kevinoli commited on
Commit
8a1cfc1
1 Parent(s): 1ed8789

Training in progress, step 4000, checkpoint

Browse files
checkpoint-4000/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7782509cca1d07a21b8c556e21997e4446816b0885fe09b7b4501df0f55c8406
3
  size 1711848436
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6469994a0d1fc083714cea684b90c729f20440980c9e3c369536781409ba2cc
3
  size 1711848436
checkpoint-4000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:662b1bb95d062a22928ca4dab2237e2b124d95d4a116054642669f0efe6430bc
3
  size 3424043887
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b05b37fb9f4d8fefe347d971c817a797b4260825a4e7d04837d0973be711f9d0
3
  size 3424043887
checkpoint-4000/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1454cbc8ef4e92d532f823e5816216a6cd4ee18fc15f9b51dc40aa621c880224
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44c9fa5b803fb1f28770699a359bea30c519543808cc61771c36f033b35e9034
3
  size 623
checkpoint-4000/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "best_metric": 1.0954023599624634,
3
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-4000",
4
- "epoch": 0.7367839381101492,
5
  "eval_steps": 500,
6
  "global_step": 4000,
7
  "is_hyper_param_search": false,
@@ -9,128 +9,128 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.09209799226376865,
13
- "grad_norm": 0.0014697719598188996,
14
- "learning_rate": 4.884877509670289e-08,
15
- "loss": 0.4588,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 0.09209799226376865,
20
- "eval_loss": 1.4194883108139038,
21
- "eval_runtime": 78.7703,
22
- "eval_samples_per_second": 15.31,
23
- "eval_steps_per_second": 1.917,
24
  "step": 500
25
  },
26
  {
27
- "epoch": 0.1841959845275373,
28
- "grad_norm": 301.76043701171875,
29
- "learning_rate": 4.769755019340578e-08,
30
- "loss": 0.4255,
31
  "step": 1000
32
  },
33
  {
34
- "epoch": 0.1841959845275373,
35
- "eval_loss": 1.3416553735733032,
36
- "eval_runtime": 77.0051,
37
- "eval_samples_per_second": 15.661,
38
- "eval_steps_per_second": 1.961,
39
  "step": 1000
40
  },
41
  {
42
- "epoch": 0.27629397679130596,
43
- "grad_norm": 35.28853225708008,
44
- "learning_rate": 4.654632529010867e-08,
45
- "loss": 0.3724,
46
  "step": 1500
47
  },
48
  {
49
- "epoch": 0.27629397679130596,
50
- "eval_loss": 1.2872973680496216,
51
- "eval_runtime": 77.5439,
52
- "eval_samples_per_second": 15.552,
53
- "eval_steps_per_second": 1.947,
54
  "step": 1500
55
  },
56
  {
57
- "epoch": 0.3683919690550746,
58
- "grad_norm": 338.9349060058594,
59
- "learning_rate": 4.539510038681156e-08,
60
- "loss": 0.3251,
61
  "step": 2000
62
  },
63
  {
64
- "epoch": 0.3683919690550746,
65
- "eval_loss": 1.234910249710083,
66
- "eval_runtime": 78.7715,
67
- "eval_samples_per_second": 15.31,
68
- "eval_steps_per_second": 1.917,
69
  "step": 2000
70
  },
71
  {
72
- "epoch": 0.46048996131884323,
73
- "grad_norm": 325.85101318359375,
74
- "learning_rate": 4.4243875483514457e-08,
75
- "loss": 0.3308,
76
  "step": 2500
77
  },
78
  {
79
- "epoch": 0.46048996131884323,
80
- "eval_loss": 1.1944907903671265,
81
- "eval_runtime": 79.0397,
82
- "eval_samples_per_second": 15.258,
83
- "eval_steps_per_second": 1.91,
84
  "step": 2500
85
  },
86
  {
87
- "epoch": 0.5525879535826119,
88
- "grad_norm": 100.85855102539062,
89
- "learning_rate": 4.309265058021735e-08,
90
- "loss": 0.3017,
91
  "step": 3000
92
  },
93
  {
94
- "epoch": 0.5525879535826119,
95
- "eval_loss": 1.1593009233474731,
96
- "eval_runtime": 77.6354,
97
- "eval_samples_per_second": 15.534,
98
- "eval_steps_per_second": 1.945,
99
  "step": 3000
100
  },
101
  {
102
- "epoch": 0.6446859458463805,
103
- "grad_norm": 8.865598678588867,
104
- "learning_rate": 4.194142567692024e-08,
105
- "loss": 0.2962,
106
  "step": 3500
107
  },
108
  {
109
- "epoch": 0.6446859458463805,
110
- "eval_loss": 1.1259427070617676,
111
- "eval_runtime": 78.2715,
112
- "eval_samples_per_second": 15.408,
113
- "eval_steps_per_second": 1.929,
114
  "step": 3500
115
  },
116
  {
117
- "epoch": 0.7367839381101492,
118
- "grad_norm": 4.393447399139404,
119
- "learning_rate": 4.079020077362314e-08,
120
- "loss": 0.2919,
121
  "step": 4000
122
  },
123
  {
124
- "epoch": 0.7367839381101492,
125
- "eval_loss": 1.0954023599624634,
126
- "eval_runtime": 78.5366,
127
- "eval_samples_per_second": 15.356,
128
- "eval_steps_per_second": 1.923,
129
  "step": 4000
130
  }
131
  ],
132
  "logging_steps": 500,
133
- "max_steps": 21716,
134
  "num_input_tokens_seen": 0,
135
  "num_train_epochs": 4,
136
  "save_steps": 500,
 
1
  {
2
+ "best_metric": 1.091786503791809,
3
  "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-4000",
4
+ "epoch": 0.7369196757553427,
5
  "eval_steps": 500,
6
  "global_step": 4000,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.09211495946941783,
13
+ "grad_norm": 20.28152847290039,
14
+ "learning_rate": 4.8848563006632274e-08,
15
+ "loss": 0.3758,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 0.09211495946941783,
20
+ "eval_loss": 1.418367862701416,
21
+ "eval_runtime": 73.4439,
22
+ "eval_samples_per_second": 16.434,
23
+ "eval_steps_per_second": 2.056,
24
  "step": 500
25
  },
26
  {
27
+ "epoch": 0.18422991893883567,
28
+ "grad_norm": 47.29538345336914,
29
+ "learning_rate": 4.7697126013264556e-08,
30
+ "loss": 0.4104,
31
  "step": 1000
32
  },
33
  {
34
+ "epoch": 0.18422991893883567,
35
+ "eval_loss": 1.3491536378860474,
36
+ "eval_runtime": 74.3404,
37
+ "eval_samples_per_second": 16.236,
38
+ "eval_steps_per_second": 2.031,
39
  "step": 1000
40
  },
41
  {
42
+ "epoch": 0.2763448784082535,
43
+ "grad_norm": 0.5450155138969421,
44
+ "learning_rate": 4.6545689019896826e-08,
45
+ "loss": 0.4325,
46
  "step": 1500
47
  },
48
  {
49
+ "epoch": 0.2763448784082535,
50
+ "eval_loss": 1.2877168655395508,
51
+ "eval_runtime": 75.3303,
52
+ "eval_samples_per_second": 16.023,
53
+ "eval_steps_per_second": 2.005,
54
  "step": 1500
55
  },
56
  {
57
+ "epoch": 0.36845983787767134,
58
+ "grad_norm": 0.11449883133172989,
59
+ "learning_rate": 4.539425202652911e-08,
60
+ "loss": 0.3423,
61
  "step": 2000
62
  },
63
  {
64
+ "epoch": 0.36845983787767134,
65
+ "eval_loss": 1.2384389638900757,
66
+ "eval_runtime": 76.113,
67
+ "eval_samples_per_second": 15.858,
68
+ "eval_steps_per_second": 1.984,
69
  "step": 2000
70
  },
71
  {
72
+ "epoch": 0.46057479734708917,
73
+ "grad_norm": 389.7380065917969,
74
+ "learning_rate": 4.4242815033161385e-08,
75
+ "loss": 0.3645,
76
  "step": 2500
77
  },
78
  {
79
+ "epoch": 0.46057479734708917,
80
+ "eval_loss": 1.1896915435791016,
81
+ "eval_runtime": 76.23,
82
+ "eval_samples_per_second": 15.834,
83
+ "eval_steps_per_second": 1.981,
84
  "step": 2500
85
  },
86
  {
87
+ "epoch": 0.552689756816507,
88
+ "grad_norm": 340.1103820800781,
89
+ "learning_rate": 4.309137803979366e-08,
90
+ "loss": 0.3171,
91
  "step": 3000
92
  },
93
  {
94
+ "epoch": 0.552689756816507,
95
+ "eval_loss": 1.149291753768921,
96
+ "eval_runtime": 76.1662,
97
+ "eval_samples_per_second": 15.847,
98
+ "eval_steps_per_second": 1.983,
99
  "step": 3000
100
  },
101
  {
102
+ "epoch": 0.6448047162859248,
103
+ "grad_norm": 196.7362518310547,
104
+ "learning_rate": 4.193994104642594e-08,
105
+ "loss": 0.275,
106
  "step": 3500
107
  },
108
  {
109
+ "epoch": 0.6448047162859248,
110
+ "eval_loss": 1.1162768602371216,
111
+ "eval_runtime": 74.6569,
112
+ "eval_samples_per_second": 16.167,
113
+ "eval_steps_per_second": 2.023,
114
  "step": 3500
115
  },
116
  {
117
+ "epoch": 0.7369196757553427,
118
+ "grad_norm": 16.604211807250977,
119
+ "learning_rate": 4.078850405305821e-08,
120
+ "loss": 0.2907,
121
  "step": 4000
122
  },
123
  {
124
+ "epoch": 0.7369196757553427,
125
+ "eval_loss": 1.091786503791809,
126
+ "eval_runtime": 76.3333,
127
+ "eval_samples_per_second": 15.812,
128
+ "eval_steps_per_second": 1.978,
129
  "step": 4000
130
  }
131
  ],
132
  "logging_steps": 500,
133
+ "max_steps": 21712,
134
  "num_input_tokens_seen": 0,
135
  "num_train_epochs": 4,
136
  "save_steps": 500,
checkpoint-4000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:417b0bacc3a8958dfa75621764230fb890db48c09ca241dff24e5264d406285b
3
  size 4847
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:595bb6f39a89ec7030e395471c015665757cb5cacaafa6bba1613fd32b629686
3
  size 4847