mlp.gate_up_proj.q_weight",
|
2130 |
+
"shape": [
|
2131 |
+
22016,
|
2132 |
+
512
|
2133 |
+
],
|
2134 |
+
"dtype": "uint32",
|
2135 |
+
"format": "f32-to-bf16",
|
2136 |
+
"nbytes": 45088768,
|
2137 |
+
"byteOffset": 0
|
2138 |
+
}
|
2139 |
+
],
|
2140 |
+
"md5sum": "858f179d9f31284e20a148ea0cd8c754"
|
2141 |
+
},
|
2142 |
+
{
|
2143 |
+
"dataPath": "params_shard_56.bin",
|
2144 |
+
"format": "raw-shard",
|
2145 |
+
"nbytes": 25165824,
|
2146 |
+
"records": [
|
2147 |
+
{
|
2148 |
+
"name": "model.layers.16.self_attn.qkv_proj.q_weight",
|
2149 |
+
"shape": [
|
2150 |
+
12288,
|
2151 |
+
512
|
2152 |
+
],
|
2153 |
+
"dtype": "uint32",
|
2154 |
+
"format": "f32-to-bf16",
|
2155 |
+
"nbytes": 25165824,
|
2156 |
+
"byteOffset": 0
|
2157 |
+
}
|
2158 |
+
],
|
2159 |
+
"md5sum": "ed50be593ce20d306d147e344640a81c"
|
2160 |
+
},
|
2161 |
+
{
|
2162 |
+
"dataPath": "params_shard_57.bin",
|
2163 |
+
"format": "raw-shard",
|
2164 |
+
"nbytes": 32063488,
|
2165 |
+
"records": [
|
2166 |
+
{
|
2167 |
+
"name": "model.layers.15.self_attn.o_proj.q_scale",
|
2168 |
+
"shape": [
|
2169 |
+
4096,
|
2170 |
+
128
|
2171 |
+
],
|
2172 |
+
"dtype": "float16",
|
2173 |
+
"format": "f32-to-bf16",
|
2174 |
+
"nbytes": 1048576,
|
2175 |
+
"byteOffset": 0
|
2176 |
+
},
|
2177 |
+
{
|
2178 |
+
"name": "model.layers.15.mlp.gate_up_proj.q_scale",
|
2179 |
+
"shape": [
|
2180 |
+
22016,
|
2181 |
+
128
|
2182 |
+
],
|
2183 |
+
"dtype": "float16",
|
2184 |
+
"format": "f32-to-bf16",
|
2185 |
+
"nbytes": 5636096,
|
2186 |
+
"byteOffset": 1048576
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"name": "model.layers.15.mlp.down_proj.q_weight",
|
2190 |
+
"shape": [
|
2191 |
+
4096,
|
2192 |
+
1376
|
2193 |
+
],
|
2194 |
+
"dtype": "uint32",
|
2195 |
+
"format": "f32-to-bf16",
|
2196 |
+
"nbytes": 22544384,
|
2197 |
+
"byteOffset": 6684672
|
2198 |
+
},
|
2199 |
+
{
|
2200 |
+
"name": "model.layers.15.mlp.down_proj.q_scale",
|
2201 |
+
"shape": [
|
2202 |
+
4096,
|
2203 |
+
344
|
2204 |
+
],
|
2205 |
+
"dtype": "float16",
|
2206 |
+
"format": "f32-to-bf16",
|
2207 |
+
"nbytes": 2818048,
|
2208 |
+
"byteOffset": 29229056
|
2209 |
+
},
|
2210 |
+
{
|
2211 |
+
"name": "model.layers.15.input_layernorm.weight",
|
2212 |
+
"shape": [
|
2213 |
+
4096
|
2214 |
+
],
|
2215 |
+
"dtype": "float16",
|
2216 |
+
"format": "f32-to-bf16",
|
2217 |
+
"nbytes": 8192,
|
2218 |
+
"byteOffset": 32047104
|
2219 |
+
},
|
2220 |
+
{
|
2221 |
+
"name": "model.layers.15.post_attention_layernorm.weight",
|
2222 |
+
"shape": [
|
2223 |
+
4096
|
2224 |
+
],
|
2225 |
+
"dtype": "float16",
|
2226 |
+
"format": "f32-to-bf16",
|
2227 |
+
"nbytes": 8192,
|
2228 |
+
"byteOffset": 32055296
|
2229 |
+
}
|
2230 |
+
],
|
2231 |
+
"md5sum": "12e022e015e09ec4c17cb20bd7c68b16"
|
2232 |
+
},
|
2233 |
+
{
|
2234 |
+
"dataPath": "params_shard_58.bin",
|
2235 |
+
"format": "raw-shard",
|
2236 |
+
"nbytes": 45088768,
|
2237 |
+
"records": [
|
2238 |
+
{
|
2239 |
+
"name": "model.layers.16.mlp.gate_up_proj.q_weight",
|
2240 |
+
"shape": [
|
2241 |
+
22016,
|
2242 |
+
512
|
2243 |
+
],
|
2244 |
+
"dtype": "uint32",
|
2245 |
+
"format": "f32-to-bf16",
|
2246 |
+
"nbytes": 45088768,
|
2247 |
+
"byteOffset": 0
|
2248 |
+
}
|
2249 |
+
],
|
2250 |
+
"md5sum": "de337359b182b6121fdc9801a084428e"
|
2251 |
+
},
|
2252 |
+
{
|
2253 |
+
"dataPath": "params_shard_59.bin",
|
2254 |
+
"format": "raw-shard",
|
2255 |
+
"nbytes": 22544384,
|
2256 |
+
"records": [
|
2257 |
+
{
|
2258 |
+
"name": "model.layers.16.mlp.down_proj.q_weight",
|
2259 |
+
"shape": [
|
2260 |
+
4096,
|
2261 |
+
1376
|
2262 |
+
],
|
2263 |
+
"dtype": "uint32",
|
2264 |
+
"format": "f32-to-bf16",
|
2265 |
+
"nbytes": 22544384,
|
2266 |
+
"byteOffset": 0
|
2267 |
+
}
|
2268 |
+
],
|
2269 |
+
"md5sum": "a6d679ed3c77a771938b7162f0bb5dd5"
|
2270 |
+
},
|
2271 |
+
{
|
2272 |
+
"dataPath": "params_shard_60.bin",
|
2273 |
+
"format": "raw-shard",
|
2274 |
+
"nbytes": 65536000,
|
2275 |
+
"records": [
|
2276 |
+
{
|
2277 |
+
"name": "lm_head.q_weight",
|
2278 |
+
"shape": [
|
2279 |
+
32000,
|
2280 |
+
512
|
2281 |
+
],
|
2282 |
+
"dtype": "uint32",
|
2283 |
+
"format": "f32-to-bf16",
|
2284 |
+
"nbytes": 65536000,
|
2285 |
+
"byteOffset": 0
|
2286 |
+
}
|
2287 |
+
],
|
2288 |
+
"md5sum": "024e3785e7a04546cbb7709a95ac89b1"
|
2289 |
+
},
|
2290 |
+
{
|
2291 |
+
"dataPath": "params_shard_61.bin",
|
2292 |
+
"format": "raw-shard",
|
2293 |
+
"nbytes": 29253632,
|
2294 |
+
"records": [
|
2295 |
+
{
|
2296 |
+
"name": "model.layers.16.self_attn.qkv_proj.q_scale",
|
2297 |
+
"shape": [
|
2298 |
+
12288,
|
2299 |
+
128
|
2300 |
+
],
|
2301 |
+
"dtype": "float16",
|
2302 |
+
"format": "f32-to-bf16",
|
2303 |
+
"nbytes": 3145728,
|
2304 |
+
"byteOffset": 0
|
2305 |
+
},
|
2306 |
+
{
|
2307 |
+
"name": "model.layers.16.self_attn.o_proj.q_weight",
|
2308 |
+
"shape": [
|
2309 |
+
4096,
|
2310 |
+
512
|
2311 |
+
],
|
2312 |
+
"dtype": "uint32",
|
2313 |
+
"format": "f32-to-bf16",
|
2314 |
+
"nbytes": 8388608,
|
2315 |
+
"byteOffset": 3145728
|
2316 |
+
},
|
2317 |
+
{
|
2318 |
+
"name": "model.layers.16.self_attn.o_proj.q_scale",
|
2319 |
+
"shape": [
|
2320 |
+
4096,
|
2321 |
+
128
|
2322 |
+
],
|
2323 |
+
"dtype": "float16",
|
2324 |
+
"format": "f32-to-bf16",
|
2325 |
+
"nbytes": 1048576,
|
2326 |
+
"byteOffset": 11534336
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"name": "model.layers.16.mlp.gate_up_proj.q_scale",
|
2330 |
+
"shape": [
|
2331 |
+
22016,
|
2332 |
+
128
|
2333 |
+
],
|
2334 |
+
"dtype": "float16",
|
2335 |
+
"format": "f32-to-bf16",
|
2336 |
+
"nbytes": 5636096,
|
2337 |
+
"byteOffset": 12582912
|
2338 |
+
},
|
2339 |
+
{
|
2340 |
+
"name": "model.layers.16.mlp.down_proj.q_scale",
|
2341 |
+
"shape": [
|
2342 |
+
4096,
|
2343 |
+
344
|
2344 |
+
],
|
2345 |
+
"dtype": "float16",
|
2346 |
+
"format": "f32-to-bf16",
|
2347 |
+
"nbytes": 2818048,
|
2348 |
+
"byteOffset": 18219008
|
2349 |
+
},
|
2350 |
+
{
|
2351 |
+
"name": "model.layers.16.input_layernorm.weight",
|
2352 |
+
"shape": [
|
2353 |
+
4096
|
2354 |
+
],
|
2355 |
+
"dtype": "float16",
|
2356 |
+
"format": "f32-to-bf16",
|
2357 |
+
"nbytes": 8192,
|
2358 |
+
"byteOffset": 21037056
|
2359 |
+
},
|
2360 |
+
{
|
2361 |
+
"name": "model.layers.16.post_attention_layernorm.weight",
|
2362 |
+
"shape": [
|
2363 |
+
4096
|
2364 |
+
],
|
2365 |
+
"dtype": "float16",
|
2366 |
+
"format": "f32-to-bf16",
|
2367 |
+
"nbytes": 8192,
|
2368 |
+
"byteOffset": 21045248
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"name": "model.norm.weight",
|
2372 |
+
"shape": [
|
2373 |
+
4096
|
2374 |
+
],
|
2375 |
+
"dtype": "float16",
|
2376 |
+
"format": "f32-to-bf16",
|
2377 |
+
"nbytes": 8192,
|
2378 |
+
"byteOffset": 21053440
|
2379 |
+
},
|
2380 |
+
{
|
2381 |
+
"name": "lm_head.q_scale",
|
2382 |
+
"shape": [
|
2383 |
+
32000,
|
2384 |
+
128
|
2385 |
+
],
|
2386 |
+
"dtype": "float16",
|
2387 |
+
"format": "f32-to-bf16",
|
2388 |
+
"nbytes": 8192000,
|
2389 |
+
"byteOffset": 21061632
|
2390 |
+
}
|
2391 |
+
],
|
2392 |
+
"md5sum": "e81c1bc7f90ff777240bee6be0140d49"
|
2393 |
+
}
|
2394 |
+
]
|
2395 |
+
}
|
params_shard_0.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:48c349a2f31f2994dd70b2150b743bccdfeb916a0e85a58bb4028cda8eca343a
|
3 |
+
size 65536000
|
params_shard_1.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:232c7f283dd7addec9c68db29fcbdef5e2395a72848ca9d36cf7fa3e91571bf6
|
3 |
+
size 33357824
|
params_shard_10.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c3fba9ed6fc447c115212a3c65d8a4e5d923028d9a00c0d58d0b984b86fb1d78
|
3 |
+
size 22544384
|
params_shard_11.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:611e189fe3725b237e2f51e91271c052d4adad12b81ed333197d0ad425161c1b
|
3 |
+
size 25165824
|
params_shard_12.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:05150375ea424ff366bf21d1f9c19cccce1dc0a712732852d5f39b4d23720d61
|
3 |
+
size 32587776
|
params_shard_13.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7e82f022911caf0a166b471f0a189635d9b9d423ebca73e41c17dd4a2675d078
|
3 |
+
size 45088768
|
params_shard_14.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:14d4ef09ba61b07db001f78339d264ef9cf3ccd17715ae4ecd4c9eac89e6fe18
|
3 |
+
size 25165824
|
params_shard_15.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7355c0a99143a510813497404b1077f643427beafb5873952dd7a3edf122ccea
|
3 |
+
size 32063488
|
params_shard_16.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3110dc095a91d9c5d7732cc8318607cb6d2d0c2048fe5f819cc446de0423a829
|
3 |
+
size 45088768
|
params_shard_17.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:66a8ccec02d8102c7b36a92bbeb1249abab8b832fd868f1a9de21261147ab524
|
3 |
+
size 22544384
|
params_shard_18.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f02bd87afce6a52c6607f1598d3dc8708d3319f952f5e2eb59287dd0645125f3
|
3 |
+
size 25165824
|
params_shard_19.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6156d8c154a59d6513f68c35618b1e97991e67d4e4ef09871affb9e907598d05
|
3 |
+
size 32587776
|
params_shard_2.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:84481b731b6fc8d0311e6ed02226c5d6d06041cac4cb6031f893966336e35fef
|
3 |
+
size 45088768
|
params_shard_20.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:892399a4da47733cfa97bfef0288ea6ee6215116040150cb44897c06d4abcd96
|
3 |
+
size 45088768
|
params_shard_21.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7b9c00929c876f71e199bf57bbe6f344d2d1fe4b50ae8f33360e6275203fcc6d
|
3 |
+
size 25165824
|
params_shard_22.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ec7c0f814b12dddb8aa4d63efbbf46630712ec776154da1f7911af56620f5a22
|
3 |
+
size 32063488
|
params_shard_23.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3d4ba67ddc5589ae47faf134e95c672dfeb93a0280c651c69fd392cb38d0ccc2
|
3 |
+
size 45088768
|
params_shard_24.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:69ea6efcb2bfcc33109e5fe15239b627d2ffa070b48cab5b3ad3fd81e1ea4093
|
3 |
+
size 22544384
|
params_shard_25.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:835f3fca438745efffdc44201833352fb64b77ec6852b9424a74c72be56f7d12
|
3 |
+
size 25165824
|
params_shard_26.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a67e986f7c8c429bc1fee8b4e47faf846dfd67756f33bd456c617ad37e9a211c
|
3 |
+
size 32587776
|
params_shard_27.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:08ab699c8abc1ca1f7e6edb6b5ed1b5628eab00d55112ff29d7ce238303f38d3
|
3 |
+
size 45088768
|
params_shard_28.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:08e7389377095a4306492451ca98b1708916e2c5750c9fa7901e651d1affe842
|
3 |
+
size 25165824
|
params_shard_29.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:807f81212ac7677891b9b4b1e5ee4cf3b95badc931299073c59772c20bebfc01
|
3 |
+
size 32063488
|
params_shard_3.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:84bf0af25cb493e072a627b2fef3bcd45b38f5a1fec1a4aedb81d4bf01bc5558
|
3 |
+
size 22544384
|
params_shard_30.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c1156799677b09b1d5a702b9e8fc9da0eaa6d111f870c46560f24ed420968948
|
3 |
+
size 45088768
|
params_shard_31.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:663052f5c09fa024aca8b319d0c90b3bcc94b444f3684e41c76751f24a3c69b7
|
3 |
+
size 22544384
|
params_shard_32.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5dbbff4dfa7f706ff1f9ea0c2b80803849b7acd6efb6a35551973a014aa50db4
|
3 |
+
size 25165824
|
params_shard_33.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0faf2b7a11ffa0c567d49aa41818c020c169b2c975fd25641814e0becddc9591
|
3 |
+
size 32587776
|
params_shard_34.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:49694160abcb370686f63fc25f0035e69c502eaaeb62a32706b6dd1013526b34
|
3 |
+
size 45088768
|
params_shard_35.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c87f1f38bfdc6f10cca5f94bd5123431c9f99b8a5b21e9c170a4306e79a7fd20
|
3 |
+
size 25165824
|
params_shard_36.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c52a7340fb688f920537938b02f343446307fef1635eba9f6463850351432c23
|
3 |
+
size 32063488
|
params_shard_37.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7a9da35485a0e730b694cd20709cc308d0b18e9b4bfc5d94fc37854f5c9abc35
|
3 |
+
size 45088768
|
params_shard_38.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a186f1cad6448451a835131292de64e0bb7dd4b16096289e2bc2799877c6884e
|
3 |
+
size 22544384
|
params_shard_39.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e726483f7a72b379a8b49c395a8ffadf257eabfcbe5d3846c27a20e02d5f42bc
|
3 |
+
size 25165824
|
params_shard_4.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:42d68728888eaaef9f650ea8ef0a07d389e0e6b8e2a46edbd18e2424db4d4cf8
|
3 |
+
size 25165824
|
params_shard_40.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c0bbcfba211e6ac42cb4b6bed068fad3a463574c47bc12c483f98dc0b79577cf
|
3 |
+
size 32587776
|
params_shard_41.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:51297bfd00a5dfcc7cf4b027df2a1f743cad8defa87ed86682b8a9bbc0c8fc4c
|
3 |
+
size 45088768
|
params_shard_42.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:323592fd493c05fc1ed6f3c66bf64cde021e896dc35ec6918e573d4115fbfad8
|
3 |
+
size 25165824
|
params_shard_43.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1cfb8866339247b8ad5a227eb98892a7ab42a3b79048a7615ff86cec90c4b5a6
|
3 |
+
size 32063488
|
params_shard_44.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cdbe9ca2a7a058ae9f6c75d0dba7fbbf2d47567f6839c5e8048a9b17c4eade2b
|
3 |
+
size 45088768
|
params_shard_45.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:16acebe162cce992660e0f3ebbb23633ee9d8a8c9bb7345f4a370ca0102f085f
|
3 |
+
size 22544384
|
params_shard_46.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0d0492626b7a04b7f5f70df9ec02726e6fbbde15b2121ab70be50a08f3dcd9ef
|
3 |
+
size 25165824
|
params_shard_47.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:65f02a016ae867514a854053ee2a9f8433d83df4d6ccf25569decb4de23fadff
|
3 |
+
size 32587776
|
params_shard_48.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5e2cf37f5c3d04df4767844e43555b91d005a267c37fa7ff12120ccce8d6d30
|
3 |
+
size 45088768
|
params_shard_49.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:192b1156be98311732c5cb4ae58aa87784c129591c0eedef7d92c08d2d352bc3
|
3 |
+
size 25165824
|
params_shard_5.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6eeeda57f2ff7cad4b1f033e2ebc2324acdc60e7d019bb5f1cff77eb07a98766
|
3 |
+
size 32587776
|
params_shard_50.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:73b8da4e82e7d005421b7f7e7cd6e5712b1ef72c7306bfbea44b2a881c96e57e
|
3 |
+
size 32063488
|
params_shard_51.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:269d3dfe9f2a6b7b75f276dd1a822e17523a8e4d14dc4234293e0c85e5bad1e6
|
3 |
+
size 45088768
|