yongzx commited on
Commit
6f07692
1 Parent(s): 2918ce0

Push to HF hub

Browse files
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
.gitignore ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ word_embeddings_layernorm.pt
2
+ checkpoint-20000/
3
+ checkpoint-25000/
4
+ checkpoint-15000/
5
+ eval_results.json
6
+ checkpoint-10000/
7
+ all_results.json
8
+ word_embeddings.pt
9
+ checkpoint-5000/
10
+ train_results.json
config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bigscience/bloom-1b7",
3
+ "adapters": {
4
+ "adapters": {},
5
+ "config_map": {},
6
+ "fusion_config_map": {},
7
+ "fusions": {}
8
+ },
9
+ "apply_residual_connection_post_layernorm": false,
10
+ "architectures": [
11
+ "BloomForCausalLM"
12
+ ],
13
+ "attention_dropout": 0.0,
14
+ "attention_softmax_in_fp32": true,
15
+ "bias_dropout_fusion": true,
16
+ "bos_token_id": 1,
17
+ "eos_token_id": 2,
18
+ "hidden_dropout": 0.0,
19
+ "hidden_size": 2048,
20
+ "initializer_range": 0.02,
21
+ "layer_norm_epsilon": 1e-05,
22
+ "masked_softmax_fusion": true,
23
+ "model_type": "bloom",
24
+ "n_head": 16,
25
+ "n_inner": null,
26
+ "n_layer": 24,
27
+ "offset_alibi": 100,
28
+ "pad_token_id": 3,
29
+ "pretraining_tp": 2,
30
+ "seq_length": 4096,
31
+ "skip_bias_add": true,
32
+ "skip_bias_add_qkv": false,
33
+ "slow_but_exact": false,
34
+ "torch_dtype": "float16",
35
+ "transformers_version": "4.20.0.dev0",
36
+ "unk_token_id": 0,
37
+ "use_cache": true,
38
+ "vocab_size": 250880
39
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86f5a3aca6924445b27152db8248e9d4d4410f7921d69527546ea7d1bc24eff3
3
+ size 3444858091
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f6efc66e73f1fd69da4f436e48befb519fdff3fe18910850c1d41bd862293a5
3
+ size 14500443
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "<pad>", "add_prefix_space": false, "name_or_path": "bigscience/bloom-1b7", "special_tokens_map_file": null, "padding_side": "left", "tokenizer_class": "BloomTokenizer"}
trainer_state.json ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.7216796875,
3
+ "best_model_checkpoint": "/home/zhengxinyong/outputs/bloom-1b7_ru_continual-pretrain_100000samples_-1vocab_original/checkpoint-25000",
4
+ "epoch": 1.4637002341920375,
5
+ "global_step": 25000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.15,
12
+ "learning_rate": 0.0001,
13
+ "loss": 2.2537,
14
+ "step": 2500
15
+ },
16
+ {
17
+ "epoch": 0.29,
18
+ "learning_rate": 0.0001,
19
+ "loss": 1.9859,
20
+ "step": 5000
21
+ },
22
+ {
23
+ "epoch": 0.29,
24
+ "eval_loss": 1.9208984375,
25
+ "eval_runtime": 7.2654,
26
+ "eval_samples_per_second": 187.601,
27
+ "eval_steps_per_second": 23.536,
28
+ "step": 5000
29
+ },
30
+ {
31
+ "epoch": 0.44,
32
+ "learning_rate": 0.0001,
33
+ "loss": 1.896,
34
+ "step": 7500
35
+ },
36
+ {
37
+ "epoch": 0.59,
38
+ "learning_rate": 0.0001,
39
+ "loss": 1.853,
40
+ "step": 10000
41
+ },
42
+ {
43
+ "epoch": 0.59,
44
+ "eval_loss": 1.82421875,
45
+ "eval_runtime": 7.2655,
46
+ "eval_samples_per_second": 187.599,
47
+ "eval_steps_per_second": 23.536,
48
+ "step": 10000
49
+ },
50
+ {
51
+ "epoch": 0.73,
52
+ "learning_rate": 0.0001,
53
+ "loss": 1.8152,
54
+ "step": 12500
55
+ },
56
+ {
57
+ "epoch": 0.88,
58
+ "learning_rate": 0.0001,
59
+ "loss": 1.782,
60
+ "step": 15000
61
+ },
62
+ {
63
+ "epoch": 0.88,
64
+ "eval_loss": 1.7744140625,
65
+ "eval_runtime": 7.2563,
66
+ "eval_samples_per_second": 187.838,
67
+ "eval_steps_per_second": 23.566,
68
+ "step": 15000
69
+ },
70
+ {
71
+ "epoch": 1.02,
72
+ "learning_rate": 0.0001,
73
+ "loss": 1.7434,
74
+ "step": 17500
75
+ },
76
+ {
77
+ "epoch": 1.17,
78
+ "learning_rate": 0.0001,
79
+ "loss": 1.6356,
80
+ "step": 20000
81
+ },
82
+ {
83
+ "epoch": 1.17,
84
+ "eval_loss": 1.744140625,
85
+ "eval_runtime": 7.2585,
86
+ "eval_samples_per_second": 187.781,
87
+ "eval_steps_per_second": 23.559,
88
+ "step": 20000
89
+ },
90
+ {
91
+ "epoch": 1.32,
92
+ "learning_rate": 0.0001,
93
+ "loss": 1.6475,
94
+ "step": 22500
95
+ },
96
+ {
97
+ "epoch": 1.46,
98
+ "learning_rate": 0.0001,
99
+ "loss": 1.6398,
100
+ "step": 25000
101
+ },
102
+ {
103
+ "epoch": 1.46,
104
+ "eval_loss": 1.7216796875,
105
+ "eval_runtime": 7.2617,
106
+ "eval_samples_per_second": 187.696,
107
+ "eval_steps_per_second": 23.548,
108
+ "step": 25000
109
+ },
110
+ {
111
+ "epoch": 1.46,
112
+ "step": 25000,
113
+ "total_flos": 1.4851359536421274e+18,
114
+ "train_loss": 1.8252051953125,
115
+ "train_runtime": 6127.1413,
116
+ "train_samples_per_second": 32.642,
117
+ "train_steps_per_second": 4.08
118
+ }
119
+ ],
120
+ "max_steps": 25000,
121
+ "num_train_epochs": 2,
122
+ "total_flos": 1.4851359536421274e+18,
123
+ "trial_name": null,
124
+ "trial_params": null
125
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f8512acb80154e87f881585e538ba4b3d802b69b3ac38ea86e2c5e252d7ec32
3
+ size 4399